satellite_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "\n",
    "import fiona\n",
    "import rasterio\n",
    "\n",
    "from shapely import geometry\n",
    "from rasterio.mask import mask\n",
    "\n",
    "from tqdm import tqdm_notebook\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.colors as clr\n",
    "plt.style.use('seaborn-whitegrid')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib notebook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "BANDS = [425, 480, 545, 605, 660, 725, 830, 950]\n",
    "\n",
    "BANDS_L = [425, 480, 545, 605, 660, 725, 830, 950]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DATA_PATH = '../data/'\n",
    "# DATA_PATH = '/mnt/MOSIN_RESULTS/'\n",
    "DATA_PATH = '/Volumes/SPACE/02_PROJECT/02_02_ACTIVE/02_FOREST_HYPE/MOSIN_RESULTS/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "hsi_pp_2 = rasterio.open(DATA_PATH + \"pp_2/pp_2_sat_pan_clipped.tif\")\n",
    "hsi_pp_3 = rasterio.open(DATA_PATH + \"pp_3/pp_3_sat_pan_clipped.tif\")\n",
    "hsi_pp_4 = rasterio.open(DATA_PATH + \"pp_4/pp_4_sat_pan_clipped.tif\")\n",
    "\n",
    "chm_pp_2 = rasterio.open(DATA_PATH + \"pp_2/pp_2_chm.tif\")\n",
    "chm_pp_3 = rasterio.open(DATA_PATH + \"pp_3/pp_3_chm.tif\")\n",
    "chm_pp_4 = rasterio.open(DATA_PATH + \"pp_4/pp_4_chm.tif\")\n",
    "\n",
    "trees_pp_2 = fiona.open(DATA_PATH + \"pp_2/pp_2_trees_h/\", \"r\")\n",
    "trees_pp_3 = fiona.open(DATA_PATH + \"pp_3/pp_3_trees_h/\", \"r\")\n",
    "trees_pp_4 = fiona.open(DATA_PATH + \"pp_4/pp_4_trees_h/\", \"r\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def exclude_bad_pixels(image_instance):\n",
    "    bad_pixels = np.where(image_instance[-1]==0)\n",
    "    mask = (np.repeat(range(32), bad_pixels[0].shape[0]), \n",
    "            np.tile(bad_pixels[0], 32), \n",
    "            np.tile(bad_pixels[1], 32))\n",
    "    image_instance[mask] = np.ma.masked\n",
    "    \n",
    "\n",
    "def get_trees(hsi_pp, chm_pp, trees_pp, hsi_radius=1.0, chm_radius=1.0):\n",
    "    hsi_images, chm_images, labels, ids = [], [], [], []\n",
    "    map_tree_speices = {'е':0, 'б':1, 'п':2, \"El'\":0, \"Bereza\":1, \"Pichta\":2, \"Sosna\":3} \n",
    "    \n",
    "    for tree in trees_pp.values():\n",
    "        tree_specie = tree['properties']['specie']\n",
    "        labels.append(map_tree_speices[tree_specie])\n",
    "        ids.append(tree['properties']['id_tree'])\n",
    "        \n",
    "        geometry_hsi = geometry.shape(tree['geometry']).buffer(hsi_radius)\n",
    "        geometry_chm = geometry.shape(tree['geometry']).buffer(chm_radius)\n",
    "        \n",
    "        hsi_image, _ = mask(hsi_pp, [geometry_hsi], filled=False, crop=True)\n",
    "        chm_image, _ = mask(chm_pp, [geometry_chm], filled=False, crop=True)\n",
    "        \n",
    "        exclude_bad_pixels(hsi_image)\n",
    "        hsi_images.append(hsi_image/65535)\n",
    "        chm_images.append(chm_image)\n",
    "    \n",
    "    return hsi_images, labels, chm_images, ids"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "RADIUS_HSI = 2.0\n",
    "RADIUS_CHM = 2.0\n",
    "\n",
    "tree_images_pp_2, tree_labels_pp_2, tree_chms_pp_2, tree_ids_pp_2 = get_trees(hsi_pp_2, chm_pp_2, trees_pp_2,\n",
    "                                                                              hsi_radius=RADIUS_HSI, \n",
    "                                                                              chm_radius=RADIUS_CHM)\n",
    "tree_images_pp_3, tree_labels_pp_3, tree_chms_pp_3, tree_ids_pp_3 = get_trees(hsi_pp_3, chm_pp_3, trees_pp_3,\n",
    "                                                                              hsi_radius=RADIUS_HSI, \n",
    "                                                                              chm_radius=RADIUS_CHM)\n",
    "tree_images_pp_4, tree_labels_pp_4, tree_chms_pp_4, tree_ids_pp_4 = get_trees(hsi_pp_4, chm_pp_4, trees_pp_4,\n",
    "                                                                              hsi_radius=RADIUS_HSI, \n",
    "                                                                              chm_radius=RADIUS_CHM)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_labels_pp_2 = np.array(tree_labels_pp_2)\n",
    "tree_labels_pp_3 = np.array(tree_labels_pp_3)\n",
    "tree_labels_pp_4 = np.array(tree_labels_pp_4)\n",
    "\n",
    "tree_ids_pp_2 = np.array(tree_ids_pp_2)\n",
    "tree_ids_pp_3 = np.array(tree_ids_pp_3)\n",
    "tree_ids_pp_4 = np.array(tree_ids_pp_4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAoAAAAHgCAYAAAA10dzkAAAAAXNSR0IArs4c6QAAOFhJREFUeAHt3Q+wZfN9APDf2/d29+3zZ/0fXTS6eLyyUmlDiCHETKtpDLtDK2KNmIiQmkq2VZWpIdjQUEmsvxWj/qwi0YkVNEvTTlMZqsSzWdZkEbWIkgr717736lw92/dv/933O/edc36fO7Nz7jnvnu/5fj/f477vPfe+q23g/VtwI0CAAAECBAgQSEZgQjKVKpQAAQIECBAgQKAhYAB0IhAgQIAAAQIEEhMwACbWcOUSIECAAAECBAyAzgECBAgQIECAQGICBsDEGq5cAgQIECBAgIAB0DlAgAABAgQIEEhMwACYWMOVS4AAAQIECBAwADoHCBAgQIAAAQKJCRgAE2u4cgkQIECAAAECBkDnAAECBAgQIEAgMQEDYGINVy4BAgQIECBAwADoHCBAgAABAgQIJCZgAEys4colQIAAAQIECBgAnQMECBAgQIAAgcQEDICJNVy5BAgQIECAAAEDoHOAAAECBAgQIJCYgAEwsYYrlwABAgQIECBgAHQOECBAgAABAgQSEzAAJtZw5RIgQIAAAQIEDIDOAQIECBAgQIBAYgIGwMQarlwCBAgQIECAgAHQOUCAAAECBAgQSEzAAJhYw5VLgAABAgQIEDAAOgcIECBAgAABAokJGAATa7hyCRAgQIAAAQIGQOcAAQIECBAgQCAxAQNgYg1XLgECBAgQIEDAAOgcIECAAAECBAgkJmAATKzhyiVAgAABAgQIGACdAwQIECBAgACBxAQMgIk1XLkECBAgQIAAAQOgc4AAAQIECBAgkJiAATCxhiuXAAECBAgQIGAAdA4QIECAAAECBBITMAAm1nDlEiBAgAABAgQMgM4BAgQIECBAgEBiAgbAxBquXAIECBAgQICAAdA5QIAAAQIECBBITMAAmFjDlUuAAAECBAgQMAA6BwgQIECAAAECiQkYABNruHIJECBAgAABAgZA5wABAgQIECBAIDEBA2BiDVcuAQIECBAgQMAA6BwgQIAAAQIECCQmYABMrOHKJUCAAAECBAgYAJ0DBAgQIECAAIHEBAyAiTVcuQQIECBAgAABA6BzgAABAgQIECCQmIABMLGGK5cAAQIECBAgYAB0DhAgQIAAAQIEEhMwACbWcOUSIECAAAECBAyAzgECBAgQIECAQGICBsDEGq5cAgQIECBAgIAB0DlAgAABAgQIEEhMwACYWMOVS4AAAQIECBAwADoHCBAgQIAAAQKJCXQkVq9yIwmsXbs2vP3222Hy5MlhwgSvIyKxCkOAAIGWCfT394fVq1eHqVOnho4O40DL4EtyIB0vSSOqlkY2/L344otVS1u+BAgQIDBMYPfddw/bb7/9sK1W6y5gAKx7hwuqL7vyl92yJ44pU6ZEOUpfX19YsmRJ6O7uDu3t7VFiphiEY5yuc+QYRyBOlCLOx5UrVzZeyOfP53EyFaUqAgbAqnSqZHnmb/tmw19XV1eU7LInuOyWxTMANk/KsXm7wXtyHKzR/H2OzdsN3rNIx/z5fPDx3K+/gA9v1b/HKiRAgAABAgQIDBEwAA7hsEKAAAECBAgQqL+AAbD+PVYhAQIECBAgQGCIgAFwCIcVAgQIECBAgED9BQyA9e+xCgkQIECAAAECQwQMgEM4rBAgQIAAAQIE6i9gAKx/j1VIgAABAgQIEBgiYAAcwmGFAAECBAgQIFB/AQNg/XusQgIECBAgQIDAEAED4BAOKwQIECBAgACB+gsYAOvfYxUSIECAAAECBIYIGACHcFghQIAAAQIECNRfwABY/x6rkAABAgQIECAwRMAAOITDyu233x6OPPLIMGPGjHD88ceHp59+GgoBAgQIECBQMwEDYM0aOpZyfvCDH4S5c+eGs846K9x7771hn332Caeddlp48803xxLWvgQIECBAgEDJBAyAJWvIeKZz8803hxNOOCHMmjUr7LnnnuHCCy8MnZ2d4bvf/e54puXYBAgQIECAQGSBjsjxhKuowJo1a8KiRYvCF77whXUVTJgwIRxyyCHhySefXLdt+J2+vr6Q/Ytxy+PkyxgxU4yR++XLFA1i1Jz75csYMVOMkfvlyxQNYtSc++XLmDFjxBKjegIGwOr1rJCMf/WrXzUGue23335I/Gx96dKlQ7YNXlmyZMng1Sj3e3t7o8RJPQjHOGcAR45xBOJEcT7GcRQlBAOgs2BMAt3d3aGrq2tMMfKds1e22ZNb9gco7e3t+WbLzRTguJlg63k4x/XAbOZmjpsJtp6HF+G4YsWKUMSL+PWUYHPJBAyAJWvIeKWz7bbbNoau4X/wka3vsMMO600rG9RiD2tFxFxvATX+Acc4zeXIMY5AnCgxz8csllu6Av4IJN3eD6l80qRJYd999w2PPvrouu39/f2N9QMOOGDdNncIECBAgACB6gu4Alj9Hkar4NRTTw3nnntu2G+//cL+++8fbrnllrBy5cowc+bMaMcQiAABAgQIEBh/AQPg+PegNBn84R/+YXjrrbfCt771rfDGG2+Enp6e8Hd/93cbfAu4NMlLhAABAgQIENhkAQPgJlOl8cDPfvazIfvnRoAAAQIECNRXwGcA69tblREgQIAAAQIERhUwAI7KYiMBAgQIECBAoL4CBsD69lZlBAgQIECAAIFRBQyAo7LYSIAAAQIECBCor4ABsL69VRkBAgQIECBAYFQBA+CoLDYSIECAAAECBOorYACsb29VRoAAAQIECBAYVcAAOCqLjQQIECBAgACB+goYAOvbW5URIECAAAECBEYVMACOymIjAQIECBAgQKC+AgbA+vZWZQQIECBAgACBUQUMgKOy2EiAAAECBAgQqK+AAbC+vVUZAQIECBAgQGBUgY5Rt9pIgEDLBHp7e6Mea2BgoBFv0aJFoa2tLUrsl156KUqcIoO8++67UcNPmDAh7LHHHuG73/1u6O/vjxb7X//1X6PFKirQNddcU1RocQkQKImAK4AlaYQ0CBAgQIAAAQKtEjAAtkracQgQIECAAAECJREwAJakEdIgQIAAAQIECLRKwADYKmnHIUCAAAECBAiURMAAWJJGSIMAAQIECBAg0CoBA2CrpB2HAAECBAgQIFASAQNgSRohDQIECBAgQIBAqwQMgK2SdhwCBAgQIECAQEkEDIAlaYQ0CBAgQIAAAQKtEjAAtkracQgQIECAAAECJREwAJakEdIgQIAAAQIECLRKwADYKmnHIUCAAAECBAiURMAAWJJGSIMAAQIECBAg0CoBA2CrpB2HAAECBAgQIFASAQNgSRohDQIECBAgQIBAqwQMgK2SdhwCBAgQIECAQEkEDIAlaUQZ0nj88cfDGWecEQ499NCw9957h4ULF5YhLTkQIECAAAECkQUMgJFBqxxuxYoVjcHvggsuqHIZcidAgAABAgQ2ItCxkZ/7cUIChx9+eMj+uREgQIAAAQL1FjAA1ru/hVfX19cXsn8xbnmcfBkjZhViDAwMRE0zj5cvYwRva2uLEabQGBMmxH1DI685W8aMPXHixEIdYgSP+d9gHitfxsgvxRi5X76MYRAzVox8xGitQNv7vyTi/vZpbf6OVpBA9hnAefPmhaOOOmrUI2RvFy9evHjUn9lIgAABAtUR6OnpCV1dXdVJWKZRBFwBjMKYbpDu7u5oTxzZq9He3t4wY8aM0N7engzqokWLotaavaZbu3Zt6OjoCPlVrLEe4OWXXx5riML3X758edRjZHbTp08PS5cuDTFfJ//4xz+OmmcRwa688spoYVP97zoa4P8FKsIxeyG/ZMmS2KmKVxEBA2BFGlXWNLNBLfawVkTMsvplecUa0obXmMWNFTvmADQ8z1jr/f39sUI14uRv+2a1x4z93nvvRc2ziGCx/5vOckztv+si+hLbsYg+F1W3uPEF4n5oJn5+IhIgQIAAAQIECEQWcAUwMmiVw2Vvof3iF79YV8J//dd/NT7nN3Xq1DBt2rR1290hQIAAAQIEqi1gAKx2/6Jm/8wzz4TZs2evizl37tzG/eOOOy58/etfX7fdHQIECBAgQKDaAgbAavcvavYHHXRQeO6556LGFIwAAQIECBAon4DPAJavJzIiQIAAAQIECBQqYAAslFdwAgQIECBAgED5BAyA5euJjAgQIECAAAEChQoYAAvlFZwAAQIECBAgUD4BA2D5eiIjAgQIECBAgEChAgbAQnkFJ0CAAAECBAiUT8AAWL6eyIgAAQIECBAgUKiAAbBQXsEJECBAgAABAuUTMACWrycyIkCAAAECBAgUKmAALJRXcAIECBAgQIBA+QQMgOXriYwIECBAgAABAoUKGAAL5RWcAAECBAgQIFA+AQNg+XoiIwIECBAgQIBAoQIdhUYXnMA4C3zve98b5ww2fvgttthi4w/ajEe0tbWF7bffPrz22mthYGBgM/Zc/0MnTpy4/h+W5CerVq2KmsmECR+8Ps7i9vf3R4v97LPPRotVVKBp06ZFCz1lypRw1113he7u7rBy5cpocZctWxYtlkAEUhRwBTDFrquZAAECBAgQSFrAAJh0+xVPgAABAgQIpChgAEyx62omQIAAAQIEkhYwACbdfsUTIECAAAECKQoYAFPsupoJECBAgACBpAUMgEm3X/EECBAgQIBAigIGwBS7rmYCBAgQIEAgaQEDYNLtVzwBAgQIECCQooABMMWuq5kAAQIECBBIWsAAmHT7FU+AAAECBAikKGAATLHraiZAgAABAgSSFjAAJt1+xRMgQIAAAQIpChgAU+y6mgkQIECAAIGkBQyASbdf8QQIECBAgECKAgbAFLuuZgIECBAgQCBpAQNg0u1XPAECBAgQIJCiQEeKRat5dIHrr78+/NM//VNYunRp6OzsDAcccECYM2dOmD59+ug72EqAAAECBAhUUsAVwEq2rZikH3vssXDSSSeFu+66K9x8881h7dq14bTTTgsrVqwo5oCiEiBAgAABAuMi4ArguLCX86A33XTTkMS+/vWvh4MPPjgsWrQofPSjHx3yMysECBAgQIBAdQUMgNXtXeGZv/POO41jTJ06db3H6uvrC9m/GLc8Tr6MEbOtrS1GmEJjxM4xj5cvC02+RMEnTIj7hkYeL1/GKnXy5MmxQhUWZ8qUKdFi57HyZazAMZ8nYuVUZJy83nwZ41gxY8XIR4zWCrQNvH9r7SEdrQoC/f394Ytf/GL49a9/HebPnz8i5ext4cWLF4/YbgMBAgQIVEugp6cndHV1VStp2Y5ZwBXAMRPWM8CFF14Ynn/++XDHHXdssMDu7u5oTxzZq9He3t4wY8aM0N7evsHjbuoPv//972/qQ8ftcVtssUXUY2dX/rbbbrvw1ltvhZRe37322mtRHbMrf9kvxuyFTvaCKNZttBdUsWLHirNkyZJYoUJ25e+WW24Jp5xySli5cmW0uDFzjJZUgYGKeH7MXsin5lhgiyoX2gBYuZYVn/BFF10UfvSjH4Xbbrst7Lzzzhs8YDaoxRrW8gPFjFmFAaioHLO4RcXOe1WmZcwhbXBdWdyYsVevXj04fCnvxxzU8gKzmDHjxn7eyfMs+zLm82OqhmXvcavyMwC2SroCx8mGha997Wvhhz/8Ybj11lvDbrvtVoGspUiAAAECBAhsroABcHPFavz47G3fBQsWhGuuuSZkb0u+8cYbjWq32mqrxvcC1rh0pREgQIAAgaQEDIBJtXvDxeafTTr55JOHPHDu3Llh5syZQ7ZZIUCAAAECBKorYACsbu+iZ/7cc89FjykgAQIECBAgUD6BuF+cVb76ZESAAAECBAgQIDBMwAA4DMQqAQIECBAgQKDuAgbAundYfQQIECBAgACBYQIGwGEgVgkQIECAAAECdRcwANa9w+ojQIAAAQIECAwTMAAOA7FKgAABAgQIEKi7gAGw7h1WHwECBAgQIEBgmIABcBiIVQIECBAgQIBA3QUMgHXvsPoIECBAgAABAsMEDIDDQKwSIECAAAECBOouYACse4fVR4AAAQIECBAYJmAAHAZilQABAgQIECBQdwEDYN07rD4CBAgQIECAwDCBjmHrVgnUSuDuu+8ufT3bbrtt1BwnTpwYZs+eHe6///7w3nvvRYmdxSv7bYcddigkxX333Tdq3GnTpkWNV0SwZcuWRQs7ZcqURqytt946ZOdmrNvBBx8cK1RhcR599NHCYgtMYKwCrgCOVdD+BAgQIECAAIGKCRgAK9Yw6RIgQIAAAQIExipgAByroP0JECBAgAABAhUTMABWrGHSJUCAAAECBAiMVcAAOFZB+xMgQIAAAQIEKiZgAKxYw6RLgAABAgQIEBirgAFwrIL2J0CAAAECBAhUTMAAWLGGSZcAAQIECBAgMFYBA+BYBe1PgAABAgQIEKiYgAGwYg2TLgECBAgQIEBgrAIGwLEK2p8AAQIECBAgUDEBA2DFGiZdAgQIECBAgMBYBQyAYxW0PwECBAgQIECgYgIGwIo1TLoECBAgQIAAgbEKGADHKmh/AgQIECBAgEDFBAyAFWuYdAkQIECAAAECYxXoGGsA+9dH4I477gjz588Pr7zySqOovfbaK5x55pnh8MMPr0+RKiFAgAABAgSCAdBJsE5g5513DnPmzAkf+tCHwsDAQPjHf/zHcNZZZ4V77703ZMOgGwECBAgQIFAPAQNgPfoYpYojjzxySJxzzjmncUXwqaeeMgAOkbFCgAABAgSqLWAArHb/Csu+r68vPPjgg2HFihXhgAMOWO9xssdl/2Lc8jj5MkbMSZMmxQhTaIyJEydGjZ/Hy5dRgws2ZoEqnJNTpkwZc515gM7OzsbdfJlvH+sydryx5jPa/jGfy/JY+XK0423utpixNvfYHj/+Am3vv9U3MP5pyKAsAs8991z4kz/5k7B69erQ1dUVrrjiilE/A5gNhosXLy5L2vIgQIAAgSYFenp6Gs/3Te5ut4oKGAAr2rii0l6zZk149dVXwzvvvBMeeuihcPfdd4fbbrst7LnnnkMOmQ+A3d3d0Z44slejvb29YcaMGaG9vX3I8Zpd+dznPtfsri3bb5tttol6rOzK34knnth4+/69996LEjuLV/ZbVa54zps3r+yU4cknn4yWY3al7uqrrw5f+tKXwqpVq6LFjf3fTbTEBgVauHDhoLWx3S3i+TF7Hl+yZEkwAI6tN1Xd21vAVe1cQXlnb09lfwSS3fbbb7/GQPb3f//34aKLLhr1iNmgFmtYyw8QM2Y20Jb9FmtIG15nFreo2MOPZX3TBapwTq5cuXLTC9rER2bDX8y4VXgLOPZzY0Yd8/mxiPw28XTwsBII+B7AEjShzCn09/eHKvzCKrOh3AgQIECAQNkEXAEsW0fGMZ/s836HHXZY+I3f+I2wfPnysGDBgvDYY4+Fm266aRyzcmgCBAgQIEAgtoABMLZoheO9+eab4dxzzw2//OUvw1ZbbRX23nvvxvD38Y9/vMJVSZ0AAQIECBAYLmAAHC6S8Pqll16acPVKJ0CAAAEC6Qj4DGA6vVYpAQIECBAgQKAhYAB0IhAgQIAAAQIEEhMwACbWcOUSIECAAAECBAyAzgECBAgQIECAQGICBsDEGq5cAgQIECBAgIAB0DlAgAABAgQIEEhMwACYWMOVS4AAAQIECBAwADoHCBAgQIAAAQKJCRgAE2u4cgkQIECAAAECBkDnAAECBAgQIEAgMQEDYGINVy4BAgQIECBAwADoHCBAgAABAgQIJCZgAEys4colQIAAAQIECHQgIDAWgaOPPjosXbp0LCHW7TtlypQwf/780NPTE1auXLlu+1juLF++fCy7t2Tfzs7OqMfJHGfPnh0WLFgQzTF2jlEL/r9gZ555ZtSwAwMD4de//nXYeuutQ1tbW7TYM2fOjBarqEA//elPo4Xu6Pjg10y2nDhxYrS4//3f/x0tlkAEUhRwBTDFrquZAAECBAgQSFrAAJh0+xVPgAABAgQIpChgAEyx62omQIAAAQIEkhYwACbdfsUTIECAAAECKQoYAFPsupoJECBAgACBpAUMgEm3X/EECBAgQIBAigIGwBS7rmYCBAgQIEAgaQEDYNLtVzwBAgQIECCQooABMMWuq5kAAQIECBBIWsAAmHT7FU+AAAECBAikKGAATLHraiZAgAABAgSSFjAAJt1+xRMgQIAAAQIpChgAU+y6mgkQIECAAIGkBQyASbdf8QQIECBAgECKAgbAFLuuZgIECBAgQCBpAQNg0u1XPAECBAgQIJCigAEwxa5vQs033HBD2HvvvcMll1yyCY/2EAIECBAgQKBKAgbAKnWrRbk+/fTT4c4772wMgC06pMMQIECAAAECLRQwALYQuwqHWr58efjzP//zcPHFF4epU6dWIWU5EiBAgAABApsp0LGZj/fwmgtcdNFF4fDDDw+HHHJIuPbaazda7eTJk8OUKVM2+rhNeUAeJ19uyj4be8zAwMDGHjLuP88MY95yv3wZI3ZHR/mfKmL3Oo+XL2M4ZjHa2tpihSosTmdnZ7TYeax8GStwFRz7+vpilRvyWPkyRuCYsWLkI0ZrBdref3Ir/2/I1poke7T7778/XHfddeGee+4J2VBy8sknh3322Secf/75I0xWrFgRFi9ePGK7DQQIECBQLYGenp7Q1dVVraRlO2aB8r+sH3OJAmyKwKuvvtr4g4/vfOc7jeFvU/bJHnPeeeeFF198cVMfvsHHZVessuN/7nOfCytXrtzgYzf1h9mgWvZbEVcAb7nllnDKKadEc5w1a1bZGcPnP//5qDlmr43feeedsNVWW0W9avfMM89EzbOIYJdeemm0sNmVv6uuuir82Z/9WVi1alW0uKtXr44Wq6hATz75ZLTQ2dW63t7eMGPGjNDe3h4lbvb8uGTJkiixBKmegAGwej0rJONFixaFN998M8ycOXNd/OwJ5/HHHw+3335744lntCed7Ek41rCWHziLFytmFQbA/v7+vPSoy5iOa9eujZpbEcGKekswixszdhXedIk5qOW9zmLGjBszVp5j7OVoz5ljPUYWM1bcWHHGWpP9x0fAADg+7qU76sc+9rFw3333Dckru7o3ffr0xpUVTxRDaKwQIECAAIFKCxgAK92+eMlvueWWobu7e0jA7DMh22yzzYjtQx5khQABAgQIEKicgK+BqVzLJEyAAAECBAgQGJuAK4Bj86v13rfeemut61McAQIECBBIVcAVwFQ7r24CBAgQIEAgWQEDYLKtVzgBAgQIECCQqoABMNXOq5sAAQIECBBIVsAAmGzrFU6AAAECBAikKmAATLXz6iZAgAABAgSSFTAAJtt6hRMgQIAAAQKpChgAU+28ugkQIECAAIFkBQyAybZe4QQIECBAgECqAgbAVDuvbgIECBAgQCBZAQNgsq1XOAECBAgQIJCqgAEw1c6rmwABAgQIEEhWwACYbOsVToAAAQIECKQqYABMtfPqJkCAAAECBJIV6Ei2coVHEdhpp53CmjVrosTq7OxsxNlll13CqlWrosR87733osQpMkhfX1/U8Llj1ptYjvfdd1/UHIsINn369Khh29vbw4EHHhgWLlwYYvboS1/6UtQ8iwi2aNGiaGE7Oj74NTNz5sywdu3aaHGvvvrqaLEEIpCigCuAKXZdzQQIECBAgEDSAgbApNuveAIECBAgQCBFAQNgil1XMwECBAgQIJC0gAEw6fYrngABAgQIEEhRwACYYtfVTIAAAQIECCQtYABMuv2KJ0CAAAECBFIUMACm2HU1EyBAgAABAkkLGACTbr/iCRAgQIAAgRQFDIApdl3NBAgQIECAQNICBsCk2694AgQIECBAIEUBA2CKXVczAQIECBAgkLSAATDp9iueAAECBAgQSFHAAJhi19VMgAABAgQIJC1gAEy6/YonQIAAAQIEUhQwAKbYdTUTIECAAAECSQsYAJNuv+IJECBAgACBFAU6UixazaMLfPvb3w5XX331kB/+1m/9VnjwwQeHbLNCgAABAgQIVFvAAFjt/kXPfq+99go333zzurjt7e3r7rtDgAABAgQI1EPAAFiPPkarIhv4dtxxx2jxBCJAgAABAgTKJ2AALF9PxjWjl156KRx66KFh8uTJ4Xd+53fCV77ylTBt2rT15pQ9rrOzc70/35wfZLGyW77cnH3X99gqXMHs6+tbX/pNbc/7kS+bCjJsp9g5DgsfZTV2rydM+OAj0vkySpLvB6mCZUdHvF8NeV/yZSzHKVOmxApVWJyYvc5j5csYSceMFSMfMVor0Dbw/q21h3S0sgr8y7/8S1ixYkXIPvf3xhtvhHnz5oXXX3893HfffWHLLbccknb2uMWLFw/ZZoUAAQIEqifQ09MTurq6qpe4jMckEO9l3pjSsHMZBA4//PB1aeyzzz7hwx/+cDjiiCPCAw88EI4//vh1Pxt852//9m/DK6+8MnhT0/ezK3+XXHJJOP/888Pq1aubjjN4x/fee2/wainvx34Vnl35+8Y3vhHmzJkTVq1aFaXm5cuXR4lTZJAvfvGLUcNnV/5+7/d+L/zHf/xH6O/vjxb79NNPjxarqECXX355tNDZlb9PfvKT4eGHH4569fP666+PlmNRgX7+859HC509T/T29oYZM2aEWFdTsxfyS5YsiZajQNUSMABWq18tzXbrrbcOu+++e/jFL36x3uNmg1qsISM/SMyYKQ6AuWPWl1i9WblyZR62tMvYg3ReaDb8xYwd65d3nl8Ry7Vr10YPmxnGjFuFc7KIXmcxY8WNFSf6ySJgSwR8D2BLmKt5kOyqz8svv+yPQqrZPlkTIECAAIH1CrgCuF6a9H5w2WWXNd7yzf7o45e//GXIvhcwexvsj/7oj9LDUDEBAgQIEKixgAGwxs3d3NJee+218OUvfzn8z//8T9huu+3C7/7u74a77rqrcX9zY3k8AQIECBAgUF4BA2B5e9PyzLI/6HAjQIAAAQIE6i/gM4D177EKCRAgQIAAAQJDBAyAQzisECBAgAABAgTqL2AArH+PVUiAAAECBAgQGCJgABzCYYUAAQIECBAgUH8BA2D9e6xCAgQIECBAgMAQAQPgEA4rBAgQIECAAIH6CxgA699jFRIgQIAAAQIEhggYAIdwWCFAgAABAgQI1F/AAFj/HquQAAECBAgQIDBEwAA4hMMKAQIECBAgQKD+AgbA+vdYhQQIECBAgACBIQIGwCEcVggQIECAAAEC9RcwANa/xyokQIAAAQIECAwRMAAO4bBCgAABAgQIEKi/QEf9S1RhkQK33HJL6OrqinKIvr6+8NRTT4WFCxeG9vb2KDEXLFgQJU6RQR577LGo4Ts6PvjP+uijjw5r166NEvvee++NEqfIIA899FDU8JMnTw4HHnhgePjhh8Pq1aujxZ42bVq0WEUF2n///aOFbmtra8Tad999w8DAQLS4WTw3AgSaF3AFsHk7exIgQIAAAQIEKilgAKxk2yRNgAABAgQIEGhewADYvJ09CRAgQIAAAQKVFDAAVrJtkiZAgAABAgQINC9gAGzezp4ECBAgQIAAgUoKGAAr2TZJEyBAgAABAgSaFzAANm9nTwIECBAgQIBAJQUMgJVsm6QJECBAgAABAs0LGACbt7MnAQIECBAgQKCSAgbASrZN0gQIECBAgACB5gUMgM3b2ZMAAQIECBAgUEkBA2Al2yZpAgQIECBAgEDzAgbA5u3sSYAAAQIECBCopIABsJJtkzQBAgQIECBAoHkBA2DzdvYkQIAAAQIECFRSwABYybYVk/Trr78e5syZEw466KCw//77h09/+tOht7e3mIOJSoAAAQIECIybQMe4HdmBSyXw9ttvhxNPPLEx/N14441h2223DS+99FKYOnVqqfKUDAECBAgQIDB2AQPg2A1rESEb+nbeeecwd+7cdfXstttu6+67Q4AAAQIECNRHwFvA9enlmCp55JFHwn777RfOPvvscPDBB4djjz023HXXXWOKaWcCBAgQIECgnAKuAJazLy3P6uWXXw7z588Pp556ajjjjDMan/27+OKLw8SJE8Nxxx233nz6+vpC9i/GLY+TL2PEbGtrixGm0BgdHXH/M8zj5csYyXd2dsYIU2iMyZMnR40/adKkRrx8GSt4Fc7JmDnmsfJlLMfY/Y6V1+A4MZ/L8lj5cvBxmr0fM1azOdhv/ATaBt6/jd/hHbksAtnVv+zfnXfeuS6lbADM/gjkH/7hH9Zty++sWLEiLF68OF+1JECAAIGKCvT09ISurq6KZi/tZgXiXnpoNgv7jbvAjjvuGPbYY48heUyfPj089NBDQ7YNX+nu7o72xJG9Gs0GzhkzZoT29vbhh2pq/cEHH2xqv1bu9MQTT0Q9XHbl7/d///cbvVu7dm2U2Pfdd1+UOEUG+dCHPhQ1fHbl7ytf+Uq44oorwpo1a6LFPumkk6LFKipQrP/+svyyK3/Tpk0Ly5YtCzGvN1x77bVFlR8t7oIFC6LFKuL5MXshv2TJkmg5ClQtAQNgtfpVWLYf+chHwgsvvDAk/osvvhh22WWXIduGr2S/KGL+ssjix4wZ8xfO8Npjrcca0obnk8WNFXvVqlXDw5duffXq1YXklA1/MWNX4ZwsIscsZsy4MXtSyInzftDYz41ZnjGfH4vIryhLceML+COQ+KaVjHjKKaeEn/70p+G6665rfP1LdsUn+yOQz3zmM5WsR9IECBAgQIDA+gVcAVy/TVI/yb74+eqrrw5XXnllmDdvXth1113DX/3VX4VjjjkmKQfFEiBAgACBFAQMgCl0eRNrPOKII0L2z40AAQIECBCot4C3gOvdX9URIECAAAECBEYIGABHkNhAgAABAgQIEKi3gAGw3v1VHQECBAgQIEBghIABcASJDQQIECBAgACBegsYAOvdX9URIECAAAECBEYIGABHkNhAgAABAgQIEKi3gAGw3v1VHQECBAgQIEBghIABcASJDQQIECBAgACBegsYAOvdX9URIECAAAECBEYIGABHkNhAgAABAgQIEKi3gAGw3v1VHQECBAgQIEBghIABcASJDQQIECBAgACBegsYAOvdX9URIECAAAECBEYIGABHkNhAgAABAgQIEKi3QEe9y1Nd6gL77LNP6Qlef/31qDlOmPDB67rddtst9Pf3R4nd09MTJU6RQRYtWhQ1fGdnZyPeCy+8EFatWhUt9t/8zd9Ei1VUoFmzZkUL3d7eHnbZZZfw85//PPT19UWL+8gjj0SLJRCBFAVcAUyx62omQIAAAQIEkhYwACbdfsUTIECAAAECKQoYAFPsupoJECBAgACBpAUMgEm3X/EECBAgQIBAigIGwBS7rmYCBAgQIEAgaQEDYNLtVzwBAgQIECCQooABMMWuq5kAAQIECBBIWsAAmHT7FU+AAAECBAikKGAATLHraiZAgAABAgSSFjAAJt1+xRMgQIAAAQIpChgAU+y6mgkQIECAAIGkBQyASbdf8QQIECBAgECKAgbAFLuuZgIECBAgQCBpAQNg0u1XPAECBAgQIJCigAEwxa6rmQABAgQIEEhaoCPp6hU/RODII48Mr7zyypBt2cpnPvOZcMEFF4zYbgMBAgQIECBQTQEDYDX7VkjW99xzT+jr61sX+/nnnw+nnnpq+IM/+IN129whQIAAAQIEqi9gAKx+D6NVsN122w2JdcMNN4Tf/M3fDAceeOCQ7VYIECBAgACBagv4DGC1+1dY9mvWrAnf//73w6xZs0JbW1thxxGYAAECBAgQaL2AK4CtN6/EERcuXBjeeeedcNxxx20w3+wt48FvG2/wwRv5YR4nX27k4Zv044GBgU163Hg+aMKEuK/D8nj5MkZtkyZNihGm0BidnZ1R4+fx8mWs4LHjxcprcJz29vbBq2O6n8fKl2MKNmjnmM8Tg8KW9m5eb76MkWjMWDHyEaO1Am3v/4Is/2/I1po42vsCp512Wpg4cWK47rrrRvVYsWJFWLx48ag/s5EAAQIEqiPQ09MTurq6qpOwTKMIuAIYhbFeQbK/BP73f//38O1vf3ujhXV3d0d74shejfb29oYZM2aEWFcLli5dutEaxvsB//Zv/xY1hezK33777ReeeeaZ0N/fHyX2ww8/HCVOkUFivyDJrtR961vfCmeffXZYtWpVtNS33nrraLGKCnTMMcdEC539t/zxj388/PjHP472bkGWXNaXlG5FPD9mL+SXLFmSEqNaBwkYAAdhuPuBwPe+972w/fbbh0984hMbJcme3GMNa/nBYsaswucXYw1puV++zOLGip19JrTst5hD2uBas7gxY1fh7fQi3hrMYsaMG/t5Z3DPy3w/5vNjqoZl7m8rc4v74aNWZu5YhQhkA0M2AB577LGho8Prg0KQBSVAgAABAuMsYAAc5waU7fDZW7/Lli1r/PVv2XKTDwECBAgQIBBHwCWeOI61iXLooYeG5557rjb1KIQAAQIECBAYKeAK4EgTWwgQIECAAAECtRYwANa6vYojQIAAAQIECIwUMACONLGFAAECBAgQIFBrAQNgrdurOAIECBAgQIDASAED4EgTWwgQIECAAAECtRYwANa6vYojQIAAAQIECIwUMACONLGFAAECBAgQIFBrAQNgrdurOAIECBAgQIDASAED4EgTWwgQIECAAAECtRYwANa6vYojQIAAAQIECIwUMACONLGFAAECBAgQIFBrAQNgrdurOAIECBAgQIDASAED4EgTWwgQIECAAAECtRYwANa6vYojQIAAAQIECIwU6Bi5yRYC9RHYc889S19M7Bz7+vrCU089FWbPnh3a29uj1H/33XdHiVNkkJ/97GdRw2+xxRaNeM8991xYvnx5tNgnnXRStFhFBZozZ0600Pn5ePbZZ0c7H6MlJxCBhAVcAUy4+UonQIAAAQIE0hQwAKbZd1UTIECAAAECCQsYABNuvtIJECBAgACBNAUMgGn2XdUECBAgQIBAwgIGwISbr3QCBAgQIEAgTQEDYJp9VzUBAgQIECCQsIABMOHmK50AAQIECBBIU8AAmGbfVU2AAAECBAgkLGAATLj5SidAgAABAgTSFDAAptl3VRMgQIAAAQIJCxgAE26+0gkQIECAAIE0BQyAafZd1QQIECBAgEDCAgbAhJuvdAIECBAgQCBNAQNgmn1XNQECBAgQIJCwgAEw4eYrnQABAgQIEEhTwACYZt9HVN3X1xeuuuqqcOSRR4b9998/HHXUUWHevHlhYGBgxGNtIECAAAECBKot0FHt9GUfS+DGG28M8+fPD5dddlnYc889wzPPPBPOO++8sNVWW4XZs2fHOow4BAgQIECAQAkEDIAlaEIZUnjyySfDJz/5yfCJT3yikc6uu+4a7r///vD000+XIT05ECBAgAABAhEFvAUcEbPKoQ444IDwk5/8JLzwwguNMp599tnwxBNPhMMOO6zKZcmdAAECBAgQGEXAFcBRUFLcdPrpp4d33303HH300aG9vT1knwk855xzwjHHHLNBjuxx2b8YtzxOvowRM8UYuV++jGEwefLkGGEKjbHFFltEjd/V1dWIly9jBZ80aVKsUIXFiXnu5LHyZWFJ1zxw7pcvY5QbM1aMfMRorUDb+x/y9yn/1pqX8mjZ272XX355+Iu/+IvGZwAXL14c5s6dG/7yL/8yHHfccSNyXrFiRcge40aAAAEC1Rbo6ekJsV/oVFskjexdAUyjzxutMhv+squAn/rUpxqP3XvvvcOyZcvC9ddfP+oAmAfs7u6O9sSRvRrt7e0NM2bMaFyFzI9huXkCRTjOnDlz85IYh0c//PDDUY+a/UJ84IEHGlfFsxc8sW4nnHBCrFCFxcn+KCzWrYjzMVZuVYpThGN2Xi9ZsqRKDHKNKGAAjIhZ5VCrVq0KbW1tQ0rI3gre2AXi7DHZv5i3ImLGzK8qsWI6rl69uvRlL1++vJAcs1+SMWOvWbOmkDxjBo3933SWW8zzMWatVYsV07GIPlfNM+V8DYApd39Q7UcccUS47rrrwrRp09a9BXzzzTeHWbNmDXqUuwQIECBAgEAdBAyAdehihBq++tWvhm9+85vhwgsvDG+++WbYaaedwh//8R+Hs846K0J0IQgQIECAAIEyCRgAy9SNccxlyy23DOeff37j3zim4dAECBAgQIBACwR8D2ALkB2CAAECBAgQIFAmAQNgmbohFwIECBAgQIBACwQMgC1AdggCBAgQIECAQJkEDIBl6oZcCBAgQIAAAQItEDAAtgDZIQgQIECAAAECZRIwAJapG3IhQIAAAQIECLRAwADYAmSHIECAAAECBAiUScAAWKZuyIUAAQIECBAg0AIBA2ALkB2CAAECBAgQIFAmAQNgmbohFwIECBAgQIBACwQMgC1AdggCBAgQIECAQJkEDIBl6oZcCBAgQIAAAQItEDAAtgDZIQgQIECAAAECZRLoKFMycqmOQH9/fyPZlStXRku6r6+vEWvFihWhvb09WtzUAhXhOG3atNIz7r333lFznDJlSiPeXnvtFWKe5zvttFPUPIsIlv03GOtWxPkYK7cqxSnCMT+v8+fzKnnIdewCbQPv38YeRoTUBN58883w4osvpla2egkQIFA7gd133z1sv/32tatLQRsWMABu2MdP1yOwdu3a8Pbbb4fJkyeHCRN8kmA9TDYTIECgtALZlb/Vq1eHqVOnho4ObwiWtlEFJWYALAhWWAIECBAgQIBAWQVcuilrZ+RFgAABAgQIEChIwABYEKywBAgQIECAAIGyChgAy9oZeREgQIAAAQIEChIwABYEK+zmCdx+++3hyCOPDDNmzAjHH398ePrppzcvQOKPvv7668OsWbPCAQccEA4++OBw5plnhqVLlyauMvbyb7jhhpB9vcwll1wy9mAJRnj99dfDnDlzwkEHHRT233//8OlPfzr09vYmKNF8ydnXv1x11VWN58fM8Kijjgrz5s0LvsCjeVN7fiDgz36cCeMu8IMf/CDMnTs3XHjhheHDH/5wuOWWW8Jpp50WHnzwQV9NsIndeeyxx8JJJ53UGKCzXxhXXnllw/D+++8PXV1dmxjFwwYLZC9C7rzzzsYAOHi7+5smkH1LwIknntgY/m688caw7bbbhpdeeqnxF6ebFsGjMoHMbv78+eGyyy4Le+65Z3jmmWfCeeedF7baaqswe/ZsSASaFvBXwE3T2TGWQHbFL7vy99d//deNkNlXExx++OHh5JNPDqeffnqswyQV56233mpcCbztttvCRz/60aRqj1Hs8uXLw8yZM8MFF1wQrr322rDPPvuE888/P0boZGJ84xvfCP/5n/8Z7rjjjmRqLqLQL3zhC40Xwpdeeum68H/6p3/a+AquzNiNQLMC3gJuVs5+UQTWrFkTFi1aFA455JB18bLvFczWn3zyyXXb3Nk8gXfeeaexQ/b9Xm6bL3DRRRc1XoQMPi83P0raezzyyCNhv/32C2effXbjxcixxx4b7rrrrrRRmqg++1jHT37yk/DCCy809n722WfDE088EQ477LAmotmFwP8LeAv4/y3cGweBX/3qVyF7y3L4t9Bn6z7D1lxDsiuo2dWCj3zkI6G7u7u5IAnvlb1t/rOf/Szcc889CSuMvfSXX3658dblqaeeGs4444zGZ/8uvvjiMHHixHDccceN/QCJRMjeBXn33XfD0Ucf3fhfZGbPl+ecc0445phjEhFQZlECBsCiZMUlME4C2Wcpn3/+eW+9NeH/6quvNv7g4zvf+U7jLbYmQtjl/wSyP1LIrgB++ctfbmz57d/+7cZ5mX2u0gC46afJAw88EO67775wxRVXND4DuHjx4sZnprP/pzTHTXf0yJECBsCRJra0UCD7YHh7e3vI/t/Cg2/Z+g477DB4k/ubIJC9dfmjH/0oZJ/923nnnTdhDw8ZLJB9HCE797LP/+W37IrL448/HrK/VM/+gjU7X902LrDjjjuGPfbYY8gDp0+fHh566KEh26xsWODyyy9vfBb6U5/6VOOB2V+lL1u2LGR/+W8A3LCdn25YwAC4YR8/LVhg0qRJYd999w2PPvpo4+sNssNlb2Fm65/97GcLPnp9wmdXW772ta+FH/7wh+HWW28Nu+22W32Ka2ElH/vYxxpXWwYfMvuLy2xw+fznP2/4GwyzkfvZRxDyz63lD33xxRfDLrvskq9aboLAqlWrQltb25BHZi9CfA3MEBIrTQgYAJtAs0tcgewzQueee27j7aLse66yr4FZuXLlkKswcY9Yv2jZ274LFiwI11xzTdhiiy3CG2+80Sgy+6qIzs7O+hVcUEVbbrnliM9NZl+js80224zYXlAKtQl7yimnNL4G5rrrrmt8fi37Wp3sj0Cyq9Rumy5wxBFHhMxw2rRp694Cvvnmmxvf+7npUTySwEgBXwMz0sSWcRDI3rK86aabGoNLT09P+OpXv9r4TsBxSKWSh8zeFhrtln2/4uC3M0d7jG0bFsi+jsjXwGzYaH0//ed//ufGd1JmV/523XXXkL3YO+GEE9b3cNtHEcj+AOSb3/xmWLhwYePjCdln/7K3g88666yQvYPiRqBZAQNgs3L2I0CAAAECBAhUVMD3AFa0cdImQIAAAQIECDQrYABsVs5+BAgQIECAAIGKChgAK9o4aRMgQIAAAQIEmhUwADYrZz8CBAgQIECAQEUFDIAVbZy0CRAgQIAAAQLNChgAm5WzHwECBAgQIECgogIGwIo2TtoECBAgQIAAgWYFDIDNytmPAAECBAgQIFBRAQNgRRsnbQIECBAgQIBAswIGwGbl7EeAAAECBAgQqKiAAbCijZM2AQIECBAgQKBZAQNgs3L2I0CAAAECBAhUVMAAWNHGSZsAAQIECBAg0KyAAbBZOfsRIECAAAECBCoqYACsaOOkTYAAAQIECBBoVsAA2Kyc/QgQIECAAAECFRUwAFa0cdImQIAAAQIECDQrYABsVs5+BAgQIECAAIGKChgAK9o4aRMgQIAAAQIEmhUwADYrZz8CBAgQIECAQEUFDIAVbZy0CRAgQIAAAQLNChgAm5WzHwECBAgQIECgogIGwIo2TtoECBAgQIAAgWYFDIDNytmPAAECBAgQIFBRAQNgRRsnbQIECBAgQIBAswIGwGbl7EeAAAECBAgQqKiAAbCijZM2AQIECBAgQKBZAQNgs3L2I0CAAAECBAhUVMAAWNHGSZsAAQIECBAg0KyAAbBZOfsRIECAAAECBCoqYACsaOOkTYAAAQIECBBoVsAA2Kyc/QgQIECAAAECFRUwAFa0cdImQIAAAQIECDQrYABsVs5+BAgQIECAAIGKChgAK9o4aRMgQIAAAQIEmhUwADYrZz8CBAgQIECAQEUF/hdRgZtDUa82LAAAAABJRU5ErkJggg==\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(tree_images_pp_2[1][0], cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Features Calculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_features(trees_images):\n",
    "    mean_spectras, median_spectras = [], []\n",
    "    mean_spectras_norm, median_spectras_norm = [], []\n",
    "    \n",
    "    for tree_image in trees_images:\n",
    "        mean_spectra = [np.ma.mean(band) for band in tree_image]\n",
    "        median_spectra = [np.ma.median(band) for band in tree_image]\n",
    "        \n",
    "        tree_image = tree_image/np.ma.sum(tree_image,axis=0)\n",
    "        \n",
    "        mean_spectra_norm = [np.ma.mean(band) for band in tree_image]\n",
    "        median_spectra_norm = [np.ma.median(band) for band in tree_image]\n",
    "        \n",
    "        mean_spectras.append(mean_spectra)\n",
    "        median_spectras.append(median_spectra)\n",
    "        \n",
    "        mean_spectras_norm.append(mean_spectra_norm)\n",
    "        median_spectras_norm.append(median_spectra_norm)\n",
    "    \n",
    "    return (np.array(mean_spectras), np.array(median_spectras),\n",
    "            np.array(mean_spectras_norm), np.array(median_spectras_norm))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "(mean_spectra_pp_2, median_spectra_pp_2,\n",
    " mean_spectra_norm_pp_2, median_spectra_norm_pp_2) = calculate_features(tree_images_pp_2)\n",
    "\n",
    "(mean_spectra_pp_3, median_spectra_pp_3,\n",
    " mean_spectra_norm_pp_3, median_spectra_norm_pp_3) = calculate_features(tree_images_pp_3)\n",
    "\n",
    "(mean_spectra_pp_4, median_spectra_pp_4,\n",
    " mean_spectra_norm_pp_4, median_spectra_norm_pp_4) = calculate_features(tree_images_pp_4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_features_pp_2 = np.hstack((mean_spectra_norm_pp_2, median_spectra_norm_pp_2))\n",
    "tree_features_pp_3 = np.hstack((mean_spectra_norm_pp_3, median_spectra_norm_pp_3))\n",
    "tree_features_pp_4 = np.hstack((mean_spectra_norm_pp_4, median_spectra_norm_pp_4))\n",
    "\n",
    "tree_features = np.concatenate((tree_features_pp_2, tree_features_pp_3, tree_features_pp_4))\n",
    "tree_labels = np.concatenate((tree_labels_pp_2, tree_labels_pp_3, tree_labels_pp_4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_features = np.concatenate((mean_spectra_norm_pp_2, mean_spectra_norm_pp_3, mean_spectra_norm_pp_4))\n",
    "tree_labels = np.concatenate((tree_labels_pp_2, tree_labels_pp_3, tree_labels_pp_4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "TREE_COLORS = {0:'g',1:'b',2:'r',3:'y'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_spruce_pp_2 = np.mean(mean_spectra_norm_pp_2[np.where(tree_labels_pp_2==0)], axis=0)\n",
    "mean_spruce_pp_3 = np.mean(mean_spectra_norm_pp_3[np.where(tree_labels_pp_3==0)], axis=0)\n",
    "mean_spruce_pp_4 = np.mean(mean_spectra_norm_pp_4[np.where(tree_labels_pp_4==0)], axis=0)\n",
    "\n",
    "mean_birch_pp_2 = np.mean(mean_spectra_norm_pp_2[np.where(tree_labels_pp_2==1)], axis=0)\n",
    "mean_birch_pp_3 = np.mean(mean_spectra_norm_pp_3[np.where(tree_labels_pp_3==1)], axis=0)\n",
    "mean_birch_pp_4 = np.mean(mean_spectra_norm_pp_4[np.where(tree_labels_pp_4==1)], axis=0)\n",
    "\n",
    "mean_pine_pp_3 = np.mean(mean_spectra_norm_pp_3[np.where(tree_labels_pp_3==3)], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAoAAAAHgCAYAAAA10dzkAAAAAXNSR0IArs4c6QAAQABJREFUeAHsXQd8VUX6Pek9IYRAEnqvShdRsJdVsWLFrmtfu64KKlZcsfeu6NobuGtbu38LKoqISO8lBQiQTtp7/+9MmMd9L+8lLyEJKd/HL7xb5s6dOXPv3DNfmQlxi0BFEVAEFAFFQBFQBBQBRaDNIBDaZmqqFVUEFAFFQBFQBBQBRUARMAgoAdQHQRFQBBQBRUARUAQUgTaGgBLANtbgWl1FQBFQBBQBRUARUASUAOozoAgoAoqAIqAIKAKKQBtDQAlgG2twra4ioAgoAoqAIqAIKAJKAPUZUAQUAUVAEVAEFAFFoI0hoASwjTW4VlcRUAQUAUVAEVAEFAElgPoMKAKKgCKgCCgCioAi0MYQUALYxhpcq6sIKAKKgCKgCCgCioASQH0GFAFFQBFQBBQBRUARaGMIKAFsYw2u1VUEFAFFQBFQBBQBRUAJoD4DioAioAgoAoqAIqAItDEElAC2sQbX6ioCioAioAgoAoqAIqAEUJ8BRUARUAQUAUVAEVAE2hgCSgDbWINrdRUBRUARUAQUAUVAEVACqM+AIqAIKAKKgCKgCCgCbQwBJYBtrMG1uoqAIqAIKAKKgCKgCCgB1GdAEVAEFAFFQBFQBBSBNoaAEsA21uBaXUVAEVAEFAFFQBFQBJQA6jOgCCgCioAioAgoAopAG0NACWAba3CtriKgCCgCioAioAgoAkoA9RlQBBQBRUARUAQUAUWgjSGgBLCNNbhWVxFQBBQBRUARUAQUASWA+gwoAoqAIqAIKAKKgCLQxhBQAtjGGlyrqwgoAoqAIqAIKAKKgBJAfQYUAUVAEVAEFAFFQBFoYwgoAWxjDa7VVQQUAUVAEVAEFAFFQAmgPgOKgCKgCCgCioAioAi0MQSUALaxBtfqKgKKgCKgCCgCioAioARQnwFFQBFQBBQBRUARUATaGAJKANtYg2t1FQFFQBFQBBQBRUARUAKoz4AioAgoAoqAIqAIKAJtDAElgG2swbW6ioAioAgoAoqAIqAIKAHUZ0ARUAQUAUVAEVAEFIE2hoASwDbW4Frd6gj0798fjz32mOfE+++/Dx5bv36951hTbBx00EG48cYbm+JWLfYebBO2Ddtod8hpp52Gc845Z3fcOqh7/vHHHzj55JMxbNgwg9PSpUuDuq65JWot9WhuuGp5FAEnAkoAnWg04LYlEfxY/frrr9Vydrvd2H///U0nfdFFF1U735wOkJiwHnfeeWe1Yv3888/m3KefflrtnB5oHASKiorw6KOPYsKECeZDP2bMGBx77LG46667kJOT0zg3rUOu//3vfzFjxow6XNGwSYkDn9c1a9YEzPihhx4yaRYvXhwwTUs7UVZWhiuuuAKFhYWYPHkypk+fjvT09Earxo8//mgwJNb8GzhwIPbZZx9ThpUrV9b7vk1dj3oXtAkuvO6667wwtlj7/k6ZMqUJSlP7LTjgcJZtyJAh2HvvvcGBE/usXemfOPjjQH358uW1F0RTBIVAeFCpNFG9EYiKisKHH36IUaNGeeXxyy+/IDs7G5GRkV7Hm/PO22+/jQsvvBCdOnVqzsXc5bKRTB111FHNsm3Ky8txxhlngB/Y4447zmwXFxdj2bJl5jk79NBDd3v78HlneXaXpuzoo4/Gv//9b5CI/uMf//D7PLCM/fr1w4ABA/yeb4kHV69ebfqUe+65ByeccEKTVeHss8/G4MGDwWeThJr9BPu3jz76CCkpKXUux+6qR50L2gQXTJo0CePHj/fcad26dYYEkVANHz7cc7x79+6e7eawweePxI+Kjm3btoEa3WeffdYMDDkwOeSQQ+pczA0bNuDxxx9H37590adPnzpfrxdUR0AJYHVMGvQItXzUjt18880ID98JNz9A7DT5crQE4Uu3atUqPPfcc6YujVVmkpnY2NjGyj6ofMPCwsC/5ihffPEFFi5ciPvvvx8kOk4pLS01H2Hnsea+zTJHREQgNLThjBFDhw4FP4gkIP4I4O+//27M+9dee21zh6dO5cvNzTXpExMT63RdTYmDeR/32msvrw96jx49jLXgP//5D84999yasvd7bnfVw29hdvPBESNGgH9W5s2bZwggyR8HqrUJCRjfsejo6NqSNuj5Pffcs1r5SOzPO+888L2bNWsWevbs2aD31MzqjkDD9bp1v3ebuIKaJJK8H374wVNfmjj+97//VfuA2wQul8uMlHjtHnvsYcwqt956K/Ly8mwS80syQI3cuHHjQFU7R1VPPPEEKisrvdKdeeaZxlxI1Tm3+YHkqJJkLljp3LmzeaE5ug9GjU+S8ve//910XuysqCVg5+UUayantuC2227D2LFjjVmcaajqpymBpJNmkJEjR5oR5cMPP2xGlVlZWbjkkktM/vvuuy9efPFFZ9Ygxo888ojRhPBa+kRxNP3TTz95pfO3Y8tlfQBtWZymDbvt9NkLtt3YKT/55JPYb7/9TFuwTagxC0aoAaA4Pwr2Omqb4+Pj7a7xJyT2vOb88883GPBZ4SiaZXBKsGXnNd9++63RPDJvlmPixIlG28ZzrMs333wDjtYtRnQhoFh3ARIzmmD5DPJZpMmS78i9995r3gmbL5+f+ppoSY6pJf3rr7/MvZ3/cfAVEhJi3gl7/J133sFZZ51lnkG+c3z33nrrLXs64C+vYz2pzXeKNY/6un+QfLItiBvrTrx4zCkFBQXGnH/ggQea95rvBT+cNWHBd8RqXC+77DJTJrvPvFmeU0891dyT1gim8TXTWrM4j1911VXGasHy1VWstWPt2rXVLq2t/rXVg33Y5ZdfDpJOkgw+e3zenGLbhNiz36Qmyj6DTMe24ntLczX7TbpS+PqU2vb77LPPTJ/KZ5X3I6b2HXTek/Xi88q6s6855phj8OqrrzqTGNNlbWX3uqAeO3QHufrqq/Hll1+a/prPMjXhVogNiSPrwrT//Oc/sWnTJnva80vsWFe+i6wPt+fPn+85X58NOzDYvn07XnrpJU8Wmzdvxt13323eOd6LGF588cVept6vv/7avJ+86Morr/T0Ldb1iO3FwR77VLYp3x0OkvkdUAmMwE6VVOA0emYXECBx4kPNjx61gZT/+7//Azv5I4880piqfLNnpzVz5kxDXtgBk4i89tprRvPzxhtvGI0Jr2Eaass4yuYvyQ39LPhBveGGG7yyJXlkB0UT4RFHHGEIKF8QmsFsubwu8LNDwvXBBx/UqgUkmTn99NMRFxdn7knNJz+mrAs7RX74nHL77bejffv25qNEjYNT2Jn17t3bjBpJPJ566im0a9cOb775punY+cFgB0fywM5u9OjR5nJiwM6OnftJJ50E+s29++67pjw8Tn+lYIWYdevWzSs5icXLL79sym1PBNtuJKasB3HnH/PiB54mtNokIyPDJOEI+tJLLzVEpqZrOBhguxPz66+/Ht99950h1zzOjtRKsGXnh5L+ZdQI03c1ISEBixYtMvmSdLHj5rPNj+xNN91ksudz4BSSX2r9SITYQXObH3YOaP72t7+hS5cu4EeBzwzN3Xx36up2wLKQ6FpNu70/6/3JJ5+Yj4zFkuf4XtEcTKLA55UfUGJCokzi1BDCQSAx43NKXz3Ke++9Zz5svD8/XJRbbrnF3J9179WrF7Zu3Yq5c+dixYoVAU3WHNzQ349mNmuS7dixo8mPbc77UivK+5aUlJh+h2ZE9iFOHHgBSQq1M/XVkNqBk68mMpj611SPJUuWmEEcy3vBBRcgJiYGH3/8sRkIsq0PPvhgU1/739SpU40JmsSApIOyceNG0x+wjYkv+xL2x3xW2ffwmFP4ntIawHeIfegLL7xg3iP2P1Z4Pd9F4k2iRLM324qkxeZX17LbvOvzy/eRhIjty2eX/SflgQcewPPPP2/6RB4n8aOrBIkd32treSGh5gCBgxT2EXwH2Gey/6YCgAOe+goH66mpqfj+++89WRArPqOHH344+L1kG/F94P347vPbMGjQINO3PP300+Y43yEKiSyF7zkHscSbzx0JOduK2mS6RKgEQEAaV6UREJCO3S3kyi0vl1tIj1tGUm7peM2dpBN2y8NttmWk4hYtnqcEc+bMMdeJ+cRzjBvSyVQ7bvNzJpSPh1s+9m5R+3sOy0thrpXO3nOM5+VldEtn7zkWaMNZRhk5u+Xlc4sW0CQX0mnylo+q53LpDN1i3naLBsBzjOmJgRBDzzGLkXRU7oqKCs9xbgiRNfmyPlaYRkZ4bumA3M8884w97JaO2S0dgVtIr+cY0zox4Ammk1G/Wzp7TzpusJ14Pyu2XDLSt4e8fqVTcR9wwAFuIZduIZbmXLDtxmuJDdtcOixPvg8++KAph7MOnpOODba5dJQmLduF7SGds1sIkyNV1SbzYt0keMdzjvfkvVkGloUSbNnz8/NNGwqhdssH1ZMnN5x1Yf4sm6/YZ0U+1J53waZhWwk5s7vml/gLKXLLx91znMdYJ7ZRbSLaIfO8OPO175F8wL0u9/cuCZFyH3bYYV7p5MPp5nEr8kE05RGNtD1kfoXsmOPElsIysN5CXMy+/Y/PD7ESgmEPuWXA6JZAFs9+sBv2np9//rnXJXxO+a7z+bcigw7zHjnfBfsMyqDKJqvx196P/QqfJb7jMkgz9RQy7V6wYIHn+rrU3+brWw/2Y6K98nqvme+JJ57olkGt5162TZje2fZMwHdCtHluIdWe9NxgPyiDR0/etgzETgYpnrRiaTDtKgMWc0wGbaYvYNvy/XCK850ItuzO6/1tC7Ex95cBoL/TbtGMmvOiwfM6z/Ky35RBq9dx8c3zOs76sI8V0uyVTgZ1brEeuIUYeh333RGia+7/+uuv+57y7PP9YVlsn8++xIkVE7K8Mkh3v/TSS57rbP/h/NbYk/7eXz7PzMP2czat/u5EQE3AAYhxQx6mxk0+cGZESM0UR1i+/lv2flRpU6vCkdKWLVs8f/QX5AiNZjQrTr8O5sv0VJ/Ly1DNvMNrnT4jDD7hKMqfOcPm7++XI13pVI2mwd95npPO05iju3bt6knC0TG1cb/99pvRUHpOyAanrQjkcyeduycp01BLIo8vnMc54qPGwlkXprUBNtK5GBOjdDjmepqn6yusHzUj1CjS3G5HzcG2G0fm1PTJB8FLeyedYlBFYptzNE7tGYUjd0YA0rTLKG1/Jg9qY63Q9Ml9lmH27NnmcLBlZ7uy3kLwQHOzU5hvsMLgFeezy+vYVtYPkBhT60Vs2a71bS+a4aiJFBLmKRo1BdQ4UtPoFGd5qMHku0QzI/2WfLXSzuuC3RYyZJ5PvvfO95qaKZri6AbB55rC959O89SE7KrQVYKRmTSVOjVy1KjQNEqtuq9Qc1QXEVJlTOc0k1IzR7zuu+8+4+Ns86lL/e01zl9iRoxoNbF9HY/RdYD3pRaJWmOnsF+xzxSPsx+gSZeaQm4724F5UMNH7ZlTiBufFyt0J6FYLSfrlZmZaTR/bDen2HeiPmV35lPXbRkgGZcZ53V8x9kn0prhrDe17dQc2+8K3XT4zrCvdqZjv8L3waZz5l3Xbb7XfNbte8W+xGLFPprvfnJystEG+nPh8Hc/5/vLfFl2ajDZl1D7quIfATUB+8elQY9ShU0/Hn582OHzoaS6259w6gp+gJjen8hoxnOYplb6xNH0y07RKczDKWlpaZ6XzB5PSkqq88tBUscPK00BJAK+whePBNSfgy9NEex4+VGiCdEKO6FA4mueYifLDoOYOoXHfQNqaN6ibyD9CEl4rNR0P5sm0K/FWzSQXmbhYNuNHwsK/WGcwvqwPYIR1pW+O/yjrx2JHOtJ8zp9AGk2t8IPoJOI87htG15LCbbs1qfL2XYmgzr+5w9/PhevvPIKRHNgPq58R6zQTFcfoR/fv/71L/PekWRxECZaJeMn5Is1fZ7oPkFzGJ9fp/BdskTfebwu23ZKGrosBBKSa7Yf25UmSboHcMDDXw7efNsxUD7O47aNbZs7z/F95LPDj7sdLPG8v/ZxXue7TZMxP7b88JJo0L/ZSbyYvi71982f+yTiFJox+edP2Dd26NDBc8q3HjR5EmM+Y/zzJ87+led9p9Gxzw3JIiWYd6I+ZTeZ1/M/33ozG+JPciWWC7+5WhcLW1brouAvMftSJyn2l6amY3xOSPjsO8Vysf+iywf7R/YFVqz52u4H+mU70LWG5njRxHol8/0Wep1s4ztKAJvoAeCIir49HKXSUdU5GncWgQ8/fUjon+dPLPHhQ04tEj8YfFnpo0ZixBETr3W+RMwnkIbN3z1qO0ZfQEb4MYikPuH8vvn7apOc530/JDwXqC5Wg8I09FWkozfLR20ZMeV1JG5OTSHTBiv0UWOd6RfDNnRKsO3mvKYhtukzQ20oR/asK/0hnQQwmHs0ddmdo3VbPvr2sAOnxoX48kPLtp82bZpHM2bTBvvLNqejP7U+9Of76quvDAHw1b5zgHCO+G6R2PKZ4UefHzimp4+U87kK9t5OAstr7PtIYkcNjT+xuLCvoC8rySq1rvTb4nNHjTM1vY0tthzB3oe+k8SZwmeQBJpaaWrLLLGoS/393ddeTw2jvZdvOl/i41sP247HH3+8GcT6Xs9932mBgulr/OXjPFafsjuvr+u2b715PctAks/+z5/Y4DGLEd8Xf4MGXhsIE3/5+jtGf1/2WzYfDqr5fNMvkQM1fhv57lsfXH95OI9xAEMLCokpfZA5uKZ/KEkh/UAt/s5rdLsKASWATfQk8APNh5EqdkbbBRISOY7KOaL29yLb62gOocaLzs828IHnrGnCpmuMX5aRWkCO2HwDOkhQ+fLxo+orjC7ki+07qvZN1xD71EJQY0J8rHmB+VLLUx9hfWjq4geOnYyvBNtuVqPJkbZTo0PNqdUq+OYdzD4JE/PzjSZm50fC6+zMbduwE6YEW3amo/AeDCgIJE68A6XxPc72YudPwucUDnRoDqqvkOzRwZyaAWrg+aFzRoQyXxI9fjxIQi1h4XGSr9rEaoRYTmrZrVhNr9232PH+gQiMTctfloMDPP5x0EizOT/edSWAto1tmzvvwfeRGjOn9s95vr7b1HLSVMvy8iNOqWv9fe9trycxDwY/3+u5z7pS68R3or55+OZry8V3gs+vP7FpdqXs/vKtyzGWgUSJGjXnM+6bh+2TSMIaCiPnPfhOURNLsmeF7774wYLBgFZIRH37w0D9CrX2fN84W4P47NoszGDPs6MbfhFQH0C/sDT8QUZC3nbbbSbCzvcD5Lwb/QWpPWCkpK9QVW7V21YzZkdsTMsXPJBpwzevXd2nFpDloXbCKRzV0X+RUZROMsqPGD/A1ArY0abzuobetqNLJz70q/KdiiaY+9JsxEhCdpw0KfrriIJtN3aq/BDQXOssGyOKgxFOBUKy6Cs09dEPykn0bBpGkFvhPbnPMlg3g2DLTvLB55gfdppTneKsCwcAdTW7sL2ceTBvRusGM+WQsxy+2yTsLA/fC5JAfiB8Nc72WXFey48PI61rE/txd/oZ8r3g4MgpjFYkGWNkovV9cp63bcprfd05SFwYOcn3u67CwRY1jhI049UmfI7oOkLzckMLn0HiznvaegVb/0BloQ8x+w5Gh/r6+vEae59A1/M4I39ZLj5X1EL5SjB5+F5DP2oO6mbMmOGFL9PZ57khyu5737ru0+eV/Ra1yL5CQmzdZ6h4YHmpkaO7kq/UByObB83QtILxfXTOD2m/ZTYdf+nbbL919jivo/j2Lfb9tXgzDb+h1N6r1IyAagBrxqdBz9L0UJvQ0faUU04xH1k6JJNM8WNNjRH9a2ha4cvM+ZmofaDJiuHyfLlp9nS+BLXda1fO88NHLSD97HyFc4gx2IFTOvCPLyg/iPyAcSqSphD6utD0x+kMuE0yyqkbOIO8vw9wTWWiFpEfDJJeElunEAe2RbDtRg0pp3whieLUHPwAM8iB5CQYTRdH0BzpchBB7Ss1GqwbP7bEl/5YTiHZoQaM2kt+hLn9zTffGC2mdScItuwk7jRhclJzmp1pqqSmgGSCHwtOxUNhwBKn5+D0C/xAsow1DXp4DduIHyfmTzwZuEBzttVIME19hISVTv8cfFB8zb88RmLLd4w+rQwcIOHn80ri5Y9s8BorNBmyjnS7oPM6/TM5dYWv8B3gXGe8B3HjSgn80JLgUuPPtmf9+dFjefmOc7oNfvTY5nxG6rvcF30K+axR60ITO59/fhzZdhzYNIbQ7YLvHwc2dEkItv41lYUDaAYwET+2E02+bB9OkUPfPX99kW9+7H9I1jk1FP+oESPZZzAHj7Mt6iKsFy077Gfop8l2JVmndpV/JFKUhih7Xcrlm5aDAD4D1HJTG8z3jc8WrQNsJ05zw+AfaoO5lCLrw3eFdeJzysAQ9ukkuxJd65t9tX1q5fje83tEcsl9ujTwPeP1NNNaofZPon2Ntph9B591lsnXUsS2olWMvsL83rH8HBRwSi9q3xkER5LJNOx/fH157f30dycCSgB3YtFstu644w7j/E3CQnMxOxlqD0i4OEKj8IPBl5kfXfpQsDPneWp12Pk2hVhfQF9/J/pSUctEZ20SHXYCJB+MDPQ1GTdWOdkR8+PADznnnCLx4/1Jomk+r4vww07hnGC+QlJPwkIJpt2YjgSZHS3bl1F1xIZO0OygaxNqsEhQSAqoweHHi23PPDiqZmSnU/jsUEvLDxDrT0LEjz47eKcEW3Z+NOlbx/nmqKWmVoVz1Z0jPnRWSPo5eOEofoZoRvjs1kYAaVZnh03Sx86bUap8dgI5/Nt7BfPL94IEkB9mX3x4PZ8NvkP0QaSGlx88Eg0SXmosahOWkaZOlpeDMpJjfpj4UXUK3022OXEjASMRY5n4TliTGO/JbbYvTWN8d2huZ/twYFgfYYQriQgHDqwj24zmSppqrUtCffKt6RrOfWo1dvTbY72CqX9NeVpNJuvBAQ+ffT6LJACcnSAYYdtyPlAO6kgy2EcwyIh9Vn3nPSSZItFlntTwUjgwdLZXQ5Q9mPrVlIZEnPWk9YEYkkSRZPHddLoWcFBKjTmfU5Itvpd8TtnPBRshzneff3zW2PbUCvM5ICa+Jmj6/NIFg30zfcs5oCKOHGg6hYSSg0q68fB943eHzzMHS3z3OMBiH01iyGMkrxwoqARGIEQ6mKq5BwKn0TOKgCLQAhGgdpgkgpOiqigCioAioAgoAk4EQp07uq0IKAKKgCKgCCgCioAi0PoRUALY+ttYa6gIKAKKgCKgCCgCioAXAkoAveDQHUVAEVAEFAFFQBFQBFo/AuoD2PrbWGuoCCgCioAioAgoAoqAFwKqAfSCQ3cUAUVAEVAEFAFFQBFo/QgoAWz9baw1VAQUAUVAEVAEFAFFwAsBJYBecOiOIqAIKAKKgCKgCCgCrR8BnQh6F9qYyzZxMlKutuBvOZtdyFovVQQUAUVAEVAEFIFGQoBL4HFJS07ezgmr26K0zVo3UEuT/HGJNhVFQBFQBBQBRUARaHkIcFk6rijTFkUJ4C60ul1Ung8Ql59pSOEyN1wPlUsIcTmvliQtuezB4Nya6teS69KSy97WnrNg6ru70rT252h34doQ923MtuESd1Tg2O94Q5S3peWhBHAXWsyafUn+uE5hQwoffArzbYkEsKWW3YBey38tuW18q9aS69KSy+7bDv72W3v9/NV5dxxTnHcH6sHdsynaxn7HgytR60qlQSCtqz21NoqAIqAIKAKKgCKgCNSKgBLAWiHSBIqAIqAIKAKKgCKgCLQuBJQAtq721NooAoqAIqAIKAKKgCJQKwJKAGuFSBMoAoqAIqAIKAKKgCLQuhBQAti62lNrowgoAoqAIqAIKAKKQK0ItGgC+Nprr+Gggw7CHnvsgZNOOgnz588PWOHPPvsMJ5xwAkaNGoVhw4bh2GOPxaxZs7zS33jjjejfv7/X3/nnn++VRncUAUVAEVAEFAFFQBFo6Qi02GlgPv74Y9xzzz24/fbbMXToULz88ssgWfv000/9TurI2b4vueQS9OrVCxEREfj6668xefJkk3b8+PGeduQ287USGRlpN/VXEVAEFAFFQBFQBBSBVoFAi9UAvvTSSzj55JMxceJE9OnTxxDB6OhovPfee34bZsyYMTj00EPRu3dvdOvWDWeffbbR9P32229e6Un4UlNTPX8kjiqKgCKgCCgCioAioAi0JgRaJAEsKyvDX3/9hX322cfTFpzMkfu///6751igDbfbjdmzZ2PVqlUYPXq0V7JffvkFY8eOxeGHH46pU6di69atXud1RxFQBBQBRUARUAQUgZaOQIs0AZOUcYZw3/X7uL9y5cqAbVJQUID99tsPJJAkjCR4++67ryc9zb/UEnbp0gXr1q3Dgw8+iAsuuABvvfVWjatxsCx2xnJPZru4YfOzv7uYXZNebstsf5v05k1wM1sv+9sEt2y0W9g62N9Gu1EjZGzLbH8b4Ra7NUtbL/u7WwvTim9u8bW/rbiqLa5qtk3sb0NWoDHybMjyNUVeLZIA1heYuLg4E/hRXFxsNID/+te/0LVrV9A8TDnqqKM8WdtgkEMOOQRWK+g56bPBNXsbS/7888/GyrrR823JZQ8GHGqR3333XbCemzdvRmJiogkyYkBSQy8NGEx5diVNS26rllz2YNqstdcvGAyaIo3i3BQo1+8e2jb1w622q1okAUxOTjYaudzcXK/6cb9Dhw5ex5w71Pp1797dHBo4cCBWrFiBZ5991kMAnWm5TXLIe61Zs8aYhX3P2/1+/fo1+AefoxM+9IxwbolrAduysx6tLZDGtg2fNboT3Hzzzca3NDMz0/iivv3223j44Yft49Gsf21dWvpz1tLekWAeipbcNsHUr7mkUZybS0tUL0djtg0VQY2pvKlem+Z3pEUSQBKKwYMHGy0eNXQUl8tl9s8444ygUeY1NAcHkuzsbGzbts0EhARKw+P8+DTWB6gx8/ZXJ0ZRP/HEE4b0xsTEgET5ySefNASX0+Tk5+dj0KBBePXVVw12EyZMMATIkrwzzzzTBOVs2bIFP/30E0iOGVV98MEHG+0r86MwH/pfvvLKKx4CvmzZMtx///2YM2eOIVZMSy0tg3Yo77zzDl588UWsX78enTt3Bu91+umnm3P+/uP5vn37mlMffPABwsPDcdppp+HKK69ESEiIOc5phBhIxMHAV199hYSEBFx88cU15mvvNWDAADz++ON2Fz179sTVV1+N66+/3pSf92sp0tTPWUPi0pLLHgwOrb1+wWDQFGkU56ZAueZ7uOV7XC4D6bL1G1Au/XypuGJFrF8H902TEZbWqeaL63iW7d3WpeV8oXxa6txzz8UNN9yAIUOGYM899zTTwJSUlJi5/pj0n//8Jzp16oRrr73WXPnMM8+YtCQTJH3ffvst/vOf/+C2224z54uKiszHnMEf1OzQB/C+++4zGkPnNDEm8S78R41RSXllrTlw5LO9woXisgohl+5a0wdKEBMR5iE7gdLY4xs3bjR4kcCQWBOTX3/91ZAZm4Zmz6ioKPz73//Ghg0bcNNNNxktKYmPFc6vSGLFeRqpdQ1GcnJyQPK+1157mbaMj4/H3LlzUVFRYS5nWz3yyCO49dZbDSldtGgRbrnlFkNMjz/++IC3mDlzJk488URDHhcsWGCuz8jIMBHk9qIXXnjBkL7LL78c33//Pe6++2706NHDyz/Upq3tt7CwECx7SyJ/tdVJzysCioAi0BAIuEXpUiHfmXL5vlqSR6JXtmG9EL4NqJDvgHxwvG4VIXuFEpgZdeqpXsd1Z9cRaLEE8MgjjwS1TI8++ig2bdpkSMHzzz/vMQFnZWV5kQ+qezlnILV6nC6G8wGS4DEfCkcDVAeTvDBYpGPHjoYAUFtktVu7CjfJ34lPz8Zva+oQWTzzi1267ajuyXjn4rFBkUDiSMLFQBhq2Cj0hXQKsZg2bRqoHaR27YorrsD06dONVs2SPZrZJ02aZDRixJUau9qEZJHEiYE3nKeRQo2alcceewzUQB522GHmEM3zy5cvNwE6NRHA9PR0M98jNX5sc7bxjBkzvAjgiBEjcOGFF5p8eU8ST6ZxBgjZctT0y+eR2tJTTjmlpmR6ThFQBBSBVokAv3GVYjUjqTPEzvxWafPMMdHuucvLa6x7RUQktiR0wIboZKyLaoc1iZ1wVO+9sNNDv8bL9WQdEGixBJB1pMYokMmXGiqnUEPl1FI5z3GbpJCaoMaWKsNjY9+lfvnTpMkpcI4++miMGzfO/FEj6pwLkYSQ5M/K8OHDQXJNwm1JI83zdRVq9LhKiyV/zuuZ/9q1azFlyhSj9bPnSFZpsq1JOEm4NfcyHVeB4RyS1LBaEwCPOYX7nFi8LkLN30UXXWR8Af/xj3/U5VJNqwgoAopAi0HAJZYho70zWjvR3vmQPJf01zWJS5QCRUkdsDk+xRC8lRFJyI5tjxz54++2qHhIp+3JolNcGC5OS/bs60bDIdCiCWDDwdA0OZGIUBsXrAmYS9vRvG2JSn1KWRcTMO9DckQN2A8//GDMvA899BAY1ECNW7DiJIi8xmoGOTq0Yk27dp8EPJCQAFLuvPNOs+qLM53N23msqbdJ/v7+97+DUeb0n/RHYpu6THo/RUARUATqg4CvH165ED0nyasMYm7c8nbtkd8uFTlxHbAmMglLQxOMRi8nrj02RyfBFertf5cQHY5eqfE4sEMcejr+uiZHY/miBRickVifqug1tSCgBLAWgBr6NElgbGTtsFdWhiA6PNSk3RUCWNfys3wjR440f5dddhkOPPBAfPHFF6DPJWXJkiXYvn270Zhyf968ecYPj6bWQNK+fXtziiZmK9T4OYWaRfrrlYt5wJdA0SeTJnn6ZR5zzDHOy2rd9l0f+o8//jB+nU5Mecwp3OeKMcEIyR+XIKRp/KmnnjL+kcFcp2kUAUVAEdgdCHj88ERz5yR21mTrzw/Pt5xusbyUdkjDNtHkZcW0xyrR4i1yx2N1ZDtsjE1GWViVG4/zukj5nvVMicMIErzUKqLXawfZax8X6WWpsdfRUqPSeAjUzkQa796aczNDgMSHQR70feOk2tynXxt956wwgIamWK6rzCAQ+ubRDF+TJo7aPZpVOeUOJ9nmdD2+06Qwmpdm+2uuucb449G0S3JJDSjvT1/Du+66y5h8GZTDcjCog9HElpzaMjp/OTULo5Dpl7dw4UITvczgIadQ4/ncc8+ZwJcff/zRrCfNoKHahOSPE4Uz+Ij+pNznH4Wk10kya8tLzysCioAi0BAINIQfXoj02e60dBS374QtiVX+eCvDE7HAFYfFIQkojtjpBuQsc6hYbru2j/XS4lmNXkZSjHwndpp2ndf5bpdtL8G27CzkbliHrKVLMFjck8LEwqLSsAgoAWxYPFt0bgzC4BQs9H8jkWG0LAMv9t9/f0+96CPIIA8SNpIwTgPD6NnahIEjJI4nnHCCCe5gpPF5553nuYzzLfK+JFKcvoWEktPAUBtJ4eTK1k+TQSecaJlTzHBN55rkuOOOMxpLXk9CdtZZZ1UL0iCBJJmk+ZYYsM7BRH6TUFrtIQNnnPLll18asus8ptuKgCKgCDQEArvqhyfTFCBCrDaVndJRmJyKzRJ0sV4CLmiqnV8hJK80HG4EJmsdE6IMyeu1Q5PXs0O82e/aPgZR4d7m3UD1rRBrT15OFrZkbcC2rExszZa/HduFW7d4XUYXpKGHHOF1THd2HQElgLuOYavJgWbPYAJhqI3jnz+hFo9qe2rvnMK833zzTechY052HmAQSk33Z3AK/+oinI6FxJMR4IGEpI9TzNRVOGUNTeIqioAioAg0JALGD08C64yJdp1E1NbDDy88NRURYnEhySsQf7yNcSnGH29RqGjySiOxemspyipdVcXmT4l3DRKi6JdnffKE4Mk2TbY95C9ezgUjLvkW5G3KqSJ4Qu6qSB6JXibyN2+sNuWLM8/w2Bi4hVAWxQIpg71no3Cm0+36IxBcK9Y/f71SEVAEFAFFQBFQBBwI1OiHJ641FTJdme98eI7LzWZoUhIiZboukjyaa/OSUpElUbSrwumPF4fleeVYtbkIhaUyl+p2uYR/RuhXV8X26JfXI8WabOMNwbP+eSkB/PJ2ZOL5YV0Ktmw2pI7Eblu2ED1q9OQvb2M2SAIDSXh0FMJTElGeGI5tcWXIityGVSE5yI0tQVnEDnIqF/fJm4NT0ne6IgXKT4/XDQElgHXDS1O3IQQ4+fTUqVOr1Zg+NvTx+/zzz6ud0wOKgCKgCDSUH15El85C8roYkheSnmGCLtZHt8cK8cdbXuQ2BI8kb3O+rGiVb3EnudupzqPbXZdkS/JEi+fR6sUhXfzywoLwy2N9ivO2CanbqcUzZluabMVXr6I88IpaYRERiE5tD3e7aBTFu7AxuhBrwzZhZWg2SiKFHPqxNEeGRqJfUg/0TOyJhO0JOKKHmn9t6zbkrxLAhkSzlefFZdlakvjOBemv7Fz+LZBwNRPOI+grNHG39TUkfTHRfUWgrSGwy3544pNMPzxq8AzRk9+wjM5Gk7dWJkFeWR6BVbnFWCkEjyRvw4qSHUrBPIGaf96Sav3ydkTWMviCZI9BGcH65ZUUFnj541Vp8kjyZHk2CXYLJKFSl7jUDghrH4/ShDBsjd2OzIhtWBayAVmh4s/nh+Qxr6SoJPRK6mX+eib1hP3LiMsQYhrmcSeKi9AAkEDY78pxJYC7gp5e26oRoG8g/3yFBHBrEHNh+V6n+4qAItByEPDyw9uxNm2VL56sbCFTUgUzH571w3OSvHAx2xYmdxQtWDxWbd1eRfA2VZG8NT8Xi19eroDEv+pi/fLoh2eja3tJAEaPDrFIiK4+9Ur1HABG2FaZasVMm2k1evzNwvYCjxqx+qUyRVhih1REpSajMikSBfGV2BhZgNXhG7G0Yi1K3CurX7PjCAldz3ZC8ESj16tdL89ve9Fmquw+BJQA7j7s9c6KgCKgCCgCuxkBl8xmEPbDj8j98UdUbMisWsKMfnhcl1b822oSpx+ek+Rxu7R9KlYXVmLhDg0etXirsuRvfh4KSv0TPN6rIfzyKqRO2yTC1gRdCMmjBs8GYBT5RNj61i8+uT0S0johJDkO2xNDsSW6GOsjtmCpez3Wl/wGl9uBCef237GyW0RoBLondjdaPGr1qM3jL4/FRkgkh0qzQ0AJYLNrEi2QIqAIKAKKQFMgwACGTJnwPurH2X51bpwPL8IEWuzwxZPpSKy5liSvIiYWa8VMu9SSPGryFhZh5f+txObCRQGrIMo08cuLES3ejsALh0Yvo11wfnk2wtZOnbLFBGAw+GKDRNjKpPuOlZd8CxKTkIjk9M6I6tAOFe0ikR9bjuyofKwMycaK4lXYsv33nZfQvc/h4pcQkWC0eU7TLbcz4jMQHqqUYidwzX9LW6v5t5GWUBFQBBQBRaARENj6+hsoFvLnlpV8ko6egCgSPBN0IYSPPnmyCpFLtFyZ20p2kjwheytnF4hf3hxs2FpizgcqmtMvz5ptOZUK/fKiI2qfL88TYSsT2nvmydsxX17expwaI2wjhZyS5CWlpYk2LxbFCZBl2IqxLmIzVpSswer8b1FSscOvj6tt+izhmxaX5jHVeky3otVLiU7xu2pHIAz0ePNFQAlg820bLZkioAgoAopAIyFQunIVNt5/v8m97LRTEX7plcYnb5Vo8VZmiyZvwRoheQuxWjR8ZRUOs6dPeeiXZ6dOqY9fnifC1vrjkeDtMNvWFmEbHhmFZJkCpp1ECDMIoywpAttiS2U6lTwsLV8nJG851hV8jUq3RNvSvc/HxS88JBzdErt5BWFQm9dDInB3W+AFNZcbRXu66v8QuvJbDMhaDHR6XlSmVYsC+MCvu7uAgBLAXQBPL1UEFAFFQBFoeQi4KyqQKSv+uGVd88IhI3Be8R4ouOfrgBWJDAtFdztf3o4Jke3qFx3i/a9j65uZJ8LWTqWyg/DVHmEbjnad0gzJS06T6NiUBDOdyqaoIqxxZ2F+/mqsyvsCm0rE7LtV7so/HyGZI7GzUbbWP69LgkQgi+/ebhUSvq2rDOFzr/w/5C1djMy8NGSVDUJm+dEorDwHR/6Zhe5ddmspW+XNlQC2ymbVSikCioAioAgEQiBX1v7ePn8+JMwfV3SbgILyEDFrVvfLo9mWJttg/fJshK31y+OvMd3WEmEbEhKKRFm5gybbdkLyaLYtTwrH1phSbAjLxTJD8hZhVf7HKBITNDb7r1nHmI7eJE8ibkn8UmNSm5fZNm8DsPo7uITw5S5eicwtKUL4BgrhOwolrjO8Khca6kZoz1Fex3SnYRBQAtgwOGoubRSBW2+9FT9K9ODGjRvN+sTDhw/HddddBy59p6IIKALND4Htsob3pieeNAV7c+wpYi5NxLBOkXj9kv0RGx1Za4E9EbYkd57AiyofvVojbNunGJJHTV6ymG1jUlPMdCo5EflYXbQGf25bKSRvHtZtXoeKTbKChx8JCwlD14Sunihbq82j2TYhUhz9mqMUCWMVwlex/Hux7q5B5uYkQ/iyyo9Euds7Qlim/0OnHolI75uMtF4J2Fi0Fhl92zXHWrX4MikBbPFN2LwrUCbTEUSKg3VrlcGDB5v1idNlQte8vDw89thjOP/88/Hll18iTCZHVVEEFIHmg4CrtBSZN9wAiAl44/B98HLsACRGh+PS0UmIcgRlVMr5fFnDtmoaFe8AjFojbBOTvEgeNXqQiN9cCcBYu32DLNEmJC/vD6zMm4mNS2Q93AASEy5RwjumUnH+dkvohoiw3Wy2DVBmz+HtMlH1mh9RuuRHZC1ah6yNccakm1N+GFzwLjs/D2l92iGjX3uky2/H7gkI39EWnHN1y7x1nmx1o2ERUALYsHi2+Nw+/fRTPPHEE1izZg1iYmIwcOBAPPnkk0a7daP4zOTn52PQoEF49dVXQXI3YcIE3HzzzR6Sd+aZZ6JPnz7YsmULLr30UvTr1w/33HMPDj74YMyaNcvkR5CYz+jRo/HKK69gzJgxBrdly5bhfnHKnjNnjsxg4DZpufpIt27dzPl33nkHL774ItavX4/OMjUD73X66aebc/7+4/m+ffuaUx988AHCw8Nx2mmn4corr/SYQ7jax8SJE7FixQpwVZCEhARcfPHFNebrvNcpp5zi2e0iUYNXXXUVjj32WGyQecRsuT0JdEMRUAR2KwKbHnkUpcuWwy1z3V3T+XBZoSIEtxzeC6ULv8K3y383y5rRJ6+2CNuo2DijwSO5o9mW2ryETp1QHO/GhvIcIXcr8VfeKiF6c7Bq1SoULC0IWG9G1TonR7YRt51iZS4+2qVbgpRJCPG6n1G08CdkLlyPrJwYZIpJN7difyl9qFcNYmPdSBeyl9G/gyF8KZ3jEepYjq5SlrUrWbIVpWvzUbomHymbZV3g9gWI6alaQC8gG2BHCWADgFinLOjwWu4Tb+8vAxn5hDJEv0z8PXZFk8QJOIPsRGjGvPbaa3H99dfjkEMOQVFREX799VdDxmwRZ8+ejaioKHCZNZKcm266CcnJybj66qttEkP0SKxee+01ebG9X35PIp+NHJl09YwzzsBee+2Fl19+2azAMXfuXBmoV5lBuC7vI488AppcSUoXLVqEW265xRDT448/3ie3nbszZ87EiSeeCJLHBQsWmOszMjJw8sknexK98MILhvRdfvnl+P7773H33XejR48e2HfffT1pgtkoLi7G+++/DxLBNPHhUVEEFIHmg0CxDCy3vPSSKdALY05FbngcDhsgkbMfPo21y5dUK2i49HPGVCskj1G2hujJNte1zXJvlgjb1VghJtvV+b/Kah4rsWbVGlS4/JttQ8XHr0t8Fy+NHrV6/ONyaC1OKsrgXv8r8ub/gsxFWUL4opBZOgD5lRzMVw3obZ2SkiqF8HVAxsBOhvAlpcZ4iK1boqvLNxQaolcmhK9sbQEqt5XaS80v7Siu4h2zTXud0Z1dRUAJ4K4iWJfrSf5elFGnjJRqEz70w5nok9pS1nK+697AeZ8GRQI3bdpkCNehhx5qNGzMuX///l43oDl32rRpRjtI7doVV1yB6dOnG62aJXvdu3fHpEmT0LNnT2MGpcauNiFZ5LJrDz74ICJk8XAKr7dC0yo1kIcddpg51FXm61q+fDneeust1EQAaZqdPHmy6XB69epl1vCdMWOGFwEcMWIELrzwQpMv70niyTTBEkCWnZpLEkBe/5J8ZFqz2du2if4qAi0FgcrCImTeNNlMjrxuzMF4L7oX2sdF4njXH1go5C8sKhpDD/kbUjK6mCAMEr7SGLcheatEk7dQ/lbm/ShTw6xCVlFWwGpHh0WbKVScJlsGYXCqlaiwqIDXNfsTrkq4NsxD7txfkblYCF92tBC+vhKwsacUnX9W3OiQUoH0/qnIGNxFCF8S4pJ21rsyrxQlC3KF6FWRvbINohmtkO+iU0TpGdEpDpE0BXeJl4mp1yNjgC4Z54SoobaVADYUkkHn03xV+gMGDMDYsWONT9u4cePAv8MPPxxJSTtHqCSENA1bYdADiU9WVpaHNNIvrq5Cjd6oUaM85M95PfNfu3YtpkyZYrR+9hy1gzTZ1iRDhw71jDaZbtiwYYag0bfE+ujxmFO4Ty1ksHLMMccYskgCTW0izcBvvPGG0ZQGm4emUwQUgcZDYOO995ol3twyncrVqQeaG90w2IWFb/3HbMccOAyrhofi6/wfsXKt+Oj9uQr5ZT6T5jmKxzVseyT28DbdijYvPS4d1Pa1eBFlRcWGhdgoFqDMxTlC+KKQVdpbAjboUsO/KgkNqUSnjuVIH9ARGUN6GF++qJgqWkHtXllmIQr+3FxF+NaIdk8IoK+ExoYjslui/CVU/XVNQKjMrUhhP105L9P3Et1vIASUADYQkEFlQ1MstXFBmID54M+XaQr23HNPD1EJ6h6+iepgAiYhovaKGrAffvjBmHkfeughvP3226DGLVhxEkReYzWD9OuzYk27dj9allwKJCSAlDvvvBMkdE6xeTuPNfU2SSj/aDZm+WjG/vzzz41/ZFOXRe+nCCgC3ggUfPMNtokLiIwE8fDo01AUHo0T+sVh80fPmIT5QxIxI/QtYK73dTIxjFnezM6f5/xtF93K/NGkby7NXI6sX35DlgSmUMOXs727BGyw39/Z90eGlYl7S5lo5NKQPqwvOkq0rg3YqBByV7ZsK7aJGddo+IT8+dXupYl2z5C9KtIX3mGnSdi7BXSvsRFQAtjYCPvmTxIYGed7tPq+EECXRIGZtLviA1g95xqP0Ol45MiR5u8yWSPzwAMPxBdffIFzzz3XXLdkyRJsl8lTLWGbN2+e8cOjqTWQtG9fpb6nhswKNX5OoWaR/nrl5eXVtIAdZDmmjh07Yt26daC2rS5CEu2UP/74AzRRW+0fz/GYU7i/K9O4kOgyQEZFEVAEdi8CFVu3Ikt8hSlLx0/AZ+Gd0Sk+HMNX/BcbCwsQkdEes7r8Dq6IcVC3g9C7XW+Pn173xO6IFrLYWqVonUzH8stcIXybkJkTjdxSiVZGxx1/VbWOjShCenq5+O8J4Rs+CCndkkzAhtHuie9eyU/ZO8y5+aLdq97nebR7Ys41Wj4x6VrtXmvFtSXVSwlgS2qtRi4riQ+DPOj7lpKSYogRo3npO2eFxIam2EsuucQEgdA3j8EbNWniSBZpVn322WdNgERubi4efvhhm6X5ZTQvA0uuueYa449HjRrJJTWgvD99De+66y6jaRs/frwhWAzqYDSxJadeGe7YyZQ1NBmFzGjdhTL/F6OXb+A0EA6hxvM5mRiWgS+c04+R0M88U6UdcCSrtklC+r///c/gRZKbnZ1t6sj67r///tXS6wFFQBFoOgQ4EMu+/Q5UbtqMyq49cH27fczNL09ajpw/lyBc3tN3+y+GSyy2p6WdguvHX+81MGy6kjb+nYhF3ppMIXy/I2tprhC+WOSXp8iN6d6z08UnKWor0jMqkDEoHemj9pQJqZOMC02FBGZQq5f/0UqUrRMNn5A/WV/Ou+Ci24igdq97lWYvSsy6YSnRXi443hfo3u5GQAng7m6BZnR/BmFwChb6vxUWFoLRsgy8cJIZ+ghSg0bCZqeBYfRsbcLAERLHE044wQRKMNL4vPPO81zGSGLe97777jPTu5BQMtqX2kjKSSedZLSO9LFj0ElsbKyZYubss8/25OFv47jjjjMaS15Prd9ZZ51lyKAzLQkkySSnvyEGrDNJZm3CaGhGSbPcJKIkzfRjpP8ft1UUAUVg9yGQ/9HHKJDBnMz/hHuGn4IyWfLszM7FyPm+KrJu/qgyWWmjBOM7j8ch7Q7ZfQVthDu7XG7krshG5q9/ImuZEL6NsSipoL80J122Ey+70CFmI9I7Vwrh64z00SNkPeF2cJdX+e6VLcvHli8zhfiJ755MzeIroXHWd2+H/14X+u7p3Ke+ODXnfSWAzbl1mrhsNHuSYNUm1Mbxz59Qi0f/RWrvnMK833zzTech0JzsFAah1HT/o48+2gSoOK+pbZtz/5F43n777QGTkvRxipm6Cs3S1ByqKAKKQPNCoFymlcq+4w5TqD8OOAE/hKaid2wF0v6Yie1ytGxoR/ySOMcskXb72NuxdvHa5lWBOpamorwSG5eJKXfuQiF8W5C1KQ7lLkbf8hPfyeQWinJ0it0ghM8lEbpdkTZmNCLbJZvAjDIJ0Cj7KRc5a1ejnL57vto90ZJGpMcjUgI0qOGLEh++sPaq3atjMzW75EoAm12TaIEUAUVAEVAE6osAzZ1ZU26GS7Ty5X0GYErsSIS4XTil6FsUFOQjOiMV/06bIyEeIZg2fhoY0btW/rUkKZV58ajZy5q3VH63Iic3Bi63/ZxX+VxHhhQhLU6WUessEy8L4es4em+ExR9izLfU6hV+mC2TLS+Dq8CPdi8+ospnT4geyV4EtXuRqt1rSc9IMGW1T0wwaTWNItCmEODk01OnTq1WZ35g6PPHSF8VRUARaF4IbJO5QYtkQneZhwm3DT4RlbJ27oWxS1Dw11Lj9zdr0DJUCpc5f8h52Dt9b2OxaF41qF6aIvHBy6Rmb/4KZC7PQ+5WBqeI052RqqmwYkO3ID12NTK6hCB9SDe0HzEGbhwqPnsy555o+HJfy0R51rLA2j1D9qrMuard2wFtK/9RAtjKG7ghq8dl2VqS0Bxdm3D5t0DC1Ux8p51hWpq4ly5dGugyPa4IKAK7CYEyWcIy597p5u6zDzoFc0PaYXT4RkQt/NocWzEmHNmRedijwx64bPhlu6mUNd/WBGxsLBHCJxHMCyRSd4UEXxREOi6qmoc1KSwT6TErkdE1VAhfdyQO2hvl20XDt1ZW1pCAjZynMkW7t8ZxXdVm6A7tXhQjc7sminZPInNVu1cNp7ZwQAlgW2hlrWO9EKBvIP98hQRwq0wvoaIIKALNBwG3vJeZN94Ed0kJtg8ZjjsjhyC+sgjjN34mExi7ETa8G76O+w5xEXG4d797ESFBIc1BTMDG+sIqwrdQInVXFqCkxGluJfmTgI3w1UL4lhvClza4B6K7jpWVQqnhE8K3KB+ZX+ZIsmzvKskauxEZcWLG3RGowcjc5CiNzPVGqc3uKQFss02vFVcEFAFFoPUgkPviiyj5/XeExMXhxj7HyrJvwNllP6C8qABxndPwTKcfTGVv2fsWdE3YOblxUyNgAjZWC2FbloesxdnIWlUo859KlIVHwmACNiKWIT16mRC+EHQa2AshSXuj1BC+IuTPzse2z7fJFfzbKaEJVb57hvCJhi8iQ7V7O9HRLV8ElAD6IqL7ioAioAgoAi0Kge0yo8CmRx8zZf784NOxBPE4onweQrNXIELm+/tkj7WoCHXhmN7H4KheRzVp3UzAxgohe+K7x0mXc9YWweWy/nssSihMwEbEYvNtJWwAAEAASURBVKRHLZUoXSCl92C4okehtFA0fOuLsfnrItHulUpaR7CKU7u3Y6LlsHaq3WvSxm3hN1MC2MIbUIuvCCgCikBbRsAlk9Nn/lMmd5dVhApH7YMHQvqi+/YN6JP9k4Ele98krAxdgm4J3TB5zORGh8oEbCzfZghfpky6nJvJpSydhC8EJmAjchEy5C8tPQwJnUeiPGw0ygoOryJ8P5TLNaLCxEZPeUMTIsWUWzUNi1lKrXM8QiKcpmJPUt1QBIJCQAlgUDBpIkVAEVAEFIHmiMDmxx5HqWgAQ2Qy+Wu7HYnYshIcs02CPsTvL25kH8yI+hLhoeGYvv904//XkHXwBGyQ8C3bZvz48nOpqXNKCEzABglfxEJ0ShWfvA5jUBayD8ryjkBZTglyxX2vSvKqfsJCECnmW8+auaLhC0tS7Z5FSX8bBgElgA2Do+aiCCgCioAi0MQIFM/9Hbk7Jq+fecCZWFcZiUn5svpHSQESO6fj+U7fmxJdNeIqDE4ZvMul2xmwQQ2fED75Kymgts4pOwI2IheKhm85OiS1R2g7IXwVB6Js61Eozq2UP5u+xGyEJUZ6llAza+YK+QuJcPoF2vT6qwg0HAJKABsOS81JEVAEFAFFoIkQcBUVSdTvjeIb58LWcYfgGXc37J3/G9rnrUF4VDS+Gb4ZxeI3t0/GPjhz0Jn1KpVXwAa1fCvzZKqVSq+8PAEbJHwROWgXnwJXrBC+sqNQnheJQir1dij2ZBIpgNo9Md8aokeTrkTmhovvnooi0NQIKAFsasT1fq0SAZqCLrjgAnz33XdmTeFDDjmkVdZTK6UINBcEcu6/H+Vr1yKkUxquSj0InYs3YPTWOaZ4xQd0wXzXl2aVj7vH3Y3QkOC0aV4BG0L4ctbkw1VBX7ydYgM2MiKXIS2iCHFRHVERNQplJSfDVRaGAllJDfzbIWFJot0z07DIVCwM1qB2Lzy48tg89FcRaAwElAA2BqqapweBMnHQjox0TmLqOdWqNl5++WWdW6tVtahWpjkjUPjd99j2xpumiK/udxYKtlfgrC0yqbsMxFJGD8YDYR+bcyR/HWI61FiVgtztWP1jCZZ9/KsEbEi0rTff8wRsdInYIHPxuRAZni4BG8NEEygrbbhDUMjFhflHcWr3dkTmhovvnooi0BwRUALYHFtlN5bp008/NRqsNTKjfkxMDAYOHIgnn3wSsbGxuFHMLfmyvuagQYPw6quviomjDBMmTMDNN9/sIXlnnnkm+vTpgy1btuDSSy9Fv379cM899+Dggw/GrFmzTH6sHvMZPXo0XnnlFYwZM8bUeNmyZbhfRvVz5syRjtVt0nL1kW7dupnz77zzDl6Uub7Wr1+Pzp07g/c6/fTTzTl///F83759zakPPvgA4eHhOO2003DllVd6yBpX+5g4cSJWrFgBrgqSkJCAiy++uMZ8fe+1aNEiU6733nsP48aN8z2t+4qAItCACFTmyXQqU6aYHHMOOQavlqXi+M0fIbKsEEkZGXgpTeb7E0vrWYPOwrjONb+Pxfll+ODh31G4hevhVq2Jy4CNjMil6BKxDe2kzwgJ6YZyjIGrMgZUBlbQ5W+H2x8DM4xWz060rNq9BmxpzaqxEWjRBPC1117DC+IAvGnTJgwYMAC33HIL9txzT7+YffbZZ3j66aexVkwGFRUV6N69O84991wcd9xxnvQkHY8++ihINEhQRowYgdtuuw09evTwpNnVDd6jpKLK8bemvFzi11Iq/itMG+qqv7kgJjzGQ3Zquh/Pbdy4Eddeey2uv/560IRZJD42v/76qyFj9trZs2fLEptR4DJrGzZswE033YRkib67+uqrbRJD9Eis2D6hocGVPScnB2eccQb22msvUJvGFTjmzp1r2ooZc13eRx55BLfeeqshhiRdbG8S0+OPP95zb9+NmTNn4sQTTzRtumDBAnN9hnwkTj75ZE9SPkMkfZdffjm+lzVE7777btPm++67rydNoI0SWXWAmLFcqampgZLpcUVAEWggBLLvvAsV0leFdO2GKxP2xaitc9GleB3CxdLw25jt2FKah4HtB+LKEVfWeMeKskp8/OQfhvylyXJxI+LnIi40Vvq7XrJyyMFybSjKXI4swum7R5+9Kr89TslCAqiiCLRUBFosAfz444+NZun2228367WSNJx//vmgBislJaVaeyQlJeGSSy5Br169EBERga+//hqTJ082acePH2/SP/fcc4bYUOvUpUsXQziYJ+9F0rOrQvJ31idnYd6mecFntTD4pP5SDu84HC//LTjzJIk0yfGhhx5qNGzMr3///l7Z0pw7bdo0ox2kdu2KK67A9OnTjVbNkj2S60mTJqFnz54ICwszGjuvTPzskCyS9D344IOmfZiE11t57LHHjAbysMMOM4e6du2K5cuX4y1Z+L0mApienm7aOSQkxLQ91/CdMWOGFwEk0b/wwgtNvrwniSfTBEMAqd0cPny4Icy2rPqrCCgCjYNAvvTv+R9+KKbWMDy1z5mIy8vB3jv8/sIPG4zvS2eBg97p+01HZFhg1xO3y40vX16EnNUFGBZTie5RnaXAnb0IHydVjuwufntdExAlvxHpceq71zjNqrnuJgRaLAF86aWXzEec5jsKieA333wDmuHsx9yJqTUz2mNnn3220VT99ttvIAEkOaM5kiTROvCT2Oyzzz744osvcNRRDTN7PIlIcxVqUceOHYujjz7amDJpzjz88MNB8myFhJCmYSskP8XFxcjKyvKQxsGD6z7dAjV6o0aN8pA/mz9/mT81t1PE7EOtnxWSVZpsa5KhQ4d6aUCHDRsGPjtcz5fklMJjTuE+BxS1CU3GP/30E6hlVFEEFIHGRaBctH7Zt91ubrL2byfif8XRmLTpvwgRp720vYbhPvd/zbkpY6agR1KPGgvzy4ersPy3jegd5RbyFy05uA3ZI9GrWjdX1syVqVlUFIHWjECLJID0Pfvrr79w0UUXedqG2ieStd9lLcjahGSPH+5Vq1bhuuuuM8npV0YNGPOwQnJBAsE8G4IAkvxRGxesCXj+/PnGpG01a7ZcdfmtiwmYhIjkiBqwH374wWhDH3roIbz99tugxi1YcRJEXmPLT9ytkLw5JVqWawokJICUO++807SHM53N23msqbZ//vlnQ0zpy+gUmpJJZmkmV1EEFIFdR4B9R/Ytt6Jym6x927c/ro4cgcOyPkFcRRHaiUvHa53noLK0Ekf0PMIs91bTHZf8nI1fP16NjmLSHRxdNQgsHBqJfifv4RkU1nS9nlMEWgsCLZIAbt261WhwfE293F+5cmXAtikoKMB+++1nghdIHKZOneox85H8UfzluXnz5oB58gS1SfwLVqJCazcnV7orwXSRIZEIC63qpILN35mOvoR1FWrA+Ee/OGpD6T95zjnnGC3p4sWLjW+gJWwki/TD69ixo8GAHbUlehYTq0HMzs72mJRJ4CksH9PRnMxAje3bt1fTAtLHkPlTC+iPiNv7+NaT5fjjjz+82oZk3gaV8DqmmTdvnlca7tNVIFC+9vh5551n/Aud9z322GNxww034MADDwx4vTP97t62dbG/u7s8dbm/LbP9rcu1LSGtrZf9bQllbqwy5r37Lgq//RYh4oJy/8jTMCB7PrqXrENYRCSWjYvEusJMdI7vjMmjJ5s+JVA5uB7vV68sRIK4Jo+Jc4t1IBTRI1OR07OwRbyvgerVWo/bZ9/+NmQ9GyPPhixfU+TVIglgfYGJi4szZl9qlBjMQF8/arZ8zcN1zZ9+ZY0lf/75Z2NlXS1f+tQxUIKBNImJicbHLje3asp6kiJG9paWluIf//iHCZ4haX722WcNSaS2klJYWAhLmJ1lZ2Qwgzh4ngE2r7/+uknPe9KvkPek2fXvf/87SKKoReS53r17g0EbPPbMM88gTyIAqZUtl3U/SfYZqOKPFNqyMFCFQRqMQl69erXRyjHYhPWhUJvMQBdqF6m1Y5npR8pAGJvGJPTzH83e/oQklhhYHPylaW7HnG3V3MpWW3lactlrqxvPt/b61YZBiJh+o+/5l5h6gV/GH42/thXghK0/m8tce/fGR4WfIkz+ndfxPKxYuCJgdtvzK7FgVgHCXSEYl1AhcwPGoKxDCHJ6yKR9Yp1p6zgHBK4ZnNC2aZxGaJEEkBohmistObHQcL9Dh8BzPlHrxwAFCqc34dQfJDAkgDaCk3lQ22SF+/SNq0k41Qm1YA0pHJ3wod9jj6YzS9DkTW0ffR5J1Ei8qM2yU620b9/emMjpB8hAEJInki/65tm5/hjIYdvAWXaakum/x78ePXqY6GGSPRJD64NHkymngWEULtuKuDNKmySdaagl5DQwb7zxhiGIxP2ss87yXO+LP8vCABFqGRnNzTzp++mcBoblPvXUU8EpaDidDQcJrDOnkAkktbUNA0lsnQLl0VyO11aX5lJOf+VoyWX3Vx/fY629fr719bfvln5w/bnnoUQGVa49huK+uGGYuP4dic91o8vokbg/6WMxIwCXDrsUE4dU+YP7y4cTPM+87ze4SkMwPqEckRLtG5Ycia4XDUWXqNAm72v9lVGPVUegMd8BKoIaU3lTvTbN70iLJID8aDPQgFo8G7DBjzz3qd0JVngNSQyFUb8kgcyD5JBCEkQTIueOq0lIRm1AQU3p6nOuMfP2LQ8JFadECST0YeTfVVddZf78peP8gHxpqT1zlp15M2LXKUtkAXencH5BErxAQi0g/4IVlpUR3ySod9xxh9/LmIbEl9P/1FWc9bPX+tbJHm/uv/7q0tzLbMvXkstu61DTb2uvX011z335FZRIoF6IDLCnDTkR+6/6GgmV4veXnoH3e/2F7QXbMSZtDM7f4/yArjKVlS589txf2LaxFKNjy5EYFosQIX0dzt0DEQnRHtNvW8a5pjZoDucao22YZ1uXFkkA2Wicw4+amiFDhnjMh5yT7YQTTjBt+s9//hOdOnUy5j8eoPmQaen/RdL3rfiTcG45aoYoJALUJj311FNGS2ingaE20JJMk1D/UwQUAUVAEWh0BLaLa80msRxQ5h19Doqzl2BEyRqEhkcg99A0LNw8G+2i2mHa+GkByR99fP/vjSXYsDQP/aIqZYJnsdSILTnl9EGI6NiwVptGB0RvoAg0MAItlgAeeeSRxieNmhv6olFr9/zzz3vMj/TPosnPCtW9nCqGgQgMYKCT/3333QfmY4VruZJEclJf+qmNHDnS5NkQcwDae+hvy0GAAwQGCvkKPyo0h3/++ee+p3RfEVAEGgABtwzSM2XlIbf4+laO2QeP5MmE71uq3reuxx2I2zc/Z+5y5753omPsTpcd31vP+2IdFn6fhbRwNwbEVAXftZvQC9H9kn2T6r4i0OYQaLEEkC1Fc28gk6/vFBxcqcK5WoW/lqYWkP5h/FOpjgCDZlqS+D4D/srOufwCCVczYcCJr9DE3dZ9R3wx0X1FoCER2CyrNpUuXIRQmYP0lu6H4bDlMyXMw4UeslLQo5XvmVtNGjAJB3Q9IOBtV87bhB/fW4bE0BDsFUfFXwjixqQhbp+MgNfoCUWgLSHQoglgW2oorWvTI8AgEv75CgkgpyJSUQQUgYZHoET8rjc/86zJ+LsJ56HTyu+QWFmIxI5p+Lz/WmzJ3YJ+yf1wzahrAt5809oCfP7Cn4iSQf04CfoICYlFVO8ktDumt3H3CXihnlAE2hACO22kbajSWlVFQBFQBBSB5oeAS1xwMm+4kZOrouyAQ/HfzM3oVbwaIWGiqzhmMH7I/RnRYdG4b7/7EBXmfz7Vwq2l+OjxuXCVA+PiZV5RIX/hKdHi9zdQ8tFPXvNrdS3R7kJA34bdhbzeVxFQBBQBRcALgY0PPIgyma8zLLUj7mo3FPtu+cmcHzDxKDyZ9YrZvmGvG9CrXS+v6+xO2fYKfPTEXBTJnH97xW1HfFgCQmS1j5RzBiM0NsIm019FQBEQBJQA6mOgCCgCioAisNsRKJIpuLbKNFKUTw8/CwPXfWH8/rqN3AuPYyYq3BU4tPuhmNjX/3x/LpdbzL4LsHl9CQZFl6FThKwTbiJ+ByIiVSN+d3sDawGaHQJKAJtdk2iBFAFFQBFoWwhUyqwLmTdNNpUu+tsxWLxiHpIqChDVPhXzRhTJUm/rkB6Xjqljpwb04Zv9/nKs/nMLOkfI0pLREvUh0u7Y3ojuqxG/Bgz9TxHwQUAJoA8guqsIKAKKgCLQtAjk3D0NFTJFV1iXrnjMnYzexavgljXQO506Hv/Z8LEs2xaKe/e7F0lRSX4L9td3G8ApX5Jkbt+RsVWftbix6YjfWyN+/QKmBxUBQUAJoD4GXghwCTQuxVaTcHqUGTNm1JSk1nM3yhxfl156aa3pNIEioAi0bgTyZT7NvA8+kK9RKN4fdzwGZ/9oKjxk4nG4f/0zZvvioRdjeMfhfoFYt2iLTPa8WCJ+gfHxZaIhjERU33ZoN6G33/R6UBFQBKoQ0Glg9EmoMwLvvvuuWYu3zhe2gAu4YgzXQ165cqWZMHz48OG47rrrzMThLaD4WkRFoEUhUCFrrWdPvc2UOffoE1G87DskyXx/KYNH4OWo/6G4qBgjO43EhXtc6LdeW7KK8OnT82Q94BDsn1iMsJAkhKfGIGUSI36FEaooAopAQARUAxgQGj0RCAGughETExPoNMpl9v6WKr/88gtOP/10vP3223jppZdQUVGB888/H1xJRkURUAQaDgGuqJN161RUbtmCsL598W7uNvH7y4c7Phm5B8dhwZa/kBiZiH+N/5ffpd5KCsvw0WNzUVYK7BNfhJjQJITGhKPD2RLxK78qioAiUDMCSgBrxqdNnuVEx3fccYdZCm/MmDF4+OGHwc7aiq8JuH///nj99ddx8cUXY9iwYWbdZaZdtmwZLrroIowYMQLUpE2aNAlr16612ZjfF154AePGjQPvw6X6aiKPjz32GI499li8+eab2H///c0qHVy1paCgwJOnNS0//vjj2Hvvvc29ubQf138ORlgerifdVz5IAwYMAFc/yczMxF9//RXM5ZpGEVAEgkQgb+YsFH75JRARjjcH7o2uBStEkReKgecdjRnLq6KBb9/ndqTFpVXLsbLchU+emIf8LeXYI6YYKeHtjENT+zMGIrxD4MFptYz0gCLQhhHQYVITNz6JlFsmO61NXELCsH07XKJ5CgkTz+Z6Soho6rjEXV1k5syZOPHEE/HOO+9gwYIFZm3kjIwMnHzyyQGzIeG69tprMWXKFJNm8eLFuOSSSwyxe/nll82KGnPnzjUaNZvJzz//jNTUVPA8iSGX6uOazjXdh+k++eQTPC1LRRUWFpr73XbbbXjggQdstpgt00lw/WYuBbdhwwbcdNNNSE5OrnUpQE8Gjg1LLpNkSSoVRUARaBgEyuW9zNnha7z6qBMQvapqvr+eR4rf34anzE1O7ncyDul+SLUbsg/96pWFyFpViG4RZei1IzAk+bi+iO4tRFBFEVAEgkJACWBQMDVMInZcayadjpLffw8qQ85ctTyolIETxYj2rftrr9aJBKanp2Py5Mnmml69epl1bxn0URMxmzBhAiZOrJqfixrERx99FAkJCXjwwQcREVE1AWvPnj29CkpSRe1cmBDc3r17G60eyVtN9yktLcX06dPRqVMnk9fNN99stIzU/JFMUiIjIzFt2jRjpqYm74orrjDXUFsYKo7mwYrL5TL5UIPZr1+/YC/TdIqAIlADAm55rzInT4GrqAjuPffET+uXI1H8/kK6D8aX6b9gs6z+0TupN64bfZ3fXH79eDWWztmIdmEuDIuLNGnix3VG3F7VNYV+M9CDioAiYBAI/muogDUMAnXUxjXMTeuWy9ChQ70II826a9askdWZRCsZQIYMGeJ1hulHjhzpIX9eJ3fs9OnTx5A/e44ELlecwmsSklNL/piOpmUStVWrVnkuo0na6aPINPThy8rK8qQJZoMmaZqxH3rooWCSaxpFQBEIAgFO9lws2n95SfFehzQkludhe1Qi2p3UB99n/oDI0EhM3386YsKrm3KXzcnBL/9dhWgT8Str/CIC0f2TkXSk9+AyiGJoEkWgzSOgGsAmfARoiqU2LhgTMMnW/PnzsaeMkKkhq6/UxwRcn3vFxnrPtE8tXG0SHu79+BEfp69hbdc35nn6QH7zzTd4VT5WaWmqWWhMrDXvtoNAqUTXc7k3yh8HHYrY9X+hUmYjG3j2Sbhr8Z3m+PWjr0e/5Ooa9+yVefhyxgJZHSQEByQUyNyA7RHeMRbtTxuAkNC6ubmYG+l/ikAbR8D7C9zGwWiK6pPkhPiQJX/3dVPbFh0t61fGInQXCKC/vGs7RuLplD/++APdu3evExHt1q0b6OPHoA5rAnbmWd9tavFycnI8WsB58+YZs67TvLxkyRJxn9xupnHhfZiGBJXaw9qEBPTOO+/E5zI3GX0Iu3btWtslel4RUASCQMAtfUHmP2+AW9w4ikaPxpoNi8EPUOTeh+Ol/BdR7irHgV0PxCn9T6mWW/7mEnz8+FyxQsh0Lwl5iArtgNA4RvwOQmi0fsaqAaYHFIEgEFATcBAgtbUkjHq95557zFx4H374odGCnXXWWXWC4bDDDjNBGtdccw3+/PNPrJYF3mfNmmXyrFNGPokZ3EF/PwaZ/Prrr7jrrrtwxBFHePz/mJwRvwxGWb58Ob799lsweviMM84Iyv+PZt///Oc/JqgkLi4OmzZtMn8klCqKgCJQfwQ2P/sstktQmUt8fz8PdSHcXYmt7Xsjd1Q2VuevRsfYjrhjnzu83E94t9KSCpnu5TeUFLsxPDZPfP86QCb8Q8oZgxCeUt1MXP8S6pWKQNtCQIdObau9g6rtcccdZzRoJ510ktH6kfydckr1UXlNmTEAhPPoMTqXq4sw+IIRvvQL3BWhZvHQQw/FBRdcgLy8PBxwwAGYOnWqV5Zjx441GkvO50cyyACVyy+/3CtNoJ033njDnGKZnUJCzOlhVBQBRaDuCJT8uQCbn3oanEzqh5EjEJm7FoXh8eh79ljcv+RuMeqGmPn+2kV7R/G6Kl343zPzsCVHon0ji9AtUsifSPLxfRHVUyPzDRj6nyJQTwSUANYTuNZ6Gc2eVqgN8ydfffWV12GaXP0JgzE4r54/4fx6vmKnkPE97rvP+QT5V5Mw8pd/dZVAdalrPppeEVAEqhBwifY8U7T2MgcUNuwzBgVC/iTmF0nHTMTTK6v8AS/Y8wKMThvtBRndMb57cynWLc5Hh/Ay7BEbZ87H79cFcaOqZgHwukB3FAFFoE4IKAGsE1yaWBFQBBQBRaAuCGx6+BGUrViBorSOmFeyzSxAn9lnP2yMl4mgNxdiaOpQXDL0kmpZzv9qPRZ8lykRvy5Z6YO6Q4n4HdgeSX/rUS2tHlAEFIG6I6AEsO6Y6RUtFAH6Nh511FEBS//RRx+BE16rKAKKQMMgUPTzL9giE71XSJTuDz26ILQkD+vjuqHDUaH4Ytl8xEfE49797kV4qPenaPWfm/HDu0tNxO/BiYWiL0xBRJpE/J7aXyN+G6ZpNBdFwARhKQyKQItAgH58tfny+TMt28p17NjRBKLYfd9fnldRBBSBhkGgUlbqyZJVeGRuJyzceyQqirehMCwO6ScfhFeX3WFuMnXsVHSO7+x1w83rC/DZs3/IZZzuJRfhIWkIjY9ACtf4jfImil4X6o4ioAjUCQF9m+oElyZuyQhw3kFOZ6OiCCgCjY9AjgROlYvWPbN3N6wX8ke/v+xRR2Fp/lMSDOLG8X2Ox996/s2rIEV5pfjo0d9k+qgQ7BW3GfFhMnUTI37PlIjf5GivtLqjCCgCu4aATgOza/jp1YqAIqAIKAI+CBR89TXy3nsfBdGRmJdYNVXLvNQxcA3+GRtLNqJHYg/cuJcEhjikvKwSH8t0L4X5LvSL2ob0iKp5O5MnSsRv90RHSt1UBBSBhkBACWBDoKh5KAKKgCKgCBgEKrZsQdYtt6BSJr3/fQ9Z0cNVgbXRXdD9uA74Mfv/EBEagen7TUdsxM7Vg9wuN758YT42rt+OtPBiDIxJNnklHNAVcSM04lcfLUWgMRBQAtgYqGqeioAioAi0QQQ4dUv21NtQKWt6LxrYC4VlJSgKi0XRfvvjf1nPGUSuHnk1BqYM9ELnp1krsOKPrYgLqcCYeHomhSN6cAoSD1OXDS+gdEcRaEAElAA2IJialSKgCCgCbRmB/P/+FwWyjOKGlCSsjRDln/j9ze5+CHLav4kyVxnGdx6PMwae4QXRoh8zMfeztSYi8aCkQjkXh4j0OLQ/RSN+vYDSHUWggRFQAtjAgGp2ioAioAi0RQTKs7ORfeddKIyKwIJuVWbbOe1Gouv4VbLU20p0iOmAO/e902uptw1LtuKbVxcZuA5KzJE5AlMRmrAj4jcyrC3CqHVWBJoMASWATQa13kgRUAQUgdaJgNvlQtbkySiXqV/m9e+OSlcl1kV3RsT+GZi96UPRA4Zg2rhpSIlJ8QCwLacYnzz1O1yuEOwbn4mY0C5i+Q1Fh7MGI7xdlCedbigCikDjIKAEsHFwbbG5cg3cu+++u8byH3TQQZgxY0aNaWo7eaMsDXXppZfWlkzPKwKKQAtAYKusoV3042wsFM1ffohb/P5iMLfvvliJGab05ww5B2Mzxnpqsr2wHB8+Ogel24EhMTmy1FuVr1/7k/oismuCJ51uKAKKQOMhoPMANh62rTbnd999FzExVVM7tLZKvv7663hDPmYbNmwwVevbt68hqvvvv39rq6rWRxFoEARKV63CxvvuR2a7eKxrFycz/AGfpR6EDoM+xYqCAgxJGYLLh13uuVdlhQufPPkb8nIr0SViG3pHVU33knBwN8QO1cnYPUDphiLQyAgoAWxkgFtj9u3bt6+xWuXl5TWeb84n09LScN1115kJoxnROGvWLFx22WWYOXMmSAZVFAFFYCcC7ooKZN14EwrE5Lugu5hwhf7R7y9t7EYsKfgTcRFxZsqXiLAIcxHfqW/+/RcyVxYjMbQEI+Pj5JIwxOzRAYlCAFUUAUWg6RBQE3DTYd1i7lRZWYk77rgDI0eOxJgxY/Dwww/Lskwc11eJrwm4f//+oObs4osvxrBhw/DMM8+YhMuWLcNFF12EESNGYPjw4Zg0aRLWrl1rszG/L7zwAsaNG2fuc/vtt8sKAIHJ42OPPYZjjz0Wb775JqiRGzp0KK688koUiJbBijUtP/7449h7773NvW+99VaUlZXZJDX+sm7Mu0ePHujZsyeuvvpqxMbGYt68eTVepycVgbaIQO7zL6Bw/nzM65WOCiF/66MzsLZ/Zywre9/AMWXMFHRN7OqBZu6nq7H4502IQCUOSJJ30h2LiM7xSD6pn67x60FJNxSBpkFANYBNg7PnLiRSFWUuz36gDZKwynI3yksr4dqFYLjwyFCvqLtA93Mep7brxBNPxDvvvIMFCxaABCojIwMnn3yyM5nXNgnXtddeiylTppjjixcvxiWXXGKI3cuyGHx8fDzmzp2LCtEYWPn555+RmpoKnicxJNkaOHBgjfdhuk8++QRPP/00CsXhnPe77bbb8MADD9hsMXv2bERFReHf//63MeXeJOuRJicnm/w9iYLYYBt8+umnKC4uNgQ2iEs0iSLQZhDYvnAhNsl7vygjBfkS+VscGo3P0/ZFSsYrKCx14eheR+Po3kd78Fj+20b89MEqs39oUg5C3N0RmhgpQR+DEKoRvx6cdEMRaCoElAA2FdJyH5K/9++bi+yVeUHfdQ6+Dzqtv4TpvZNw/HUj6kQC09PTMVki+kJkJv9evXph6dKlJuijJgI4YcIETJw40RSBxOnRRx9FQkICHnzwQUREVJl/qFFzSlJSkiGXYWFh6N27t9G8kbzVdJ/S0lJMnz4dnTpVTTNx8803Gy0jNX8kk5TIyEhMmzbN+CnSbHvFFVeYa6gtDA2tXem9ZMkSnHrqqeC9qP174okn0KfP/7N3JvBRVWf//82eTPY9ISEJ+74LKooCogi4U7CtS7XW2tZarb7Vtm9rtRt1aV9ra2utdbdVW8V/VXCv1AURJOw7IYSQfV9mkln/55zhhgSTkJBMMnfmd/zc3O3cc5/zPcP4m3PO85zRHU3nMQlENAGf6FEvveuHKIuxoTg1QbF4J+08DJ/+IUraKpEbl4v/PSPwY1DerChqxLtPbBdHBsyPOwyLYTQMFunxOxGmBHr8KoD8QwKDTODk/zccZIPC/XVCU4V8kkOrUvxpSQ7rHj58GFLYdZcmT57c6ZbML4eQNfHX6eaxEymqpPjTkhRwNWIFgZ6SFKea+JP55NCyT4SgOCQmomtJDkl3dFKReWQvXllZmZalx70UqnLu30svvYSvfOUruOuuu3DgwIEen+FNEogkAtXiB17t4UPYfize36aEGXCPrUWJawPMBjPuO+c+Nf9PMmmqbRVr/G4S3x8GzLAXI8EU+DElh32tOXGRhI11JYGQIsAewEFsDimqZG9cb4eAt4m5NVOnTu0kkvpq7qkMAff1HTK/7CnrmGQv3MmS2dz54yf5dJxreLLng3Vf2p6Xl6eKl8J2+/bteOaZZ9S8yGC9k+WSgF4IOD7/HJVPPIGCUdnwGA0otWViY2YeEhIelUt/4Hszv4fJqYEfhK5WD974/UY4WoCRtgrkWgOjAPHn58E+NdBjr5d6004SCDcCnf8PHG61C8H6SJFjsR3v9erORKPobDNZAnk79pJ1l38gr0vh2TFt3bpVCaK+2JGbmws5x086dfTUC9jxPb05lr14FRUV7b2A0jlDDut2HF6WQ7itra2IiopSRco8UqDK3sNTSbKHsbdOJKdSPp8hAb0Q8LW0oFR4/e7JEvP+7Da0mqLwZsYCDBv3D9R52nBm1pn42qSvqer4fH68/ZcC1FS4kWxqxNSYFPh9RkRPS0PcwuOOIXqpO+0kgXAjwCHgcGvRAahPaWkpVq1ahcLCQrz++ut47rnncO211/ap5AsuuEA5adx+++2qB62oqEgNq8oy+5Okc4ec7yedTDZt2oRf/vKXWLJkSfv8P1m2FGvSOUQO265btw7Se/jqq6/u1fw/6UyyceNGlJSUQApJef7ZZ5/h4ouPT2bvj/18lgT0TKDi/gdQ3FiLw8fm/b2duhBxo9cL8XcYyVHJ+PW8X8NoCPxv5eOX9uDw7ibYDG2YJ6YJ+n3RsIggz8lfGtNpiomeedB2EtAzAfYA6rn1gmT7ZZddpnrQVqxYoYafpfi78sor+/Q26QDy5JNPKgElVxeRvXTSw1fOC+xPkj2L559/Pm688UY0NDRg/vz5+NnPftapyDPPPFP1WF511VVKDEoHlVtuuaVTnu5O5BxEOeevsrJSObHI+YQyVM1ZZ53V3SO8TgIRQaD5v/9F6eqXsX2sjPcHfJ4wHeWZDlij1qlzuc6vXO9Xpu0flGDbB2XC5QM4P1HM6/XlCWcP4fF7zUTh/HHyERBVCP+QAAkElYCuBeDzzz+v/udcVVWF8ePH46c//amaM9cVMTmhX07sl7HpZJo0aRJk75ScY6cl2bMkQ6B0TDJGnRQAkZJk6BQtybh8XaX333+/02XZU9ZV0sRTV/d+85vffOGyFkLmCzdOuCDjCcqtpyQ9f+XW1yS9h5lIgAQ6E/DU1eHo//4EBXkZ8AjHrcroLGxIG4/k4X9Cq5j3d83Ea3BOzjnqoeKdNfjwRfmdYMCi+AMw+Scoj98UscavSYR9YSIBEggNAroVgGvWrFHDlFKkSK9VGUvuhhtuUHHbUlJSvkBXzkdbtmyZCgwsJ/k//vjj+PrXv4433nijfT6ZfGjevHmqXK2A3jgzaHm5JwESIIFwJFDxi19gh8WHBnsUvOYovJG6AMkjXxbirwkTkifgtpm3qWrXHG3GW2Len99vxBkx+2A3TlLXk68cB6sI+MxEAiQQOgR0KwDl8KKMF6fFnpNC8IMPPsDLL7+Mb37zm18g3DFQsLwp54699dZbKmiwHPLUkhR8Wjw57Rr34UFAzm2UPwK6S/LHgAx4zUQCJHCcQIP4d7Hvo3UoGhFwolqTsgDerAK0mvch2hytQr5YTVY4Gl3C4/czMe3CiPG2w8iwThAT/4D4xfmInhwYGj5eKo9IgASGmoAuBaCc5L9z504VAFgDKOeYzZ07FwUFBdqlHvdOp1OtSiGDEXdMcsK/nEMWHx+vlhK77bbb1CoSHfPweGgIyHl8J5vL19XQsmZtenq6mgagnZ+4l/eZSIAEjhNwV1Si8Ne/wrbhgX8b25OmozjZiNjUt6W2w4/m/AgjEkaI0FZeEevvMzQ1ApnmSoyPGa48fu0z0hE3PzBn8HipPCIBEggFAroUgHViPooMSnziUK88762X6YMPPgj5P3wpGrUkh3+lg0FOTg6OHDmiVrGQzgYvvvhij7H4pC09BUnWyu/LXitP2/fl2aHOq9ms7YfaHu39MgSPbNueUm9s1vJo+57KC/V7Wh20fajb29E+zWZt3/FeOBxr9dL2g10nGZOzRKwItDnJDo/ZhOa4LPw3aRoS8/4Etwj4d2Hehbh4xMXwuD14929bUXHEhVhjM85IsMPvtcGSG4f4y0aqQO2DbXtf3qfx1fZ9eZZ5g0tAaxNtP5BvC0aZA2nfYJSlSwHYXzCPPfYY5BxCGdxXhhXRUsfhQenAILdFixapMCCyV7C7JJdKC1aSQYj1mvRse2+Yh1P99FwXPdseyp8z83vv4WDhHtSnJ4qgpFb8K2EBoof9P7iN1Ui1pOIS+yWQMUKPbGzB0a0emIQsXJDYIMRfPrx2AyqneVCyo3NM0d7Ud6jyhPvnaKi4DsR72TYDQfGLZehSACYlJakeuROXDZPnqak9zzWRHr1SAMo5hNJzuKc0fPhwNfwrlzXrSQCOHTv2Cyth9FRub+7JXyfyQz9lypQeex97U9Zg59Gz7b1hFU7103Nd9Gx7qH/OXMXF+Ozfq3EoO+BQ91b6eXCm7kd0/FaYDCb8duFvMS1tGvZtKMfRLYEoABckHYLRNwkGqxEZN0xBTmZMb6o55HnC/XM05ID7YUAw20YuDxrMzpt+VHvQHtWlAJSOGjKMy/r161UPnaQlV2uQ5zLgb3fpr3/9Kx599FEV1kUKq5Ol8vJy1NfXn9QpRK6Q0ZdVMk723o73g1l2x/cE41jPtveGRzjVT8910bPtofg584sfn4fE0O/WDNHzJ1Jp9izsj41BfNaTcqU33Dz9ZszMnInSA/X44Lnd4ooR58Rth9U/U0Z+QfKXxyMqO14+qqsU7p8jXTXGCcYGo21kmZGedCkAZaNdf/31KmCvXKtVxvKTYWCkY8cVV1yh2vTOO+9U4V3uuOMOdS57/R4WC5hLb+Ds7GzI2IEyySXCYmJi0CKWOPrjH/+IxYsXq15EOQfwgQceUAGF5dxAJhIgARKIBALVYpRkfVMV3DFRMCVlYbV1GmJzHoNPrOgxO3M2vj7562iocmDtHzeJH95GTIvejSTTDIUmYckIRE/8YhiuSODGOpKA3gjoVgAuXboUtbW1StRJMSdXmZCx/bQhYLlmrPQM1tILL7yg1qU9MTjwd7/7XeVZKn8NyO5gGSy6qalJOYjI1R9uvfVWMBagRpF7EiCBcCbQuncfPnrxOdSnxsNstuCZ+EWwpL8D2I4iwZaAVWevgtvpwxsPbRCrBRmRZzmMEfYJYt6fAfZZGYidlx3OeFg3EggrAroVgLIV5HBvd0O+HVe0kHlPXL1CXuuYoqKiImrFj45173gsl22TcyN7WpVj4cKFam3g6667ruOjfTqWq640NjbiT3/6U5+eY2YSIIHgEPCL8Fqbf/QDFArxJ1PBqGWotxyFPeUjdf6LuWKpt6g0vPbbT1FX40eiqQYzEzPgc1tgzY9H0uWjucavIsU/JKAPAse7yPRhL60MAQL/+te/+rw2cAiY3WcT5LQB6Qn+q1/9qs/P8gES0BuBot/9DpsMLmW2cfRsrPPGwp79T3X+5XFfxvzh8/HfZ7fjaGErrAYn5id7hPiLgyk5CilXi+XezPzfid7anPZGNgH+i43s9j+l2icnJyM6OrrbZ91ud7f39HJj27ZtkNMGpABkIoFwJ9D8+Sb856N34Rbx/uISUvAn33REDXtJhH9pxpikMbjjtDuw5e1D2PVpjfDz8GJx8hH43Tkw2ExI/dpEmGK5xm+4f0ZYv/AjQAEYfm3a7xpJ1/uf//znmDVrFk4//XQ89NBDYm1PGfc/kOQQ8FNPPaWdKpH097//Hd/61rcwffp0/OUvf1H39u/fr1ZrmTlzJmbMmIGvfvWrKBbhJTomGZbn7LPPVu+Ry/n1JB7/8Ic/4NJLL1XC7Nxzz1VrQMs5mnLOppbk0PJ3vvMd5dBzxhlnqLWf7777brE8VaBnQ8vX0146BP3gBz9QywWeuFJMT8/xHgnokYBPhMN4/+c/RZ1w+jAbjHg9/3Ig6ROYY/cjyhSFB855AKXbG/HJ6kOqeuclboXROyng8fvV8bBk6CPcix7bhjaTQDAJUAAGk24XZUsh5W5t7dXmFaKlt3m7y9dRuHVhTpeXVq9ercLa/POf/1RzAaXYk8c9JelBLVdRee2115QntnTQufbaa5UDjfTQfuWVV9S6zR6Pp72YDRs2KEEo78sl3OR75dZTkgJy7dq1KpyPdPrZvXs37rnnnk6PyHBABw8ehJwH+jsxrPXOO+/gkUce6ZSnpxMpfqXA7LhKTE/5eY8E9Exgyz0/xf6oQEgM9xlXYIerAlHpb6kq3TnnTsQ3puOdx7eKcwPmxGxGjGGOupewbCSixyXrueq0nQQimoCunUD01nJSjL1w950o3SdjZ/Uufdi7bN3mGjZuIr587319mpydlZWFH4s4YHLptJEjRyrvaCkCV65c2e17LrroIiXwZAbZgyhD7sTFxSkBZrFY1HMjRozo9LzsXZO9c9IDe9SoUUp0SfHW03va2tpw//33qxA/srCf/OQnqpdR9vylpaWp8qXX9q9//Ws1TD1mzBhIz2/5jOwt7OgZ3smYYydviIXvd+3aBTnPkYkEwp1AxZtv4uMDuwCLCcNGTcaPq+Jhz39YaD0vFuUuwoVpF+PlX34Ej9eEMbY9GBZ9GvweA2LmZCL2rGHhjof1I4GwJsAewMFuXiGqQj1Nmzatk2CUw7pyNRQp7LpLMh5jxyTzyyFkTfx1vKcdjx49ulMAbSngTlzdRcur7aU4zcjI0E7V0LIMAn7oUGB4St6Q8/Y6zlGUw88y6rsMDdRTkvelw4eM/9hxicCenuE9EtArAbfopV/7l9/DJcRfos2OxxLOgy39/8ForUFmTCb+d9ZPseahT0WMVCPSzSWYnDhKiD8TbCMTkHjJqE7fEXplQLtJIJIJsAdwEFtf9qjJ3jiP6MU6WZJiSzoiyCDX/YlYbhZrHcv3BjvJgNodU29iJ5rNnT9+0s5TGbLu+N7+HO/cuVMJUC2YuCxLtsPGjRvx/PPPq6X5+tMW/bGNz5LAQBN4767bUWM1wSxGJo7OvwFHyjcgOnEzjGIe4KqzVmHDk/tQXeGD3diAs1Pt8LbaYUqJQvJV9Pgd6LZgeSQwFAQ6/x94KCyIsHdKkWMRMQdPloxCeJjEUKbMO9iiQwrPjkku+J6Xl9cnO3JzcyHn+Emnjp56ATu+pzfHspeuoqKivRdwy5Ytali34/Dy3r17RZDaVsjYjjLJPFKgyt7DnpJ0GpFzGDumH/3oR2oY/MYbb+xT/TuWwWMSCDUCu5/4K3Y2VImhXjHNY97F+N6hCsSMeFWZedPUm9C2LgpFuythggsXpFUL8TceBjFPMPVrk2CKCUzpCLU60R4SIIG+EeAQcN94RUTu0tJSrFq1CoWFhXj99dfx3HPPKYeOvlT+ggsuQHNzM26//XbVc1ZUVKRWWZFl9ifJoVk532/Pnj3YtGmT8tRdsmRJ+/w/Wbb0+JWBrA8cOIB169ZBeg/LgOEnm/8XGxuLsWPHdtqkcExMTFTX+mM3nyWBUCHQcGA/3l0jnK2k+EvNwqqWUYjOfgEGUxtmps/EvPql2PJBpTJ3cep2+NvGy+V+kSJ6/izpnXv6Q6VOtIMESKDvBNgD2HdmYf/EZZddpnrQVqxYoXq9pDfvlVde2ad6SweQJ598Uq29LFcXkeJLLtcn5wX2J8meReltLHvkGhoaMH/+fPzsZz/rVOSZZ56peiyvuuoqJQalg8ott9zSKQ9PSCASCXi9Hvz7Zz+Ey2REvA/YMPcGVJc/D1v0EcRZ4nB7+k/w0eMHBBoj5sWvh8VzjsKUePEoRI1JikRkrDMJhC0BCsCwbdpTq1jHJfRkXL6u0onL6skh166SdMaQcf66SjLsy4mpp+XnOuaV8QTl1lOSnr8nrvvcU/7u7nXk0V0eXicBvRBY96t7UOlpg0k4Tg1f/m38Zt+niM5dp8z/8bh78dlTRfD5TZhi/xwp1nnC6QOIOSMLsWfS41cvbUw7SaC3BCgAe0uK+UiABEhAxwQKP3gPBTsK1NDvaRNn4bZDXkRlvyRGgv1Ykftl1L1kgMttRK51P0bHz4Sv1QDb6EQkXjxSx7Wm6SRAAt0RoADsjgyvhx0BObdx2bJl3dZLxgAcNow9Hd0C4g3dEmiurcHaPz+kxF++wYqXxlyKlvr7YDY3YWTsGEz4eC4qG/xINJVjdloWPC1WmFOjkSJW+jCI4WImEiCB8CNAARh+bRq2NZLz+E42l6+roWUNSHp6unJE0c5P3Mv7TCQQbgT8Yrj3tZ/8AK3wI7bNDfs3f4y1219BVOYeWIQYvLb0FpSX+GE1tGBhlgue5iQYos1IuW4SjHZ6/Ibb54H1IQGNAAWgRoL7sCcg4w7KcDZMJBBJBD7526MorREhXbw+nLXoUly9dS9smWsUgps9P0b5Tr9Y5M2LCzP2wds8R/h/GAIev6IHkIkESCB8CbBvP3zbljUjARKIcALF2wrw6TsBsTczJhm/jZ0KV8qzMBi9uMj3VbR+FvDsPS/lExhahfgTKfFS4fEr5v4xkQAJhDcBCsDwbl/WjgRIIEIJOBob8PoDvxTz/oCc5lbUXfM/+KT+CZhsVRjdOg25G09TZObEfohYw0J1HDt3GGJP7zlgeoTiZLVJIOwIcAg47JqUFSIBEoh0AnLe3+ur7oHT1YbYVhdOu+YmXFbwFqzpGxHXmoILd10Nj8+E0bYCZMfPh88B2MYmIWHZyEhHx/qTQMQQYA9gxDQ1K0oCJBApBDa88iKOFO6HUQjBs3LH4kctdvhT/gWrJwpfPXCbWI/cinRzIaZlThbizwhzuubxK7oLmUiABCKCAAVgRDQzK0kCJBApBEr27MQn/3xeVXdKYxu2XXojdrj+LFbjceGyg9+BoSkeMcYanDM8Fp4Gu/D0Nas1fo1RHBCKlM8I60kCkgAFID8HfSawcOFCPPXUU31+jg+QAAkEl4A2788vXjOsrgkjvnsXfrX9KZiiD+OcQyuRXJsHs6EVi4eXw12XHfD4vXoCzCn0+A1uy7B0Egg9AvzJF3ptEvIW/etf/0J0tP7/h/H222/j0UcfRXFxMTwejwoRc/3110OuhcxEAnojIOf9rX34AbQ0NyFGzPubO2cevlNWCkPSe5hSfi4mVJwpquTD4syN8DYsVNVLunw0bCPp8au3tqa9JDAQBCgAB4JihJWRnJwcFjVOSEjAt7/9bYwcORIWiwX/+c9/8OMf/xgpKSmYN29eWNSRlYgcApteX42i7VvUvL/ZbQasmXcxio7cgfz6CZhbFPhRc07yf2B2LxFQRFDos7MRMzszcgCxpiRAAp0IcAi4Ew6eXHPNNfj5z3+utlmzZuH000/HQw89BL9fDioF0olDwOPGjcM///lP3HzzzZg2bRouvPBCfP7551p2td+3bx++8Y1vYMaMGZg7dy5+8IMfoLa2tlOejievvPIKTjvtNLz77ru44IILMGXKFNxwww0oKytrz/aHP/wBl156KV544QWce+656t233normpqa2vP0dCDrdv7552PUqFHIzc3F1772Nci6nGh7T2XwHgmEAoHSfbvx4d+fUqZMLK1B3B0/wR8P/B/SXDG4YN91IhKMEVNiPkJK1IWAx4+occLjd+mIUDCdNpAACQwRAQrAQQYvhZTP5e3VJr+oe5u3u3wdhVtvq7p69WqYTCYl6v73f/9XzfeTAq+n9Mc//hFLlizBv//9b5xzzjl45JFHUF9frx5pbGxU4mrixImQw8ePP/44ampqcNttt/VUJFpbW/HnP/8Z9913H/7xj39AlvP973+/0zNy+Hbt2rVqKFeWu3v3btxzzz2d8vTmRHJav349Dh06hNmzZ/fmEeYhgZAg4BRDvq//32/Uj7QsMe9v6iXLcVOhiO1nPYIle74Js8+GXOt2jE2fp8K9mDPsSP6KWONXrPjBRAIkELkEOAQ8iG0vRUbVo9vgOtzYq7dmiFwVqzf0Km93max58Uj71lQYDL3/ss/KylJDofIZOTwqe++k08fKlSu7ew0uv/xyXHTRReq+FHbPPfcctm/fjvnz56tjKf5uv/329ud//etfq147KbhGjOi6J8LtduPuu+9WPXvyQbnO79KlS7Ft2zZMnTpVldXW1ob7778fGRmSFvCTn/wEN910E374wx8iLS1NXevpj+wtlILV5XIJL0kjfvazn+Gss87q6RHeI4GQISC/U97600Noqq2Bvc2FWbZ4PDF5KurK7salu7+DWFcSEk0lOD1/BFyVZhhjLPT4DZnWoyEkMLQEKACHln9Ivl0O43YUjNOnT8eTTz4Jr9erega7MloOnWrJbrcrJxFtiHfPnj3YsGGDGv7V8mh72YPXnQCUa/fKoV8tyaHa+Ph4HDx4sF0ASrGqiT+ZTw4x+8RkeCkseyMAY2Ji8Oqrr8LhcKgeQCkyhw8froa+tfdyTwKhSmDzmv+Hg59vEPP+/JhxtAat9/8Bf993NxYf/jLSW3IRZWjE+aPcQvyJJd9MYo3fa4THb3JUqFaHdpEACQwiAQrAQYQtRZXsjfO7fSd9qxRbWk+XHI491WSwGDuJuVMt52TPSSeKjknWVQoxmaS4WrBgAf7nf/6nYxZ13BuR9oWHBvCC7PXLy8tTJU6YMEGJy8cee4wCcAAZs6jgECjbvxf/fe6JwGe3tBojrr8Rl+5/DmdUzMao2uli1p8bF+btEuIv4NCUdMUY2PITgmMMSyUBEtAdAQrAQW4yKYwM1pMLOr9XGGYWU7dFXmM/BOCpVE8Kz45p69atSiSdqhCdNGkS3nrrLWRnZ0P26vU2ydAsO3bsaO/tKywsVPMAZU+glqRTSEVFRXsv4JYtwgtSiLruehW157rbS9Eqh4OZSCCUCbQ2N+P1h+5TP7Iy65sxLicfv86KQs4BF2aWnq9MX5S1Dv6mpeLYj7hzcxAzKzBNIpTrRdtIgAQGjwCdQAaPtW7eVFpailWrVkEKrtdff13N4bv22mtP2f6vfvWraGhoUHMApbiUw74ffvghfvSjH6lh5e4Klr2Kv/jFLyAFqBSCMr8cjtbm/8nnbDabmu8nh5k3bdqEX/7yl8oZpTc9i3/5y1/w8ccf48iRI6rn74knnlBOLJdcckl3JvE6CQw5ATXv79GH0FhdKeb9uTG1ugmHv/Mt7D78Js4pvFLZd3riu2K1j2WAV3j8TkxB/OL8IbebBpAACYQWgd53x4SW3bQmiARkIGTpgbtixQo150+KvyuvDPyP5VReK+foSS/eBx98UIVykT1sw4YNU7H2ZG9ddykqKgo33ngj7rjjDtXLJ8PC/OpXv+qUXYZvkaFcZD4pMucLpxPpyNGbJIem7733XpSXl0O+Szq8PPDAA8rRpDfPMw8JDAWBgjdfw4GNnwbm/R0uR9ptd+CW7c/g4v1XweRkote3AABAAElEQVQ3YXT0pxiesgTuGh8sWTFIvnIcPX6HoqH4ThIIcQIUgCHeQENhnhymleFfpDjqKr3//vudLu/du7fTuTyRIVlkb52W8vPzIUPF9DXJGIBy6ynJHka59TXJkDInhpXpaxnMTwKDSaD84H6sezYw7298WTWGzTgNPxBevucfWIwobwwSrfswc+RstB0Vq7zFWpDytYkw2k4+5WQw68B3kQAJhAaB7rtfQsM+WkECJEACJCAItLbIeX+/gc/rQYaY9zeiTcSu/MrFyN2aisTWdJjMVVg8PlmIP+HlK+YPp1wzEeZEevzyw0MCJNA1AfYAds2FV8OAgAwJ013661//qlYa6e4+r5NAKBGQ8/7e/svDaKisQLTLg6lHqhD1s5/ik3UHMLZpJnxGJ740pgKtxYHPfPLysbCJGKBMJEACJNAdAQrA7shE6PVnn302JGp+xRVXQG49pVtuuQVy6y7J+H7dpY6xA7vLw+skECoEtrz9BvZv+AQGsSLjjKJyJC9ahPsKizG++jT44MUFuevhrhTLvEmP3wXDYZ+RHiqm0w4SIIEQJUABGKINQ7P6T0CL79f/klgCCQwdgcpDB7HumceVAeNFvL+UmDi8PGUGxn+eq66NS38Lca4rxNCwF9GThMfv+YG4lkNnMd9MAiSgBwKcA6iHVqKNJEACEUnA09aKNx6+H14REzOjoQX51Q2ovuFmYHOm4mFO/A+mxi8Xa/x6YRkWgyR6/Ebk54SVJoFTIUABeCrU+AwJkAAJBJmAnPe3963X0FBRjmgRz2/KkUqYL/sy1m+OgtlvRWP8Nlw6fBE81R4Y46TH7yQVOD7IZrF4EiCBMCFAARgmDclqkAAJhBeBHe+/jao9O2AQ1ZpeeBS2zFyscU1ElDse9dEluFysWe05Ir7CzUakXjsJ5gRbeAFgbUiABIJKQNcC8Pnnn8fChQsxZcoUFbT4xCXMOpJ76aWXVKy42bNnQ27XXXedWmu3Yx75i/v3v/89zj77bLXahMxTVFTUMQuPSYAESCDoBPxiScL1L/9DvWdcaQ0SnW5smH0dLI4UtFgaMC2zHLFHA44eySvGwjo8Lug28QUkQALhRUC3AnDNmjVqubKbb74Zq1evxvjx49UqEzU1NV220IYNG7Bs2TI888wzeOGFF5CVlYWvf/3raoUJ7QEZGkR6wd5zzz2QgjE6OlqV2dbWpmXhngRIgASCTqDswD446utgEkIwr7oehxbfhub6NLiNLlgyP8D0xrnKhrjzcmGflhZ0e/gCEiCB8COgWwH45JNPYuXKlVi+fDlGjx6tVq2Qy3m9/PLLXbbSb3/7W1x11VWYMGECRo0apdaM9clf2evXq/yy90+Kw29/+9tYJEIsSEF5//33o7KyEu+++26XZfIiCZAACQSDwMHPN6hi0xsdqJx8KYpaR6nzkpx/Y7lrpVrjN3pqKuKFAGQiARIggVMhoEsBKNeS3blzJ+bODfwKlhWXa8rK84KCgl5xcDqd8AjPuoSEBJW/pKQEVVVVncqMi4vDtGnTel1mr17MTCRAAiRwEgIHPgv8MLV7E7E75TyVe3vOGlznXgFjm1jjNycWcujXYJQzBJlIgARIoO8EdBkHsK6uDl4R8yolJaVTjeV5YWFhp2vdnTz44INIT09vF3xS/MnUVZnV1dXdFaOuS1vkNpBJK0/bD2TZwS5Ls1nbB/t9g12+Vi9tP9jvH8j3aXXQ9gNZdrDL0mzW9sF+32CVX19ehtrSEhHT2Y+y/GvFa03Ym/opLneficQ2M4zxViRdNV6s/iFuDfD3zmDVMZTeo31+tH0o2Rbptmhtou0HkkcwyhxI+wajLF0KwP6CeeyxxyDnEMohX5vN1t/isG/fvn6X0V0B27dv7+5WyF/Xs+29gRtO9dNzXfRse1efsyMbP1GXLcZ0+MyJKI3fjym2aExqSYffBFTNMaKscFdXj/JaPwiE2+eoHyhC7lG2TXCaRJcCMCkpCSaTCSc6fMjz1NTUHkn97W9/gxSAcg6hnOenpbS0wERqWYbsGdSSPO+YT7vecT927FjY7faOl/p9LH+dyA+99HCWddVT0rPtveEcTvXTc130bHtPn7P9/+9FddsfNUkt82ZJ3Y8ltZeqa0krx2HY5M4jHz2VxXsnJxCun6OT1zz0cwSzbRwOR1A7b0KfroggpQcjT7TRarVi0qRJyoFDOmzIpDl0XH311Sdmbz+XXr6PPvoopAiUwqpjysnJgRSB0ilEOorI1NzcjK1bt+IrX/lKx6xfOJYCLVgiLZhlf6EiA3xBz7b3BkU41U/PddGz7Sd+zhyNDSjdt1tdNlqE40diCa6rvUSdx1+Qh9hpx3+cnvgsz/tHIJw+R/0jEXpPB6NtZJmRnnQpAGWjXX/99bjrrrswefJkFbPv6aefhnTsuOKKK1Sb3nnnncjIyMAdd9yhzmWv38MPPwzpDZydna0cPuQN2XMXExMDg8GAa6+9Fn/+858h15CVglDGBJS9gZrIVAXxDwmQAAkEicChgk1i6p8fJkMi4iwJmGuIE6t+GBA9PQ1xC4YH6a0slgRIIBIJ6FYALl26FLW1tUrUSQcO2Wv3+OOPtw8Bl5WVKc9grVFl7D+3243vfe972iW1/+53v4tbbrlFHd94441KRN59991obGzErFmzVJkDMU+w00t5QgIkQAJdEDiw6dPAVds4zLAbYfdbYBkuPH6XC49f8SOViQRIgAQGioBuBaAEIId7uxvylQGdO6b333+/42mXx/IL9tZbb1Vblxl4kQRIgASCRMAjwlsVbflclZ4QNR4pZhO88CP9K+NgsOgyYleQSLFYEiCBgSDAb5WBoMgySIAESKCfBIp3bIUUgWLAF7lRAae0plQjTFzjt59k+TgJkEBXBCgAu6LCayRAAiQwyAQObgqs/mG0jUWO9dhw7yhdD9IMMkG+jgRIoC8EKAD7Qot5SYAESCAIBPxiWcoDGwOrf6RGT0KMyQiX0Y+2YRSAQcDNIkmABAQBCkB+DEiABEhgiAmUH9wPGQIGsCA3OktZ4x2dKAJ10fFjiJuGryeBsCVAARi2TcuKkQAJ6IXAwc8Dw78my0gx/BuITzZs7jC9mE87SYAEdEiAAlCHjUaTSYAEwovAgQ2B5d+GxUyD1WhAs8UL2yjRA8hEAiRAAkEiQAEYJLAslgRIgAR6Q6C+vAw1pSUiqwG59pzAI+NTYRBCkIkESIAEgkWAAjBYZFkuCZAACfSCgBb82WrJR5Y14PSRf05eL55kFhIgARI4dQIUgKfOjk+SAAmQQL8JHNwYWP1jeNwssQScAVVRbthyYvtdLgsgARIggZ4IUAD2RIf3SIAESCCIBJxNjTi6d5d6Q250YK1f48RMLvsWROYsmgRIIECAApCfBBIgARIYIgKFmzfC7/cj2pqLNItFWTFuQf4QWcPXkgAJRBIBCsBIam3WlQRIIKQIaKt/5MXNVr1+RTGtiEqzh5SNNIYESCA8CVAAhme7slYkQAIhTkCu+3uoYJOyMs+eq/aWSdkhbjXNIwESCBcCFIDh0pKsBwmQgK4IFO/cCo/bhXjrcCSarfDAiynnjdRVHWgsCZCAfglQAOq37Wg5CZCAjglow78j4k9XtdgT54A9IUrHNaLpJEACeiJAAain1qKtJEACYUHA7/PhwGfrVV2GHxv+jZrC2H9h0bisBAnohAAFoE4aimaSAAmED4Hywv1wNDYg1ZaPGJMFDoMLpy0cHT4VZE1IgARCngAFYMg3EQ0kARIINwLa8G9+/BxVta3xLYiNtYVbNVkfEiCBECZAARjCjUPTSIAEwpPAgQ2fwAgThtsDwZ9jp44Kz4qyViRAAiFLgAIwZJuGhpEACYQjgfqKctSUliDLPhJWoxnVxmaccS6Hf8OxrVknEghlAhSAodw6tI0ESCDsCBzcFFj7N0+s/SvT5oQWJHH4N+zamRUigVAnQAEY6i1E+0iABMKKwP4N62Ex2jDs2Nq/cVPHhVX9WBkSIAF9EKAA1Ec70UoSIIEwIOBsbkLpvl3IsY+DyWBEkbka55zF4d8waFpWgQR0R4ACUHdNRoNJgAT0SuDQ5o3w+/3Ii5umqrAp3omM+Gi9Vod2kwAJ6JgABaCOG4+mkwAJ6IuAHP6NNsUh3ZalDI+fOklfFaC1JEACYUOAAjBsmpIVIQESCGUCHrcbhwo2Ijd2AgwGA7ZbjmDxGWND2WTaRgIkEMYEKADDuHFZNRIggdAhcGTHVni9HuTHTlFGfRbfhuxEDv+GTgvREhKILAIUgJHV3qwtCZDAEBHYt/5jJFhSkWhNhcfvQeLkgBAcInP4WhIggQgnQAEY4R8AVp8ESCD4BPw+H/Zv+Bh5sYE5f59bC3HxaeOD/2K+gQRIgAS6IUAB2A0YXiYBEiCBgSJQUXgAba2OdgH4UYILI9NiB6p4lkMCJEACfSZAAdhnZHyABEiABPpGYO/HHyItajjs5ji0+VuRPmFG3wpgbhIgARIYYAIUgAMMlMWRAAmQwIkE9ny0rr3371Prblw6Y+KJWXhOAiRAAoNKgAJwUHHzZSRAApFGoKGyHI6mBgwXq3/I9F6CFxOy4iINA+tLAiQQYgQoAEOsQWgOCZBAeBHY898PkBU9ClZTFFq8DRg29jQVBzC8asnakAAJ6I0ABaDeWoz2kgAJ6IrArv+8L2L/BYZ8P7ZuwxXTJ+vKfhpLAiQQngQoAMOzXVkrEiCBECDgbG5CU20tsuyjlDWvJwHTcxJDwDKaQAIkEOkEKAAj/RPA+pMACQSNwN51/8HwmHEwGcxocJVhxMjTYTQagvY+FkwCJEACvSVAAdhbUsxHAiRAAn0ksOOdd5AXc2z417IFX5o2vY8lMDsJkAAJBIcABWBwuLJUEiCBCCfgcbvRXFWP9Ohc+P1+vJJixOx8MQbMRAIkQAIhQEDXAvD555/HwoULMWXKFKxYsQLbtm3rFun+/ftxyy23qPzjxo3DU0899YW8f/jDHyDvddwuvPDCL+TjBRIgARI4GYH9H64Tw79jVbY6ZyFGjzgbZpOuv3JPVmXeJwES0BEB3X4brVmzBqtWrcLNN9+M1atXY/z48bjhhhtQU1PTJX6n04mcnBzccccdSEtL6zKPvDhmzBh89NFH7dvf//73bvPyBgmQAAl0R2Dr2jeF929g7d/1ps1YPnVmd1l5nQRIgAQGnYBuBeCTTz6JlStXYvny5Rg9ejTuvfdeREVF4eWXX+4S4tSpU3HXXXdh2bJlsFqtXeaRF00mkxKIUiTKLTk5udu8vEECJEACXRGQQ77OihYkWNPg83nxzzQT5o5K6Sorr5EACZDAkBAwD8lb+/lSl8uFnTt34qabbmovyWg0Yu7cuSgoKGi/dioHhw8fxtlnnw2bzYbp06erHsNhw4b1WJTX64XcBjJp5Wn7gSw72GVpNmv7YL9vsMvX6qXtB/v9A/k+rQ7afiDLDnZZms3aPtjv60v5hz5dL1b+CIR+qW3cibxp82EWzr99sVXLq+378n7m7T0Bja+27/2TzBlsAlqbaPuBfF8wyhxI+wajLF0KwLq6OvVFmpLS+Re1PC8sLDxlbrKXUA4rjxgxAlVVVXjkkUdw1VVX4bXXXkNsbGy35e7bt6/be/29sX379v4WMWTP69n23kALp/rpuS6haPuef72C2bGL1Mdog/lzzIiZhi1btvTmY/WFPKFYvy8YGQYXyDl0G5FtE5y20aUADA4K4Nxzz20vWs4pnDZtGhYsWIC1a9cqJ5P2myccjB07Fna7/YSr/TuVv07kh146uMhhaT0lPdveG87hVD891yWUbd/3yIuwp8fB42nDv1LMWL14DqKtfft3HMr1682/E73kIefQbalgto3D4UAwO29Cl+pxy3QpAJOSkpQoOtHhQ56npqYer10/j+Lj45Gfn4/i4uIeS5ICLVgiLZhl91ipAbipZ9t7U/1wqp+e6xJqtpds2ojs6Dz1Eaqr2YyMqYsRG939vOOTfdZCrX4ns1ev98k5dFsuGG0jy4z0pEsnEOnEMWnSJKxfv769/Xw+nzqfMWNG+7X+HrS0tODIkSM9eg339x18ngRIILwIbF69Fjli9Q+ZNpk3Yvnk08OrgqwNCZBAWBDQZQ+gJH/99dcrr97JkydDzt17+umnIUO9XHHFFaph7rzzTmRkZCgnDnlBOo4cPHhQ3ZPHFRUV2L17txq6zcsL/Fq/77771JCvdPqorKyEjAsonUsuuugi9Rz/kAAJkMDJCHgqAWuKDa2uJrwm9s9NzDjZI7xPAiRAAoNOQLcCcOnSpagVi6w//PDDymFjwoQJePzxx9uHgMvKypR404hKQXfZZZdpp3jiiSfUNmfOHDz77LPqenl5OW6//XbU19er8C+zZs3CSy+9xFAw7dR4QAIk0BOBiq1bMCwqR2VpKv8MtkVLEB9l6ekR3iMBEiCBISGgWwEoaV199dVq64qcJuq0ezII9N69e7XTLvf/93//1+V1XiQBEiCB3hDY/PJbmGgPDPkWGD8Vw7/X9eYx5iEBEiCBQSegyzmAg06JLyQBEiCBXhDwVVlgMpjQ5KzCu2lxOH9iVi+eYhYSIAESGHwCFICDz5xvJAESCEMClVu2YpgtEDTeWfopnGOXICXWFoY1ZZVIgATCgQAFYDi0IutAAiQw5AS2v/oB0qKyIZeB2+Vdj0umHo8rOuTG0QASIAESOIEABeAJQHhKAiRAAn0l4BdhqHxVgd6+upYj+DAzCUsmZ/e1GOYnARIggUEjQAE4aKj5IhIggXAlULWlANm2wHy/tpL1qBx7PjITosK1uqwXCZBAGBCgAAyDRmQVSIAEhpbA3n9vQrw1GR6fG8XODTh/+gVDaxDfTgIkQAInIUABeBJAvE0CJEACPRFQw7810SpLdVMh1menYNmUQCzAnp7jPRIgARIYSgIUgENJn+8mARLQPYHKTZ+J4M+B1T5cR9bj0OgFyEuJ0X29WAESIIHwJkABGN7ty9qRAAkEmUDh2j2INtnR5nWirvlznD3n4iC/kcWTAAmQQP8JUAD2nyFLIAESiFACfq8XqLWr2pc37sWm7GQx/Ds8Qmmw2iRAAnoiQAGop9airSRAAiFFoPyjT5BpS1U2uYT3794RczEmPTakbKQxJEACJNAVAQrArqjwGgmQAAn0gkDJf4phMVrQ7K6Ht2YHJs/9EgwGQy+eZBYSIAESGFoCFIBDy59vJwES0CkBv9sNY0Ogt6+0cRe2ZSfgoqn5Oq0NzSYBEog0AhSAkdbirC8JkMCAECj94COkWRNVWW0lG7Aj/zRMESKQiQRIgAT0QIACUA+tRBtJgARCjkDFR1UwGoyobSuHraYQuWd9mcO/IddKNIgESKA7AububvA6CZAACZBA1wR8rjZYmuMBC1DSuBNH06KwdMaYrjPzKgmQAAmEIAH2AIZgo9AkEiCB0CZQ+uaHSLLEwOf3obV0I7bmTsWsvKTQNprWkQAJkEAHAhSAHWDwkARIgAR6Q6B6Y6PKVtl6GEk15Ug4+8swGen92xt2zEMCJBAaBDgEHBrtQCtIgAR0QsDjbEFUq3D2MAHFYvjXFW/AhXOm6MR6mkkCJEACAQLsAeQngQRIgAT6QKD0tQ8Ra7LC43OjrWILNg8fjzNGpvShBGYlARIggaEnQAE49G1AC0iABHREoH5rm7L2qGM/UuvqYD5jJSwmfpXqqAlpKgmQgCDAby1+DEiABEiglwQ8TQ2IcQdi/R1u3gWnxYMF807v5dPMRgIkQAKhQ4ACMHTagpaQAAmEOIGjqz+GzWhCm9eBtqpdKMgeiXljAmsBh7jpNI8ESIAEOhGgAOyEgyckQAIk0D2B5j1edbO4ZTfSG5vgmnUJoizCG4SJBEiABHRGgAJQZw1Gc0mABIaGgKu6GjFeEfxZJDn8a/G04YxF5w2NMXwrCZAACfSTAAVgPwHycRIggcggULr6U5jF0m/N7jq0NhRhe2Y2FkzIiIzKs5YkQAJhR4ACMOyalBUiARIIBgFnYeDrskj0/mU0tKBh2gWItTGUajBYs0wSIIHgE6AADD5jvoEESEDnBNpKyhDji1W1ONy8EwkOJ6YtuVjntaL5JEACkUyAAjCSW591JwES6BWB0lc/h9FgQE1bKVzOKuxPScZ5U7J79SwzkQAJkEAoEqAADMVWoU0kQAIhRcBdYlH2SOePjEYHKibOR6LdGlI20hgSIAES6AsBCsC+0GJeEiCBiCPg2FcEu/jP5/ehuHk3UptaMGbp5RHHgRUmARIILwIUgOHVnqwNCZDAABOoeH2HKrHCWQSPuxnlMXacN2fUAL+FxZEACZDA4BKgABxc3nwbCZCAjgj4/X74KqOUxdL5I7XZgeJxc5EeF7imo6rQVBIgARLoRIAxDDrh4AkJkAAJHCfQsmUfbOI/j8+No479mNjggG8xh3+PE+IRCZCAXgmwB1CvLUe7SYAEgk6g8p396h1HHfuECHShxWzEwnOnBv29fAEJkAAJBJsABWCwCbN8EiABXRLwe30w1MQo22Xw56SWVhSOmImcJLsu60OjSYAESKAjAQrAjjR4TAIkQALHCDR9shMWgxmt3lZUOA+J8C8tSFl0GfmQAAmQQFgQoAAMi2ZkJUiABAaaQM1/i1WRxcL5wy/+M3k8OOvCMwf6NSyPBEiABIaEAAXgkGDnS0mABEKZgK/NC2Pj8aXfYlpdKMyZiNHpcaFsNm0jARIggV4T0LUAfP7557Fw4UJMmTIFK1aswLZt27qt+P79+3HLLbeo/OPGjcNTTz3VZd6+lNllAbxIAiSgewKNH2yByWBEk4j7V+sqU8O/9nkX6b5erAAJkAAJaAR0KwDXrFmDVatW4eabb8bq1asxfvx43HDDDaipqdHq1mnvdDqRk5ODO+64A2lpaZ3uaSd9LVN7jnsSIIHwIlD3aZWq0OHmwI/KeEcrZl92XnhVkrUhARKIaAK6FYBPPvkkVq5cieXLl2P06NG49957ERUVhZdffrnLBp06dSruuusuLFu2DFZr12t49rXMLl/EiyRAArom4G1sg9kR8PQ93LwdVrcHh1PzMHF4sq7rReNJgARIoCMBXQaCdrlc2LlzJ2666ab2uhiNRsydOxcFBQXt1/py0J8yvV4v5DaQSStP2w9k2cEuS7NZ2wf7fYNdvlYvbT/Y7x/I92l10PYDWXawy9Js1vYD9b7adwpgMBhQ46pHs6ceOY0OOM69HD6fb6Be0atytHpp+149xEx9JqDx1fZ9LoAPBI2A1ibafiBfFIwyB9K+wShLlwKwrq5OCa6UlJROjOR5YWFhp2u9PelPmfv27evta/qcb/v27X1+JlQe0LPtvWEYTvXTc10G2vbEzXVi7Y9YFDVvUR+D1CYHqqaOwZYtgfPefDYGMs9A128gbQunssg5dFuTbROcttGlAAwOilMvdezYsbDbBzY4rPx1Ij/00sHFZDKdunFD8KSebe8NrnCqn57rEgzbPZVC7HnFcm9iDeAjTdthFL1+VbGpWL70HBiNht58PAYsTzDqN2DGhVFB5By6jRnMtnE4HAhm503oUj1umS4FYFJSkhJFJzp8yPPU1NTjtevDUX/KlAItWCItmGX3Ac8pZdWz7b2pcDjVT891GUjba98POH1UtNWizedARpMTrtOWwmIZuq/Kgaxfbz7XkZqHnEO35YPRNrLMSE+6dAKRThyTJk3C+vXr29tPzs+R5zNmzGi/1peDYJTZl/czLwmQwNAS8IteP8cupzKiqCUwlzi9oQVjrlg2tIbx7SRAAiQQBAJD97O2n5W5/vrrlVfv5MmTIT18n376achQL1dccYUq+c4770RGRoYK+yIvSCePgwcPqnvyuKKiArt371ZDt3l5eer6ycpUmfiHBEggLAm4DjfA5LXBI4Tg0SbREyj2TrMdC08bH5b1ZaVIgAQim4BuBeDSpUtRW1uLhx9+GFVVVZgwYQIef/zx9iHgsrIyMWfneAdnZWUlLrvs+DqeTzzxBOQ2Z84cPPvss+pTcLIyI/ujwtqTQHgTqH9XDv8acLS1Bl6/G0ki9l/TlHkwm45/j4Q3AdaOBEggkgjoVgDKRrr66qvV1lWDaaJOuyeDQO/du1c77XbfU5ndPsQbJEACuibg9/rQdlAs/wYzihyfq7pkNDgQf8lSXdeLxpMACZBAdwT407Y7MrxOAiQQMQRa91TD6DejVcwlLm/cqupt9gCzz5sdMQxYURIggcgiQAEYWe3N2pIACXRBoOE/u9TVI2L4V0z+Q0yrC3Vj5yBqCL1/uzCTl0iABEhgwAhQAA4YShZEAiSgRwK+Vg/cJYEYf0Wtm1QVMhpbkHHhYj1WhzaTAAmQQK8IUAD2ChMzkQAJhCsB57Zy4fphRJOYB1gtgj/LFOdwYc4lC8K1yqwXCZAACYhvPSYSIAESiGACDf/dr2p/pLVWrPzhh9XtQUPOJMTEREcwFVadBEgg3AlQAIZ7C7N+JEAC3RLwNrbBWx1YEeCQ+zOVL73RgYRF53f7DG+QAAmQQDgQoAAMh1ZkHUiABE6JQMvnR8XwrwE1Hh/qmwOOICli+bdZK7j6xykB5UMkQAK6IUABqJumoqEkQAIDTaBp/WFVZElrPcxuEQdQhIFxJOchKTVxoF/F8kiABEggpAhQAIZUc9AYEiCBwSLgrmiBv9EMn1jyrci/Qb02TfT+RZ+7cLBM4HtIgARIYMgIUAAOGXq+mARIYCgJtGwsUa+v8PjR0LJHHaeL8C9Tr7x4KM3iu0mABEhgUAhQAA4KZr6EBEgglAj4hbdv86YyZVJJWyNsTpeI/+yH15aMjBHDQ8lU2kICJEACQSFAARgUrCyUBEgglAm4ihuBVhPcQvQdNQaGf5McrTCdcW4om03bSIAESGDACFAADhhKFkQCJKAXAo6NR5SpZW4/aluPDf82ODBxxUV6qQLtJAESIIF+EaAA7Bc+PkwCJKA3An4R8qVlq1zzFzjqaoa9wamObT4rsmdMVsf8QwIkQALhToACMNxbmPUjARLoRKB1bx3gMaJVzAMss25UcQBjWsUcwOlnwmAIrAnc6QGekAAJkEAYEqAADMNGZZVIgAS6J6AN/5a4fKhx7VYZM4T378hLl3T/EO+QAAmQQJgRoAAMswZldUiABLon4Gv1wLlXOICIVOp2wF7TpI7jnV6MXHiWOuYfEiABEogEAhSAkdDKrCMJkIAi4NxRLcK9GNDkFc4fsVth9hlhFSuA+MZMh8FsJiUSIAESiBgCFIAR09SsKAmQQMsx798jYvi33LNDAZHBn4dfxOFffjpIgAQiiwAFYGS1N2tLAhFLwNvQBtfhgMdvuacFUZUBT+CU5laMWXZexHJhxUmABCKTAAVgZLY7a00CEUfAsbVK1NmAGhEGpjl5L6JdZhh9PhiHjYEpNjbieLDCJEACkU2AAjCy25+1J4GIIaCt/XvEJVb/8G5V9U5tciLtgvMjhgErSgIkQAIaAQpAjQT3JEACYUvAXd4CT5UbPrH0W7WvBcaKUlVXGf5l/OVLw7berBgJkAAJdEeAArA7MrxOAiQQNgQcWypVXSo8frRlHEJis014A/thicuAJSM9bOrJipAACZBAbwlQAPaWFPORAAnokoBfrPjRsinQ4ye9f4v8BaoeSY5WJM6n84cuG5VGkwAJ9JsABWC/EbIAEiCBUCbgKmqEr9kHt+jxa0Az3GVFytz0BgfGLV8WyqbTNhIgARIIGgEKwKChZcEkQAKhQMBRUKHMKBXOH96cUmTWRqnzGGM0oseMDgUTaQMJkAAJDDoBCsBBR84XkgAJDBYBvwj54tgWmP9X4vbhgKFABIIxIKbVhfgzz4HBYBgsU/geEiABEggpAhSAIdUcNIYESGAgCbTuqYW/zQ+nnAdobEJz6V5VvPT+HX0ZvX8HkjXLIgES0BcBCkB9tRetJQES6AMBR8Gx3j/h/GHIrcbwisDwb6IQhfGnzexDScxKAiRAAuFFgAIwvNqTtSEBEjhGwOf0wLk7sNxbiRCAe20FMPtMsLq9iJ0yGwazmaxIgARIIGIJUABGbNOz4iQQ3gSc26sBH9DoFc4f1mZUH9mmKpze1IKRDP4c3o3P2pEACZyUAAXgSRExAwmQgB4JtBzz/pWx/wz5dcgvCwz/pja3IemceXqsEm0mARIggQEjQAE4YChZEAmQQKgQ8NS3wnWoUZlzVAjAPbHbEO2ywOjzISZ/Aox2e6iYSjtIgARIYEgIUAAOCXa+lARIIJgEHFuqVPHVIgyMOaoJRw9vUuepTU7kX7IkmK9m2SRAAiSgCwIUgLpoJhpJAiTQFwLODt6/ptHNyC21qsfTGx1IPZ/Lv/WFJfOSAAmEJwG6wYVnu7JWJBCxBFxlLXBXOOAVS7+Vuv04nLgDyc02QJzHpWTDnJoasWxYcRIgARLQCLAHUCPBPQmQQFgQ0GL/VQjxZ7c34eChj1W9Eh1tGH7hBWFRR1aCBEiABPpLgAKwvwT5PAmQQMgQ8IsVP5xbAmv/yqXfzGOdyCkNDHRkNLQgfQkFYMg0Fg0hARIYUgK6FoDPP/88Fi5ciClTpmDFihXYti0Q56s7omvXrsWFF16o8l988cVYt25dp6w//OEPMW7cuE7bDTfc0CkPTyKbgBQYrsONMNd64W10QZ4zhQ6BtkMNol3ccIl2kT2A21N3IaMmWhmYaI2FbeTI0DGWlpAACZDAEBLQ7RzANWvWYNWqVbj33nsxbdo0PP3005Bi7c0330RKSsoXkG7evBl33HEHbr/9dixYsACvvfYabr75ZrzyyisYO3Zse/558+apcrULVmtg8rh2zn1kElA9SzurUfP6IRga2iA/YZXvbYLfABjsFpgTbbCIzZRggzHeqvYmuT92bLSaIhPcINdaG/6Vc/8S4hrxbuEHmIt42NtcGDZ/4SBbw9eRAAmQQOgS0K0AfPLJJ7Fy5UosX75c0ZVC8IMPPsDLL7+Mb37zm18g/swzz0CKu2984xvq3m233YZPPvkEzz33HH7+85+355eCLy0trf2cB5FNwC8cB1r31qH2jUL4q5wQeg9ucU0sLgGbFH/ySosbHrkdbe4WliHKJMSgFIhSFIq9EoYdj4VwjLHAYJRvYDoVAn4x5KtW/xAPy6XfLJPcyNwRGOTIaHAgc+kFp1IsnyEBEiCBsCSgSwHocrmwc+dO3HTTTe2NYjQaMXfuXBQUFLRf63iwZcsWXHfddR0v4eyzz8a7777b6dpnn32GM888E/Hx8TjjjDMghWJSUlKnPCeeeL1iOFBsA5m08rT9QJYd7LI0m7V9sN8XrPLbChvQsLYI3tIW9QqPEH4H2zywGt9CtLEaDd4sOLw5YrgxEwZDAqIMBkQJvREtRFyU0HFRYh8tzs3iur/VC0+rA55KR/fmivzGOEtAHAqBqHoS1f6YYDx2bSB7E7U20vbdGxd6dzSbtb1zVzX8bV44xPBvjVDoFRm7kPteYPg3xQ1YxVQRLW/o1eaLFmm2avsv5uCVgSCg8dX2A1EmyxgYAlqbaPuBKTVQSjDKHEj7BqMsXQrAuro69UV+4lCvPC8sLOySW3V1NVJPCP8g88vrWpI9hOeffz5ycnJw5MgR/O53v8ONN96IF198ESZT90N4+/bt04oY8P327dsHvMzBKlCvtltqvIjZ7oatSiwkK5IUfofafKLnbwMm5n+OxulfgtuWgmHOclgd5bA5PoaxuQau+ja0NokOQYcNDe5UFLszxTq0mXB602Azmr4gDKVAlEJRCkbVmyhe52twqU3olW6TzyLyRRvgFVvHfeDYqK75RNQToUq7LePEG3ptK1kPzfaEj9sgF3uTK38kxNRi7YF3scCXAKvHi6gxE7BVp/+WtPqd2GY8H1gC5DywPAeyNLbNQNI8XpYuBeBx8wf2aNmyZe0Fas4gixYtgtYr2H7zhAM5h9A+wEtLyV8n8kMvHVx6Ep8nmBISp3q13S16+preLUabGPKVySeEX5EQE7Xeg5idtw4Zl16P9ba5ePXAq/DUeHDW6LMwMWUxhiWOgtXUYa6oeA6OGqChGKgvhq9mJ5rKa9BY2YLGOi8amqxoFALxiCtT9CJmQgxWqqFlKQJV76HoNZS9ibL3UPYoxph8iDZJkWiCUUw6NAp1aBRz3MyNPTigfKE3sXMvojY/0Sd+14TD58wgBHrFy4HVPuTavzFzjEjfHfjnkNbYgvHfWonY6dMDF3TyV6//jnSCt91Mcm5HEXIHwWwbh8OBYHbehBzMLgzSpQCUQ7JSFNXUiP/Jdkjy/MRePu22vN6xt09e7ym/vD98+HA1/Hv48GE1LCyvdZWkLcESacEsu6u6DOQ1vdjuFsOyje8cbp8/Juf9Fbv8Yh5ZJWZkrMG8yy/Czoz/wd0Fv8eWI5uQWxENp82LDyv+A7dFCDGjGaMTR2N88ni1TUyZiHFJ42CPny0+RLMh+46Tj22KryYQ6w/DX1eM5tIyNJTVo7G6FQ31Qjc2R4sexDSUiB5Et19bs1aKPQ/kP9iOw8xxZjfirT7YLUbYTGaYfWYhDoWSFMOgvelNlHMTk6N8aDpQCGtWLCxZMWozxXYQtMro0P0jP2fOXeK7QNS5wSvqInpSizL3YuR7AXZpLS7Eiekexh568UO3dlDfLcH6fgnleg+2bXr5vhpsLqHwvmC0Df9NielJodC4fbVBOmpMmjQJ69evh+yhk8knFnmX51dffXWXxU0Xv/4//fRTXNdhHqB0ApHXu0vl5eWor6+nU0h3gHR+3VPbisZ3D0N5jh7rTJPOAwfbRI9RwhtY/qVpKJnwY9yx7c94b9M9GH84DssPZMPmFt1yIslHGuM9KEt0ojLpCN5NOoBXo1+VniHKOSQvPg8TkidgfMp4tZfHiVGJgaHZGLEahdgM2bMQNxmIUyUe+3NMIPrriuAsK0Hj0So0lDehodaDhkYTGhyxQiCmo8qfIDxSxD9hZ8eHPe29iXFmF5KtTsRH+xFrsyDKYhPzF4WzidcCv9MPv6irnJtoaRVFFFSpTStJzkW0HBOE1syAKDSnRcNgCtRdyxcqe837t0QI9/SkRqw5/BYWumJhFN8LGeOnwBgdmAsYKvbSDhIgARIYagK6FIAS2vXXX4+77roLkydPxtSpU1UYGKfTiSuuuEIxvfPOO5GRkaFCv8gL1157La655ho88cQTOPfccyHDyOzYsaPdA7ilpQV//OMfsXjxYtWLKOcAPvDAA8jLy1Pew6pQ/gkLAl4RxqXx/WK0bBQBg0WvkUxlwoN0r9ON4fb3cPmyRDSfeTt+tec5rH5tBXLKbLh87zDEOcTkO5GMienweVwwNNcjodEstjiMLw5IOI/dhKokF47ENwhRWIo364uwtmitek7+yYzJDPQSJk9U+wkpE5BhzxDT9TrM15PHUhyKzZ5zGuyiIzGzvQRxoARiLdrKi9BYfFT0HtaiodKJhjqf6D20obE1Ac3eFLS6ragSGwI+LMdKkM5KXtEr2YYkay1S7C7EWnzITE1DtDUF/mYLPLVt8DW50dZUh7Z9geFw9bAYhrak24/1EsreQnkcC5PwXh7K5KlrhauoUWCRvbY+pM02InFvYBZlapMIBP3lC4fSPL6bBEiABEKSgG4F4NKlS1FbW4uHH34YVVVVmDBhAh5//PH2IeCysjJIz2AtzZw5Ew8++CAeeugh5dyRn5+PRx55pD0GoOwOlvMBXn31VTQ1NSE9PR1nnXUWbr31VjAWoEZR33tvswtNH5Sg+dMyMZoacPCoFMJvd6sPieaNWHq2CPC8+Ab89fAbeO6NLyOu2o/Fu1ORXi9dC4Tuio7DR3GnoQypqI+LgynJixG+Skw01SLdIT5vtUdhdniR5TAh66gc9BXJLKRWuhWlCQ4ciKkU4rASH7SU44MjH6jb8k+SLSkwfCx6CiceE4a58bkwGo5/ftszywMlEFNgG5WCtFGzkNbppjgRQsjTUI3GoiI0HKkQvYfCm7nGhcYGA+pbYtDkShQS0IZqV5bYjj1cJfdtSDAVISu+GsOE0EtOyRKf/Qy46y1wlzuUh61brrMrNhEF8diDQhAL72SL6CW0Hhs+lsPI5lS76C3sIGrbcw/8Qeu2gCNXtUeE7BF6fkfWXoz4T2D4N73Jgdj58wf+pSyRBEiABHROQLcCUHKXw73dDfk+++yzX2iaJUuWQG5dpaioKPztb3/r6hav6ZyAz+lB04dC+H1UKoY9A+F6aoQA3O30CS11EOdO3oO0y6/DCzWf46/vXAdfbQvO2JuI/PKYQM3NVhQkTkdDWzyu2fwOptQUwivE2aGEYdiZnI/dyXn4IHkuaofHIaOtErmeCoz2VSOx+SgMLidspU6MKDVgBDJUeaa0eLSkm1Esegl32kpQ56/D+rL1atNQ2832gCgU8wplL6EcPh6ZOBIWYy9624RANCemIXm63LQSj++9wiu2+agQhoePoK6kCiX7KtDYGI9aR5IKbdNQl4U9xzr+rIZyZNoOIjOlGZl5SUhMHymGvkXomzqTEIUt8Na0widWRGmTW8feQrPoLcwQQ8fHho/V3EJxPOC9hULsOrco9SrmTMqezCa8VvImzhfzKKUQTs/KhbmLwPDHafCIBEiABCKTgK4FYGQ2GWvdWwI+EROu+ZOjaFp3VMx186jH6qXwEz1+LUKgnZ73KUavXI43vKl45KPvo66uEtMOJIjh3Gwxd0xkF0Jqf8JEHPZn48odH+D0it2BV4vrJr8Po+tL1HZp4UeBsmOTsCMxDzuS8vBZ8jgczFyAeG8TstrKkNNWgeFCGNqdIoRRVSOihGYZK54aC9HLFieGUXOSUZ8qws3E1mCrsRAOjwObKzerLfBSKPE3JmlMYF7hMYeTsUnCA92iOYpoOXvem0SvZELeMLVlC29zj4iReaGYC+sR8wHLtx9E+c5DKC9yoLwmVsQ4jEFx61SIkWYRX0XObfQi2fwZsuzFyBzmRebcDNiSJ4lQOXlwV4tQObKHUPYWCqHtFoGx5dYxSe/jgKPJ8SFkc4qcW3hqvYXmBtHbWSlEqGiPUjH/b/hkK2L3yViL0Uh0tGH4ZZd3fD2PSYAESIAEjhGgAORHIewIyBUhmjeUieHeI/A1B+aCNYrAwHuEwKkSwmpm2oeY9qWz8XHiMvy44AEcqjkgHDzicd7BHFikB61IpbH52GkZg0v3f4pbj/47wEhME0gUc0yTbvomdu3ahZEiIHnr1m3CeaIArbt3ix6/Opwtt5ItKr/HYkVJej4K4odjqxjS/TR5DtxpBmS1VihROKytHOlt1XA1CZG0u1kMygLjxTbJkivEWQ782fGoTnZjX3Q5drXsR5O7CbtqdqktYJAYfhU9kfnx+e29hJoncoItQcvS670txoq8MyaoTT7kEx61NUcaUb5tP8pEL2H5UT+anDGo8YxATeMI7GgUmfYAdmMtMi3rkZlYjaxcGzIuyAcSZ8DtGQ5XlfuYKBS9hcLpRq6fLDe5ukp7MhtFb6E2tzDQayiHk41iib2TpajDAWFfJprZI/omd2TuRt5/Aw4fGSL8S9x5552sCN4nARIggYgkQAEYkc0enpX2C8HSsqkCTe8VK5Eha9kiru0RPX5H3R5MiF+PxRfnY9/oL+EbWx5GweYCjCiz44q9OYhxymAtImxfVCoKoiZjYdEOXF/8mOrpk9fjxZzTtO/dAquYOypjU/krKhAnes0SL7pI3oZPxJRybt+hxKAUhA7Rq2ZuaED+0X1q0/qhmtKGoVAMo35qz8aGuLkoTU9GmrtGiMJyJQqlOIx2t6L2QCFwQIQAEWVPENtZ2dOQMDIX7sxolAuv4z3+w9hduwc1rTUobChU2xuFb0hTVMqOzQ44mYihYzmELIVhWnRaZ2cTLXM3e6Pw+E3LTxTbbEw5lqdFBLsu31eJsh2il7CwCVU1VrHyRjIK285EYYXIJDbTRhfSLZuQaf07MtOdyByVCPuiKfClClHozRY9hM52USiHkaU3cpe9hWJdZS0sjTaEbE4VvYUixqFMcn3mqOLAkL50/hiW3oJHS9diUU1AAKZFJ8A2YoTKyz8kQAIkQAKdCVAAdubBMx0SkELAsaVShHQRwk/0MsnkFOE/pPA7IoYFc6K2Y+UiE+rnXo4fbn8MH7z9ADJqbbh4TxZS6gPx7pyWWGyOmYpZJUX49aGnYfEFepZihcd42m23Iko4GfWUjCIQeMzpc9Qm8/nF+13CCUOJQSEInQVb4Dp4EHFVpZgmt2OFeWNiUZ07FntS8vGRfSLeS16IKL9TiEEhCIUolL2ESe561B49ojbNhslx8bhw/CLEj8iFI92CI/Y67G3cL0ThbhwVcw+17b3i97RHkBKV0ikkjZxXmGXPar/fm4OYRBtGzRmuNpnfI4Z6K4ubUL6nTGylKD/igbPNKryqJ6gNh0QmsSWYSkUv4YvIjD6ErBwjkkYLJ5dzZ8E/bCY8njQ1bOwuE0PGx5xMvEJoSm9tubXuqW03zSDiHZplb6GYT2iwGWESXh9uvxcVwgFk1LRoWPY3QshW2NtcyDvv/PbneEACJEACJNCZAAVgZx480xEBKfycO4W3qwji7BFhUGRyiblge0WMO7mCR6K5GMvOqIJ12VI8tPcfeG3t1YhtMmLh3jQRzDkwb84jnCq2xE7BqMpa/GTLS4j2tKly7KedhrTbvw+78B4/lWQQHui2kSPVlrh8uSrCK2JKOrduhUP0PEph6BQrvZhampGxe7NwD9mMc2Uu8ZwrbxTKcsZge0oe3jfNwSExLJ0l5hBqojDdVSWWnGvEgY2fAnITyWA2Y2b+aFw+8Uokjs1DU6oRB1yyl3A39tTswaHGQ6q38OOjH0NuWooVwne4dTgWGBfgjOwzMDllMiymkw+9as+brSYMG52oNlw0QYVikSFpyg/Wo0wKwoN1qK0xCueSYWrb27oAEKO/1h0tyLDsQ5Z1DTJjy5AxIgH2vKnA3FlA9kz4jPHtYlD2ErqEMPTI3kIxvO8uEUJRbFoqUZ7MPhGsezdyPwr0/mU0OBB/LEaolo97EiABEiCB4wQoAI+z4JFOCMh4b3IOWePbRZDLt8nkEQuq7RPCr1AsCWY11uOc8XuRtWIJnih9F/9481oYhCfwbOXgEQ+DCBUi/H+xO3ac6AH04vsb3kCcSzoOiFU2Jk5E2ve/j5izz+rTcKl6+CR/TImJkD2KcpPJ73aLeuzrMGxcAI9YFcR6aD/y5CbyyAFmQ1oanGMmojhrND6PnYy1nni01pS1C8Ks1jLYPa2oPLBHbbJsmWypmThz/ERcNXkhUmaORGVUM/bW7lWiUArD/XX70exuxm73buzetht/2vYnRJujMTN9JuZkzcGczDnK4cQk1jHubZLxDBNFD53cxs8dph5rc7hRcagRZQfqUL63HOXFIm6fJ0b0zs5QG4SWM5QL55KNxciyvCWGjh9GlvDmjcsfAVuOEISzhAjPmga/sM1Tc2z4WAhCl+gxbNx7EAfa4pCd5cRjZW/iHLFKi0xpPiOipwlByUQCJEACJNAlAQrALrHwYqgSaBU9S41vH4brsPRAkCGNfdgvhgEPiuFev4hjNz17CyZ+ZT5ectbiiXXfgaO1GROL4jC9MBkmsX6uTIeixfCj044bNn+EZGeDumYVvXVp3/se4hZfMODCT72giz8GiwXRkyepDddcrXK4xeozTjF/0LF5sxo2ls4lfhHnMqpqnfAYlhvwVZsN5gkT0TBqAg6kn4HPooZhfWUdfOWHMEzNJSxHslsEca4ux4GP5PZ+4O1RMYjPG4N5kybjG9NWIEXMZzzUfBhvbH4DpZZSfF7xOera6vBxqeglFJtMcZY4zMqYhdmZs3F61umQXsjdxicMvOULf23CmSN3UoragNEB5xIh3MsPNqD8QC3K9tegqUEs7SidS8S2w7lErIcnnEsOSeeSvUIQ/kn0FO5H2jAxJ3D4NFhEDyGmzYJ3ll2s0FIt5iDGIXdGPFwHKmEWaytbRJib/LPmCc/i3gvXLxjNCyRAAiQQ5gQoAMO8gcOlem3FjUr4tR2oV1Xy/f/2rgO+qur+f5OXHbIT9hREiGwQBBUXKopiHWidxaJ1oWKtWq2to+5W66it1tU/QquoddRirQsXIluW7DBCEhJCQshe9//7nnCf72Xxkrwk9yW/8/ncvHvPPffcc77n5r3v/U0R46VJaJfNYvInZn6Se3cVxs0Yho+jB+Hulb9BTnEOjsiIxoQtfRFuhHsWssOSUVSZjAvWLkH3olzTT0jPHkiZfRPipp9j1KjtjVdodyEwU6ciVjaWasluUyoZa4rFhtCojUV1TFVyxepViJKNMi5uYf36wTViJPaJreKmpDOxvCxUyNVmBOXsQPeSLHSTvMYhpUUiMVuN1dz+JX0HuRCU3BthKT1w1RkX47HzHkZa0S4sy1qG77K+w4qsFcbzeFH6InBjiQ+PN2SQ0kFuA+IGNJkwG+cSiZmYItvwk3qbfo1zyfYDyJQta8t+5Owu+tG5RBxMWFwSzLrrRolJGLZCiOF8RLgKhTA+JIS0ChtSNqLPNzXSv67i/Ruv6l+Dmf5RBBQBRaAhBJQANoSM1jsCgfIMUfOJxM92BLCE+O0uK8MG8dotE+LXK3ILJp2ViO8HDccVq5/FjgIhPPsicL4Qv9i8Gm/Rg64uyA7uhbM3iddvwXdmXi4JDpx83XWIv/giBEtuaacW5rCNOuYYs3GMVH/XOJeQEK4UYrhKJF/bRCK6E5CNoatptThOMpVEjhwJ1/AR2Nv3BKyRdHMb08SRZPsmhJAUinNJdJXYTeaI/aRsH29Ygv8KIbQkbE3P1BGYfdx1OGLykdicvwVLs5YaQrhy70rkl+Xj450fm43jSY5MrpEOdp9gCGHvmN5NJoTsxziXjOmKgbKxVFaIc8lOcS6hlJCkUNTHJUUeziWmVc2fXr0r8FLmhxifcSj7h6icoyfVkEaPZrqrCCgCioAi4IGAEkAPMHTXOQhUZBeLV+9OlBxK80Xl7V6J4bdG8vGWVEs8vpBMnHpCBbInpWLOmr9i7ZdrEXcwFFM390T3vTVODGVBYdgjMfVO2b4Rqbnfm8kFCzFKmjULiaJyDY4+lOnDOdM+7EhoY8fQJtziz68JLlMl4WaMc8khb+OSNWskl+9BFH39NSAb5WITxLlk8lFHIWr0KLgmTkF6j4FYnVeCtPXrkb91PRILdiGmSuwp925DtmwffP4OKlwR4qU7CH2OHonbT/g1+pzc08QgJCFcmrkUq3NWY1/JPnyY9qHZOPge0T3c6mJKCJn7uDklJNTDuUQ6IPE9kEPnkkNSQvncL3aAovdHwgk9kbdmDyLLe0oA72r0GzEKwZLZR4sioAgoAopAwwgoAWwYGz3TDghUShiXAonjV7xyr/lx5xByUYRVBWGSvSMUEcEFOGHkXoScPQqPbXwV33z2BCLKgnH8VpEe7ZIYcUIIqiQQSLp4th6zezfOy/rIzCJICEHiFVcg6epZcMU1PUhyO0Dh8y05ny6TJ5uNF1mVlZKWbbORDjL8TInYE1ZkZKBM7Am5Af8E6dFxKcmYMnIU9vfqiqSf3Y5lxUFYL20Ltm1A/IGdCK8S/fruddgr27v/nY/isFgE9x6M/sNH4beTH0L3HvFYk7PGSAhJCNfsW4PMoky8v+19s3EsfWP6ugkh7QgpMWxOMc4lkp84XrYhE3uYLooPlmHVsu/xfcxi8equUf8mFZYg6ZLTm3MLvUYRUAQUgU6FgBLATrXczp1sVUEZCj7bjaJlWcLgKO8DDrqKsDwvFAXVYULpKjB6wHZ0nzEWL6R/i4Wf/RmuqiCMSkvAyLR4BEl4EJaMsB4YnLkf56QfcnwQR4uEGeLwcN21CO1ao140DTvwH4aEoTczN1x2mZlpIP/U5gAAMRxJREFUhQSuNmRQpITFYjtYuuEHSUm3D4WffAIqwA/O/wdSu3TBWPG4jRaVc3nqT/FtXhk2StiaorQNiCvYg6hycbzZvlxUssvx1nsv4YAElg7tOwQDR43GQ8dfjsQECamTvbqGEIqUcH3ueuw6uMtsb29524xjYNxANyEc120c4iPim70S4VEyzzgXPtn1CVIPqX+7FZSgy8knNbtPvVARUAQUgc6CgBLAzrLSDp1nVWG5pGxLR+GSTDH8qiFxpeElWJ4racgqa2zzjkzZgqNmHI1/lORjwTfXSiaOSgxKj8ax27ojpJgBm6VtSKIEdy7HrLRvJMCLEEhRecZNn47k2TcirHeNo4FDIWiTYYV26ybOJWeIc8kZ5n7VpaUoFfVv0fIV2Pv55wjdulXS5hWi6IsvzcZGqRLcetzo0cb+sGzIpfg2twhb165B6a6NiCnKRlyJJDTelIPMTV/hH28EIy+mFyL6DcGRo8fi8YlXIy7GAu0GjcpYCCFD0Gw7sM1sr296XdYpCEdJzmTjYSw2hPQ27hLWpUl4ZIktY3rWdhxXWLPGffoOQEhCQpP60MaKgCKgCHRGBJQAdsZVd8CcqyUu38GvhPh9nSGpwGrSeVVGlWH1vhLsya+xzesevRtjz+2JD2Or8du1t6FYbAB75ETgxK39EJFH4leJQlc0Ig+G4JJtKxFSXdNPzGmnIeWWmxE+aJADZurMIdBGLmrsWIRLOrtd48ZiyPDhqNiyBcXLlsm2HMXLl6NabAuLvvnGbJzFMLnmGGkfdcw4FA86CktzC7B9/VpU7t4kavgCJB/cDazbLWniPsbc+WHIlRzI0QNSMXTsKbh40g2Ija7A8r3L3SpjksGNks6O22sbXoNLnFBSk1KNMwntB0d1lXuF1jh2NITi8oLl6JNdo/6NLypFt/Nrgm431F7rFQFFQBFQBGoQUAKoT0KbIlBdVoXCxXtw8Is9sCSMC4vVpVLUhbnYlp8oRxKrLnQfxp8SjmWDQ/CzdQ9iv8SDiy8IxWnbBiAhk1JCsXETBw+rLBrTt65BRIVJBSGen5MkiPMcRAqZ0dI0BBgzL/JoiUkoW9LMmSaVXRkJ4VISQtmEEFbt34/iJUvMxt6Hi/f0+BEjEHnMBBRITMEVQgh3/bAeVRlbES6BqXvmbwNWbUPGqn/jFfHEzo3rh9iBqRh2zLm4/ARZp0hR8XsQQqqL1+5ba7aX172MkOAQjEge4Q5KPSJlBMJd4V4TW35guaj8awggw7/EnHqK13k9UAQUAUVAEagfASWA9eOitX5GgCm8Cr/LFHXvblE1Vpjeg2Ig6sB0rE3vJseJCA8uEhs0kQBOCsecDc8jfXk6IktdOC2tL3ruEIWhpHmjg0dZVSxO3/YDYktqsncw3InJ3nHsBD+PuvN2x1R2EeI1zI0e0yb8jOQyrpEQLhNbzWXGhpDEkBvLCLG3nDBsGCLGTUJer574fv8B7N60QTyLd4iHcSFi9q8HZMtY9ib+GpqI/fH9kXjkMAwfewl+NvlXCI8owLK9y4yHMdXGdChZmb3SbM9//7whf6NSRrkJYWxoLNIlkPXk3D7m/r1ik0w8RHOgfxQBRUARUAQaRUAJYKPw6MmWImBVVYud2V4c/GwXqg7USOqCY4ORUZKGpbu7ibVeN6F0lRh2ZDaqz+qGh7fOxcalGxFSGYSJu3rgyK2RCJbMDixFViwm70hDt4It5jh88GCkzJljjP7pJaql9RAw4WdEpU61esIll7jjEbpVxkIIK5nFxISiWWUGMlKkihNEohg2ZiL2deuKdXkSwkXS1QXl7kFyxX4k5+yXOIQrkb54Hp6J6IY8IYRJg4dh9JhZmDn5TiGE+W77QRJChpxhgGpuLMxI0idHng+xJYwqq0Cvk9T71wCjfxQBRUAR8AEBJYA+gKRNmoGAxG0rXpWNws/SRXUo4USkBMe4kGelYfHueMneUZMn9ojumUg+txeez/4cS5cuRZBoeIdlJmLU5iSElJAwVqFY1MLj0jPRP1dUilJC+/ZFyk03IXbaWaCkSkvbI+AZjzDhoosMIaxIT/9RZSyEsGLPHpRKTEJudOcZQ6nikCEIHj0BOcmJ+CEvH9nbNsJVmIdeks+4V5Y4AmV9i51fhWFxpMT2E0LYbchwjB5xPa4+8Tdwhe9zE0JmK2FQajv8S7cDRYidcmrbA6F3VAQUAUUgQBFQAhigC+fEYVvVlqQpK0NJWj6S/leKAwVbzTCDo10oidyNxWnBKKiqieHWNSYHA6d3w7yqFfh4zaMm5l+/3C6YtKkXwg+QMJZLZt8IHCU5bodmbhMZDxAiYVySb7gB8RecD+bR1eIcBEgIw/r0MRvXh4WxBykhpLqYnxU7d0n4GVEJy0Y3n3G8ZvCRqB4xGjmJcdicl4fcHZsRXlaCI4p3ANwyFiHti2h8Gdkb+UIIe6QOxzGpN+OaExOQVbgF3334gPRUDfEHR4TYI2pRBBQBRUAR8A0BJYC+4aStaiFQVVSBiqwiUfsVoWJvsdmvyCp2e/TywQqKlNRiSXuxeHM+9pb3Mz10CTuAo0/rgoVJm/DQ1kck5F+VZKEIw8lbj0BMVom0KZWIfyHonVeK0bvXI1giujDQcdIvfoGEyy7VDA+11sHJh6E9eyLu3HPNxnEyFqHxMD5ECMu3b0f5ps0SSmYzEuQ8LThDBw1E2bBUZMfFYLs4neTvSUMXyVAytHATwC39I2z8PAEfCyEsDe2CY8UuNFRMBPodd4JKgwmyFkVAEVAEfERACaCPQHXWZpbkZK3ILjlE8ITskfAJ0as+WGPPVwcXVxBCUiJxsHortmZkY3vm0dIkDqHBpRg2wcLyobtx/ebXUCpq4SjJ53vyjiFITisRCV+JyHGCkFBoYfyOLQgV28FgiUOXeNVVss2ES4IUawlsBBiLMO7saWbjTCr37TMOJLanMb2OKySvcbBsTCDHzSXexUWpRyE7Nho79ueicO8eJFXkmY19sHQtKFb1bw0U+lcRUAQUAZ8RUALoM1Qdu6FR3wopqyF4P0r1KiUuH+Mq11dcCeEQZ064oqS9lYvi4n0oOJiHnKwgbM4fI4QuRShdFYYMLcLOY/NxR5pkj/jhAEIrgnDynkHovbEKrmpK/YCI0mAcm5Ym2SYqESThRRKuuFSkftcgJJGhYbR0RARCkpMlMPVUs3F+laICNl7FRkK4HGUbN6Jqxw5EyNZXznOr7tsHB4cciZzoCOzKykB54QH0LSxF9MSJ7EKLIqAIKAKKgI8IKAH0EaiO1IzZN2xJnk34KkWNy1At9ZXgcFHnRpehylWA0ooDKCgpQF5xOfIlG0fBzq4oriZJs7cfe+jbMx+lUyw8kvE8sjZmGQeP8Xv7YdC6MIQdit3nqhBP0bSdiC8pEybpQrykbUu+4XqE9ujxY0e61ykQYAaPWAnizY2lSgJRF69Y6Q49Q/vB4F27EcdNzg+Sje8mMVOmIDhcHlItioAioAgoAj4joATQZ6gCr2G1ZNggsbNJnm2rZ8fhqzMjccGtDi1GmSUpwSqKkCdBm3MqopArAZrF0kqaM29r/blbw1zliIkuRVx8ELokhmNj7B68HPVvbNsinrvyKz00rzuGrU1AdBFj94n6uMolNn4Z6H6g0Dh4xE6bJp69sxEmKj8tigARoO1nzCknm43HVZKqrmTlIUIoAapLJJVdUGUlYn/yE57WoggoAoqAItAEBJQANgEspzal+rYy17bTqyF8dM6oZPiVetS3llRWis1dcVUJDkiIvdyKCOyvCkOhEQBSksItyWu6wUFViIksQaykZYtNCkNstziEd49HmVjvF0TuR3ZVLrIkcO+aoizx5tyMTXlisH8A6FUYj3HreiNh/0HpT6SMVjBSs/ahn3j3MoBLlxNPNNk7GB5EiyLQGAK0A+0yebLZ2K6i4CDWLl5s4kA2dp2eUwQUAUVAEaiLgBLAupg4tobZGKoP1njfuqV6dMrILmJ2tHpLhVUhoVeqkV8ZggIhigVC+A5WSbgWI9HzDqUSFSbG9DEiUUkMQZfkKASJM0dpQlANwQvKxI6STCF5WSZDAz8L0gqAtHpvi7iSCBy7eTB67BEWiINC/IIwQFKFHZm5D6HV1YgaNw4pv7wVUWPG1N+B1ioCh0EgODoKVkryYVrpaUVAEVAEFIH6EFACWB8qDqhjztzQXAmCvDQLVXv2ozzjACr2VUmYlfoDH1cKOSSxI8GrIXrctyTgMifDa0SJK2ra2C5l6BsXhAiR4lUnuoTgAflR+chx7cHOsgxD8EjusouzJdWXdJbTOBixwTHoXp2ChIp4RJZEIaw4VDI9VCBl1x6EWCR/4s1ZUIKh6dmIrKhERGpqTdq2448DY8dpUQQUAUVAEVAEFIG2R0AJYNtjftg7Fq/chNwFmeJW4RIt6naP9sEm4wJVtSR33A4aqZ6FIqmjmrZLZCliEqsRLaZ6EbKVxIv0L6oA2SHpyKjcjb3Fe40EjzZ+opGt2TzuYHaFNIZVBCOuPBwpVUmIr4iRnLyRQu5CEFIsYV6KqxBSWo7w8jKEV9dk6xBFslzK7ccSV1KBYbuzRBpYjrAjjkDKzTcj5ozTlfj9CJHuKQKKgCKgCCgC7YKAEsB2gb3xm+5eugkRVryEVqkheTbZo2SvylWCiJgKMZAXBw/ZimMrkB9dgL0iwUvHDmOHl1OSI1Z+h4z/8uVe3KQwzVpkuUvInAsJZZGIr+yC2PIuchyO8JJghElEltBSkdKVS/8SYPfHwswc3OovVO+6RMoYxv6rqhBdUoLuuflIOViM0J49kPLbmxA3/RwEhejjVj+CWqsIKAKKgCKgCLQtAvqL3LZ4+3S3nheciP++9glyynJhdXUZgpcVsge7sA0ZpekorfIgYyKAo1QuskyI3aFtSFkXRJeFIqY8SshdmImxF1ZajXAJ6lxX6UqiVxOLr87gqoMQKjwyvFJs9ioqJHCzkE8hhxGSeSFc1Lnh5lOkgWLTV7tfKzYWXe++C4k//SmCJa6fFkVAEVAEFAFFQBFwDgJKAJ2zFu6RLCr4Eg8k3ickKxiRxULs8lyIEnLHbYRI7iINwQuXT8l/WiokTVTBDRex4xOXD3eRpsHC+cLEMYQ2eVFlZSIVrDhE5mpIXYQQxTAJr+FqoNvg6Gi4krsiJClZtiTZT6rZl0+XHAdLPLdN5eVImDABwRLbT4sioAgoAoqAIqAIOAsBJYDOWg8zmtD0IFz6cT8hYb4M7hBLE1LH9GkRQuqihXzZEjqSuXAhc5T+UWIXJlttaR3vEiwSu5CkbocIXQ2xCzlE6JixwRA9Ej6pC46IaHRgVaIGxurVjbbRk4qAIqAIKAKKgCLQfggoAWw/7Bu8876theKEIaeFqblEukcyFykq2NpkzpA8IXYRlNaJfWBtYueKj29UUmckeLbUTtW0Da6HnlAEFAFFQBFQBDoaAkoAHbiiU6ODkbluO4LFCSTYUw0rYVNckhvXSOoMcfOU1NVI59ySusQEBIV6x/lz4FR1SIqAIqAIKAKKgCLQDggoAWwH0A93y8Rzz4ErLAQ7d+/GERIoOawr7e3Evk5s64LUpu5w8Ol5RUARUAQUAUVAETgMAkoADwNQe5wOjoxE7Lnnokrs6KJHjYJLSV97LIPeUxFQBBQBRUAR6LAI1J9WosNOVyemCCgCioAioAgoAoqAIqAEUJ8BRUARUAQUAUVAEVAEOhkCSgA72YLrdBUBRUARUAQUAUVAEQhoAjh//nyccsopGD58OGbMmIE1a9Y0uqIffvghpk6datqfc845+OKLL7zaW+J1+/TTT+P444/HiBEjMHPmTOzYscOrjR4oAoqAIqAIKAKKgCIQ6AgELAFcuHAhHnnkEdx444145513MGTIEMyaNQu5ubn1rsnKlStx22234cILL8S7776LU0891Vy7efNmd/sXX3wRr732Gu677z4sWLAAkeKMwT7LJFuGFkVAEVAEFAFFQBFQBDoKAgFLAF999VVcdNFFuOCCCzBo0CDcf//9iJAMFW+//Xa9azN37lyccMIJuPrqqzFw4EDMmTMHqampmDdvnmlP6R/bXH/99ZgyZYohlI8//jiys7PxySef1NunVioCioAioAgoAoqAIhCICARkGJhySXW2fv16XHvttW7Mg4ODMWnSJKxatcpd57mzWkKqUKXrWajqtcldeno6cnJyTB92m5iYGIwcOdL0OW3aNLu6zidTn5n0Z3XONL/C7s/+bH5PbX+lPWb7s+1H0Lp3tOdlf7bu3Vq3d3sO9mfr3s2/vdtjtj/923v792bPy/5s/xF1zBHY+NqfHXOWgTkre03sT3/OojX69Of42qKvgCSAeXl5hnAlSXBkz8Lj7du3e1a59/ft24dkyWnrWdie9Swkfyz19Wm3MQ3q+eOpRq7ndIuq1q5d26Lr2/PiQB67L7h1pPkF8lwCeeyd7TnzZb7t1aajP0fthas/7qtr4w8U6/YRkASw7jTat2bw4MGIiory6yD4dsKHng4ugRYIOpDH7ssidqT5BfJcAnnsne0582W+7dWmoz9H7YWrP+7bmmtTXFyM1hTe+GP+rd1HQBLABEmJRlJU2+GDx7WlfDaArK8tyfNsn5KSYpqyrqukXrMLj+lg0ljhWFqLpLVm343NyR/nAnnsvsy/I80vkOcSyGPvbM+ZL/NtrzYd/TlqL1z9cd/WWBv22dlLQDqBhIWF4eijj8a3337rXr/q6mpzPHr0aHed584oSam2ZMkSzyosXrwYrGfp3bs3SAI9+ywsLMT333+Phvr06kwPFAFFQBFQBBQBRUARCBAEApIAEturrrrKhGphCJht27aZ0C0lJSU4//zzDfR33HEHnnjiCfcyXHnllfjqq6/wyiuvmPbPPvss1q1bh8svv9y0CQoKAtv89a9/xaeffopNmzaBfVAaSK9gLYqAIqAIKAKKgCKgCHQUBAJSBUzwzzrrLOzfvx/PPPOMceAYOnQoXnrpJbcKODMzE/QMtsuYMWPwxz/+EU899RSefPJJ9O/fH8899xxov2eXa665BiSRv/vd71BQUICxY8eaPsPDw+0m+qkIKAKKgCKgCCgCikDAIxCwBJDIU3pnS/BqrwQDOtcuZ555Jrg1VCgFvOWWW8zWUBvPeqqdWUga/V1o/MpCQ9VAs1UI5LH7so4daX6BPJdAHntne858mW97tenoz1F74eqP+7bm2ti/2/bvuD/GG2h9BEkAZCvQBu2U8dJBRFPFOWU1dByKgCKgCCgCikDTEKA2sHb4t6b1ELitlQC2YO0qKytx4MABUEXsqW5uQZd6qSKgCCgCioAioAi0MgKU/DHNa1xcHEJCAloZ2myklAA2Gzq9UBFQBBQBRUARUAQUgcBE4EcvicAcv45aEVAEFAFFQBFQBBQBRaCJCCgBbCJg2lwRUAQUAUVAEVAEFIFAR0AJYKCvoI5fEVAEFAFFQBFQBBSBJiKgBLCJgGlzRUARUAQUAUVAEVAEAh2Bzun6EgCrVlFRgdDQ0AAYKUwgbqbMs3MtM+/yyJEjTWq9gJhAJx4kn7M9e/aYMAgxMTGdGAmdemdCICcnx6T51O8s5626rk3brYkSwLbDut47LVy40KSaY35jlnnz5uHll19GVlYWYmNjccUVV2D27Nn1XtvelQxSzawpnAODaNOdnoWhcRhectq0aXjggQcQGRnZ3kP1+/0DiaDbk3/xxRfN8xQREQEGWGVmHAZM5z7DGE2fPt2sl5NfPNasWYNVq1Z5vWwwV/eIESPsaXaoz0B8zpy8AJ35O8vJ68Kx6dq0wwoxELSW9kNgyJAhlryFmgG89dZb1vDhw62nn37aWrRokfWXv/zFGjVqlLVgwYL2G2Ajd7777rut008/3fryyy8tiYnobsl9ybtszv3mN79x1wfizn/+8x9LYkW5hy6EyTrppJMsrtv48eMtySntPuf0Hc9nTdImWsccc4zFZ27Lli3We++9Z02cONH629/+5shp8H/kkksusY466ijr5JNPti688EKzcZ91PGf/HzlyAocZVEd6zg4z1XY93Rm+s9oV4BbcXNemBeA181JKarS0IwL88bJ/uPijJlIar9HMnz/f+slPfuJV55SDcePGWStWrGhwOMuXL7fYJpCLJ2kKNIJeG3fPZ43P1Ouvv+7VhCRQpLZedU45uOmmm6yLL77Y2rZtW50hsY7n2CZQS0d6zpy8Bp3hO8vJ+Dc2Nl2bxtBpnXOqAm4HqWvtW1J9yrJ7924cf/zxXqd5TFWdEwsjqTemLuS5QM+zKP92buiFMOHmm2/G1VdfbepOPPFEo/b+xz/+gRkzZrjbOXnHftYyMjJA1aln4XF6erpnlWP2RaIMeRnCEUccUWdMrLvnnnuMervOyQCp6GjPmVNh7wzfWU7F/nDj0rU5HEL+P69ewP7HtMk9igoVn376qUkpZyeotjthqhr7R9uuc8qnqEKNDeCGDRvqDIl19913H0RFV+dcoFXY+DdE0Hfu3BkwUxJzAsydO9cQd9pqepaioiLYtqie9U7Y57gKCwsbHIqTx97goGud6EjPWa2pOeaws3xnOQbwJgxE16YJYPmpqUoA/QRkS7r59a9/7b58yZIlXpKZ1atXo2/fvu7zTtqhA8htt92G888/30jCEhMTzfD279+PgoICI81km0AvJOj0kGXO50Ai6LVx79mzJ0gAWUio1q9fD7EDdDf77rvvMGDAAPexk3bOOuss8P/krrvugtgqokuXLmZ4JIXffvstHnnkEZx99tlOGnKTx9JRnrMmT7wNL+gs31ltCKnfbqVr4zcofe5IcwH7DFX7NPz8889NouoTTjihfQbgw13FBgskqmLLaFozDIw4r2DgwIE+XO3sJmKb5TXAW265Bddff7277s033wRVwO+88467LlB3uIYkhqmpqY6bQnl5OR566CG8/fbbxmvZNj2gl6zL5YLYz0KMyB0rwTwcoJ3pOTscFm1xviN/Z7UFfq15D12b1kTXu28lgN546JEi0CQEAoGgN2lCDm9Mid+6deu8XjaGDRvmlgg6fPjNHp4+Z82GTi9UBBSBBhBQAtgAMG1dTQNYxmKrXWgcnpmZCarvnFg2btxofpAnTJiAPn36QEKKGGN9zue0006DkyWXTsSztcdESdonn3xSR2JLB5BTTz01YCVorY2b9t9xEMjLy8OmTZtAqWt8fDxosiIe/uD/xplnntkhNBcdYbX420ezlF27dpmkAnSItCX/HWF+TpiDEsB2XgVKNCRWHviGT7smCWdhAj9TrcVCtSpJ1A8//NDOI617+//973+YM2eOsY/jl+dzzz0HqkgpkSGZpW3WY489hnPOOafuxQFec+WVVxq7s169egXMTOisMmvWLGRnZ5tMLUlJSWbsubm5JitC9+7dwWDR/fr1c9ycGBid6mnbzlRCDOGf//yn++Xosssu87KdddwEfBxQoL4I+ji9dm/GQOI///nPjUMRA+2/8sor5jsrJCTERCzg/wZNOo4++uh2H2tnG8A111yDJ5980vye5Ofn4xe/+AW4XgkJCeBx//79jXDB/g7obPi0xnzripxa4y7aZ4MIPPXUU6AU7fHHH8ett94KicWGG264wbyN2hd5hoiw65zw+fzzz0Nir5m3tAcffNB8kc6cOROvvvqqyWbyq1/9ynw6YazNHQO9s+vbSEAWLVrkPtfc/tvyOnplDx48GIsXLzYZQPjscWM2ENYNGjTIZAJpyzH5ei8+Z0w3yEIJJjPkMHPAmDFjUFpaao75EhWohS+CfHmi7eykSZMgweCNraM9H5J0Smi1tAyBP/3pT5g6dSr4/3vttdfixhtvNE5FH330ET7++GPQ2UgC8LfsJnp1sxBgqCcKElj4vUTPfq4JBQn832ZGKf5faPEjAkIutLQjAuL6bonnr3sE8kVvMhzIW6rJQCF5EU3WCXcDB+0wS4mERjEjEsmFJW/NlpBZ9whFdG8ymbgrAnCHwZMZpJefDW08HwhF0qVZovpqcKhcO7ZxYuGzxueJRWIuWi+88ILXMJmhxakB070G2sDB73//e5M558MPPzSZf5jhRCQg7iw0/B7g86elZQgw+83WrVtNJ0I2zP+2vFi4OxX7Uks0Lu5j3Wk7BPh820kRzjjjDEte9LxuLi+p1imnnOJVpwctQ0AlgH4k083pivYnnvZ9FG9Tgsa3H4rEa4cdac49Wuua6OhoI5pn/wz7Iing3Meso60N2wRyod3J5MmT8c033xhJLaW13Kii/+CDD8y+E9Xz9WHOUDZ79uyp75Sp4zm2cWIh3vyfYGGwaq6JZ+FxWlqaZ1VA7VPKzLzZlE4xqDht0vjdQI9zWypixwkMqIk5bLD0Gmc4JxbakzEvNlWMdrHVjfaxfrYtAvYzzt+T2uHPeEwVvRb/IaAE0H9YNqunHj16YPv27V7X0hbw5ZdfBoNAz5492+uckw4Yj40/Wu+//z7uvPNOHHfcccaGg278nNMf/vAHo6Jz0pibOhbJmWtURBdccIFRQzT1eie1J7HgOv397383xJX2pdxIaFnHGHu0QXViYbxCyZdrhjZ06FAsXbrUa5g0Fu/WrZtXXSAdBPKLYCDhTDtXBnS3C1XCKSkp9iFE0upFCN0ndKdNEGCsT/7mUZhQOysRv6tot6nFfwhoIGj/YdmsnihhYmwzphXzLJSckXzQYNmphWTijjvuwL333muInm1TJvlkTfYSvrExdlugF9o10suZNo20RSFRCsRCGzPa0fC5evTRR90ZZkSJAMZuZIo7Sp2dWIj9pZdeaiQAY8eOBX+4165dazw2KflbuHAh7r//ficO3acx2S+C9KS3i/0iSMcdJ78I2uMNhE9+N5Fs24XZJzzLZ599BjGD8KzS/TZC4LzzznPfifautbVftNOsHS/TfYHuNAsB9QJuFmz+u4jpuCjWPvLII7065Y8yxeE0DmdatfHjx3udd/IB37D5z8scrfSu6yiFzgYPP/ywcXrhHCn5pONEIBaOn2/ULCR/nsTDqfNhOAi+ZND5hg4gLHy+6HVO8jplyhSnDv2w4xIbQCN9euaZZ+q05XcAXwRJeAPF3KDOJAKkgt9bNDegx7kWZyHA/3muja3Cd9boAnM0SgAdum78UaNHcCBm0wjksfvyOFBKwJR99CK0Q6n4cp228Q8CfDmiVyxDptBmqyPEBmvoRdBGLBBfBO2x66ci4CsCFIYwvNOKFSvMCxHDifHllBJBphwlAdTiPwQ6jnjGf5i0aU/MYVpfqaqqwt/+9jcTqJTnnah2DOSx14e5L3V8C6VzC1Wp4rEJqpQ8jch96aO92jD3L21obGnfu+++i9dff90dS+/yyy8382mv8fl6X0rGKbXsSIWSPwYhbqhQHRxIWoCG5uGE+nnz5pn4cjS74f8v/w/4XcsXitNPPx0333xzh9JcOAFzX8ZACfdVV11lnD/onMO4pczvTccdhkmjqRTNV/i/oMU/CKgE0D84NrsX2jRwq+19uWzZMqPaItHgD97cuXObfY/WujCQx+4rJowLxsCwzBjAjCwMOEwPtf4SlJRqVL6RvvHGG25S5Wu/7dFu+vTpoJE148wxhzFjN9IxhFJm2tGxjkHJmVfXaaWjkNeGcOX/Ev/PaTdLhyPaQ3k6JzR0ndY3DQHG+COJoO31ypUrwYDudLibKXa+lDbRGeqSSy4xJLBpPWvrliJA3OlIaNu7UgM2f/58LFiwAJSQ/+xnP8O4ceNwzz33tPRWer2NQMuiyOjVLUWA8cwY24gxjjxLamqqJWnVPKsctx/IY/cVTM/YVLfddpslXrKWEEBzuajlLPnhsH75y1/62l27tmOMP/GsM2NgzDwhrl7jEZtGSwivV51TDiSbjCWheMxw5AfBxCtk7Dwh55Y4Gpl4k0JgnTLcJo+Dzxm/A4SUW+JwZGJqXnfddZaYG1iiDWhyf3pB/QiInaglzgTmpNhTWuJRbgnRcDeW7EaWpLB0H+tO2yHA7yc71ifvyueesWUZA5Pl66+/toS4m3394x8EVAVsM+F2+mS6m2OPPRa33347hAhCyETA2DQF8tibs9yrV682nqa2tJae2sxQwTULhEK1CtXXTF+3d+/eOt6OI0eOrBN6wSnzojrITlFHiSwllRdddJF7eMOHDwcz0zhReuke5GF2mKWFoZXoWc8MCFR5MVMF7Uxp/8TNxuAwXenpBhCgjRltlFkodaXUj2GF7CIv3hprzgajjT/5nHN9bBMVOqkxHIyt8uWzT0mgFv8hoHEA/Ydls3ti2IF//etfJjwB1T+bN292h+hodqdtdGEgj91XiKiaY2FcxtpqOcae8wwr4Wuf7dGOwZJpYM3CuHr//e9/vYZBm8bawVe9GrTjgU1eOYRAI69NhY1OLTQ9oGqSae9IdP/973+bINFN7UvbeyNA21HJBGIqd+zYYdLt2ces5L7mmvXGrK2O6OjBdJVffvmlcbJj6Cd+T/F/n4VmKoEc67OtcGzKfVQC2BS0WrEtpUmPPfaYCXZLQ1g6gQRKCeSx+4IxbU8YboSemPwSoqTGLhkZGW5HHbvOqZ/8QqWdDZ09KAVhxhkGVLZtACnhfO655xw5fJu8Mq6kTV4pwbGLk8mrPcbmfDJLEKXMtItivmYtLUNATAlMMHSSDeaYZfggOhjk5+ebl25KkSUNWctuolc3C4E5c+YYyT6z3/D3j3mxmUzALnwRDxRtiz1mp3+qE4gDVygrKwuSk9IY60dFRTlwhA0PKZDHXt+s/vznP3tVU00quULddSTtlEg9+eST7jon79CBhR6PDGhNJxZ6Pnbt2hWjR482hvBUpTqxEGOSVxIikldKMsU+qA55rR1Q3YlzqW9MNP+gyjdQPMrrm0Mg1PF55/PPlx0+8zRjYRBxEg3GAOQ6/Pa3v0Wgfe8GAva+jpGaFqp+KVjQ0roIKAFsXXy1d0VAEfATAoFKXv00fe1GEVAEFAG/IqAE0K9wameKgCKgCCgCioAioAg4HwF1AnH+GukIFQFFQBFQBBQBRUAR8CsCSgD9Cqd2pggoAoqAIqAIKAKKgPMRUALo/DXSESoCioAioAgoAoqAIuBXBJQA+hVO7UwR6LwISJYRSEYLSIaFBkGglyXTbbVHYRq8G264oT1u3ex7OmnMvqxvsyeqFyoCikCbI6AEsM0h1xsqAjDBo++9916cdNJJJqwJc2DOmjULK1asaFN4SNgYbFiLIuCJgJOIp+e4dF8RUAT8h4AGgvYfltqTIuAzAgzuW1FRgUcffdSkPsrNzTWBaRmQ1mmlvLwcYWFhThuWjkcRUAQUAUWgBQioBLAF4OmlikBzEGA8u+XLl4OZOZgHmrl5mVLv2muvBTMU2IXSOea9ZbYCnue52unbMjMzccstt2DcuHEYP348GEWfqjrP8tZbb2HatGlG0ijJ1PHAAw+Y01THsjDfLO9lHz/77LM499xz8eabb5o63puFKZoYjJn3mjBhghmvJG8355ryp6ioyET0Z6R/BtWeP3++1+XMUMKMDTzPwM5MD8Vr7MK0iRzDV199hTPPPNME9KX0lHlE7cJMAo888oh7rMz2IOnT7dN1PnmOa+GJLzEgXnbhmjEINQMGszQ2TmaNIW5ffPGFfbn5ZI5fBiC2+/Bl/Tw7YCDjF154wb0u06dP9xrzd999Z9aSWS6YO5iBy3/6059i+/btnt3gL3/5i8k7zLEwr/If//hHs+ZsxPV/55138Omnn5q++GywX7swgPgVV1xh+ub9V61aZZ/y6dOWLjLVHfHls3T//febFyK7Az6LHCPzInOMJ598shkP0y7yGWcdn5G1a9fal+inIqAINBEBJYBNBEybKwItRYBZBrhR9UrpWmPl6aefNqmp3nvvPfODx1RI27ZtM5dQgkjiw4j5JFHMjsF+SRjtfkkgSfjsfLL8UbXz/ZIYspAoff3117CPWUdi99FHH4GZUN59911WGdLCNIXMWEE7PqZmInkkKWlK4Q8/07iRZDATA9O7ffPNN+4u2C9JyQcffGAkpEuWLPFKCcWGpaWleOWVV0war3nz5oFEillZ7MJz7P/hhx82JJpJ5Em+Giq8J1PMMTUeC9sTZ97HxnvZsmVgppTIyEjTprFxMoE91fucg2dhTt8pU6aYPnxZP89ruU/yx/UgYfrPf/5jsrfcfvvt7nHb7f/0pz+BRItr5XK5cPfdd9un8P7774Mpz/gCQjLdo0cPd45oNvr5z39uiDXJOZ8LbiRcdmHffO44jv79++O2224zmRvs8758klDyGfu///s/s8ZcK26ehefGjBlj6vkiQDLIjaST4+ZzfOeddzZK7D37031FQBGohYC8+WpRBBSBNkZAJE2WEA5LCIV18cUXW0888YQlzhNeo5Ccw9bvfvc7r7oZM2ZYYjto6uQH2JK8pZYQMHcbSaNkieTJEumYqRMJiyVp6tzna+/wHkKMvKqfeeYZS9KsWaKW9qqvfcDzvH7Tpk3mlEiGzPGGDRtqN3UfiyTHEvLgPuaO5AC1hLR61XkeSJ5fS6Sb7iohNeY+O3fudNcJCbQmTZrkPhabSuvFF190HwvZsiSfsCXSI3dd7Z25c+daIik11cSEWLO9kGhTN3PmzEaxrD1O9iFSTKu4uNhcf/DgQbPeIhU0x76snxAc95i5tiLRs1auXGmut/8IubPkxcAcClk22EjeYPu0tWjRIlMnZNbUcV5CIN3nuSNSQkuIlbvO8752pb2+CxYssKusLVu2mL63bt3qrjvcDvvmcyDpvtxNb775ZvMc2BU8LwTVPrREumvu89RTT7nrRPJo6nhOiyKgCDQdAbUBrEWI9VARaAsEmHCeEiKqFZmXlOrMl156CQ8++KBR3dlj8JS8sI5qUdvLduPGjUaKQimJZ2EuTUpXaFdItejEiRM9T/u0z5y7iYmJXm137NgBIYf4/vvvkZeX55a8UPomRNCrbWMHnINn4TGlPXYR8mIkXVRbUpVKdS7nRLWpLX3jpy3J5HXMZ8z5sgjRQk5OjlFRmgr5ExISYtS38hVpV9X5pASQ0kiqGSnto0o9OTnZSNcuvPBCo+qkdNUuhxunEE6Ehobis88+Myp4SlQpGRSiaro43PrZ97E/hfAaDCih8yyUJA4dOtSzyqhu7YqUlBSzS3y4rmlpabj00kvt0+aT6mpKWn0pVAnbxe6bmA0cONCuPuznoEGDjGTSbsh+Nm/ebB+aT8/7cB1YPJ+zpKQkU8d52eMwFfpHEVAEfEJACaBPMGkjRcD/CISHh4Pev9yoSqXak/ZXtN3ypYhkCSKpM/ZbtduTvFFF2dxiEy3P66+77jpjr0iSSsJF1e/ZZ5/tZbvl2b45+7RfpC0kbQ1vvfVWxMXFGc9oYkOiY4+LhM6zcK6NkTvPtg3tk3DwflQDkwCKZNIQCxJz2poxQb1NyH0ZJx1nSPSp9qUNJtXBZ511liGjHMPh1q/2ONmehWrgbt26eZ2u7aTjiY/9HDRVVe91A48Dklq7NLdvz/Gxr/rWz7ONfZ/67t3Sdbfnop+KQGdDQG0AO9uK63wdiwClIvaPvD1ISgc9C6VvtqSF5I9SIUpC+vXr57XFxMQYaRMdTOgQ0FDhDyolbIcrlPhRckQDfEoUOQbayTWncA6exXNO69evN0SO9muUDA4YMMDLucPzuob2OXdKhDzvQ/LGvhsrJBl0LqHzg6g2MXbsWCNJoz3lG2+8YSSItLFk8XWcdFSgDR37o4SNx3Y53PrZ7exPYk6il5GR4bXWXHva8flaiGlt54nax3wu/EUYfR2XtlMEFIG2RUAJYNvirXdTBIz69MorrwQdO6gGpFel2I8ZFbCnFzCholcqnTNIvqh+XbNmDS6//HKDIslEQkKCIWVUJbMfGtdTQpeVlWXaMNwMvVXFvg1U4ZK4vPbaa+Yc/9gEkSrTxggdJWPx8fGGCJF0klQyhE1zitiwQezzzJzovMI5Eg8WkhlK+jhGzoeOBq+//nqTb8P+eA862tCJg04T9L4+XKHal84VVKnSuSY4ONiQQkrxqCK2i6/j5DVUX9Lhonfv3l5qaV/Wz74fP6k+pvqXTjt0mKCa317P2g4UntfV3ufzw2eK1/CZoGOQ2HF6SYz5XLCOaniqd7kmWhQBRaBjIeCtR+lYc9PZKAKORIDEguE5aPfGH3FKp7p37w4xzgfVrJ6FBG7hwoWGwFCqJc4ioKSQhepQesAyhMfs2bNNqBSqBimhI1lgOe+884z93N/Fa5ehUEjipk6das7xD70oSeQY8oXX0l6tvkIiRO9PkkuqfSlFuueee0w4kPraN1ZHT+J169bhueeeM+OktI8epyz0Dr7rrrsMeRPnFUO+6PnMcTalkCiR1PI6jv2CCy7AaaedZuwDG+uHBJASUX7ahfuUCnrW+TpOShWp/qUamWp+z+LL+nm25z7V0lTvUw1MNTSlnampqXWem9rXeR7Ti5bkml7TtK1kKB0+J55SQHqNUxVO3CiV5gsESaEvhap0klRfTRl86VPbKAKKgP8RCKLfiP+71R4VAUWgpQjwh5QkiWFDtCgCrYkASTkllX/4wx9adBsSS75gUIraX0LEaFEEFAHnIqASQOeujY5MEVAEFAG/I0BvaqrVGYSZ0lGSNXo001SgpYXBwik9VPLXUiT1ekWg9RFQAtj6GOsdFAFFQBFwDAJUSzNDCYNBUwVMdT69z+3wNC0Z6GWXXdaSy/VaRUARaEMEVAXchmDrrRQBRUARUAQUAUVAEXACAuoF7IRV0DEoAoqAIqAIKAKKgCLQhgj8PyndqKXyLGMZAAAAAElFTkSuQmCC\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(BANDS, mean_spruce_pp_2, label='spruce pp_2')\n",
    "plt.plot(BANDS, mean_spruce_pp_3, label='spruce pp_3')\n",
    "plt.plot(BANDS, mean_spruce_pp_4, label='spruce pp_4')\n",
    "plt.plot(BANDS, mean_birch_pp_2, label='birch pp_2')\n",
    "plt.plot(BANDS, mean_birch_pp_3, label='birch pp_3')\n",
    "plt.plot(BANDS, mean_birch_pp_4, label='birch pp_4')\n",
    "plt.plot(BANDS, mean_pine_pp_3, label='pine pp_3')\n",
    "plt.legend()    \n",
    "plt.xticks(BANDS, labels=BANDS_L, rotation='vertical')\n",
    "plt.xlabel('Spectral band wavelength, nm')\n",
    "plt.title('Mean Normalized Spectral Values for Reference Tree Data')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(len(tree_features)):\n",
    "    plt.plot(BANDS, tree_features[i], c=TREE_COLORS[tree_labels[i]])\n",
    "    \n",
    "plt.xticks(BANDS, labels=BANDS_L, rotation='vertical')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.utils import resample\n",
    "\n",
    "from sklearn.svm import LinearSVC, SVC\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "from sklearn.feature_selection import RFE, RFECV\n",
    "from sklearn.model_selection import cross_val_score, LeaveOneOut, KFold\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix, cohen_kappa_score, f1_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### LOO Cross Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc = RandomForestClassifier(n_estimators=200)\n",
    "knn = KNeighborsClassifier(n_neighbors=3)\n",
    "svm_rbf = SVC(C=1000, kernel='rbf', gamma='scale')\n",
    "svm_linear = SVC(C=1000, kernel='linear', gamma='scale')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "285fb369707f475b9555d0603bec8a77",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1046), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "y_preds, y_trues = [], []\n",
    "\n",
    "loo = LeaveOneOut()\n",
    "\n",
    "for train_index, test_index in tqdm_notebook(loo.split(tree_features), total=len(tree_features)):\n",
    "    X_train, X_test = tree_features[train_index], tree_features[test_index]\n",
    "    y_train, y_test = tree_labels[train_index], tree_labels[test_index]\n",
    "    \n",
    "    clf = svm_rbf\n",
    "    clf.fit(X_train, y_train)\n",
    "    \n",
    "    y_pred = clf.predict(X_test)\n",
    "    \n",
    "    y_preds.append(y_pred[0])\n",
    "    y_trues.append(y_test[0])\n",
    "\n",
    "y_trues = np.array(y_trues)\n",
    "y_preds = np.array(y_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.49898564066555895"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(y_trues, y_preds, average='macro')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6692160611854685"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_trues, y_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support. ' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>');\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option);\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width / mpl.ratio, fig.canvas.height / mpl.ratio);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>');\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"250\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(2.5,2.5))\n",
    "g = sns.heatmap([np.bincount(y_preds[np.where(y_trues==i)], minlength=4) for i in range(4)],\n",
    "                annot=True, cbar=False, fmt='d', cmap='Greens', annot_kws={\"size\":8})\n",
    "g.set_xticklabels(['spruce','birch','fir','pine'], rotation=0, fontsize=8)\n",
    "g.set_yticklabels(['spruce','birch','fir','pine'], rotation=90, fontsize=8)\n",
    "plt.xlabel('Classified as', fontsize=12)\n",
    "plt.ylabel('True class', fontsize=12)\n",
    "plt.title('SVM RBF')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-Fold Cross Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6972448703719514 0.08627023540494913\n"
     ]
    }
   ],
   "source": [
    "clf = RandomForestClassifier(n_estimators=200)\n",
    "cv_scores = cross_val_score(clf, tree_features, tree_labels, scoring='accuracy', cv=10)\n",
    "print(np.mean(cv_scores), np.std(cv_scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6617732434047638 0.0772765181962358\n"
     ]
    }
   ],
   "source": [
    "clf = KNeighborsClassifier(n_neighbors=3)\n",
    "cv_scores = cross_val_score(clf, tree_features, tree_labels, scoring='accuracy', cv=10)\n",
    "print(np.mean(cv_scores), np.std(cv_scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5393435504469988 0.04583384770482453\n"
     ]
    }
   ],
   "source": [
    "clf = GaussianNB(var_smoothing=3.111)\n",
    "cv_scores = cross_val_score(clf, tree_features, tree_labels, scoring='accuracy', cv=40)\n",
    "print(np.mean(cv_scores), np.std(cv_scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6652549679277341 0.038303060005334996\n"
     ]
    }
   ],
   "source": [
    "clf = MLPClassifier(max_iter=3000)\n",
    "cv_scores = cross_val_score(clf, tree_features, tree_labels, scoring='accuracy', cv=3)\n",
    "print(np.mean(cv_scores), np.std(cv_scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7304088761774665 0.08045600700960982\n"
     ]
    }
   ],
   "source": [
    "clf = SVC(C=1000, kernel='rbf', gamma='scale')\n",
    "cv_scores = cross_val_score(clf, tree_features, tree_labels, scoring='f1_micro', cv=10)\n",
    "print(np.mean(cv_scores), np.std(cv_scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7179807336577592 0.09149813909587959\n"
     ]
    }
   ],
   "source": [
    "clf = SVC(C=1000, kernel='linear')\n",
    "cv_scores = cross_val_score(clf, tree_features, tree_labels, scoring='f1_micro', cv=10)\n",
    "print(np.mean(cv_scores), np.std(cv_scores))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### UpSample Firs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "UPSAMPLE = 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_features_upsampled = np.delete(tree_features, np.where(tree_labels==2), axis=0)\n",
    "tree_labels_upsampled = np.delete(tree_labels, np.where(tree_labels==2), axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "fir_upsampled = resample(tree_features[np.where(tree_labels==2)], replace=True, n_samples=UPSAMPLE, random_state=26)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_features_upsampled = np.concatenate((tree_features_upsampled, fir_upsampled))\n",
    "tree_labels_upsampled = np.concatenate((tree_labels_upsampled, np.repeat(2,UPSAMPLE)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7488033488033488 0.0856365043651621\n"
     ]
    }
   ],
   "source": [
    "clf = RandomForestClassifier(n_estimators=200)\n",
    "cv_scores = cross_val_score(clf, tree_features_upsampled, tree_labels_upsampled, scoring='accuracy', cv=10)\n",
    "print(np.mean(cv_scores), np.std(cv_scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 491,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "931cfc22efaf44c0b0b97abc357d1773",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1102), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_preds, y_trues = [], []\n",
    "\n",
    "loo = LeaveOneOut()\n",
    "\n",
    "for train_index, test_index in tqdm_notebook(loo.split(tree_features_upsampled), total=len(tree_features_upsampled)):\n",
    "    X_train, X_test = tree_features_upsampled[train_index], tree_features_upsampled[test_index]\n",
    "    y_train, y_test = tree_labels_upsampled[train_index], tree_labels_upsampled[test_index]\n",
    "    \n",
    "    clf = rfc\n",
    "    clf.fit(X_train, y_train)\n",
    "    \n",
    "    y_pred = clf.predict(X_test)\n",
    "    \n",
    "    y_preds.append(y_pred[0])\n",
    "    y_trues.append(y_test[0])\n",
    "\n",
    "y_trues = np.array(y_trues)\n",
    "y_preds = np.array(y_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 492,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9255141610653322"
      ]
     },
     "execution_count": 492,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1_score(y_trues, y_preds, average='macro')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 493,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9183303085299456"
      ]
     },
     "execution_count": 493,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_trues, y_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 468,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"250\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(2.5,2.5))\n",
    "g = sns.heatmap([np.bincount(y_preds[np.where(y_trues==i)], minlength=4) for i in range(4)],\n",
    "                annot=True, cbar=False, fmt='d', cmap='Greens', annot_kws={\"size\":8})\n",
    "g.set_xticklabels(['spruce','birch','fir','pine'], rotation=0, fontsize=8)\n",
    "g.set_yticklabels(['spruce','birch','fir','pine'], rotation=90, fontsize=8)\n",
    "plt.xlabel('Classified as', fontsize=12)\n",
    "plt.ylabel('True class', fontsize=12)\n",
    "plt.title('Random Forest')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "estimator = RandomForestClassifier(n_estimators=200)\n",
    "selector = RFECV(estimator, step=1, cv=10, scoring='accuracy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "789e6affec9f416ea43742f3f1098a2a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "feature_importances = []\n",
    "\n",
    "for i in tqdm_notebook(range(1)):\n",
    "    estimator = RandomForestClassifier(n_estimators=200)\n",
    "    estimator.fit(tree_features_upsampled, tree_labels_upsampled)\n",
    "    feature_importances.append(estimator.feature_importances_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 5, 3, 1, 6, 7, 2, 0])"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argsort(np.mean(feature_importances, axis=0))[::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "estimator = RandomForestClassifier(n_estimators=1000)\n",
    "selector = RFE(estimator=estimator, n_features_to_select=1, step=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "selector = selector.fit(tree_features_upsampled, tree_labels_upsampled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support. ' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>');\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option);\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width / mpl.ratio, fig.canvas.height / mpl.ratio);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>');\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "g = sns.barplot(BANDS, np.mean(feature_importances, axis=0))\n",
    "g.set_xticklabels(BANDS, rotation=90)\n",
    "g.set_xlabel('wavelength (nm)')\n",
    "g.set_ylabel('importance value')\n",
    "g.set_title('Feature Importance from Random Forest')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support. ' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>');\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option);\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width / mpl.ratio, fig.canvas.height / mpl.ratio);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>');\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(selector.grid_scores_+0.01)\n",
    "plt.xlabel('number of bands')\n",
    "plt.ylabel('accuracy score')\n",
    "plt.title('Accuracy for subsets of most important bands')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 425,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([19,  3, 26, 20, 18,  2, 27,  1, 28,  4, 21, 29, 16, 17, 30,  0, 14,\n",
       "        5, 22, 31, 15,  7, 13,  8, 23, 12,  6, 25,  9, 24, 11, 10])"
      ]
     },
     "execution_count": 425,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argsort(selector.ranking_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=200, n_jobs=None,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = RandomForestClassifier(n_estimators=200)\n",
    "clf.fit(tree_features_upsampled, tree_labels_upsampled)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pixel Based Approach"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "pixels_features_pp_2, pixels_labels_pp_2 = [], []\n",
    "pixels_features_pp_3, pixels_labels_pp_3 = [], []\n",
    "pixels_features_pp_4, pixels_labels_pp_4 = [], []\n",
    "\n",
    "for i in range(len(tree_ids_pp_2)):\n",
    "    tree_image = tree_images_pp_2[i]\n",
    "    tree_image = tree_image/np.ma.sum(tree_image,axis=0)\n",
    "    tree_image = tree_image.reshape(32,tree_image.shape[1]*tree_image.shape[2])\n",
    "    tree_image = np.swapaxes(tree_image, 0, 1)\n",
    "    pixel_features = np.array([pixel.data for pixel in tree_image if not pixel.mask.any()])\n",
    "    pixels_features_pp_2.append(pixel_features)\n",
    "    pixels_labels_pp_2.append(np.repeat(tree_labels_pp_2[i],len(pixel_features)))\n",
    "    \n",
    "for i in range(len(tree_ids_pp_3)):\n",
    "    tree_image = tree_images_pp_3[i]\n",
    "    tree_image = tree_image/np.ma.sum(tree_image,axis=0)\n",
    "    tree_image = tree_image.reshape(32,tree_image.shape[1]*tree_image.shape[2])\n",
    "    tree_image = np.swapaxes(tree_image, 0, 1)\n",
    "    pixel_features = np.array([pixel.data for pixel in tree_image if not pixel.mask.any()])\n",
    "    pixels_features_pp_3.append(pixel_features)\n",
    "    pixels_labels_pp_3.append(np.repeat(tree_labels_pp_3[i],len(pixel_features)))\n",
    "    \n",
    "for i in range(len(tree_ids_pp_4)):\n",
    "    tree_image = tree_images_pp_4[i]\n",
    "    tree_image = tree_image/np.ma.sum(tree_image,axis=0)\n",
    "    tree_image = tree_image.reshape(32,tree_image.shape[1]*tree_image.shape[2])\n",
    "    tree_image = np.swapaxes(tree_image, 0, 1)\n",
    "    pixel_features = np.array([pixel.data for pixel in tree_image if not pixel.mask.any()])\n",
    "    pixels_features_pp_4.append(pixel_features)\n",
    "    pixels_labels_pp_4.append(np.repeat(tree_labels_pp_4[i],len(pixel_features)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "pixels_features_pp_2 = np.array(pixels_features_pp_2)\n",
    "pixels_features_pp_3 = np.array(pixels_features_pp_3)\n",
    "pixels_features_pp_4 = np.array(pixels_features_pp_4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "pixels_labels_pp_2 = np.array(pixels_labels_pp_2)\n",
    "pixels_labels_pp_3 = np.array(pixels_labels_pp_3)\n",
    "pixels_labels_pp_4 = np.array(pixels_labels_pp_4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "pixels_features = np.concatenate((pixels_features_pp_2, pixels_features_pp_3, pixels_features_pp_4))\n",
    "pixels_labels = np.concatenate((pixels_labels_pp_2, pixels_labels_pp_3, pixels_labels_pp_4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "07d550be9cf347c8a661b52737d670f0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=10), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9168260038240917\n"
     ]
    }
   ],
   "source": [
    "scores = []\n",
    "preds = {0:[],1:[],2:[],3:[]}\n",
    "\n",
    "cv = KFold(n_splits=10)\n",
    "\n",
    "for train_index, test_index in tqdm_notebook(cv.split(pixels_features), total=cv.n_splits):\n",
    "    X_train, X_test = pixels_features[train_index], pixels_features[test_index]\n",
    "    y_train, y_test = pixels_labels[train_index], pixels_labels[test_index]\n",
    "    clf = RandomForestClassifier(n_estimators=100)\n",
    "    clf.fit(np.concatenate(X_train), np.concatenate(y_train))\n",
    "    for i in range(len(y_test)):\n",
    "        pixels_preds = clf.predict(X_test[i])\n",
    "        y_pred = np.argmax(np.bincount(pixels_preds))\n",
    "        y_true = set(y_test[i]).pop()\n",
    "        scores.append(accuracy_score([y_true], [y_pred]))\n",
    "    \n",
    "print(np.mean(scores))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Delineated Tree Crowns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from skimage.feature import peak_local_max\n",
    "from skimage.filters import gaussian"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/vasilii.mosin/.virtualenvs/forestry/lib/python3.6/site-packages/ipykernel_launcher.py:2: RuntimeWarning: invalid value encountered in true_divide\n",
      "  \n",
      "/Users/vasilii.mosin/.virtualenvs/forestry/lib/python3.6/site-packages/ipykernel_launcher.py:5: RuntimeWarning: invalid value encountered in less\n",
      "  \"\"\"\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAABQAAAAPACAYAAABq3NR5AAAAAXNSR0IArs4c6QAAQABJREFUeAHsnWusLklVvxuYCwNEEBAweGBgIDEoRLmIl0QSRC5+ICCiDjcDIvABE9HgJQQ0EkkMAWKMGgYkfwdJBB0hSogoEkj8oHKVW6IyI3I0AgqIDjAMtz9rj2vP2nWququ6q6qrqp832VPdVavW+q1nVffbb827z77F17/xmnhBAAIQgAAEIAABCEAAAhCAAAQgAAEIQAACQxK45ZBZkRQEIAABCEAAAhCAAAQgAAEIQAACEIAABCBwQoANQBYCBCAAAQhAAAIQgAAEIAABCEAAAhCAAAQGJsAG4MDFJTUIQAACEIAABCAAAQhAAAIQgAAEIAABCLAByBqAAAQgAAEIQAACEIAABCAAAQhAAAIQgMDABNgAHLi4pAYBCEAAAhCAAAQgAAEIQAACEIAABCAAATYAWQMQgAAEIAABCEAAAhCAAAQgAAEIQAACEBiYABuAAxeX1CAAAQhAAAIQgAAEIAABCEAAAhCAAAQgwAYgawACEIAABCAAAQhAAAIQgAAEIAABCEAAAgMTYANw4OKSGgQgAAEIQAACEIAABCAAAQhAAAIQgAAE2ABkDUAAAhCAAAQgAAEIQAACEIAABCAAAQhAYGACbAAOXFxSgwAEIAABCEAAAhCAAAQgAAEIQAACEIAAG4CsAQhAAAIQgAAEIAABCEAAAhCAAAQgAAEIDEyADcCBi0tqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBgA5A1AAEIQAACEIAABCAAAQhAAAIQgAAEIACBgQmwAThwcUkNAhCAAAQgAAEIQAACEIAABCAAAQhAAAJsALIGIAABCEAAAhCAAAQgAAEIQAACEIAABCAwMAE2AAcuLqlBAAIQgAAEIAABCEAAAhCAAAQgAAEIQIANQNYABCAAAQhAAAIQgAAEIAABCEAAAhCAAAQGJsAG4MDFJTUIQAACEIAABCAAAQhAAAIQgAAEIAABCLAByBqAAAQgAAEIQAACEIAABCAAAQhAAAIQgMDABNgAHLi4pAYBCEAAAhCAAAQgAAEIQAACEIAABCAAATYAWQMQgAAEIAABCEAAAhCAAAQgAAEIQAACEBiYABuAAxeX1CAAAQhAAAIQgAAEIAABCEAAAhCAAAQgwAYgawACEIAABCAAAQhAAAIQgAAEIAABCEAAAgMTYANw4OKSGgQgAAEIQAACEIAABCAAAQhAAAIQgAAE2ABkDUAAAhCAAAQgAAEIQAACEIAABCAAAQhAYGACbAAOXFxSgwAEIAABCEAAAhCAAAQgAAEIQAACEIAAG4CsAQhAAAIQgAAEIAABCEAAAhCAAAQgAAEIDEyADcCBi0tqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBgA5A1AAEIQAACEIAABCAAAQhAAAIQgAAEIACBgQmwAThwcUkNAhCAAAQgAAEIQAACEIAABCAAAQhAAAJsALIGIAABCEAAAhCAAAQgAAEIQAACEIAABCAwMAE2AAcuLqlBAAIQgAAEIAABCEAAAhCAAAQgAAEIQIANQNYABCAAAQhAAAIQgAAEIAABCEAAAhCAAAQGJsAG4MDFJTUIQAACEIAABCAAAQhAAAIQgAAEIAABCLAByBqAAAQgAAEIQAACEIAABCAAAQhAAAIQgMDABNgAHLi4pAYBCEAAAhCAAAQgAAEIQAACEIAABCAAATYAWQMQgAAEIAABCEAAAhCAAAQgAAEIQAACEBiYABuAAxeX1CAAAQhAAAIQgAAEIAABCEAAAhCAAAQgwAYgawACEIAABCAAAQhAAAIQgAAEIAABCEAAAgMTYANw4OKSGgQgAAEIQAACEIAABCAAAQhAAAIQgAAE2ABkDUAAAhCAAAQgAAEIQAACEIAABCAAAQhAYGACbAAOXFxSgwAEIAABCEAAAhCAAAQgAAEIQAACEIAAG4CsAQhAAAIQgAAEIAABCEAAAhCAAAQgAAEIDEyADcCBi0tqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBgA5A1AAEIQAACEIAABCAAAQhAAAIQgAAEIACBgQmwAThwcUkNAhCAAAQgAAEIQAACEIAABCAAAQhAAAJsALIGIAABCEAAAhCAAAQgAAEIQAACEIAABCAwMAE2AAcuLqlBAAIQgAAEIAABCEAAAhCAAAQgAAEIQIANQNYABCAAAQhAAAIQgAAEIAABCEAAAhCAAAQGJsAG4MDFJTUIQAACEIAABCAAAQhAAAIQgAAEIAABCLAByBqAAAQgAAEIQAACEIAABCAAAQhAAAIQgMDABNgAHLi4pAYBCEAAAhCAAAQgAAEIQAACEIAABCAAATYAWQMQgAAEIAABCEAAAhCAAAQgAAEIQAACEBiYABuAAxeX1CAAAQhAAAIQgAAEIAABCEAAAhCAAAQgwAYgawACEIAABCAAAQhAAAIQgAAEIAABCEAAAgMTYANw4OKSGgQgAAEIQAACEIAABCAAAQhAAAIQgAAE2ABkDUAAAhCAAAQgAAEIQAACEIAABCAAAQhAYGACbAAOXFxSgwAEIAABCEAAAhCAAAQgAAEIQAACEIAAG4CsAQhAAAIQgAAEIAABCEAAAhCAAAQgAAEIDEyADcCBi0tqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBgA5A1AAEIQAACEIAABCAAAQhAAAIQgAAEIACBgQmwAThwcUkNAhCAAAQgAAEIQAACEIAABCAAAQhAAAJsALIGIAABCEAAAhCAAAQgAAEIQAACEIAABCAwMAE2AAcuLqlBAAIQgAAEIAABCEAAAhCAAAQgAAEIQIANQNYABCAAAQhAAAIQgAAEIAABCEAAAhCAAAQGJsAG4MDFJTUIQAACEIAABCAAAQhAAAIQgAAEIAABCLAByBqAAAQgAAEIQAACEIAABCAAAQhAAAIQgMDABNgAHLi4pAYBCEAAAhCAAAQgAAEIQAACEIAABCAAATYAWQMQgAAEIAABCEAAAhCAAAQgAAEIQAACEBiYABuAAxeX1CAAAQhAAAIQgAAEIAABCEAAAhCAAAQgwAYgawACEIAABCAAAQhAAAIQgAAEIAABCEAAAgMTYANw4OKSGgQgAAEIQAACEIAABCAAAQhAAAIQgAAE2ABkDUAAAhCAAAQgAAEIQAACEIAABCAAAQhAYGACbAAOXFxSgwAEIAABCEAAAhCAAAQgAAEIQAACEIAAG4CsAQhAAAIQgAAEIAABCEAAAhCAAAQgAAEIDEyADcCBi0tqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBgA5A1AAEIQAACEIAABCAAAQhAAAIQgAAEIACBgQmwAThwcUkNAhCAAAQgAAEIQAACEIAABCAAAQhAAAJsALIGIAABCEAAAhCAAAQgAAEIQAACEIAABCAwMAE2AAcuLqlBAAIQgAAEIAABCEAAAhCAAAQgAAEIQIANQNYABCAAAQhAAAIQgAAEIAABCEAAAhCAAAQGJsAG4MDFJTUIQAACEIAABCAAAQhAAAIQgAAEIAABCLAByBqAAAQgAAEIQAACEIAABCAAAQhAAAIQgMDABNgAHLi4pAYBCEAAAhCAAAQgAAEIQAACEIAABCAAATYAWQMQgAAEIAABCEAAAhCAAAQgAAEIQAACEBiYABuAAxeX1CAAAQhAAAIQgAAEIAABCEAAAhCAAAQgwAYgawACEIAABCAAAQhAAAIQgAAEIAABCEAAAgMTYANw4OKSGgQgAAEIQAACEIAABCAAAQhAAAIQgAAE2ABkDUAAAhCAAAQgAAEIQAACEIAABCAAAQhAYGACbAAOXFxSgwAEIAABCEAAAhCAAAQgAAEIQAACEIAAG4CsAQhAAAIQgAAEIAABCEAAAhCAAAQgAAEIDEyADcCBi0tqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBgA5A1AAEIQAACEIAABCAAAQhAAAIQgAAEIACBgQmwAThwcUkNAhCAAAQgAAEIQAACEIAABCAAAQhAAAJsALIGIAABCEAAAhCAAAQgAAEIQAACEIAABCAwMAE2AAcuLqlBAAIQgAAEIAABCEAAAhCAAAQgAAEIQIANQNYABCAAAQhAAAIQgAAEIAABCEAAAhCAAAQGJsAG4MDFJTUIQAACEIAABCAAAQhAAAIQgAAEIAABCLAByBqAAAQgAAEIQAACEIAABCAAAQhAAAIQgMDABNgAHLi4pAYBCEAAAhCAAAQgAAEIQAACEIAABCAAATYAWQMQgAAEIAABCEAAAhCAAAQgAAEIQAACEBiYABuAAxeX1CAAAQhAAAIQgAAEIAABCEAAAhCAAAQgwAYgawACEIAABCAAAQhAAAIQgAAEIAABCEAAAgMTYANw4OKSGgQgAAEIQAACEIAABCAAAQhAAAIQgAAELgIBBPYk8O///u/Ta1/72ukd73jH9IlPfGK65JJLpnPnzk2Pecxjpic/+cnTZZddtqc8YkMAAhCAAAQgAAEIQAACEIAABCAAge4J3OLr33h1nwUJdEng7W9/+/T85z9/uv766736L7/88umqq66a7nnPe3rH6YQABCAAAQhAAAIQgAAEIAABCEAAAhBYJsAG4DIjLAoQ+MhHPjJdeeWV0w033DDd5ja3mZ797GdPD33oQ0/O3/KWt0xveMMbTqLKJuA111wz3e52tyugApcQgAAEIAABCEAAAhCAAAQgAAEIQGB8AmwAjl/jJjOUX+9997vfPV100UXTH/7hH07f/d3ffUbnq1/96umlL33pSd9zn/vc6Wd/9mfPjHMCAQhAAAIQgAAEIAABCEAAAhCAAAQgEEeAPwISxwmrjAQ+8IEPnGz+icsnPOEJF2z+Sf8znvGM6YorrpDD6eqrr56+/OUvnxzzHwhAAAIQgAAEIAABCEAAAhCAAAQgAIE0AmwApvHCOgOBt73tbadeZAPQ97rlLW85Pe5xjzsZ+p//+Z/p7/7u73xm9EEAAhCAAAQgAAEIQAACEIAABCAAAQgsEGADcAEQw/kJvOc97zlxKv/233d8x3cEAzzkIQ85HXvve997eswBBCAAAQhAAAIQgAAEIAABCEAAAhCAQDwBNgDjWWGZicC111574uke97jHyb8BGHJ773vf+3RI55x2cAABCEAAAhCAAAQgAAEIQAACEIAABCAQReCiKCuMIJCJwJe+9KXps5/97Im3u93tbrNeb3/725/8heAvfOEL0yc+8YlZWx183/veN33ta1+bbnGLW8xuLqo9LQQgAAEIQAACEIAABCAAAQhAoCUCX/nKV6avf/3rk/zTWO4fzGxJJ1r6IsAGYF/16l7t5z//+dMc5FeAl16XXXbZJBuA8hPzks0/ecnNkj8cEkMMGwhAAAIQgAAEIAABCEAAAhBokYB+vm1RG5r6I8CvAPdXs64VyzcA9XXxxRfrYbC95JJLTsZuuOGGoI0dkG/+8YIABCAAAQhAAAIQgAAEIAABCPROgM+3vVewLf18A7Ctegyv5tJLLz3NMeYbejfeeOOJ/a1vfevTeXMHF1100ek3/5761KdOn/70p+fMGRuQwJ3udKfpta997UlmrIG6BT5//vx07ty5ukE90XpdA8LPvixLHbN91pbjswR6XQNns+BsKwHWwVaC/c9nDfRfw60ZsAa2Eux/fq9r4Oqrr57ufOc7889a9b8Em8qADcCmyjG+mNve9ranScb8Wu8Xv/jFE/uYXxc+dfx/B7L596lPfcrt5vxABFgDdYttN/jrRg5H62UNyD9b4L7k/uX2c09zKS2f97IGljPBYgsB1sEWemPMZQ2MUcctWbAGttAbY25Pa+CrX/3qGNDJoikC/ApwU+UYX4xsENzhDnc4SXTpD3t87nOfO/23/5b+YMj45MgQAhAYkYBs8LmbfJpnqF/HaSEAAQhAAAIQgAAEIAABCMQSYAMwlhR22Qjc5z73OfH18Y9/fJK/bhR6XXfddadDV1xxxekxBxCAwHEIzG2QuRTUVn9d1h1v8Zx/16XFqqAJAhCAAAQgAAEIQAAC4xFgA3C8mjaf0YMe9KATjfIrwB/+8IeDet/1rnedjj3wgQ88PeYAAhA4BoG5b8DNjY1Kh83CUStLXhCAAAQgAAEIQAACEChPgA3A8oyJ4BB4xCMecdpzzTXXnB7bA/lz529605tOur7pm75peuhDH2qHOYYABA5MQDf/9Bt/2vqQyLcB1d43vndfirYU273zIj4EIAABCEAAAhCAAAQg0BYBNgDbqsch1DzgAQ+YHvzgB5/kKhuA73vf+y7I+zWvec107bXXnvQ/7WlPmy6++OILbOiAAATWE5jbNFvvtc7MNd+Ea3HzrEVNdSpIFAhAAAIQgAAEIAABCECgNgH+CnBt4sQ7IfCCF7xguvLKK6cbbrhhesYznjE95znPOfmWn5y/5S1vmV7/+tef2F1++eXT05/+dKhBAAIHJDC30SdjbKCNsyhCtZxbA+NkTyYQgAAEIAABCEAAAhAoT4ANwPKMieAhcL/73W96xSteMT3/+c+frr/++unlL3/5BVay+XfVVVdNt7vd7S4YowMCEICAuzkU2kSCVLsElmrmG3fr3m52KIMABCAAAQhAAAIQgEA7BNgAbKcWh1Py8Ic/fPqzP/uz6eqrr57e8Y53TJ/85CdPftX3Hve4x/ToRz96espTnjJddtllh+NCwhCoQWDETRTJybdhpDx1rIXcVYtqi211Xgs5xGoO2WkuoXFf/wh5+/KiDwIQgAAEIAABCEAAAqUJsAFYmjD+Zwnc/e53n37lV37l5GfWkEEIQAACEQRkg+hLX/pShGW8iWxUsfEUzyvFUrmGNgN1PMUnthCAAAQgAAEIQAACEIDAhQT4IyAXMqEHAhCAAAQ6JnDu3LlZ9bLZZH9CxmrjG9cxaVNfa+a4MTS+27/2PLe/GB12c88ex8zFBgIQgAAEIAABCEAAAhBII8AGYBovrCEAAQhAoFMCssnk22gKbchZW90g07YVBLn15PaXwsnyTpmHLQQgAAEIQAACEIAABCCwTIANwGVGWEAAAhCAQOcE7OaSPda03I0v91ztfK3Yxr5SbGN9ip3qXevfZaL+UjRstV2rfWvcLfOVU4/at+TNXAhAAAIQgAAEIACB/giwAdhfzVAMAQhAAAKJBNwNGtnwsj/qTjd09DymdTfPfHPW+PX5ielzc90yZ42vmHjWpkYMGy/Xca+6c+WPHwhAAAIQgAAEIACBvgiwAdhXvVALAQhAAAIrCMRs0vncrp3n8xXTlyteb5tTvemVWkqt9CemtthAAAIQgAAEIAABCEBgTwJsAO5Jn9gQgAAEIFCUgPxBkC2banMbU7k3f7bodCHm9OX65vxCAvC+kAk9EIAABCAAAQhAAAJtEWADsK16oAYCEIAABAIE5jbjAlOKdads+MTqVrsU38USHMjx+fPnT7JRvgOltnsqwhSuu5cBARCAAAQgAAEIQCCKABuAUZgwggAEIACBvQn0ujGWoruVDRXRrD+l657CZ60W4Vojzlp9vc6Daa+VQzcEIAABCEAAAkckcNERkyZnCEAAAhCAgBIIfYNJNjd8YyNveoycm9abNi+BuTXjXj9ztnlV4Q0CEIAABCAAAQhAwCXANwBdIpxDAAIQgMAqAvJhX39WOdhhkm5QyMaEbk6EjrfIU99bfJSca3MuGUd8K/OYOCm2IX9b/x3IkN+a/Tk41NQrsXrUXJsR8SAAAQhAAAIQgEBNAmwA1qRNLAhAAAKDEnA/7PvO3b69UaieWptzo8VZUz9lrq34sMdrfDKnbwLUv+/6oR4CEIAABCAAgX4I8CvA/dQKpRCAAASaJBD6AO/r175am2FzwHwafH3iQ/pV+5zPFsZEZyiPNfrcvF0Wa2O5fn3a1vq2vuSPgFx66aW2a/HYpy2HlsXAgxi4/HzsXJuY1GWOz1fMXGwgAAEIQAACEIDA0QnwDcCjrwDyhwAEILCBwJoP8RJu7bwNUjdPlY0HNh9uwqgctN0MFwezBHJzLn395darcEr5Vf+0EIAABCAAAQhAYGQCfANw5OqSGwQgAIGKBOTDeemNhYrpZA9Vk00o1poNlNCcUP8cOF0ja+bO+V0aEx4pMVXnkt9ex1NYrM2xRoy12pgHAQhAAAIQgAAEjkiADcAjVp2cIQABCGQi4H7I13N3A0r7JayOpW7KbJFcM9YWnaXntsDBroXS+ar/NTHXzNF4tMsE4LvMCAsIQAACEIAABCCQkwC/ApyTJr4gAIHuCcgGif50n0wHCQjr0q8aMUrnkNP/6DwkP/l3/470kpzdnyPlT64QgAAEIAABCEAAAssE2ABcZoQFBCAwIAE+LNcrqnzTx/J2I8tYjVetOKFcWvrG094sQoy29ofyCvVvjdfq/JbWWquM0AUBCEAAAhCAAASORoBfAT5axckXAhA4/RXUORR8gJ6jEz+mHLUNbcRIv9rEe+/bUvIN8VibmTKM8Tsa85ic13LtYZ7WvgetaLyZgF23qTW0c2/2GD5K9R/2xAgEIAABCEAAAj0SYAOwx6qhGQIQyEpAPhSlfpDKKmAwZ8qyhQ+b9ldBW9DjatBzZbZ2Kagfna/nS351XO11Pm1fBKhfX/UStXrtWeXSV7KWvpg2/txxSV1zcRmDAAQgAAEIQCAfAX4FOB9LPEEAAh0S4ENN3qLpB8wYrjE2edW1660Uixi/Phupo/vTLr1jKNNr6xjZjp+l77rz9YVIsB5CZOiHAAQgAAEIQCBEgA3AEBn6IQCBYQmkfMgqCWHkDRbNLcRPayCt/oRsj9KvHJSNm7cdd49dW3vu2rr+tVa2tfNLHEusudfS+Nxcd+zcuXNFv1XlxitxnpNHCX34XEfAvTZjvdRaD1ZfrDbsIAABCEAAAhBolwC/AtxubVAGAQgUJOBughQMdYHr0Ic37d9T2wViEzpU/9KUXvNbyivneGlG6j+2ZjlzE18aP+R3adw3T+bslY9PD30QgAAEIAABCEAAAhBoiQDfAGypGmiBAASGJSAbE/qzlKTaabtkn3NcYq59+TZtfDn4+tbGZN42Ar6abfOYZ/badSj5yI9862+kl+Y1Uk7kso7A2msjNVrte4O+L9g2VTP2EIAABCAAAQjME2ADcJ4PoxCAAAQORWDrh76t83PDthtBtT44586hpL9WmbS2jkrWAN8QiCWQ+3oNXWeh/liduexy55tLF34gAAEIQAACvRLgV4B7rRy6IQCBogTkg0cLH4Ja0OCCdj+UuRrl3LVxz905boyc56qnZsyc+lvx1co10QoPdECgRQJ6n3PvuapVx91za1/rWrcxVQ8tBCAAAQhAAALlCLABWI4tniEAAQgcksDShzr3A6hA0jm+sRwQS/nNoa1lH3DLUx1d3+INpnmYHs2LXUNu7u6a8tm6Nj4fdp4cL81xfaSc21gp87CFAAQgAAEIQGA9AX4FeD07ZkIAAhA4PAH3A+KaD3V2jj0+PNyGAEhdqM36grjXyXpPzBydgF5r2q7Jd+16c+dxza+hzxwIQAACEIBAuwTYAGy3NiiDAAQqEpAPPi1++NnyIbA0PstrrU7fB8y1vkrne0T/Wh+ttZ4fkcWWnJWbctzii7njENB7nW3d7HTM7dfz3GvK9adrV+PRQgACEIAABCDQLwE2APutHcohAIECBOTDj/6I+6UPXwUkDO3S/XAZSjbWLjSf/jwEbB30mA2BdWyV37rZzIJAPQJ2rdrjegqIBAEIQAACEIBACQL8G4AlqOITAhAYgoB+8NFNQD1PSS7XZon6WaMhRW+MrU+D9KlG6yPUb2302OdXx2jrEvDVwtZXjn02dVX2Ew1W/dSqN6WlrsWSa9beS+Z4l9QwF5cxCEAAAhCAwKgE+AbgqJUlLwhAIDuB2A8tElhsU+xjxbo+NY7bH+uvpJ18eLO6fB/mfH0lNeE7HwFb23xe8QQBCIxMgPvGyNUlNwhAAAIQaJ0A3wBsvULogwAEqhGQDya+DSlf35yo0h9wSvufy82Oxeiw7HLxtRpqHYe014pfM46t2VLcGC52naT4XorNOAR6J+C7HmKvF2sXcx22wMpqXtLjY7M0h3EIQAACEIAABOYJ8A3AeT6MQgACByKQ4wNHygec3GhrxZY4NpaPm/aprbXPnXcNf5pPjVh7xpjLMzSmNfbp7r3uvpzog0BJAnqdaRuK5Y7ba80eh+bX7o/VJHm5udXWSjwIQAACEIDAqAT4BuColSUvCEBgFwLywSX2g04JgRK79IenFP9qq0xq6FviqlrETvUtzRl1XPK3PHLkOeevhfrnyBEfEChJQK/LpfuTjus1p21Jbfhul4DWX9fFnFK1DdnE+AjNpR8CEIAABNolwDcA260NyiAAgU4J7P3gvPRgXxqrL39fX2kdrn/hsjcbV1ML57Y2S3ysrasdvi4RziGwnoBca0vX43rvzByZwNy64T49cuXJDQIQgMAyAb4BuMwICwhAoCEC+mA7txHRgtw9P7ztwSYm5p5MdN3o2ojRq7YprcQp5TtFR6xtLS5WT2+MrHaOIVCTQOy9RO3c61m16rie79FaDSGde+gaJaZlalnb/KyN7bfHobnWhmMIQAACEOiXAN8A7Ld2KIfAoQnEPMjuDWiPB+k9Yu7NOSW+8CnFSNektim69rBdqzMHv7WxYziV9B0THxsI7EVA729uu5eeUFxXnz0PzaF/mYByXLa80GLL3Au9re/h/r2eHTMhAAEIxBBgAzCGEjYQgECTBORBsfWHRXmodl+lHrR9sdzYe5/vqbFkbLsOS8bJVT+rN5fPVD8lr9/z58+fyrHH0lky7mlQDiAAAQhUIhB7P+/hvakHjZXKShgIQAACRQiwAVgEK04hAIGaBPQDfexDcE1tEkseaO2P9OXWutdDs7KXnFp+leaj/rVtgYXWRlvVFFp7sdpD89X/3q3kce7cuVMZ9lg6ZTw211MnHEAAAhBolEDKPS1072v9vu5Dr+9t2vps6IMABCAAgbME2AA8y4MzCECgYQJbH1C3zm8YDdIaIBD6YNWAtBMJcx+SWtM+p7UVnuiAQG4CvEflJnrWn/tt4LOjnPVKIGUDtNcc0Q0BCEAgFwE2AHORxA8EIFCMQMpmgPuAr3P1g5W2xcQuOFY9C2YnwymbMnvkZWPa45jcsMlPQNeWtvkj5PcYWuOh/vwK8AiBNghwD61Th9Y4175f+/LvdQPNp9uXX52VRRQIQAACfRBgA7CPOqESAoclsOVhzt0MFIh7biyk5KI6tV1aALF2S362jKfktyUOc28mIMz15+bePo6W1uzSeMtZci20XB20HZWA/HMAPd9Xjlq3lLypbwotbCEAgSMSuOiISZMzBCBwXAL6wbz1h0TVuVSpWnmE9Ej80JholzHXppbmJXYlxpVFLzmm6tT81rJLjbc2zp7zlJGu/T215Iit+Ryhdjl44eMsAV0/2ss6UhI3tzWYjHI/upkaRxCAAAQgsIYA3wBcQ405EIAABHYiIB8U7E8NGe4HOI2pH1q0lX6xVXu31XlHaDX3lnO1dWtZZ2/aRuLawzrubX2gFwKWQO37hcSzMe2x1cUxBCAAAQiMSYBvAI5ZV7KCAAQWCMgH270efH1xQx+0fbYLqZ0MW39rfYgj62cursaw9tJnz3W+9ukc7e+91bw0D/dc+6XNkfucfxtD7bbGVD82j9jjrbFj47Rid7R8W+GOjvYIyLWg9w6ui33qI/xd9u75PsqICgEIQAACtQmwAVibOPEgAIFmCPgeikuKm3vgnhsrqUk/mOWOEeO3Nv/cOW7xVzP3vdbWFj7MhcDRCMh1GnPf7JEL96D9qjbqmqpN1HJkPdemTzwIQCAnATYAc9LEFwQgEEVANz/sA5VMtA9V7liUY4xOCLjslLfF49rYMd9xix9ONQe7bnzaa/VZHaptLravLnP27lioJlaHO6fWeQsaSuQ6V9dRcxaOI+dWYp2s9QnnteT6nefeU3KtAdfv1vebUoRTdaXar9Gt7LQWeq6+amjQWLQQgAAEchNgAzA3UfxBAAKzBOyD09zD1ayThEGN4U6xOtyxEc/dB9jYHEP8as33xXFz0fOtWn2x1va5WlTjWn+15m3V6eZdS3eNOEtsZHy0/EfLp8Y6IQYEUgjINWbvLVvvI9aX1dHCtRzSZnW2eOzWqAWWLXJCEwQg0AcB/ghIH3VCJQSGIeB7cJI+7c/5gHju3LkgN42nBhI3Z2z123s7x2QPZiE9bj1b415bX4jTHJc1c+b8MQYBCEAAAssEcrw/zL0f5/C/nEUdi9K5+N4Hta907DoEiQIBCBydAN8APPoKIH8IVCQQ8xAlD1hqF5KmD2FLdqH5oX71GxrvpT+GYWouc6x9Y7lZujFy+0/lscberUuJHNSntmt0rp2zR8y1WkvNk3UKh1J08QuBcQnIfUPf5+buI74xnRei45sTsi3Zr/fGJb0lNcz5dnXpueqem8sYBCAAgV4IsAHYS6XQCYEDEdCHLX34sqnrmO0rcezGrhW3RC65fSoLl5HG0XE939qG4mz1O9r83Nxb5SPr4Si5tloDdEFACbj35x6vTc2htHaNI+zcWHZM2bqt2kjrzndtOYcABCAAAQj4CLAB6KNCHwQgkJ2APrjOOXYfapcecGU8xu9cTB2zsVyfdkztW29zsnHr4su9JCM3l5KxfLnl7HNzyek7pk454+3hy702c2vI4V999LxOc3PF33gEdJ27mUl/b2u/Bb2qQblqq3x1XM/dce1vsV3SquNujrG56Hy1X+tH5uXypVpoIQABCLRGgH8DsLWKoAcCENidwNqHx92FOwJy5uE+FDuhqp3mzKma6IKBcvKI9SV2vp+CaVZxnXuNiz/9qZIAQSBQkYDeAyqGPESoENe196eQv9Zgis6U1xyPubGlGFaHPdZ5vj4do4UABCDQAwG+AdhDldAIgc4JxD6MrXmw0jm+GOfPn58uueSS7r6N0Gq5hbHw9rEurVnrXDpODf+5c8lRk6Wa5tZcg3NsjKXcY/2MZGeZjFz7kWq2Vy66Puyascc6vpe+FuPGMBEbyzE2jxjfsb7W2q3RHRvL+rbH7nwda4GHq41zCEAAAnsS4BuAe9InNgQOQEAfwkqmuhRjabyktr195374PTLLvWu5V/zca2hLHqKlJT1bcmEuBEYiELouj/aeIfmGck69f4WYhtZNqn3IT+3+XnXX5kQ8CEAAAjkIsAGYgyI+IACBIAH3gTf3g17oQdsKyh3T+m75eO6DSG7dMXXIHTPWX8vaYnOItTtSrrFMluxq3x/0umy1VlZXbTZLtTrSuF0neuzmH+p37Wqdh9aL6pS2pZfVtUWb9aP5aZ+er22FaYjrWp+l56XoTbEV3an2Ugf5bZDYl66D1Dix/rGDAAQgsDcBNgD3rgDxIXAgArUfqM6dO3cgumdT1YfYs71lz/aIuZRRyoP/kq8exmtfYz0widFYmpteG9qqJt+59Ln9ap/Sqh9fG/KTI27IN/3bCbi13OKRWm+hd/PcWhzn7lEyNjd+s9o6R7FM9tDsPhOErqnYHOoQJQoEIACBfATYAMzHEk8QgMAMgT0e9ETO2rgyT39m0upySPNy2y6T8Yh2H/A9JsN1aS2HS6xiQj6G2ift1tfSB0p3XD+Ybo3rm6++bUx77JtD3zgEStY6x7VSi7SrVa+LXPFd/1v9ij/fz1a/OefPrS2rXWLO2YY0rZkT8kU/BCAAgSMSYAPwiFUnZwgMRCD3A/ZAaE5T0Yfu047AgWWpc7QNTLmgu4WH89a/+SmMWuB0QfHoOCGga95eDzIQ6ldsrr32L7Vz62Gtz6WYdnwuvrXjeJlAzuu6ZO3Vd069S3QkpsZdsmW8HIHS13tqjWPXYC7d7vOBrstU3eUqhGcIQAACZQmwAViWL94hAIFOCeR62NwrfX2Y1VZ02OOQLrGJsQvNl/7YB/o5HznHtuaTU4v11fsas7kc7di3plLWvV5nPj9HYzlKvin1j8251/Xh6nbPY/OvYafatK0R86gx9D1vy7UidQrVam5sDfMtOtfEc+fk4OX65BwCEIAAG4CsAQhAYFgCa38V1D702ePeQPkekn19veWVotf9v/0pc2vZ9rzGajEaLU6o5voBVts1eYvvkP9Uf7n8pMbt0X5LzebyVb9z7dz80Jj6C41v7Vf/0rb+Uo2q2dWr14FeW7bVue4cOdd5vrGR++aYuHnH2rq1sfN8Y9onbexzgPWpOlupYSs6lAstBCDQLwE2APutHcohAIFvEMj9UOTz5+s7EvxQ/r6H5SNxGSnXuVqG6j9S/mtymWO25M8319e35IdxCGwhINc21/c8QeWjrWsd6le7pXG1q9XKfWaUe02JXNSnbWvVZilOa2tpSS/jEIBAmwTYAGyzLqiCwNAEcj7ELD3IpsRKsR26QE5yLmN7rsfS2h/HRdKp1GHkWigzC6WFfG39XI219GkcbS2j1o5zaxR/a33q3LXzla3W3a4FHWulzZVrK/m0oGPrumkhh1Ia9Joo5R+/EAgRsPdhPQ7Z0g8BCEAglsBFsYbYQQACENhKwH7IkOOWHqyttq15Hm1+jjoejb8ys3m3dk2469hqVf2uDedpBCzTtJkXWktNcvhrubY58ruQXB89NveWa9QHTVS2RsCub9HGGm+tQuiBAARGIcAG4CiVJA8IHJjA0gdfebCce5h0HzwV5dwctdmjVb019bmx5HyJ6xIbzcO1c2O54yOdK0fNaStT9VO6dWuXq2bqR9uSeWxh7eafqnPtfJ3n8tFzbdUuRZfOTZlT03ZNTjX1lYpVMm+peUn/pZjU9BvLp/Xrpyaz2FjKLJZxrF/sIAABCEAgTIBfAQ6zYQQCEOiIgD5IhiTrA6a0+iO22h+a11p/S3qXmLfG7oh6aqyXHtdBDS5r1tscyxTN4sf+LGmZi7s0t9a4q9E9r6WjZpyUmq/VpRxrxFqrsaV5wkuZtaSrNS0xjEJrLmZua/miBwIQgEAvBPgGYC+VQicEOicgD3p7P9S5D5vueQnEbowtDFxfJfSW9mlziGUhc2JtS+uv4V8Z5ch5iw+Zq1pCeW/x7/PZS6017yU+vhxDfeozNL6l3+d7rXad5/O5RWPM3D1ixugqZSP5Ku9SMcTv0bimsAyxCfWn+Mb2wv8JC9f4VbHm3gDfeL5YQmBUAmwAjlpZ8oIABJok4Htgi3kgc+fFzGkSAKJOCbg1PR3o5KDEGizh04ezVhxf7LV9olnWTC7tuf2tzYt5EIDAmAT0XtX7e92Y1SErCEDgqAT4FeCjVp68IbCRgDzQpTzU6YPgxrCz02vEmBWwclBZzrWu6xT27ty151bfWh8yT+qUUqsU2y26WpmbyqcV3a3oqHltlK6Vm0vua2GNPzfnXPeFVtZPSzrc+rekDS19EGAN9VGnWJXUM5YUdhCAQIgAG4AhMvRDAAJBAi0/gJw7dy6oe48B98PyHhpKxJQ1oD+x/lteN7E5lLRrba3M1WvNxlFJdiP6Vv7atphjy9pa5JWiCbYptLANEdjrXi3rV39C2tb2q1+3Xeuvh3maq2iVY3npM8Nce2LIfyAAAQgYAmwAGhgcQgACcQT0YSPOGqsSBPZ6qHdzSdHBunHp3XyewvHmWcc70g8+c5mPwlLz0HYu573GWta2F5MacWOugxo6iAGB3ASW1vbSeG49rfqL5SD3aO7TrVYRXRDYhwAbgPtwJyoEhiCw9qFCHlxiH15ygNqiM0f8nn1ordx68VC5vaq6LrXd7jGfB7feruelcdd+67muQ/FTO7ZqVw22La2nxbWhPLRVjdLuVRvVcqQW1keqdl+5bl2beo/tK+t21Wo99F7drlKUQQACNQiwAViDMjEgAIEzBOQhJPQgog8qZybseJJDTw4feyCY0z03tofWXmOGroNe8ymhu+W11rK2ErUI+dR1rG3Ijn4IQAACsQR891fuMbH0brYTZnC7mQdHEDg6Af4K8NFXAPlDoDECpR5SxK/vYbJ0+jljlmJjGcTqraHF6uL4WASW1qGM116DGm9J25pKqe81c5lzDAJ2jZRYg8egSJY1CMhaLblGfdeC7auRIzEgAAEI9EqAbwD2Wjl0QwAC0QRKPohGi9hoWOPhNoVTiu3G1JneKIFSayDWb6xdo/hOZdW4tk+DcdAFgaW1LWuGddNFKQ8rUteo26YAiVnj6j/Fbwu2S9f4Xhpb1bUXD+JCYEQCbACOWFVygsCBCZw/fz6Y/doHRXkgWvNQtGaOK36tZtfP0vkarWvmLOlgvB8CMR/OUrNJXVOp9ql6atiPkEMNTkeKoff9EtfYkTiS6z4EQve0UL9P5ahrXxgoBz3Wc+Wg/XJuj+W81L3BjSOxeEEAAmMS4FeAx6wrWUEAAoaA+yAp5+4DlzHn0CHg8rPs5Ngdd6Zz2iEBrWnN+tp11SEyJEOgKIHa16LeA4omhfMmCei9eI81sEfMEkVQhuJbc7J9bkwdU1u3de1LnrtaSsbCNwQgUJ8A3wCsz5yIEIBAIQL60LLGvTxs2Z81PubmqO85GzuWam/nrjlOiefabuG+Ritz6hGQWrf+GmH9LeWg49Lqcet1QV9eAjWuRV1fNWLlpYO3XATs/cUep/p357rn1p8+U/S47iQv/bE52eOlcWsrxzk4pMQUW98rxYdvPn0QgECbBNgAbLMuqIIABDIQOHfuXPBBSh+w9MHTDaf9aueOL53bByrxoQ9Sc/62xlzSFDOuGsTWHvvm6ri0vMYioOu1l6zs9daLZqtz6RrScW3tXI7HJiA1r1333q7/sVdAveyk7rretC0dXde2rjltS8fN4V+0hl4p/NRWWYR8xvZbXTl45vARqx07CECgPAE2AMszJgIEINAogdiHLbWzD1UpKbnz1F+Kj9q2PWiszeQI8Xp+0Hevs17qlXqtpdrPceiV2VxOjEEAAusI+O4ta94TxI/PV0iV7z60Jm7If4l+n+Y1cVI4rfVvYyhXbVN8rpmT4h9bCECgDgE2AOtwJgoEIFCBgDzkyLf+SrzEt32ISo3hzre+3LFU39hDAAI3Ecj1oSyFp72WQ/P0Gtc2ZFe7P0Z7bU3EgwAE2iFQ8p6l9x9tfVn3vuk0l5svX+nTnNe+n2nNYmKvibFmTihX+iEAgfoE2ACsz5yIEIAABE4I6EMaOCAAgWMTSPlAlWJ7bKpkDwEI9EAg5p6mNtq2mNcabUtzlsa3cuA5dCtB5kOgPwJsAPZXMxRDAAIOAXlAKv2Q5ITkFAIQyEwg5tsKe35Yce8zubTovcv17+JdGnftOQ8TgGWYjR3RtWn7OIZADgL2fm+P53zremzh+k3VvJSXzW3OdsuYxtjiQ+bG5r41DvMhAIEyBNgALMMVrxCAAATOEMj14HXGKScQGIhAy9fIkrZcH4hi/MTYDLQssqdia2mPswfCIQQg4CXgu4f5+ryTIzrluq5xbfs027j2OEK2V7MvRoyvJRvxG+tbbbVd8s04BCDQNoGL2paHOghA4GgE5IFJHjLsg1PMQ0qMTS2WVruNqbnZPo7bImBr19KaaotSfTVuLex56ZpZ/5K5jS3n7rj0zb1S7dWXG1f7abcRkHrANryOc/Bx1zy8t63ZHmdLzZfWkq4Ld72k5GvnLsVL8RuyVc2+cR2zmnx2tfpUT614xIEABNokwDcA26wLqiBwWAL6gCKtHi89PKldbWhLulw9e+l0dXAOgd4J2GtP7xUlri8bR5htjWH9bfXVcw2FQ+gnZ17nz58/dSfHlv/pAAfFubhrnTocc9G56yBEIdbOnT/iuto7p7W1cGvDOQQg0A4BNgDbqQVKIAABDwF5+GjxAUQfyqTVH5Gv/W4qLebgauS8HwK65mybU33N605zSNG/5XoKXaMaX/X47OyYHuu8uTbWVu2kXftSHynzt8RLiRNjm1NLzF+F37KWYvIZwWZrTXzzfX0jsCKHPARC70Har9etrCP9yRM5vxfVrJ5Vu54vtan5pdjP2abqXMqDcQhAoA0CbAC2UQdUQAACHRGQBybfK9Tvs6WvTQL6oN7jg+/cg3wK7Vy5L10PKXqXfG3NT7XkiBPLL0eslLx9tpq3jNXUo9eZT5PLTzWu0We/ARgTy2dD33YCbk3n6r89Gh62EFhznW2JV3qu5LN0Hyitwfp3rwU7lvt4qZZz4yGdc3Ny68cfBCBQhgAbgGW44hUCEBiQgDz4pD788EFnwIXQaUox6zf00J+asl4noZg6Hus3ly6Np7q01X5t11y3do4eL7Uaz9emMhIfdo7mZlsbx9pKv2it/XJjuueuxlAuId1z3wB0Y4V8HKG/NAuto8QpHesI9SqZo9RH61UyTi7fuqZKrCu932zVav3Y4xS/ufNbU+M1c1JyxBYCEKhDgA3AOpyJAgEIQAACCwR4uFwAZIZjP/Tohw3L1tdnXBc5tDGtlthga+bE+rZ2ylX73HPt11Y/lGmr/blazdvym/OdokNz03bOb6kxyUvj+7T7+lSLMnHPtV/alr75ozp7boVpzEtrYG3namntOD5LwMfyrEX+sz1q5ebpnmuW0h96Wd1yrD9z/yMg5Ctnv9W15Fc1u+3SPN94SlzffNtnudtja2OPQ/WzNhxDAAL7EGADcB/uRIUABCAAAQ+BmAdLz7RDd7kP+frgHcMyxqYVuHtqtR/GXB4uf3d867mbt3tu/YfGVL+1HeVYcg7lPUqONfKIWccxNqLVtXPPa+RztBg9XwNW+9brWdYa62376rc1ifWmtbNz7XGsH+wgAIGyBC4q6x7vEIAABI5JQB5Ae3/wCekPPVy79iG7Y66IY2Xd+vrPsTa3+nCvl60rJORvq86tukLzQ3pD9tKfY12l8hCdqXPmcjjKGMziK63XwlpmOl9b8dPTulW98cT8lrlyVo4axfVrx9fWTH3bNpevnDytvrljy2TOjjEIQGB/AnwDcP8aoAACEOiEQOjhTPrtj6Zj7eXhSH90vOV27mHON+b22dxbznMUbZa3by2G8rTzQjat9PekdS2zpRy1tq5/9/pzx3s+d5m4525uIUbya4BLc11fwjV1juujp3MfO81fWIy8zvaqUwxTrUEpjS3UNleOpXLROmkrtUjRnGKbq85Way6fqX72yDtVI/YQOBoBNgCPVnHyhQAENhHQhxlp9cd1KP36EKqta9PCuT4caqua3HPtt63mpa0dU0a2L+Y4xDNmLjY3fRhx2StT2yor11b7c7W+mCm+7fzSWlN0iW0OPa4P99zVNDc+N+b66fFc8rM52mPJR8dtvxzn+Le/fPe4HhmmaHZ5Wq5H5JHCLsVWWNqXsrWtZW9t7bFrI/PdPmvvHqfYunNznsfoiLERTZZhqkaJYePoufiUl55v9Zs6P8Xe6rfzlIvtizlem7P1rfxsH8cQgEB9AmwA1mdORAhAoHMCoQcrTauHhxzVaFs51nPNhfYmAiNxyfEgf+R1sXT9p7BRX9r65sbWa86Hz28rfVuvrVJ5b9XVCt+WdMC0pWqc1VLqOjobZflsiw7fXF/fsoqbLGSunW+PY32k2JW4PsSn+tU2RVNu2xY05M4JfxDojQD/BmBvFUMvBCBwAQF9oCj9cHZB4ECH6FBNrkltjSEdrq5c57Xzy6U7xo+wHDm/GAZrbeauCdfn0Ri7+brnLp+5c51b+7qf07Q0ppqX7NzxtfNcP77znvj59Jfoy8E7h48SudX2GbO+crPSmLn9lmAnGl29er4Ur3R+pf1LnqViWK5LHN3xOf5zY9ZPqbxsDI4hAIFlAnwDcJkRFhCAQMMEYh88Wkih9sNPT2xaqM+Shtr1W9Iz6jjrdntle1irorEHndurgQcIlCdwhGvpCDnKSsnxHuj6cM9zr0i9nx+lRrn54Q8CNQmwAViTNrEgAIGiBOQB5/z580Vj9OBcOJR42LMPeKHjHvis0chD7RpqZ+ekMCyxfs+qGf9Mr9HxM82X4dy9kzWZjzOe/AT0mk25V/o8ufN17br9vrkt9Kle1aLn0uqxjuVo1W8J36Jvjf8tWkJzQ/2xDHV9ak7auuvKPY/1jx0EIFCHAL8CXIczUSAAAQhAAAKHJyAfDLZ+CDk8RADsQoAPtbtgP0zQHOsr5CPU3wPc2u8XEi8Hr5DuUH/LtZjjMTfWck5og8CRCbABeOTqkzsEOidgH6TkIcSe29R8/Ty0WELLx/BaZrTGQtbmnmz12qipQWJp3Dlme7OZ08aYn4DWteZ68ivJ0ztKHnlobPcy2vqIJXLUvGP5qJ1y0vNQq3a5rk/3PWmrf53v6le9oXHXXux0jjs2d+6bExvT+vX5kfFQf8xca8MxBCCwDwF+BXgf7kSFAAQyE/A93MivA/v6M4f2upMHpJiHJO/kxjpHyaMxrCdyWmEr18le10qoLq2wCemj/yyB1tbPWXVpZ7L2WH9pzJas59aH3n/mbJb8tzY+l5Our6OvMVtvexxbyzVzQr59tdAahub4+kOarH+tv29+S32hXFrSiBYIQCCdABuA6cyYAQEINELAPlCJJPe8EZknMmpqqxmrJcZogQAEILCFgNw7R7t/rtnE2MIwNNfHtRVtIc01+kdcczHcdHNJ25g5pW20Fr61KrGXtC6NW/0ptnbe2uNQTq6/rQzUn+RXO0eNTQsBCMwTYANwng+jEBieQO9v0u5Dzblz55qqmT5M1RblcnHjz+nSMW3duZyPQcD3cF7rfqBra64dg3KdLHy1rBP55ii2ljf35j+qtUbzK2/Lo3Kca0WxO54jC/WZw1cvPuT64HUzgVz3i1JryadP+ubixdZYfCy9NP6SXcp4rD716bOXvlj96ocWAhBoiwAbgG3VAzUQqE6gxENG7SRGyKEEM9/DW4k4+OyPwNID/NJ4fxmPrfgo1zrrMn0dh9ZGqN9G8PGWPl+/nRc61pjS6nHItvd+zVHb3vMprX/reli7JtfOW8ujdjxX51bO4i/WR6ydq5FzCECgLAH+CEhZvniHQPME5GGkxzfpXnXXXhA91rY2o9zxfA/4rdTBp83mrzrFTm21z9pxvB8BrYurILZOdn7sHDdW6XOrsXSsI/j31XkrY51vfWufMLX9sYzXzIn1jV37BKT+dg2VVqyxpI1de3N26i+ke2lc5s35D/mlHwIQgEAKAb4BmEILWwgMSKDHhw19iNI2VBb5deAe85N8lnIL5Zza3yuf1Dz3tG+F8dKaakXnUq2W8liaz/jNBHpl2ctavZl0W0c56y6+9Cc1S1tHe5zqB/uzBHLWVzzb+trjs1G3n23RvWb9rJmzJstacdZoqzFH18yW+tbQSQwIHIUA3wA8SqXJEwIDEoh9qBI7++ARO28PZKqzpkbLp2bcPfjWiGl51ogXE0PXlc/WV3Obg8z12fh81ehrSUuNfEMxLIe5+rrzdZ47xz1XO3d+jXOJ7eqRuHtqqpF3jRg+rrnixtZHNKittqpB9bn9Ok4bJqDsxMIyDs9IG7H+02YuW2/x3fJasdrs8TKRMhZLnHOvGxuvhfzLUMUrBPoiwAZgX/VarfaDH/zg9M53vnN673vfO330ox+dPvOZz0wXX3zxdJe73GV64AMfOD3hCU+YHvzgB0f7F19veMMbJvErvu54xztO97///acf//Efnx72sIdF+fnKV74y/fEf//H053/+59N11103feELXzjR8/3f//3TU5/61Om+971vlB+MjkdgzUPEmjl7kBWd8sCU+yFsKZde+Czl0cp4LzxjddZej63UsTUd9sOU1RZbx7k54qPlOq/J0ebL8TIByzi01kJe7NyQTYrPltdiKL89+31sSzGMqXUKC5/2lPlztuL7rne965xJtjHlovnoebYADTrSXEXaEfJtsARIgkAyATYAk5H1N+HJT37y9O53v/sC4V/+8penj33sYyc/f/qnfzo97nGPm1784q3kMi0AAEAASURBVBdPl1xyyQW22vG1r31teuELXzj9yZ/8iXadtJ/85Ccn+Xnb2942PfGJT5x+/dd/fbrlLcO/YS6bhs961rNONhCto/Pnz0+vf/3rpze+8Y3Ti170ohNfdpxjCByBAA9R/VRZHn57qJdoTHlQt7b9VGO7Upt3y3Utpa2U35TKaA1a0JKiu7RtaS4ubznXmEu5uXPn7JdsdVxiy4+ez/k8+thcnXIwLFkDn3Ybzze+VG93jny2kC8s+F42lm98TZ/4dDWs8bPXHNEew8XN0T23PuzxXnkRFwIQuIkAG4AHWAmf+tSnTrKUb/s9+tGPPvmm37d+67dOspn3/ve/f3rNa15zsnn3pje9aZJv5b3sZS8LUnnFK15xuvl3v/vdb3rmM585yb+zJm+ur371q6ePfOQjJ9/qk28E/vzP/7zXz1e/+tXpuc997umb8SMf+ciTjb473OEO0z/8wz9Mv/d7vzd9+tOfPtkAFM2x3yj0BqMTAjsSiH2I2lHiIULrQ2nuB1D120udJf9cWnP5aXUB5l4rufJsVVeu/PSayuVvJD8la7/Vd+z9ICVOiu1IdU7NpddrplfdsfVZWr+x10xsvJx2S9olVkz91CbGX079+IIABOYJsAE4z2eI0Xvf+97T8573vOlRj3rUdKtb3epMTt/1Xd81Pfaxj52uvPLKk28CvvnNb55+8id/cnrIQx5yxk5O/uVf/uVks1COv/M7v3N63eteN9361reW0+kBD3jA9PCHP3x6ylOeMn3oQx+afv/3f//k14rvec97nozb/8i3+97znvecdD3pSU+afvVXf/V0WPz84A/+4PSjP/qj0/XXXz/9xm/8xvQDP/AD00UXsVRPIXGwikDthy198FkllklNEJhbM2597XnLD7ux2ubsNFdt52ybKGSkiNHyiUy7mJnwXLM21swplsQBHKfyTrU/AMLqKeq9Kjbw2msx1n9uu9T8bPzYuT47Wdu2v9Ra1xjSlophmcQet6TF1azMWtboauYcAq0SCP+OZquK0ZVM4JWvfOX0Iz/yIxds/qkj+bbeL//yL+vp9Na3vvX02B78wR/8wck3BKVPfg1YN//U5rLLLjvpl3P5JuH/+3//T4fOtPKNQ3nJN/5+8Rd/8cyYnMim4bOf/eyT/n/913+d/uqv/uoCGzogkEqAh4ZUYmF7eRDTn7BVOyNS+7X11zzdtp3s6itxeSqb+kryRnTzyuv9mN7ctaHnvlYIrb1Oj0k3T9ZSi9Cr9DUxFzuk6ej9a5npNdcKP11bOa55zW0tG2GiOnLqaoW11WHzc4+t3dKxzl2yyzG+pa454uMDAqMRYANwtIquzOehD33o6cyPf/zjp8d6IDffv/7rvz45lW8UyjcHfS/pv9e97nUyJPbuTVu+RXjttdeejMuvI8umoe/1+Mc//rRb/l1BXhCAQJsE3Gt8jUrx4f6s8ZNzTo68cuppyZfWqiVNaGmbAGumnfrIB3d56Qd4PW9HIUqORkDXYq28Q+/vte5Tmu9Rrr0Q75h627lH4RXDBRsIbCHABuAWegPNvfHGG0+z8f3xjn/7t3+b9N8S9P168Onkbxx8z/d8z8mp/FEQmWdf+qu/0qd2dlyPv+VbvmW6/PLLT07lLxfzgkBvBPQBrzfdtfXah7vasefibXnQbDWnuXxzjB0pb8nV/cnBcBQfa9bCmjmj8KqdR8r9za2Le75Fe4qOLXGYezOBnPW72Wueo73Xg8TfW0Meku15WbvutCbUpb2aoqhfAmwA9lu7rMrf9a53nfq74oorTo/14KMf/ageTvINwLmXHb/uuuvOmOq3/6TT2p0x+r8THf+P//iP6Qtf+ILPhD4IQGAHAu6D2NoHu1TpteKori0PnrW1quYarVv/GjF7iDFyzXvgj0YIjE4gxz3mSPfv2Fxj7UZfX+QHAQgcgwB/WeEYdZ7NUv4a8FVXXXVq85jHPOb0WA8+8YlP6OF0t7vd7fTYd2DHZfPOvqyfu971rnbogmP5S8XykgcemacbghcYBjrudKc7BUaO1S1/oVn+UvNRXrbu9vgo+dfKU741bNeV/MXuNS+5D8gatS/x6/oTG4n5pS996dTUxj/t/MaBrbs9tjapxz6dSz7cHJbsexoP8dD6hGpTK0dbd3ucK7791rz12Ur+VtMexyE+S1qEX861Y2tvj5d0MH4TAb3vKg+tTYl7W6lnFVt3e6w59dSG7ruhHLRedrxE7az/rcdr7x2+uJKrMJO6v/a1rz0xkWOXY+tMfLm10ics7ct9nrNjue/v1vfSsb327fHSvL3H3T/eubce4o9B4Bbf2FwJ/+u/Y+RIFgsE5I9y/OZv/uaJ1SMf+cjpt3/7ty+Y8epXv3p66UtfetL/qle96uQv9V5g9H8d73znO6dnPetZJ2e/9Eu/ND3jGc84NZV+GZfXBz7wgenSSy89HXMPJJ7Eldc111xz8peHXRv3XHx++ctfdrs5hwAEIAABCEAAAhCAAAQgAAEIdEXg4osvnh7wgAd0pRmx7RLgV4DbrU0VZX//938/vexlLzuJJf9H5Nd+7de8cfVbDTIoN6G51yWXXHI6fMMNN5wey0EuP2eccgIBCEAAAhCAAAQgAAEIQAACEIAABCAQJMCvAAfRjD/wz//8z9Nzn/vc6Stf+crJN/F+67d+68yvzlkC9pt6S9+ws1/fv/Wtb23dnPnGn/ixfs8YfuNkzo9r6zt/6lOfOn3605/2DdE3MAHZyNZf9fCtgdCvJ/h+VUYxlfrVJPVP6yewlvvSGvBHi+sNrZ/Q7Ll1FZpD/3YCJdeAqEtZB741EJrvs91OYx8PSzmGxlVtDhal14FqrdG6vHLwqaHbxtgjh5HWgLJ0OWq/r+11nYhum6fmYft8+bp9Ms+ugW//9m8/+SKD+nPtj3wubHNwcWuUw+fWutg14PtssNV/qflXX331dOc737mUe/welAAbgActvNyc5VdzP/e5z03y7wu8/OUvn+b+uu9tb3vbU1JLf5Dji1/84qntbW5zm9NjObB+Pv/5z89uAM75OeM0cCKbf/qXiwMmdA9OwLcGZNPZ9y8fzK2VuY1qF6H6ln9UWo57/MelNQebW6481HeMvxTuVqs99q0BO556HFo/IT/y19DlFZNvyAf92wjErAFdlxIpplYp60DXQEwWPtsYPTG+a9vY3wbQ2DYX37jaSass7Bw7nnocsw5Sfdayt+tTYwqfXGzUZ8nWzWEP7T2vAVubpWtHbPfgazVuOdb3fnuf1Wc0HfP5j1lj8ltMwk/9+fwctW+ObQoTd33mvpenaPHZ9nQf+OpXv+pLgT4IbCLArwBvwtfnZLkRP/3pTz9585MHhJe85CXTIx7xiNlk7B/2sH/IwzfJjusf8lA760ffEHTMbfUPiIhGO8+14xwCKQTcB0SZm/tBWfz54qTo3Ms2pDvUH6NT52obM2ckm9zrayQ2a3M56lpay6v2PF99uA7WVyHEzsd5fZR6M0P51FNQN1KvdapLaT5aToZHW38+sjl5+vz7+vaI6dNBHwSOToANwIOtgM985jMn3/zTr2e/8IUvnB73uMctUrjPfe5zanPdddedHvsO7Lj7l3uvuOKK0ynW7rTTHOi4bCK63yQ0ZhxCIEhA1rk8cNiHDvfBzz0POoscUH/S6nHk1F3NXE4+MWrjttbWHZNzeWlrbXs71txidGv9e1oDMXntbWNroMc51pbWS/Jb8qdxa7FY0lNLx5Y4vuvA12dj6Hht3lZDrmPNQds1fnWNuu0aX6lztuiWWDJfX1pXPafNT2BUxroOtXXJxVwbpX4d1dXknrta9z4vuUZsHdw8hQsvCEBgXwJsAO7Lv2r0//3f/52e+cxnTh/96EdP4v7CL/zC9OQnPzlKw7d927dNd7nLXU5s3/Wud83O0XH50/Ayz74e9KAHnZ7KHyAJvf7zP/9z+tjHPnYy/MAHPjBkRj8ELiCgm9vugD6M6cOHPqC4dkc9X/MwaBkq36Pyc/PWdeb2cx4moGtIW9dyjuncmOtn7lzXdC5/c7FkLOa6i7FZirPn+Bb9Wo8tPvbKXddxrbVUKk+rX3NKjdVzHVNz9dnnXr+5/fk079mn68yuPZ8etfON1epzNfjOa2lpMc7oa7VF5miCwBIBNgCXCA0yLv+e3rOe9azpwx/+8ElGz3nOc07OY9OTG/gP/dAPnZjLN/Pe//73e6dKv35zT+zdG/+97nWvSb8F+Bd/8ReT/Xf+rMM3vvGNp6dLv558asjBoQi4D1mSvPTxgkAJAilrS+97OkfbErpG8Cl8lhgtjefmIDVUXW6bI5b413Wibchv7dxDOlL6Nb+l3JbGU2K2ZLtUs6XxVnIJ6Qz1t6IbHccgoPeZPbKVa4DrIJ78qPf6eAJYQqAdAmwAtlOLYkrkr+nKX/t973vfexLjaU972vS85z0vOd5P/dRPnfzBEJn44he/eLrhhhvO+JBz6ZfXRRddNIm97yV/fERe//3f/z299KUvvcDk4x//+PTKV77ypP+e97zn9MM//MMX2NBxbAL2ocsep1BZOy8lxui2wlB/Rs917cOrrrOjcFqzDkJslZkyXOO7hzmaX+kPs8qzNSa+vENrojXt6IHAXgR8181eWnLG1fthTp+5fbkatRa2lZh6njt+b/6EAy8IQKAdAvwV4HZqUUyJ/Krv3/zN35z4/97v/d7px37sx6Z/+qd/CsaTv5Al39RzX9L30z/909NVV101fehDH5quvPLK6Wd+5mdO/mS8/Nrlq171qukjH/nIyTSxu/zyy10XJ+ePf/zjp2uuueZkQ/J1r3vd9F//9V/TE5/4xOn2t7/99IEPfGD63d/93en666+fbnnLW04veMELTjYTvY7oPCQB98FLIYT6ddzXypxeH0xKaBcWazj62B65b45hibodmXXJ3H3Xg94vbB3n6r2kz861x0vz1o6r/rXza8zrQWMODnYN5fCX20fselQ7vV6OUr/cvNf6U+4yH/Y3/SbIHAf5vLLlr93qetd6hWKF+nXeUVu4HLXy5N0SATYAW6pGIS1/+Zd/eer5b//2b6fHPvaxp+e+g7vf/e7T29/+dt/QyTcH5c+nywaebPb5vkkoG4w/93M/550vnbe61a2m3/md3zn5FeQPfvCD01vf+taTHztB/oT8i170oulhD3uY7eb4QARCH47k4cF9AHPPe8QUysF9WArZ5cq5VjzJw42VK4dSfqzepTqIrc/G+iilsze/Pk6pOZTiautoY4SORXdMPj4b9ekbkz4dT2UTq2mNX+b4CUitfHV0rbfW1fWX6zxGu8bSXHWOtu561X47T49HbDVfl0OJXGvEKKF7jc+lXJfG18RkTh4C1CYPR7xAYCsBNgC3EjzYfPlW3kte8pLpUY961PT6179+kg28z372s9M3f/M3T/e///2nn/iJn4jatLvjHe84/dEf/dH0hje8YXrzm988XXvttSf/HqD8oZHv+77vm+TXlO973/sejC7pugTkATr3A4P40wdzN94e50ta5saVj7Yl9JfkVVJ3CRbqc64m1sZll3stayzasgRS6+bWPVbd0vVg112KJjvPp0XHU3z6/Gzt2zv+Vv09z9c1kJKDzJlb6zouPn3+7XhKXGwhsIWA/hVgXZNb7jtb5m7JgbkQgAAEthBgA3ALvU7m/uM//mN2pfLNvK3fzpN/J/BJT3rSyU92gTjsmoA+mEkS9tj3YcPXt5T8mjlLPvceV04lHkhdnxorNedeuUu+LoNQ7q6dex6aRz8EhEDstZWyJmPJlvAZGxu7PgnErtej3gePmvfa1RxaT8pRxq2N9q+Jt3auzBMNa+ev0cocCEAAAjkJsAGYkya+IACBogTsg58G8vXpmG3dhzX33Nr2dOzmX+PB1GXnanD5WXs5tvY19Lp61pxbzb75NkffOH0Q2IuAXnOhNbo0Xku3XmMhnbV0jBqnl3vtKPxZx2mV1Ot/bpbeq+ZsaoxR2zyUtebwzMMTLxCIJcAGYCwp7CAAgW4JjPpwoXnpQ1Rqgew89ZXqQ+xlrvWlPrb4VB97tb58fFp6ztGXT8t9lrVbHzm347nyWOtXtbg6c+kSPynaVE8o/tJ4aF7OftGwhlcKh5x6Q3FTc1D7UjVQ/3O5r2U/55OxeQKh9TM/a9xRXf+h9errh+EY60Frq2tAstI+m6Edt/2ubcjOzuEYAkcmcMsjJ0/uEIBAmwRyvnnn9FWCVgl97sOQT3eMjW9eqM/Nwz0PzZP+3FrmYsWOpeiP9YmdnwCsL+QSw2TJRq6rFq+tC7M929OT5q1al2p4lsy6M4mhP3MetmjxzfX1zcUfdazX63CvevjW6tx1Nje2Vw69xq3N0t4j7HGv/NANgV4I8A3AXiqFTghAIJmA/mPPyRO/MUEehGo8kKx94FrSZvXPxVjyk8Iu1pfaqS49T4lV2la1SRxXn465/aU1jexfWQpbPZ7L19poPbS1Y3M+Ysa2+pL5qkviqT+bpx23Nmobo9O1sT7tsfXvzvGd69w5LWIzN+7zG+pb42fNnFD8Lf3KKsaH1WyPY+a2ajNKHqX4+tZHzmunlO69/Mp68jHz6anJ0Woabc3vkU9KzNg6x9r51hJ9EDgCAb4BeIQqkyMEOiGgD1badiJ7tczYPOUBae4hKTQm/kMxlnyuTipxYkh7opti5j59rbArlvSOjn28l+S4c2TNnz9/fmlatXFdL1ane+yz2SJQ/Wu71pfq8s2395fQfcY3T/vsfO3rqVU2vefRE3O0HotAyv1rzT0olWaNGKmafPZHuyfpvVhZpKwbnUMLgSMRYAPwSNUmVwh0QKDmA1bNWDnQhx5qestDWYTy0fE9W/eBMoeWXuuUI3d81CdQ8vraspZlrp3vntcnRUQhYGuSQqTkOkvR0ZMtzHqqFlr3JhD7PMZ1tXeliN8LATYAe6kUOiFwEAL6Rl/jjXwpxtoPRGtLZXO3x9bfkmZrGzrO4SPkm/4LCdgNDj2uvbYuVDVWT8k1Ta3OrpUtPLbMPauijbPR8mmDarsqpN6pNQ/dm6Q/NNYugbaVaX1SaxSblVsvjWPjzh3Hxtlip5rEhz3e4rOVuS5/n66QTUu/FeDTTR8EahJgA7AmbWJBAAKzBEJv3HbSks2Wf/fPF6fFByjLIFWfnWvz5bgMgdT6lFFxLK+yxnPdB4Qc10ye9TN3LRyN8RyLPLTLe5Gaba1brxy26N7KrHxl24mwhfOWuSkEasWxmiSm/rj99pxjPwE2A/1c6D0OATYAj1NrMoVAVwR8D1X64KztUkLyJu/zszTPHc/hw/Up55qHtHrss/P1+R7+fHbSp/5TY4T8Ha0/hfXR2LSUr67zljS1pqXkPUCvE21t7tIXeqVqmvMVitFif2t5xNYh13XWWv4xa8QyStGvzOz8mHjYQGCJgKxD/fHZpqxT3/wW+ubyi9E3x0B9z9nExMAGAj0RYAOwp2qhFQIHIeB7I977wdmnKUc5SudV2n8OBr34KLUGeskfnfUJ5Fhzcg/Qn5oZiPYc+n2a1a+2auOeaz9tuwR6rlnP2ttdEWMqy/UsxppLXx+52KdHZgYE2iTABmCbdUEVBA5NwL5Zp35wdb/ab32lQnVjt/TgFZNXjE0qkyPYS531R/LVdZDKs6X1coS6jZajrJ/UNVeawZo1vWbOXB56PapfafVH5mn/nI9cY6n1Ue3a5tJR0o9q1bZErJo126pfONiXrr2ecrD6OV5PQK8Jd02ox6VxtYtpdZ3F2FqbkDZr08uxMlhzrdl/FkSO1ZfmPhInzYkWAiECbACGyNAPAQjsSkAfnEIifG/Wbp99ww/5iel3/cbM2dumR817M5P4cw+Wc2Ou9iVb6uMSa/98qaY5M5BYra2RmvlvZbmnVqmb/mzNY+/5pdagrz6+vr3zJ36/BOzadf/H8JasrF/rx/bLsT23djWPS2mQa1V/SuXjMnTPc8QtxSeHNnxAoCQBNgBL0sU3BCBQlID7QCAPJLk2/axw+8FEH3psn7Wteezmr7F5qFES21u33kt1V/u5yNRnjk67Y3vVbWnN1SCWS4Pes2y7Vv9e9XD12lxcTXNjrp+jn7vsWubh0+rrazmHHrStZbp2XiwT8e+LEeqP9dujXcwzz5a8XKbu+VrfWr9c/tbqYB4E9iDABuAe1IkJAQgUIaBv6Dmdz33w1QefOZucWmJ8lWAQE3cUm5iHQV+9dS2MwoE88hDQdWHbJc+6vrRdsu9hPOa6yp3HSPxys/H5c3nVfC+pGcuX+5q+Pdb0Gp3M2U5Arg33+tjuFQ8QgAAE9iFw0T5hiQoBCEAgH4FSHx56e+ArxSFfpdZ5snXYO0dffKtvXYbtzJJcfDm2o7AfJaF1Eer3Zebauuc91Ko1jcKwNU2+2tfsc9eVxC7JaST+kouPX836EesmAr51JbW58cYbsyAqeU1kEfgNJz4GuXyP5EevWXiNVFVyiSXANwBjSWEHAQhAIECAB4gAmEzdrfDNoSOHj0xYvW5a1+cV3VCnfqgQSTlZil/rW1MO9et4rjZnLjk0LeUtekOaQ/05dPXow7euNI+5MbWp0S7Vu4aGUIxWGIX0relX3m67xlfOOXpdayu+7fWsem1MGbf/NIz8e4BLNfP5sT5bPLZMcuhLYWBrkCN2DR+5edXQTAwI5CLANwBzkcQPBCAwHAF5QJCHoNDDzdzYcDB2TihUg1qy9o5fK0/irCcg9wNeZwnscd3sEfNs1pzlJGCvq9bec622nDnv5avlfHzXta9P2MWsE83V+tA+5e+ea3/tdkmHzSGHtph4apM7dg79+IAABOYJsAE4z4dRCGwiIG+QvDluQrj75FD9WqptSOPu8DoWANOOi7eTdFkz+qFoJwlFw4auiT1yDmkpCqAT51qPJUZqt5SW2C35ivEhNmv8rJmjua2Zu5SLjmsMe74lnvrb4kO1rGk1/pq5Jeek8ki1bzVvZSr6JCdttb9UGxsnlXMpvfiFAATSCbABmM6MGRCIItD6Q0VUEhhBAAIQGJRA7AedQdPPlpb7QRCu2dBucuQ+g9g6uWObAjU2udfcbH32QCrxe2bXq/alWuu60HbJfut4bBzlHWu/VRfzIQCBfAT4NwDzscQTBM4QkDdF3hjPIMl6Ig8f+gCS1XGkM6ntnvEjZWK2QKBmDbkfLBSj8nBqPWLWiuvTPS+VYoy2UrFdv7VyduNyPk9A1oj+zFuWGy2xTjUnbUW9rMFe12EJRjkq2jrPnmueoz41fdg1ao9raiAWBCCwngAbgOvZMRMCEGiAAA8f+xdBauD+7K8KBRBog0CtD87cC9uo94gqcq2tGtdCjRhS41JxXL+52MesSzd2zBxs9iFg14Uc2/N9FPmjtqzNr5heCIxPgF8BHr/GZAiBIQnog6o+XOh5zWT3iFkzv6VYcw+cdqxlTjm0iQ+bb4ib2OSIF/JPf1kC1K4s3y3e7fVHnbaQ9M/dynTrfFFlayzn6lNb6av90tiqTVvRoWOqyY5pn2uj/baNsbH2W48lnqu1lAaJU8r3Vg6h+ZbNnto1ttUT0py7X2JrXNURirE0HppXo7/H9VeDCzHGJ8AG4Pg1JkMIDE2g5YeLocFHJEdtIiBhMhSB2A9FQyW9czLKfGcZw4bfeh+39dniS+ZaX60DF61zmrewaD13q8/WrHbOWgOrZ7Tj2kyV315xNf7WVteltr3ns5UH849FgF8BPla9yRYCQxLQN/Ahk2swKeG9xJyHKX/hlrj5Z9HbGwG9RrTtTX9Pet17DddYvuq5bJc863q37dKcteNaZ23X+tk6by5+aCyV61aNqfOtPnuc6sfNX85tn/i256n+Q/ZunJDd1v4S2rdqYn4cAbuu7XHcbKwg0DcBNgD7rh/qIQABCFQnwMPShchhciETeiBQiwDXXy3S6+JIfXLXSPz1tgFTgsO6iizP2lqvUG1cv+75sjIsIJCPAOsvH0s89UOADcB+aoVSCEDAQyD0kOkxpSsjAR6a0mAKL/1Jm4l1TwT2vC72jN1CjfT6CnE4f/58dxtGa7gqB5mrxyEmc/59c+beb31jGn8uTsqY+vNpS/GT0zZGi+rOGbe0L189t8Tci4HkIdd+qVduTqV04vdCAjHX7oWz6IFA/wT4NwD7ryEZQOCQBNyHLjkf7c3c5thibqKpdY01Lw6XR83YxGqHgL1W7fVRUqGNWTLOCL7dmozKzs1Lz938QzUVO50jNjpP29C8mv1WX824bizV0RIbV+Pac3cdxPhZMyfGb6s2R8u3Vh3gWos0cY5GgG8AHq3i5AuBAQjIQwGvNgjoB5821OyrgnW5L/8Wo8v1UfoaKe2/JFeumZJ0/b7XrBfq5Gc5au+WesvcNWtsVJbklU5A1tCWNZgekRkQOBYBNgCPVW+yhUD3BEIPBa09cG59gAnl2WIBhX1r/PfgBIM9qBOzZwItXDM93Wtr13rr+1htvXvH861nX9/eOpfiW832eGmeXkvaLtkzDgEloPca1o4SoYVAOQJsAJZji2cIQKASgZQH1NKS9CFG4/AwoySO0ba0Fo9BnCxj7zHuvelo5M6dO3e0lIP51rpPHXHN1WIbLG6mAcljbS5r52WSjpuBCMS+vw2UMqlAoDgB/g3A4ogJcDQC9s2Kh6Dy1W+d8Vp9a+eVJ342gq73XvSeVV/mbAsL4bllfpmM8FqDgNZdr6kcMXP6yqFnbx/KeG8dLcS3LEqvE59/G78FHjk1aG6+vHPGac2X5t2arhg9oj2mXqk5qs/UeTGaR7KJ5T9SzuQCgb0I8A3AvcgT9xAE9I3/EMmS5BkCPOydwcFJAgHuGwmwBjLVuqfcO8TWtVc/ITRL46F59EPAt96gMk+A622ez96jdk2799Ic2kr4zKGrJx+1rqGlODK+ZNMTV7QelwDfADxu7cm8AAH3jYE3/vyQe2Lak9Y1lXLX+xofvc2xOZeobwmfvTEeUa/W1a4fX55L43aO+rR9cmz77bFrN3fu6ljrZy4GY/0SkPXgrpF+symjfAQ+kkNv1/5W7nNrey0L9bl2fpkVut1r7vWhnHzKtK4lGYZ8a2zVlTtv9UsLgVoE+AZgLdLEOQSB0JvHIZIfMEn3TT82xdA6UH/S6nGsz5btQvm2rHmNtpFqtiZ/5kAgBwGuoxwU8dEDgaO8N7ZSi5bvLXNroWXdtWs7x0m0CCt41a4K8UYjwAbgaBUlHwhAYDMB+4CR+qAR8/CiAm0c7euplVyX8u0pH7RCoBcCPV93PWvvZX2U0pnzPSv1vbVUTqX86jrXtlSckn570b5lXfrWoS9vX19O9j4d4l9zC43n1JDqqxQT8bvkuyYP1aOatE3lhT0EWiHABmArlUDHcAR4g+izpDUfKvokhGoIQGANgVz3lqO8t+TitaZWzJknQG3m+fhGYeaj0m5fa/fZlvTUWsuSs/60slJaqkMrTNDRHwE2APurGYo7IVDrDbITHMkyhd8ahufPn0+OJRPWxnODhTSH+t35LZ8ro6255PLTMiu0QUAJ6HrXc231g4222t9Su4c2y2vrvaYlliNq0fUhLS8/Abue/Rb0QuDsv91qebR4be2hyd5r9Nhy4hgCEIgnwB8BiWeFJQSCBPiQEkRTdUDqcOONN57GlM3ASy655PQ814HW2/cQJH06nive3n7W5mPnWVbKyPbtneNR40uNqEO56ttrIFeUGjXLvSaUg/Xr5qE2ysnaah9tWwTcmrWgTjXtvX5URwtMRtfg1joXe/Ervlz/uXku+V8az60Hf/MEfOuLGs0zY7QtAnwDsK16oAYCENhAoMQb8JxP30OAyJ+bsyG94lMlH5uTe24FWDvpV1vt11bnuOe9MtJ8RmjdmoyQU085WP72eCmH3q4dm5s9dvOQc+3TdokF4zfde+HQFgG7zq2yUL+1aelY9PamWfjlvH+Irx4ZtLSO0AIBCLRFgG8AtlUP1AxGQB4acj6IDIZnNp0WuC1pWBqfTfAbg1vnL/lPGbcPuPZ4zkfIztffUq5zOdkxyUN1uzlpv7XPeVw6nvVv88yZw1F9WbZLDJS9rKeYeaXX3ZLe1PGYnFyfveXo6q95rnx1Ha2NHbv+1vqvOa9mLspf84uJbee0utatRslNz2vo1VjKVNo1cWNqEeNb9Ui7RofNY8vx3vGtdmXi8mtJo9Wb89jmbv1q/55rxOrhGAJzBPgG4BwdxiAAga4I6BtwDtFLb+JL4zk01PKRk1stzbXiCBvlM1LNa/E7Uhy7VlLy1vWVMgdbCAgBvSdpC5WbCOzFQ65liW1/5moy4rW/9j4onPbgMRfTHXPP52pbYmzv+FtrJPpbyKFEbfAJgZ4I8A3AnqqF1i4J6JvdXg+kXUKrJFprIjXS45jQS7Za8zlfSz7m5jJWl4CtlT2uqyJfNHd9jpBTPjrpnlye6R7CH3xHrI17v3XP1/A78pwcayTHGl5Tg95qX4KT+sxRxzU18M0RLapLxmO12Tn2OGa+tXc1yViMDztvzp+1k2PXf8pc19fcufWbms+c35pjc7p9Y8rW5l5Tb+5YkuMoueRmg79+CPANwH5qhVIIHIKAvLHqT2rC8sZ87ty5M9PsA4kc2x81tDbal9qq5tgHg1i7VB2p9q3oSNWNfR4COdZ+HiV4cQmMUBu937q5ufcd99y15zwfAZe1e54vUpynvePHqQxv0rvz3Xx6vY712s2hX5i4XCy3uTFrF3u8xp/O0TY21lq7WnHW6ts6T/LTHLWN8WnnxdhjAwEIpBPgG4DpzJgBgQsIyAPS0hucjLsPUr6+C5xn7nB1upoyh9vkbisf3QxsIUffGtma3xa47jrY4mtubgvs5/TpmPLQOkm7l3bVotpo2yRAneLroteSZWaPxZN7rnPio5SxdHVJlFa0+TJWvVaj9qm9nouN2kmfHrt2ek4bR8DHcm6my33OtuUxyUPXVozOWNtafGL0qE0OTeIr1k+sXQz3Vm2Ubav6VJfWwurVPrWhhUDLBPgGYMvVQVtXBOTmH/sGIG8a+sahbSvJtqanFS4ldMSulxKxa/hsMT/32rPnwkTOW9Wdo2bu9e2e54gxug93zczlK2tp7Xry1SYl9pyuvcaUh2330nL0uO66dM+Fj69Puc2Nqc2RW9/16+NxBI6S49o8U+fFcvfVIqUv1724lt6U3Pa2Ta353nolPnVsoQpoiCXANwBjSWEHgUgC9o1L3hDsuetibsy1LXXue9PSvtr6NO6aXJdYr/GZMkdZzeUgNjqu9ikxcthq/By+5nzsld+cppixPXXb2uhasW2M/jU2EnfPvNdorj3H1iY2tmWqxyE/Oi6+rY3WxvZZGzsvVldrdpqDm2MLOlvUNMfF6tVj4Ss/ej433zdm66PHarfFr/qQVrS5vu14zmPlkBovNlfrV2Pl1B/ry41tdcX6WGOnnHLF0zzUn5zrsatPbd3+kuc25hpdoTklNS/5nmO8NFfGLRNrv5Tr0rj11dJxr7pbYoiWugT4BmBd3kQ7GIG5N4W5sYNh2pyuPmxou9lhQQdSd2pfEPCC65746zrRVlLbssa3zF3AOvTwGm62ZhaOr9/tc8/XxLcxezmWvO1PC7pVj6tFaqI/7tio5+66HDXPUF4lrsNSTEtoDXHx9cfmpddXjL293nz5+fp82kr2qcZYLTF5l9Qb8l1KVyyXkC76IQCBPAT4BmAejniBQBKBUm+uSSKMsdXT2hu06rEajfRNGyLWT65j0amarU/pC+Vg7Uoe+3Stjae5hHy2kO9Sbm6tNKeleWvGQ5xUg7Qx8WNsQvo0lju+xafra8TzELe1ucb405q46yZm7lpdrc1r5R6iNdCaKCft19YdV7uarW99qD7R0YJGH4+autbEsgx9+kNs3Vjqx+0P+VzT78bQ8zW+as5xmSzpXhqvqd0XS/S5OfnsRuyTvFuvTynuNu+j1r8UW/zmI8A3APOxxBMEuiDgviG5560mIW+qvh9X7/nz592uZs7tg8EWUZZDrJ/U2LIu5n407tz6UZ1qe8R2jsEcu1KstKbqfw8NGvvIrXLXNsRCx6V1j/U8NLfnfrluesqvB62lNJby28v6lfxjGaTYrs3fjeGer/XLvHQCoff/FmsiWnO+NEdp9WWPtS+1DTFN9bPVXnm5Ofny3hqL+RDITYANwNxE8QeBDgjYNyx9E/PJnhvz2dN3EwHfA4D25WAUWz+NlVLHNTqtHo1p25T4dl7t4xI6Q2y0X2Lqce1894pbO89W48HfX5nY61DsYm39kbb1Sv30Z5un/mdvWctb5vZPjgwgcAwCOa7zPe/3bpW497tEOO+JAL8C3FO10AqBQgTkTVXezFp6cy2U6ma3lpF9oLH9EkTHtE0JbH2F5of6fXHU1vqds/ONLfXZGHosc5ZiLvktPS5aRaPVnDvmnO+5sdw6rL+94loNvRyvWcM6JwfnHD56YS06Y/JVvmJf8vrVODGaREsLL9Wq2kWT9pXUpzFsXI0XGtN+tYtp1f+auTH+sYHAEQmUvJ5y+NbrXmvjnueIob5pITA6Ab4BOHqFyQ8CkQTcN9PIac2ZnTt3rrimpQeNpfEUgVKXXLXJqSuUQ40Yodhr+/fWnKu+a/M/4rxazGvFOWINNedS1+8otSvFR/m7bUw8sYmxc33vcT63DnrJYQ9uW2PCdivB9Pmy1n3r3deX7r3sDKvRHpeNincI9EmADcA+64ZqCFQjUOONVGLUiJMDmj6USquatc3hfy8fmlfu+KX85ta5pz8Y1aWv16u2S9FD9ZF++7PkJ+d4rPacMVvyFarJnEb592GFm/0J2bt83fPQvJb6dW3uqcnV4J6v0ZbDR0xcXSchW9HBqyyBXIzt/xiWY11Dti2bSV/ede1rK+pT74E6N3VeDKnQutCYqrdE7Bh92ECgdQJsALZeIfQ1SyD1jSXVvqXES2ov6bskQ9WtrcYKPZjoeEybw0cojutbzt2+0Ny1/aX9r9XFPAjEEohZwyEb9x4RG3PJLpdf8aM/bsxcMVy/Oc6Fd4j5Fv8t57wlL50bqrWO52hL1CWHLnwcj0CLa3H0e0zJVSb1bIGfatC2ZM74hkBuAmwA5iaKPwgECLT4EBKQekF3z9ovSCZDR403fGFufzLI9rqYy0XG5sa9Dp3OHD4cl5xCYDMBvbZSHOkcbX1z58Z89nv3ude3e71KPj2+NI+lvwqvdrYN5Ss2vOIJ9Lp24jMc29JeE75jX/bWzjdes6/1e7GyqslkKVbOa1Z9absUe2lc6ync9n6pBrfdWxfxIRBLgA3AWFLYQWAwAilvyqUeVPTNMxfalJzWxsytea2OLfNcTjVy0jVUI9YWNswdn4C7/mtkXGrdl/JbgwkxINAjAbl/yM/I115MbmojrR5rPXPcY31+1f9SmyP+Uow147p21szNNUe5auv6zckupy9XJ+cQgMB6AvwV4PXsmAmBk4eeXt7g9AFtTq9vTOdtLXcuPyEdPu1qK7HnxtXuyK3LyNbLHYvlZH3EzsEOAiMQkPuNu/7XXkdLPDQO97ibSfn43zx6zCNdJ8fMPj3r0BqyHPV4tGtP89H8QvR0XO21Ddnn6NcYGtv61DHbF3ssc30+Y+cv2fm0+fqW/GwZT8nP8qitcy7HFnRZDXNaGYNAqwT4BmCrlUFXFwRaelPcCsyXi/uw4J5vjVlrvi+3NbHn8s8VY42uXHPm8kuNkdNXamzsIdACgdr3hDXXXG2NLdQFDRCAwDwBuZfE3k9K30PEv/1R5aXjapxSbW39sfW0dsq9FIO1flvVtTYf5kGgNgG+AVibOPGGIVD7zXsrOJ9e7bNv+EtxxFbnLdnGjFtfKTpifNe2sbnUjp0rnq2B5mP7csVRP6H1pDFVg9rTQqAXArqGVW/OtSy+XP8Sx9en8bXNqUN9HqEdjZusldI51YhRcu2FrjM3Zu952nxi7iFiX3rtxPiPsbG5LR2rv1gGS/50XP2GzrV/bevqdeOJX+lz7dx4S+Ouve9cffg0+Oz37sulV/z0kvPezIm/PwE2APevAQogsDsB901L3xBDwnzjrg93roy789w57rn4cOe4fvVc7Hzzdbx0qzr31JAzR5uHPV6KoRy22MX6WIrBOARaIJBy/aToFb9rrhWdM6crxiZFK7b7E3DXy1z9c6qVtVQrVk7d4kuvA+vX5uIbt7Y9Hs+tE1tLe9xjniHNOWtq10ooXon+UG3m9Lh56/ncHFe7zpH+kAadY22lLxQn1k79prahuEt+ZJ7VttbPUhzGIVCCABuAJajiEwIdErBvZB3KR3IDBFhDDRQBCc0QGOEDgVzTI+SxdlH0kntsnWrlM/J7gWUtPHPlqn5q1WjumghpsP32eM5Xb2OSl9aiN+2qt5Xa2GtFtUnr4+uz9dlZP3sft8J5bw7E748A/wZgfzVDMQR2ISBvdHNvdvJGHftmPefHTS7FVuKfP3/edVH1PJZBVVEVgqXk7drKuf1ZWmsV0mk6hMuvabGIK0og5f7oCtFrzu13z3teb+fOnTtNxx6fds4cbGE747bYUGw9cwvQuLZ1Y+iY2z/C+db3K5eNez4CIzcHzTHUuvY1z0VTjy9dh+59KzYfd54w8PXNsbEa9HjO3h1z14M7zjkEIJCHABuAeTjiBQKHIBD7IBEDI/XBIsZnSZucuZfU2aPv0FoI9feYY07NvXLhGsq5Cvy+ZG3oj1j41oqvz++t715fnr6+vrO88Ns0+iG6dF614pTOI9V/rTU08v2yFsPU2ua2L11DvQaXeMbqUD/S6nFuJviDAAT2J8CvAO9fAxRAoGkC8hCgDxlLQpceGJbGQ/7tvNgHmZCv3P2qrTVdW/OUfDS3rb7c+T6/vj53Hud9Ehjt2mitCqFrR/u13aJ7zf1gzZwtGmPnCo/QmszBKlZH63YhRlt0q88eOKvWLflunSsaemC1Jk+bl8t6pLxL5OLyCvG3dvbYsg/N9fXnykXjW02+eG6fznP77XkOjSFdMfGtFo4h0CoBvgHYamXQBYFGCITeCGPlyfytPmysPd6AffpFh9Wi57bP6m7x2GrWY6vTl7cd53hsArnqb68J8ZnL79j0y2Zna1IiEjUuQbWeT+rnZ136uvFHHb+3Fa69r3vRb39CK8fNU8+1Dc1b2x/yK3WPrX1uu7W5MA8CIxDgG4AjVJEcIJCBgL65ht6o3RCuvZ6rXawftW+19eXh5tqq9jldsTlo/rH2czHXjqkGmb+njrX6e52Xi7WtX68scum2LHLxzaFNtVh91q+O276541T7OV8lxkSf5Nq6zpjcNZcY2yWbUP2X5rnjyjXkT/vVzp3fwrlytW0uXZp/Ln+9+4FHfAV9rHx9Sx7dOfZcjmtemzVjLXFhHAJHIMA3AI9QZXKEQAKBmDdia2OPNYx9kNC+nK0vZk7/IV8Sd6/YIU2j95deS6PzazW/o9e1tfuI1qM1XSXX75FyjeGoayDG9ig2uka0TclbeOqPznPPtd/XSkw3bsp8n0/65gm4vOet40ZL+IyLvM7Kt+7UU2+5qG5aCEDgLAG+AXiWB2cQ6IaAPAjaV803Zolt49ljV5fVmOPY+te4ts/GkL8IfOmll9qu5GONkTyxswluTV35ytjHQ8fcOfbcN8+Oh45lXoz/0Hz6IdASgbXXQakcQnq0X689ad1rUW1KacNvPQJa53oRjxUphe/cdeWO6XXZM80UNj3n2Yv2XGvK+rHvHe4aTuVi18tWX25s9e361X5t3XHXD+cQaJ0A3wBsvULog0AkAX1jijSvbpZbn30DtsduYhJ36Ufn5Naofltubc72OKRZbdxW7aUWvh8dX9NqfbVd44M5+xGgbvux90XWazdUl1C/+Job88Wibx8Ce9dJ19g+2bcRNXcNRmOq+eTmtKX6qmmLD3duTp85fbk67bnE0R/bn3Jco67KQ7XqeYpObCFwRAJsAB6x6uQ8BAH75irH9nxLgq2/gYZy3ZK/PjwIN3u8hWMPc7XW2sZqVntt7Txfnx1fexxTX4ktP/LNT14QgMBZAnp9nO2dPwvdb2WW9WeP5z0yWoKA1knblBhSuxKvXH51beXyVyLXkE+th7Yhu5h+8eG+fH2uTavnLWi3a6vk+orx7Wrxne9RS9WxNbbWO5c/1SP+7EvPtbVjS8eiUXXO2a7xPeePMQjUJsAGYG3ixINARgKxb1YpIdWntilzxdbOk2N92WPtO2rbysNDiZqU8Bm7TvaMHavxqHZSG62PPa7No5Vrr3beW+NJzYSdj5+vb2s85m8nEFOXGJvtSsIeUuOn2ocj7zPiu/dpn7YhZb3nHspL+iV391U639L+U/KprcXVVuPczdE9j9EQO0fttI3xLTZi7/uJnY8dBHohwAZgL5VCJwQaIuB7WHPlqc3SQ607r/S56tG2dDzXvzxcKBt3bI9z5aDtVg2pD1xb47nzJY9z58653Zw3QmDvtb93/BbKsIbB3td1C9x61DBXt7mxVnLtQeMaVmuuwTVz1miLnbOmNjJnzbxYTa3baf7KQM9b1y36VPNeWlPjt3a97MWNuBDwEeCPgPio0AcBCJwSGOVNNJSH7U99wDiFlHBg4yVMq2Yq+rZykPmt51kNKIEg0BABrsuGilFQir2H++7HdjxGhn1fsGso1Y8by/pyx+TcxtVxXz46RnsTAVuXEOMYGx9POy+2FnaOz6f0qU611fOQ/R79Pk2qN1XP2nmpcVqxF3ZuznLuMg31pebh85PqY86+tP+52IxBYCsBNgC3EmQ+BCDQPAH3ASMkWB9QYu1DfnrvVw5zecTYzM1nDAIQqEtgy32N671urXJHkw+rqS93vbjn4m/LuvD5W9K4Zs6Sz73Gt7DbS7NvHeXeCGmlxq3o2KvWvrg5au2ue59P7ZOWFwQgkJ8AG4D5meIRAhCAwK4E9OFpiwj3IY2H4S00mQsBCECgHgH3/h2KrPd19z2j9AdvjRvSZftTbO28pWPNsZT/pfgyXiO2m6eeW32+Pju+dCzz3Vy0b6vvpdhrx0VvSJuby1yMOT9z83odU2YpjGyuOn9Nn51T4lhz8mn0xRM7neMbpw8CrRLg3wBstTLoggAEshGIfTPXgKn2Om+0Vh9stB0tv6PnwzofdwVIbfWnZpYxMVl3dSoi9+3Ye7dr557XUVwvil2D9riegrKRfPWLuTa3qFL/2oqv1tn6OK1hkMvPmtgtz9G1oG3LWtEGgSMRYAPwSNUmVwgMQsD3UOnrs+kujautPMjJT6y9zmulza079cFW+bXCo1UdueuUmufe8VP1Yj9PQOqpNdVWZ7jn2l+ytXrcOHpPmbNx53C+nkDonqx1WO/5ppkh/1v92vlr17DMO3/+/KkrOV7r69RJJwc16pKKojVNOddCrusplele9i47Obd99jhWY48MNW+3ncvZ2s7ZMQaBUgTYACxFFr8QGJiAvnntlaL7kCB65OX276XvKHHhva7Sul7XzWZWywT2vje2zMZqs9eAPbY2HOclIPfro92zj5Zv3hWTz1tLa083gbnvbK+vj6Gvb3uk/jwIBx8LX19/2aG4dwJsAPZeQfRDYAcC+lDdwhuZalBN/5+9M4/Z5irr//zYWhYVyi68acsWJLQSEAhoqMpqTIyVnVBIWIp/AAIxkIigghBjVSSyhDWsypr+wRIIEG0wwbCFLRWVvYS9gAYKlCU/rrt8n17P9Z59mTkz872T+z1nzrnWz3Vmzjzz3s9zL4BjKJdz32hb7lIP/R4KjiMYrB/HVLehuWtkE1kiZxvDVo+Xqq32i75uW/CGvRxb+lqg+9ZG7prMlbf+eh6PHJvkXVLHFF6wK23LVy3PEydOHIUjfcR5NLjxTut65OAagbWsH/0pUBt/Sz4tbdk4XcfgO6df7VP6wteeo/bYFbtrrFTPZatmrCVP8EFbExd1SaAVAT4AbEWSdkhgZwRwEzBK2q0315Y3AKMw6hFH6IZtVIZYK6HYe7CiTRLYAoFe53XO+dgrhhb1kdhycmnhczQbwiD0zo23R721za3XS+eq+7l1yJVfmmuq/zmZ5DIcSX4NnCRGvGvYrSHXmvyou28C/Bbgfdef2ZMACRgCuGHk5m/AOA7ByjF1uAFzjY84hjxGqDliEU694+ltf8RaM6Y2BGTt6LXaxurxLw3A+nT5wVwrvz3sSNxriLNH7i1suurewq62odfxUvWK5dlqDWk7Om/No0e/NVcXL52b5OCSceWm9XTfJZs61optip3UPHNiF1mfXYz7WGEccql+XXKwhbmYTSsPvdFbyWutsY/OlvH5CfATgH42nCEBEhiYgGyasRuCgcPfdGij38xIfHiPVIja9Zx7TtT6G4kdY9kWAVxDfGvUNz4KhRGvL6OwicXhuo71rDdqhTUXi6/1/BL+e/JszSfFHtYMWuiArf41cJnDOORat63si505X5afzzfkpNUvjOuxVn0wdTFxjbXym2oH8aXKU44EliTAB4BL0qdvEtg5AXvzkIqjVC/VPuXKCYxwI5YTPW7a1hY3cuS5ABLjt6xVWo1wLpJXGq89SGFNrD1XWdP6vfZ8JH7soaFcRjqX7cPAUNy1c7XrdiRutSxc+qV8SvVcMYTG5vITioFzJNCDAH8FuAdV2iQBEkgmIDc4OZtszQ1Riq+cWJKT3KggWbUtLHimrFN9HkAe+m2jorUWBFCjFrasDb0WtrAGdD42V33ck6n2wz4JzEVAzl+9/nucz9p+aV6uuGzs1nar89XnxxWTjWGJYx1XDvsc2SXyKvGpWUDfV0/k79KB7igtYkTMo8TFOEjAR4CfAPSR4TgJkMAmCGBjlmR0fxPJMQkSKCDA86AA2qAqe/iBQ9arb83uIf9Bl95sYflqP1sAjRzlrtUR8w6diymYchmk2KTMugnE1kRsPpZ9rX7MvsyLjzn8pMRCGRJIIcBPAKZQogwJkEAXAjU3uC5d1waMMZd8l6RolAQqCaSs1RSZlDBwfohsK5spflvJIP5esbey3yM+xNaKZW87wkDH7GKi5yUeLWP1Ea/oaDmMs21LQNemJ29d555+2tIps+Zauzk5oyapOpAvizaupWunpVPj0zqhvs9PSIdzyxPA+tProaSWsJOSkfaVIk8ZEtgDAX4CcA9VZo4ksBECOZu+Tpk3AJpGfl+4452vvW6N0jW3xqzXlKuOdY7zW/tDbWUMb4yhdcljrlWLvKXFu5Xt1nZsrCX252BaEtcedCx7e9yDAdZMD9tL2HTl4xrrFVttzeaMNcbAlYtrLGYnNO+y5xoL2eCcm0CMo11rMXm3l/WMSn5bz3E91dhHpHwAuI86M0sS2AUBuWmwf+DZ3kjsAkTHJPd4k7K1m7NQDXWuIbmOSyzJNM5rtElKlUJgg1abwxhamdN9Lduy3zt/5OBqxbd929ww7xq3Y3Js8xG/8kJ7OHD8E5t3qHAogwDqiPqgzTCRJdrbflYwDYU1xzXlmBOrlkW+DRHOZkrnAaeuMcyltLX6KT7WIqP3FFfMllVM3mWjdoz7Si1B6o9KgA8AR60M4yIBEqgmYB8GVhvcqQF9I6b7e8Ox1ptBxI0b6JS6WR0cp+jOITNaPKGc1xRrKA8758pLrg/6bXV6He/5utSLqc8uWfvItB93nWPtvbS3iDXSI/4eNlMILOU3JbYtyWDtSE66H8pRasP6hAhxjgSOE+DfADzOg0ckUEwAm0/qhlXsaMeKwlY4k/H8i2CvzLHmQHyt6w/XJ+RR0sLGCGthrhhCfsCjhOWoOrGcNA+Rhbwer81NbMGu2NJ9OW7pS+zxRQJCwK4zGVt6rcX8x+YlB7xEFjlCD8eQad3CT4pdxGJ1MJ5io6eMjaunr7XY1kxa1MlnT8ZT7IuMtjEKR8QUygEyo8TMOLZNgA8At11fZkcCmyPATXJzJR0+IVlzoRuqgwbhAABAAElEQVS30RNwnTNrzmcU3mtfF5ZjbE241pHY8I1b+znHlm0PHznxUJYEtkDAnkf2PGuZo/WVajt2HUq1M6pcT+a5OaNGLZi3tGXzyIlPZBGLtTP3sY1DjnUudn7u+OhvvwT4AHC/tWfmDQnoC3pDszRFAlUEsC55k1GF8aC8NYY2H6yVECmr45OFLZGXvm59OlsYd/EBi7Xn58rNNdYyT9gXhuCIsZZ+aIsEhICsLawzHO+BjM1b57y2803q1yLmVnY0yx59m6tevzF/qLu1EdLLkQ3Zic3l5OGyBf254nXFkOJ7LevMlR/H1k2AfwNw3fVj9AsTkIu33mhSLvgLh0z3OyKA9ajX6Y7SZ6oNCWAtxUzieihy6Eubqh+zP9p8LC+Zj8mMlpOOZ7T4eS3T1WG/NQGcq2hb21+LvbnPe+wVS/LBtWUN+1VNfbRuzjq3suDVsmalNnVOrnjE7iWXXOKaajqGOCyrpk5ojAQaEOADwAYQaWK/BHpe5GXD4osESIAEliaQep3zXbNS9ZfOk/7nIeBbJ/N4pxcSCBOQ65Ws0T2tU8lZv8OE2s4uzdnWevT9qkV8lnmNTc3P2m27UuLWfHn4xuMWKUEC2yTAXwHeZl2Z1YwEem4s2Ex7+pgR1cEVcpKDLeU1N8dUf2ScSopyPQjsYf1tJUfJY47rcw4vHY9enxjPsaX12d8vgdDawRzoyDHXGGi0azVXfd0Jsba1cUUT0rfyKfasTu1xb5+apS9WxCBtiry1A/3UcSvX+1jnhBx7+8yxr+OTPl8ksAQBfgJwCer0SQKZBHwbbqaZxcW3ksfiIGcKQOrFms0Ee2Nu5MaWN7frLSprt97aMXI/gTl+DdDvnTMgYK8vKfuFTwbj1iZ8jdK2uJfqkWOOzRY59KiHjUvnJHN2PiWGEp0UuyKj40vVoRwJtCTATwC2pElbJNCQgGwQPTeghqEmm9I5cQNMxraoIOu0KP7FnbP+i5dg1gBGqnfO/ueSHSmXWYtIZ1ECqWsjVS7qkALNCKTURF8PXPJ63gYmcy4dK7eF45xcQ8xGYGFzkRqmxpwqF8sTdlLWT4pMzB/nSaCUAD8BWEqOeiQwA4EtbhCS0xbzmmE50AUJDE0A5/aWzm/c0A8NfqPBpawjqY+vRr7xjeJiWoqAr/YnTpw4dv/hk4Op2Dzk2I5LQGqoP/mp+zZqfc3x1V7G8bb6Sx/7Yl46rr35L61Dqd7e+DLfegJ8AFjPkBZIoCsB/FDd1QmNk4AhgBsRadE3IjzcEAH9g4+kxevOFcUVDq3PAdhD23oZ9bLbOs4Ue1iHdn2m6IpM6If9VBuUWw8BvfalH3ulrKsUOzE/nE8noGuYrnVcMqWuxzWuPBJdiSHFRsrayMkn5hO2fO2VWaT3UnJIt7asJLhIFMgrxlRkU2RELuelYwnpQQ7xhmQ5RwKtCPABYCuStEMCAxLgxjJgUVYaEm9OVlo4hl1FoMe67/HDRlWSK1Eu4Saf+OJr+wRwr6MzLVkvWl/3Xfb1PPvjEcitv5bX/VhmvrXhG4/Z43w9AWGvXzn11Hot+lgHaMUm4kPbwg9tkEAOAT4AzKFFWRIgARLYAQHelOygyCpFfWOqhnfVBQPbCgT54aHnDxCtz7fe8S65MHLqkCO7ZE703Z5Aae1jeq3P1faZr9+i1CBWhxGzdO0dI8a515hG+c8gXEPQ2npgHdlxHpNASwJ8ANiSJm2RwKAE1ngzNSjK4rCwqfs2/WLDjRVd8a31hrwxGprbMIGlr5Gu824E3CPGpa9HqBvG9PEI/BjD+ghgDa0vckbcm0Dv62Fv+735jGhfmJLriJVhTEsS4APAJenTNwl0JoAfinxu1roprjFu/UMFbkh066vR0uM67iVjKa05/waYu2prWHvuyPuN1qx1rE+0sShrfMVst5wfOU7EhhZ522OMs90HgRb1Fxst7OyD+JhZxq7FqDHaUBbYL0MynBubAO8Fx64Po5uXAB8Azsub3khgKAJy47PGG5vSG/PYDWGP4mifoRtNLdcjjlKbS8cl/hED+jhOyWmUX/tIiXUumRC/0BqdK74l/ej8dT8Wk8jKC21MHrLwkaOXYpsyJLAnAr3OI56X61xFeo+TGur7AOnrumpZX7Za3ifD8XURkLrr97qiZ7QkUEeADwDr+FGbBEhgZQRSbvZWlhLDJQESaEyAP/A1BkpzJLBCAvZ+YYnrgo1hhRhnDRm8pFYp9RIZ0cHbF2yKLZ8ux8ciwE8DjlUPRjM/AT4AnJ85Pa6MAG4mVhZ2cripN0kxg7Gbp5h+j3nZ5BGXrqPu9/ALmzl+Rri59MXrG0eec7cjsJo7597+Wl0Hesc5p/2adV+jO2eOLXxJrvrdwiZtkAAJ8NtCS9dAr3uEXnZL86QeCZAACZQQuFqJEnVIgARIwEdAfhDkTdIVdISDfRBgj8GR3EDieLvWtTRiPX1r7zjxsY+QQ691AfugoI97+YSvtbb6OkdGa60i404hINeDuda4vvakxEaZKwjk1oecuXJIgAT2RoAPAPdWcZPvBRdcML385S8/Gn3Na14z3fWudz06dnUuuuii6U1vetP0yU9+cvr2t789nXbaadNZZ501PehBD5rOOeccl8pJYz/5yU+mN7/5zdPb3va26XOf+9x02WWXTTe60Y2mu9/97tN555033frWtz5JZ4kB3BjMedO3RJ4tfebefLX0nWNrzppiHY3ORuJDrDksa2Wtz9E5xfK1+cTk55oHV4kPtcbYXDHU+EnlKnKXX375YY+q8Wd1tX/wExkXQ9eYtVd7jDrW2mmhH8tXsxN/Vn6kXFrwoI1xCei1aNfhqFHXxsnzK62yljPWih0PWdOy0A/Jc44ESIAE5ibAB4BzEx/I33/+539Or3rVq5Ij+tnPfjY94xnPmN7ylrcc0/n6178+yfu9733v9MAHPnB61rOeNV3lKv7fLpeHhueff/5JP5zJr2u+8Y1vnC688MLpmc985sHWMUcLHMhGjg2cN1ALFKCzS9RW37C1dtnTdm2svvwRM+bFT6/1r33U5kP9NAKoL9o0rWWl5lonwiTFl5ZBf008l62m27twJEM3G462I5B6jrfzmG8J1xTRrD0nYGsP51dJjr71AG5SA9jVYzIee/lsx/Rk3tY913eKD8osR6BmbSwXNT1vhQAfAG6lkpl54GGefBLv+te//nTppZdGLTzvec87evh3u9vdbnrMYx5z+GYteXAnnyK8+OKLD5/qk08EPuUpT3Ha++lPfzo9/vGPP3r4d5/73OfwoO+6173u9PGPf3x68YtffIhFHgDKJwJTP1HodNZhEDcBHUxvxuRIjOTb3r7xjW8csfXdQNnxlBsvK3PkJNIJbfo6jlL7EffHpl0+9JjuH1OsONA5VpgJqso1ST6hvNQrhRs4pMgulceofnszE/uoz6gMRo/PckPN1ha3zYPH2yCA9SjZyJrUxzpDjGPd4ljL9OjP5adH7LU2S1hDp9S38BYbITuhuZBf2A7J2DlX/TFWGof1weNlCKCO4h191BTHy0RGr3si4P+Y1p4o7DBX+VVf+QH5Fre4xfSABzwgSuDzn//89MpXvvIgd/vb3376l3/5l+n3f//3p7PPPvvQ/vM///Mk4/J6xSteMX3xi1889O0/8um+j3zkI4fhhz3sYdM//dM/Tfe4xz0OduRXf8Xuda5znUkeUD7nOc+Z5AHlki9clBEDL84g4W9HZCR1tLX0Z3DFDwTQ8elhPmTHNyeMYpxq7Pv8rmE8xmUNObhiRD3RQkaO+YoTwLpAG9egxIgEfPWTcd/ciHkwpn0RWOvatPuLPV6yitgLbUxLXQta1Nhnwzeu+SPvFFmtxz4JkAAJ5BLgA8BcYhuQ/8pXvjI9//nPP2TyV3/1V9PVr371aFavfvWrjx7Gya8Bn3rqqcd0rnnNax5+PVgG5aGd71eL8RBRPvH31Kc+9ZgNOTj99NOnxz3ucYdxeYj4nve85ySZJQawMS/hey0+98rI3rzm1AvMfDd8vnHxgZvnHH+jyer8wKI2xpp61Pou0dcMSvRLdNa2dhBvKiuRk0//zvVqtXZL4l3bepccU+tYwoM6JJBLIHU9psiNdD76YvGN53LLkRef9q31W8WUUiPtt2U/5lvm8dZ+XWN63vYhL+1SLx0D+kvFsga/mpFe6zgnkIM9xjhbEmhNgA8AWxNdgT35G33ypRvnnnvudJe73CUasVyQ3ve+9x3k5BODd7jDHZw6Mn7mmWce5kReX+RkUD5F+NnPfvYwf7/73W+Sh4aul8SFl/xdwaVeNv6l4qBfEggRWOM61TdDuh/Kc8S5HPaSp325xqzM3o/BeElWWKO6lbrgeIkaLcljiXzpkwRIYJ0EcA1vFb3Ywxs2l7wWI4acFvHWXMdbc82J3yVbk4vL3lbGXFzs+t1KrsxjPQT4AHA9tWoS6Tvf+c7pX//1XyffJ/BcTr785S8f/R21O9/5zi6RozE8UJQvBRE9/cKv/soY5PQ8+je84Q2nM84443D40Y9+FMOzt9igZ3dMh6si4NrcSxKwduxxic3RdOa4YRVuvT/9VXrzJrHp95L1Wcv6Aq8lWbl8j8BvVDYuXnZsjmuB9cljEuhNYITrQkqOpXtYim0ts6bzvDbWFH1wR6tZ5fRTfOXYayU74vrHPtn7vjDGMKdmObIxv5wnARcBPgB0Udno2P/93/9Nz33ucw/Z/emf/ukkX9aR8vrMZz5zJCafAAy99PznPve5Y6L49J8MarljQr84wPxXv/rVw6cVXTJzjY24oc2V+wh+am+UeubQem3gRiXFLmTQ9syzlW3k18reUnbAHO1ScZT6XUvco9wEI46Rr0Wla6GlHviAV8j2WtZgKAfOkUAJgZTzo8RuiU7vWOQ8rznX9TUF/ZI859JBjC6uqWO9Y62pR0psOfaxPnJ0UmLQMqiJfDncEi/4X8I3fZKAjwC/BdhHZoPjF1xwwfTNb35zuuMd75j0xR9A8LWvfQ3d6SY3uclR39XR8/LwTr+0nRvf+MZ66qT+TW9608OYXDhFDw8ETxIMDMi3G/O1XgLYrC+//PJDEvKt0CkvXXf0Zb3BXoqNmAz+JzE1ppi90nnJa+kYSmPvqYe6iw/db+mT7FvSdNuSc/9HP/pR0Sc6dd113+3p5FFcd2RGzneJA2OlMZ3sZVsj4CNZCSN54Vp5OFjgH1173V8gFLpciICuu+4vFM7hOjLHNUSfj6Fc57iHwD1/6D7MFUfs3s2l48pV1133IStxpfKCTqy1sfnsl6wFn61YTDIvccW4huzINd3mZuVT4rN2UnSsn5zjH//4x0fiun80OFgHjK961asOFhnD2QKB//fzByz8CsItVDKSw4c//OHp4Q9/+CQXEvkm3tvc5jZHGvJNvC94wQsOx/LtwHe9612P5qTz8pe/fJKHh/J62ctedvjW3sOB45+LLrpoOv/88w8zT3va06ZHPepRR1IyLvPy+sQnPjGdcsopR3O2I/7Er7ze+ta3Hn3DsJWzx2J3DRd2GzePSYAESIAESIAESIAESIAESIAESEATkC/sPPvss/UQ+yRQTIC/AlyMbj2K8r8q8s298qz3kY985LGHfylZ4H/yRTb2jcHXuMY1jkz+8Ic/POpLp5WdY0Z5QAIkQAIkQAIkQAIkQAIkQAIkQAIkQAIkECTAXwEO4tnG5Ete8pJJ/h7fr/7qr06Pf/zjs5PSn9SLfbpOf4T71FNPPebL2tHHxwR/fhCyY2V9x+edd9506aWX+qY5PiCB0K+HpP4qmfx6x2tf+9pDdnYNuOzDrmvOIoKsHefxWARCa2CsSBmNi4DvXMw5/1qtAV8sEndOPK48tzbmY6U5iQyOXfKYa8Wm1TpoFU/IjmYTkht1DvVsXcPafEdcA71ZwX6I3ZJ10vEhDj2GuDGHY7QuWcy5dGJrIGQPdnu3rritzxZx+vzEbPv0esRobZYeS8zIS35+/fSnP30wddvb3jb6gRafT80Btn2yJeOwD9vyxZ03uMENSkxRhwS8BPgA0ItmGxPyxRvyAFBef/7nfz5d61rXyk7s2te+9pHOZZdddtR3dX7wgx8cDVtf2s73v//94K8Ah+wcOYh05OHfN77xjYgUp0cioD9BKnHJHwbGXykoqaVdA9a++IBd15zM6xdk9Rj7YxOwa2DsaBmdEPCdi/Lt8iV/LLxmDfhikTh5PRAKV75crGy95D/+cE2/UvPKntRYXlbvSonyXs06KPearin8euafHkmZJOo/8nkxyhoQVj3WOCoXO89Ebsk6IT7NAGPIQVrf+YC1pmXRj+0TrjUQsge7vdtYPULXzZzYfH5CDHSdYr5CdmK6reclV9e6kt9mK41Tr69SG6E8UR/Y/ulPfxoS5xwJFBHgA8AibOtRevWrX334m3jyPwryK7nveMc7Tgr+f/7nf47G/uM//mP61re+dTj+nd/5ncMDQ/3FHvqLPI6UVEfP44s8MK3tyAU09C3E+AIR2XS0Hmyx3R8B3Pzk3IhoStDXY9L3jVu5Ur/WDo9JgATcBFLORZEZ5VwcKRY30XlHpS4pNUT9UmTnzWBZb+AHPktGg9qEYoEM4gzJQobtFQR6s7K1GZF7CgORkVxy8xH5FPtLcEFOaBHDUjHH2CLOUXmCn211vNLXf4ZKfibGgzarF+Mh8iky1m7uMeJ3/dyea4vyJGAJ8AGgJbKxY/wqrXyU+ClPeUo0uxe96EVHMu973/sODwBvdatbHY3JrxKHXnrefnPvLW95yyNVkfu1X/u1o2PbgR15iGg/SWhleUwCvQlgI+7th/ZJgATCBHguhvksPSv1wQ9HJbUq0Vk655b+15I/atwyd9oiAVw/as+DXH34nasCiG9uv7n56ThzdSE/So7y0O+d73wnwvK2o8TrDZATJNCAAL8EpAHErZu4+c1vfvSV7x/60IeC6WJevmJe9PTrTne609HhBz/4waO+7Xzzm9+cvvCFLxyG73jHO9ppHm+QgPww4fqBInUshsRlJ6aj52v1ta0l+muPfwlm9EkCMQL44Sgmt7d54VLCpkRna2xd12oZc40vlbsvFt/4UnFu0W8Lxjg/W9jqwdheBxCvbn1+IWPn8ffUZFz6kjveVrbnsc1NfCFm11zPWHrZ1vkgJ9dYL/9iF357+uhle4l12SsX2h2XAB8AjlubJpH9zd/8zfRf//Vfwbf+YpDXvOY1R7J4gCcX0nve856HeOSTeR/72Mecsck4Prkn8vYCfOaZZ074FOC73vWuSf+dP23wwgsvPDq8173uddRnZ5sERr0J1bTtWtZzI/dxI7HW+Edmy9hIYA3XLlaJBGoI6DWO/aTGHnVPJpDLVdfkZGvpI2u+L5DY8ZaMQ7nEeOXyTydMybUSCK2ntebEuElAE+ADQE2DfS+BRz7ykdNVr3rVw/yzn/3sw98T1MLy9wVlXF5Xu9rVJpF3vR71qEcdhr/73e9OF1xwwUkiX/rSl46+tOT000+f7n3ve58kwwESmJtA7AZy7nhS/ekb5FQdypFACYG1niOhXHNyypEN+dzTHK5PaPeUe0qusqb0usKxbsWOlkmxG5Ox9uAvpifzVjdFZ88ymlcq55SHEykyo3DXDEpiQq5oS2z4dMSmfUO2hz/YnqOt5T5HjEv6WHt9l2RH3+MT4APA8Ws0RITy6b1HP/rRh1g+9alPTQ996EMPf0vhk5/85KGVYxmXl8idccYZh77959xzz53wa72vf/3rpyc+8YnT+9///ukTn/jE9LrXvW56yEMeMn3ve9+brnKVq0xPf/rTDw8TrQ0eb4tAySaLG2W0MSL6Bi4m65vnzZKPDMdrCKSu4RofvXUlh5LzuGdc+le+Svyknu9bqF8JH+r0ISDnUWhN4TzDniZRYKxFRKnrvoUv2riydqhnrJax+bUybbnucP6gzWGi+eq+tiHjmNN9LePqQ8c1N+eY5qK5u+LT83PGSF/8zxSugb4E+CUgffluyvqTn/zk6dJLL53e+ta3ThdffPEkx/b1gAc8YHrSk55kh4+O5VOEL3zhC6fzzz9/koeH7373uw/vI4Gfd+Srz5/5zGdO55xzjh5mnwScBOSmBTcprhsYp9JKB/eS50rLUxT21tdsERQqkcBOCeAaH0qf14wQnfXNsZ5j1KxnHXrabkUv5drTyldvO8hlDdx7s6B9EnAR4ANAFxWOOQnIp/Ke+9znTve9732nN77xjYcHeN/5znem613vetNZZ501PfjBD056aHfaaadNb3jDG6Y3velN09vf/vbps5/97OHvAd7oRjea7na3u02PeMQjplvf+tbOGLYyyM2prpJ2U7fHIesiC/4huVHm1hTrKMwYx3wEsD7R5pyLNkrowpadTz0W/csvvzxV3ClXcp0Qv8jBaZSDixKQT4Wecsopi8ZQ43zUtTVqXDWs166rr6Gj16d1fGJP559SS1cMOdfzmE/E4/KTEt+SMqHYQ3NLxgzfiE+O5Z5APnhiX1pGz41QK4kt5ZuLddzsk0AKAT4ATKG0cZknPOEJk7xTX/LJvNpP58nfCXzYwx52eKf63ZIcbhbk4j7CJrM0WzDwbcRLx7eE/xALrpslKkKfMQI4j2NyofnQug/p9ZjDdbqHbdokgREJ5Kz5Fuf7iAwY0/4I4J5qpP1ntCqAEeJqyQrXkpY2EaerTfnPoLliccXHMRLoTYAPAHsTpn0S8BDAhueZ3vQwNtYcBjmyMXhiCzHEZOeez4kLsj42mLc5+OStHI9JYG4Cteemb83PlYf45/k1F+2T/aD+rhqcOHHiZIWBRhC7KyRXPi65FmPwZePBeAsftNGXwN5rpfO369hH3sppGz6dvY1bRsgf42tihpiRg25deYTktS77JLAGAvwSkDVUiTGSwIYI5G6ishG7NuMlkOTGnhtja/ujcMvlQPl1ENDrVdaaPl4iA+2/1cMenkNLVDLfp9Re1z/fwvIavrXmG+8dMfxKi36uT9Rl7bXJzTtXnnxyiaXLl67ddA9XSPb209t+br4u+ZR1nCLjsr302Br4L82I/tdDgA8A11MrRrpxArIput5bSRu56XzsmGyw9q3lW/ZLN3Mbc6uYxG7OC/KxPDCvueb4oSwJ+AjYNYW15pNPHS+1Y+NJ9ReT03ZjfbGFczNml/NtCGydt+SHHHW/Db2wldJz0VptZcfa3cox+fStZO5/CJXWQ/TwdmWE89g1t5WxUI6huRHyD9UuJT7o6zZFLyRz85vfPDTNORIoIsAHgEXYqEQC7QnIhrHV16ibfg3zljm1tOVaQzV5uuxxjAR6E3CtWddY7zhof2wC9topa2St68TmEiIvsjnyIVucIwESIIGWBPZ4ffLtO77xlrxpiwRyCfBvAOYSozwJdCSAjWJLN/a+XJCrDyf0YnI+/TnGESN8LRErYljCN/IepRUWIQ6x+VHyYBxXEgjV80qpcXpri3cccmWR7Ik3rvWalB0bjcdo8Wh27JNATwKy9u35Kf5K70Nctmrj1+dnD/s9bNbmDP2U3FvEn2vDt24QN1sSaEGAnwBsQZE2SKAxAdkA8G5smuY6Esjd6FuGUutb9GtttMynxJa+oXPpx+ZdOhwjARIggd4E1n7tjfHZen6u/Lewp7ryWtMYaiDf+jr3K7TmEZcrppCeS750zN4P2eNSu7l68Nsyb7Fp3zou+yvhkEVd0GqdUB85uGRgO7UVG1r2y1/+ssssx0igigAfAFbhozIJkECMQGhjDOliAwzJcO44gZQbqNwbm+MeeEQC8xCQdVp67ZgnQnohARLIIZCyP6XYa2UnxVcLmbXF2yJn2kgjIGvDtT5842lWy6WW2HNDua7l5wBXDcurQE0S6E+ADwD7M6YHEtg1AdfGuMRNhq8II9xgzMnDl69v3MeN4+MSwA012ppIXedvjb1U3TnPidSYKEcCvQjg+lu77pc6X0NcdEy6H9KJzbWyE/NTM19byxrfo+pK3fCeI0ZXDeTTX/p8c/VrY8tdn2Di0sNcbUyp+i5mqbot5XQcuh/ygVqGZHxzLvY+Wdd4TH/uOrpi5BgJgAAfAIIEWxIggdkIxDbK2QJRjkI3GKE5ZaLqE0u4cbGtth/rp8YZs8P5dRLgDeY668aoSWDLBOy+NOL+vxR/XLP3yGSrObfKC+cN2pZrtIfNlvGFbEnseIfkQnP2V8Jb1cxlR8Zc46H4OEcCvQnwAWBvwrRPAjskgA3Pt+nh5iMmNzc6xAW/+iZDz+m+yGo56LZqe9q2MY5WDxsfj+cnYNf6/BHQIwnsiwCu+WjXnr29D6i9pqyJC3LVe6vuo7aQw/HaWp2TrbfOxeYJPS3Tui8+7d98a+1jbfZCNZJchJmt1dI5umKqiRFrL8YiJ29XPK6xHJuUJYEeBPgAsAdV2iSBHRMo3UxL9XqitjG5NnLXWM+YWtuWHPHWtm3ueo79dRFgLddVL0ZLAiSwHwJbuIew1QrtOWvP1+ba+zjEsrfvNdiX9YT3GuJljCQwAoGrjRAEYyABEtgOAdzchW5afHMYh40lqIhvxOHyj9jQumR6jNm4tP9QvD1ioc3lCPhqLevBrhFEqdcKxtiSAAmQQG8C9nq1t2uRzb8377XYxzrYIp9WuW2RTc76lPzBMkdvdNmt5jU6d8Z3nAA/AXicB49IgAQaEZCN27V5u8asS9z4oLXzPY+X8JmaD5imMEy1KXKt7eX4pmwbAli3rGUbnrRCAiRAAjUEcE2O2UiVi9lZYt4VO+5TloiHPtMIpNwnuGobs55iN2ajdl7ixlts6T5sz/Ur4T6GvnHEx5YEehPgJwB7E6b9TRPQF/ERNr4RYYOLsEI/JU6wzdVLse2TgU/f/KjjwtXGnssNtcnVG5XJVuNy1Rq52tpB1o5DPqcVG/aFNWPHeUwCJFBPwHXO1Vvtb8EV95auFcjPlRPmcimLnsterp1e8qV5xeIZOWdX7JZDKH6Zs/Ium0uMxdZbSdxggbbEhmURi9Mlr8d8MdgvAdE6NX2XP9eY+MjNrSYu6pKAJcBPAFoiPCaBBAJy4fZd1BPUj4m0snPM6IAHuCkoCW0ORnP4KMk9VaeGr/bRyo62yf78BFBHadFvHYWcM2s/b1ozoT0SIIF9EcB1sPZaWKvfi3pqXD33ml657dlual1TGWl7up+qTzkSIIH5CPATgPOxpqcNEZAbHbvByXHoB23XPGygDelvAZ/NE8dL5zZKHC05pK6lVLmWsdFWOQHfWtV11P1yT9QkgWUJYK3vdT1L3mCwbCXSvNtY11w3m4sQsPVwyaSRckuNxislvzliRhxz+HJXxj8qsYXikjnE77ey3Iwr/pJ4kWeJbix7V4zQ6eEPtnu0WCtri7sHC9pclgAfAC7Ln95XTAAbnk7Bt1HhYo9WdLARaP2t9nXekqM9tnmDTUzO6rU+Rhyt7dIeCeQS8J0LJWtU2wrpazlfvCF9nw7HSSCFANfWyQ+dUrgtJbP1euF6iDxx3II3bNba0jHV2NR2XDHV2HbZk7GYT5/eyONry6k0XtHDmii1Eaoj7MdsSwwxmZCfuebAai5/9EMClgAfAFoiPCaBDAL2Ii4bDzYfzOHYmrXjkNdyIuMa1zIj9m1uuTHW6uf688mvlb/ks+bYffXg+HECLa4NrnWSev618H88Ix6RAAmQwPIE5Nrmug6OfM1zxVtL0sVhTgZz+qplNYc+eLSs9ai2Sni2zKXEf0gHtQvJcI4E5iLAB4BzkaafXRBIvcCnyuVCw+bXy35uPCnyOlbEn6LXUmaEGFrmQ1vbIuA6L/Sazc0WurArLcbElvQxl2t7K/LIX3PJya1WP8cXZUlAE8Dak7HS9avt7bWvr4MujhjTvEtZiQ3YK7URizfHro5F93Ns1Mou5bc2bujnrgubr0u/xTpBfK1bV7ytfcAeWM3pE75btzoH5NXaB+2RgCXALwGxRHhMAo0J2Au6PQ65y5EVOyKfqxPyXzKnN7OYfkqsS+eUk08s3znnU9jqeCTPteaq82C/ngDXAR+c1K8iWhiBAK/r/auQe48C+dw9OpZJ7nWbayNGdL751mthvsjn96RZ6f78kbTxuIUc2pCglTkJ8AHgnLTpa7cE5AKP99YhpG5mLjnX2Nw3qXupk16H+gcH3dcy7G+LgNQZa92ed/Z4W5mnZ5PKAdcotPCQqg95tiTQi4Bdm738xOyOEkcsTszLOZx6HodkMWdt2WP4LWl9PqwtWwN7bOWXOG7JpVX8wsn3AkPd+mRTx30MQnGk2s6Vw9rK1auRh0+0YsvFxDVW4zeki1hKfIbWBuxq35DXY+yTQCsCfADYiiTtkAAJOAnojRKbnB6zSlrGJTf3puiKwca8teMlbjC3xrBVPq5atFiTMRux+Vb5bdGOq2ZbzJM5zUtAzkn7zolgpHN6y+eI5WyPc2oWkxWOYIk2psP5tgRacHfZcI31XEs+Kkv49MVSOi45nDhxolT9oNeSg9RW11f3q4KkMgkkEuDfAEwERTESIIF0AnajtMfplq78Hz+7QeK4xnYojl52Qz6XmpNcwXOpGOjXTWDJdRhbF1gzS8bopjbGqPAhmzFqsdYoYueYXV+QHzHfkWNrzcvWJWQ/R1bbAU/RRx9tyKaW1/aW6iNm+A/FDpm5WxtjD//w4asPuOh5jEG3V1zw08N+C5s2PvCw4zW+xCbswX6NPdHVdnS/1i71SSBGgJ8AjBHiPAmQAAmQAAmQAAkYAnLDzpt2A4WHzQhgbeGHzhTDWtbXT7HTQ2a0eHrkOJdNrA3xp/up/nUtRMcep9qhXHsCvnrqcVsvObZjNZHBXkubNfGMois10HUYJS7GQQK5BPgAMJcY5UmABGYnENpwsSGjveSSS47ik35I90iQnWM3j7zp44IAgZHXwhrO7TXEiFqz7U8A+5RutVffuJaJ9fHDu8jpfkxvjvnR4pkj59Y+Wl1TUAtpY68W6zLmY8R55J0TWw7XHLsiq2vvqpv2ref1eMgn5LSMtqPH0bfzsKFbyM7RakYl9ZsjRvoggaUJ8AHg0hWgfxIgARIYiIC9mRsoNIZCAkcE9E3+0SA7JDAQgdQ1Cjm0A6XAUAYjkLpGUuXmTE9iwht+bZxrvf9o/bAL9sCptHXZcY2l2hdd/ZJjW0M9v0TfxigxjBbjElzokwQ0AT4A1DTYJwESGJKAa0NPCVT+6G+pbor9rcmMxMr+oLA11lvKZ+6ba/Ena3WE9RqLITbvWwdY/2h9chwfk4DvnPCtB8jLvH2PmWF+VMgxX3NdGj3PWd/6WRehKx7IbGk9+HLBuYz62GOM61bXOEVe6+b0tW3dj9mArLSuvF1jWgf9mJ/W8/q3g1rbpj0SWBsBfgnI2irGeEmggIDekGXz5as9AWFMtm246vXaxiKt1BCQdR2riW++xznRw2YJH8SBtsSGT0dsaqbS7+HH55/j9QRK6oWal+jWR9zXwhZz8hEbMdctry1fHeYYT6m1lcFxak0g3yMfbTs1HokDsqUxid+QDZnTsaX4CdlL0acMCeyFAD8BuJdKM08SIIEuBOSGgzcdbdHKTR/ebS3TWimB3BvxUj9L6410Lu+F+dI17+E/VLvQXI9YaHNeArX1bXUNamVnXnr78oa1MkKtdAzS18e2KqE5kY3NW3u+41Q7Ipcq6/PFcRLYEwF+AnBP1WauuyeAm43dg2gEwN5wyDEZN4JLM6snsKZzAefyaDH3imfUfFe/6BMT0HVFLbQq9xJNYz99rAVf/TEvRLCG9FgKKchrffRT9EtlxAd8axs9fYu/Gvs+XVceOqecvs9Hjo0a2VAumCuJ0cceNlNjtvIlsaT6ohwJ7IUAPwG4l0ozz10TkA2Tm+a6l4C9CVp3NnnRI3e0edqUXoIArzdLUE/zyfMojdOSUqOdP1wz/VfDyIxxD9l6Xba2179KV3hYW9wjr625akY/JEACVxLgJwCvZMEeCZDAQATWesMicbe6ObQMWtoeqNTRUIRnj9wtXwTSqn6wt/UW9UGe5AcSY7es0xj10XXocZ2rzdJ3nay1S/0wAXDH+sCx1ZJ5rBufjOjAjtWXY+iFZFx6NWM1vpCvzz/y8c23GAf3Fra2agN1QK1x3Crf1vZY01aVoZ3RCfATgKNXiPGRAAnskoDvxkbGfXNbBSX54gayZY49bLaMbzRbwsv1ljjBEu1oscfiQV4xOc6TQE8Co50/S+01S/ntWduY7dzaa0ZaV/etT62j+1au1zFiQ1viB3FLi3fMjusbYKGLNmbDNV+Th8te7ZjkMuKrhrHNB7ZGzdXGy2MSGJEAHwCOWBXGRAI7JYANHe3oGHw3f7hBKY1/LfmX5per5+Oca8cl77JN/i5S8TEXy7gWJeYmIHVireamTn8xAtg3W11/Yc/18CcWi2++VWzavraJc1Ofn8hD66CvdWUMerADObSYx7HVx3jP1sbg8xXK2+pAFi3mXb5kPSyRN2JKaZFHapyQF9s+Hde4a8wVX6qcS3dNY3vJc001Yax9CPABYB+utEoCJFBIIGUD1je3rhs8uG554w+bOa3kYt+i78sRsjk+KFtPAOsJawltvWVaIAESIIH1EsC1EO16Mxk38lHYjhKHrZTvfsnKhY5PnDgRmj7MtfATdZIgEIvDzttjcSFjvnGE4JrHHFsSIIFtE+ADwG3Xl9mRwBGBNWz2uAFFexT8zzsyhjfGIYdxafWNnu5Dp3WLGFLtog7S2neqDZGDbo5OrSxir7Uzsn5uPUfOhbGRAAjs4dxFrmtvl7i2h5jNdU0UP/odiillDnG3vA+AzRT/tTJg4bMTm/fpYRz6aDG+dOta/xjLuY5JXngt/Z/BiKOktTnrvMQejlFHtDIHbtLHyzWGOV8LH5jXPjDGlgRIYD0E+ABwPbVipCRQTMDeQBQbmkERNxr6BgNjM7jPdrEmttnJUYEESIAESIAEVkpg5HuHGqRbzavX/VTqQ+Be/mtqnaJbsh5SdUJyobmUuClDAiSwDAE+AFyGO72SwKwEZJNe60a91rhnLXBHZ/qGmLXoCJqmSaAjAZ67HeE2Ni21Yr3SoMr+hHeaxpVSem8rtXGltX4911oYOd5+JPIso77S5nwCEHp53vpK65h03+VV5mMyoicyuNZgjelj9F0+ONafAPn3Z7xnD1fbc/LMnQS2TgAb/Nbz3FN+uFGTnHGjp8dassBNJNpeflrGTFskQAIkMBcB7rFzkb5ivwt5K6kF9raQXc6tl0BpfUvWUi9K+r6rNB8bm7aJOdcY5kKt6LWKK+SHcyRAAu0I8BOA7VjSEgkMR6B0Qx8uEQbkJCD1ZY2daDhIAiSwAQLyg+WoP1xKXLz+zrPIeqwBWzt7PE9maV5Gji0tg7hUjxrHvY4lEWMg6yC2Fux87HgsAoyGBEhgDgJ8ADgHZfogARLYLAHckNmbrCUSlpvH2A1kalyt7KT6oxwJjEig5Tk1Yn6jxmS5j3g9annNHzG/UddGy7hqa7hk3Wpjb8mxhy3JT797+Mi1ietSqO6huZg/W1N7DP1UH9DvzRF+EB/bcgKoVbkFapJAnAAfAMYZUYIESIAEkgjMcRPkuvGTMT2OYz1mEwjNWVk5niM3l9+lxnL5LBUn/fYlwJvxvnxd1vd47u3t+uqqe4sxWTtbXT9bXiMpNVv6WpwSI9Zwjix00KbkmbIWUmTgs7atybfWN/VJgATyCfABYD4zapAACZDAogTkZku/S4PRN23anh6HbdcY5tiSwNYJcP33qXAO1xzZPtHuz6owH4H7HA8zSvMs1StZTXg41JIHapyTR4lOSb61OuCV+i3A2l8OD62HvtbXvNDXchjTOjU1rtFFXLHWFXNMh/MkQAJjEOADwDHqwChIgARIYBYCuGmDM3uM8b23c9xA750x8ycBS0D/AGzn5JjXKxeVfmN4gNLPQ5rl2LpIs9JPKhRfaK5fROmWc/e61vn47Mm4by49u2UlQ/FjLpe/KyOwQisysO+SH3lslGvOyIwYGwnUEuADwFqC1CcBEiCBgQnghhCtL1TcdIVuRmM2fLY5vj8CWCu61RRkfE2v0HmxpjxGi1W4+tZIKNa1rZ9QLpyLE8g9/1LWB9ad9p6ip+V1XC3sadvo58YEvbna0vhcvFwxl9p32ZIx+TSgrptPTsa1b9HRenpO20iVc+lo+3o+t4/Y0Obop+j4ZGzuOX5FtlX+OX5rY87xRVkSGIUAHwCOUgnGQQIkQAIkQAIrJyA/GPh+OJDUML/Ejf7K0TJ8EiCBDAKh61DITKkebEIfLcZL27mulTV+JFe8Jc9Y7rF5zaokLtEp0dN+U/qhPFL8h/RT/KfI5PqIxR2yF5qLxRrzG9PX82JLv/Wc7rf0qe2yTwKjE+ADwNErxPhIgARIYAYCcuOGd8xdzU2etp3qT+uwPxYB1BCtLzq50RYZ3HBLny8SEAL4QY00tkVghHPcF4NvHBWQef3GeGpr7dvjVDsiV6Ob6gfX5Zh8SiyQkRZ92LXHGHe1vph842Kjx7XElYeONycnrYc+7KPFeE6LvH1sSmNETLqNxVXiyxe3y5fOFX3I2WM9jr5uc/LSeuyTwNoJ8AHg2ivI+EmABIYikHMj4wu8hQ2f7RHHS24YR8xjbzGl1g3r2bZ748V82xHAWmpnMWwpda2HrexnFrykRb82e6l5Sd0vueSSWtdHDwN9hhBbSXw+m1sZxxrAOkAbyg8c0YZk55oLxR2ay41PbNXYG4lZbu4hecnLl1toLmRzqTlfHkvFQ7/7I3C1/aXMjEmABEhgTAL6pgB93AjiGJFjHMetWu0n5EPmtGytf/hqabM2JurXEQjVMjRX55XaJNCeQKv1uubrXM41X/MSPeQtldFzqJSex5htoSdtirzVl2N5GHiNa1zDNZU0VuIXcSc5cAjV6jtMFg/lsEfcJcxEB/quYENzLvlWY75cXPH4ZFNjiTEI2XHFE5J3zdXG77LpG5vTF2LQjOb0D7/SzukXebMlASHAB4BcByRAAiTQkEDOho4bgZD7FJmQfs1c6AZlybhcOQn30WJyxbnlMfLvV11cV/bOeAv5rzEHrL/c6yz09JnhGtPzvj58l+r77LYY17EtWV+wWTKGFjzXHD9q0IID1lULWzk2WuaQ49cn23o9yBfBLPUaje1SHOh3WQL8FeBl+dM7CZBAAwKyoa5tU629oRF9vBsgTDaxhM/k4Cg4KwGsYa6Jvtj1tU33+3qd3zrW0/ye5/Xoq6GM++bmjfBkbzjHc2uUK3+y5+Mjo/KRKEeO7TjF40drjft4FjwigTICXP9l3Ki1bgJ8ALju+jF6Etg9gTVt3qU/RMWKDLupcpBHq/UsT8ig1bK1fdjUbYpNiRFvyIsNvuYnQO59mdvzUby5xvpGMZ91XAtcHrew1nz56Zraa5uLxVrGdF4tYvbxa2F77TbA2nWeYE1BRufqkse8Sx5zsbZGN2bbzrf0FeJh/eJY/OONsd5tSs4pMq44fecZxn2ty1aLsdI8tO8lP/Wn42CfBEYgwF8BHqEKjIEEVk5Ab84lN0+l6Wu/pTZa682Zf2rssZhkXrOUfkwn1XdrOR1na9u0RwI1BEY+b2ryWkJ31OtPDxa8pqVR1WtC90fhJzFJLGu/DoCnZqwrhDz1WOv+kgyRf+ucWtmLxWfna1j61oArF+vXJRMbm2NtSQzyt0BPOeWUY/e9c/mOMeA8CcxBgA8A56BMH7skULPprg3YnBtni5uMtfFNjRc3a2CE41R9LbeG9VuTn86VfRKoIdDifPP5n/Pa6ouB430JuGrMa1s6c80K52K6dr0k/MM3justl1mwceA41xr00ObqgwPaUjvWr7Uj9uHDyuYeazvWT64tnzx8pNqHHPSklTEcww/kcLzGVue1RD5gu0Z2jJkEcgjwV4BzaFGWBDII2M05Q3WVokvnu7T/VRZNBb02fmuLV6FmdyME5AeUJX5I2Qg+pqEI6OuZ7iuR1XXXdG4I87Vzt/GX8rd2Wi28XLs2fjm2YzWx2XjscY3tFN0af1a3JZeU2K1MS/8tbdk49fFcfrRP9klgFAJ8ADhKJRgHCWyAgL0pmSuluf3KjQPec+U4h5+5OZbkJDGOHCdvKkuqui6duc/9kdf7uio3drSjXNuwvmuuZTW6NVUCQ92G/vYX5OBTjl0vPY6+1YUe+OF4ybZVHZBzai4+eR8zn12wbJWH9YN4dLzw5Zqz+jXH8JNqAyy0vGtMz6Of6wt6S7W+eH3jtXGi/r3s18ZHfRJoTYAPAFsTpT0SWBGBNW52uClbCvMameWwAl9pt55rDpdUWdxIpspTbj0E5HxY6pzQ60r310OPka6BwFLreyQ2vvNLxjEn7ZZZIU9dF9eYnkc/RQ4y0tq3tgM5jPVsQ9d3xNjTf67tEdZfiFluPjF55Is2Jm/n5e/+8UUCJHAFAT4A5EoggR0TmPPmqjVm3JDNmUPpjUfr3GP2wCYmF5ufk20sFs6TwNIEYudD7+tDq/N6aY70HyfQey35IsAaQ+uTC40vFXsoJszJpwFxHqPFnG7tnORk3yKvx7T+Un0bt40jNq/lRRZvPR7rg4lLDnPS+l45PnPy8flDTL55PY7Yavzm+HP5lrEQP62Dvi9exKJb6MTa3Bhi9nLmU3zrnGLykM2JwSeLNeKb5zgJjECAXwIyQhUYAwmQAAmQAAmQAAmQAAn8goD8UOr7wZ2Q6gjEuMYeGNR5X6+2cMtlkys/N52SnOaOUfzVcBRdrPmYHS3ryzNmw6c31/gc8bl4YmyuPOmHBEoJ8AFgKTnqkQAJDENAb/Y3vvGNm8el7Tc3ToPDErB1583dsKWaLbDQGrDrpVVQYjfkt5Uf2hmHQMt6Y122tOkjBV+++di4xFhrI+YjNL+k71BcOXM96yx8XPZd3FxjOXlA1rcmXHFAJ7dNtZWSE2y5ZENzNmbI6nGxKeMu21rO1c/R0bKuOFz27VhpnNZOyXFP35aHPZZ4Nb+S+KlDAr0J8FeAexOmfRIggVkJtPw7H7KJ730jB4O9c5BFvBcGe8mz9YUJPwi05ge7reOlvfEJyFrSb0SMNYY5jLtaWT/yho5LptVYyVpFfNDVxxhrFZ/LDhi25tPaniv2ucda1qPGVo3uXMx0jNLXx7EYfLK+8Zi91vPIR7fWh6z/JeOt8a3zqrFjmcSO5/QVi4Xz2ybAB4Dbri+zIwESIAESaEhgiz/UWTy8CbVE8o7JL4/XHqXlOoK3K3/fnB6XPl56HGNLtDqmVv7lfOI51YpmHzu59amV77HOYmRKfJas3RQ2JXZj+fWYL2GWGkdP26kx+ORSaujT5TgJzEGAvwI8B2X6IAES6EpANtseNwO97HaF0dE4GO/h5ga5unDquT2wcDHgGAmQQBsC+nrSxqLbyojXqiVjmou7qxra95IMXLHNMSb55+Stec0RX2sfufmG/IOF8ANDjIX0Sud07PADvyGbIgP5kFyvuRrf0LU5pOTdKx/aJYGWBPgJwJY0aYsESGAxAtyY+6AXrmTrZ4sbRb/E+maQE9r1ZTBOxMKQHMepByPpS4B7xXG+vnOfnI5zWsNRbc2wF/jWhDDI8aHtiB7evVhqf718tLALzq1saTs5DHJqqX2wTwJzEOADwDko0wcJkMAiBOTvAeZs2K4ge99UuXzWjOHmpzZvGwM47OWmBvmCQyjv1qzhs6atiUlyFf1QzjWxUZcESGB+AjXXhJxoU68bqXI5vlvISlz23cKuz8ZcdfH5bzUOZin2amtfq58SYwsZqW1qfVNzisnF5kvycuWA3NBauy4dK1N7DN+6rbXZUl9qgXrofksftEUCJQT4ALCEGnVIgAR2RwCbuE18jpsc6zP1GDdFqfK5cr3t58ZTK69raXPTcy4/sXmXTs8x33pN9Vmrn+pn63LkuPUKl+XXY12k2JzrOpUSSxm5dWktzWGueqdWRXgszSQ11lQ55IQWej3Zg2FPH8jD1db6Rfwu26OPleSu89X90XNlfNslwL8BuN3aMjMSOEZANq2tbjwlG/IxOIkHwm8uX4khnSRmayzxImY7d5Jy4YDYt7Z7+ywM1auGeNF6BQMTmoPPjuUUMMepjRBgzccvpOt87V03se/yW0IrJ1b4zNGpiQn+SmzMrdOLSYhBL5/CrqdtX216+BSbmqH0W/spsdk6Bh9T17hmYmPXcy7dJcfAbOQYY3ws75g850lgNAL8BOBoFWE8JNCRgGxae3y1zBs3L5qj2Mdbj4/Qd8U7QlxbjWHUdbBV3syLBHoR8O0bOMd98znx8PqcQ2ubsi3WUYgM1mtIpvVc75wk3tY+9LnoY6ZlWjMrsYd40JbYmFtH121Ncc/Nif5IoCcBfgKwJ13aJoGBCPTYaPVGblPt4c/6kONQDC75FmOu3CQO13jIn449Vzdk1871sg27Og/4xhyO19xKLrEcXfNrzpmxkwAJzLO/+K4vOfzl+hO75lo/KTo5MfhkrV/I4ZoZixvyvdtQHL4cQjFpezF9sBB7Wi9kP3durnojrpb+NB/Yb9XCNrjj2GcftYS8lYvpW3k59tlyyaaMlcSQYjdHRudk45FjzEtr53P8LCWLmJHHUnHQLwmUEOAnAEuoUYcEBiEgGxA2oblDivmNzbeIdw4fqXHu+SZgy7kjN7Sp62ErciOdY1th2joP1qg10TR7I10TUmOxcrJ27Dste0qFCFjOIitjeId0dT1CcilzS10bXPmnxOuT8XGryQ+c4TPHVuv8EEOotfFCNidu6PRsXbWK8YrN94yXtklgjwT4CcA9Vp0574KA3BT02lRhO3bjAbmlgffi0CKvkWNrkd9abKAOWNM4bhl/D5st4/PZGuU8tvGhVhgP8dWyITnYYksCMQJYU1tbT7G8ZD43Z5GHXXDNtQG9Xm0sL1cOrlhS8oKMZWLt2XnoQS42Dzlpra6eQz/HHnRsKzbEV4o/qxs6trGFZFPmJD5r0x6n2NEyLn2XH60T61ub+jiVs66F1o/5zpnXPlx6mId/aTEm8tLHnEt/1DEds85n1HgZFwkIAT4A5DoggY0SwGbaY0OqsYnNssYGSoYccexqW/hx2W0xNnJsLfLbmg273vRatnM297XWesS4wd0ylnFXvJdccskxUei7ZI8JruRgK3mMjFszxvpBvL51h/lYa+3F5F3zsKHjdMnljPnyKvVRqpcTc46sxANu0PPlLPNWFjpztjYGVw46nlrmIR7aj/QRW61PaxfH2m4LX7AB+7rVvvQ4+tCFHI4xLy3mdB9yek7r9Oj39lVjX/NAvweDuWwihxomc8VKP/smwF8B3nf9mf3KCcgmE9toZEPCptQ6XfhG67Lv890zLsRx4sQJdNnOSGCO2s6YzjFXobV+TJAHixDwXW8WCWZgp1s+R1tgl/Ncv1vYnMtG6BwIzc0V3+h+wAhtTrxL7A+9fZZwyGFWIotzs0RX65Sw0zykr4+1bVe/VdxiuyR2V0x6rIdNbd/2wSOHobYxd7zad6hfmk/IJudIoCUBfgKwJU3aIoHBCGBz7LUZwa604gvHFoOet3LQQayiizHp63E55mtsArp2Y0eaH50rt5T1mSJjoxFfJXrWztaOXTXQOaYyS5XTtrfY5zrzV9XHZuS1Ezs//Nm6Z2Bv5JzdkaeNSl7IUWu4xnyy0MthpO1DT4/Bpq+FDuZxrG1gDDK+1qejx9EP2QzN+XwvMS6fCD/llFMOrpGXjQPjrpxkDPNWb65jV1zad0mMsFmiq32X9Et8It4Sf9CxNlrWtaUtxMuWBFoR4CcAW5GkHRIYmIDd5EYMVTZLvHV8rjHMc4MFiTHbNdZHzhXf+WLHU46tTGqlSvVS7a9VLsQlNNc63zWubc1AWOGtx/fW9+0va6jvXOt9Lj9rXnujMWoRj9jAG7VZw3mBWENtbR4t+Ibi0UDV3AAAQABJREFUazHnitGXt8jKHN4t/PtsuOLyyWK8RAe6vjbFpsikyPl8cJwERiXAB4CjVoZxDUFgjs1wiEQbBCGsQhulzMsLba5L1EK3Lhtr+7XfUh6u3Ocec8Wu14Duzx1bL3+SE94uH5gbKXecM654RxjzxedaXxKvi61rDLnJNSE0D7nUFnH54k61Q7nlCGyhdliHLoq+uZbngcvv1sY0R92vzVPqgDds2WOM2zZUw1Qb2iZ0YnahIxxasoDdnBYxoM3RtbI6f9u3svo4xEvLte6H/IIH6iOyGJM4kB/6iA3yOM5pa3V1fD6/8IH4Qwx8NlLGrd2U2FLsUoYERiPAB4CjVYTxDEUAm81QQQ0QzKibot28B0C1yxBYh7HKvpbrGG7yhZ70Q+sIOaFNIQ57aFN0KLNNAqE1gDUVklkzla3m1asmo97v9Mo31e5IXPTekRp/SC4lt14+4Rutvh7Fzl09b/V1vlpOj4/cd8XcugYj58/YSKAlAT4AbEmTtkhg4wRwQ+FKE5uztPrtkuXYlQSE1VpvYlDzK7O58hOerjktx/4+COCaYVtf9pCT+V7nRa1dXN98OXB8HQRcdRz1uoVYdRujrM8lK5uTZ+35Yn2PdpzDwhf7XIxaxOrLITQeyi80F7JZOufzFxq33wQf8w3OPpsx/dx5lx+JAXHk2oO86Ittl309Bl+l/rQt+Pa1VrbUp89+yvgcvym0RF4puVOGBIQAHwByHZAACQxJYE+bZyxX3DChHbJgDIoECgjMuaZb+5LzNnbuFiChyswEUEO0cN96vcCuba3f3Hkrn3oc85tqp5ec8NfvUj+5D39K/dTqIddaOz30Za2Mvl5ceZeew1pP910+XGMuVqivq3XZaD0Gv2JX52RjtcepcWibPh2fTKlPn5/YuFwTNI+YfOn83HmVxkm9/RHgtwDvr+bMmASKCchm5tvAfeO5zko2zBKd3LiWkgdXtEvFkep3y7VIZUC54z9ghHiErik91xLOp54+QnmvZU447YGRK0fXWK+6iS+sSe0jFINPR+vH+qk2llgHqbHFctTz8oP/Na5xjaZrOlQj7TvUd9XeyrfwY23mHusYUmLOtV8r3yKmmI3YvCuHEh2XndiY9WOPRV+P6XrGbKfMa9siD/t2XNuSOZGTd0wu1aa2H+vDJ2KNycfmrZ1YXmLP6sR8cJ4EagnwE4C1BKlPAjsjoDcq6et3LQptu9bWVvQtE9ysbCU/5rF+ArIm7TuUlV3TIVnOkcAcBPZ8XU09H/U5PkdNWvlw/bof6o2ccn0JM7xzda08YrHj+ji1RlqH/TYENPuUWrm8ahuu+SXG5oip9PwK8SitQcimzOXYzWWXKx+LlfMkUEuAnwCsJUh9EtghAd9mJuPYRHU/hsjag42Ynsxb3RSdtcnsIcfUmsjaII9UWsvLpdRKZOw5n6KXm53Lj/bbw2dujKPJ74XJWvN0rWmsobXmhPilDeWn5Ur6+tyP6fdimRJDL9+xnHPmJY81xBnLydbDl5Mdt3r22MrH4tjbvOUVyz9XPmYvd77Ev6yBEr3c2ChPAikE+AAwhRJlSIAEkghgc8PNDlooY16O7RxkXHNaLySn59jfJoHQutlmxuNnJTWx52hNnWp0Y7Rg28YrehiDTMyWb76VHZ99jpOAJiDrFWsO47VruLUd2CtpW+WS43sunyE/tqY58S8hi3hDOdXEBfs1Nmp1fbnpcVecrrHaWHrq63zEz9riL2GDnKXtlS98IL5efmCfLQn4CPBXgH1kOE4CJFBEwG5w2ojM4a3HY33ooBV56fNFAq0IyI0Yb8bKadaej7X65ZGfrFm7DvR16mTrHBmdQG39l8hPnz+6XxvLCCwkBrxr81lCv5ThqNeRNdfC1t+eK/pY961e6LhUL2Szx1zpuuwRy2g2pYZz1HEuP6PxZTzLE+ADwOVrwAhIYDME5tgwBdZcfjZTGCYSJcA1FUWUJFBzQ1ujmxTcL4RSfvBJkcnxSdl1EVii/rXXoJrzB7q6lYrVxjRC1VO+BVjn3SpnWUNYR+jjeAQuvWKYO9dQvVBX/XcgpY/xXgx89jFuW1cc4OiaazEmMWzh1SKPFjaEZSs7W6gLcxibAB8Ajl0fRkcCJEACJDATAd681YHeEr8t5VJX1X1p7+EBTWpFRzkHauJYsp41cafWiHJXEGjFuud6aRWjZFz7cLBlLEuuQeSBdslY6JsE1kSAfwNwTdVirCRAAiRAAiRAAsUEev6AVxwUFYckIGtl7h8s5/Y3JHhHUD259LTtSIVDnQigjrnnrW9PyLUTSguxhWR8czYOX7w+fTuOWGAHx1ZOjiHjmms9JnHk+LNx4zjHBnKADmxgnC0JbJUAPwG41coyLxIgARIgARIggWwC/CEgGxkVSGA4AvihfrjAGFAzAq5rtWusxGGtHay/Wjs6dtjUY1vp5+TWkunS/HLyXjpW+t8OAT4A3E4tmcnGCcgmYd8bT3mV6aFGqwyeQZPAzgls6QeLnZeS6e+cgJzL8tZ//21uJK7rCX/gD1cB91C2tVoutlYm9xhrpsY24obvlHq7/LligRzm4COntfHl6C4pi5zBwBcL5HzzofFaNrX6odg4RwItCfBXgFvSpC0S6EDAd/MQ2wQ7hEKTGQSkbqxRBjCKkkBnAr5rKdzyfAUJtrUEeP2vJdhff67zXfzErj39s92Wh161C523qGHIN2RKaIfsanupclon1EfMKXZFBvIum7ARknHp9RiLxRryWRu/1gcTlz8t55rnGAn0IsBPAPYiS7skQAK7IiAbud3M7fGugDBZEhiMgO9GXMZ9c4OlwHA6EXBdv8VV6TW8Zj0hFt360i6Nz2dv6+NzfANsjGHN2ojZnmMe10u0IZ+56zNXPuQ7dc7WA+ed1fflC3lX7D4da3uEYx0/ctJxgRNaPSd9Pa77Vm7O49o4RF+/W8QOtpp3C7u0QQI5BPgAMIcWZUmABEjAQ6D2RsNjlsMkQAINCeibefQbmqepFRII/SA293XdF4sdl2M7tkL0XUIekc3c66gLWI/RWG6p6zRVzhNG8+HR4klJMFaLmI3QuQMe0oqfmK/YfCwWOx+Kzcq2OG4VP7hJTLrfIkbaIIFSAnwAWEqOeiQwEwFstNiMcDyT+9W4GWFjRW10uxqADJQEFiYw9w3+wunS/cIEUtZbikwsjVZ7E2KRlnvMceqaDWYwhuOlW6mZvEaLqwUXrMcWtpawgdpY31Ir+0L90Np5OV6Shy8XV5y5Y+CBtqcvX2zgrlufbOvx0nzBq3U8tEcCpQT4ALCUHPVIYAECpZvPAqEu4rLFJrvETUUJrNw4wUbroV/inzp9COTWBPLSrvml49f9NefE2MckgHMmZz+FDlqbWWjN5vhJkU2RsfFt+TjEXvK+5JJLhkg/FucQQVYGUbI2fedUZSjF6pKDfpcY0jmhP2f9EX9J7Lk6IV8yp18hWS3Xu6//FEBuTDan1FjnrH9qTJTbLwF+Cch+a8/MSYAEIgRkwy7d7COmZ53GjQfaWZ03dLaVeoSQbGG9hfJzza19Xbpy4tj4BGrWnUt3juvTHq8PS6+k2rq61srSOfX0L2s0NedUuZ7xwnbPcyvVdu1aQy5oc2oBndatK3eMtag/bLWOO9We9t8in1S/lCOBGgL8BGANPeqSAAkMQ0A2Yb0RlwTm2rxdYyW2e+pIjKG39u1itJYcXbHr3LbSRy1tPq46Yd3vhY1lIscuLi45ju2bgHwSTNYKzpkeNFqsRcRnW4m3hf3UvMXXnP5S48qVk0/7lL4k/9Jr61b4lbLTemBhWy2zVF/HZGPAnB635yWOtYz0sW7Q2nnXcaos4kLrstVzTOIU37UvsZOas8uXTxdccmK0tnJ0EZu1gXG2JDAaAT4AHK0ijIcESGARAiWb/SKB/sKp3Gjod04s9ibFHufYmktWYlxbjWrZ2HzXUKfanHP1LaNcfcrvhwB+7asmYzkH8XbZ6XmOuvzK+u91DiCXXvZd/EYbA4PR4ho9ni1w0+u+53lWUkt7LdCxWns9axHya+MIHbeMsSamlnGE8uUcCSxNgA8Al64A/Q9HABu9bocLkgE1J+Db+H3jzQPoaFBywBvrGu7WlN+aYgXf2rbmZrbWdwt9rLecPHLqjHXdIlba2BeB2nW29NrLOadKK7t0jilx++o4euxz1C+FXy8Z8PfVp5ffHLu2BnKMd44dkdX5tsg5JQ7tMzfeGnnLrcYWdC0zeww5aZG3HmvVh+2Qf5+vEi4lfnz+OU4CqQT4ADCVFOV2TaDkor43YFtgZDdie7xkTYVviPFIsS7Jib7HIRBar74ouY59ZDg+KgGsWbRzxFlybs0R15I+hL9+LxkLfR8ngLocH73yQU7uuYP7Id1a27HjtZxDLeLM5RtjN8d8i7x7x7lGrr2Z0P46CPBLQNZRJ0Y5AAFsRnu44CNXjd2Xt5bVfZ+8tjlif5S4NcsSTr48ZFzblr5PtsQvdeoJoB5rrA1i1mtMiCAnFx27Jl0yHCOBHgSwLu16hS/M49jXpsr59GvGl/RdE3eprq4VckdbarOHnsSkY3X5kPkRY3fF2mIslCvmYsx8cWg92PLJcvwKAsJJc1sDl1htdU663zK3WAwtfdEWCfQgwE8A9qBKmySwIgKy+du3K/zcmwTYdNniWF8CsZsTO59b277R0zoI2DphfNQ2tI5Cc5LP2nIdtQaMqw0BWY9ck21YtrQSu4609FViS+IbPcaSvNakk3LerqFGoTwQP9qU+oTshfRL9UI2fXPiC28r4xu3cjwmARKIE+ADwDgjSpDAMQI5G+4xxUEPcjZ3yd2+B01rl2Hl3CDlyO4SJpPOJqDXFPrSygttyGiKTEifc+shMNo+qtcr1+G61tEoa0nHgfukFJI5sin2esjo3Ertp+bpOxf1uC+G1HM3Vc76acHB2gwdI2crg/jR2nl9DO6u2F1jWrdFv9SHL/fUmCybmm8DL/WJHNCm2qEcCfQmwF8B7k2Y9kmABEiABEhgxwTsjfiOUew6dfuDoBxzbdQtCfKr40ftMAF7zoal47Ml69XqtIpJ7LayFc+8n0Tv62gpI62n+7aeMTK58tqe9qvH2SeBvRPgJwD3vgKYfxYB2YhqNqMsZysXJqu6ArrWmWuszsuVf4S71g71SYAESMBHwPeDmG/cZ0fGRQfvkNwW52J5Y163tRzE1givUByhORt7SzbWds0xctDx+fo1flJ04TdFNlWm1f2Ly46MucZTY8uRQ51ydHrKxvKOzZfEBt4u27G1E5tPjSdUB4kL84g11a5LDjHDpkvGNZYr77LBMRLoQYCfAOxBlTZXTUBvHKtOhMGvkoDvhsF1o7XKBBk0CZAACWQS8F0XZbz22qht19qStFrElIknKI78SnMr1QsGVThZG8sll1xyzPOotUKQOl/UEXM9Wp8PHUcPv7k2W8Ujdnw558Y0gjxysXz0sci48vaNSV6wq3N0jen5WB9xxORC84hB5wd55FPrBz5gN2YPfkMxwRZbEliKAD8BuBR5+l0dAdfFfHVJzBQwWZWBtjcaZVaoRQIkQAL7ITDidXPEmPayInLYt7hXEX85PlPrEIqth78eNlNzpdx8BLCupEUf3l1jmNNtq7Uidkpt2dh1fK36I8fWKkfa2ScBPgDcZ92Z9Q4JYKNFCwSlGxz0bTvHpmx97uHY1m0POTPHNAKtz+E0r/NL7SXPEFlcB9CGZNcwV7tfpP7Amsqi5RpracsXv/ah+z753PEeNnNjgHzqWmktB/++1vqzxz690LjLhh7T/ZCd1DlfncVPa1+pMaXI+eJO0YXM6DkizpQWueRwgY627xrT8637OfFq3744xZ62KX376V9tx9XX+nY+NAdZkYnJSfx8kcASBPgrwEtQX9jnV77ylektb3nL9G//9m+T9L///e9Pp5122nSzm91suutd7zr93u/93nSb29zGG+VFF100velNb5o++clPTt/+9rcPumedddb0oAc9aDrnnHO8enriJz/5yfTmN795etvb3jZ97nOfmy677LLpRje60XT3u999Ou+886Zb3/rWWpz9SgKuTUjGWm8+re1Vpr0qdVeNXAlYubUw77HeXHw4tk0Cdt1vM8twVltjkHPt0rKtryVie81swSOUh+YXXmVXzoIJ7F85M96vOevYbF9+8Jf7Vf0q4aH1bb+1PWtfH8d8oW4xOdiEPI7ZkgAJpBPA+aPPN4yJFenrOVh2jWGOLQn0JsAHgL0JD2b/ta997fQP//APhwduOrSvfe1rk7w/8pGPTN/73vempz/96Xr60P/Zz342PeMZzzg8PNSTX//61yd5v/e9750e+MAHTs961rOmq1zF/+FSeWh4/vnnn3RDJjdpb3zjG6cLL7xweuYzn3mwpf3M2ceFGRdxHM8ZQ44vxAmd1HitHvRt67JndV0y1g6P90uA66Nf7bfMFteZLeeYsjLAQcuulUlt3LX6miH6rWy66gQfvVodu/RtDHo+NQZrQ45dtlPtzSVn4xa/l19++ZH7EydOTN/4xjeOjkfo5NQnJuvKHzliLrWOMV+wu3SLtVkbB/IFp5C9Vj5DPkrnEH/LGFPXTCxmMNZyLeNE7tq+qy8/b55yyimuqaOxWM7Wlz0+MvTzjmtOxlw8tB77JNCLAB8A9iI7oN0XvehF0/Of//xDZGecccbhE3vyyb1f+qVfmr773e9OF1988fSe97zH+/Duec973tHDv9vd7nbTYx7zmElupuRC+vKXv/ygL5/qk08TPuUpT3ES+OlPfzo9/vGPP3r4d5/73OfwoO+6173u9PGPf3x68YtfPF166aWHB4DyicDUTxQ6ne14EJsNNpfYRuZDBX3XfGjOJc+xMAFbI/BFLcPa657ljdC669czente9PS1Jtu4Pqwp5j3Eatdr62tbbt1z5WM1Qn6t7cb8rml+tD3bxmOP18TWxrrUOoz5BeOYnM1Hjmt0XfZajCGmGls+Fr7xGl+pupKXy3+LfFNjoBwJLEGADwCXoL6Azw984ANHD//+8A//cPrrv/7r6epXv/qxSO52t7tNj370o4/9bykEPv/5z0+vfOUrD4e3v/3tp9e//vXTqaeeejg+++yzp9/93d+dHv7wh0+f+tSnple84hXT/e9//+n000+H+lErn+6TTxnK62EPe9j0F3/xF0dzYuce97jH9Ed/9EeHTyE+5znPmX7zN39zutrVllumro3hKOAFOr7NyheKlpdcUje10fL25be18b1y32veW1u/rfLBdQrrAm0r+2u3s0Yea4y5dJ1IrljDYkPvw6U2c+yUsNbxumLUNnXfJdtzrBXLnjHOaTtWN4nFV6+Qrk9nztyW8iW5h9ggrpS1CDs5POFf24cd8R2zBX3Euac2J3fNVzPSrPV4j74vhh6+aJMENAH/72lqKfZXTUB+dfcv//IvDznc9ra3neTBmn34pxO8xjWuoQ8P/Ve/+tWT/N0+ecmvAePh32Hg5/9c85rXPIzLsci96lWvku5JLzxElE/8PfWpTz1pXh4aPu5xjzuMf/GLXzx8IvEkIQ6cRMB3Q6DH59rUxM9cvk4CscEBXcMNpseUSMBLgNeRk9HwenAykxFHUCdp0a+Js+e50NN2Tc4+XYkXb5/MUuNzsvT5ql1zLdZrKX+bE+qs21LbrfTAB20ruy47yFvPucb0vPRr14C1t9VjYckXCeyRAB8A7qDq//7v/z594QtfOGT62Mc+NvsTdXKBfN/73nfQv8UtbjHd4Q53cFKT8TPPPPMwJ/L2wiqfIvzsZz97mL/f/e53eGjoMnTuueceDcvfFeTrCgLCs/aGAzcFvlY8tfIxUt2EnV2PI8XHWK4gwDpteyXo+vrOR4zjGrVtImnZgUmaNKVGICA1q91LdR4tzgeJyfXWftbWH+ncKIkltkZsvXR9oItW5nRfy7r6Imvl7bFLr9cY+OmcXb4ghzl7jPGcVvsM2UvlkyrnizGmH5v32e01LvHot8tPiKtLfq4x1B7+RmOLuNiSQEsCfADYkuagtt71rncdIpOL2m//9m8fRSl/908eDEoben35y18++qPJd77znUOi013ucpfDvHwpiOjpF371V8Ygp+fRv+ENbzid8fO/USivj370o4eW/+Td2Ll4pWxqKTIu20uO2c3bxjLqTYeNk8cksHcCOFfXeB3ae+2Y/8kEsJ5PnuFIKQG5Ntjrg+bsmi/1laOnY8jRy5UVP9oXWCyVd278OfLILaSjWYTkQnOpNlLiCflZ05xlkpt7rnwOGxtbjm5IVtuV+PEO6XCOBNZKYLk/rrZWYiuMW75cQ143u9nNputc5zrT2972tumlL33p9N///d9H2cgDtwc96EHTeeedN9lfAf7MZz5zJCefAAy99PznPve5w5eEQB6f/pNjLYd53cq8PJz86le/evjG4mtd61p6mn0HAdms9AamRWS854asfaX0fXFCNyVWawPH8o1/n/zkJw+m5Atq9Gs0Djq2vfdRv71z2HL++rzWfeTsGsPcnlvhwvNjPStglFq1iGPUczIWF+ZjDOw89FyrTWRl3uq4ZFPGfL5g3zcvthFLip+YDPzAb0x+zvkU3oi/JK7UnEt8lOjoHKBvY8S4lm3V177Qhz9pMWb9+catXItjxIG4rE07nhObyFp9HOfYsTHxmARGI8AHgKNVpHE88vf/5EGcvK53vesdvvzjta997Ule5GHb3/7t3x7+5p48HPzlX/7lI5mvfe1rR/2b3OQmR31XR8/Lwzv90nZufOMb66mT+je96U0PY3LBFb3YA8OTDPx84PrXv75reNNjwlU/9PrRj350yFceism3Ko/yknh8L/lm6ZRYba6w9+Mf/xjdSfcxmGIbsnO1umaSv37JXIjXiPno+FP7+ppQm5M+93U/NRbKrZ+ArrvurzUznB+158Za8y+NW9de90vtpejp6/WS9cIe6dpTUvIQGdxDWBup+kvLCQOpO+577RrQtdL5uuIWWeFhdVyyKWN2beA+APbhS9hjDnZlzOpjrrQVVq1t5sSCvLHWEAvGtS3Mpcas645+7N7K5U+PzdW35zFyb+3fx0PWIdagqxY5cbSKPbXuEhv46Z8H0Mda0+cX8rV5iR390jp6PLcfY3LVq1411yTlSSBK4P/9/AEL/wJmFNN6Bf73f//36NdtTznllMPNi/yKrXwBxznnnDPJmHxa6u/+7u+mj33sY4dE733ve08veMELjpJ++ctfPl1wwQWH45e97GWHb+o9mjSdiy66aDr//PMPo0972tOmRz3qUUcSMi7z8vrEJz5x8H00aTriT/zK661vfesk3zyc8hK7uLCnyFOGBEiABEiABEiABEiABEiABEiABEYkIF/eefbZZ48YGmNaIQH+DcAVFi0n5B/84AdH4vK/GvJtva95zWumP/iDP5h+5Vd+5fBtvvJ3/eRbfuUbguX1nve8Z8KvDcux6OEV+vZgkdG/PvzDH/4Qaoe2lZ1jRnlAAiRAAiRAAiRAAiRAAiRAAiRAAiRAAiQQJMBfAQ7iWf+kfiAn2TzgAQ9w/jrtqaeeOj35yU+eHve4xx2Sfuc73zn9+q//+qEvnxLEK/bpOv3xcLGpX9aOPtZy0g/ZsbK+Y/l7hpdeeqlvevPj9uPp+Kj7KInb+CSunBhd+mJD1uinP/1p6R4earseWuf4ORjq/I/NJRSfldWhhfS03Nb7+le+9n4d2HqtfflxDfjI7Gt87nUw2vXZxmP3CDvvWx1Wzyc30jhys/cE+JMymI/FbHNP1YvZDc1bn5AV3745yKylreEYY6Bti2zqdUDrteAYi7OFjxIbWEep+SKPVHlfTLDjm+857lsDrpxCcbrkS+MO+YFN+dDODW5wAxyyJYEmBPgAsAnGcY3Il37o12/91m/pw2P9u93tbtPVrna16Sc/+cnRlyiIwLWvfe0jucsuu+yo7+roTxzaL+7Qdr7//e8HfwU4ZMfl1zUmD/++8Y1vuKaqxuS35vFHYasMdVa2D397sKhJwcYntnJidOnbeOThn5UbqXb2LzCkxIYH51Y3l59ltdXjXteBrfLqndcS10+ugd5VXYf9OdaB3W80ma9//euz3zvIfqH3ColBXthr7LyOV/dz9matp/tzn/vYK+1vn4CBji3Ut7mHahyykzqH2rh4IadUW5DDGoBtjM/RuvIQv6UcYzkgV+Rm6xe6DqSeD7AdamNxhnRT5nSeub6wjlLzzT1nJB4dH/KxtcD43K1eA3Ydhli6cqqJPYXHT3/60xoX1CUBJwH+CrATy3YG5cJ22mmnHSWkv6TjaPAXHdkI5ItC5PXtb3/7F6PTpHX0F3kcCaiOnscXeWBa24ltJvgCEbkQaz3YWqrFxR/tUnHk+g1taLm2lpAX3va9RBwj+Vx7TUdiyVhIgARIoCWB0a/P+h5m9Fhb1qWFrRpeKbq415FYdZ1qYk/xW2Pfp4v4dU4iK8cSU+u44M8Xz1bGR84TNUU7OnMdp+6PHjfjI4EaAnwAWENvJbq3utWtjiKVbwUOvfA/DfJJQLy0Pr5RGHO21fP4NQvI3PKWt0T36JuJjwZMB3bkIaL9JKERLT4s2UBlc8C72PFMiohzxA0th32O7Exom7tBrXIN69rqfq4dypPAXAS4TvuQluvkHq6Vfei1s6qv5ejrtp2ndEs+/3OvF4ljK6+SXKCDNpXFms9tyVXni1z0mO4LE+ig1ZysLOZgF8fSuvT1vK+v9Xz+fLp7H7e87PHc15zUeuiaWx2srVFjt/HymARSCPABYAqllcvIl3zgFfrbBd/73vem73znOwdR/XXnN7/5zSd8TfmHPvQhmHK2mBd90dOvO93pTkeHH/zgB4/6tvPNb35z+sIXvnAYvuMd72inmx7zgt4UJ40tQMDeYC0QAl2SAAmQAAmsgIDdL/DD7QpCXyREy6s0iFZ2Sv0vrSf5gwFaG5OWsXNy7NOrlXXpa385fmGrRAe6bEmABEigNwE+AOxNeAD797nPfY6ikG/49b1kDg/E9MM62cjuec97HtTkk3kf+9jHnCZkHJ/cE3m7AZ555pkTPgX4rne9a9J/508bvPDCC48O73Wvex31W3ckPhtjax+05yYA9mghxR9GQCKv5TrO40VpEtgaAVwDsIdvLT/m044A1oq16BsXubWtK9xL5MYtDPTbMtLH4KXlMQY5HOt4cmOCrdoWsdTaqdH3xeAbF18yF5q3MjHZmvipGyaw1NoOR1U+u7V8yklQc2sE+ABwaxV15HPb2952usc97nGYecc73jF94AMfOElKPnX3j//4j4dx+eKE+9///sdkHvnIR05XvepVD2PPfvazpx/+8IfH5uVYxuUlvz4s8q7Xox71qMPwd7/73emCCy44SeRLX/rS9JKXvOQwfvrpp0/3vve9T5JpObCGi7vEOEKcPWPgDVvLVb09Wz3X3vZoMaM9E+C1dM/VL88d6wZtuaV9abp4yRjeQsPK6LkUWrnyKTa3KmNZbzXPkfPSP7Pw3m3kSjG2PRPgA8CdVP/P/uzPpl/+5V+e5G8APu5xj5v+/u//fvrwhz98+Lbf17/+9dMDHvCACV/g8Sd/8ieT/hVgQSSf3nv0ox99oPWpT31qeuhDHzq9853vPOhLK8cyLi+RO+OMMw59+8+555474dd6xe8Tn/jE6f3vf//0iU98Ynrd6143PeQhD5nkV5GvcpWrTE9/+tMPDxOtjVbHa9uYlowXvtG2qoG2o29ye/kZ7eZwtHh0PUbqk9NI1WAsoxLgeTJqZcaLK7RWQnPjZeKOSHLA+8SJE26hmUb1/Yxmq/s2FMQekrE6PPYTCP35I7/WFTO6fjHZueaxPuBvxBh9MfnGkQtbEiCB/gSu/KaH/r7oYUEC8gDvxS9+8SQP9771rW9NL33pSw9vHZJsKH/8x388Pfaxj9XDR/0nP/nJk3x1+lvf+tbp4osvnuTYvuRB4pOe9CQ7fHQsnyJ84QtfOJ1//vmHh4fvfve7J3nrl3xz8TOf+czpnHPO0cPd+rIZjXyTJbFhw1wiVvhGAXrH0KsWveyCS0lr2ZbYoA4JkAAJkAAJtCLAfakVySvshO499P0dvIbkIZPTop7w1dp+KBb4hsycvuFTWnkILB9WyH3Z+HP1e8ujpuJHYg3xjc33jlXbHykWHZfta561a0Hbsn54TAJzE+ADwLmJL+jvN37jN6a3v/3th0/avfe9752+/OUvTz/+8Y+nG97whtNd7nKX6bzzzptud7vbeSOUT+U997nPne573/tOb3zjGw8P8ORLQ653vetNZ5111vTgBz846aHdaaedNr3hDW+Y3vSmNx3i+exnP3v4e4DyRSN3u9vdpkc84hHTrW99a28ce5zQm/we899qzrwh2GplmRcJkAAJ7JPAWn6433p17AMLe7yH/EvvsWpZzXkO4OeDlFxtXLV55q4hxDi339w4KU8CWyfAB4Bbr7DJTx7WPeEJTzi8zVTyoXwyr/bTefJ3Ah/2sIcd3smOOwrazQibVEeX2aaxyWcrVios5Rdh21rYWkGOLQmQAAmQAAmQQB4B7LEpe6vIQN560fohOas30vFa485hiDr56phqq5WdVH+5crn5IZ9cPy75XN8uGzaekM3QnNjWtnTf5bd0TGKI2Y7Nl/p26cFXjI1L147Blh0vPYa9FrGVxkC9fRPg3wDcd/13nb1ceNd28V1bvCMtMGy4I8XEWEiABEiABEhgKQI59xTYQ6XVb1/skPfNc/wKAjk1ALMU/pD1tTX10bq67/OVM47ccnR6y0qN8O7tq4f9kjWGOHLyjvmBrZgcfJe2ek2G1lNoDr61LYyVtPDVyl5JDNQhARDgJwBBgu1uCeiNSC7M+rgWiu9C39JHbYzUJwESIAESIAES2BcBex8ix757lhCZkE7re6pQHK3mdD6WUSsf1o5mr/1bOddxrry2kVsfn69cOzoG25+Lufj15WNjCh3r2oncnPGH4moxp3OxecK+lpExewy5OVtfrIhBvhRG/t48XySwVwL8BOBeK8+8nQRablyhGwuZC807g+NgNQEyr0ZIAyRAAiRAAhsl0PIeaKOIuqUl7Jfg3+q+KNVOqlw30MpwKm9fzNBHq0xXdX3+XEZFFm/XPMZaxwi7a2l93wQNLjnMa3KGvxob1CWBWgJ8AFhLkPok4CCQupFg09atmNPH6DvczDZkNyyJaa2vNce+VuZbi3uEc3JrTEfLhzUerSKMZy4Cdr+v9bvmPRfXgd45zOXHV8tW+YXsYF1Ji3x98Yw47opZ5yt5Icea+LXNmJ0cWbGVG5+1n6sfi7/nPOqBVr4JGi/dlxzlLXLo67x1zroPW7ktfOXqUZ4EWhLgA8CWNGmLBBoQ0BtPA3M04SBAxg4oHEoioNeO7icpU4gESIAEVkjA/uArx3ivMJ2hQi7ZR3qwL4kjFaRdP9Dr6RM+Rmljucp8TGaJXHRMur9ELC192jUZyk2fb7rfMh7aIoE5CfBvAM5Jm742R0A2jJxNxAfA2tByoU1Jy7n6Wjfkw6Vrx0Rf23PlbnVGPta5pLDR8sgrRQ+ybLdHgPXfXk2REWsLEmy3TAD7ml7vGEPemEOLcbQYt3oyjznIsq0noJmi72Jf4gl2Su1Cz+Xb2oaMjIf0INerhW/El+snR68kVx0f+rkxanmxkROz6ObKa39z9G18qZwgZ/VTYq7RFfvQT/FFGRJoTYCfAGxNlPZ2QwAbBlpJXPdTQcQ2gdh8qp8eciX51sQxtz/EupRf+Gc7BgGugzHqwChIgATaEuC1Lcyz531Yju0c2XBG4VlZD7VrQuv7+uEo9jOr+cyRdct1NHfsKXxyY2rJIyU+ypDA0gT4AHDpCtD/pgiUbCKpG5XI+WRdc1bWHueCr9XP9Wfle/ovte3T843bnHi8XgL6XGe911tHRk4CSxNY6vph/eJYX9t0P8ZJZO07prOG+RwGpfnU+qjVL43b6mEN2fHYcalezG7KvPgO+ddrOsVeTCbkK1dXx7bkGqjJKZZzyXwJC6szWk4lHKhDAj4C/BVgHxmOk0CEgN0sIuLe6RQ7IuPajFxjcGR1UvxA17YhPzJXY9v6WuK4Jn6bP1jZ8SXyos+2BPQ5hTrXrJ220dHaiAR4HRixKsvHhOsHIllinejrGeJAu9frWu+8Q3W2vu0akdpgzMrKnB2DrMzpF+R881p2y/2U/MEKHHCcq5siDx+21brwb2V4fPL6b8FE2IeY69q08EcbJDAXAX4CcC7S9EMClQRkEwptRC7zJTouO7GxvW+CyB9tjBfn10tAn4Poo11vVoy8JwFeF3rSHd+2rb89bpGB2MS7hT3a6EMgZ68IybLWfeqTalVqgzd0QvWyc6gfWthAa+UxPnIruSz1asHLF79vfKlc6ZcEWhDgJwBbUKQNEvg5gSU2idRNL1WutJC97Utc4qMH41axu2KTsVb2S2tDvT4EWNc+XLdmtdd1a2uctpoP9gW0oTxb7Rcldua8nlkWc/rW/OFXx4M+5rT8En2JAzG18t/ano2rhX2x0bsGMbap/rWc7lsuei7GKDZvbecel9pHDqX6uXGmyEssiCtFvpXMSAxa5UQ7+yHATwDup9bMdFACa9hE7OYqx3psDTm4yq9zcM1zjARcBLhuXFQ4RgIkMAKB3OtTrnxNjmu9V6jJmbppBPTakDWpj9MsrEdqznPORUX84+2aj40tHb+Or1Us4JFib8trU7Nlf7sE+ABwu7VlZiQQJZCzieVsjlHHBQI5sYbMz5lHyo1EKNbUuVZsUv1RjgRIIJ3AXNeB9IgoOSKBmnVSuq/V+GzBcG7/J06cOOnBh2aHfovcWtlwMUKcdu+XY7xb+Z/TDmK3ec0Zg/hyMW8dQ4kP8GkZS04c8O+rj2+8Zbxz2ML5hXYOn/RBAnMS4APAOWnTFwk4CORsvg71RYd07Lgx6BWQ9lXqo4WNHN9z3gzN6SuHAWVJgASWI4DrMq8P89eAzOdnviePWF9o95R7q1xd94SusVb+rB3x5Xpbud7Hc+bcOxfaJwESiBPg3wCMM6IECSQRwAaaczMGnSQHHYS0f8Stx2IuoROTW3o+JydXrNAfOV/E6IqfYyRAAiQg1y9eJ7gOliIw8v65BBPNI3ZeuuZlDDbQ+vLQsj6ZEcclL5u7ztXOleTQwkaJ35AOYkKu9hjj2gZk9FhOX/RddnNs7EU2hxXqQrZ7WR3ryJOfAFxHnRjligjgYh8LOVUuZmfJ+TXksIYYl6whfZMACZAACZBATwL84fdKusJibh5z+7sy2/qejl33cyxfcsklR+LSL7VzZKSgU+JT7l/1Pazu2xBK7FsbIftW1nVcq++yyTESIIH2BPgJwPZMaZEEjm3YwKE35y1tkpKL5DZiTvL3fvbwGpH9HrgzRxLIJbDktZLXidxq7U8e9ym1awV29kcwnrGw1XxqWcc95ksgJh1nvpV2Gr44ZByx+ryJzOWXX+6b7j6uY0c/FrMrKOi65mSsxKbLltiJ+dJ6rfxqm2vrg0GIG2TWlhvj3SYBfgJwm3VlVgMSGP3iL/GVxliqFyqTbKR4i1xoYw3ZaTnXI8+S+DSXEn3qkAAJ7IMAruujXLv2Qb1svxphj+tZH6zFnj7WaLum7r3P67XVDPdGaGU99GYUWnM1tbV2Q3mE5qydlGOfPTtuj1Nsl8i05FjiP6aD+ObiEYuH8yQQI8AHgDFCnCeBTAL6xsOqyubADcJSiR9jc/VJgqu0o33qL7QefPmExjUL3Q/pcI4ESIAESGBsAlu+N9C56f167IrkR4f9PrY3awaaTb7Htg+4EJeNwzdu5ZY4TmGu7wvxTdBzxFpbWxsj7KEeaK1c7XFo/cInYqn1laI/p6+UeEIy4AOZEEvIsCWBuQnwV4DnJk5/JLAwAdmM1rSZxnDNnYv4672h59jfWj1j9eY8CayZwNzXqzWzWmvsOdfvnBxddlutJ+wjrezpvHrY1PZH6bvqM1dswrjGf2qNav3MxWMkPy5mON9K4kytVYlt6MBHzZqCrS215LGlau47F34CcN/1Z/YdCXCjqIMrNyD6XWetrTbiyrHK9ZBDi7IkQAIksE4C+OF5ndG3iRp7JFm04Ukr2yOwlnvCPZ/DUqO11Gl7Zwgz6kmAnwDsSZe2d09ANo7RNs+54nFtmvCt5zAWWyypcjE7S83Xxm/1NUP0rcxSudIvCZAACeyZAK7FuDansBBZ6Gn5HBtaL7fv8y92XHO+uFw5hGKxdnL1Q7bnnFs6bvFvWbbOv7f93HgtcxufHFuZXB8t5HUMNsYW9nvb0PH39jWi/ZKaCTPood07xxFru9eY+ABwr5Vn3t0I4KK/5ws9NjsL2TUuY2tlhVrbPFse+9jM4btlHrRFAiRAAiQQJ+DaJ0ULe4GeRx9zLuuQwZyVtfOQQ4t5tBhnOx4BXdst1kvn56Iv86PljXgQO1pX/K3GxGetH8TdKqaWdlrk1zIen60R16MvVo7viwAfAO6r3sx2JgK1G+9MYXZxM/JNQ5eEGxrV7FLWEGS0Xm44a7mRys2L8iTQmwDOO5yH4o/n0xXUXWx612ML9sFN56LXl4zrY8i71h3mtC3p+8Zjc9aO61jH5pp3jYkOYirRd9nc45hdA5ol+Pq4WF2XXMyGS6flmM6npd0WtlLYpDBuEUuuDRuXzmVE5ojPxp2bd2t5xAVmOG7th/ZIoAUBPgBsQZE2SIAESMBDADcDnulmwyV+cIMy2o1UMyg0tBsCS6xhnHNL+F5DYcFnDbH2iBHX1xrbMYYyDz96HWKsxneKbiy+FButZZD7iLG1zBV5ptgEixwdbbdUT9uYq6/PiTl8prJBDeaIKdeH5DA3t9wYXfJg34MtbLv8hsZ8ej1iDMXBORIIEeADwBAdzpEACXQn0GpTxA1M94CNg5bxw3SuzVx58aNvUkr0EStbEhiJwBLXAX3+6P5IXGpiwbUiJ7cc2ZrYRtUFs5r4UhmmytXEYnVb+xR7wqzm/LXMa2zZfEc7trmmxqfrZm3gWMvALsYgg/G5Wvify5/PT2n+ofjFZmjeF0toPMceZEtz03FoG7Cr53v0e/jRefSImTZJYGkC/BbgpStA/ySwEAHZ4Obc5GST1hu1PV4Iw+7d6prsHgYBrJqAXM+4nvuVcM79ol8W/S2PwKnneSC2R8ixfyX36cFVWxlzje+TUHrWcq70PBfTI9mO5NJMWc/trKU9Z8JPAO65+sydBDoQkM1R3yjazdIetwqhl91W8cXsrD3+WH6cJ4GeBPZ2/oSusa05741ta3659lrxhh29VmKxiA7koe/SCc255FPGam3q2MVfrb2UmJeQQX2s79x8LS9tz+dDy/Ts5+bSM5Y5bIN377zFT4qPFBkXl1I9ly3XGDjJnMsX5l1zLnt6DLp6zNcX+znyPjscJ4GlCPAB4FLk6ZcEBiEgm1jJZhkKv7W9kC/O5RPAjQvrlM+OGiQwGgGczxJXj3Ma9nNsl+iMxrVXPJYjWI3iz8bXKy7abUegpGa9111pdiW5lPrK0SvhJTqxfGQetlPka2OGr1hcOX7mkEXcMV9ryyuWD+dJoAcBPgDsQZU2SWBwAq1vMgZPl+EpAriJ4k2SgsIuCayUgP7hsVcKNdeKvV1vkK+vFi6WrjGffotx+IvF2sLX3DaQ29x+5/QnObaoXSs7LXIfvW41vKGbmiPkhWuqTkkNxI+2r/t2LsW+jhvy2ibGcluXXbFhY4ScHY/5g15MjvMksCUC/BuAW6omcyGBRAKpm7JsjKnvRNcUW5iA1D61/guHunn3OLc2nygT7EIA5zHaLk4aGd3DWpccY68UmZiNVvN23djjVn5opz2BVrUSO61spWYJn7pN1S2Rk3PukksuKVE90kGsRwONO74a9Lxe+HxKajInvvEuTbdn/BKTjs/mo+dc8bvmxYbrbfV752X98ZgEWhPgA8DWRGmPBDZCIHeDy5XfCCamQQLFBHCjyXOnGOEuFWW92B927PGoYCR2rvfj1VmSCa5BxyMa96jF2unFG3bRgmKLmGFLty3P+Za2dIxb6veqYy9GtfHmrolc+Vje9jwKyVtZnTvm9ENg3Yfd1Pi1beiyJYG1EeCvAK+tYoyXBGYiIJshN7qZYNPNZgjIOeM7d3w3mJD3zW8GDhNpQsC1TnCtRqtlMAbneg5jrVv4sL5b+xnBXm6OufJz5Ih6zeGr1Ae4SVsTr+jCVmksohezoedrY7ZxatuIpYaJtd/zuDULV6ytaqxtl9jMqQlkW9TW2rB56GP0QzquOcQb0xddl6wec9mHXbYkQALtCfATgO2Z0iIJkAAJrJaA3Ijhvdok/j97bx5rW1ZVD58fFiUqidIIdi+ASjQIRDF2QIQAoon8Q6sSmoCARkhUTNQEwQSDMWJEY8CGTiQkAgUEMYoRoviHiR1YQIgNFGiBUCCICkiVwPcx73Pcmne+1fd773GSU6ubc8wxx1z77PV2nXvvZOIlh1l9GJ5Mf3fh5f/2l9Rky0L48l1hn63AYURtJU+8R8RjjHQF9rgHfdd8qip702RWPoiLNlV/n10uTuk+yI2j+cZiyrq2qYml47JPBahAmQL8BmCZbvSiAodQQG7S+qYdS1psj3Bj15rsLV9d86PUM7avW6zvbZ+00GQ0xp73s95fNk+9NlrzmbF756rvAzqWK2fM+Xy0P/vrK7DXOh5hn8r/ELr66quLzqqiT0rtoWPJToZvShyLH/MBtvXDODU/sY/FAqZutY/mkhNX47Xsgxt4YdwyBrGowCoK8BuAq1SCPKgAFdiEAvZQYMebSIIkhyqAA+XQoAzmVOAI1+tK++0Iejs3WubkSjXLpD7MXGuk95XuDyOTGUhzz3TNMm+lhfDtxbkXbpZQhcYxfVvmlluDGLfUlFvmEIspnMF7ZNwQL80pZMc1KrBlBfgNwC1Xj9ypQEcFcFPuGCILGnxmHRIQ30Va1mbxcvFJmUM+ljfmgbHF3MB9Rmv1nMHBFZN1vFkVanGzFr17q14PLfJGbvozE3MufG2n10M+2o79iwpoPdEfpaXEQcyLrC7/BVXMic0oToi5cjtai1CdbP1C3KytT+NZ9W7NL6SbL/fcec1Z93NxaE8FqEC+AvwGYL5m9KACu1dgazfj3nx748/YUHLAwyFP8sN7BhfGpAKjFdjjNT1aQ8a7qEDoAcJFS472qADuqTo3uydG3Gf52aYrcPmB7KVLly5OOkaoDfSzrcNliSnwbE3G7t3W+MSjAlRgngJ8ADhPe0amAlSggQI4/Egrb/n9Lny1VQDatkUlmlYAGkvb48XD/JWq9tL6ykic2bMCrgc/Nl+91+y1qK99bWcxOL747TqXHiX6QX8XXmgOdUcbsvWt1cT2Ydp5xEBr1482Tn0YKLqIZnt7leZkP7egC/a/bx12bKkAFVhHAT4AXKcWZEIFllCg9HAA8rX+wFmtjR1uJO8t5m7zwmFuNf3JhwqUKiD/4LP7vBSLflSgtwJbvI/01iQVv/Q6L/VL5WXtWpwXQpyB79pLrjnLb+/jkHZbyn1kLVNi7UXXLe0BcqUCJQrwdwCWqEYfKrBDBVJu7qlpC9YeDgLQZA+56NrZ+rjyc81pDPapwNYU0Hsa1zZavba1vMh3fQVy95f9jF4/w3EMRUtct+Oi1kcKcS6td6kWmkvu3oQS2k/jYd222t6uzRiXajeDqy8mNE3RX2OU7LdUvVLtNJ8991GjPefI3LanAL8BuL2akTEVOKQCchN13Uh98zkiyWHI9RYMHVP3XfixdZfPyDnkODImY6UpoPcO65SmWcgq9qsARG/9DmFxjQq0UkCu7dQXPwfcSuVo6EaYM6s/410MSvOK4bpi6TlfXN+8+OqYITsdZ8W+vgfonMAV6xiH2hzbEE7JGmLbtgRL+wgeXlJnPca8bXvvhxQOlhPHVIAKXFSADwAv6sERFTiEAvYGbcctROiBGeKFg0/K73fROLk87eEDce28jrFCPzfPFTjP5pCqWapdbj69cHN50J4KUIE6BeRalnfJfQK+dQzo7VJgtc/YWXxq4pbsaVctVpjTuaBfo83snJBDDY8Qhl0bqZWNXZMjfanA0RTgA8CjVZz5Hl4B3KDRiiC9bqQ6RorwYp/rk4JbatNLl1I+pX5W073kVapHih80Q+vzwTpan13KvNTFvlP8aHNRgRa1uIjIERXIV0D2Id7wtmPMx9q9fGaX5q/1aaWFcOn18mG34t6Dt+as+7FYObYxrBXWcQ8GF9/YzsN+z63UGnlLq1++fWDttE9pH5g+LhYX9nY+Nga+tcO8bq0Nx1RgZQX4AHDl6pAbFeiogL0h2nHH0FFo30Ei6lhg4MvbN69DpNho+xX6W+Q8UzfqNVN9xqYCVGAvCrS8r8vnMt6ij+6X6AVu0uJdgiM+GqsEA/6pvrn2qbh7snP9Sohc3Wr32Ep6Si6lrxzdEAdtacwUvxYxBAPvUEytgfT1OOTHNSqwigL8IyCrVII8qIBHAX1jaXWD84Q6u/HpeCE7vRbyCa1pDN1vkafGi/Vz4+Xax+L3Xt8a31562L2Yoov4hOxkDbhoQ/a9ciPuzQqI/jfeeOPNE+xRgQkK6M+GCeGXConPRpDCGBrVfGbW+IKPtOAk/RJM7S8Yvhdy9q3LvGClcEiNGYql11LxUu182Cm5ad/SPnjedNNNV0CgDqO4XEFg0ETv/KCxTad3XBtPxohpOcnYtQcshtgBA621wdjGwDxbKrAFBfgNwC1UiRypwEIKuG6KMueaL6HdCqckNn2oABSo2dM8GELFeW3u7wKdx5SRqcBFBfDZg3sh2otW9SP5nOrxWZWDmWNbnzERqMBFBVpeW9zLF7WdNUqpg5wPXJ+zej8ITgrWrDwZlwrUKMBvANaoR18qMEABuSHJTUjfmAaEdYaIcXCt59xAXf5OIpycogD2IdopJCqCyv7Cfkzdayl2GlfopfhUpEFXKkAFNqKA/WzIpd3jswSfgeDS8vMc2GgRw9f2yM8Xq/d8Ti6wTdWpN/e94ovO1Hh8dbG/deTetaipM/gKhuVZg6vzZ58KrKQAvwG4UjXIhQp4FMDNybPcdNoXyzefG7wVTm5c2rdVwHUocs21jVqPJvuv5x7siV2ffRsEqfMWat0m222hsC7r1SvnMyHHtiRT3/7wzZfEyPGZFTeH4wzb1H0gdqm2M/LwxRxZd9En9I1w4dKCz6g6tODqq0vJfE7eObY5XLQmNoaM7ZwPO9XO5895KrAVBfgAcCuVIk8qsFEF7A1V36h1StZOr7E/XwGpm9RI10/6ejyfJRn0VkDXm/XvrXY+Pj9H8zUb4bFCXfS1OyLnljHwWdMyhxVqAo2Ei35jvke7St4ta1mjE/Qo5YO61XBI8cU1kGKrbUr9gOHSxc5pDaAn/G0bW7f2dox8dKttZF7zwZp+CCx/FMbmIHbAhA9wajkDjy0VWEUBPgBcpRLkQQUWUmD0zW50vIWk3gQVHJTQ+kgfvY57zz9Wf9++4DwVoALuXw0gnxn6PUunUZ9dLeLYf6TP0qxV3BaaYA+14kScfShQe8/O9U/dy6l2tVWIxdnbZ0mtXvQ/jgJ8AHicWjNTKpCsQO5NPwTswsJhFW3If09rezxsxA5Ye6pfKJeeOriuoRCX1muxfYv12Txb5008KtBaAdzz0LbGn42Hz0G0lo98RvjWrC3Gvs+VFp87rjrk8gPPkrZ1LB+envfpWcJ/Lz6ij9Zoi3np6wF9nYdrTq+7+lvVBLnqVr7153uJnevlm3fZco4KbEkB/hGQLVWLXA+hgNxwVrzprsprxKZwHQJKa1TqNyJPG8OVt8tmSzlZ/iuPU/TvwV9f67kctG8PbsSkAqsrsOo1gM9pe01jPlXXmD/w0ApuzMcX2/r57BSp9E0AAEAASURBVGrnNddarNn+kotLN+ToWpvNeWZ86AF9ZnIJxfbVVXyQQ8w/tO5bc+mSEs+Hh/lQPrAZ2fpy8s2P5MZYVKC1AvwGYGtFiUcFKhTAjUZa9Cvgil1dN/xisA07tq7DlnSduf82vGU2Tx11r9n7wNi8GExg9wq03Ks118wsoUvuSaU+Pf1a1nFWLUbFLalDL24r1W0lLrl6x2oaW8+NV2u/xc/K2pzpTwVWUoDfAFypGuRyeAVwk555EJHYwkPePXggx9WLHcsdOq2eRy6/WN65eLQvV2D0tdKq9qN5lyu8b09fPVmfi3Xf62f5xSxvHpXWHzr59tXNEfy90th+xLVXoNXovHU84WB5yDrmZisoPDTfmXw0F63PCvw0H5dG4KjtMOeybzWnNUvFFF4xP51HKi7tqAAVSFOADwDTdKIVFRiqwIibtk7I3mjtWGxbcGqBoXn36rvy7xVr67hbqWlPnWW/7E0H5JN7LcCvp97EpgJUYLwCM69tHdv3mbTHz+EWVRbtoBnaFrh7x9C6rZBrKh99rYzijWsPbWpcux8199R8U2Ol2M2ImcKLNlSgtQJ8ANhaUeJRgY0pYG/ALemHsF1r+ubfkkcOlotXjv+RbFeo10y9sVf2ooMrD9cc8ob2LhussV1PAdSPdWvzP7Zshfeu68z8ZO8iPvax1l+v6/nZfcsVOYzgZWOPiJkTY6QWKbxW3UPCXbQSfitwxL7SnIQj5sFXWtcLfq41zMVsZF1eOiZ8S9oUnBgnHRf89Bz7VGAFBfg7AFeoAjlQgQ0okHJjjKUhGHjHbFusy1/9yonXIscWvGdi8MASVz9nT8XRaEEF+ijAa7mPrkSdqwDu06797Zqby/bK6CM5QqsrWXDGp4Cuj/T12Oczch58tlDb2RxRP2kvXboULRO01YYaA+totR37VGBLCvAB4JaqRa5UYLICsZu5rOt3Dt3eN1TwspwwLy1faQrgQJRmvS8r7pPt1RP/I2B7zOsZH/larVcvD4Fa5+nVwhqao9WYuLfL9T/rpXnp/ig+ErPnqzd+T+7A1nXZSj6aM/KY1YbORD49cW2Csy+fELb4WpwYHtZ9rSueL4YPg/NUYAsK8AHgFqpEjlSgowK+G7QvJG6GrtbnE5vP5RDCix32wTuEwTUqkKJAy32bEm8VG31YP6oGq9QilYeuk65fqj/tqAAVKP9RQ3397UVH5LSnz5Pa82Gqv9gd4ZWqxxG0YI5UYCUF+DsAV6oGuVCBCQrMPIjgADkh7fOQwmGmBudE2NmEAnbP2vFqSWBv9+DZA7OnfqLF1ji31OPIubfUcQtYPa/72vzBDTgl+3LWtVzCFXnOaK3WrThYHXrFacW3FY7OU2vgmw/FFR+NEbJdbU1465xj/GK2el1rgr5ej8WCT8yuZj2HT00c+lKBXgrwG4C9lCUuFdiIAnKzxDuV8ogbbCoXbZd6U7b87VhjQhttgzltt5d+qoZ7yXfveex5r+69dsyPCpQosLXPcOELzmhjeeN+HLLHZ1/K7/6KxdvLOnSrzcfqXoNrsWq59fAXjpYn5nzzPXishInrC20rblZPwU3dX6l2JVx99S7Bog8VmK0AHwDOrgDjU4GAAq4bYcC8aiknVo6tj1TPG7UvZu685ShjO5eLuaJ9i3qumJdwiv1I+Kq8yau9Anu8dturtE1EfIZJi/42M9k/a991iLrZGmKMdauQb97aHX3s071El5ZYJfF7+4zaU6Pi9NarNX6pLuKX4sv/KdC6YsTbmgJ8ALi1ipHvoRTY0iFLuKbyTbXrVezYASEnl14cR+Dqw5LubzF/8Le15UFvxE7qF8PWsySS7IHZnzklvOmTrgDqi7bFvkmPvo7lVvJGnULKSS46H/HB2/ppO7t25DH0gt7UKW03QK8068vfUIPW8EnVOtUOuLNb4bs1zima5dY8BZM2VGBVBfg7AFetDHlRgcEKyM0v96a+0g0zl7tP3pk56Rxm8dBx0de8fLrNnheuW+A5W6etxcce3Bpv8qUCMxTY0ueg79rWn+M+G5e24pdj78LoMVeaT2sumkcNdg+NZ9dOa+PKT+a0jdbPZa/X99YPadEyV70nfNq3jEcsKnAkBfgNwCNVm7lSgYgCqQcZsfPZ+uYldMgvQi24XHM4EN8a/yCxzMWQdplQSeaoh26to0sbaIbW+swaI49Z8Rl3Pwqstrf3o2z/TPg5cLPGrs/vm1fX7KF+0sZesBW7FHvgjdLFxrFj8Ondjoibo3/vfHPxsY9COcAmB1t0d2lv4/jsXL458bdu69MllleJ39G1jmnK9X0pwG8A7quezIYKVCnQ6gaIw43FwxjrVWT/zxmYuVji15JHbnyfvXAqzcmHWTPv02gljjX50ZcKWAW4t60iHFMBKkAFLiuw6tlpVH1Kzmipmm3x3uPjXKJT6xqm6t46LvGowOoK8BuAq1eI/KhARwXk5qjfrUOFHh75Dg2pHMA71d5lV8vBhdlizqdbKnYLbWKxhCPeMdtV10fotGru5EUFqMAxFDjK51zO/VzuXdTl8v7HeQNtylXh0i5H/5QYW7Gxebt0TJ3z5Wxj+OxGzbvyccUWO7xd6ylz8E+NaTFd2vGPw1mVOD6aAnwAeLSKM18qoBTQN1TcZGUZ83oObljDeFa7Co9Z+afEdR18UvyOYANttriPwP0IdZqRIz73trg3ZujFmNtQQD43XO9Z7MGlZfySz8YSnxrOo+OlcC39rEMNV8wpJW+xaZGDzV/rKX27nspN24GnnluxL/nq/Es5unBKcUU7PvQrrQT99qgAHwDusarMiQpUKIAbLFqBwo1Yz6WGCPngQIPDEcap2DG7o/0FWJd+0Dam1dHWsae3mHfomtpiPuRMBajAPAVm3SPwOYb7FtpcJSx/O7Z4rjgxH4tROx4dz8XXdQ9ETVz2e5wrrUOKTlpf6bti+XDg61p34WyhNjYXnaNrzZXTSrlbzi6+nKMCqyrAB4CrVoa8qMCOFMCNPpSSvrHrfsinx9rM2D3yISYVoAJUgApQASqwlgI8a8yvR++HOKgx2ljGrrOyy9c1F8Putd5KwxjOSjn30pK4VGCUAnwAOEppxqECCyoQuqGG1kpTcR1ufFgp8e2BQcb2W3/WxhdPz6fE1vYr90vyXzkfcqMCVIAKUIFyBeSeYN/laBc9ce+UFm9YYE3Gdg02WNPjWD/nHuf7McAcjBgfrOt8MbeFVmuh98kWuI/kqHWSuL56++bBNaYx1mGPNoYLuxEtOFpNdGzL145hG8KATa82J7aPfy9uxKUCLRXgA8CWahKLCmxMAdy0LW3c2NDa9dpx6k1W4oc46LUQpi/P2jxW87d5hjRZjTv5UAEqQAWoQF8Fet4TcD9Gi0xkjDndx7qrhb1rzTVn730um9AceOXGDWGmrPWI1wITekIXyaXn3hF8HUvGI141Wlnf3vqM0KM2BmpotbFjxNHz6KOFDdpR+raIIxgtcJA7WyrQWgE+AGytKPGowAYVsDcqO56dkutA4JprybM3fkuuFouHD6sIx1SAClCBYyjgu3/75vemyip5zuBhzy12jFrP4IbYsdZylrGdi2GkrNdiQkNp0U+Ju2cbrYOum563+Ws7uyZjvR7CcflyjgpQAbcCV7mnOUsFqMBeFJCbZ81Ns9Y/VUfNUWLal53T9tbWNxYfi+OzHT2veZXkNppvr3iiw5Hz76UrcakAFTiOArWfobgf1eLkKN4iVgxDfkXIhz/84XNayBMTMX/YpbYpeC3ueTYPYEp8uybcsR7LA/xT7WN4qes1nCUG/MHfFdenjcvWNYcYo7UBl1BusJnR+nj55qGjcLU2Lm3FRvv0yLEkRolPD+7EpAIpCvAbgCkq0YYKbFQB3CTRrpQGbvTSor8SP+HSSzeLq8erajGiNlqHEfEYgwpQASpABY6rAM4faGcpIfe+0fe/3Hi59j201Bx0vzSWPW+1wAxxaYVveYdirr4WysW3JvN6zY5n5aw5zeLAuFQgRQE+AExRiTZU4GAKtDqkuGQTbLx9N8tVbuYu/q3nemrdmivxqAAVuPlHknjtcjespgDurbV7c8Y92HLX41g+dl37rlajEB/N2+YU8ktds3XtESOVS6kdNBJ/H3/kCVv9x1+k7/MLYfr4Siy+6hRAvXJRrJ8d5+LBPrQ/YONruR98ynB+JQX4AHClapALFeikQO0NqeZmqFPKxanlrWOj3wMT2DmtaGH1WIVbTh60pQKrKGCvpxG8ZsQckRdjbE8BuxfteEsZlXAXH7x1rvrhj54P9Uvih/BK12p4uLQADzlr4I05X1vDwYfZcr6UH/QpPXfF9PPx8s231IRYVIAKUIGQAvwdgCF1uEYFNq5AycHGdzjBfC4m/EqlzI2XEkdj1vJLiWdtJL7EBY8ZHCynmWOdPzSZyYext6lA772j9ykU6h0TcdhSgZACrr0p9jK/lz0aygU5+nQIaWfXgIFW1oGvbfW6z0bbl/YRx8UhB7PWPydWja2PJ3RwYWNN+2LOZW/nxC/H3vrnjiWW5prqX+KTij3SDlr3ykfjIlYsP+0Ts8V6iQ982VKBmQrwG4Az1WdsKkAFDquA6+CQelA5rGhMnApMVMBes3Y8kdrU0Pzcmir/MsFXuB6EA94lwuTsZZtvjm8Jtxofy7UGa5ZvSg6oAdocrtg3KXFycFvZrsqrJr+SOtXE8/nuUVtfrpynAqIAHwByH1ABKpClQIsbNg5aWYE7GKfkkmKTQ03j6T4wXHNY21srudp87XhvOTOfbSuAzy7+g+FyHXG9ot12dY/Lfo/1a3GN4nr37Qwbo0RHxLCtL6ael3ixmJaj9k/pl/BKwbU2iGPn9Rg2toWNzMf0gK20Yj/jNSvujFxdMXX+OfVyYbnm9HUhsXQ8ax9as7YyBl6unwuLc1RglgJ8ADhLecalAgsqkHIjzrnppeDNlCEnl5Y89eFkdY1a5m2xrP4ytnPWh2MqQAWoABW4qIDvc9M3r71xD0Kr17bel/wvXbp0nkbO7wNM0U6ArZ2+v58HzuzEapEaw3LLpDHMvJan+AMD7TDynkCr8PDQmz6tazadDAlQgYMpwN8BeLCCM10q4FMgduD0+fnmW+P54uxxXrQ7yuHxKHnucZ8yJyog1++RPq9WrjhqAY4zPlstB+Giecw+F+iHgdDJ1WrOrvXYXK2/C3+2di5OtXM9dBJOgpuiV8lnl8UN5RBaq9FOc+gVo4Zfjm9r/lqblPrCXvOQPuZ1LtpGz7NPBbamAL8BuLWKkS8VmKCA3PTwjoWXm6bvxpmKEYvRcj3GSdZbvVy6xLChJ9qYPdepABWgAiMVaPkZOZL3XmPF7mk271x762/Hof2QGgt2trWxOA4rUHLmCCNuZ1XvHf3gV/clG5ytoJVu0fdlLTFmvErjxvKZkUsspnBO5W3tcnTScSxOjCPXqcDWFOADwK1VjHypwAYVyLkJz0pPOGqedtyCl8aP4enDiLb1zWsb9qkAFVhPAf6jYr2a7I1Rzj3G5l7ja7FC49h1MIpHiGPumuas+7k4PnvB7IHrizdifmY+vtjYm2hH6FAaw5dDCK/EJ4TXc01qkFsHyQ85ok3hqG1DMUNrKXFoQwVWUYA/ArxKJciDCkxWwHUD1HMp9Fw3x1yMlDi9bFz8W8ZqiQ+sLenbUktiUYGtKcBrdWsV2xZfuSessMdwb4J6dox5Vzubv8QXvuCcwyfH1pV7ypwvBvimYKxiI5x9+azC0cXPNTeL70pcSjTAvk3JQ+8X3XfFtXhif9NNN7lMz/YgeDgNzKQrtva3sY07h1RgCQX4AHCJMpAEFVhLAbmB6RvaWuz6sTlq3v0UJTIVoAJUgAr0VmAP9+sV/+FsdV2RY++9tVd8nvf2WtmLedlr+OLq5R8Bt3Mp4xhuCgZtqMAsBfgjwLOUZ1wqsLgCJQdd8YGf7s9MVW7SeKfyQA6p9il2vQ4Lubi5WqTkRhsqQAWoABWYrwA+33PvCy2Z594/VzkrQINc/vDr2ZbWs0cuPTB7aYfrwYePvWdbsd9Snr78Vp+H7rk8U2tTet3E+PTCjcXlOhVopQC/AdhKSeJQASpwrkDqzfnc4fMduaGW+GmMlD7i4AaOmJgXDN1PwVzRBvlpbpIr8tXz7FMBKkAFqAAVGKXAFu9Dq3J23euljqvyHbXHauJsWTvshy3n4Krd3vJx5cg5KjBKAX4DcJTSjEMFJiogBwK8J9IIhu51cxdcvHEwQmsJYb4XFxuv1djWFnkAH/ljLK210Wtb6G+dPzTeSx7Ih22aAqx7mk602ocCuAehXSkruRZj1+PIM4HmAr1C8bWN7q+k8UguqKe00GNk/Nmx7F6BHq159cJtzZN4VIAKXKkAvwF4pSacoQKHUkBu4vLCoQGHpr2KgDxtfpjHoQZja7fyGLWLcUfNYb9yTj5usRx9fqvN7yWP1XQlHypwZAVafa7gXpGLJ/bw3UodVrgfbkGzlTmuzG3kdYDrdS96IJ9RGtp4e9FxlH6Ms74CfAC4fo3IkAoMUeAINzh7Ux8ibOcgkpPULjU32B6h3p2lJzwVqFIA12DqtVsVjM6HUaDlfgIW7ht7FBE5IjdclxhLOyJ/G9fy0nzYjyuwd/1S9ktPDXpi29zi1XZbCMdcLFdemMvFcrPiLBWYrwB/BHh+DciACiynwFFvclvIWw4i+i2bB4eTlI2UY5uCRxsqQAXKFOC1WKYbvcYrULJXxQfvEGO576507y3JNZRf7lqKZrmYq9qPqLvEuP76688lkD7ioj1f3FBn9j7tJZWrJq65kfER60jXJnJmu08F+A3AfdaVWVGBCwrITSvnBrrXg8UFUcwgRx/jWjSExq64WCsCTnQaESORCs2owOEUwPWH9nACMGEqoBRY7ToQPvbeLOOePHtiK6mX67bU1Wpoa6iTz1mzuBont4+4LTFzOdA+rEDLPRmOxFUqMEcBPgCcozujUoHhCvgOGzKPA0kuKfgBe4s3TeSQm3uuPTSyfr55a8cxFaACVIAKUAGXAjX3cRfeyLlV74ExTe15x54lrP+qeebU2uaY4wtbrUtvTXQsxC9pba1LMMRH61eDqXHARc/11hUxdU69Ytbilu6BkJ6lmFo39qnATAX4I8Az1WdsKrCIArU32EXS6EZDHwS6BSHw4RWQfca9dvhtQAF2psCoa1ru47yXr7N5bN3teB2m45hgf664V6U+eFtFwNvOzxivtI9yuOTYiq61eyQ3XqiWLqyV9kSIO9eogEsBfgPQpQrnqMABFXDdzOxNz2UjUmk79H22q0krPMHZctPz0t9KTjYPjrelAPfatuq1Glvun9UqcvM9stU9BPemVnjrKbYmI31tHU177LmcyqyukeUn45I8UzSxuDZ2CMP6hmxlDfY5MUKYwIvZhOKF1jRuqp32QV94ir/lK3M33ngjzIpai1kEQicqsIgC/AbgIoUgDSqwNQX2cjPcSx5b2z/kG1ZA9iXeYUuuUoGLCtT8A+oiEkdUgApoBVwPF7Cect0d5TNdtEjRA9qt0IIv2p6cRsToyX9VbNFVn+l1f1XO5EUFZijAB4AzVGdMKrC4Ar5DKuZTbqraNsV+NUm2yHk1DcknTQFcKy5re6B12RxxLqTZEfVgzmsq0Po+wn3fr87QFjVDayPi4Q3s9bqsYV3mdV/buXz1+mp9nxaaJ3L35axtR/aFz6VLl64I6eNpc/XZWUDUFK1dl7FgpeK5/Fefs9rN4Gv1DdWjJb+917alVsSarwB/BHh+DciAChxCAbkJ2xvzIRJnklTAo0DqYZnXjUdATlOBAymQ+nlxIEmapFqrq+tsc7TP7L3m2yqv2j3m2uiaWw98V0yZQ9ySmNYHWL5YnKcCVKCPAnwA2EdXolKBXSmgb9L2Bp6TKHw1Xo6/tRW8VlgWm2Mq0EsBXAcxfO5tv0LU5kpt8HmI/UWNrtRo5Azq0Cqm1FNjsr6tlHXjaL1jWuu6CJq2131r5468nVmd23ZYu5mW1ibHT2xzNXPhWww7dmdYPpuD7+JbHpmeVIAK9FCADwB7qEpMKrBxBXJu9r5ULYY+FJQcgnxxaueFp+ZWi0d/KtBCAXv9tMAkxn4VwGcY2v1meuzM+LnQvv6hM0Cp3ilnnFLs9goQ0aVAj/qUYMInZU+58uAcFaACVMAqwAeAVhGOqQAVqFIAhxULIvP4x6nPxvqExsAK2aSstcJJiUUbKsD9xj3QQwF8vrb4bO3B72iYvM63VXFcP2At9au5lny+mN/a/tC80YdWK7XQNcaxdj0351i8FLwWGClxtA30lDkdX89r+5y+xsvxoy0VoAL1CvCPgNRrSAQqcBgFYjf92A1d1mM2qWK2xEqNSTsqMEKBVtfICK6MQQWsArH7hLU/ypjX9dqV1vXR/bVZj2WndZHrfKVrHVw0R5c6rj8IInYxPxeWz68US/CQhy/erHnUe1V+s3RhXCqwRQX4DcAtVo2cqcAiCuCQIwcC9BehFqXBQ0xUIhpQga4KbPFzo6sgleCtPoNtXew4RrMVj1ickev6fhHLT9uCY8wHdmznKuCrk6umlqn4wk5aH5b4hdY0LnB8rbYd0Ud+OhbmUnPSvi360EawajmU+pf6ufJvieXCX2VO120VTuRBBY6iAB8AHqXSzJMKNFZAH1J0v3GYpnA4qJaAbiXHktzoM04B7KPQXhx5MNY8wG2UGqPjjcpry3H0fkAeR6gT8nblijXoIWNrZ21gy3b7CtjaovaYx1gylT7mpdVruUpoHPF1jWvwU/ggJuLo/LQ/1vXcqD44zeQwKtfcOLV7EJpiH+TG36q9zhv9reZC3lTApQB/BNilCueoABUIKrDFG2LNAWaL+QYLyMXpCsiesm9Nqma/ahz2qUCqAthz9vMO86k4W7PT+el+KA9tp/shH65tWwF8XoeyaLUXYjgpXEI8W67Zz4uW2KlYK3BI5TrSrpUurXBG5s5YVIAK+BXgNwD92uxu5aabbjq9/vWvP73xjW88/eM//uPp4x//+OmWt7zl6Q53uMPpXve61+mRj3zkWRtL/C1vecvpVa961ekd73jH6WMf+9jptre97eke97jH6VGPetTpfve7X8z9bP0zn/nM6dWvfvXpDW94w+m66647fepTnzrjce973/v02Mc+9nTXu941CYdGYxSIHUbHsCiPMou/jctDVHkN6UkFWikg12XLa1Ff5y1xW+WbgqNzcOnjmkvB3YON1FTrg5xcc1hjux0FdB1d169rTrLDfOjaCK3lKoR4uX5bt0d99p4/8kS99p6v5Imca3MFDrQrbS0POy7FpR8VWE0BPgBcrSKd+HzgAx84/ciP/Mjpn//5ny9E+N///d/T+973vrP3a1/72rOHb894xjPODzba+HOf+9zpmc985umaa67R06cbbrjh7P2mN73p7CHis5/97NMtbuH/cqk8NHzKU55y9gBRA11//fWnV77ylafXve51p2c961lnWHqdfSrQQwF9g291iOjBk5hUAArIPtX7FvNsqQAVyFdAriV+9ufrFvNY/XOqVc1jn8UtdIjFiNViD+stdNyDDsgB+3fU3uDnJJRnSwW2rwAfAG6/htEM5CGffvj3Dd/wDacnPOEJp7vc5S6nT37yk6e/+7u/O730pS89+xbey1/+8rNv4skDOvt63vOed/7w7253u9vpSU960kn+mpY8uHvRi150ete73nX2rT75RuDTn/506342/uxnP3t62tOedv7w78EPfvDZg74v+7IvO1177bWn3/zN3zx99KMfPXsAKN9MTP1GoTMYJ6sVwAGjGmhBANehCXMt8nZhyBxiLCgJKU1WAHsjtE9c+6qGtsQMxavBXtUXOrfi1xqvFa8cHOwDl09ozWW/9Tl9jaG2tRoAE3hb1yjGH/miXSlvcIrlULJu90lp3hqnFKOEv/bROkkfPNBq25790fF65hLCRp5a9xR72MAPOJhv2Qo24rTE1fsrFzeVj/x79Y/+6I9y4WlPBXapAB8A7rKsF5N685vffP7Nv2/5lm85veIVrzh9wRd8wbnRfe5zn9MDHvCA0w/+4A+e5GHhC1/4wtMTn/jE01VX3bw93vve955e8pKXnPnc/e53P8O41a1udTa+5z3veeb/mMc85vTOd77z9OIXv/j08Ic//HSnO93pPAY68u0+eeAor0c/+tGnn//5n8fSSXC++7u/+/Swhz3s9IlPfOL0nOc85yTcNI9zY3aGKICDhL7BYm4IgUZBcg8tufaNaBKGCpwpsMVrjKXbvgKhfRda237m7gxScxY7fY90o12erfmHbgh3tTVot1q+8j+s7Sunfta39xg69o5j8V37ebVaWs57GdfWfKt1KuHt2qcz9kEJ9xk8GZMKQAH/z2nCgu3mFXjb2952noN8s08//MOCPNS7//3vfzb8r//6r9N73vMeLJ21L3vZy07ye/vkJT8GjId/ZxOf/88XfdEXnc3LWOx+93d/V7pXvPAQUb7x99M//dNXrMtDQ/m2orz+5V/+5fSnf/qnV9hwYrwCciDBe3z09IhyE3Yd8AUB/NHGUFPtfDjwt63PnvNUIEUB7CfYyp6vPQQLJl99FEB9XG2fiGuiIv812aWxstdemteVVrXX65WI686s+NmCOqLV6mGfrlCjFbXTWm217zsjbjUf8B61X0bFQV6t2h51x+dFK47EoQKjFOADwFFKT4wj3+rDS74C7XvpNe0jH3DyLUJ5fe3Xfu3pm7/5m50QMi8/ViwvsbcHKPkWIR4sft/3fd/ZQ0MX0EMf+tDzafm9gnxRgRQF7H5L8YnZ9MCMxey9vsecemu2Gj5ruFpF3HxYp5t1CT1sudlqTi+3TsgF7RzWjNpTga0+5OipSU9suQZzr8Nce/DX/9bB3GptaW6Sx4i92+uzLyfvHNvU+uZgwraXFqmcaUcFShTgA8AS1Tbmg4dyQjv0f0CwJh9md77znc+zfP/733/68Ic/fDb+tm/7tvN5V+fbv/3bz6blD4OIn37hR39lDnZ6Hf0v//IvP4//1re+FdNsqUBQAdyEexzu5EaPm32QxAYWRSe+tq0AariXPbntapSxRw3LvPfhhc/s2dmAB2qCcSovXoepSs23850PUHO0NUyBIe1eXsipZz4hvXAG020pF/m3zgrXrOag88I82pw8Qxrm4KTa9ogXyxtapXLMtYvFF7wUm9y4tKcCIxXgA8CRak+K9f3f//2nW9/61mfR5ff7yR/isC/5Ax5//ud/fjb9kIc85NxeJt797nefzct/5BuAoZdev+666y6Y4tt/MqntLhj93wDrH/zgB8/+OInLhnNUoKcCrhu8a64nB2JTgZACOHyjDdlybawC/KwYq3eraLyWWilJHCqQrkDoumv9WSoPgUPx0lmXW7bOKYeJxG4ZX7RsrWeIX+tYOdrRlgrsRYGb/8rDXjJiHlcoIH+V95d/+ZdPP/VTP3WSb9Q94hGPOD3+8Y8/+5bdpz71qbM5+d188mO/3/RN33T62Z/92QsYH/rQh87HX/EVX3Hed3X0ujy80y+Nc8c73lEvXdH/yq/8yrM5uQmIHx4IXmEYmLjd7W4XWN3+Er6xiUx8/2cb60dpdd11Pzf/m2666QoX0fjGG288/+vX1PwKiZaY0HXX/SXINSYhn6XyF9P5uqiArrvuX7TqN3J9fiCafG6wZlCjb6trr/sto8o1aO/HMfye9bdcfPcpbeezieWxhXVdd90fwR0ab0Vf8NWfX7M/rzQXWzOcx+y8Heu66761azmGlsAUHTGHnIS/vDCGrbS9dJdYPbBdOeh8cvvQBlpIa/XLwdS/2kr3XRixz2eda8zWhZ8z5/q9/Tn+tKUCLgX+3+cfsPx/rgXO7U8B+QbeS1/60tM111xzxf/9uf3tb3+SPxDyqEc96orfzfeiF73o9NznPvdMEPkGofylXt/rLW95yxmOrP/Mz/zM2V8Thq3gy7q83v72t5++8Au/EEtXtBJP4srrNa95zUn+SEnKS3BjH+wpOLShAlSAClABKkAFqAAVoAJUgApQASowU4Fb3vKWp3ve854zKTD2jhTgjwDvqJihVOT/Vrz+9a93/nEO8fv3f//30x/8wR+c/vIv//IKGP1/YeQDKPS6+uqrz5c//elPn/el0wrnAigHVIAKUAEqQAWoABWgAlSAClABKkAFqAAVoAJBBfgjwEF59rEoP+b75Cc/+fS3f/u3J/kq8ZOe9KTTwx72sLOvgMuDwWuvvfb0/Oc//yR/pOOpT33q2Tf3nvCEJ5wnr7+pF/t2nf5a9K1udatzDOlYHD2+YPj5QQjH2vrGj33sY08f/ehHfcu7m5evxsvX+uWFr8nrJLGGOZeNrFk72PdofRxqeMiPd7z85S8/o1uyB0KcrAYjtbKxOfYrULsH/Mh9VrDnVtpPwmklPrnKr7QHVqxvrp6r2ENLy8e3V3vvAx8fy8+OfXytXcq4FQeN05JfSg6tbSQX5NBiD0AbYIb4wtZlk+Lv8quZC/HJwd0yd/m3yz/8wz+cpfuN3/iNJ/kyQ2k+0NPnj/UcbbWtD1fbxPqaQwu8WDxZ1zFT7FNsXNxL4+g9IP82+Pu//3snhdyYLnsncOHk7/3e753kp/T4ogItFeADwJZqLor1G7/xG2cP/4Tec57znNNDH/rQc6byjb373Oc+p+/4ju84+3Hdv/qrvzr7fYHf9V3fdZKbpLy+5Eu+5NxeHiaGXv/zP/9zvvzFX/zF533paJxPfvKTFx4IXjD8/CCEY219Y3n4h79e7LPZ07x+oKq/iYkcrRYuG7G1dvAvbe1vGcAv8LXzFt/HQ/yAYX3suGQP+HSx2KkcrB/HYxUo2QNjGV78nyOjY9t4+rr0XYPWZ/Xx7D2gP5tX12p1fr7P5xtuuCF6X+ixD3x8Yjq2vLZacZB9iuu/JT9oAeze907EceVQugdyruFQPVycoE+vNsQnJ+aWues85eGfaFKaD/T0+WNdx8zp+3BTMLD3tW3KZ6O2L+3X5m3j+j4nWsSRzwHgSBytm0sv2FqOMq6plwvPzrn+cKe14ZgK5CrAHwHOVWxj9vKh9trXvvaM9Z3vfOcLD/90KlddddXpx3/8x8+mPve5z537yIT+wx76D3lof/T1Ov6QB9Y0jnzAhl74AyLywaz9Qj5cu1kB0U3fPHUfVq45rLVoZe/pmyowffNYRws728q6aw5+LVutI/poW8YhFhWYrQCuKc1D5uzLZWdt9jRGvmj3lJvOxVVrvb6F/ugalWrW8t5byiFWzx64yLsFtmD4cFa9RyP/mPat11vF9endmq/Ga8FdMPS3tKTfAlfz1H3B1m+9FurDJ2RTujajdpYr8pM29IKdzya2Dr9QHP0tQtHGZytrIe18fuDAlgqsqgAfAK5amUa85Hf7ffzjHz9Du9vd7hZE1X9o47rrrju3/fqv//rzvp4/n1QdvW7/cu/Xfd3XnVtqu/NJ1cG6PES03yRUZuxGFIjdnGLrEfglllvnIHh4I8HWMYCb2+IwEjqQ5GLS/tgKYE+lqoBr4ah7MFevVF1n26Gus3nE4qfw1HtT/0Mvhj1iPYX/CB6+GJqf1tFnnzq/1+smNX+xa6lnTtyt2+o9aXMJrYltbN3i9RivwKFHXimYkjve2r6FJsB1YbnmdPxYn59XMYW4vnUF+ABw6xWM8Nd/Pjz2NWL9+/3kG4F4fc3XfM0Jf+b8b/7mbzDtbLF+xzve8SR++vWt3/qt58O//uu/Pu/bzkc+8pHT+973vrPpe93rXnaZ40wFYjdCfRON2WaG7m6+Nb4pguDg4WpT/GlDBXoqcKR/xIZydV2fmOup/9GxQzXR2qxWC9xnNcfafqoWvjg+f3BteX8Fpo9L7nwIz5dXboxSe3BrqV8pl1l+uP7Q1vKwWkJjwUVft7XxRvm35ix4rpdvvve1EsK3nKCFi39sDr5oYW/HmLdtiKe15ZgK7EEBPgDcQxUDOXzZl33Z6da3vvWZxdve9rbTZz7zGa81Ht6JgX54Jx+gD3zgA8/85Jt5vl+cKvP45p7Yi59+3eUudznhW4BvfOMbL/yeP233ute97nz4oAc96LzPDhXQCtj9pde22k89hOwx963W7Ci8ZW+m7s+9aFJ6nR1NpxXrXVq7nFxCMWRNv3NwU2xb7bFWOCmcxQaapNqX2IXqUoJX4zMi3xi/0TV2xXPNxXjXrPfSvTQP354UvFJMnz42lh1bv9bxLX5oDG5oQ7ZcowJUoJ0CfADYTsslkW5xi1uc7n//+59xk19U+lu/9VtOnv/5n/95+pVf+ZXzNfhg4vGPf/zZXxCW8S/8wi+cPv3pT2PprJWxzMtLvj0o9q7XE5/4xLNp+bHk5z73uVeY/Ou//uvpt3/7t8/m73SnO52+53u+5wobTmxDAd7Q29dJNKWu7XU9MqJvP9m95rLr8Y+XvdRi5j+q9qKhKw/XPnTZ7V3/VB1c2sTmRmiHzw4dC3Mxfltc71kvq4fWUetr7UrGwEZrMULxQmsWJzRupSVyQOuLifVa/sIbbxurFtviIU5IKx1T9y1WbCy+Pn/E99n4/GIxZ69D39k8GJ8KlCpw8895liLQb3kFfuzHfuz05je/+ewbd/IXgd/5znee/TEQ+UW4N9544+naa689vexlLzv927/921ku8heA73vf+17IS76998M//MOn3/md3znz/6Ef+qHTk5/85LNfrCu/Y+eFL3zh6V3veteZj9jd+fN/cMT1kr9A/JrXvOb01re+9fSKV7ziJL+j8JGPfOTpS7/0S09vf/vbTy94wQtOn/jEJ07y4PIZz3jG2cNEFw7nqMAeFdCHJZsf1uw8x1RglAKyB/WBnXtylPKMoxXAvtN7Ua+zn6eA1lG0hb55KO2shc9sDu2yuRkpllPrvHVdb2bRtjciRlvGfjSX/iPyk30hcVzx/WzXWGmljy9333xO9rkcUY+cGLSlAltTgA8At1axAr7yY7fyYO3pT3/66T/+4z9Of/Znf3b2dkF953d+5+nXf/3XXUunn/zJnzzJn06XB3jysE/G9vWIRzzi9BM/8RN2+nwsv5Pw+c9//ukpT3nK6R3veMfpT/7kT87e5waf78ifW3/Ws551ut/97qen2d+YArjp4tCLsSsN2GAtZAubPbdWjz3nytzWVECuQdc+dM3lZKCv7VqsnLiltsJRc87B8WmYg0FbvwJ6/5TWyI8eXwnFdK1pvnH0uEXN3gyha+7grOfgizWMS1qNofPR8bRNSYzWPpobsC1H2Nh52NsW9tKm+lgMO47hIKb12/tYvrQg/85wvUZp4qtNy/q78vPN+fj47FPmXZg5+sJ2hibgDg7IF/MYs6UCW1WADwC3WrlM3ve+971Pf/zHf3y65pprTn/xF39xeve733367//+77Mf67397W9/usc97nF6yEMecva7/nwfcPKtvF/8xV88fe/3fu/pla985dkDPHmgeJvb3ObM/wd+4AeSHtrd9ra3Pf3+7//+6VWvetXpD//wD0/vec97zr6dKH9oRL59+LjHPe5017veNTNDmq+ugG9fad72ZqvXdD8FS9uzP18BXVvWb349UhigZjX1AkZKvFVtdP65+Yi99l81xy3zyq3JrFzBs9V+AF7LfCymHbeM5dJB5nrGrOWfy03sXXnm4sR4WzxXTIsBG+tr7VYZg+8qfKBba17AlTx7YkPH1jGAq1vE0LnpdenH1oBh/UrG8hNw8mux5KXj2hgy1uvoW7sSDvShAjMV4APAmeoPji0P6uTHduVd85Jv5tV+O09+T+CjH/3os3cNF/quqQBukrns7M1W/HGjBSbGudi0X0cBqSXreLEee9XEdU1L5keqf8vatsS6uAM5ylFA6uB76T2v+7BnDfOu/y3opTmG9gb2AFsqkKKA3lcp9jEb1+dRzCe07sOTed8La77rxIVpfTD2xWg5r/nofssYxKICoxXgHwEZrTjjUYGdK+C7qaemLTdY/YYf5jBmux0FXHvCNbedjNZhKjriHWMFO6s9ri1pW79sLMHvEac1bx8etPKt++ZdOvhsQ/Nb1i6UV+laK11L42s/vTdc/T3UDnn1zAUxtLZS59m1DsX36dFqXmvh6kMztC6bmXMr8ZJvf7V6+erbCr81jt3DLfijtrr18dbxXPZYRwsc3/Xvm4eftIjjq7uNpX3RBwbwMM+WCmxVAT4A3GrlyJsKUAEqsGEFUg5dG05vKequQ7L9h8BShEnGqQBr5pSFk4MUmPWZreO6PssGpe8No/mJkR17HQ+yQD3WKbTUQr/XYXaZSegeF+Id8muZ46g4LTkTiwq4FOADQJcqnKMCVCBJARzGdZvkSKNpCsw4wOiDG/rTBFg0sOhS8tJ64jrUOKF6u9Y0HnBcdljztS4usC3Bg69uQzHkF72v9gLfWOvjXbpHfHiuec3Ntb7SnHANvUQv37c+Qn4pa/Y6Sa0N/GLcUzmk2G3dxqWVa26lPFHnECeXzep5hfI52lrqNZ+iSwhL9gT2BdoUzC3ZhPJHHrHcY+vAsa32031tJ/NYQ6vX2acCW1OADwC3VjHypQJUgApUKJBy0KqAp+sEBXAgldrm1he+Idq5mCGsEWspOZXyEOye+KW8Sv32lAs0KLkO4Jvblsba2jWVq8sI+9l7t7T2I7RhjP0pgP2Odm8Z2s9EXF+6bZ2zS0vXXOu4xKMCsxXgHwGZXQHGpwIbVGD2DVLi28PCBmUk5f9TQO8n1jV/W4hmoiF01BpiLYSq/aQvPvDTWCEMvQY8PdejD46CDd49Y1tsHb9Hfj0xbS52XFL3nnxj2FvjG8snZd3WLMWnxkbHi+kN25hdiI/4Aidkt+U1V34yV6Nbrh6lOmuOrjzAo0U+wNcxgd+zHRkPsZBrz7xWxYYGwq9k3+T6SDyX3q65VTUjLypQogC/AViiGn2oABWYpgBuzNLiPY0MA1crgHoCyI4xzzZNAX2AFo8aPS1WGoOxVuBo254sJFaveMDtyd9iu/aIzOm39Zk1BieJjzrM0GxW/ojrqhnWXO1WNUKNXTnNnOutZ259R2uh81+dq0sbzd+1zrk6BWr3RIq/z8Y378uIe8GnDOf3rAAfAO65usyNCjRWQG6suTfXlhR8sVN4+Xxb8tNY4DQ6ruawhT4PX+tUqdVejdU0tl6riOD3+N1vLt6tNKvNOddfcnHlIziYRytz+DyTfutXjoYh3q15rYaHGuToJTnoOrbIKYaXyy+FE+pu2xTfVjbIG20rXOCEctO176EvOMRacIzZ2fVazqVxLQ89BqZusY45jHu3Pn3AQ1q8dB9zK7XIRVr0c/hpH90HRilu6PcCr64pcmdLBVopwAeArZQkDhXYsQI5N1wcWFaSw3WIGMkvR7+RvFaJxcNXm0pQxzY6hlB6azz7swr5oYUWMrZzWKttY7j4/JytTW2eK/qn1hU1QrtiLiM4Sf7Yjy3iQf+Qrqvsex9H37zVp6VuFrvFWOs8kquOG8oDeyVks+Jaan7CPcc2lKutXyvcUEyuUYEtKcDfAbilapErFVhYAXsItOMW1AWz5kYuvj14uXJDnBq+Lty9zkGvvebXOy+ffjIf2oPaD9cH2lLOoXilmPTrowDqj5qh9hj3iRpHnR3fMoQudn6rY9Qd/DEO6Q4b+MTaXPsY3hHXffVwzffSOwVXbFycXDUTuxRMl+/IOZ1TT74h7FKtUvx0vUIccjTXmNoP87E4WE+11zFcfeC41jhHBY6sAB8AHrn6zJ0KJCpgb8raDWt6rmdf4rlu6j4e2tZn05tvD3zJa0Y+PXLZKubsvVWrm90/GKOtxff5b3HvotZWGxljzZdv6rzFTvVrYafzsPnM4KX5oD+Dh9WihdapGKWxQzqF1lJ5xexGxIhx6LVeWpMSPjmxSjXPiVGSA3zAD/GkxRxs0MIGY58d1nNb+VHQq6+++kJ8G1Mw9Zzut+YT44/YsbjaLmQLOx1X5kI+2lb6LnsXLvxysMUn1x5xdAsMFy/sAW2f2gduqj3tqMCKCvBHgFesCjlRASoQVEBuwHiLYeiGHFoLBll8ca95LS77OT3XofJ8cfGOvnZ6UR0Roxd3F27oegutubA4tx0FZtW2x+fLiFyEdw/uK+yYveXVMp8QVu2+C2G32he1HFvxCOUaWkP8UB7in4IBLF8LDLRip/vWL8TJ2rYaz4jZijtxqMAIBfgNwBEqMwYV2KkCK9xkUziITeiAstPyMK1OCmxtL6VcIy2lGh2vJfcQ1qy8sN96xwc+4kELGWMNcyNaiSmxwWcWD8l1VP7IdYS+OTFmap/Ds5ftyLqUxHLVBziuvYu1FL18tsD1rQs2bBBHxmJv52U9hKPXXL7Ar2ktLnii9WHXcLO+4KDndVwXF9jCV9u37iMWcO0Y87p1ccZ6aA02tkWeKbHFF/Y33nijhcoeAyvbkQ5UYDEF+A3AxQpCOlRgVQVSb7ar8ueNe9XKXOS1+j6z/GRfcW9drGFsRL1iCs1dX6k+K3BZgUPqjujJ1X72pXKiXboCIY1zayv2Lp9QjHSmlx/YhbBcsQU/dR78ffY5XEtsJW4oPxdmrj0wZuWI+LFW8irNzYWt8UpxV9fMlTfnqMAqCvAbgKtUgjyowMIKlN6gF06J1BZVYOVDnb0OVua6aHnPaImOe9NO8rH7I6UGKTqk2KTEKrGZGbtEz5IcV/GpyXdUnVI4juLSs24peYpNbq4u3BBGaE3nH7OzXMXexUVjlvZjXEK4Pl/N1+YSwmux5uMk2D00dGH6OPTWwsWlRNNWODY2dInhy/pNN910wV3vqQsLgYHgIGbAjEtUYHkF+A3A5UtEglRgrgKxG+tcdmOiU4MxOq8eRR/8dH913pbfyP28ZZ2sbq6xaIm3a32rc1K32bWbHX+l2oW0CK3l5iBYNXg1vrlce9jnXsuuz1IXhmsO/LFmsVpoCWyJBXzgSos+uGyhHck5Fsuu23FvPW081Lgkbo1vSbyWPqKD1qIkF+3fkhuxqMCKCvAB4IpVIScqQAWWUQAHCbTLECORKQrgkLjqfkjhhRxGCeiKl8JzFL/cOMId71xf2IsmLl2wzvayAtCot16+evrmU+oDX7QhH7HRL+SLVtaghbXT41Z9HTcV08Uv1XekHerhakt46NrZPmKU4IqPxivBsDVx4aHW1rZFvBIMnw94+tZT5y9duuS8llL9rR145eqna4F9oucQJwUX/uKj+8BIbeGLNtWv1k7itXi5ahHC1vYhuxbciEEFVlKAPwK8UjXIhQosqIA+fOAGqecWpDyFErWZIvuwoKjvsIAFgcAR7erX6VZ4FpSCLhkKyD4I7dXQWkaYbFPsT3GMcUwBB15NPvBtwSeFs8QDb2sPLnae43YKtNY4hheqdyyrGHbM/yjrvuvpKPnH8mzx2Vaicc3ej+XEdSqwmgJ8ALhaRciHClCBzSjgOmRgDocJtEiKh2QosV6ra6fZYV7PrdZ3cZS5Lew3cN8SV1/9kQNyEjvXnM9/5jw4g28tF+BpHIvtstH2o/shPlizOfg4ajv4umztmvZz2cuc2Ihfiq0PI3V+RIxULrV2119/fS2E09/W0GmUMWk19+FbO1eIFBvtJ/a+eNoO/Vx8+LFto4CvVjKfW8s2jMpRkAtajVS6z+Bn/wowYmAdYx2TfSqwVwX4AHCvlWVeVIAKNFcg54AAW7QgI2McODDHdr4Ctk7zGeUxkD3lyoH7LU/HVtaua9zOrVqbUbywX60urWqwEo7v+hSOyD9Hd2g3I8ccnqX8dH7QpxRra36xfGPro/JdhUdpviP2sYub6IbYep/7bF3zqXMx/FSc3nZb30u99SE+FWitAH8HYGtFiUcFdqqAPkhIX493mnLTtOSAg3dT4EKwPdSvdQ4r1aewrGduyANtDRZ9LyogmrZ6tcRqwak1Hx+evX/Y69iOW+QWwxCuPr7Wt5RfLEZJ/FIuNqfQWGIgjm1dfrBxraXMWX8Z27kUnK3ZuPaHK/cRWri4iJ6YT92rK9ZAvgUKDdGO5gn9oKcvfiq/GI4Pf5V5yTOUK/RK4atxtJ/87kfXKxZb+2g8Pc8+FdiaAnwAuLWKkS8VmKCAvqFOCD81ZIsbfguMHiJsva4tdc3FyrXvUT+LKZxW5CU8U3htZT/6cvHN2zq1GPfSKiUHiY13LBfB05g+3sDzrcfitFq3fF24Oh/X+p7mQvXwrUEfXVPbz9EopSY5eFux9ek7g3/vGrTONQcvx7aX9rhmgA+97TzWdattdF/buPqIgbUcX/j0bi3H3vGITwWOogAfAB6l0syTCjRU4Ag35RUOhQ1L5oRa8cDnJMrJZAW2vm+3zj+5UA0Me2kV+lywMe24QVpnEL1wW/E7Gg72BFrkH6qT2Fp7+LG9UoGQlto6VdNUPI09s9+Sr9Uo9rsfW8au1dByz8EL5SG4LmzffE7cVWxD+a/CkTyowAoK8AHgClUgByqwMQXkJrv3G60+KNUckDTOxsq8JN0Ze8/W0I5XEsrFbfa1OqNmvWsiOtu3K2av3BHbFbPnnGt/xeKV7L+SODEeOevQtweP2j2hOel+Tn4jbXM46r0ifjm+I3PqFQt7Q+tQGstqp7F1P4ZvcWL2JeuodYu8XfF9P/7psh05F6qDXkvRJWSjawitbZ4hf2s7Yqw5l8ZbLafSPOhHBVorwAeArRUlHhXYoQK+GzFvrv5i+w5Zfo8xK1Iz1A3tmMjbjrJqPbWqtp5brbXNQ+fYuj8q1qg4rfVx4elrwXdv0H4pNtpe+ivr1YpbKxyr3YpjvQd0fzTXVR8EpeowQ7uR+3RGfqnaj7ITvVtqfnRNW2o5ag8wDhXorQAfAPZWmPhUYOcKHOnmKgcp/ZbS6rHur1j21Wsl/FbnKDVe8QXdoCHGmivW9FzvvotHSsxSvxRsa9Mzlt4vPePYnEaMdW4t4+FztCVmayypJd652C10G6GR3q/IVc8hb9cc1nSLvH32dl7G+q2xavrgUYIB3XVbgrO6D3RflSf42TaHr65hzZ7IiVlqG+MnOrheMT/tY23tWNuO6KO2ObGEs37DV7BiPwYOW7ZU4AgK8AHgEarMHKlAZwV8h4/OYafDzz4gTRegIQG9h3TfhsDhzs4feRzSa5QuwsH1rom/Ql41/F2+e8zJlaed2/NnJfa9zTk03tvnWMq+TrEJaVbrr7FL9qPPZ2+1FJ18uWoNt9qXb4HuOT/UryRH7OUS3177oeTztRcX4lKBvShw1V4SYR5UgApc/rGpHjfulgdv1mkdBXrslZLs7P5ahVdJLr19rFa946Xg9+Qk2FvfD8Jfa6T7ou/W80vZI8jT5p7quwU7ndseamr3bawGkr8vb62NCye27vKZNae55mqk9fHppfFb5OjDs1wklp7TuQFDr7fgZjF8mlg717g3N1fM2jmtcQgrVZdWGqDeIU46Vop9CCtlDTF03JAf7EM2XKMCR1WA3wA8auWZ924V6HHTS7nhStwesXdbqAUSW6VeKftrAblIwSgw6ppfZZ+a9LOG3ONpv9dvNZ1K996oayNnE4JTaU4psXpip8TPsWmx13z5CnYMP7buysUXz2Urc6n2KXx9MThfp0DqPgjVMhUjhWkoDvxbxgNmaiv88E71oR0VoAIXFeA3AC/qwREV2IwCoZu0rJXcoDWm9bdjK5T2tWscz1dg9frI/lqd4/wqXv6GRq5O+toVX+2v13Ly0xg5fke29e1x0bK0DkfWs2furfb3CrV15dKTlyteq1rFeMfWLQ/fNWntZJyal76Wdd+FaedSYuTmaGPkjK0+vWPn6pWTyx5sQ/rXapey93wa2n1i7cCtJobFlHFMj1bxwN/FgXNUYIsK8BuAW6waOVMBKkAFMhRodQjKCLl505U1Sz2Mil3MtiRP64M4sVhqLrwiAABAAElEQVQrbgpwH8XNp5FoanUdxWlUHF/uo+LPirNqXUO8Ru7HEA9bsx57qAem5T1rbHOTMd4pnHJqk4JHmysVWEHjURzsfrxSjfyZEHfs9R5x85nSgwqsowC/AbhOLciECjRToMXNTm6qOTg5ts0SJVBQAXswWr1Gq/CzugVFnrQoWlmeKfq5/CyOpOTCgh3WMNYSuPD1+tH7Ls2Orgnyx77CeE+t1H3F/PR+HMlPx21Z59IcxK+WU2ls5J8TX2xT4mnMFHtw2Vqbk5v8Ndirr776PMUc33OnRh1dn0aQWTAt44uOKXguvVP8shJTxsC+6aab1Cy7VODYCvAB4LHrz+w3qABuZi7qrhury843V+vvw+X8GAV67o0xGcyPIhpu5TroyTO0l+yaHc+v4rYY9KzjFpTw5a/3lc+mR346bkt8wR2ZB7gjpi8vrMN+tTaHn84xx2+1nFP4zNpPKdxqbFDDUP1Ca6mxW2CkxoIdcsM4p23Nt4ZLDu+ZtkfIcaa+jL1dBfgAcLu1I3MqQAWowAUF5IDIA88FSYoGomHrw3YRETo1U8BeF6xvM2mLgViDYumKHPeut73Gi0QqdKrVtgd3jVnKT/w0jsgj41K8HHlHxcnhVGNrdRQs6Ohaq4kV8+0ZLzWnnhxi+XOdChxdAT4APPoOYP6bUoA3zE2Vi2Q3qAAOryOoy/U8Mh5yQkwdf8+fLfIjX/aFfKGFXe81RrxZ8XvllYqL/FPte9npvd8rhsVdJXfLyzUGV+xTl83e5iTn3Hyh0ywtwBc8MG7Jx6WLxEHMlrE0/x74mmuvHHQM9HVemNP5QWO0LhvM1bYuLrmYKdrp/DR+i/gaz9eviWPr4IvBeSqwVQX4R0C2WjnyPpwCrpsZbrDSon84YZjwBQVc+8A1d8GJg80qIJ8Lrs+GlIT0vpC+Hqf4b91mdr6z469eP+zJXjq5rhvXXAudeuXQglsIYzbvnPjaVvdD+dWsjYiRyi92H6jd15Krzlf3UzmuZrdaDpaPHc/SryWP2n2YqkFtnFr/VJ60owKzFOA3AGcpz7hUIEMB3IxcN2LXXAY0TTMVCNUiE6q5Obg1ByZgFwVqrt1etbacesWJCWp5xOxz13vjx/jMjh/jt+d1vad1v2XOe6qv5NJaJ+gTwoVNTl1KfDQ+/Fvz0jFsPxTL2taMbRzkmoNZ4pODL7YpMWwuiJHie+nSpdMNN9wAl+6ti2uIZ2itlqyLiw8zl0cKNjC1rWsOnLCGsbTaF/OuOawJRmgddrZ1xbY2HFOBrSrAB4BbrRx5H0YB3LhWuRkJn1W4jN4EqIXE3YoOR63V6L3BePUKcK/Wa0gEKrAlBeSa1/fVVbhbXq0/m1bMeQXtW+vsy2lUHImv99LIuDb3nD2nOVucHmMXt1ytWnMWTrkcemhDTCrQQwE+AOyhKjGpQCMFXDfFRtDFMD1uiMizFHvWjXpWXFfxoKFrjXN+BaxupXvQH6H9iuUsEVrvRVeM9pncjDhSd8Ry5Yi1m5nts4fcj5JvbRVTdNqzppI/8qvVEv6t8YDbqk2peW6s1XPOzcdnL3m21k9jArtUT43ly6HlPPi2xOyJlaJrio2Po+gBf7TaNqQX1sQPfe2LPtZc+LBhSwWOqgB/B+BRK8+8qcAiCuibs+7n0qvxzY01wj4nnxzbEdwZY9sKjN5POKi3VM31hz8sfo+4Nsaq4yPn3qsmoil1TVM35TMmxSYtGq1GKzDqOsA1hzY1z1H8UvnQLl+B1Bqm2uUzoAcV2K4CfAC43dqR+WIKyGG19YEVhxrewMLFhk6t9Q9HvbxaG1P7Sx/vlNgxG+wbHSPmw/X1FUBdse/RtmIOPN22wrY4EqP1K7TfbTw7bs1lZbyj5i55r557aA/P2FM9+KTUoEfcGfqtEhN7P0X7VTjLHui5D3rju3REzJ552bijao6cbKv5YE3Ptej79rfMy+9+5IsKUIHLCvABIHcCFWisQK8bWy/cxulnwblycs3lgKZ88ycHb5Rtad4uPxyChLtrfVROjNNPgVGHeclg5T2kuel+P+WJTAX6KrDaPp7NpyR+iU9uVSVGKE5oLTdWC/uR94wWfEdhjNZltX3RQ2fkqFv0e8QjJhWgAnkK8AFgnl60pgLDFZCb5pFunC0PYy0fBoZqEFoLbRj4oZXc9TvkK2vwi9nl2qbgrWgjeuRoIjlA7xXz2Tun2msdtdZ1x5xLO1887AHfegjTFYdzYxVA/dG2ju7bF63jAA/xkM/M/dcrNnJEi9x9LbRwrWNNt2KHscun5Rzi6NaFL+srv1bkpzml7pUSjXtix/joHGO2teu+PDEvLfq1sehPBajAugrwAeC6tSGzjSmAmybaVvRxQ26N24pfDY7NyY5rsOErh6sRByzEQYv4OW2ub65eI3TIybelrc6t5MGvaJmrZ0v+W8GCTrot4d5Ca3CIxZcf/SmNh32FaxPjWEyuz1FA6pxao9I9MSIzyy01pxHcbAzhavlam9AY/rZ1+ejrEH2X3Yi5UM6oFzhiPIJXTYyVeK7EpUZT6xvaN9a2xxjXGbDBB63Mw8a21gfjWAsca4d5tLIu/ZYvXIMl58KWPIhFBVZSgA8AV6oGuWxegdY3rs0LkpAAbvzULkEsh4nWDVo6zDhFBYoVwB5Da4F889au1xj7Xrf8fT+91F4PV++/PT00wH4WxXWOIyvQI24PTJ8mo2O1jgc8adH35Rqaj/nq68bXD+FzLU8BXc9YbfKQ0601hxQv2Re5PoI7Kz+JrfeyjPmiAlTgsgJ8AMidQAWowG4UkIOG7x/+chAoPQyU+pUKC666DWGFDmUzD18hzlzblgLcR9uq19HY4jMQbUr+qZ/rqXYpMUtsZl97reP79NT3O59Njn6tebtiY7/1iAVMtDWaWG3t2JWbzCG2b33EvHBYgUePXCWvmrr24BTC1PsmZIe1UN1ca645YOW04JnjQ1sqcCQFrjpSssyVClABKlCiQK9DGg47cljpFaMkX/pQAZ8C+McK9q7YYc7nY+e1r11beax545pdmS+5uRXI3a9uFM5SgbEKHO2MoD9vxyo9NtrKeeZ8Vtr9OTqvHK6xCo/mHuPDdSrQWgE+AGytKPGoABVYToEWN3N7uAklmRsP9qEYfOAQUvzyNxVaHgDD0Y63arW1Y6sI9rSdj/lZ+9ljXx6zeTF+mQJb23/IErxn7cfQvQkcc1rJx+bSOkYOn5a2yAs1a4ntwkI8rNXG1XguLL2OmGy3qYDrOpRMXHWX+VjtZd3nK/4tXhp/RLwWnIlBBVZTgD8CvFpFyIcKUIGmCsQOLE2DfR5sdLzW/IlHBXooIId2HNzR9ogzCpPX+SilGUdfL/o6Gq2M7Hm77+24llNrvFo+Nf6zcqmJW+NboxV95ymQ+vkS2xszP5vmqcfIVGCbCvAB4DbrRtZUgAokKOD7fYAJrsUm+jCVC1Ljmxurp/2sPOSAKu8Zde+p52zsFvW0/3hogTlbl73G33pthD/eukZb+yuQrhx0PrP6+JzV13TJnvH5aHwdIyffUj8bw8fR2rUe1/LXGvq4aRv0fbac36cCep+59jrW0eaoIHguTMEowXPF9uG7bDlHBajAzQrwAeDNWrBHBagAFQgqIIcWHFzQxzjoOHlxNNctaDK5JJsJX3LAZv03U94o0ZL6R0E7GmyNr5YC3KVFX6/rfmxd2/bqg2dvLrmfJ7n2vfQRXOGSywf2vXUtzXtVXqX5HN0P+010wH7Vc7X6cL/UKkh/KtBeAT4AbK8pEakAFchUQA4IWzok2MORHWemn2y+JY2Sk6JhtgKj9kFJnNC1IGv6nZ04HborgJpLq/vdAzcOgH2WAwsfae0LWtj5VmPgo22F2wNHOLbg6dLZxzfFFvXzYZTMp8RNwc3hpvVtpXNO/JR8wNG2Kb5bttH55uZR45sbq8Te7nWM0bowU3IK+bsw7VxKDOvDMRWgAmEF+EdAwvpwlQpQASoQVUAOKPol49pDj8Zjnwr4FMDe437zKcT5HAWwj7CvcnxXsUUOuXy29rm9Nb66HqU10hjSb4VjcVcat7wWc/RqGXclPUu51OhhfWWcU4tSzj39bE49Yo2I0YM3ManA6grwAeDqFSI/KtBRAdxcZx5EwKFjmt2hS/XLzb00TncBGGCYAtgzaBFYxi32h8UFPts1FdD1Sql/q32yphr1D4S0njbHXtqFYloOWx+n7FFfjuLbqwY2Zq84yH/1mq/Oz9ar99ilR8oecfmBa4o/bFdscT36uGGv+9Zj8y7tXHMxHK5TASpwpQJ8AHilJpyhArtSwHfI0DdSn82uhFgsGa1/CrXaw1RKDNq0UwD1bV03i+eL0/uadvGwc+3U7I/k0zE3cilOTr0QQ3NL1V7sYrEsfiq25rOVPvRI5au1WUGXWC1T8wrZIedcrTRmK61a4Whurn7vODVauvjWzKG+NRj0nasAath73+osEVPPSb8Vh1nXSCv+VheOqcBKCvB3AK5UDXKhAoMU8N24B4U/fBjqv+8toOur+6Oz7hXbdUB2zY3Ot0W8Vpq1wmmRUw6G5b2VulreOTmX2s6I6eIqPEZwKY2xlT3k0rbnXCtdSusiudX49tRmC9it6tcyV1tPO24Zi1hUgApsVwE+ANxu7cicChQpsNqBQA5RePsSWo2zj6drXrjbt8uOc/tVQNe/dZatrx3fP2oQx7feMq+eerl4Sjz9smO95upre62Pnnf56Tntp+dd/Rxbnz+46RZ9l8/KczW8a3xbaVJbz1Y8euDMzE1qu0J9Q7r69JF5+w7hlKzFtNHxU/G3oHluLjGdfHghP2jr882dFzzXyzfvsm0xNzpeC87AaF0T4LKlAisqwB8BXrEq5EQFGiqgb8i+A4m2aRi6CZSPcxPwjiBb5d1REkJ/XgHZFytfb64ibY2vK4fcOVy/rtz1Gvq5+C3tXRxj+OID7mhjPinrGquEV0oMbYN4I2LpuOhL/FmxwaFnC317xhiNjZxWqRt4xHiJHWxqNbM44FCLuxd/q09OXjW+OXGs7cga2hxHxrZ5c0wFqEC+AvwGYL5m9KACm1TA3rAlCblpr3zj1px1f5MFyCCNuug2w52mkxWIXVOyl3vvZ3BAmyOJ9tH9HIwa29ExQ/FCdcKaz983X6MNfIGNFvM5bYov9iraED70gI0dY75lKzmk5CExL1261DL0rrFS6r1FAa6//non7dS9mmrnDBKZTNnLPptcXvqa0X0fxRQb8fXx8+H65nvuP98eABeJnZovfKSNcYY2Jdg6zsy+K0fJxzXfgueWtWqRPzGoQC8F+ACwl7LEpQKLK7D6jVUOFPblmrM2HFOBGQrgAIw9mnJ9wUfzxRxw9FpuP4VDLuYoe+G+df69tarVp2SP+XxC87KGd29NjopfuxeOqpvkjb2LPYrWpcnKOiMPF2/fXMt8UrByOKbg+fKKzYd4IC7aFKwQnvinYsVi9VoX/qEcQuvaT/d7ce2Fu2XuvTQh7n4V4APA/daWmVEBpwJyEHEdRnCDR+t0HjQZuhGvwK+1DKgJ2tb4xBurQO4ehb3d95gfy/6Y0XDtudpcRQRjxGv0/oA2Nje7b+26Ho/mrGNLf1RtbNzYuJSXryaxeDnriKHbmD9sY3Y912McYnsxtt6TewxbcrMv15y1kbHOS/ddtrVzwEeL+BZX1vUrlgvwrJ/GQF9sYt/6g61tsYdcrbXd6lj0gd7Q1eaCdTvvGqfUxOWXOqdrkepDOypABS4qwAeAF/XgiArsWoGUm3iKza5FapBc7wNQA4qEaKiArXfra8jiN6TuhcIh22uwsQXRsJeOwEW7sjTQwXJFvW2LXLCn4Y/50tbGL8XZi9+KemAvuDTGfrBrIR9ru5Vxr9rUXku+GszSNScfl22qztbOjnX+oTVth77Y5/rAd9VW54S+znG1fZSj45a55+RJWyrQWgH+EZDWihKPCiysQOimr2+kuj8jHRs/xHsGv1hM4Q/ONpeYb6/11fj0ynM2LuqNFrrX8hIcYOZglfrlxFjdFt/+KNHPlRtwUFsZj9YZsYVfamzt48rLNYdcsabHJXjAmdW25Ky1qMknl1OruCHOqTFS7UKxfGvQpWcMX2w735OLYJfmKH4uf/C1efjGsC/loXGBpedCfW2PfoiHXoO9Cz+0JvZYh4YaQ9Z0HL221T7yBX87xry0rvxdOmkf9Hvp5sNN5QV+rtaH7bLlHBXYgwL8BuAeqsgcqEAHBeQAgAMC2g5hkiBxc0ab5DTZSLjO5osa6vphbrI8uwmvtXUlNXsPzI7v0mT0nPzxhx46CCZw0Y7KrSReiU8oH8HTb7FtHSMUn2tzFRhV6yPcs2L3kVClR9Shhp+Pe4h3ajzBkLe1t2MfB5n32frmQ1h7X0PN0O49X+ZHBfaqAB8A7rWyzIsKVCigDz7Sdx2wKuCLXFsfOCQvnWcRqQWdkNcec1tQ7iRKrfduUlAa7V4B2Vd4p17vob2YiuETNoRtfXJsrW/pGN8ELfWHHzTHuKYt0bzEp4bjKr6SN96tOM3Yh624AwealO4L7OeQFqE18MhpgRfiXJsX+Eis0r8ADg7AkhZzaPXain1onctN8gu9LK6M7VzIn2tUgArMU4APAOdpz8hU4FyBEQeJ1Buz66bvmjsnv7GO1XpvuW2sHJuni0Mv2s0n1CkBXHe27RTucLCpn+8+YfA5iPr47FLma7mkxNiajU9Xl1b4LLHt1nIu5evTqhQv5md1xtj6yXzLl42DazAnhvWx4xCWjR+yTVkDXo5OqbbYE6H8XDap+LH8XNgxnxI9Ypil67U62Pwx1nxcc3q9tA8dS/1dfqUPhF1YnKMCW1SADwC3WDVypgKFCuBGGjoM+NZ884VUprjJAWWPr14Hrz1qtZWc7PVmxz3y6LWPQtddaK1HjsTspwBqibZfpO0gixY5eviu8xyMGnVy+dbEqvFtoYdPa/CKrcOuZZuTl8sWnNG25NYKU+O4cmjJWbAkRotv/wpvzb01zxXxRLsRNarN3dYGY1+99H7w2dRyoj8VWFkB/hGQlatDbodRgDeguaWWA85Wa5BzONtqjnN3R5vo0D5UL9ggoh1jvlfbI14o3155EHeeAvgsDdW9xz4bmTFy8+WBdRen0JrYy7rgxuxc2LVzOib6vhxrY7n8EdO15poT+1J+qX66FuCX6uvi7JorwQMXFx7mNHeZQ5wUX7FPtQOu+OBlY+t59Etb4eWKqfGwnpqD9g31W+OFYrVe89WkdRyrEWrROo4LzxVrVN4uPpyjAqsqwG8ArloZ8qICVIAKFCogBx7XuxCObp0VQK06h1kW3v6DYVmiByBWuhdtDQUn52X9c3xn2fbi3As3pJMvpm8+hDVyTfiBI9pY/Ny9ae1T48R4zFpHPlo7y2XrOdp8Woy3rsks/rPitqg5MajAXhXgA8C9VpZ5UYEKBXBABIQdY35P7VYPKVIb+95TXfaWi72W7Hhv+abkQw1SVCq3kc82vAUFnxca0TWn12N9Vw1dczGc1uv6R71aY/vwoGVK/jm2vngt5sHDYqXkYH1SxtiPre67wAFuCocaG8RB3BqsEl9fXTQvzU3b63mJbX3seoifxoUd8DD2tTlxLAZioPWt2/kjj121GqWHrZMd9+Yhudvf+ydzMzXpnTPxqUBIAT4ADKnDNSpwYAVwczzSDVIOJVt7bZHz1jRuzRfXFNrW+Ll4oT0UWsuN47PvEWPGgx9ffivM4/Ncc8H+Q6vXSvqIIfXEOxWnFQcdr8e+EnzkaTkj59S41l9zD/VL/UKYWENuGI9qUzVL5eOrhTwEcMVyzaXGyrEDrxwfn23OPkB+aH2YtfOp+Kl2uXxyNMnFDtnjutFtyH7W2mr8ZB/Y9yxtGJcKHEkBPgA8UrWZKxWgAkEFZh0eg6Qii6M49zqwR9Lb7fKouqUIKFxwCNf2qLlrTduF+oKNt89uJS18HLc8H9I3tFaac6ze2Fel+Kv7aU19WmibnHxK/XJi9La19bfj1vH1/wxA336mia4hHi7dZc417+Kv46X6uHBq5kL5leIiL7QpOK15oA6zdJ0VN0XrkI3WDf2QvazBTrcxH72esk9gg1b7t+rbbwO2wiUOFdiKAnwAuJVKkScVoAJnCvQ8FFBitwKi+YjXqDgjcukdo9d1ANwetZB/NNiXa87alIx5wC9Rra1PqLY99pdm3xtfx0rph7Sw/mKr33a95xjXv8TI4ZzKabW6aN6Sb4ifrklIG2jow/LNay577R8595qaYk/VYNT6+va8b742Hvy5Z6AEWyrQTgE+AGynJZGoABXYoAL6UL9B+kMoQ6PewXofJHvzb43vO/j65lvHB17L+gMLLWK42hX+0ePixbl0BWyd9bhXfUdfH+lqhC21NtpS5vHqlZuthY6J2LVtL+41vCynkryhHVrNx875aqx99tRfMV/5n0PghXqjXVF7u0dX5Nhbvy1osGJdyIkK+BTgA0CfMpynAlRgWQV6HzaWTZzEDqWA7HP9D8hZh+AZcRGT13qbLQ89LZpv3trVjqWOrGWZirZGdlyGSi8oUKNnji/3PxRv17bQtAajxredCnVIsofx9iGtkGfOtebLg/NUgApcVuAqCkEFqMCaCsjNboWb7mrqtNCkBcZqupDPvhSwh139eSD7V49bZA5MH1breL44Mq9zHxk3xKnXGnLt/Znkw/fN98rXhatrrPsu25S52F5OwbA2KTqFbFBn4NqxzIf84VfSxmK51kvihHxaxnDp1BI/lIdey4np4qyxttz35Sb65GiUo0EoZg5Oja2PQw2my3dUHFfs1LledbbxbZwtaGNz4JgKrKAAHwCuUAVyoAJUYIgCWz0s2EOPiLXVXIYUegdBbH3tHrDrpSlb3BCO2LaKG4ojMXJ4hbBkTX7x/9VXX31mNoJ/jA/WdY7or8QPPNGmcIQNfKS1OblsXHMaY2bf8u/FBRroeJizMWVe29l1jEP+sHG1qfguXzvn4wA75GHtMA87X2v9fHat54VfKHYq/9a8LF6IR4i/xXGNQ9h6rTaOjS14Gt+uh8a4H5T6h7D3upajt9W1de33qjHzogIjFeCPAI9Um7GoQIYC9iaa4TrVVG72vOG3K4FLy63ujXaqHAvJtQf2rgD2ONo95mtzs+NVcrb7z441T1cOYg8ftNrH9lNsrI8du3hYm9XHLXSYnWNODrpmuh/KIQc/FScH08fTNx/i0GOtN48crVrn1yq24MSweutYq01KDq4Ysby1j8/WN699Z/ZLtZnJmbGpQG8F+ACwt8LEpwJUYAkFWh0CgIO2V3I+/NUPor30ODKu1DxWd+wXaXu+euNr7rGcte0K/ZIaoLYr5yrcbN3tGPr75pEf8oV9r9bHo1e8lrjCPYV/ik0Nr1R88HW1vvi+feCbd+Gk8nP5uubAXzjUYIv/Cq8UHik2sVygW0gzXxyZxzsWx7WuY4fih3xda6458ETrsuk5Z3PVY8TFHMahNlcvi5UTy/rGxqJxi5fOUb75yRcVoAKXFeADQO4EKkAFmiow63DUNIkAmD2Y6ANGwM25hAOUbp2GnKQCnRTY+/WqZbPXrl5jP6yA3Se1Wlo8G732c9XilY5r8yyN28KvRkPEb4EBLNsCG61dD41LfEJ4WEO90WJ+xTakwSz+winES+toOdqxtj16P1VT6BSqQ2gN/r42xqNVDX1xWuH78uM8FTiKAnwAeJRKM08qkKCA76ab4Hpu0gLjHKxRpzenFHyxsW9XerDBmhx49BvzbKmAT4GU/ah9S+xzfXS8Wf1SzuKX65trP0uT3Lj4LHL5leQc+gdda90vXboU/SatzivETdul9FtiIV4PTGDrFnXwtdo2pS+8Q3sFcVKwVrAZVQeba6u4rXA0P1cNJY5+a3v0XVy0D9bRwq9la+PFYoX2ckteJViam6smpZgaNwUjpqELA3WQNfhLXD3v8pM5m6v4yOc/X1SAClxWgA8AuROoABU4UyD3hm5lww0XN2q7Pmts87LjEbxmxByRF2NQgVUViP24Dz6v0PrySL12rd1qn4O+/GbPW91cfFJsXH6+OfwDMtb6/EvnW++J1rqU5pXrl8pb7FJtczm47FvXxxVjxNzW87DXpdUM+cHOrss4tOayxxywMY612J/YqxjH/HLXc3nl4reyb50/8NAKT+nrseYeWtN27FOBoyvAB4BH3wHMnwr8nwKlByZxl5tujX+vIsQeApTGtYcxO9a4voOKtvH1Q7g+nxnzPHTNUP1iTFx/2DM1++4isn+EuqP1W45ZAQ+bu/4//7BpzUjrjxqEYoCjbsHNtiGcLaxZPXTOqfyhibbHnG71+hH60FLnCj30nO3bmtj1kjGugRJfVx4xnJp4Mezc9R56pnCABrpN8dM2Jdprf19/lCY6d/R9nPQ8bHWr10v7uP5a6wqeJbxacynhIPxzXjmcoXkOPm2pwJEV4APAI1efuR9OgZwbao44uTf2HOwaW/0Pf43TUodeuffC1Tq06gvXLfFtlfeWcXBgbnEtrF7/Xv8joLb+0B6tC2/r15WPfyhnlw6YEz+8Mbel1qdHTQ5aS93XmLhG0cpaay4ldQEHtKm8rL2M9ZzOfYU+tEG7Aifh4Nsv4Add0WJ+y23urwLw5Zqinc939DzqN+IaWW2Pj9aa8ajAFhTgA8AtVIkcqUAjBUbc/BtRbQbTMmccbHDwS8EWmxQ7nXCuvfZlnwqU7Lka1ex1UYM10hc6ha43vYbrvpYjcBDf4mFex7Y2K441X+Tg4qnX0Ne+Lp/SOd//BCrFa+EXyxWa5MTCNejyCcVLjZVj5+Jg5zSeix/WXWsWS49L/QQD16XGK+mn4ITqVRKz1Cekb2itNB796hXIrYvdj7n+tYwRz/LQuGKDt8zDR9vYfopNKpbF5pgK7F0BPgDce4WZHxWgAlSAClCBAyiQ+g8CSBH6BwlspNX/MNHzrfuj4rTmrfFScxDtc+ul46T0e+P7OEhuePtsRnFrEUdj6L4vtxHz0FfaVV7QBu0qvGI8hK/lbMfA8M1jXbe6Rujr9RF95Lbi/wioyT+nDjm2NZxCvjnX6Qp8Q7lwjQrsQQE+ANxDFZkDFShUIOemXBhiCTccAnU7mhhip8Q9Sl1StKANFeilQO4/NHLtXbxd13bOZ4PFFDy87dqKY3B1aQkd0Jbyr/UvjYvcUv3B07bwl3l5YR3zsRb28I/Zyzp8UmxDNqJB6OWKk6Oby9Y1JxwQSz/80X3LM0cv64txCwxg5bYx7WN40KtnDrUcYzn0XsdeS8kjxaaGb2q9evPw5ZCiFXLwYYTmU/BD/lyjAkdW4KojJ8/cqcDRFeh50Du6tsyfCqysgFz7rn8Y+OZTcwHm6p8tJfyQm2iR6699U7Xco12ObmK7Jd1ycovVFnlrzBQ9tH0shmvdFQNcXPYt5lI5t+LhyjGVQ2m+rpilWC4/aCNt71xc8XPmtsDRl09uHVGX2poAR3i5sDCn7ZBDLmf4zWxdeWg+yBdzPntrB3u2VODoCvAbgEffAcyfClCBYQr4DikuAjm2Ln/ObVMBqTvevTOwh2M7rok/KgfLMSWHFBuNW5pLqZ+OndrPzSkVd7bdXvMK6Sr7xvfamx64RkI5w8anScm81lH3S7BSfEL5pfjTZg0FSutY6idZ1/iuoVoei1i+rutV5lzzeZFpTQWOowAfAB6n1syUClABpYAcMmIHDWU+rKsPMatyLBVjb/mU6uDzm7EfcXDW+073fVxT5lvhpMQK2ciP/bnyDPnE1mK10uu6r3FlXr/1WmofeaXa0259BfR149o7qLm2Q1Yue6yltLX+KTHEBvs+ZJ9iE9NCMGJ/BTwlToinXQOebq0NeNv5FmOJ2/PVgju0ac2zF24LnqV1SdHbhy2+8kIbyyMlFmxcbQwf6/DFOKXN8RFbvqgAFfArwAeAfm24QgWowAEU8B2cdOqzD5Wu+K45zZn97SmAAy7a7WWwBuNRh39cg2h19pqD7msb9tMUyNVP6rH1V2rOqXar6eHivYe6zdJZ66n7s/jUxsVnaumeqPWv5W/9pSat6pKiSct4NhffOIWXzxfzPoxc7QQn1wcc2FKBIyjAB4BHqDJzpAJU4IIC9pBhxzCWed+a2ITWgJHbhjDBR9u45nJjjrKfcSgdldsR46CevnamJqFf9p/LKzU/fS3aazQWU2Lw1U4BrX871DiSbw/EPa+0GL0najQr4YrrymYODe08xvAriSnfBvTh1+QPbq4WfF1rLedaxvFppPmmxIONrwUe4qHFvLSY061vXc+jL36tXjqPVpipOKHYsiYv2LgwQ2vWHnh2PjRO9Um1Q6xUe13nVB/EYEsFjqYAHwAereLMlwocXAF9SGghBQ6lMazWcWPxuE4FShWI7VU5XB/9gB3TKEV7aIg2xYc26yrQYk+0yG4UD3wOYP+iLckBWDHfmhgxbFmv0U58rb/mm5pjCs8cG81J91MwUux1jhrTN69t0M+xFR/hhTcwQm3sx8BDvr41H2fXfIqOvjgz5kN8sQb9MU7h6dImxY82VIAKtFWADwDb6kk0KkAFFldADiCut6Wdc6gRXxyGXDi5WBYjZcyDVYpKtIkpMGKvxjisvu7SCJ8pqdxxvaJN9aNdugKuOqV751n6Ysm8by0lQsr+SLGpiSX4rrcLE3auNZnzaQGdXOvAlNb3gr9eh5+e033fuouD9svp+3gJBtZaxtPcgI9Y0vo01Law8+kj6/alcXP8LM7Wxr689Txygsa96o04I9rUHLAX0I7gxhhUgArEFeADwLhGtKACVOAgCuhDjesAlyKDPuRpvBTfUptSrqXx6LdPBUbt1xnq9cyN19+MisZjjqpLyt5KsYlnNN6it4atdXHh2V8HIDnpvOw4prLEsHHsOIZh12v9Y3g6X2urx67c9Hqon6tjCItrx1Og5hqAL/fg8fYNMy5T4KoyN3pRASpABfangD0kY4zDxf4yZkYpCqD+2A8pPluzQY4x3lvWQOfYKo9cnFz7WD22to4a5OgAH+Qa8hXb0DowWrWWWwh3NLcQl95rUgOfNr55H6eWutmHgTpmyr7R3HVf47j6sEUM5IR58cGcy79mDjFrMHr7ah1yYrlyK8XKiattXRxkXeZ9XPQ8/GUOfY2f2k/1d8XQfFLjaTvrb2PkrruwQ5h2TfuzTwWowJUK8BuAV2rCGSpABajABQVWPlyszO2CiIMG9qBZG7Y1Xi0f+ucr4Pr9T1JX/c5Hdf84nb0eZazfOXHAL8dnK7al15XV1+YbW7f2NePSHGpi9vS1+YzUMpRXCg+xwTuEVbuWwiUUw17TIziH+GCtNi/grNCuoqlokaKr3hP2GhylJ3hqLjmx4Z/qo+Ok5gwftKmxaEcFqMCVCvAB4JWacIYKUAEqcIUCuQecKwA40V0BORi2rJM9mNpx94QWC9BS28VSO38YmMILOoT2g9jALgVz7zaiVUgv5A873WJtlRbcQnx0/fU+SNEghNtrzfLSnEtjtsAojR3zQw3Rxuz1empeYueztfMhWx07p5+LCXvLLRazREONafeeXgv1Yzx1PqFvfoZitFiL8bQxSvQo8bFxoZfM6761c41tfD3WfesbWrO2vnGuvj4czlOBIynAHwE+UrWZKxWgAlMUkANKi4POFPIHDeqq11EPmjPy1vrPiJ+67YWn5WfHqVi1dtBsVvwc/iUcS3xyOMVsoW/MTq+7OAPHtaZ9a/qufQk8xMe4R9syBrCsXqEcQzkBL2TTY83y7xEDmFuKJVxzazIyP2g6soUeOk/M9eahY6JfErvEJzc38Mv1oz0VOLoC/Abg0XcA86cCVKBIATl4pB4+RhyEipLYmVNqPUrS7oldwoc+8xRY8XrWnHR/nkoXI6deP6l2F9HTRvKj4KO0SYmTYpOWmdtK8HvHcEUeEXNEDFdupXM997WL06zau7ikzI3Ux/UrIVI49rBJzXv2fk/l2UMjYlIBKtBeAX4DsL2mRKQCVGDnCujDEPquA1poDRLBBmPdujD1Ovt9FDiS7rFcQ/uzj/qXUSWucOsdvzd+jUY13EZol5tbST4lPjFepdqAC/ylDb1i6yHfraxBi63wdT38QV1Tc0itq7ZDPzeW5qS1TsXTPhqrZz+Vm+WgtQGGtdnq2Oamx9L35eubtzpoPLtmxxYzx9dicUwFqMA2FeA3ALdZN7KmAlRgcQX0oUr3hbaM8Q6lYf1ctvYw57LhXJ4CWnf0j6gzcs9Tr511LH5tTWL47TIZiyR5bTk3cEfbSj35PWAttAGvFli5e1js8U7VBfZoU/1S7YCLNtWvxK5ljJL6SfwZL8RFKxywD2fwkZiaSw4H8Yv59sxt5u8DFJ1sbjEtcrTdoy2uU93uMU/mRAVGKsAHgCPVZiwqQAU2rQAOIK4ksIbW2mBe2tQXD4apSvWzK6lBiU+/DMqQc/ZpWYR6r1KOuBZLGei4ul+K18qvJi/Zs3jX8AGGbQVT5lJfI3TVHFN5uexquWoeORppW9S+losrP9+cju+y0ZxCfZdvqznL0fUQWGzsW8fXa3re9m2Odl3Glo/Lxs5ZH4zRWns9HrkfJK7mJH3XG/y0LeY0htZTr++p79MgJ8dYjW0NcrDFFv65fi3sfblhb7SIQQwqcEQF+ADwiFVnzlSACiyvQIuD4fJJLkzQd/BMoVzj68PvdQjfw0Ea10ovjVw1WUk35O/imTIn/nrPluKF/EJrKRx72wi/Go7YD2h78LX8dM16xKvBzOFWqllODF8uVlNrF1u39nrs4uea0z66L7F98WOa5cTRMXP6NoaPq8UEd/hrP93XfrCVOd3XNkfsU4sjVp05U4F6BfgAsF5DIlABKkAFpiogh2bfwXkqsf8Lvjo/rRG4uvTEmrYf0ddcdL9lbPmHhP7HRK84LTlrLM1X+q7f9aXt99K3eefkJb7wRwt/vYa5mtburxqsnr6t83Zx1deZa903p/10H/Ytudv9gBgz21xOsA99FkBHtCX5ia/PX6/5bGzMVnVMjWfj9xy7OOk56eux5hJa03Zb6vtybZlDbYxa/5a5EIsKUIE2CvABYBsdiUIFqAAVGKLAlg9jrf5hM0TohYLYmuMftr0pjopTk8cWOLry09eC7rtse83ZfbX1OC35z6pJLAepma9uvvkYpm/dd22tqo0vj9g8dEMbs89d9+mYi7MHe2pxZRVl3+m9p/tXWo+fmVkzG9uOx6vBiFRgHwrwrwDvo47MggpQgZ0roA+F6NvDkB1DEthjPLqV+D5uo7kwXliBrdUKeztnf4kt/MJqtF+1PPV4Bi9fvUP6gLO1seNW6vXSJcQXObbKwcbCWMfBXKuYLXA0v1S80jzEryReKq+YnY1tx9Y/lqf2j9lq7JgOOVgat6aPXGbEruG9RV9obbnXaG99dQzp23XE1naYa9HqeLEYWBcf7deCBzGowFEV4DcAj1p55k0FqMDSCuCwg9ZFtvdhCAcvtC4OqXO9uabyiNlB7558S/TUfHQ/lg/XqUCuAthfsX0aW8+N67JHDLQumxXnREP99nFMsfH5cr6fArP3m29fyPyKrxy9cmxXzHUGp951H1kTm4sdz9CXManA0RTgA8CjVZz5UgEqsCsF5PAUO0DVHO5qfK3Q4NoS08ZoOfbpKvxrcvDhxrhDv5hd7Tr4oa3Fgz90Q4v5mrYUq6Z+JXxLefpiufBq62X3l4wRR7c+Tq3nJaZ+aQ7o6/WWfatFDBt8LOeYX+/12J6IrffmV4qvefu0Rw3lL//GXr3rZvE1Z/R9HJGHztlnO3o+xn00ny3GC2mI2rfOy7WXhEevF/KwcZG7ne/Fg7hUgApcVoA/Arz4TvjoRz96evvb3372fsc73nGS98c//vEz1g996ENPv/RLv5SVwVve8pbTq171qjOcj33sY6fb3va2p3vc4x6nRz3qUaf73e9+SVif+cxnTq9+9atPb3jDG07XXXfd6VOf+tTpDne4w+ne97736bGPfezprne9axKOxH/5y19+etOb3nT6wAc+cObz1V/91acHPehBp8c97nGn29zmNkk4NKICR1ag96HtyNq2zF3XaQuH3dYcdf7QVeZq44i/CxsxQm2L+CH81DVoIHxSOOl84Zsay9ppLFfsmL4uHxsjd6w5xXxhW6tDLE5sHfHBJ2Y/aj2FT6zGo7i2iCP5ohYt8IABzBQ9xcfawR94rrYXdx2rNIbmb3PT+FvsIx+d48g8JC7qAi4j4iOuK1ZPHsCW+FpzHx9t4+LKOSpABfIV4APAfM2GeshDtRavz33uc6dnPvOZp2uuueYC3A033HCStzyEe+QjH3l69rOffbrFLfxfDJWHdk95ylPOHiBqIPkra6985StPr3vd607PetazzrD0uu1fe+21p6c+9amnj3zkIxeW/umf/ukkb3nA+IIXvOB0z3ve88I6B1SAClxWAIeoFD1WO0CtxidFwxqbnFrVxPH5In5Id7EJrfuwtz4PbXx55GoCPO2HOV8MzGsfzK3QCq/UHGr5jopTy9PnD61G1VL0GhXLl/PM+ZL9Ap9U3bQd+sAozV1wajFKY8f8NC/kG/MJra+aq85T+GPcIueQHq41V0zXnMu3Zs7WRjToEdfi+mJYPjW50ZcKUAG/Av4nPX4frkxS4Ku+6qtO973vfYuiP+95zzt/+He3u93t9Ku/+qtnD9mklbG85KHbr/3ar3nxP/vZz56e9rSnnT/8e/CDH3x64QtfeOb3cz/3c6fb3e52p5tuuunsAaB809D3+uAHP3j60R/90bOHf1ddddXpSU960ukVr3jF2Vv6MicPBsXmQx/6kA+G81TgsArgoHpYAQYlHtLZd4AdRK0ojOTjewtgKF9XQG0PXOBgrG18GKm2Pn/X/Ow55I3WxUf2UMk+gl+Jr4uHnnPVQuZCsVw+FlPG2k76+oU1PZfTt3g5vq1tRavafHI41eQ+kmdOTjHbEO/QmvzParxgJ23oBTttE7oexE7W9Vv7om8x7Bh2LdtYjJgWwiWEIWv2DR89X5uTqyYhTF136YfyzMUOxc1dC2mbi5Vqb+sS0iYVM2bn09jGnqFHjDvXqcAeFOA3ABevonxLTn5EV963v/3tT+9///tPD3zgA7NYv/e97z295CUvOfO5+93vfvag7Va3utXZWL5h94AHPOD0mMc85vTOd77z9OIXv/j08Ic//P9v7zzArinK+70Cgr0CloiAJbagBktiNCJij0ZF0IhRI1H0CmrEnmj4mxgNBjT23k0xRmwx9hpLjAoasWDEgi3GHkUFFL//+1t9js8778zubD27e+69rveb2ZlnnnI/c863O2dLsf/++++woav7TjnllLL9qKOOKv7f//t/KxnpuelNb1ocfvjhxVlnnVU84QlPKG584xuXi3kroV9WtBipKwm1nXTSScVtb3vbX/YUxfWvf/3iWte6VnHccccVuv1ZC5JNb3NeKaMCAQhAYCACOlBte3DaZWzTcMID6qbj6+RD/eG+xsfaUnolm8O1ic6UraHb5+BjFYM+/K/SUdVX5ddc+nLmcddYZGMpHPuOw/Tph2k9Pkebngf4zW9+sxUz6es7p33r6zqfwvGpmGPzbh2xpPwL42izP6TuNv5s0ph1zKVN4kusEOAKwInPgQc/+MHFoYceWi7+tXX1ZS97WaHn9mnTbcC2+Gf6LnjBC5bt2pfcS1/6UuvaVtoi4iUucYnikY985LY+7WjR8P73v3/ZfuaZZxZvf/vbd8joyj49O1Cbrmb0i38mfLvb3W51pePrX//6HbcJmxwlBCCwnYAOmmJ/26W27+kgd8htaP1D+m4su9rwB7O+3kSvOPq/nLFN2bf1LccXkzGmXW2ZHtM7RNnVxyF8mqJOy4VKm6Ohn03nYjg+Z38MGzl+SGbMueNtWS5y/VyaXG78xszKpXFoE0/q85P6TJuNXOYm33eZ8tvs1PWbXKysiz02pmvbunl29T813lhafPbZs/YueUrZpB0CEIgTYAEwzmUxrfpCfec731nGc6UrXam47nWvG41N7QceeGDZJ/nwi1hXEX7+858v+29zm9sUWjSMbXoxiW16rmC4vetd7yr0PEJtutIwtelKQm2S1Ri2+RGw/9Tn5/n0PbYDJ/PUDqhsf2pl+H0yNf/G9ifMXxv7VUyn8NnLiXEK89Z8iJVt8jLUGMtpVd7b2M7Jk/Tm2jWObXzpY8xQnPrwDR1pArnzK63hV4uuubps7ltZpVt9uXJ1euban8t16Pim4kffcS41Ls8p/AxtQsw+fuoQmAoBFgCnkomB/NAtw7rdQdsNbnCDSis3vOENy369FETj/Ga3/qrN5Hy/1ffZZ5/igAMOKHdPPfVUa16VXk+VP74vpmelkMqkCOg/c/szx8J9a19n6X2yeliu078c2zqQsr8c+ZRMeECWkmvb3oePbW0POa4ptxgHP+fqfI3Z0/hwC9vMbqrU+JjuUG9sv2pc6EeVndC3mK1UW+hDuF81Llc2paNLe45tmx+hnZBtSi4ct459xen/xvRBXPyzv8a0vW5bsfnl8xDr78vncH7W6W0qn9KX8zloMh/G4pWKp6o9J9bYeBsXK2PyYZsfF/ZNbd98zfHLcp2SbaIrpaOuPfWZlO1N2lIcNokBsUJgaAIsAA5NeM36zzjjjJUHugKwavP9X/jCF7aJ2tV/avRy24R+uWP9etnHj3/8420i5s9FL3rRQouFqW3fffctLnKRi5Td3nZKnvb1E5jjQUqVz1V966eNB0sjoPlWN+diB8ZVY2LyKW5NZL2O3HG5cl5303oTGzm8m9pPycuv8C8la+0+r75u/WHZJPZwbNf9HP+62mB8PoHY3I7Nj1hbvpV+JNcxd2J8wmhibNbha+hXl/25+18Ve1+xKe/+L7SZM3fCMX3s9xVfH770pWOJMfXFBj0QGJoALwEZmvCa9fu36F72spet9Mb3a/HOb17PZS5zGd+1o365y12ubNOXu8bZgqAadXWhNm+rbIj8Iz2f+9znWr8JWG8lZhuPgB60XbVpUbfJFv5Sr4d3pzYv+9Of/rQ4/fTTS9HYHJBsna9mp6nPNo5yvQR83n1dua+aRymvY/PlnHPOaaXL29B3qfdJ9Tq9db747+ec+Sv5HDnvd1g3mz6WUKarjVCf3zf7siEfxEjfA7bpERd6qVS4SbZvv8wXb6vKhnywzc/NWJ5NTqXmiW2SrbJhcrHSeMX6mrR5fzROscRiq4uric0cWT8P/HdBztgpyxhbP2fM3xjj1PdKTNb0dCn9fPA+mt+meyj7pl+lnwO+7j8zoR++z3T18V1putqWxs/8jfkZ6g7HhP1d93N86GrDj1ceLCbf7uec2o2R6j7vvq4+zc9UDPZ9HtpLfZ6kr+/N4kj52Kc9s9WnTq/L5yhmqyoXXk+buv/+9/U2usYcs/vuu49pDlsbQuB8W4s0m3Vt8cwT698CrOft1b0h94UvfGFx4oknllG/4AUvKN/Um0Lw3ve+tzjmmGPK7kc96lHF0UcfvRJVu/q1feITnyj22muvVV9YkT3Z1XbyyScXevOwbToR+8lPflJc5zrXKV71qldZc7Q88sgjS1sXutCFio997GNRmbBRvoX/uYcy7EMAAhCAAAQgAAEIQAACEIAABKZO4PznP39x7Wtfe+pu4t9MCHAL8EwS1dZN/2uLvjyqtj333HPVffbZZ6/qqvStp84X2TR/vG21s0EAAhCAAAQgAAEIQAACEIAABCAAAQjkE+AW4HxWs5T0V+rVXRnnL8e+wAUusC3eUI/f3ya4tVOnR1cA1vkinaanylZo2+/f8573jN725WWoD0MgvF1CVvxtQE2sxnRVjdfcsluAr371qxd1i81t/arygb71EtDtHa94xStKJ9p+D9TNu3DepORDufWSWbZ1nwP/PdB2DvRBy/uUq09zRuOq5k6ot0o2167kQr1NxnpZi8HamvjXlw+yrbsOun4XWAzrLmNcQq4xmSZ+962vynZoq0rW+sL4TEfYbvLqz/3/wOswvaana+l1e10xOyYb9lm7H696KBf2L20/xaEqTv//gT8uNHZ1Ok3ObFTJh7I2pmkZ2uhLb5Ufoc0q2aZ98j+lf4zYcr8HmsY1tPzLX/7yYu+99x7aDPo3jAALgAtP+IUvfOFVhOELOVYdv6xoYc423XbrN6/nRz/6UeUtwHV61F/ni2ybHm/b+1RX1zOf7A3IdbL090tAi7b+6QKxB2rnWrQrQXPlvZwW/6rGd/HL26E+XQJNvwf8vE1FFc6bqjF8B6Uo9t+e+qw3nQN9epbyqcqG5kzdD19ebzgfq3RX9VXN46pxsb6cGGLj1OZjS8nktvtnP65zHuT6m5JL5Sb8funKLtRn8zBlP+VvVXuX+RrGZ/6G7WZf/Vp0OO2008qmqjlgOrr4Z3bD0nSH7XoutrfnOVtsNiamw481uaWXMQ5NYvbHhfZc8rrxYS7CY1wb31c+/Dww3aEP1t619La8/769qw2Nl/+p3A0VW8rvqu+B1Jh1tZ933nnrMo3dBRPgFuAFJ1eh+Zdt+Bd5xML2/fYiD5Pzeur+w7QXiOg/Ej9OuuyBut6W2QhL0xPqCOXYnyYB5d/+unjYhw5v3/SpHGrTgZP9DWUDvc0J5BzQ2vxoor3NmCb6kZ0fgfDznztHcr+XTF+ufEjQ/PNlKKN9s2N1K+vs5nzWpGvIrc7HIW33qXsMlj7PMd/7YOlttI0p5UeqXbHkXl3k/YsxaNtWF6v67a/Kho+xqa91PlTZnVqf51DnWxPZKl25/HLlUraq5kFX3aHNKluhbNf9VB6azuOufjAeAhD4BQEWABc+E65ylausIvzCF76wqscqvt+/uVeyV77ylVdDvNyq0VWsX4uI4ZWE5s8Pf/jD4lvf+pYbtb2qX4POOuusstHb3i7FHgSmR6Dvg7TpRbhcj7rmzg5m7WDXyuUSm25kuSf9Q0TQdR4N4VNXnTaXreyqr834ddpu4+/YY1LzTtyq2FX19RmD+ReWTWzY2NiYseKI2Y61ydcqf2Nj6trCXPatv86+71+nbe9HVd779DHUVWXX+9dnPfShT91DxRPTq7ZYe5/xoAsCEEgTYAEwzWYRPVe4whVWr7f/yEc+UhmT9esqPY3z2/Wud73V7oc//OFVPaxoUe9LX/pS2XzwwQeH3YXXY/Z2CG01+L6YntgY2pZNoMsBg40d6oBDB2X2t+wszCO61HNm6vLv+309jDrnILxqfKhv7P0c/8f2qU97U2Nf54/67a8rh5zc5sjk+FEXV46OmEyMRcqWycbKmO6ltSmX9tcmthTXNrpsjM+FtVnZde7V+Wu2ZU912Uv9f2A+9V12yYf5UhenyeWWfeszu13zaXq6lKEPNgeGitn76m1Z3fdX1W2eeP+b6qjS30dfF39srI+vD5/QAQEIdCfAAmB3hpPWoC/gww47rPRRV+Z9/OMfj/qrdrtyT/Ia57cDDzxwdRXgW97yltXz+byM6q997WtXTbe4xS1Wdavc/OY3L3bb7RfT7uSTT7bmHeVrXvOask2yGsMGgTYEdBVQOJfb6KkbM4aNOh/ojxPocvKnvM49t/4kw+pxUvNundJJxrrmzNIYLOHzN4VPVepz33Se5siHMuG+51HV5+XCettxoZ6qffssWVkl6/uayvuxVm8SX45snU/qr5Mx33wp2zn2/Zi+6im7YXu4n2NfY9qMy9Hdh0ybXNXZjcXr23y9Tles3+aY9HTVFdNPGwQg0IwAC4DNeM1S+t73vnex++67l74//vGPL84+++xtcWhf7dr22GOPQvKx7eijjy6bv//97xcnnnjiDpEvf/nLxfOe97yyff/99y9uectb7pDZZ599ijvc4Q5l+/vf//5Ci4nh9uY3v7lQn7Y73vGOhcawwnmPUQAAQABJREFUQQACEKgjYAeZXq6PW0HtoNVK6R/iIHwInZ6Fr8dY+f451xVbl4XfprHH8mZzxZ/spOSa2ovJx3TH5MZqkz99+eQZev9T7V5m6nXj5Mup+9zUP58n1e2vqR4vbzqMm+/zdcm1+T9A42yrmsdm30obEyvNZ9NtZUyWtnoCxlOSvu5HVuXOy9l4nxNfNxt+TJO6zQ9f+vG5ftoYr8fq1tektLibjEEWAhCYPwHeAjzxHH70ox8ttLBm2/e+9z2rFmeeeWZhV8pZ4+GHH27VVamr9/74j/+4eP7zn1988pOfLO5+97sX97vf/cqDIp0kveAFLyg+/elPl/KSO+CAA1ZjfeXOd75zoav2Tj311OIf/uEfim9/+9vFkUceWVz84hcvPvGJTxTPfvazy+f26aq9xzzmMeVioh9v9eOOO6543/veV3z3u98tHvawh5U+3exmNyu73/Oe9xQveclLyvqlLnWp4iEPeYgNo4RAScAOypoeMK0Ln/w0n9flA3aHITCH3GruzeWz0jZL64rP2w3ngvXZZ9/2fYzW59umVje/p+LrVPzweQpz7/uq6sY2lGmrL9QzlX2Lc4jcpXSm2psyCfVYLF31hHqb6utTfkq+9BVX2zyl7I/JqK3vNi7X11w5MTHdKT5hu+luOi7Uwz4EIDAMARYAh+Ham9ZXv/rV226r9Yq1EKc/v8UWANWvRTe99lwLeFrs0364HXHEEZULbrqK8FnPelZxzDHHFKeddlrx1re+tfzzevSK9+OPP7445JBDfPO2ul4O8tznPrc49thjyxeBaAFSf37TVX+yxRuAPRXqdQR00LHvvvsWb3rTm+pEe++XbQ52ese6FoXKox3ArsWBno1aLMzPnsEG6mJ8Y23BsF52LcdSNoTNUGfO9533KRWk9ObKSUeObMrWkO0hH7NV5a/1xcbG2kxnX2Uue29PPlf5Zn2hnMXqdc2tbrE19TsVe0qftafG9WW/qZ6pysc4hW1Vx4WhrI+zqs/L9VG3vHfVldIzdCxef8qHrrExHgIQ6IcAC4D9cJy8Fl2V98QnPrG49a1vXfzzP/9zuYCnqwkveclLFgcddFBxt7vdrXLRzgLUVXmvfOUri1e96lXFG9/4xuLzn/98+TxA/ed6oxvdqLjXve5VXPWqVzXxZHmd61yneMMb3lC8/OUvL975zncWX/3qV0tZvXxEzyDUbcjyjQ0CKQL+YCMlM3a7fOLAZ2zqv7C3Du59z8G+9a0nE92tKpf2WapjkpN3Xem+1157dXesJw11MfVkZlA1ddz7iNHmwaCBdFBexyBXtbHqS1+uXck1ZZzrYyhn+xZrEx+nImu+Wyxd/TJ9oR7TrzIlE44Zat98ifkxBf+6xh2Lq6vOqY4fOl82V2LxbxLnWPy0QWBqBFgAnFpGAn9OOOGEQn99bboyr+rqvBw7ek7gUUcdVf7lyKdk7BZfbvNNEaIdAhBYF4Gqg+W5Hsya31UH6uvi3cbuXOOomlttOIRjlOe5slEs3vehWYXs+ty3z1ufOmO6vB3PLiYbaxuT8Zi2YrH20TbW58vntQ+/u+jwefNzzLd30b+OsVPiO1b8c87XWIywA4FNIMBLQDYhy8QIAQhAYAAC/kSgifq246ps6GDe/to89D3UveSTA+M0ZozKeU7eTUal+Wdtm5SjMNam+z7HVm+qo418aCuVO9Mtecn4P7XFtlR7THaoNu9nlY02vmpM03FV8k301eVJsdbJmL0qn3L0VHGdSp+PtS7epj6b7qbj+pS3ee51xtrUb+25pddJPY9AzpwwmbA0C9Zu+2OVsssGAQhMiwALgNPKB95AAAIQmBUBHfQ32ZrKN9Edk23zNlgOWGMku7UZUztJlDZf99rDExUb62WoT49ALE+xttBzk1Fpdcn4fd8ejl/S/tjfj55d6vPoZahvJgH/WZw7Af1AOKfvE++rr/s8pNolo76qfq+HOgQgsBkEWADcjDwTJQRmT8BOTnw5h6CWfOA1hwPLPq4GnMM868tH+3z1pc/r8Z8F2YltXibWH7blyA81B2z+Wxn6VrWfir9qTNc+89OXXXVqvOnroks6/Bbu+77cuhj7HwBUN+4qrW76wn1rn1PpucVizIkl5JCrx+TC8Tk21y3T1XfP3ceSavcy66772K3uffIx+LqXya13HZ9rZ8lyYmh/Y8Q5x8/zGFywAYE5E2ABcM7Zw3cIbDgBDkzmOQE4CZh23qryw2eu39w15Sn5pmOqPPa59vWqMXPoizGKXfljclZWxZYjo/Fjcsz1qSquWJ/ptTIm06VtKL1tfAp9Cffb6JzLmLpY+5rL0tOXriZsFV9djE30DSFrbKxsYqPNmCb6u7Ab2rcmcSALAQhsJ8AC4HYe7EEAAhCAwIAEdFDYZLMD+LoD0br+JjbHlrUYx7Ybs1eXn7r+mE7f1nW816X6lPLeJjaNaRNDmzEhO9uXD+a71X1pcqmyT19SNoZs9/6rbvtWt3JIH0y35cH2fVnV5+X6qhuHOn02V+rkwv4q/cY8JhNrC3WzX0+gLUfLty/rra1Xwl8JvA5Pws9uFTvfF44by3f7/LWxZ/63GcsYCEBgHAIsAI7DGSsQgEAHAm0PVDuY7GXoug7eenF+RCX+YHOuuW6Ly8frObTV1+c4883KvnSHn4s2+qfGqgsb8WjLwNtto8OPT9XDfMXkhrIdszVGW5t4xCmHVRv/vd5UvY3eujHeVp1sm/4Y51ib6bY+lbE/k2tSKkb707ihY27i2xRlLQcx32LsYm2xsUtuq2K2rrjDz8+6/MAuBCAwPgEWAMdnjkUIQKAhATs4Dw8kw/2GagcTN38HM7BQxXaQ7A9Mc0Od6lzI9V9yU5s3xtTKJrFUyVqeq2RifW3mhfT4Z7/F9HZta+pXLM9tGYe227KtY2A+ez99XeOrbJufdXb66jd/Y89/ND9NJsdmGKuNMR2pfpNrUppOG+N1qy7/QxmT7bM0u1amdMdyG/pn+zFdVeNj8ik/fLvpVNnk6q+29mS7y1jv+1h1+dvUZ/HU5vlq3+vydfWtczNfYt8DXf0yBlV6cmSqxo/ZZ76qbLs1nU9t7TAOAhDoRoAFwG78GA0BCEBgGwEOgLbh6G1n07l2OSjvLQlrVNQ1/tz540+CrF4Vdle/qnRX9SmeVExD++Rthz4YMyurYhijL7b44/1v4oON82WT8bmydfkLmXu9dWO97Bh1Y+Vtpfy3OZPq9zqsHsqaDusPy9h88DJd+YX+eN1zrBuPOq5VsZmOKpkl9qXiXtocWWLuiAkCSyewx9IDJD4IQGBZBDh4Wk4+UwfIPsJNyrfxCGO2dpVhn2fVpm66w7HeTt92pTu0G+6H9kP/muzb1R91cYQ+1NloKm/6/Dgfp/VXlaF8jKXG+1jNXji2yk4ffd6er/ehO0eH4j733HNrRWMMfdvYvps9K2sD2BKwHFfJxvSlxsVkq3T7vi5jvZ66utlRmYojpkOyNlb9TcbG9I3VFvrdl13PwnR6Jr5u/VVlTD7WFrNbpXcJfVOKWb7E8tKG85TiauM/YyCwaQS4AnDTMk68EIDAZAnoYMz/VTna14FblY0x+nTg2OTg0fh09c30WNlV31LH2zyzcqw4LS992zW9sTj8PPT10IcqHTG9VW2h7irZWF9qvPdfdb8f09Olzev3deOU8jFm04/xdS+bq0++2OKvHx+re7/Vn2sjpmvMNmPU1mYYt/SozW9NWIRjvZ6cet1487dOrs6Wj6mrrjpbbfvNx645bms/d5z5Z2XOuKkyz/Hdy8TiaMLB61pHPeZ/Uz/60NHUJvIQgEA3AiwAduPHaAhAAAKDEKg7qKrrH8SpnpW2jaHtuJ7dn706naiktjmdxKRi6NJexaaL3jHGruvzEdq1fSv7iH0deVmHzT5Y9aWjz/z15dOQeqaQb/PByiHjRXc7AqnPRaq9nRVGQQACEOifAAuA/TNFIwQgAIFWBOzA0cqmSnSyMLcThqb+VrFpqqsp37HkwziqYm7rk2w01asxoW9t7XcdJ9/Dv5TO8CUgGtcmDou/KbeUX13avS/eH19vo9/0WtlGhx/T1B/J25/XE/oT7ntZ6jsJ1OUhxty0iPUYW52P8iHlS85YjdcVobFYc8dLR7jZXLTS+lM6QzmTt1L9VWNNbqhS9ue8GV9f+mc/hv8f5MZq+nLkbY5ZmTNmbjJLjm1uucBfCDQlwAJgU2LIQwACEBiQQOrAP8dkl7E5+oeQMZ+t7Gpj7icvdfH3FZ/xVmn1Ott99XeNoau/Zl+l1eti83K+Xjeuqr9P9n3qqvI5t69rjmSnDx25/i5ZrgnHJrJjM7PPnZV9228au/yo8iXW59ti463fyliMVX0x+aHbxM3/DW2vL/1NOHpZX+/Ll3Xqsdw18aHpZ6WJbmQhAIHhCfASkOEZYwECEIDAaATswMwfpFrbaE5kGjIfzT+V1mYqtG/91ubLUN73VdWrdFaNG6vPxx3G6Pvq/GkiW6dL/WNy87b6jsPH6nXLpvbH2rztpjY9n6ZjU/JD6EzZymk3f2I5sb4cPV1kxrLTxkfzLcanjb5wzFB6Qzux/Zhtizcm30dbjv6YX3W2q8ZU9aX02pgcf1M6Yu2mN9YXtqVs+/Ym+kL9bfab2jN573OOXY1rOiZH77pkjMO67GMXAhAYlwBXAI7LG2sQgMBECfgDINXtb6Lu4tYWAZ8zA1J1UB6Tt3FTLLvOQY0XjynF3daXqrxa7iSTI2fysbKtfzFddW1tbGlMm3F1vsyxv08OcN05A/rk67Wn9FoOrPRj6uopnXXjcvq76E7F0vV7KsdvZIYj0DZ/XebScNHk/7DXx/+xQ8aBbghAII8AC4B5nJCCAAQ2gEDsYH2qB2xN0jHFGLxPVrfSx5Y60A5l6w5MQ3lvY531VHyhTxaflWF/bN90WxmTsTbTW1eafNsyxxfpzpWL+aGxuW9/Dcd3sWt+S0dXPdKlOWt/oZ/WH2tfYpsxtbLPGKf63ZAb4xD+t9VZN66uvyrm1Geqi84qe9bn7Xa15eev12u2mpZd/fH2UrrM57D0Y5vWU7aa6vHyXXQ2GdtH3rzf664r9rr4Lffr9hX7EIBAPwRYAOyHI1ogAAEITIZA3cHcZBx1jszRZ+d+56odYFvZWeGMFFjMKrVZOaMQVq7ayVTX+RyON70rQzOpDOF3yKYLij51dfGj6VjjOif/2/pq3w8xRtJZ932R+2NAnX91/TH/5tK2jtjWYbMqH2P4UzdXq/zr0qfY7K+LHsZCAALLIMAzAJeRR6KAAAQ6EvAHZuGBoPZ9f0dTgw4PfTdjU4tBPFO+ms8qY37njPM6VJ9y/mK+GZ9Yn8VmHKpkTHaKZczvWFvfvtfZsH7j27f9HH3yIWY/1Z6jcwyZmM992+2LwRi+9h37HPQZ17afIxtXFWtoo8ucMF2yZ/XQB2uv8qmuz+us0uflvE8x/aFsTKaurYkvdbqW2t+Vc9fxXbiG+Q3363Sv0/c63+iHAATaEeAKwHbcGAUBCEBgcgSaHthNLoDAIR14hgefYYwxmUDN7HYVYxinD6Ku38tOqe79DvM6pJ+5VwCFPtjcCn0N98NxTfardPk+1W3fyiZ2libreSwttpx45hC//7znxFQnY/pUNtm+8pWvbBM3Pdsag50cmWDIjl3LkZU7BIKGlJza57g1zVOXGM2WMbSyqU7T03Qc8hCAAATmRIArAOeULXyFAARGITDXA+5R4PRkpM2Bth+zKTlSzD5WY+DbekrJYGrMZzOwDt/7tGk5CeOy+KrKpn40la+yXddncdXJpfpjPHL8bzsu5Udde8xeakyO/6mxbdvNvxzbJmNjYjbVZ3Kx/nW1yacc36piM99zZCTrFwNzxuTImA85jE1fjqzpVdlU3o9N1c2XVH/XdstvSk9fMVkcMX3Wds4556Tc2NEufTZuRycNEIAABBZAgCsAF5BEQoAABKZBwA5Ep+HNdL2o46SD79gBeKwtjLJOdyg/hX357P+8Tzkxe/mp1FPxyL855sjmpOVjrBiMY5hX8yds77JvscV0DBFvKjbZH8JeLK6qtioeVePa9lXxkM66/pTdvuKwOWdlyl6s3Y8J6zF5a6ubB9bvddrYVGmyKsfczNcxbVbZisXv2cT6q/TF+kxf2NeHbuk0pnX6/FXgvh76pf06XbExS22DxVIzS1ybToAFwE2fAcQPAQh0ImAHoGHZSWmLwWa/xdBFDVnSAeuYsWj+2N+iJsRAwYyRmy6f6S5j+0AmPvqb2pxaN5cmbI2dlTZ26Bhi+vua733pMRYqQ39t7nkZq6cWf2xMF/9CP8ymyqo+Lzd0PeWHxT+U/S5cq3xq6/dQ/lT5uq6+TYp1XYyxC4G5EWABcG4Zw18IQGBSBOzgyg5Ebb8vJ3XAHv6FulMH9aEc++sj4HOUqpt3NpdsPyy7zrFwPnl/Qltt97v62NbuGOMsNstTrIz5YeNifb7N8lPXpv6UrO+LyXjdOfWY76bXSq8nbAv3vazVjaPt2xiVY2+xeIfwoW2MNk4+pfiEPHP8T+nKGRuTSfnQtx3Pw/ww26lFP8mZjI3p2y/TKzvaQnvWP0ZpsVk5tj/GYIxYc2w09Sc2x3LszFFmnfN0jrzwGQJzI8AzAOeWMfyFAAQ2noAORO3g1R/MzwWMfJ+j3235Wqx95M3y3tWX1Piu+k2vxWz7Syl9DqtiCud4X1y9Tc/Y173MlOspJtbuY/J1xWQyU4nP/Mv1y+S7+t9VT+54yfnYcsel4jNdod6U/FDtY9q3mIeKJVfvVPzI9XcKcl3n+xRiyPGBuZFDCRkIzJ8AC4DzzyERQAACG0hg7gekOtC0GJocdDaRXfe0sPi8H74tjMX3+TF91YfSb3p9PL7el//r1mNx5voxBIOmPuT62lQu5ofFG+trqn8oefkY+hdrG8r+nPWG3LrGYvpU2twJdcbabVwo23Tf9FTZD3VW+RPrC8evc38K/skH8Y4xt3xU+Zkj04Sx+WNjYrbNpsnEfLe+OZex2OccD75DAAJpAiwAptnQAwEIQGBRBKZ24Lr0A87w5CI2mcKTi5jMWG1d8zG1+TUWt7Ht5MyrmE9+rnXNtdcf6kr5F8p5HU3qfemRzZQuY5Xqb+JvW1nZNj+a6lin3zm++rj68jVHj94CvOeee+5w0fsTdtbpresP9W36foyX56+6yfh2a+uLn9dtOr1ttcVkTJYSAhCAwFwJ8AzAuWYOvyEAAQhAYNIE6k4eUv1qT/U1Ddh09aXP2zfd1uZP0MI+k1ln2ZWB4vMxDhVLjo0cmZR/XcaaTtNhpbXHSsnkyMXGhm196DF/qnSZTGg/tZ8rX/W5MB1WypbVVc5t8z7H4rbYmsQV+wzX6dGiX5utTm8bnYz5BQGbD1aGXGLtYe7D/VBHal/zoWqs2Q5lbD6onNtW5bPFNbeY8BcCEGhPgAXA9uwYCQEIQGDtBGIHb1UHe2t3GAeyCYQnINkDWwj2MWdi/vqTqVh/C1dbDZHtPmJsZXygQeuOZ932+8Bq81Ox+L8+dMd0dGGWOzZXLuZfTlsf+vvQkeNr3XdOzI9YW44tZMYjYHm1z+94lrEEAQhAYP4EuAV4/jkkAghAYMEEck9GQjnt20HyuvGEfoS+5vhnOtqMzdHfVcb8kx7zsY8cSK/pa+qj96npWMnb+NC+tcd0VvXF5MdqC2MYy66349nE/FG/tVup8VXjTM7LeJuxeihrOmKyU2vzjPr0zZiMwcLbMLuKJSc2G+vHdeHQVp+NM79DH3y/+sL9UL7LfopbHSPzSaXJWt36uvjFWAhMiYDN6dTnZUq+4gsEIDAsAa4AHJYv2iEAAQisjYAd8K3NgS3DdmLVlw996+vLryH19BVzk/nQl80hucxJd4yn2qw9LGOxNclfbHyqzWyn+jehXWyH4ts3P/LVjOh+++0XHSCOIUvbt1IDfT2qiMZGBIbgOYTORkHNSHgu33MzQoqrEJgdARYAZ5cyHIYABCCQT8BObKdy0Of90EG7/eVG5MfnjhlDbizOIa9w38fqfVJdsmzrIWC5MOvKhbVZ3frC0uTC9j7zKV196vO+ttFbNaaqz9ttUs/RKZkcuSZ228hqPqS2qr7UmKbtxsHKcHyfPnThHfvcxBYDUzZ8fFZPyYYMNnnfWDVlEMtXEx1mNzdHVfasz8omfiALAQhAYOoEWACceobwDwIQgEBPBHQwu+4tdZBu7WP4l3uCMIYvTWzE/F7HCUpbm33PvxiPJjzHlg391b7++uDSNichg9DHsH8q+8bOyqZ+heP6yEFTHyS/LrttfO17TO5c83JdecVeCNLXZ6dvPnPU53OV8t94h6XJp3Js8iZXVVb5oUVgbyPU6/vMRqzN+ighAAEIzI0AC4Bzyxj+QgACEJgQgaoD7TZupvSFB+ltdZt+lVZvo2sdY9qehHSNNcWpiT9NZD3blO/mk5V+zNzqYYy5McWYxtqa8si130RvU51N42iqv4nvY8quI46+beboMxkr6xhXyTWdKylbfelJ6V96u89RF5Zej5hJl9fn60tnSnwQgAAEhiDAAuAQVNEJAQhAYEEEdEDu/yy08EDd2q2s6ze5PkrvX+wqjz5sNNXRZ/ymKzwZaupTnbxxDOXMftie40+bE7bQD9u30vuR8s3LTKHehIPFGZYWh2+3tj7LOqZ1/TFf2oyJ6em7rUleZHsKcTT1OZeZ9Mb+csebnBilOKXabayVdXLmp8m3LevshHolb39hn99vqtePnVM9ZNF2boZ6xCClK9XelVsTvTF/u9pnPAQgAIExCLAAOAZlbEAAAhCYMQEdFPs/C8XabJ9yWAJDnlB2PZlJnTil2qtINY2zjY0q+0P2dfW1Lk9N2VXF2qeurnbErY5drr+5cl19rho/xb4qvlV9VbHksJZMKBfux2zExsXkcttybIa62owJdcx9388NX28SV4xjW11N7HrZNvbajPE2qUMAAhBYB4E91mEUmxCAAAQgsHwCOQfHsQP/pmSqdOhqwD333DOpUmNz/EwqGLkjjDXmeyjT1cWYjZjOXLnYWLX17XfKTtges9s1ltCG35fumE0v06Tep67QrnTHWMTaNLbOl1R/zE7MRp/sYvrD+Kv2LRbTY/vhGOsP24fYNx+a2MyVNTmzkeu/5G1s1Zg6van+mH4ve+655xannXZalenaeVs5OOj0toOuxe7m5LdJ8Dn6JJNinTO+iT8x2TFsxOzSBgEIQKArAa4A7EqQ8RCAAAQgMGsCqZOILkENoTPmj+yEfzG5tm1TP8mZun9tuU9p3FhzeUoxN/HFPn+pMUPwq9NZ12++5sp5+bl/5uT/3GOwfOSUTXOco3NIma656Tp+yNjQDQEIQGAKBLgCcApZwAcIQAACG0qgj4N16ZjKSc5U/JjjdMrJY45MH7FbHvuYn7n+VNkyf0yXyYbt1t+07INrF18snjq/UzZyx9fpH6rf/B7TT7NZFVNTf0zeyjY2NDYcZ/rka9hX5b/15Y7RG2BtM5u5Y8Nxth+WpjdsH3vfx+Xr5sdU/DR/2pQWg8Vn+210MQYCEIDAphBgAXBTMk2cEIAABCDQCwE72ZAyf8Lh23sx5JR4O2oe0pYzO0hVvofx5BiyMVbmjGkrIxtTYmz+9B2712c2YsyqctYHpyr95k/Kjo/BZKvKpvJVupr25cRZpzPFoW7cVPvDfPj9pcW67hyI7RKZ+jkzNGPjN6bNoWNCPwQgsFkEuAV4s/JNtBCAAAQWSUAH4/bnr/JQ3dpTZRcgOhmwvy56mo71sTQdmysvG31vdvKU0puymWpP6enaLj+HslnHIOV7zJ9YW2p813ab52HZVa+Nr+LSts90dy2r7DfVXccvtFUnP+YcaBpryje1p/qa2mgrb1xzx0/JZ/O9qrS4zG/jbfsaO5XNfDIfp+JX6MeUmIW+sQ8BCEAglwBXAOaSQg4CEIAABBZJoOlJR0x+aScGsRjHSr5se57el1T7WL4tzY7nOYXY5I/P9xR8WocPxqEuP8Yq/Mzk+mx2cuW9XJ1vXjZVN/9T/UO0V/lt/ngZaxvCl6Y6m+a5ynfTVSXT1D/kIQABCEBg+gRYAJx+jvAQAhCAAAQ2nEDVibqdyIWIwhM7f1Ibyo61H/qUshvKxXyvYpLSG7bH9JqM7wv9MRmVuXI2RrqajrGxsTLUF5Np2+bj9j631ZczTnbWYTfHt1yZPnJSx9szCv2K9aX0hbxDXV32vc2YTyndflxKpk273gpftRmLJr5W6euzrw0TiyflxxTjTPlKOwQgAAEI9EOAW4D74YgWCEAAAhDYYAI6ker6V4evyQlg3Yld6Gtou4ktG6sxbcbZ+FRZpbOqL6XPt9dx8rIWX1iGMn7f10NfLQdepm3ddMXKLjr9WNNtbeG+tfsyR8bLW90ztrYhyzp7isM2i8mXYZ/2/Rjr76M0uyldKbt141L6+mr3jMO6t6G+PjaL1z8SIkdvX/ZzbMVkjE2sj7b1EVj3vFhf5FiGAASWRoArAJeWUeKBAAQgAIFZEkiduNcFk3tiIv25snU2U/1tY0jpy2lXTF3sVnHpotd8N+Zel6+b3FCl2TI/utoxfabH9qv058iYvimVVTF5Py0+3zZEvaud3HiG8B2dEJgrAT43c80cfkMAAjECXAEYo0IbBCAAAQhAYCYEtChgf+Zy6oQllDN5a/el9Y1Ryl/78/ZScXgZ1XPlwnG2r7ht68LA65E+71cqPi9jPgxRNokrjCPHHz/GbOWMm5pMXT7q+hVPLNdDx+n5x2xV+W3+Vsl4nblyfkxV3exXybTtq7rt1+ZpWLa1tc5xYQy238QnnwerW9lEz5JkFb828WSDAAQgsAQCXAG4hCwSAwQgAAEIQMARWMfJSl827YTLhTNKtav/ueMVn2R9nNY2SqADGqljEMY9oCvZqn0ebJDlo62/bceZ/b7LWIwpGxZ7qn/I9iZ+VvlRNw81NkemysZQfeZXExY2pi+fYrbDtr5t9uX7EHosViuHsIFOCEAAAmMRYAFwLNLYgQAEIAABCAxMYI4nKOGJZQ4ii7PN2Bz9MZk+bJqOUH+qPZTra9/siZ/qxtHa+7IT02M2zGZMZqy2HB9i/hq3sfyUHfOjzmZOTHU6Yv1D6Y3ZatsWMpLP5557blt1sxgXxtzFaa9rDvnuEms4VvH6+K0/1mZ9lBCAAATmSIBbgOeYNXyGAAQgAAEIQKA8YdMJWniSVnXb3zqx5ZxU58j0FYNxizHsy0ZKT8qm9yk1to/2oTlbfBZPHz5X6RgyniF1V8XUpC/kPAefY/HJ7zrfw1hjerq2jWGjq499jt+0ePtkhy4IQGBeBFgAnFe+8BYCEIAABCAwOwJ2UmulBVB3omtyVuokLedErembP01/Thn6EO7X6TAGNi6MJ9wP9TVlFo5vsu9tmd++rYmulKznYLH70up+fDjG++brfiFY9S6+hza9P03rpit3XIxB7thQzmyHZShn+12YmY6qMtcP0xHK2771W6n2OW45fltOcmS7MDC2VnbRxVgIQAACEJgGAW4BnkYe8AICEIAABCCwSAJ2suqDU9vQJ6/e3lLqxnJMfmbTMxzTvrerORPzx8vk1vvUVWWzq799fk6qdPm+rj5X8fB9ZnMse962bJp9377O+rp8CvlPjcs6c4JtCEAAAksjwBWAS8so8UAAAhCAAAQmREAnk7E/uRhrr2obKiydANtfjo3whDkc00RXODa2n9JnflgZG9ulzXLhdcTafP8QdW9TddtSXKw/VtqYOmZ1/THdsTbvb9hf1RfKdt0f25bs2Z/5bvtWqr0JZ41rskneXw3s6011NbHbRNbmox9jfHzp+/usx/jH2kKbOTLhGPYhAAEIQGD9BLgCcP05wAMIQAACEIAABCZCQCe2YywO2Al0jq0qmbH8VXqq/Bg7fcZvbLt92GuSM4tT7K3ehw9D6UjNkVT7UH6k9E7FD/PP57XKNy9nY9uWOfMoZ47myLT1kXEQgAAEIDAMARYAh+GKVghAAAIQgAAEZkKg6sS7SQg5J9amrw+btigw9Il4H75a3F3LJoy72mozvmkuLId1tqYQd9N54H1uyqWOx5L6m3JtKi9WPhd9sjO9bXzq0w90QQACEIBAHgEWAPM4IQUBCEAAAhCAAAS2EdBJr50Ab+tI7AxxkjyEzoT7a2uuYtw0B7lBtOWaMy6UGSqG3Fhz5EKf68bEcqa2pnrq7Cy53zM0blb6uE3O93nW1u/H5NRtnOlVmdLr26t0m07JmN4qefogAAEIQKBfAiwA9ssTbRCAAAQgAAEILJyAP3H19TBsf7Ib9rHfD4EhGFfltKvXQ+ru6tsY48OFIs/D+mI59XJj+LluGyEDY5Pjlx9rdfGzuukwpmG79VupfpO1Un2hTi/nx1o9LGPyoQz7EIAABCDQLwEWAPvliTYIQAACEIAABCBQEvAnyzEksRPvujFejx/fZJzXQX07gaE41ulVv8/ndq+675nuOj+6W2qvYcq+tY+qv5HKYZN5YvLmQYqvtYfyNi63ND258shBAAIQgMD4BHgL8PjMsQgBCEAAAhCAAAQaE9AJuv9rrGBBA7TYEP7lhufHaUy4n6snlFNuwi3WFspMZd/m1lT82WQ/NCfDLdZmMlV9JmNlbE7G2kw+p4zZ76ozx26VDPO5ig59EIDAphLgCsBNzTxxQwACEIAABCCwVgKxk+Yqh2LydpKtMtYf09dENjZ+7DbF1ZfPIaNwv01slgON9fWmuizOpuOayLfxrw2jMWJpEvccZXO5h3LhvsXuc291k7XSZHNK05EjiwwEIAABCEyDAFcATiMPeAEBCEAAAhCAAAQGJaATdv21Odkf1LEK5U0WGaYaVxO/JNtEvgLdLLua5HuWAa7BafvcD2W6y3wl30NlBb0QgAAE4gRYAIxzoRUCEIAABCAAAQhMnoAtGOWchJvs5INyDlpcVrquVdUWOHIWE3JkVopdxdsI605scdUq7k2D7TL/2uYt5eOcc+hz4uthrBajb7ccWOn7mtardPSdr6a+Sb7Kvzb6GAMBCEBgCQRYAFxCFokBAhCAAAQgAAEILJSAX0zw9TbhalGg6dbFZht7Tf0bSr6r7368rw/lb53er3zlK6tnaIayXXIc6hpjXzybMm0qP3QcbWIY2if0QwACEFg6ARYAl55h4oMABCAAAQhAAAIzJmALF7ZIY+UcQpqTr8azz4WZPnUZS5X2Zz6nSi36beImPva5WXf8NgfCct1+YR8CEIDAJhJgAXATs07MEIAABCAAAQhAYAEEbFEoN5Sm8rl6ketOICc3OTJNPbGFKT9uCDtefx/1Kh/Xvfi3bvt98EUHBCAAgSUSYAFwiVklJghAAAIQgAAEILAwArZQ4xcXfFtY32+//aIEqhZOogM6Nsre2Da7uGz+WtlFV5OxPq+pcWGOU3I57aYrJdtXzoyjlSl7vt3b9nUvo3pVn/rrmMbGqy38k66mWx3fpvqQhwAEIACB7gRYAOzOEA0QgAAEIAABCEAAAhMmoMVAW5CoWxRZdxixRZl1+mSLQWP4UJWbqr6Ub34R2NfrdKm/TiZls6q9ic66eZCrq0pOfZZfK6v8pw8CEIAABOZNgAXAeecP7yEAAQhAAAIQgAAEBiJQtwgzkNlJqh2LRbhgpf2wrS2gPnXl+OCZNYnB+6m615NjN5RpYjscu7T9riyXxoN4IACBzSKwx2aFS7QQgAAEIAABCEAAAhDIIxAuvqQWUuoWFVLjQi/q9ITyS93P5TX1+MP509bfFI9Ue64dP99Ml28zPWqzfmuba7mUOObKH78hAIH1EmABcL38sQ4BCEAAAhCAAAQgMHMCtqgQWzwJQ/OLKTny4Xj250fA5sf8PMdjCEAAAhBYEgEWAJeUTWKBAAQgAAEIQAACEOiVQJPFmzpZW/CzsldHOyrzvqf8U7uX62iy83Dv55T88oFN1S/vY119CTHUxUg/BCAAgU0gwALgJmSZGCEAAQhAAAIQgAAE1krAL1b14YgWZdrqtAUdG2/7OX7ZmJRsE10pHTntdX7k6NhkGc+vLmeSrZPZZJbEDgEIQGAuBFgAnEum8BMCEIAABCAAAQhAYLYE+lpA8Qs3bWB4P3zd61J7VzteH/VixTPFfGxGqRybfyz6jZ0R7EEAAhAYngALgMMzxgIEIAABCEAAAhCAAAR6IWALNFLm622UxxZ55rjw15VDG3ZLHwPTpWeY+CAAgU0kwALgJmadmCEAAQhAAAIQgAAENp5AbJEn1pYLaszFQ/k5pr1cBim5LlxTOru2T9GnrjExHgIQgAAE0gRYAEyzoQcCEIAABCAAAQhAAAIQmCgBFrAmmhjcggAEIACBSRJgAXCSacEpCEAAAhCAAAQgAAEIzIvA3Bbk/BWEc/N9XjMDbyEAAQhAYAoEdpuCE/gAAQhAAAIQgAAEIAABCEBgDAJa+POLf7IZ7o/hBzYgAAEIQAACYxJgAXBM2tiCAAQgAAEIQAACEIAABCAAAQhAAAIQgMDIBFgAHBk45iAAAQhAAAIQgAAEIACBaRHgFuBp5QNvIAABCECgfwIsAPbPFI0QgAAEIAABCEAAAhCAAARmQaCP25/70DELWDgJAQhAYMYEzrf1Zb1rxv7jOgS2ETj11FNXz3D59re/XZx33nnb+tlZPoHdd9+92HvvvctAmQPLz3csQuZAjMpmtTEHNivfqWiZBykym9POHNicXKciZQ6kyGxO+1zngM5n5LuuTj744IM3J2FEOigBFgAHxYvysQmccsopY5vEHgQgAAEIQAACEIAABCAAAQhAYBAC17ve9QbRi9LNI7DH5oVMxEsmsNtuuxU///nPy19K9tiD6b3kXBMbBCAAAQhAAAIQgAAEIACBJRL42c9+Vt7ZpvNbNgj0RYArAPsiiR4IQAACEIAABCAAAQhAAAIQgAAEIAABCEyQAMvJE0wKLkEAAhCAAAQgAAEIQAACEIAABCAAAQhAoC8CLAD2RRI9EIAABCAAAQhAAAIQgAAEIAABCEAAAhCYIAEWACeYFFyCAAQgAAEIQAACEIAABCAAAQhAAAIQgEBfBFgA7IskeiAAAQhAAAIQgAAEIAABCEAAAhCAAAQgMEECLABOMCm4BAEIQAACEIAABCAAAQhAAAIQgAAEIACBvgiwANgXSfRAAAIQgAAEIAABCEAAAhCAAAQgAAEIQGCCBFgAnGBScAkCEIAABCAAAQhAAAIQgAAEIAABCEAAAn0RYAGwL5LogQAEIAABCEAAAhCAAAQgAAEIQAACEIDABAnsMUGfcAkCrQh87WtfK17xilcU73nPe4pvfOMbxZ577lnst99+xW1ve9viHve4R3HBC16wlV4GDUPgtNNOK9773vcWp556anHGGWcU3/3ud4vzn//8xb777lscfPDBxV3ucpfi+te/frZx6XrVq15VSK90XepSlyoOOuig4q53vWtxyCGHZOn52c9+VvzLv/xL8a//+q/FF77wheLHP/5x6c/v/M7vFPe85z2Lq171qll6EOpO4MQTTyxe+MIXrhS9/OUvL37rt35rtR+rMAdiVObX9vWvf7149atfXX6Xq/6jH/2o/Dz/2q/9WjkH9J3+67/+68nAmAdJNJPvOPfcc4vXv/71xVve8pbis5/9bPH9739/2/8LRx55ZPn/Q10gzIE6QuP2f+c73yk+8YlPlH/6P1p/yq22O9/5zsUJJ5zQyKEp5VfHGzr2fMc73lHoOFSbvqtucYtbFPe6172KS17yko1iW7JwH/PgJz/5SfG+972v+MAHPlB88pOfLL785S+Xx2oXuchFigMOOKC4yU1uUvzBH/xBsc8++2ShlL6///u/L79zvvKVrxT6DrrsZS9b3OxmNyuP+5TLnI1zkBxKRdHHHEhZUi5vf/vbF1/96ldLEeXuXe96V0p81c4cWKGgsiEEzrdra9uQWAlzwQT0Bf+IRzyiOOuss6JR6qDg+c9/frH//vtH+2kcl4AWZD/60Y/WGr3Tne5UPP7xjy8Xc1PCP//5z4u/+Iu/KBcMUjI6afyrv/qrYrfd0hc96yD+mGOOKU9MYnq0oHz88ccX0sU2LIHPfOYzxRFHHFFoQda2qgVA5oBRmn+pE+mnPOUp5QldKhqdVD/mMY/Z0c082IFkVg06gb7//e9ffO5zn6v0Wz/GKP/nO9/5dsgxB3YgmUTD1a52taQfTRYAp5bf//qv/yqOPfbY4lvf+lY0Pi1CPfvZzy6ufe1rR/s3rbHrPDj99NOLu9/97pX/P4ipFgN17Hi7292uEvGZZ55ZHvd96UtfispJz0knnVQceuih0X5r5BzESNSXXedAlYUnPelJxYtf/OKVSM4CIHNghYvKBhHgCsANSvZSQ/30pz9dHHfcccXZZ59dXOhCFypPIHSlkPbf9KY3lVeF6T93Le6cfPLJ5YHBUlnMJa5vfvObpau62u82t7lNeaXf5S53uUIH9x//+MfL/8D/93//t3jd615XLgI9+clPTob2d3/3d6vFv2te85rFfe973/LKT/2SqyvIND90VZ+uCHzoQx8a1XPeeecVD3zgA1eLf7e61a3Khb5LXOIShQ7wn/Oc55S/WmoBUD7nXlEYNUZjJQE7wdPi36UvfemSe+WArU7mQB2hefTrRPlpT3ta6ax+tNHVu7qK96IXvWh5tZA+y29/+9uTC/nMg3nkOeblT3/6022LfzpJvM997lMceOCB5RWgp5xySvGSl7ykPPHXIrG+h/V/ergxB0Ii09u//OUvX1zpSlcq3v/+9zd2bkr5/Z//+Z/iAQ94QHnHwR577FH80R/90Wqh6N3vfnfx0pe+tFwYlMxrXvOa8qqyxgEveECbeaAf+XVnhjbdKaKFud/4jd8odKymH3Hf9ra3lcd7knv4wx9eXPjCF04er0lG3yG2+Kf/b7RgeIELXKD4z//8z+J5z3teeVGBzi/+6Z/+qbjGNa4RzQbnIFEsWY1t5kBKsfLwspe9rNhrr70KfR5150DdxhyoI0T/YgnoCkA2CMyZwFFHHbVr63awXVuLP7u2bifdEcoLXvCCsl8yT3/603f00zA+ga2Drl3/9m//tmtrkSdqfOsWgV1bi3CrvH34wx+Oym3dplvmXbk9/PDDd21dxr9NbutAsWy3+bF1oLet33a2FghXth73uMdZ86rUuK2DzVLmlre85a6tk9VVH5V+CWyd5JectxaGd20t/K7y8qEPfShqiDkQxTK7xg9+8IOrXD/ykY/ctXUbVjKGc845Z0cf82AHklk1vPnNb17l/253u1v0/4at20Z3Xeta1yrlth4PseN7mDkw3ZRvLezv2rpKatfWlXKlk1s/0K3y/ahHPSrL8anld+uuk1UMWz8274hBxzg69tBfbow7lCysoes82PohYNef/umf7tq6SjhJZutHol1bPyCU3Lduw9619aNiVPapT33qKj86Twg32dJ5hfL3h3/4h2H3ap9zkBWKrErXORAzonOJrSuJy1w985nP3LW1MFzWVVZtzIEqOvQtmUD6frjFLnkS2JII6Jkydiupnhn3m7/5mzvCO/roo4srX/nKZbtuI9SVBmzrJaBfVvVL6+677x51RFfrPfrRj171vfWtb13VfUW/9tltoroNWL/c+k3PfVS7NsnpF/nYZrcM6FfkrcWHHSK6dVy3pmnT7QK6ComtfwJ63ptdAfaXf/mX5bO/6qwwB+oITb9fV31uLbyXjl796lcvnvCEJ1TmXrfjhxvzICQyr/2PfexjK4d1VU7s/wZd6aPncmn7wQ9+UHz+858v6/YPc8BITK988IMfXF6ttffee7d2bkr51S2/elawNj1zTs8lDTcd46hPm55rmbpNOBy35P2u80BX/W0t2hRXucpVkpj07EXdxaFNzwfUlWHhpvMAXUmsTecHOk8IN3sWtdq3foQun18ZynAOEhKp3+86B2IWdG73qU99qrxi/H73u19MZEcbc2AHEho2iAALgBuU7CWGqocu26YFwNim577pWXLadNKgS/vZpk/Av/BBB3HhtvXLTPHOd76zbNbtRNe97nVDkXJf7bqNTJvkNc5vX/ziF1cnkrodOfWyGD2nyDY/76yNsjsBPadRt/eI9Q1veMNahcyBWkSzENCtgFtX2Za+6uBdt+802ZgHTWhNU9b/MKeXd6U23+fHMAdSxJbRPrX86plv+uFCW+rYU31bdyaoKGU1hm0cAnXHjzoP+OEPf1g6o/OD1POhLX8SjB33+bbUPOAcZNic69mxW3d3lUb0w3HsB8KYB8yBGBXaNoUAC4CbkumFxqnnAmnTs/+2bg1KRnmDG9xg1ae3zrJNn4DexGZb7OBMb/myZwn6/NoYX9pikp4raG8Hs36bQ9o3OevzpR7mreeSaWMOlRh6/UfP69Rzk1JXYcaMMQdiVObXpje+atNLHewKL+3rLaFaGLS3haottjEPYlTm1WY/0shrPb81tVmf5soBv/w+lixzIEVsGe1Ty68/bqg6/vB9HDeMNxf98WPsamKfv6rjPl11bD8Kx/JnejgHGS+3oSUt+umH4zve8Y6FX/gN5cJ9y53amQMhHfaXToAFwKVneOHx2S1AV7ziFSuvGtEVYrbZGNunnCaBj3zkIyvH7BbuVcNW5Ywzzljt+vyuGl3F9289R8j1FKur/9To5bYJ/XLH+vXwb3sQdUyOtmYEdGXuE5/4xHKQHtytW8BzNuZADqXpy+hFO9r0xj69dVG31t3hDncoD+Zvfetbr8oXvehFhT+xs8iYB0ZivuXv/d7vrV7QtfU8rkIvZgo33cr3nve8p2y+/e1vv5JXA3OgxLLYf6aWX/NHLyjSj4OpTS+r0XeaNo49U5T6b9ctu7bFjh99Luy4zuR9qavRdX6hzY8xGWvjHMSIjFtuPWezeO9731tc/OIX3/bYoBwvLHeSZQ7kEENmSQRYAFxSNjcslq0HwRff+973yqgve9nLVkav/xz0C522b3zjG5WydK6fgG6tef7zn79yJPZ8HZ/Huvz7fi3e+c3rucxlLuO7dtT1pmJtuh3Jj9shSEMjAieeeGL5fCQ9c+eII47IHutz4HMcU+D7mQMxQutp02fdFuUveclLFn/9139dvr3xv//7v7c5pCsB//Zv/7a4173uVT7KwXcyDzyNeda16K/86mobXWmj7wG9BV5vhd96QUyx9WD3YutB/OUzfHW1v39GrCJmDswz77leTy2/uptAm/9/JRWLHTf4GFKytHcncPrpp5eLQtK09QKP1TPAvWbLhc4LLnaxi/muHXXLn94y7H+A4hxkB6pRG/7v//5v9cPxwx72sOwfjs1J5oCRoNxEAs0etLOJhIh5sgT8K95tca/KWZ1Y6KotrtyqojSNPr2sQw9X1qaHOes2jHBrkn+7hUM6wvx7PRe+8IVDM9v2q/RsE2Qnm4Be4rP1FubyCl7dyqFb+3I3n7u674Cq3Hk9zIFc+v3I6TlM9iwtLfptvem1vKJGL+M55JBDir322qtsO+mkk8rFIL0s4s///M/LBSHzwOePeWBU5lcedthhxcknn1xsvQm8ePWrX11svTl1WxB6gcTWG0CLu971rqvb8kyAOWAklllOLb/mT933jbJh//fYmGVmaBpRaYHuMY95zOoK4uOOOy7qmOWiSf6kSOPsGXOmQ+25ejgHEa1+Nv1g9O1vf7t8+aP+T2i6Wf5yc2f6mQNGgnLOBLgCcM7Z23Df9eubbec///mtmiztP+2zzz47KUPH+gno1o0nP/nJpSOXvvSlV28HDT1rkn/LvXSE+e9LT+gf+/UEdLCutzTrisp73/ve5a/19aN+JdFX7vrS8yvPqOUS+MlPfrISVR50sqw3+v3+7/9+eVuP3uyt52jpDaB6Q7A2vYXbbhvWfl/560uPfGJrTkDfB3pbauxlTdKmk703vOEN5RWBofa+cteXntA/9rsR6CsvfetpcuzpbXejwegUAb1I7JOf/GTZrZeJ3fzmN4+KWi6a5E+KbFxYb6InPAaNOkhjJQE9Ikg/FukW7cc97nGNfjg2xZbLJrnTWBsX1pvoYQ5YFijXRYAFwHWRx25nAroyxDb/NkBrC0u7dF8nlGzTJPC5z32ueOADH1j87Gc/K6/8edrTnlZoETC2Ncm/5V56wvz3pSfmI23VBJ73vOeVt39e/vKXL/NeLb2zt6/c9aVnp4e01BHwi/OS1a2fsefx6HPrr+bQS2Ns6yt/fekxvyjzCejKmPvc5z6FvhN0a9d973vfQjnWFaF6WPuLX/zi4nrXu155cn/ssceWVwl67X3lri893jfq3Qn0lZe+9TQ59vS2uxNBQ0hA3x26m0DbQQcdVBx//PGhyGrfctEkfxps48J6Ez3hMejKKSpZBPwPx3okiP0wmDXYCVkum+ROw21cWG+ihzngEkF1LQRYAFwLdoz2QcDfqhfe1hnTb1ea5FzuHRtP27AE9HbHo48+ujz501vbnvKUp5RX/qSsNsm/5V66wvx7PXZLQMpmlZ7UGNrjBPQAZh2wa3vsYx+7Iy/xUdtbfe7qvgOqcuf1MAe2Mx56zx6Qb3ZucpObWHVHeaMb3Wj1sictDNnm88c8MCrzKp/xjGcUehyAtic84QnFIx7xiPLZXVog1hy58Y1vXF4Zqrc86oph3f6lZ33ZxhwwEsssp5Zf86fu+0bZsP97bMwyM7TeqF75yleWx4zyQj8g6RnS4bGe99By0SR/Gm/jwnoTPVV+eR+pxwk85znPKb74xS8Wejbjgx70oLhQRqvlsknupNbGhfUmepgDGQlCZFACPANwULwoH5KAfoW5xCUuUXz/+9/f9gDwmE1dUWBfzjkPbY7poG04Anqgtq7++OY3v1leyq83wt7iFreoNOjzaA/zTQ3w/fZAZ5P1euRH1Rto7eURek6dH2e6KPMJ6JZO/WK63377lbdl621u4aYrQm370Ic+VN4CqP1DDz20PLj3OfA5tjG+9P3MAU9mvXUt8Ogzpwesa/M5DT3Td75eFPKtb31rJR+O8XkOx2vf9zMPYoTGb9OC3mte85rS8AEHHFDo1r3Yptu99AzAo446qnxupMboeZDa/LzxOY7p8f3MgRih6bVNLb96YZhuSfdzKUXNjht8DClZ2psTeOMb31jo+cHa9CZ5PUO06jhOcsqFHiOh84If/OAHlS8CsfxJp79inXMQkRx/01vitekHwXe/+91RB+x8T6UdWyp/GmMbc8BIUG4iARYANzHrC4r5Kle5SnnVwJe//OXytlGdIMQ2e8uk+q585SvHRGhbEwGd+OvKP10BqE3PhLvTne5U641yb5vPr7X50veHtxf6+SC5a1zjGn7otrrp0Ukjv+BtQ9N4x27LVt4f+tCH1o5/9rOfvZLRM8LEnzmwQjLrivKoZ39qsxeCpAI677zzyi7/Xc88SNGaR7sWUvRDnrZrXvOalU77F0LZ97EGMAcqsc2+c2r5lT+f+tSnCr3ESD9I7LPPPlHG+lHzrLPOKvv8sUZUmMbGBHQsoJcF6f8N5UAvkMtZaPW50PfIda973ahtPY7Gjk39GBPWPNCVy5yDGJHhS7vVVj8A2Q9HKavf+973VseXN7zhDbctAPp8MgdSBGlfKgFuAV5qZjckLj0TSJt+5dHBWGrTA2NtO/jgg61KuWYCOnjWs57OOOOM0pOHPexhxT3ucY8sr65whSsU++67bynr8xsbbP361V7j/GZzSG22COH7ra6D/C996UvlLnPIqKy3ZA6sl39f1vWSD9vsZMv2fakTaR3Qa9Nn2TbmgZGYZ6lHPthmC7y2H5Z28qd2vwjMHAhJLWt/avn1xw12fBEj7vs4bogRat/2H//xH8VDHvKQ8sd/3Q2kK/+ueMUrZin0+as67qCR7LsAAA+8SURBVNMLRexqslj+TA/nIFnYJyVkuZNTzIFJpQZnRiDAAuAIkDExHAF/m6jeCBXb9Mvg6173urLrYhe7WKFnCLGtn4Cei3PMMcesFm4f8IAHlPu5nuk23MMOO6wU1693H//4x6ND1W5Xikhe4/x24IEHrq4Kfctb3rJ6Xo+XUf21r33tqsnPu1UjlUYETjjhhOKzn/1s5Z9eCGOb3gxr8joZ1MYcMDrzLm91q1utAtAbflOb+nS7qDZ/8M48SBGbR7tO3u1ZkB/72MfKE/qU535Bxb4HJMscSBFbRvvU8qu3y+622y9OoVLHniJvVyhJNvVG2mVkaNwoTj311OJP/uRPCt1JcNGLXrR40YteVFz1qlfNdkJXg2mcNp0f2P8roQLLn9pjx32+LTUPOAcJqXbbt+PAqlK3gmtTaXKveMUrthlmDmzDwc6GEWABcMMSvrRwr33taxfXv/71y7D0n69OHsJNbw/UCwe06Y1ROa9qD3Ww3y8BHbRpcUcHcdqUF/+Gz1xr9773vQu7euTxj398+Sw5P/bss88u1K5NV4tIPrbpFmRtug3txBNP3CGi2zvshRX7779/cctb3nKHDA3rIcAcWA/3Pq3qLX43velNS5V6Xo+u7Ag3XYH71Kc+tWzWd/hd7nKXbSLMg204ZrWjxZGb3exmpc+6ZfK5z31u1H89y/ekk05a9dkYa2AOGIllllPKr243vcMd7lCCfv/731/ox8Nwe/Ob31yoT9sd73jH5G3C4Tj2qwl85jOfKe5///uXV+bpUSA6NvOPBqge/YtePcvvnve8Z7mj8wMtIIabzidsUU+LRTrfCDfOQUIi89lnDswnV3jaP4HdH7e19a8WjRAYj4Ce2fb617++/CXwTW96U3klgBZ7dLum3gSmP216uPjf/M3fbHuI73heYskT0G0b9vDe3/7t3y4P5vQswO985zvRPz2kWQ//Dze1aZHvlFNOKV8g8u///u/lw5ytTQ+It1vD73e/+xW3ve1tQxXl/tWudrVy0UEPe9bbRfXyCV2RohNOXXWkZ8yorhNVvX0yfI5gVCmNnQnotgy7NUMvBvBX/Jhy5oCRmHd50EEHFW94wxvKz7NOnHVLlQ7Q9Z3wjne8o/wM2gP39cxIWzC0qJkHRmKepZ6lpatt9Mwtfeb1va2FXv1Y9NWvfrV461vfWvzZn/1ZceaZZ5YB6mHuD37wg7cFyxzYhmNSO3pOml7kpMUb/enWyg984AOlj1rE0f+t1qcy9izeqeVXi076ztLdDPqOUqnvrK9//evFP/7jP5bHCrr6Sy8f0I8XdpXrpBIzsjNd54F+jNUPxvbM0Ic//OHl8z9Tx45q1xZ7ZrPy97a3va3U9cEPfrB8luMFL3jB8uUuOqfQ86jPOeec4gIXuEDxtKc9LbmAyzlIs0nUdQ7kWNNL5vSIId31pR8OUhtzIEWG9qUTON/WZc+/uJ9m6ZES36IJvOtd7yoe8YhHrB62HAarxT8tBOrqLbb1E9CCW5NNl/Erx7FNB9iPfexjV7/UxmSOOOKI8kpAnWSkNi1A6pZkLQDGNh3YH3/88cWRRx4Z66ZtAALPeMYzimc+85mlZt0CnLp9nzkwAPw1qNSJgd7yqpdCxDbdBqhHBegHhNjGPIhRmU+bTsK1uGvPeUx5rh+Nnv70pxcXv/jFd4gwB3YgmUTDox/96G2P0ahzSrftxbap5Vdvkj322GPLxaOYv7pS8FnPelZxnetcJ9a9cW1d54F+JNAPAU023W3yoAc9KDpEPyjouM+e7xwKadFWVx0feuihYde2fc5BtuGo3Ok6ByqV/7JTt9t/7WtfK28BTp07mB7mgJGg3CQCXAG4SdlecKx6jtvtb3/78uo//TKoX+30i5+eCaJfC5/4xCcWe++994IJzCs0W9TJ9brqVzx7NpBuxdBVQz/60Y/Kq0aUb50o6mBRt/hKrmrTL7+HH354+SuvXjagX/P1+8jlL3/54ja3uU05h373d3+3SgV9PRPIuQJQJpkDPYNfkzp91vQZ1BUXuupXV/Iqt3rrtp61pO9xu+0u5iLzIEZlPm377bdfeWu3ngmoKwH1HaxSP77ozZ43vvGNy0dFaJFQcyS2MQdiVNbfpivkTj/99GxHUgs2U8uv5uWd7nSnco7qO0tzVleu6pj0rne9a/GkJz2pOGDrB2i2XxDoOg90daje/Ntk0+27qR8P9V2jx0noeYD2f45+KNZ3kfKqZxXHbv0N7XMOEhJJ73edA2nNv+rJvQJQI5gDv+JGbXMIcAXg5uSaSCEAAQhAAAIQgAAEIAABCEAAAhCAAAQ2kED6frgNhEHIEIAABCAAAQhAAAIQgAAEIAABCEAAAhBYGgEWAJeWUeKBAAQgAAEIQAACEIAABCAAAQhAAAIQgIAjwAKgg0EVAhCAAAQgAAEIQAACEIAABCAAAQhAAAJLI8AC4NIySjwQgAAEIAABCEAAAhCAAAQgAAEIQAACEHAEWAB0MKhCAAIQgAAEIAABCEAAAhCAAAQgAAEIQGBpBFgAXFpGiQcCEIAABCAAAQhAAAIQgAAEIAABCEAAAo4AC4AOBlUIQAACEIAABCAAAQhAAAIQgAAEIAABCCyNAAuAS8so8UAAAhCAAAQgAAEIQAACEIAABCAAAQhAwBFgAdDBoAoBCEAAAhCAAAQgAAEIQAACEIAABCAAgaURYAFwaRklHghAAAIQgAAEIAABCEAAAhCAAAQgAAEIOAIsADoYVCEAAQhAAAIQgAAEIAABCEAAAhCAAAQgsDQCLAAuLaPEAwEIQAACEIAABCAAAQhAAAIQgAAEIAABR4AFQAeDKgQgAAEIQAACEIAABCAAAQhAAAIQgAAElkaABcClZZR4IAABCEAAAhCAAAQgAAEIQAACEIAABCDgCLAA6GBQhQAEIAABCEAAAhCAAAQgAAEIQAACEIDA0giwALi0jBIPBCAAAQhAAAIQgAAEIAABCEAAAhCAAAQcARYAHQyqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBYGgEWAJeWUeKBAAQgAAEIQAACEIAABCAAAQhAAAIQgIAjwAKgg0EVAhCAAAQgAAEIQAACEIAABCAAAQhAAAJLI8AC4NIySjwQgAAEIAABCEAAAhCAAAQgAAEIQAACEHAEWAB0MKhCAAIQgAAEIAABCEAAAhCAAAQgAAEIQGBpBFgAXFpGiQcCEIAABCAAAQhAAAIQgAAEIAABCEAAAo4AC4AOBlUIQAACEIAABCAAAQhAAAIQgAAEIAABCCyNAAuAS8so8UAAAhCAAAQgAAEIQAACEIAABCAAAQhAwBFgAdDBoAoBCEAAAhCAAAQgAAEIQAACEIAABCAAgaURYAFwaRklHghAAAIQgAAEIAABCEAAAhCAAAQgAAEIOAIsADoYVCEAAQhAAAIQgAAEIAABCEAAAhCAAAQgsDQCLAAuLaPEAwEIQAACEIAABCAAAQhAAAIQgAAEIAABR4AFQAeDKgQgAAEIQAACEIAABCAAAQhAAAIQgAAElkaABcClZZR4IAABCEAAAhCAAAQgAAEIQAACEIAABCDgCLAA6GBQhQAEIAABCEAAAhCAAAQgAAEIQAACEIDA0giwALi0jBIPBCAAAQhAAAIQgAAEIAABCEAAAhCAAAQcARYAHQyqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBYGgEWAJeWUeKBAAQgAAEIQAACEIAABCAAAQhAAAIQgIAjwAKgg0EVAhCAAAQgAAEIQAACEIAABCAAAQhAAAJLI8AC4NIySjwQgAAEIAABCEAAAhCAAAQgAAEIQAACEHAEWAB0MKhCAAIQgAAEIAABCEAAAhCAAAQgAAEIQGBpBFgAXFpGiQcCEIAABCAAAQhAAAIQgAAEIAABCEAAAo4AC4AOBlUIQAACEIAABCAAAQhAAAIQgAAEIAABCCyNAAuAS8so8UAAAhCAAAQgAAEIQAACEIAABCAAAQhAwBFgAdDBoAoBCEAAAhCAAAQgAAEIQAACEIAABCAAgaURYAFwaRklHghAAAIQgAAEIAABCEAAAhCAAAQgAAEIOAIsADoYVCEAAQhAAAIQgAAEIAABCEAAAhCAAAQgsDQCLAAuLaPEAwEIQAACEIAABCAAAQhAAAIQgAAEIAABR4AFQAeDKgQgAAEIQAACEIAABCAAAQhAAAIQgAAElkaABcClZZR4IAABCEAAAhCAAAQgAAEIQAACEIAABCDgCLAA6GBQhQAEIAABCEAAAhCAAAQgAAEIQAACEIDA0giwALi0jBIPBCAAAQhAAAIQgAAEIAABCEAAAhCAAAQcARYAHQyqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBYGgEWAJeWUeKBAAQgAAEIQAACEIAABCAAAQhAAAIQgIAjwAKgg0EVAhCAAAQgAAEIQAACEIAABCAAAQhAAAJLI8AC4NIySjwQgAAEIAABCEAAAhCAAAQgAAEIQAACEHAEWAB0MKhCAAIQgAAEIAABCEAAAhCAAAQgAAEIQGBpBFgAXFpGiQcCEIAABCAAAQhAAAIQgAAEIAABCEAAAo4AC4AOBlUIQAACEIAABCAAAQhAAAIQgAAEIAABCCyNAAuAS8so8UAAAhCAAAQgAAEIQAACEIAABCAAAQhAwBFgAdDBoAoBCEAAAhCAAAQgAAEIQAACEIAABCAAgaURYAFwaRklHghAAAIQgAAEIAABCEAAAhCAAAQgAAEIOAIsADoYVCEAAQhAAAIQgAAEIAABCEAAAhCAAAQgsDQCLAAuLaPEAwEIQAACEIAABCAAAQhAAAIQgAAEIAABR4AFQAeDKgQgAAEIQAACEIAABCAAAQhAAAIQgAAElkaABcClZZR4IAABCEAAAhCAAAQgAAEIQAACEIAABCDgCLAA6GBQhQAEIAABCEAAAhCAAAQgAAEIQAACEIDA0giwALi0jBIPBCAAAQhAAAIQgAAEIAABCEAAAhCAAAQcARYAHQyqEIAABCAAAQhAAAIQgAAEIAABCEAAAhBYGgEWAJeWUeKBAAQgAAEIQAACEIAABCAAAQhAAAIQgIAjwAKgg0EVAhCAAAQgAAEIQAACEIAABCAAAQhAAAJLI8AC4NIySjwQgAAEIAABCEAAAhCAAAQgAAEIQAACEHAEWAB0MKhCAAIQgAAEIAABCEAAAhCAAAQgAAEIQGBpBFgAXFpGiQcCEIAABCAAAQhAAAIQgAAEIAABCEAAAo4AC4AOBlUIQAACEIAABCAAAQhAAAIQgAAEIAABCCyNAAuAS8so8UAAAhCAAAQgAAEIQAACEIAABCAAAQhAwBFgAdDBoAoBCEAAAhCAAAQgAAEIQAACEIAABCAAgaUR+P9AqB/D3e4SFQAAAABJRU5ErkJggg==\" width=\"640\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "image = hsi_pp_1.read()\n",
    "ndvi = (image[24]-image[15])/(image[24]+image[15])\n",
    "\n",
    "ndvi_mask = np.zeros_like(image[0])    \n",
    "ndvi_mask[np.where(ndvi<0.5)] = 1\n",
    "\n",
    "plt.imshow(ndvi_mask, cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tree_detection(input_layer_path, output_shapefile_path, smoothing=0, min_distance=3, mask=None):\n",
    "    input_layer = rasterio.open(input_layer_path)\n",
    "    input_raster = input_layer.read(1)\n",
    "    if mask is not None:\n",
    "        input_raster[np.where(mask)] = 0\n",
    "    input_raster = gaussian(input_raster, sigma=smoothing)\n",
    "    tree_tops = peak_local_max(input_raster, min_distance=min_distance, exclude_border=0, indices=True)\n",
    "    \n",
    "    schema = {'geometry': 'Point', 'properties': {'id': 'int'},}\n",
    "    with fiona.open(output_shapefile_path, 'w', 'ESRI Shapefile', schema, input_layer.crs) as c:\n",
    "        for i, tree_top in enumerate(tree_tops):\n",
    "            tree_point = geometry.Point(input_layer.xy(tree_top[0], tree_top[1]))\n",
    "            c.write({'geometry': geometry.mapping(tree_point), 'properties': {'id': i},})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_detection(labeled_trees_path, detected_trees_path, circle_radius=1):    \n",
    "    labeled_trees = fiona.open(labeled_trees_path, \"r\")\n",
    "    detected_trees = fiona.open(detected_trees_path, \"r\")\n",
    "    \n",
    "    detections = []\n",
    "\n",
    "    for tree in tqdm_notebook(labeled_trees.values(), total=len(labeled_trees)):\n",
    "        trees_detected = 0\n",
    "        tree_circle = geometry.shape(tree['geometry']).buffer(circle_radius)\n",
    "        for detected_tree in detected_trees.values():\n",
    "            tree_top = geometry.shape(detected_tree['geometry'])\n",
    "            if tree_top.within(tree_circle):\n",
    "                trees_detected += 1\n",
    "        detections.append(trees_detected)\n",
    "    \n",
    "    return detections.count(1)/len(labeled_trees)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_detection('../data/pp_1/pp_1_chm.tif', '/Users/vasilii.mosin/Desktop/ForestHype/test_shapes/test_shape_01',\n",
    "              smoothing=0.0, min_distance=3, mask=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1790"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(fiona.open('/Users/vasilii.mosin/Desktop/ForestHype/test_shapes/test_shape_01'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 375,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a4bda4693dcf4b76be9bfea326d116bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=327), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.7675840978593272"
      ]
     },
     "execution_count": 375,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "evaluate_detection('../data/pp_4/pp_4_trees_h', '/Users/vasilii.mosin/Desktop/test_shapes/test_shape_01',\n",
    "                   circle_radius=1.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Final Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "def classify_detected_trees(hsi_path, detected_trees_path, output_shapefile_path, clf):\n",
    "    hsi = rasterio.open(hsi_path)\n",
    "    detected_trees = fiona.open(detected_trees_path, 'r')\n",
    "    \n",
    "    map_tree_speices = {0:'spruce', 1:'birch', 2:'fir', 3:'pine'}\n",
    "    schema = {'geometry': 'Point', 'properties': {'id': 'int', 'specie': 'str', 'proba': 'float'},}\n",
    "    \n",
    "    with fiona.open(output_shapefile_path, 'w', 'ESRI Shapefile', schema, detected_trees.crs) as c:     \n",
    "\n",
    "        for detected_tree in tqdm_notebook(detected_trees.values(), total=len(detected_trees)):\n",
    "            tree_point = geometry.shape(detected_tree['geometry'])\n",
    "            tree_geometry = geometry.shape(detected_tree['geometry']).buffer(1.0)\n",
    "            tree_image, _ = mask(hsi, [tree_geometry], filled=False, crop=True)\n",
    "            exclude_bad_pixels(tree_image)\n",
    "            tree_image = tree_image/65535\n",
    "\n",
    "            tree_image = tree_image/np.ma.sum(tree_image,axis=0)\n",
    "            mean_spectra_norm = np.array([np.ma.mean(band) for band in tree_image])\n",
    "            prediction_probas = clf.predict_proba(mean_spectra_norm.reshape(1,-1))[0]\n",
    "            prediction_specie = np.argmax(prediction_probas)\n",
    "\n",
    "            c.write({'geometry': geometry.mapping(tree_point), \n",
    "                     'properties': {'id': detected_tree['properties']['id'],\n",
    "                                    'specie': map_tree_speices[prediction_specie],\n",
    "                                    'proba': prediction_probas[prediction_specie]}})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e328194908824599989e93e33ccb8fcf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=1790), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "classify_detected_trees('../data/pp_1/pp_1_hsi_clipped.tif', \n",
    "                        '/Users/vasilii.mosin/Desktop/ForestHype/test_shapes/test_shape_01',\n",
    "                        '/Users/vasilii.mosin/Desktop/ForestHype/test_shapes/pp_1_classified', clf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "spruce \t 1275 \t 0.7742156862745098 \t 0.1492484852257338\n",
      "birch \t 297 \t 0.8103535353535354 \t 0.15214059463088517\n",
      "fir \t 1 \t 0.44 \t 0.0\n",
      "pine \t 217 \t 0.6997235023041475 \t 0.13160097856110636\n"
     ]
    }
   ],
   "source": [
    "classified_trees = fiona.open('/Users/vasilii.mosin/Desktop/ForestHype/test_shapes/pp_1_classified', 'r')\n",
    "\n",
    "classified_trees_statistics = {'spruce':[], 'birch':[], 'fir':[], 'pine':[]}\n",
    "\n",
    "for classified_tree in classified_trees.values():\n",
    "    properties = classified_tree['properties']\n",
    "    classified_trees_statistics[properties['specie']].append(properties['proba'])\n",
    "    \n",
    "for specie in classified_trees_statistics:\n",
    "    specie_stats = classified_trees_statistics[specie]\n",
    "    print(specie, '\\t', len(specie_stats), '\\t', np.mean(specie_stats), '\\t', np.std(specie_stats))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}