SliceFinder_demo.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SliceFinder Demo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div class=\"bk-root\">\n",
       "        <a href=\"https://bokeh.pydata.org\" target=\"_blank\" class=\"bk-logo bk-logo-small bk-logo-notebook\"></a>\n",
       "        <span id=\"07fc1343-da38-4a49-8f32-95b2c52247b6\">Loading BokehJS ...</span>\n",
       "    </div>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "(function(root) {\n",
       "  function now() {\n",
       "    return new Date();\n",
       "  }\n",
       "\n",
       "  var force = true;\n",
       "\n",
       "  if (typeof (root._bokeh_onload_callbacks) === \"undefined\" || force === true) {\n",
       "    root._bokeh_onload_callbacks = [];\n",
       "    root._bokeh_is_loading = undefined;\n",
       "  }\n",
       "\n",
       "  var JS_MIME_TYPE = 'application/javascript';\n",
       "  var HTML_MIME_TYPE = 'text/html';\n",
       "  var EXEC_MIME_TYPE = 'application/vnd.bokehjs_exec.v0+json';\n",
       "  var CLASS_NAME = 'output_bokeh rendered_html';\n",
       "\n",
       "  /**\n",
       "   * Render data to the DOM node\n",
       "   */\n",
       "  function render(props, node) {\n",
       "    var script = document.createElement(\"script\");\n",
       "    node.appendChild(script);\n",
       "  }\n",
       "\n",
       "  /**\n",
       "   * Handle when an output is cleared or removed\n",
       "   */\n",
       "  function handleClearOutput(event, handle) {\n",
       "    var cell = handle.cell;\n",
       "\n",
       "    var id = cell.output_area._bokeh_element_id;\n",
       "    var server_id = cell.output_area._bokeh_server_id;\n",
       "    // Clean up Bokeh references\n",
       "    if (id !== undefined) {\n",
       "      Bokeh.index[id].model.document.clear();\n",
       "      delete Bokeh.index[id];\n",
       "    }\n",
       "\n",
       "    if (server_id !== undefined) {\n",
       "      // Clean up Bokeh references\n",
       "      var cmd = \"from bokeh.io.state import curstate; print(curstate().uuid_to_server['\" + server_id + \"'].get_sessions()[0].document.roots[0]._id)\";\n",
       "      cell.notebook.kernel.execute(cmd, {\n",
       "        iopub: {\n",
       "          output: function(msg) {\n",
       "            var element_id = msg.content.text.trim();\n",
       "            Bokeh.index[element_id].model.document.clear();\n",
       "            delete Bokeh.index[element_id];\n",
       "          }\n",
       "        }\n",
       "      });\n",
       "      // Destroy server and session\n",
       "      var cmd = \"import bokeh.io.notebook as ion; ion.destroy_server('\" + server_id + \"')\";\n",
       "      cell.notebook.kernel.execute(cmd);\n",
       "    }\n",
       "  }\n",
       "\n",
       "  /**\n",
       "   * Handle when a new output is added\n",
       "   */\n",
       "  function handleAddOutput(event, handle) {\n",
       "    var output_area = handle.output_area;\n",
       "    var output = handle.output;\n",
       "\n",
       "    // limit handleAddOutput to display_data with EXEC_MIME_TYPE content only\n",
       "    if ((output.output_type != \"display_data\") || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n",
       "      return\n",
       "    }\n",
       "\n",
       "    var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n",
       "\n",
       "    if (output.metadata[EXEC_MIME_TYPE][\"id\"] !== undefined) {\n",
       "      toinsert[0].firstChild.textContent = output.data[JS_MIME_TYPE];\n",
       "      // store reference to embed id on output_area\n",
       "      output_area._bokeh_element_id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n",
       "    }\n",
       "    if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n",
       "      var bk_div = document.createElement(\"div\");\n",
       "      bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "      var script_attrs = bk_div.children[0].attributes;\n",
       "      for (var i = 0; i < script_attrs.length; i++) {\n",
       "        toinsert[0].firstChild.setAttribute(script_attrs[i].name, script_attrs[i].value);\n",
       "      }\n",
       "      // store reference to server id on output_area\n",
       "      output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n",
       "    }\n",
       "  }\n",
       "\n",
       "  function register_renderer(events, OutputArea) {\n",
       "\n",
       "    function append_mime(data, metadata, element) {\n",
       "      // create a DOM node to render to\n",
       "      var toinsert = this.create_output_subarea(\n",
       "        metadata,\n",
       "        CLASS_NAME,\n",
       "        EXEC_MIME_TYPE\n",
       "      );\n",
       "      this.keyboard_manager.register_events(toinsert);\n",
       "      // Render to node\n",
       "      var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n",
       "      render(props, toinsert[0]);\n",
       "      element.append(toinsert);\n",
       "      return toinsert\n",
       "    }\n",
       "\n",
       "    /* Handle when an output is cleared or removed */\n",
       "    events.on('clear_output.CodeCell', handleClearOutput);\n",
       "    events.on('delete.Cell', handleClearOutput);\n",
       "\n",
       "    /* Handle when a new output is added */\n",
       "    events.on('output_added.OutputArea', handleAddOutput);\n",
       "\n",
       "    /**\n",
       "     * Register the mime type and append_mime function with output_area\n",
       "     */\n",
       "    OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n",
       "      /* Is output safe? */\n",
       "      safe: true,\n",
       "      /* Index of renderer in `output_area.display_order` */\n",
       "      index: 0\n",
       "    });\n",
       "  }\n",
       "\n",
       "  // register the mime type if in Jupyter Notebook environment and previously unregistered\n",
       "  if (root.Jupyter !== undefined) {\n",
       "    var events = require('base/js/events');\n",
       "    var OutputArea = require('notebook/js/outputarea').OutputArea;\n",
       "\n",
       "    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n",
       "      register_renderer(events, OutputArea);\n",
       "    }\n",
       "  }\n",
       "\n",
       "  \n",
       "  if (typeof (root._bokeh_timeout) === \"undefined\" || force === true) {\n",
       "    root._bokeh_timeout = Date.now() + 5000;\n",
       "    root._bokeh_failed_load = false;\n",
       "  }\n",
       "\n",
       "  var NB_LOAD_WARNING = {'data': {'text/html':\n",
       "     \"<div style='background-color: #fdd'>\\n\"+\n",
       "     \"<p>\\n\"+\n",
       "     \"BokehJS does not appear to have successfully loaded. If loading BokehJS from CDN, this \\n\"+\n",
       "     \"may be due to a slow or bad network connection. Possible fixes:\\n\"+\n",
       "     \"</p>\\n\"+\n",
       "     \"<ul>\\n\"+\n",
       "     \"<li>re-rerun `output_notebook()` to attempt to load from CDN again, or</li>\\n\"+\n",
       "     \"<li>use INLINE resources instead, as so:</li>\\n\"+\n",
       "     \"</ul>\\n\"+\n",
       "     \"<code>\\n\"+\n",
       "     \"from bokeh.resources import INLINE\\n\"+\n",
       "     \"output_notebook(resources=INLINE)\\n\"+\n",
       "     \"</code>\\n\"+\n",
       "     \"</div>\"}};\n",
       "\n",
       "  function display_loaded() {\n",
       "    var el = document.getElementById(\"07fc1343-da38-4a49-8f32-95b2c52247b6\");\n",
       "    if (el != null) {\n",
       "      el.textContent = \"BokehJS is loading...\";\n",
       "    }\n",
       "    if (root.Bokeh !== undefined) {\n",
       "      if (el != null) {\n",
       "        el.textContent = \"BokehJS \" + root.Bokeh.version + \" successfully loaded.\";\n",
       "      }\n",
       "    } else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(display_loaded, 100)\n",
       "    }\n",
       "  }\n",
       "\n",
       "\n",
       "  function run_callbacks() {\n",
       "    try {\n",
       "      root._bokeh_onload_callbacks.forEach(function(callback) { callback() });\n",
       "    }\n",
       "    finally {\n",
       "      delete root._bokeh_onload_callbacks\n",
       "    }\n",
       "    console.info(\"Bokeh: all callbacks have finished\");\n",
       "  }\n",
       "\n",
       "  function load_libs(js_urls, callback) {\n",
       "    root._bokeh_onload_callbacks.push(callback);\n",
       "    if (root._bokeh_is_loading > 0) {\n",
       "      console.log(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n",
       "      return null;\n",
       "    }\n",
       "    if (js_urls == null || js_urls.length === 0) {\n",
       "      run_callbacks();\n",
       "      return null;\n",
       "    }\n",
       "    console.log(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n",
       "    root._bokeh_is_loading = js_urls.length;\n",
       "    for (var i = 0; i < js_urls.length; i++) {\n",
       "      var url = js_urls[i];\n",
       "      var s = document.createElement('script');\n",
       "      s.src = url;\n",
       "      s.async = false;\n",
       "      s.onreadystatechange = s.onload = function() {\n",
       "        root._bokeh_is_loading--;\n",
       "        if (root._bokeh_is_loading === 0) {\n",
       "          console.log(\"Bokeh: all BokehJS libraries loaded\");\n",
       "          run_callbacks()\n",
       "        }\n",
       "      };\n",
       "      s.onerror = function() {\n",
       "        console.warn(\"failed to load library \" + url);\n",
       "      };\n",
       "      console.log(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "    }\n",
       "  };var element = document.getElementById(\"07fc1343-da38-4a49-8f32-95b2c52247b6\");\n",
       "  if (element == null) {\n",
       "    console.log(\"Bokeh: ERROR: autoload.js configured with elementid '07fc1343-da38-4a49-8f32-95b2c52247b6' but no matching script tag was found. \")\n",
       "    return false;\n",
       "  }\n",
       "\n",
       "  var js_urls = [\"https://cdn.pydata.org/bokeh/release/bokeh-0.12.13.min.js\", \"https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.13.min.js\", \"https://cdn.pydata.org/bokeh/release/bokeh-tables-0.12.13.min.js\", \"https://cdn.pydata.org/bokeh/release/bokeh-gl-0.12.13.min.js\"];\n",
       "\n",
       "  var inline_js = [\n",
       "    function(Bokeh) {\n",
       "      Bokeh.set_log_level(\"info\");\n",
       "    },\n",
       "    \n",
       "    function(Bokeh) {\n",
       "      \n",
       "    },\n",
       "    function(Bokeh) {\n",
       "      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-0.12.13.min.css\");\n",
       "      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-0.12.13.min.css\");\n",
       "      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.13.min.css\");\n",
       "      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.13.min.css\");\n",
       "      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-tables-0.12.13.min.css\");\n",
       "      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-tables-0.12.13.min.css\");\n",
       "    }\n",
       "  ];\n",
       "\n",
       "  function run_inline_js() {\n",
       "    \n",
       "    if ((root.Bokeh !== undefined) || (force === true)) {\n",
       "      for (var i = 0; i < inline_js.length; i++) {\n",
       "        inline_js[i].call(root, root.Bokeh);\n",
       "      }if (force === true) {\n",
       "        display_loaded();\n",
       "      }} else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(run_inline_js, 100);\n",
       "    } else if (!root._bokeh_failed_load) {\n",
       "      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n",
       "      root._bokeh_failed_load = true;\n",
       "    } else if (force !== true) {\n",
       "      var cell = $(document.getElementById(\"07fc1343-da38-4a49-8f32-95b2c52247b6\")).parents('.cell').data().cell;\n",
       "      cell.output_area.append_execute_result(NB_LOAD_WARNING)\n",
       "    }\n",
       "\n",
       "  }\n",
       "\n",
       "  if (root._bokeh_is_loading === 0) {\n",
       "    console.log(\"Bokeh: BokehJS loaded, going straight to plotting\");\n",
       "    run_inline_js();\n",
       "  } else {\n",
       "    load_libs(js_urls, function() {\n",
       "      console.log(\"Bokeh: BokehJS plotting callback run at\", now());\n",
       "      run_inline_js();\n",
       "    });\n",
       "  }\n",
       "}(window));"
      ],
      "application/vnd.bokehjs_load.v0+json": "\n(function(root) {\n  function now() {\n    return new Date();\n  }\n\n  var force = true;\n\n  if (typeof (root._bokeh_onload_callbacks) === \"undefined\" || force === true) {\n    root._bokeh_onload_callbacks = [];\n    root._bokeh_is_loading = undefined;\n  }\n\n  \n\n  \n  if (typeof (root._bokeh_timeout) === \"undefined\" || force === true) {\n    root._bokeh_timeout = Date.now() + 5000;\n    root._bokeh_failed_load = false;\n  }\n\n  var NB_LOAD_WARNING = {'data': {'text/html':\n     \"<div style='background-color: #fdd'>\\n\"+\n     \"<p>\\n\"+\n     \"BokehJS does not appear to have successfully loaded. If loading BokehJS from CDN, this \\n\"+\n     \"may be due to a slow or bad network connection. Possible fixes:\\n\"+\n     \"</p>\\n\"+\n     \"<ul>\\n\"+\n     \"<li>re-rerun `output_notebook()` to attempt to load from CDN again, or</li>\\n\"+\n     \"<li>use INLINE resources instead, as so:</li>\\n\"+\n     \"</ul>\\n\"+\n     \"<code>\\n\"+\n     \"from bokeh.resources import INLINE\\n\"+\n     \"output_notebook(resources=INLINE)\\n\"+\n     \"</code>\\n\"+\n     \"</div>\"}};\n\n  function display_loaded() {\n    var el = document.getElementById(\"07fc1343-da38-4a49-8f32-95b2c52247b6\");\n    if (el != null) {\n      el.textContent = \"BokehJS is loading...\";\n    }\n    if (root.Bokeh !== undefined) {\n      if (el != null) {\n        el.textContent = \"BokehJS \" + root.Bokeh.version + \" successfully loaded.\";\n      }\n    } else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(display_loaded, 100)\n    }\n  }\n\n\n  function run_callbacks() {\n    try {\n      root._bokeh_onload_callbacks.forEach(function(callback) { callback() });\n    }\n    finally {\n      delete root._bokeh_onload_callbacks\n    }\n    console.info(\"Bokeh: all callbacks have finished\");\n  }\n\n  function load_libs(js_urls, callback) {\n    root._bokeh_onload_callbacks.push(callback);\n    if (root._bokeh_is_loading > 0) {\n      console.log(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n      return null;\n    }\n    if (js_urls == null || js_urls.length === 0) {\n      run_callbacks();\n      return null;\n    }\n    console.log(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n    root._bokeh_is_loading = js_urls.length;\n    for (var i = 0; i < js_urls.length; i++) {\n      var url = js_urls[i];\n      var s = document.createElement('script');\n      s.src = url;\n      s.async = false;\n      s.onreadystatechange = s.onload = function() {\n        root._bokeh_is_loading--;\n        if (root._bokeh_is_loading === 0) {\n          console.log(\"Bokeh: all BokehJS libraries loaded\");\n          run_callbacks()\n        }\n      };\n      s.onerror = function() {\n        console.warn(\"failed to load library \" + url);\n      };\n      console.log(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.getElementsByTagName(\"head\")[0].appendChild(s);\n    }\n  };var element = document.getElementById(\"07fc1343-da38-4a49-8f32-95b2c52247b6\");\n  if (element == null) {\n    console.log(\"Bokeh: ERROR: autoload.js configured with elementid '07fc1343-da38-4a49-8f32-95b2c52247b6' but no matching script tag was found. \")\n    return false;\n  }\n\n  var js_urls = [\"https://cdn.pydata.org/bokeh/release/bokeh-0.12.13.min.js\", \"https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.13.min.js\", \"https://cdn.pydata.org/bokeh/release/bokeh-tables-0.12.13.min.js\", \"https://cdn.pydata.org/bokeh/release/bokeh-gl-0.12.13.min.js\"];\n\n  var inline_js = [\n    function(Bokeh) {\n      Bokeh.set_log_level(\"info\");\n    },\n    \n    function(Bokeh) {\n      \n    },\n    function(Bokeh) {\n      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-0.12.13.min.css\");\n      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-0.12.13.min.css\");\n      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.13.min.css\");\n      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.13.min.css\");\n      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-tables-0.12.13.min.css\");\n      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-tables-0.12.13.min.css\");\n    }\n  ];\n\n  function run_inline_js() {\n    \n    if ((root.Bokeh !== undefined) || (force === true)) {\n      for (var i = 0; i < inline_js.length; i++) {\n        inline_js[i].call(root, root.Bokeh);\n      }if (force === true) {\n        display_loaded();\n      }} else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(run_inline_js, 100);\n    } else if (!root._bokeh_failed_load) {\n      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n      root._bokeh_failed_load = true;\n    } else if (force !== true) {\n      var cell = $(document.getElementById(\"07fc1343-da38-4a49-8f32-95b2c52247b6\")).parents('.cell').data().cell;\n      cell.output_area.append_execute_result(NB_LOAD_WARNING)\n    }\n\n  }\n\n  if (root._bokeh_is_loading === 0) {\n    console.log(\"Bokeh: BokehJS loaded, going straight to plotting\");\n    run_inline_js();\n  } else {\n    load_libs(js_urls, function() {\n      console.log(\"Bokeh: BokehJS plotting callback run at\", now());\n      run_inline_js();\n    });\n  }\n}(window));"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import pickle\n",
    "from sklearn.preprocessing import StandardScaler, LabelEncoder\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from slice_finder import SliceFinder\n",
    "\n",
    "from ipywidgets import interact, interactive\n",
    "from IPython.display import display\n",
    "\n",
    "from bokeh.layouts import widgetbox, row\n",
    "from bokeh.models import ColumnDataSource\n",
    "from bokeh.models import HoverTool\n",
    "from bokeh.models.widgets import DataTable, TableColumn  \n",
    "from bokeh.io import push_notebook, show, output_notebook\n",
    "from bokeh.plotting import figure\n",
    "output_notebook()\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "matplotlib.rcParams.update({'font.size': 17})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Toy Example: UCI Census Data Set\n",
    "For the demo, we use UCI Census data set and the classification task is to predict who makes above 50k a year."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Workclass ['Federal-gov' 'Local-gov' 'Private' 'Self-emp-inc' 'Self-emp-not-inc'\n",
      " 'State-gov' 'Without-pay'] [0 1 2 3 4 5 6]\n",
      "Education ['10th' '11th' '12th' '1st-4th' '5th-6th' '7th-8th' '9th' 'Assoc-acdm'\n",
      " 'Assoc-voc' 'Bachelors' 'Doctorate' 'HS-grad' 'Masters' 'Preschool'\n",
      " 'Prof-school' 'Some-college'] [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15]\n",
      "Martial Status ['Divorced' 'Married-AF-spouse' 'Married-civ-spouse'\n",
      " 'Married-spouse-absent' 'Never-married' 'Separated' 'Widowed'] [0 1 2 3 4 5 6]\n",
      "Occupation ['Adm-clerical' 'Armed-Forces' 'Craft-repair' 'Exec-managerial'\n",
      " 'Farming-fishing' 'Handlers-cleaners' 'Machine-op-inspct' 'Other-service'\n",
      " 'Priv-house-serv' 'Prof-specialty' 'Protective-serv' 'Sales'\n",
      " 'Tech-support' 'Transport-moving'] [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13]\n",
      "Relationship ['Husband' 'Not-in-family' 'Other-relative' 'Own-child' 'Unmarried' 'Wife'] [0 1 2 3 4 5]\n",
      "Race ['Amer-Indian-Eskimo' 'Asian-Pac-Islander' 'Black' 'Other' 'White'] [0 1 2 3 4]\n",
      "Sex ['Female' 'Male'] [0 1]\n",
      "Country ['Cambodia' 'Canada' 'China' 'Columbia' 'Cuba' 'Dominican-Republic'\n",
      " 'Ecuador' 'El-Salvador' 'England' 'France' 'Germany' 'Greece' 'Guatemala'\n",
      " 'Haiti' 'Holand-Netherlands' 'Honduras' 'Hong' 'Hungary' 'India' 'Iran'\n",
      " 'Ireland' 'Italy' 'Jamaica' 'Japan' 'Laos' 'Mexico' 'Nicaragua'\n",
      " 'Outlying-US(Guam-USVI-etc)' 'Peru' 'Philippines' 'Poland' 'Portugal'\n",
      " 'Puerto-Rico' 'Scotland' 'South' 'Taiwan' 'Thailand' 'Trinadad&Tobago'\n",
      " 'United-States' 'Vietnam' 'Yugoslavia'] [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24\n",
      " 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40]\n",
      "Target ['<=50K' '>50K'] [0 1]\n"
     ]
    }
   ],
   "source": [
    "adult_data = pd.read_csv(\n",
    "    \"data/adult.data\",\n",
    "    names=[\n",
    "        \"Age\", \"Workclass\", \"fnlwgt\", \"Education\", \"Education-Num\", \"Martial Status\",\n",
    "        \"Occupation\", \"Relationship\", \"Race\", \"Sex\", \"Capital Gain\", \"Capital Loss\",\n",
    "        \"Hours per week\", \"Country\", \"Target\"],\n",
    "        sep=r'\\s*,\\s*',\n",
    "        engine='python',\n",
    "        na_values=\"?\")\n",
    "\n",
    "# drop nan values\n",
    "adult_data = adult_data.dropna()\n",
    "\n",
    "# Encode categorical features\n",
    "encoders = {}\n",
    "for column in adult_data.columns:\n",
    "    if adult_data.dtypes[column] == np.object:\n",
    "        le = LabelEncoder()\n",
    "        adult_data[column] = le.fit_transform(adult_data[column])\n",
    "        encoders[column] = le\n",
    "        print(column, le.classes_, le.transform(le.classes_))\n",
    "\n",
    "X, y = adult_data[adult_data.columns.difference([\"Target\"])], adult_data[\"Target\"]\n",
    "\n",
    "pickle.dump(encoders, open(\"encoders.pkl\", \"wb\"), protocol=2)\n",
    "\n",
    "# Train a model\n",
    "#lr = LogisticRegression()\n",
    "#lr.fit(X, y)\n",
    "lr = RandomForestClassifier(max_depth=5, n_estimators=10)\n",
    "lr.fit(X, y)\n",
    "\n",
    "sf = SliceFinder(lr, (X, y))\n",
    "metrics_all = sf.evaluate_model((X,y))\n",
    "reference = (np.mean(metrics_all), np.std(metrics_all), len(metrics_all))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SliceFinder Recommendation Example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "degree 1\n",
      "crossing\n",
      "effect size filtering\n",
      "degree 2\n",
      "crossing\n",
      "effect size filtering\n",
      "sorting\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Sex:Male \n",
      "---------------------\n",
      "effect_size: 0.489027069358\n",
      "---------------------\n",
      "metric: 0.415072526958\n",
      "size: 20380\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Martial Status:Married-civ-spouse \n",
      "---------------------\n",
      "effect_size: 0.923021720221\n",
      "---------------------\n",
      "metric: 0.54597411274\n",
      "size: 14065\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Relationship:Husband \n",
      "---------------------\n",
      "effect_size: 0.837430353161\n",
      "---------------------\n",
      "metric: 0.543190034329\n",
      "size: 12463\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Occupation:Exec-managerial \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.468894755867\n",
      "---------------------\n",
      "metric: 0.525271237101\n",
      "size: 3453\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.479851077333\n",
      "---------------------\n",
      "metric: 0.532079519063\n",
      "size: 2333\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Loss:0 \n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.410478506987\n",
      "---------------------\n",
      "metric: 0.503320568135\n",
      "size: 1482\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Loss:0 \n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.410478506987\n",
      "---------------------\n",
      "metric: 0.503320568135\n",
      "size: 1482\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Relationship:Wife \n",
      "---------------------\n",
      "effect_size: 0.693774765497\n",
      "---------------------\n",
      "metric: 0.596982222682\n",
      "size: 1406\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Masters \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.525841538236\n",
      "---------------------\n",
      "metric: 0.54213274679\n",
      "size: 1368\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.525841538236\n",
      "---------------------\n",
      "metric: 0.54213274679\n",
      "size: 1368\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Country:United-States \n",
      "---------------------\n",
      "effect_size: 0.402075450765\n",
      "---------------------\n",
      "metric: 0.503786882237\n",
      "size: 1300\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.498045502735\n",
      "---------------------\n",
      "metric: 0.539630623007\n",
      "size: 1225\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Workclass:Self-emp-inc \n",
      "---------------------\n",
      "effect_size: 0.421223761818\n",
      "---------------------\n",
      "metric: 0.515127201504\n",
      "size: 1074\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Education:Bachelors \n",
      "---------------------\n",
      "effect_size: 0.40727476323\n",
      "---------------------\n",
      "metric: 0.504898479083\n",
      "size: 688\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.40727476323\n",
      "---------------------\n",
      "metric: 0.504898479083\n",
      "size: 688\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Occupation:Exec-managerial \n",
      "---------------------\n",
      "effect_size: 0.431487778222\n",
      "---------------------\n",
      "metric: 0.520904812107\n",
      "size: 665\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:40 \n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.409749735485\n",
      "---------------------\n",
      "metric: 0.511567183022\n",
      "size: 628\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:40 \n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.409749735485\n",
      "---------------------\n",
      "metric: 0.511567183022\n",
      "size: 628\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Age:47 \n",
      "---------------------\n",
      "effect_size: 0.4239286605\n",
      "---------------------\n",
      "metric: 0.51819380336\n",
      "size: 600\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:55 \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.466883529327\n",
      "---------------------\n",
      "metric: 0.533366944316\n",
      "size: 573\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Country:United-States \n",
      "Age:50 \n",
      "---------------------\n",
      "effect_size: 0.42523742325\n",
      "---------------------\n",
      "metric: 0.517751793938\n",
      "size: 529\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Age:50 \n",
      "---------------------\n",
      "effect_size: 0.491393616011\n",
      "---------------------\n",
      "metric: 0.542508406273\n",
      "size: 509\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Race:White \n",
      "Age:50 \n",
      "---------------------\n",
      "effect_size: 0.44910745663\n",
      "---------------------\n",
      "metric: 0.526896020101\n",
      "size: 501\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Race:White \n",
      "Age:51 \n",
      "---------------------\n",
      "effect_size: 0.409709239172\n",
      "---------------------\n",
      "metric: 0.514116484746\n",
      "size: 497\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Occupation:Exec-managerial \n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.41583399533\n",
      "---------------------\n",
      "metric: 0.505310609317\n",
      "size: 491\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Occupation:Exec-managerial \n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.41583399533\n",
      "---------------------\n",
      "metric: 0.505310609317\n",
      "size: 491\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Age:48 \n",
      "---------------------\n",
      "effect_size: 0.407535961687\n",
      "---------------------\n",
      "metric: 0.518190972947\n",
      "size: 473\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:45 \n",
      "Education:Bachelors \n",
      "---------------------\n",
      "effect_size: 0.408315772404\n",
      "---------------------\n",
      "metric: 0.504084231462\n",
      "size: 451\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:45 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.408315772404\n",
      "---------------------\n",
      "metric: 0.504084231462\n",
      "size: 451\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Occupation:Sales \n",
      "---------------------\n",
      "effect_size: 0.435338449041\n",
      "---------------------\n",
      "metric: 0.530730005321\n",
      "size: 438\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Prof-school \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.675026275195\n",
      "---------------------\n",
      "metric: 0.602555140143\n",
      "size: 403\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Education-Num:15 \n",
      "---------------------\n",
      "effect_size: 0.675026275195\n",
      "---------------------\n",
      "metric: 0.602555140143\n",
      "size: 403\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:45 \n",
      "Occupation:Exec-managerial \n",
      "---------------------\n",
      "effect_size: 0.433975900256\n",
      "---------------------\n",
      "metric: 0.525845188398\n",
      "size: 388\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Workclass:Self-emp-not-inc \n",
      "Education:Bachelors \n",
      "---------------------\n",
      "effect_size: 0.583924669039\n",
      "---------------------\n",
      "metric: 0.558327329103\n",
      "size: 387\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Workclass:Self-emp-not-inc \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.583924669039\n",
      "---------------------\n",
      "metric: 0.558327329103\n",
      "size: 387\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Workclass:Self-emp-not-inc \n",
      "Occupation:Exec-managerial \n",
      "---------------------\n",
      "effect_size: 0.480790311888\n",
      "---------------------\n",
      "metric: 0.530712725758\n",
      "size: 383\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Doctorate \n",
      "---------------------\n",
      "effect_size: 0.518529059992\n",
      "---------------------\n",
      "metric: 0.559417172446\n",
      "size: 375\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education-Num:16 \n",
      "---------------------\n",
      "effect_size: 0.518529059992\n",
      "---------------------\n",
      "metric: 0.559417172446\n",
      "size: 375\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Workclass:Self-emp-not-inc \n",
      "---------------------\n",
      "effect_size: 0.446396234521\n",
      "---------------------\n",
      "metric: 0.537494831946\n",
      "size: 369\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Workclass:Self-emp-not-inc \n",
      "Occupation:Prof-specialty \n",
      "---------------------\n",
      "effect_size: 0.416353755384\n",
      "---------------------\n",
      "metric: 0.519951877944\n",
      "size: 365\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Country:United-States \n",
      "Age:54 \n",
      "---------------------\n",
      "effect_size: 0.427271174807\n",
      "---------------------\n",
      "metric: 0.530716962159\n",
      "size: 356\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Age:54 \n",
      "---------------------\n",
      "effect_size: 0.479406119121\n",
      "---------------------\n",
      "metric: 0.553494859139\n",
      "size: 351\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Race:White \n",
      "Age:54 \n",
      "---------------------\n",
      "effect_size: 0.406045478297\n",
      "---------------------\n",
      "metric: 0.520299088448\n",
      "size: 349\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Occupation:Exec-managerial \n",
      "---------------------\n",
      "effect_size: 0.441691595113\n",
      "---------------------\n",
      "metric: 0.522266334232\n",
      "size: 314\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Education:Bachelors \n",
      "---------------------\n",
      "effect_size: 0.501483796123\n",
      "---------------------\n",
      "metric: 0.53781868672\n",
      "size: 305\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.501483796123\n",
      "---------------------\n",
      "metric: 0.53781868672\n",
      "size: 305\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Workclass:Local-gov \n",
      "Occupation:Protective-serv \n",
      "---------------------\n",
      "effect_size: 0.496073880609\n",
      "---------------------\n",
      "metric: 0.590139505206\n",
      "size: 304\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Age:59 \n",
      "---------------------\n",
      "effect_size: 0.412457194204\n",
      "---------------------\n",
      "metric: 0.525254911607\n",
      "size: 300\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Race:White \n",
      "Age:59 \n",
      "---------------------\n",
      "effect_size: 0.413192456134\n",
      "---------------------\n",
      "metric: 0.524586042239\n",
      "size: 300\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "fnlwgt:130557.24 ~ 158681.4\n",
      "---------------------\n",
      "effect_size: 0.443453690807\n",
      "---------------------\n",
      "metric: 0.535150863557\n",
      "size: 280\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Workclass:Self-emp-not-inc \n",
      "---------------------\n",
      "effect_size: 0.556089859065\n",
      "---------------------\n",
      "metric: 0.560511026595\n",
      "size: 274\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Occupation:Prof-specialty \n",
      "---------------------\n",
      "effect_size: 0.405867383357\n",
      "---------------------\n",
      "metric: 0.502799729557\n",
      "size: 260\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.465863372828\n",
      "---------------------\n",
      "metric: 0.52675537386\n",
      "size: 253\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:50 \n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.465863372828\n",
      "---------------------\n",
      "metric: 0.52675537386\n",
      "size: 253\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:70 \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.425118259343\n",
      "---------------------\n",
      "metric: 0.51756272665\n",
      "size: 245\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Occupation:Sales \n",
      "---------------------\n",
      "effect_size: 0.487427063753\n",
      "---------------------\n",
      "metric: 0.532853973323\n",
      "size: 241\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Masters \n",
      "Martial Status:Divorced \n",
      "---------------------\n",
      "effect_size: 0.477058490718\n",
      "---------------------\n",
      "metric: 0.52510673471\n",
      "size: 229\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Martial Status:Divorced \n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.477058490718\n",
      "---------------------\n",
      "metric: 0.52510673471\n",
      "size: 229\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Occupation:Craft-repair \n",
      "Education:Bachelors \n",
      "---------------------\n",
      "effect_size: 0.4353831562\n",
      "---------------------\n",
      "metric: 0.509473588338\n",
      "size: 220\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Occupation:Craft-repair \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.4353831562\n",
      "---------------------\n",
      "metric: 0.509473588338\n",
      "size: 220\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:65 \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.465551569963\n",
      "---------------------\n",
      "metric: 0.555859502867\n",
      "size: 211\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Loss:1902 \n",
      "---------------------\n",
      "effect_size: 0.430036366897\n",
      "---------------------\n",
      "metric: 0.486885713868\n",
      "size: 194\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:55 \n",
      "Education:Bachelors \n",
      "---------------------\n",
      "effect_size: 0.465482797253\n",
      "---------------------\n",
      "metric: 0.527393954411\n",
      "size: 188\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:55 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.465482797253\n",
      "---------------------\n",
      "metric: 0.527393954411\n",
      "size: 188\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:55 \n",
      "Occupation:Exec-managerial \n",
      "---------------------\n",
      "effect_size: 0.422919597991\n",
      "---------------------\n",
      "metric: 0.517172638113\n",
      "size: 188\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:40 \n",
      "Age:54 \n",
      "---------------------\n",
      "effect_size: 0.438788872949\n",
      "---------------------\n",
      "metric: 0.539983076589\n",
      "size: 186\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "fnlwgt:13769.0 ~ 65741.1\n",
      "---------------------\n",
      "effect_size: 0.421148381567\n",
      "---------------------\n",
      "metric: 0.509972185834\n",
      "size: 183\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Race:Asian-Pac-Islander \n",
      "Occupation:Prof-specialty \n",
      "---------------------\n",
      "effect_size: 0.402777435448\n",
      "---------------------\n",
      "metric: 0.540610264079\n",
      "size: 176\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:158681.4 ~ 178425.0\n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.465134318068\n",
      "---------------------\n",
      "metric: 0.523408697512\n",
      "size: 175\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:158681.4 ~ 178425.0\n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.465134318068\n",
      "---------------------\n",
      "metric: 0.523408697512\n",
      "size: 175\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Workclass:Federal-gov \n",
      "Occupation:Prof-specialty \n",
      "---------------------\n",
      "effect_size: 0.49146345462\n",
      "---------------------\n",
      "metric: 0.569795372722\n",
      "size: 167\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:32 \n",
      "---------------------\n",
      "effect_size: 0.404705084625\n",
      "---------------------\n",
      "metric: 0.497366318171\n",
      "size: 166\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:32 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.404705084625\n",
      "---------------------\n",
      "metric: 0.497366318171\n",
      "size: 166\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:196343.44 ~ 219896.34\n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.407727284087\n",
      "---------------------\n",
      "metric: 0.505975291781\n",
      "size: 165\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:196343.44 ~ 219896.34\n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.407727284087\n",
      "---------------------\n",
      "metric: 0.505975291781\n",
      "size: 165\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:13769.0 ~ 65741.1\n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.401197507518\n",
      "---------------------\n",
      "metric: 0.49456387879\n",
      "size: 162\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:13769.0 ~ 65741.1\n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.401197507518\n",
      "---------------------\n",
      "metric: 0.49456387879\n",
      "size: 162\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "Occupation:Craft-repair \n",
      "---------------------\n",
      "effect_size: 0.45353445388\n",
      "---------------------\n",
      "metric: 0.530707047921\n",
      "size: 159\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:39 \n",
      "---------------------\n",
      "effect_size: 0.490028617657\n",
      "---------------------\n",
      "metric: 0.526821855873\n",
      "size: 158\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:39 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.490028617657\n",
      "---------------------\n",
      "metric: 0.526821855873\n",
      "size: 158\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:43 \n",
      "---------------------\n",
      "effect_size: 0.599400768658\n",
      "---------------------\n",
      "metric: 0.568950812411\n",
      "size: 158\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:43 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.599400768658\n",
      "---------------------\n",
      "metric: 0.568950812411\n",
      "size: 158\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:34 \n",
      "---------------------\n",
      "effect_size: 0.401864612929\n",
      "---------------------\n",
      "metric: 0.496774518592\n",
      "size: 157\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:34 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.401864612929\n",
      "---------------------\n",
      "metric: 0.496774518592\n",
      "size: 157\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Loss:1887 \n",
      "---------------------\n",
      "effect_size: 0.708850393259\n",
      "---------------------\n",
      "metric: 0.571615601483\n",
      "size: 155\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:35 \n",
      "---------------------\n",
      "effect_size: 0.501868646368\n",
      "---------------------\n",
      "metric: 0.547803001436\n",
      "size: 152\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:35 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.501868646368\n",
      "---------------------\n",
      "metric: 0.547803001436\n",
      "size: 152\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:260347.72 ~ 329012.98\n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.401465321718\n",
      "---------------------\n",
      "metric: 0.518132752662\n",
      "size: 149\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "fnlwgt:260347.72 ~ 329012.98\n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.401465321718\n",
      "---------------------\n",
      "metric: 0.518132752662\n",
      "size: 149\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:45 \n",
      "---------------------\n",
      "effect_size: 0.598457572268\n",
      "---------------------\n",
      "metric: 0.567434388696\n",
      "size: 147\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:45 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.598457572268\n",
      "---------------------\n",
      "metric: 0.567434388696\n",
      "size: 147\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:45 \n",
      "Education:Masters \n",
      "---------------------\n",
      "effect_size: 0.545796923811\n",
      "---------------------\n",
      "metric: 0.554524260359\n",
      "size: 146\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:45 \n",
      "Education-Num:14 \n",
      "---------------------\n",
      "effect_size: 0.545796923811\n",
      "---------------------\n",
      "metric: 0.554524260359\n",
      "size: 146\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:44 \n",
      "---------------------\n",
      "effect_size: 0.429639633324\n",
      "---------------------\n",
      "metric: 0.50488621172\n",
      "size: 144\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:44 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.429639633324\n",
      "---------------------\n",
      "metric: 0.50488621172\n",
      "size: 144\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Hours per week:60 \n",
      "fnlwgt:130557.24 ~ 158681.4\n",
      "---------------------\n",
      "effect_size: 0.438975553556\n",
      "---------------------\n",
      "metric: 0.522871454514\n",
      "size: 144\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Occupation:Exec-managerial \n",
      "Education:Assoc-acdm \n",
      "---------------------\n",
      "effect_size: 0.437233135819\n",
      "---------------------\n",
      "metric: 0.568473559096\n",
      "size: 143\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Occupation:Exec-managerial \n",
      "Education-Num:12 \n",
      "---------------------\n",
      "effect_size: 0.437233135819\n",
      "---------------------\n",
      "metric: 0.568473559096\n",
      "size: 143\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Bachelors \n",
      "Age:33 \n",
      "---------------------\n",
      "effect_size: 0.445588285129\n",
      "---------------------\n",
      "metric: 0.520303499631\n",
      "size: 142\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Age:33 \n",
      "Education-Num:13 \n",
      "---------------------\n",
      "effect_size: 0.445588285129\n",
      "---------------------\n",
      "metric: 0.520303499631\n",
      "size: 142\n"
     ]
    }
   ],
   "source": [
    "\n",
    "recommendations = sf.find_slice(k=100, epsilon=0.4, degree=2, max_workers=4)\n",
    "\n",
    "for s in recommendations:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    for k, v in list(s.filters.items()):\n",
    "        values = ''\n",
    "        if k in encoders:\n",
    "            le = encoders[k]\n",
    "            for v_ in v:\n",
    "                values += '%s '%(le.inverse_transform(v_)[0])\n",
    "        else:\n",
    "            for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                if len(v_) > 1:\n",
    "                    values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                else:\n",
    "                    values += '%s '%(v_[0])\n",
    "        print ('%s:%s'%(k, values))\n",
    "    print ('---------------------\\neffect_size: %s'%(s.effect_size))\n",
    "    print ('---------------------\\nmetric: %s'%(s.metric))\n",
    "    print ('size: %s'%(s.size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Sex:Male \n",
      "---------------------\n",
      "effect_size: 0.489027069358\n",
      "---------------------\n",
      "metric: 0.415072526958\n",
      "size: 20380\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Martial Status:Married-civ-spouse \n",
      "---------------------\n",
      "effect_size: 0.923021720221\n",
      "---------------------\n",
      "metric: 0.54597411274\n",
      "size: 14065\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Relationship:Husband \n",
      "---------------------\n",
      "effect_size: 0.837430353161\n",
      "---------------------\n",
      "metric: 0.543190034329\n",
      "size: 12463\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Occupation:Exec-managerial \n",
      "---------------------\n",
      "effect_size: 0.468894755867\n",
      "---------------------\n",
      "metric: 0.525271237101\n",
      "size: 3453\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Capital Gain:0 \n",
      "Hours per week:50 \n",
      "---------------------\n",
      "effect_size: 0.479851077333\n",
      "---------------------\n",
      "metric: 0.532079519063\n",
      "size: 2333\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Masters \n",
      "Capital Loss:0 \n",
      "---------------------\n",
      "effect_size: 0.410478506987\n",
      "---------------------\n",
      "metric: 0.503320568135\n",
      "size: 1482\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education-Num:14 \n",
      "Capital Loss:0 \n",
      "---------------------\n",
      "effect_size: 0.410478506987\n",
      "---------------------\n",
      "metric: 0.503320568135\n",
      "size: 1482\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Relationship:Wife \n",
      "---------------------\n",
      "effect_size: 0.693774765497\n",
      "---------------------\n",
      "metric: 0.596982222682\n",
      "size: 1406\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education:Masters \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.525841538236\n",
      "---------------------\n",
      "metric: 0.54213274679\n",
      "size: 1368\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Education-Num:14 \n",
      "Capital Gain:0 \n",
      "---------------------\n",
      "effect_size: 0.525841538236\n",
      "---------------------\n",
      "metric: 0.54213274679\n",
      "size: 1368\n"
     ]
    }
   ],
   "source": [
    "slices = list()\n",
    "with open('slices.p','rb') as handle:\n",
    "    slices = pickle.load(handle)\n",
    "\n",
    "pickle.dump(slices, open('slices_ver2.pkl','wb'), protocol=2)\n",
    "\n",
    "encoders = dict()\n",
    "with open('encoder')\n",
    "\n",
    "slices = list()\n",
    "with open('slices_ver2.pkl','rb') as handle:\n",
    "    slices = pickle.load(handle)\n",
    "    \n",
    "slices_ = sorted(slices, key=lambda s: s.size, reverse=True)    \n",
    "k = 10    \n",
    "for s in slices_[:k]:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    for k, v in list(s.filters.items()):\n",
    "        values = ''\n",
    "        if k in encoders:\n",
    "            le = encoders[k]\n",
    "            for v_ in v:\n",
    "                values += '%s '%(le.inverse_transform(v_)[0])\n",
    "        else:\n",
    "            for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                if len(v_) > 1:\n",
    "                    values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                else:\n",
    "                    values += '%s '%(v_[0])\n",
    "        print ('%s:%s'%(k, values))\n",
    "    print ('---------------------\\neffect_size: %s'%(s.effect_size))\n",
    "    print ('---------------------\\nmetric: %s'%(s.metric))\n",
    "    print ('size: %s'%(s.size))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SliceFinder User Interface with Minimum Effect Size Slider"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div class=\"bk-root\">\n",
       "    <div class=\"bk-plotdiv\" id=\"74ad42d1-871e-4fe2-9358-a4cd796f40f6\"></div>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "(function(root) {\n",
       "  function embed_document(root) {\n",
       "    \n",
       "  var docs_json = {\"c4338bdf-c63b-4f2b-8fe3-300626de2b8d\":{\"roots\":{\"references\":[{\"attributes\":{\"callback\":null},\"id\":\"6c007a60-a238-4be3-a80f-fe4381d5ae89\",\"type\":\"DataRange1d\"},{\"attributes\":{\"callback\":null},\"id\":\"be132bc0-a99d-4974-a2dd-049031f75792\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"c3ecec06-e4bf-4746-983a-0959820a9b15\",\"type\":\"LinearScale\"},{\"attributes\":{\"source\":{\"id\":\"f4c2a054-30fb-4b8b-9556-cb4778ef4d95\",\"type\":\"ColumnDataSource\"}},\"id\":\"b1aec665-c927-4ee0-bd66-a0f6232e1f82\",\"type\":\"CDSView\"},{\"attributes\":{},\"id\":\"3db75ccd-b7c6-439b-9a01-af4f24375ac1\",\"type\":\"StringFormatter\"},{\"attributes\":{\"axis_label\":\"Effect Size\",\"formatter\":{\"id\":\"67265ce6-2dcf-4b8f-9b77-b3d70458a843\",\"type\":\"BasicTickFormatter\"},\"plot\":{\"id\":\"746c3466-c6b5-4111-8e98-3ec46c90fe0b\",\"subtype\":\"Figure\",\"type\":\"Plot\"},\"ticker\":{\"id\":\"2f706e44-ab34-4130-aeb8-951ab6c835e9\",\"type\":\"BasicTicker\"}},\"id\":\"2fbfc7ca-820f-43fa-9e89-91a522dc4702\",\"type\":\"LinearAxis\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.1},\"fill_color\":{\"value\":\"#1f77b4\"},\"line_alpha\":{\"value\":0.1},\"line_color\":{\"value\":\"#1f77b4\"},\"x\":{\"field\":\"size\"},\"y\":{\"field\":\"effect_size\"}},\"id\":\"95d50826-eff7-4ec1-b3b9-9ef63fae2ee3\",\"type\":\"Circle\"},{\"attributes\":{\"callback\":null,\"tooltips\":[[\"desc\",\"@description\"],[\"size\",\"@size\"],[\"effect_size\",\"@effect_size\"],[\"metric\",\"@metric\"]]},\"id\":\"b8292cf7-214a-4e60-a659-f0487f13e8a4\",\"type\":\"HoverTool\"},{\"attributes\":{\"columns\":[{\"id\":\"9d00c0e1-dd6f-461e-a226-ba3b04e8e6cb\",\"type\":\"TableColumn\"},{\"id\":\"a8cf714d-8a8b-41d6-a3ce-dcf27f2850ac\",\"type\":\"TableColumn\"},{\"id\":\"5aeb34fc-c514-4586-a9b9-f98d0547d5ad\",\"type\":\"TableColumn\"},{\"id\":\"9961adef-8bd1-41d1-8bdb-274611d16ef1\",\"type\":\"TableColumn\"}],\"source\":{\"id\":\"f4c2a054-30fb-4b8b-9556-cb4778ef4d95\",\"type\":\"ColumnDataSource\"},\"view\":{\"id\":\"b1aec665-c927-4ee0-bd66-a0f6232e1f82\",\"type\":\"CDSView\"},\"width\":600},\"id\":\"5e2eec85-4416-43d4-a09a-188fd5163a2b\",\"type\":\"DataTable\"},{\"attributes\":{\"active_drag\":\"auto\",\"active_inspect\":\"auto\",\"active_scroll\":\"auto\",\"active_tap\":\"auto\",\"tools\":[{\"id\":\"b8292cf7-214a-4e60-a659-f0487f13e8a4\",\"type\":\"HoverTool\"}]},\"id\":\"8a1d502b-5482-4d4f-9285-4467a7f1af11\",\"type\":\"Toolbar\"},{\"attributes\":{},\"id\":\"67265ce6-2dcf-4b8f-9b77-b3d70458a843\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"fill_color\":{\"value\":\"#1f77b4\"},\"line_color\":{\"value\":\"#1f77b4\"},\"x\":{\"field\":\"size\"},\"y\":{\"field\":\"effect_size\"}},\"id\":\"18d92379-cf85-416d-925d-66e7222ea4cb\",\"type\":\"Circle\"},{\"attributes\":{\"data_source\":{\"id\":\"f4c2a054-30fb-4b8b-9556-cb4778ef4d95\",\"type\":\"ColumnDataSource\"},\"glyph\":{\"id\":\"18d92379-cf85-416d-925d-66e7222ea4cb\",\"type\":\"Circle\"},\"hover_glyph\":null,\"muted_glyph\":null,\"nonselection_glyph\":{\"id\":\"95d50826-eff7-4ec1-b3b9-9ef63fae2ee3\",\"type\":\"Circle\"},\"selection_glyph\":null,\"view\":{\"id\":\"affcb953-26fb-4b45-9501-5f07952e78f9\",\"type\":\"CDSView\"}},\"id\":\"abc7e422-4aab-4649-ad6e-e12d938a123f\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"editor\":{\"id\":\"c6bd8594-5348-4b35-af08-078ae2923a13\",\"type\":\"StringEditor\"},\"field\":\"size\",\"formatter\":{\"id\":\"3db75ccd-b7c6-439b-9a01-af4f24375ac1\",\"type\":\"StringFormatter\"},\"title\":\"Size\"},\"id\":\"a8cf714d-8a8b-41d6-a3ce-dcf27f2850ac\",\"type\":\"TableColumn\"},{\"attributes\":{},\"id\":\"a409eb01-82dd-4b6d-810c-58605f848fef\",\"type\":\"BasicTicker\"},{\"attributes\":{\"below\":[{\"id\":\"f87a2fe7-5530-4895-a0e9-55bef30d213d\",\"type\":\"LinearAxis\"}],\"left\":[{\"id\":\"2fbfc7ca-820f-43fa-9e89-91a522dc4702\",\"type\":\"LinearAxis\"}],\"plot_height\":400,\"plot_width\":400,\"renderers\":[{\"id\":\"f87a2fe7-5530-4895-a0e9-55bef30d213d\",\"type\":\"LinearAxis\"},{\"id\":\"5aee2955-41aa-4456-9a68-463c83426d67\",\"type\":\"Grid\"},{\"id\":\"2fbfc7ca-820f-43fa-9e89-91a522dc4702\",\"type\":\"LinearAxis\"},{\"id\":\"361aa0bd-5280-4aea-9aeb-5c9ca06a6fab\",\"type\":\"Grid\"},{\"id\":\"abc7e422-4aab-4649-ad6e-e12d938a123f\",\"type\":\"GlyphRenderer\"}],\"title\":null,\"toolbar\":{\"id\":\"8a1d502b-5482-4d4f-9285-4467a7f1af11\",\"type\":\"Toolbar\"},\"x_range\":{\"id\":\"be132bc0-a99d-4974-a2dd-049031f75792\",\"type\":\"DataRange1d\"},\"x_scale\":{\"id\":\"de7dcdd5-9486-485d-a1db-fc25e6fa2308\",\"type\":\"LinearScale\"},\"y_range\":{\"id\":\"6c007a60-a238-4be3-a80f-fe4381d5ae89\",\"type\":\"DataRange1d\"},\"y_scale\":{\"id\":\"c3ecec06-e4bf-4746-983a-0959820a9b15\",\"type\":\"LinearScale\"}},\"id\":\"746c3466-c6b5-4111-8e98-3ec46c90fe0b\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"children\":[{\"id\":\"746c3466-c6b5-4111-8e98-3ec46c90fe0b\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"id\":\"96ddc4f9-d7b2-497e-9d5d-30520db947fe\",\"type\":\"WidgetBox\"}]},\"id\":\"a55137e4-0802-47b5-a0e0-9f8f73667083\",\"type\":\"Row\"},{\"attributes\":{},\"id\":\"643901d9-d907-498f-a7e0-eb77911cc3a5\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"dimension\":1,\"plot\":{\"id\":\"746c3466-c6b5-4111-8e98-3ec46c90fe0b\",\"subtype\":\"Figure\",\"type\":\"Plot\"},\"ticker\":{\"id\":\"2f706e44-ab34-4130-aeb8-951ab6c835e9\",\"type\":\"BasicTicker\"}},\"id\":\"361aa0bd-5280-4aea-9aeb-5c9ca06a6fab\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"c6bd8594-5348-4b35-af08-078ae2923a13\",\"type\":\"StringEditor\"},{\"attributes\":{},\"id\":\"c2939265-fff6-48ad-9a6f-623efd747e1e\",\"type\":\"StringFormatter\"},{\"attributes\":{},\"id\":\"0e7b27a2-0c3f-4963-bcd6-44c928ce4ab2\",\"type\":\"StringFormatter\"},{\"attributes\":{\"callback\":null,\"column_names\":[\"size\",\"metric\",\"description\",\"effect_size\"],\"data\":{\"description\":[],\"effect_size\":[],\"metric\":[],\"size\":[]}},\"id\":\"f4c2a054-30fb-4b8b-9556-cb4778ef4d95\",\"type\":\"ColumnDataSource\"},{\"attributes\":{},\"id\":\"2f706e44-ab34-4130-aeb8-951ab6c835e9\",\"type\":\"BasicTicker\"},{\"attributes\":{\"editor\":{\"id\":\"7330607b-ee1f-4282-84cf-0c99733431ed\",\"type\":\"StringEditor\"},\"field\":\"description\",\"formatter\":{\"id\":\"65ed03fa-4492-491e-9864-a845b16dfa3f\",\"type\":\"StringFormatter\"},\"title\":\"Description\"},\"id\":\"9d00c0e1-dd6f-461e-a226-ba3b04e8e6cb\",\"type\":\"TableColumn\"},{\"attributes\":{\"editor\":{\"id\":\"f5521908-b3e2-4ffd-bfa5-7ec63bb8212d\",\"type\":\"StringEditor\"},\"field\":\"metric\",\"formatter\":{\"id\":\"c2939265-fff6-48ad-9a6f-623efd747e1e\",\"type\":\"StringFormatter\"},\"title\":\"Log Loss\"},\"id\":\"9961adef-8bd1-41d1-8bdb-274611d16ef1\",\"type\":\"TableColumn\"},{\"attributes\":{},\"id\":\"de7dcdd5-9486-485d-a1db-fc25e6fa2308\",\"type\":\"LinearScale\"},{\"attributes\":{\"plot\":{\"id\":\"746c3466-c6b5-4111-8e98-3ec46c90fe0b\",\"subtype\":\"Figure\",\"type\":\"Plot\"},\"ticker\":{\"id\":\"a409eb01-82dd-4b6d-810c-58605f848fef\",\"type\":\"BasicTicker\"}},\"id\":\"5aee2955-41aa-4456-9a68-463c83426d67\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"65ed03fa-4492-491e-9864-a845b16dfa3f\",\"type\":\"StringFormatter\"},{\"attributes\":{},\"id\":\"7330607b-ee1f-4282-84cf-0c99733431ed\",\"type\":\"StringEditor\"},{\"attributes\":{\"children\":[{\"id\":\"5e2eec85-4416-43d4-a09a-188fd5163a2b\",\"type\":\"DataTable\"}],\"height\":400,\"width\":600},\"id\":\"96ddc4f9-d7b2-497e-9d5d-30520db947fe\",\"type\":\"WidgetBox\"},{\"attributes\":{\"axis_label\":\"Size\",\"formatter\":{\"id\":\"643901d9-d907-498f-a7e0-eb77911cc3a5\",\"type\":\"BasicTickFormatter\"},\"plot\":{\"id\":\"746c3466-c6b5-4111-8e98-3ec46c90fe0b\",\"subtype\":\"Figure\",\"type\":\"Plot\"},\"ticker\":{\"id\":\"a409eb01-82dd-4b6d-810c-58605f848fef\",\"type\":\"BasicTicker\"}},\"id\":\"f87a2fe7-5530-4895-a0e9-55bef30d213d\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"33fbfcc8-694e-4812-a9c4-59baa6afa253\",\"type\":\"StringEditor\"},{\"attributes\":{\"editor\":{\"id\":\"33fbfcc8-694e-4812-a9c4-59baa6afa253\",\"type\":\"StringEditor\"},\"field\":\"effect_size\",\"formatter\":{\"id\":\"0e7b27a2-0c3f-4963-bcd6-44c928ce4ab2\",\"type\":\"StringFormatter\"},\"title\":\"Effect Size\"},\"id\":\"5aeb34fc-c514-4586-a9b9-f98d0547d5ad\",\"type\":\"TableColumn\"},{\"attributes\":{},\"id\":\"f5521908-b3e2-4ffd-bfa5-7ec63bb8212d\",\"type\":\"StringEditor\"},{\"attributes\":{\"source\":{\"id\":\"f4c2a054-30fb-4b8b-9556-cb4778ef4d95\",\"type\":\"ColumnDataSource\"}},\"id\":\"affcb953-26fb-4b45-9501-5f07952e78f9\",\"type\":\"CDSView\"}],\"root_ids\":[\"a55137e4-0802-47b5-a0e0-9f8f73667083\"]},\"title\":\"Bokeh Application\",\"version\":\"0.12.13\"}};\n",
       "  var render_items = [{\"docid\":\"c4338bdf-c63b-4f2b-8fe3-300626de2b8d\",\"elementid\":\"74ad42d1-871e-4fe2-9358-a4cd796f40f6\",\"modelid\":\"a55137e4-0802-47b5-a0e0-9f8f73667083\",\"notebook_comms_target\":\"32253aa7-a654-42e7-a2a4-eb01693eeda5\"}];\n",
       "  root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "\n",
       "  }\n",
       "  if (root.Bokeh !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined) {\n",
       "        embed_document(root);\n",
       "        clearInterval(timer);\n",
       "      }\n",
       "      attempts++;\n",
       "      if (attempts > 100) {\n",
       "        console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\")\n",
       "        clearInterval(timer);\n",
       "      }\n",
       "    }, 10, root)\n",
       "  }\n",
       "})(window);"
      ],
      "application/vnd.bokehjs_exec.v0+json": ""
     },
     "metadata": {
      "application/vnd.bokehjs_exec.v0+json": {
       "id": "a55137e4-0802-47b5-a0e0-9f8f73667083"
      }
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6c76b3aa200d4ab8bdaba0e107a483a1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(IntSlider(value=30, description='k', min=1), FloatSlider(value=0.3, description='min_eff…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "slices, uninteresting = list(), list()\n",
    "with open('log/slices.p','rb') as handle:\n",
    "    slices = pickle.load(handle)\n",
    "with open('log/uninteresting.p','rb') as handle:\n",
    "    uninteresting = pickle.load(handle)  \n",
    "\n",
    "def get_top_k_slices(candidates, min_effect_size):\n",
    "    description_ = list()\n",
    "    size_ = list()\n",
    "    effect_size_ = list()\n",
    "    metric_ = list()\n",
    "    for s in candidates:\n",
    "        if s.effect_size < min_effect_size:\n",
    "            continue\n",
    "            \n",
    "        description = ''\n",
    "        for k, v in list(s.filters.items()):\n",
    "            values = ''\n",
    "            if k in encoders:\n",
    "                le = encoders[k]\n",
    "                for v_ in v:\n",
    "                    values += '%s '%(le.inverse_transform(v_)[0])\n",
    "            else:\n",
    "                for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                    if len(v_) > 1:\n",
    "                        values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                    else:\n",
    "                        values += '%s '%(v_[0])\n",
    "            description += '%s:%s '%(k, values)\n",
    "        description_.append(description)\n",
    "        size_.append(s.size)\n",
    "        effect_size_.append(s.effect_size)\n",
    "        metric_.append(s.metric)\n",
    "    return description_, size_, effect_size_, metric_\n",
    "\n",
    "# load interesting slices\n",
    "description_, size_, effect_size_, metric_ = get_top_k_slices(slices, 0)\n",
    "    \n",
    "# load uninteresting slices (candidates)\n",
    "description_c, size_c, effect_size_c, metric_c = get_top_k_slices(uninteresting, 0)\n",
    "    \n",
    "data = dict(\n",
    "        description=description_,\n",
    "        size=size_,\n",
    "        effect_size=effect_size_,\n",
    "        metric=metric_,\n",
    "    )\n",
    "source = ColumnDataSource(data)\n",
    "\n",
    "# scatter plot\n",
    "hover = HoverTool(tooltips=[\n",
    "    (\"desc\", \"@description\"),\n",
    "    (\"size\", \"@size\"),\n",
    "    (\"effect_size\", \"@effect_size\"),\n",
    "    (\"metric\", \"@metric\"),\n",
    "])\n",
    "TOOLS = [hover]\n",
    "p = figure(tools=TOOLS, plot_width=400, plot_height=400, \n",
    "           y_axis_label='Effect Size', x_axis_label='Size', title=None)\n",
    "r = p.circle('size', 'effect_size', source=source)\n",
    "#show(p, notebook_handle=True)\n",
    "\n",
    "# data table\n",
    "columns = [\n",
    "        TableColumn(field=\"description\", title=\"Description\"),\n",
    "        TableColumn(field=\"size\", title=\"Size\"),\n",
    "        TableColumn(field=\"effect_size\", title=\"Effect Size\"),\n",
    "        TableColumn(field=\"metric\", title=\"Log Loss\"),\n",
    "    ]\n",
    "t = DataTable(source=r.data_source, columns=columns, width=600, height=400)\n",
    "\n",
    "show(row(p, t), notebook_handle=True)\n",
    "\n",
    "def update(k=30, min_eff_size=0.3):\n",
    "    desc_, size_, effect_size_, metric_ = get_top_k_slices(slices+uninteresting, min_eff_size)\n",
    "    idx = sorted(range(len(size_)), key=lambda x: size_[x], reverse=True)\n",
    "    r.data_source.data['description'] = np.array(desc_)[idx[:k]]\n",
    "    r.data_source.data['size'] = np.array(size_)[idx[:k]]\n",
    "    r.data_source.data['effect_size'] = np.array(effect_size_)[idx[:k]]\n",
    "    r.data_source.data['metric'] = np.array(metric_)[idx[:k]]\n",
    "    push_notebook()\n",
    "    \n",
    "w = interactive(update, k=(1,100), min_eff_size=(0, 1, 0.05))\n",
    "display(w)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preliminary Results\n",
    "Evaluate different baseline techniques as well as SliceFinder to see if they yield large problmatic slices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'slices' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-3-a3d8d0bc372c>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      9\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     10\u001b[0m \u001b[0msf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mSliceFinder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 11\u001b[0;31m \u001b[0msf_recs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0ms\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mslices\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0muninteresting\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0meffect_size\u001b[0m \u001b[0;34m>=\u001b[0m \u001b[0mmin_eff_size\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     12\u001b[0m \u001b[0msf_recs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msorted\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msf_recs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mlambda\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreverse\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'slices' is not defined"
     ]
    }
   ],
   "source": [
    "import copy\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import KMeans\n",
    "from risk_control import effect_size\n",
    "from decision_tree import DecisionTree\n",
    "\n",
    "min_eff_size = 0.4\n",
    "dataset = 'UCI'\n",
    "\n",
    "sf = SliceFinder(lr, (X, y))\n",
    "sf_recs = [s for s in slices+uninteresting if s.effect_size >= min_eff_size]\n",
    "sf_recs = sorted(sf_recs, key=lambda x: x.size, reverse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "numeric_cols = [\"Capital Gain\", \"Age\", \"fnlwgt\", \"Education-Num\", \"Capital Loss\"]\n",
    "X_ = copy.deepcopy(X)\n",
    "X_[numeric_cols] = scaler.fit_transform(X_[numeric_cols])\n",
    "reduced_data_train = PCA(n_components=2).fit_transform(X_)\n",
    "reduced_data_test = PCA(n_components=2).fit_transform(X_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "'list' object has no attribute 'as_matrix'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-11-d508408f8dbf>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      6\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m         \u001b[0mdecisions\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 8\u001b[0;31m \u001b[0mdt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mDecisionTree\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdecisions\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      9\u001b[0m \u001b[0mdt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmax_depth\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m6\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmin_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m100\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/slicefinder/decision_tree.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, data, model)\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mSliceFinder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m         \u001b[0mmetrics_all\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mevaluate_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmetric\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mlog_loss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     22\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreference\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmetrics_all\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstd\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmetrics_all\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmetrics_all\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     23\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/slicefinder/slice_finder.py\u001b[0m in \u001b[0;36mevaluate_model\u001b[0;34m(self, data, metric)\u001b[0m\n\u001b[1;32m    157\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    158\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mevaluate_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmetric\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mlog_loss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 159\u001b[0;31m         \u001b[0;34m''' evaluate model on a given data (X, y), example by example '''\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    160\u001b[0m         \u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mas_matrix\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mas_matrix\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    161\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mAttributeError\u001b[0m: 'list' object has no attribute 'as_matrix'"
     ]
    }
   ],
   "source": [
    "decisions = []\n",
    "for x_, y_ in zip(X.as_matrix(), y.as_matrix()):\n",
    "    y_p = lr.predict([x_])\n",
    "    if y_p == y_:\n",
    "        decisions.append(0)\n",
    "    else:\n",
    "        decisions.append(1)\n",
    "dt = DecisionTree((X, y), lr)\n",
    "dt = dt.fit(max_depth=6, min_size=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sf_k, sf_avg_sizes, sf_avg_eff_sizes = list(), list(), list()\n",
    "cluster_k, cluster_avg_sizes, cluster_avg_eff_sizes = list(), list(), list()\n",
    "tree_k, tree_avg_sizes, tree_avg_eff_sizes = list(), list(), list()\n",
    "\n",
    "for k in range(1, 11):\n",
    "    sf_recs_ = sf_recs[:k]\n",
    "    if len(sf_recs_) == k:\n",
    "        sf_k.append(k)\n",
    "        sf_avg_sizes.append(np.mean([s.size for s in sf_recs_]))\n",
    "        sf_avg_eff_sizes.append(np.mean([s.effect_size for s in sf_recs_]))\n",
    "    \n",
    "    kmeans = KMeans(init='k-means++', n_clusters=k)\n",
    "    kmeans.fit(reduced_data_train) # train with mis-classified data\n",
    "    y_predicted = kmeans.predict(reduced_data_test) # classify all data\n",
    "    sizes, eff_sizes = list(), list()\n",
    "    for cluster_id in np.unique(kmeans.labels_):\n",
    "        X_cluster = X[np.array(y_predicted) == cluster_id]\n",
    "        y_cluster = y[np.array(y_predicted) == cluster_id]\n",
    "        sizes.append(len(X_cluster))\n",
    "        eff_sizes.append(effect_size(sf.evaluate_model((X_cluster, y_cluster)), reference))\n",
    "    cluster_k.append(k)\n",
    "    cluster_avg_eff_sizes.append(np.mean(eff_sizes))\n",
    "    cluster_avg_sizes.append(np.mean(sizes))\n",
    "    \n",
    "    dt_recs = dt.recommend_slices(k=k, min_effect_size=min_eff_size)\n",
    "    if len(dt_recs) == k:\n",
    "        tree_k.append(k)\n",
    "        tree_avg_sizes.append(np.mean([s.size for s in dt_recs]))\n",
    "        tree_avg_eff_sizes.append(np.mean([s.eff_size for s in dt_recs]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "matplotlib.rcParams.update({'font.size': 17})\n",
    "\n",
    "plt.figure(1, figsize=(6,5))\n",
    "ax = plt.subplot(1,1,1)\n",
    "plt.plot(sf_k, np.array(sf_avg_sizes)/1000, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, np.array(cluster_avg_sizes)/1000, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, np.array(tree_avg_sizes)/1000, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Size (k)', fontsize=17)\n",
    "plt.savefig('%s_size_%s.pdf'%(dataset,min_eff_size))\n",
    "\n",
    "plt.figure(2, figsize=(6,5))\n",
    "ax = plt.subplot(1,1,1)\n",
    "plt.plot(sf_k, sf_avg_eff_sizes, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, cluster_avg_eff_sizes, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, tree_avg_eff_sizes, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Effect Size', fontsize=17)\n",
    "plt.tight_layout()\n",
    "plt.savefig('%s_effsize_%s.pdf'%(dataset,min_eff_size))\n",
    "plt.show()\n",
    "\n",
    "plt.figure(3, figsize=(12,5))\n",
    "#plt.suptitle('Quality of recommended slices from UCI Census Data Set', y=1.02)\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(sf_k, np.array(sf_avg_sizes)/1000, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, np.array(cluster_avg_sizes)/1000, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, np.array(tree_avg_sizes)/1000, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Size (k)', fontsize=17)\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(sf_k, sf_avg_eff_sizes, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, cluster_avg_eff_sizes, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, tree_avg_eff_sizes, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Effect Size', fontsize=17)\n",
    "#plt.subplots_adjust(left=0.2, wspace=0.8, top=0.8)\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Adjusting T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/yeounoh/git/slicefinder/risk_control.py:26: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  sample_b_mean = (mu*n - np.sum(sample_a))/(n-len(sample_a))\n"
     ]
    }
   ],
   "source": [
    "\n",
    "k = 10\n",
    "\n",
    "sf_k, sf_avg_sizes, sf_avg_eff_sizes = list(), list(), list()\n",
    "cluster_k, cluster_avg_sizes, cluster_avg_eff_sizes = list(), list(), list()\n",
    "tree_k, tree_avg_sizes, tree_avg_eff_sizes = list(), list(), list()\n",
    "\n",
    "for min_eff_size in np.array(range(1,9))/10.:\n",
    "    sf = SliceFinder(lr, (X, y))\n",
    "    sf_recs = [s for s in slices+uninteresting if s.effect_size >= min_eff_size]\n",
    "    sf_recs = sorted(sf_recs, key=lambda x: x.size, reverse=True)\n",
    "    \n",
    "    sf_recs_ = sf_recs[:k]\n",
    "    if len(sf_recs_) == k:\n",
    "        sf_k.append(min_eff_size)\n",
    "        sf_avg_sizes.append(np.mean([s.size for s in sf_recs_]))\n",
    "        sf_avg_eff_sizes.append(np.mean([s.effect_size for s in sf_recs_]))\n",
    "    \n",
    "    kmeans = KMeans(init='k-means++', n_clusters=k)\n",
    "    kmeans.fit(reduced_data_train) # train with mis-classified data\n",
    "    y_predicted = kmeans.predict(reduced_data_test) # classify all data\n",
    "    sizes, eff_sizes = list(), list()\n",
    "    for cluster_id in np.unique(kmeans.labels_):\n",
    "        X_cluster = X[np.array(y_predicted) == cluster_id]\n",
    "        y_cluster = y[np.array(y_predicted) == cluster_id]\n",
    "        sizes.append(len(X_cluster))\n",
    "        eff_sizes.append(effect_size(sf.evaluate_model((X_cluster, y_cluster)), reference))\n",
    "    cluster_k.append(min_eff_size)\n",
    "    cluster_avg_eff_sizes.append(np.mean(eff_sizes))\n",
    "    cluster_avg_sizes.append(np.mean(sizes))\n",
    "    \n",
    "    dt_recs = dt.recommend_slices(k=k, min_effect_size=min_eff_size)\n",
    "    if len(dt_recs) == k:\n",
    "        tree_k.append(min_eff_size)\n",
    "        tree_avg_sizes.append(np.mean([s.size for s in dt_recs]))\n",
    "        tree_avg_eff_sizes.append(np.mean([s.eff_size for s in dt_recs]))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x116c89748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11cbc6b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12075e940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(1, figsize=(6,5))\n",
    "ax = plt.subplot(1,1,1)\n",
    "plt.plot(sf_k, np.array(sf_avg_sizes)/1000, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "#plt.plot(cluster_k, np.array(cluster_avg_sizes)/1000, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, np.array(tree_avg_sizes)/1000, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Min. Effect Size (T)', fontsize=17)\n",
    "plt.ylabel('Avg. Size (thousands)', fontsize=17)\n",
    "plt.savefig('%s_size_%s_2.pdf'%(dataset,min_eff_size))        \n",
    "        \n",
    "plt.figure(2, figsize=(6,5))\n",
    "ax = plt.subplot(1,1,1)\n",
    "plt.plot(sf_k, sf_avg_eff_sizes, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "#plt.plot(cluster_k, cluster_avg_eff_sizes, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, tree_avg_eff_sizes, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Min. Effect Size (T)', fontsize=17)\n",
    "plt.ylabel('Avg. Effect Size', fontsize=17)\n",
    "plt.tight_layout()\n",
    "plt.savefig('%s_effsize_%s_2.pdf'%(dataset,min_eff_size))\n",
    "plt.show()\n",
    "\n",
    "plt.figure(3, figsize=(12,5))\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(sf_k, np.array(sf_avg_sizes)/1000, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "#plt.plot(cluster_k, np.array(cluster_avg_sizes)/1000, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, np.array(tree_avg_sizes)/1000, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Min. Effect Size (T)', fontsize=17)\n",
    "plt.ylabel('Avg. Size (thousands)', fontsize=17)\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(sf_k, sf_avg_eff_sizes, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "#plt.plot(cluster_k, cluster_avg_eff_sizes, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, tree_avg_eff_sizes, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Min. Effect Size (T)', fontsize=17)\n",
    "plt.ylabel('Avg. Effect Size', fontsize=17)\n",
    "#plt.subplots_adjust(left=0.2, wspace=0.8, top=0.8)\n",
    "plt.savefig('%s_slider.pdf'%(dataset))\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fraudulent Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import copy\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import KMeans\n",
    "from risk_control import effect_size\n",
    "from decision_tree import DecisionTree\n",
    "\n",
    "#main_df=pd.read_csv(\"creditcard.csv\")\n",
    "main_df=pd.read_csv(\"data/creditcard.csv\")\n",
    "main_df.head()\n",
    "feature_size = len(main_df.columns)\n",
    "class_index = feature_size -1 \n",
    "\n",
    "fraud_raw = main_df[main_df['Class'] == 1]\n",
    "normal_raw = main_df[main_df['Class'] == 0]\n",
    "\n",
    "# Undersample the normal transactions\n",
    "percentage = len(fraud_raw)/float(len(normal_raw))\n",
    "normal = normal_raw.sample(frac=percentage)\n",
    "fraud = fraud_raw\n",
    "cc_data = fraud.append(normal)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#msk = np.random.rand(len(cc_data)) < 0.8\n",
    "X, y = cc_data[cc_data.columns.difference([\"Class\"])], cc_data[\"Class\"]\n",
    "#X_train, y_train = X[msk], y[msk]\n",
    "#X, y = X[~msk], y[~msk]\n",
    "\n",
    "reg_model = RandomForestClassifier(criterion='entropy',n_estimators=100)\n",
    "lr = reg_model.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "sf = SliceFinder(lr, (X, y))\n",
    "metrics_all = sf.evaluate_model((X,y))\n",
    "reference = (np.mean(metrics_all), np.std(metrics_all), len(metrics_all))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V17:0.512362754277 ~ 1.17193723459\n",
      "---------------------\n",
      "effect_size: 0.730149341476\n",
      "---------------------\n",
      "metric: 0.104595746695\n",
      "size: 99\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V18:0.487793259179 ~ 1.01170928247\n",
      "---------------------\n",
      "effect_size: 0.45922869468\n",
      "---------------------\n",
      "metric: 0.0793451792101\n",
      "size: 99\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V21:-0.100400181768 ~ 0.0137410881749\n",
      "---------------------\n",
      "effect_size: 0.427219694911\n",
      "---------------------\n",
      "metric: 0.0773811381036\n",
      "size: 99\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V3:-0.643387104927 ~ 0.0259153271539\n",
      "---------------------\n",
      "effect_size: 0.452310831997\n",
      "---------------------\n",
      "metric: 0.07986624206\n",
      "size: 99\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V3:0.60905131525 ~ 1.20643458984\n",
      "---------------------\n",
      "effect_size: 0.471839184916\n",
      "---------------------\n",
      "metric: 0.0825898737282\n",
      "size: 99\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V5:0.774036846205 ~ 1.51151331132\n",
      "---------------------\n",
      "effect_size: 0.499900215842\n",
      "---------------------\n",
      "metric: 0.0816519366196\n",
      "size: 99\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V9:0.45975256166 ~ 1.00923259801\n",
      "---------------------\n",
      "effect_size: 0.426599826816\n",
      "---------------------\n",
      "metric: 0.0768673275498\n",
      "size: 99\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V10:-2.12689408851 ~ -0.871535563396\n",
      "---------------------\n",
      "effect_size: 0.685037384736\n",
      "---------------------\n",
      "metric: 0.100822349331\n",
      "size: 98\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V12:-0.796971157322 ~ -0.287248272598\n",
      "---------------------\n",
      "effect_size: 0.689830991085\n",
      "---------------------\n",
      "metric: 0.101132046926\n",
      "size: 98\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V14:-3.65342823258 ~ -1.19501632601\n",
      "---------------------\n",
      "effect_size: 0.933568628984\n",
      "---------------------\n",
      "metric: 0.117196723339\n",
      "size: 98\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V14:-1.19501632601 ~ -0.47276337572\n",
      "---------------------\n",
      "effect_size: 0.656142962932\n",
      "---------------------\n",
      "metric: 0.0987573824644\n",
      "size: 98\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V2:-18.5017229552 ~ -1.02092630363\n",
      "---------------------\n",
      "effect_size: 0.427461157031\n",
      "---------------------\n",
      "metric: 0.0750236843637\n",
      "size: 98\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V27:0.00395815659095 ~ 0.0499363242282\n",
      "---------------------\n",
      "effect_size: 0.412696829214\n",
      "---------------------\n",
      "metric: 0.0774061853577\n",
      "size: 98\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V28:0.0308575374649 ~ 0.0818269775365\n",
      "---------------------\n",
      "effect_size: 0.44278667524\n",
      "---------------------\n",
      "metric: 0.0797402435978\n",
      "size: 98\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Amount:268.36 ~ 4097.4\n",
      "---------------------\n",
      "effect_size: 0.444362412666\n",
      "---------------------\n",
      "metric: 0.0802118633321\n",
      "size: 97\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V10:0.674547742092 ~ 6.91375329521\n",
      "---------------------\n",
      "effect_size: 0.540916163595\n",
      "---------------------\n",
      "metric: 0.0832807242556\n",
      "size: 97\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V16:0.868552978087 ~ 4.87036097505\n",
      "---------------------\n",
      "effect_size: 0.465183492195\n",
      "---------------------\n",
      "metric: 0.0788714967035\n",
      "size: 97\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V17:1.17193723459 ~ 6.73938438478\n",
      "---------------------\n",
      "effect_size: 0.729892849399\n",
      "---------------------\n",
      "metric: 0.0991177165815\n",
      "size: 97\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V18:1.01170928247 ~ 3.79031621184\n",
      "---------------------\n",
      "effect_size: 0.422838950729\n",
      "---------------------\n",
      "metric: 0.074006691686\n",
      "size: 97\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V7:0.944240710124 ~ 19.7753572833\n",
      "---------------------\n",
      "effect_size: 0.709151367024\n",
      "---------------------\n",
      "metric: 0.102639111879\n",
      "size: 97\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V28:-0.0398842606352 ~ 0.00788012753403\n",
      "V1:1.83508064351 ~ 2.37960943156\n",
      "---------------------\n",
      "effect_size: 0.521962595151\n",
      "---------------------\n",
      "metric: 0.0880254990073\n",
      "size: 36\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V8:-41.0442609211 ~ -0.726205400649\n",
      "V20:-15.4489859436 ~ -0.506508975195\n",
      "---------------------\n",
      "effect_size: 0.752769118439\n",
      "---------------------\n",
      "metric: 0.108373951898\n",
      "size: 30\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V4:0.217491690115 ~ 0.709334440778\n",
      "V28:-0.0398842606352 ~ 0.00788012753403\n",
      "---------------------\n",
      "effect_size: 0.618122660616\n",
      "---------------------\n",
      "metric: 0.10348192948\n",
      "size: 26\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Amount:19.075 ~ 40.9384\n",
      "V28:-0.0398842606352 ~ 0.00788012753403\n",
      "---------------------\n",
      "effect_size: 0.476109855784\n",
      "---------------------\n",
      "metric: 0.0917675529151\n",
      "size: 25\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "Amount:2.28 ~ 8.7688\n",
      "V2:0.950025213823 ~ 1.41606246591\n",
      "---------------------\n",
      "effect_size: 0.461576137921\n",
      "---------------------\n",
      "metric: 0.0851428264828\n",
      "size: 24\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V6:0.336493830306 ~ 1.15197950482\n",
      "V12:0.781529939641 ~ 4.06062850579\n",
      "---------------------\n",
      "effect_size: 0.407656875029\n",
      "---------------------\n",
      "metric: 0.08373890123\n",
      "size: 24\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V7:0.420629513659 ~ 0.944240710124\n",
      "V16:-0.206176897948 ~ 0.118558369548\n",
      "---------------------\n",
      "effect_size: 0.613620090303\n",
      "---------------------\n",
      "metric: 0.100878442754\n",
      "size: 24\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V26:-0.532686620666 ~ -0.363908862687\n",
      "V4:1.28773551195 ~ 2.33007026464\n",
      "---------------------\n",
      "effect_size: 0.505440800848\n",
      "---------------------\n",
      "metric: 0.0836104931481\n",
      "size: 24\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V8:-0.118273335038 ~ 0.016036877881\n",
      "V27:-0.041784292854 ~ 0.00395815659095\n",
      "---------------------\n",
      "effect_size: 0.419422526766\n",
      "---------------------\n",
      "metric: 0.082660718755\n",
      "size: 24\n",
      "\n",
      "=====================\n",
      "Slice description:\n",
      "V8:-41.0442609211 ~ -0.726205400649\n",
      "V3:-4.15534639703 ~ -2.29974072867\n",
      "---------------------\n",
      "effect_size: 0.409842928311\n",
      "---------------------\n",
      "metric: 0.078447634564\n",
      "size: 24\n"
     ]
    }
   ],
   "source": [
    "recommendations = sf.find_slice(k=30, epsilon=0.4, degree=5)\n",
    "encoders = {}\n",
    "for s in recommendations:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    for k, v in list(s.filters.items()):\n",
    "        values = ''\n",
    "        if k in encoders:\n",
    "            le = encoders[k]\n",
    "            for v_ in v:\n",
    "                values += '%s '%(le.inverse_transform(v_)[0])\n",
    "        else:\n",
    "            for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                if len(v_) > 1:\n",
    "                    values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                else:\n",
    "                    values += '%s '%(v_[0])\n",
    "        print ('%s:%s'%(k, values))\n",
    "    print ('---------------------\\neffect_size: %s'%(s.effect_size))\n",
    "    print ('---------------------\\nmetric: %s'%(s.metric))\n",
    "    print ('size: %s'%(s.size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "slices, uninteresting = list(), list()\n",
    "with open('slices.p','rb') as handle:\n",
    "    slices = pickle.load(handle)\n",
    "with open('uninteresting.p','rb') as handle:\n",
    "    uninteresting = pickle.load(handle)  \n",
    "    \n",
    "min_eff_size = 0.4\n",
    "dataset = 'Fraud'\n",
    "\n",
    "sf_recs = [s for s in slices+uninteresting if s.effect_size >= min_eff_size]\n",
    "sf_recs = sorted(sf_recs, key=lambda x: x.size, reverse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_slices, rejected = sf.filter_by_significance(slices, reference, 0.05)\n",
    "print (len(filtered_slices), len(rejected))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'copy' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-11-a3f8d95bf7eb>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mscaler\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mStandardScaler\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mnumeric_cols\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmain_df\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;36m28\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mX_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdeepcopy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m \u001b[0mX_\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mnumeric_cols\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mscaler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_transform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mnumeric_cols\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mreduced_data_train\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mPCA\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mn_components\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_transform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'copy' is not defined"
     ]
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "numeric_cols = list(main_df.columns[1:28])\n",
    "X_ = copy.deepcopy(X)\n",
    "X_[numeric_cols] = scaler.fit_transform(X_[numeric_cols])\n",
    "reduced_data_train = PCA(n_components=2).fit_transform(X_)\n",
    "reduced_data_test = PCA(n_components=2).fit_transform(X_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test root, size: 984\n"
     ]
    }
   ],
   "source": [
    "decisions = []\n",
    "for x_, y_ in zip(X.as_matrix(), y.as_matrix()):\n",
    "    y_p = lr.predict([x_])\n",
    "    if y_p == y_:\n",
    "        decisions.append(0)\n",
    "    else:\n",
    "        decisions.append(1)\n",
    "dt = DecisionTree((X, y), lr)\n",
    "dt = dt.fit(max_depth=20, min_size=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sf_k, sf_avg_sizes, sf_avg_eff_sizes = list(), list(), list()\n",
    "cluster_k, cluster_avg_sizes, cluster_avg_eff_sizes = list(), list(), list()\n",
    "tree_k, tree_avg_sizes, tree_avg_eff_sizes = list(), list(), list()\n",
    "\n",
    "for k in range(1, 11):\n",
    "    sf_recs_ = sf_recs[:k]\n",
    "    if len(sf_recs_) == k:\n",
    "        sf_k.append(k)\n",
    "        sf_avg_sizes.append(np.mean([s.size for s in sf_recs_]))\n",
    "        sf_avg_eff_sizes.append(np.mean([s.effect_size for s in sf_recs_]))\n",
    "    \n",
    "    kmeans = KMeans(init='k-means++', n_clusters=k)\n",
    "    kmeans.fit(reduced_data_train) # train with mis-classified data\n",
    "    y_predicted = kmeans.predict(reduced_data_test) # classify all data\n",
    "    sizes, eff_sizes = list(), list()\n",
    "    for cluster_id in np.unique(kmeans.labels_):\n",
    "        X_cluster = X[np.array(y_predicted) == cluster_id]\n",
    "        y_cluster = y[np.array(y_predicted) == cluster_id]\n",
    "        sizes.append(len(X_cluster))\n",
    "        eff_sizes.append(effect_size(sf.evaluate_model((X_cluster, y_cluster)), reference))\n",
    "    cluster_k.append(k)\n",
    "    cluster_avg_eff_sizes.append(np.mean(eff_sizes))\n",
    "    cluster_avg_sizes.append(np.mean(sizes))\n",
    "    \n",
    "    dt_recs = dt.recommend_slices(k=k, min_effect_size=min_eff_size)\n",
    "    if len(dt_recs) == k:\n",
    "        tree_k.append(k)\n",
    "        tree_avg_sizes.append(np.mean([s.size for s in dt_recs]))\n",
    "        tree_avg_eff_sizes.append(np.mean([s.eff_size for s in dt_recs]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "matplotlib.rcParams.update({'font.size': 17})\n",
    "\n",
    "plt.figure(1, figsize=(6,5))\n",
    "ax = plt.subplot(1,1,1)\n",
    "plt.plot(sf_k, np.array(sf_avg_sizes)/100, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, np.array(cluster_avg_sizes)/100, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, np.array(tree_avg_sizes)/100, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Size (hundreds)', fontsize=17)\n",
    "plt.savefig('%s_size%s.pdf'%(dataset,min_eff_size))\n",
    "\n",
    "plt.figure(2, figsize=(6,5))\n",
    "ax = plt.subplot(1,1,1)\n",
    "plt.plot(sf_k, sf_avg_eff_sizes, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, cluster_avg_eff_sizes, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, tree_avg_eff_sizes, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Effect Size', fontsize=17)\n",
    "plt.tight_layout()\n",
    "plt.savefig('%s_effsize_%s.pdf'%(dataset,min_eff_size))\n",
    "plt.show()\n",
    "\n",
    "plt.figure(3, figsize=(12,5))\n",
    "#plt.suptitle('Quality of recommended slices from UCI Census Data Set', y=1.02)\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(sf_k, np.array(sf_avg_sizes)/100, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, np.array(cluster_avg_sizes)/100, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, np.array(tree_avg_sizes)/100, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Size (hundreds)', fontsize=17)\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(sf_k, sf_avg_eff_sizes, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "plt.plot(cluster_k, cluster_avg_eff_sizes, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, tree_avg_eff_sizes, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Numb. of Recommendations', fontsize=17)\n",
    "plt.ylabel('Avg. Effect Size', fontsize=17)\n",
    "#plt.subplots_adjust(left=0.2, wspace=0.8, top=0.8)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/yeounoh/git/slicefinder/risk_control.py:26: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  sample_b_mean = (mu*n - np.sum(sample_a))/(n-len(sample_a))\n"
     ]
    }
   ],
   "source": [
    "\n",
    "k = 10\n",
    "\n",
    "sf_k, sf_avg_sizes, sf_avg_eff_sizes = list(), list(), list()\n",
    "cluster_k, cluster_avg_sizes, cluster_avg_eff_sizes = list(), list(), list()\n",
    "tree_k, tree_avg_sizes, tree_avg_eff_sizes = list(), list(), list()\n",
    "\n",
    "for min_eff_size in np.array(range(1,9))/10.:\n",
    "    sf = SliceFinder(lr, (X, y))\n",
    "    sf_recs = [s for s in slices+uninteresting if s.effect_size >= min_eff_size]\n",
    "    sf_recs = sorted(sf_recs, key=lambda x: x.size, reverse=True)\n",
    "    \n",
    "    sf_recs_ = sf_recs[:k]\n",
    "    if len(sf_recs_) == k:\n",
    "        sf_k.append(min_eff_size)\n",
    "        sf_avg_sizes.append(np.mean([s.size for s in sf_recs_]))\n",
    "        sf_avg_eff_sizes.append(np.mean([s.effect_size for s in sf_recs_]))\n",
    "    \n",
    "   \n",
    "    dt_recs = dt.recommend_slices(k=k, min_effect_size=min_eff_size)\n",
    "    if len(dt_recs) == k:\n",
    "        tree_k.append(min_eff_size)\n",
    "        tree_avg_sizes.append(np.mean([s.size for s in dt_recs]))\n",
    "        tree_avg_eff_sizes.append(np.mean([s.eff_size for s in dt_recs]))\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10cd7ecf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "matplotlib.rcParams.update({'font.size': 17})\n",
    "\n",
    "plt.figure(3, figsize=(12,5))\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(sf_k, np.array(sf_avg_sizes)/100, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "#plt.plot(cluster_k, np.array(cluster_avg_sizes)/100, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, np.array(tree_avg_sizes)/100, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Min. Effect Size (T)', fontsize=17)\n",
    "plt.ylabel('Avg. Size (hundreds)', fontsize=17)\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(sf_k, sf_avg_eff_sizes, 'o-', label='LS', linewidth=3.0, markersize=10)\n",
    "#plt.plot(cluster_k, cluster_avg_eff_sizes, 's-', label='CL', linewidth=3.0, markersize=10)\n",
    "plt.plot(tree_k, tree_avg_eff_sizes, 'v-', label='DT', linewidth=3.0, markersize=10)\n",
    "plt.legend()\n",
    "plt.xlabel('Min. Effect Size (T)', fontsize=17)\n",
    "plt.ylabel('Avg. Effect Size', fontsize=17)\n",
    "#plt.subplots_adjust(left=0.2, wspace=0.8, top=0.8)\n",
    "plt.savefig('%s_slider.pdf'%(dataset))\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Top-3 slices for LS and DT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#sf_recs = sorted(sf_recs, key=lambda x: x.size, reverse=True)\n",
    "for s in sf_recs[:3]:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    for k, v in list(s.filters.items()):\n",
    "        values = ''\n",
    "        if k in encoders:\n",
    "            le = encoders[k]\n",
    "            for v_ in v:\n",
    "                values += '%s '%(le.inverse_transform(v_)[0])\n",
    "        else:\n",
    "            for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                if len(v_) > 1:\n",
    "                    values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                else:\n",
    "                    values += '%s '%(v_[0])\n",
    "        print ('%s:%s'%(k, values))\n",
    "    print ('---------------------\\neffect_size: %s'%(s.effect_size))\n",
    "    print ('---------------------\\nmetric: %s'%(s.metric))\n",
    "    print ('size: %s'%(s.size))\n",
    "    \n",
    "for s in dt_recs[:3]:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    print (s.__ancestry__())\n",
    "    print (s.__str__())\n",
    "    print ('---------------------\\neffect_size: %s'%(s.eff_size))\n",
    "    print ('size: %s'%(s.size))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Simulation Study\n",
    "Emphasize that we are not just accuracy-hungting (i.e., give me slices with lowest performance?? The challenge is automatically defining slices, because there are just so many ways to slice the data); to this end, show how the performance distributions change over differnet numbers of clusters--connect this to the original motiviation, where the overall perofrmance differs from smaller slice performances.\n",
    "\n",
    "Comment: model accuracy distributions? What happens when we add two additional dimensions, like size and interpretability? We need an interactive tool to assist users to quickly browse through problematic data slices for model validation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature variance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_hat = lr.predict(X)\n",
    "X_mis = X[y != y_hat]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "min_max_scaler = preprocessing.MinMaxScaler()\n",
    "X_scaled = pd.DataFrame(\n",
    "    min_max_scaler.fit_transform(X_mis.values), \n",
    "    columns=X.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13]),\n",
       " <a list of 14 Text xticklabel objects>)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e457c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams.update({'font.size': 22})\n",
    "plt.figure(figsize=(12,8))\n",
    "ax = sns.boxplot(data=X_scaled, palette=\"Set3\")\n",
    "plt.ylabel('Normalized Feature Value')\n",
    "plt.xticks(rotation=45)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Accuracy-hunting approach\n",
    "If we simply look for data slices with the lowest performance metrics, then we might end up with 1) slices that are too small or 2) too complicated to summarize and manually interpret. This is aside from the fact that we want to avoid presenting a numerous number of data slices to the user; to this end, we propose both a metric to identify a problematic slice for model validation and efficient top-down data slicing strategies that bring user's attention to a handful of the most problematic slices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "adult_data2 = pd.read_csv(\n",
    "    \"data/adult.data\",\n",
    "    names=[\n",
    "        \"Age\", \"Workclass\", \"fnlwgt\", \"Education\", \"Education-Num\", \"Martial Status\",\n",
    "        \"Occupation\", \"Relationship\", \"Race\", \"Sex\", \"Capital Gain\", \"Capital Loss\",\n",
    "        \"Hours per week\", \"Country\", \"Target\"],\n",
    "        sep=r'\\s*,\\s*',\n",
    "        engine='python',\n",
    "        na_values=\"?\")\n",
    "\n",
    "# drop nan values\n",
    "adult_data2 = adult_data2.dropna()\n",
    "adult_mis = adult_data2[y!=y_hat]\n",
    "#print(adult_mis.dtypes)\n",
    "df_edu_sex = pd.crosstab(adult_mis['Education'],adult_mis['Sex'])\n",
    "\n",
    "#mosaic(df_edu_sex.stack(), title='Education-Sex')\n",
    "#plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113d40ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(2,2,figsize=(8,6))\n",
    "adult_data2.plot.scatter('Age','Hours per week', color='yellow', label='Correct', ax=axes[0][0])\n",
    "adult_mis.plot.scatter('Age','Hours per week', color='red', label='Incorrect',ax=axes[0][0])\n",
    "\n",
    "adult_data2.plot.scatter('Education-Num','Hours per week', color='yellow', label='Correct', ax=axes[0][1])\n",
    "adult_mis.plot.scatter('Education-Num','Hours per week', color='red', label='Incorrect',ax=axes[0][1])\n",
    "\n",
    "adult_data2.plot.scatter('Capital Gain','Capital Loss', color='yellow', label='Correct', ax=axes[1][0])\n",
    "adult_mis.plot.scatter('Capital Gain','Capital Loss', color='red', label='Incorrect',ax=axes[1][0])\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x113d34940>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdIAAAG2CAYAAADLBQd6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzs3Xe8FNX5x/HPFxBRUBTBbsSusZdoYosaSzT2/NRoQNQYYqLGGns3lthjw4DGjsaGLVZsWGMBS6xBxQYiilgQEO59fn+cs9xx2TJ37/b7vH3Na7kzZ2ae5eI+e86cIjPDOeecc6XpUusAnHPOuUbmidQ555zrAE+kzjnnXAd4InXOOec6wBOpc8451wGeSJ1zzrkO8ETqnHPOdYAnUueccw1PUi9Jp0q6T9IkSSbplHZeY1tJ/5E0TdKnki6W1KvYeZ5InXPONYO+wEnAGsDo9p4saSvgXqAFOBS4GhgM3FHs3G7tvZlzzjlXhyYAS5jZeElLAh+18/wLgLHA5mY2A0DSu8AwSdub2b35TvQaqXPOuYZnZjPMbHwp50paBVgNGJZJotF1wLfAHoXO90TqnHOus1snvj6f3Glm3wMvJ47n5E27DoBu3ZdoutULlp5/kVqHUBGTp39d6xDKrn+v5vxdfTr9y1qHUBETpryhjl5j5ufvpf7M6d5vuT8QnldmDDWzoR2NIWGx+Dohx7EJhNpqXp5Im4CkzYDHgIFmdkONw3HOubKKSbOciTPbPPF1Ro5j0xPHc/Km3QppT1dsSfNLujR2t54m6bnYgyy73D6S/lzx4J1zrtJaZqbfKm9afJ07x7EeieM5eSKtnFRdsSUJuAf4HXAVods1wH2Sfp5VfB/AE6lzrvG1tqbfKi/TpLt4jmOLAQU7MXnTbuWk7Yq9K7ApsK+ZXQMg6Vrgv8D5wHpViNU556rKrCoJMq1MZecnwBOZnZK6A2sBdxY62WukFdKOrti7A1OA2c82zWw6oXa6rqTlACSNA34OLBebiS3uS5KkIyR9IGl6bCIu2NvMOedqokY1UknzSlpZUt/MPjN7E3gD+L2kZPPu3kAv4NZC1/Qaae2tA4wxs1lZ+59PHH+X0OR7FrAgcGQ89m3WOYcQ2vgvJvxu/wKMkLS8mVXlQYNzzqVSgRqppIOABYD5465NJZ0Q/3y9mX0ArE/onHkqcEri9COAfwOPxlbB/sDhwCOEx295eSKtvcXIGrsU/aDN3szulHQoMFeBnrl9gFXNbBqApLeBEcDWhH8gPyBpMLFLubr2pkuXnh15H845l15rSyWueiSwdOLnzeMG8BTwQb4TzewBSTsQEuzFwFfAlcCxZlZwqI4n0tqbh/xdrjPH07omk0SjTFv/srkKJ7uUN+M4UudcHWvJboTrODPrn6LM40DOcbBmdh9wX3vv64m09qaRv8t15nhaP/i2ZWZfhk7B9CktNOecq4w662zUIZ5Ia28C+btcQ5Fu11nytZV0eBYS55wrq+oMa6kKT6S1NxrYWlK3rA5HG8TXMYl93vzqnGsOTVQj9eEvtXcboZfZgMwOST2A/Qi9eccmyk6NZZ1zrrG1tqTf6pzXSCsoZVfs2wm9yYZIWp4wccM+hK7X2dMEvgT8StJ58c/fmlnBbtlpLd6r+R6jfvD1xFqHUBFzd5ur1iGU3WuTx9U6hIqYf+55ax1C/apAZ6Na8URaWUW7YptZq6TtCWNEf09Iuq8BvzKzx7KudwGwMmE6wSMInYvKkkidc66qmqhp1xNpBaXpih3LfQX8KW7Fys2xwGyR7tze0cg5V3+8s5FzzjlXOrP6f/aZlidS55xz1edNu84551wHeNOuc8451wHVWbC7KjyROuecqz5v2nXOOec6wJt2nXPOuQ7wGqlzzjnXAV4jdc455zrAE6lzzjlXOvNeu84551wH+DNS12zGfzu51iGU3d6L/6zWIVTEiM9frnUIZTeD5qmdJM3dtflW6imbJmra9fVIm4Ck/pIssUSbc87VN2tNv9U5T6QVIqmXpFMl3SdpUkx0p+Qot4WkKyW9Jek7SR9I+pekFXOU3VnSSVV5A845V0mtrem3OudNu5XTFzgJ+AQYDWydp9w5wEKEBb7fAhYHDgRelrSRmY1JlN0ZGACcVqmgnXOuKnxhb5fCBGAJMxsvaUngozzlDgeeMmtrv5A0nLC494nArhWP1Dnnqq0BapppedNuhZjZDDMbn6LcqGQSjfvGAq8CP87sk/Q4MAjoGpuJTZJlX0/S3rGZeIak1yRt1eE345xz5dZEz0i9RlqHJAlYFPg4sfsMwhefjQgJNZddCE3KVwDTgEOBEZJ+ZGbN1y3XOde4mqhG6om0Pv0W+BFwXmaHmT0s6bfAhmZ2Q57zlgdWNLOJMLsWOwbYE7gsu7CkwcBgAHXtTZcuPcv5HpxzLr8GqGmm5Ym0zkhaFbgceBEY0s7Tb88kUQAze1nS18CyuQqb2VBgKEC37kvM0UzsnHMV4zVSVwmSlgLuB6YAO5tZe7u1fZBj35dAn47G5pxzZeW9dl25SeoHPATMA2xiZp+UcJmWfJcvOTDnnKsEr5G6cpLUG3gQWALYwszeylPUm1+dc83BmufjzIe/1JikeYH7gFWAHc3sxQLFpxKGv/SqSnDOOVcpZZ7ZSFJ3SadL+lDSdEmvStoz5bnzSDpO0htxhrnxku6WtEGa871GWkGSDgIWAOaPuzZNzId7vZl9ANwIbAjcBCwpaUDyGlk9dF+Kr5dIGgm0mNnNFXsDzjlXKeVv2r0K2IvQWfNVwmQ2wyV1MbMbi5x7C7AtcCVwIdAPOAB4Ms4w90Khkz2RVtaRwNKJnzePG8BThM5Ba8ef94xbthuy/vxTwj+QQYRnn55I87hu/LO1DqEiFu/VfH3Hvvl+Wq1DqIipM6fXOoT6VcbhL5LWJUyfeqqZnRL3XQmMAs6TdIuZ5VxiSNKywPbABWZ2RGL/ncDr8bqeSGvFzPqXo0yi7EzgD3FL7h9Hng5F7bm+c85VTUu+vpEl2Z3Qh2T2eHkzM0mXA8OBTYFH8pybaTGckLU/8/N3xW7uz0idc85VX3mfka4DjDOzSVn7n08cz+ctYCJwZFxha0lJ6wDXAZOIY+0L8UTqnHOu+tqRSCUNlvRiYhucdbXFmLNGSWLf4vnCMLPphMdl3wIjCAuMvASsAPzMzN4v9la8adc551z1teMZaXIWtjzmAT7LsX964nghUwjLXd5EmFVuaeBo4N+SNslR0/0BT6TOOeeqzlrLOo50GjB3jv09EsdzistcPgv82cyuTex/mND79zjgsEI390TqnHOu+so7ReAEwqId2RaLr4WWtPwdocPRiOROM3tT0pvAJsVu7s9InXPOVV+rpd+KGw0sHadaTdogcTyfzPPTrjmOdSNFhdMTqXPOueorb6/d2whDAA/M7IjrOh9A6JE7Ku7rLWnlOC1rxtvx9QeT4UhaD1iJtolw8vKmXeecc9VXxpmNzOwFScOBEyX1oW1mo02AQYnJGHYBrgb2Ba6J+64BDgculLQGYfKF/oSkPI3EutD5eCJ1zjlXfeWftH4/YBywN2HSmneAAcWmBzSzybH2eSKwFTCQMK/5Y8BJZvZmsRt7InXOOVd9ZZ5r18xmAMfHLV+Za2iriSb3f0qiWbi9PJE655yrvvJOEVhTnkgdAM8t/JNah1B2B7R8UesQKuLlL96rdQgupdYmWnOz7Mo7jrSmvNduiST1l2SJZdFqGcvjksbWOg7nnEvLWltTb/Wu0yZSSZvFRJhrG5Aot7Okkyocy7ySTpP0TlyQdqKk+yT1T5TpL+mU2KvMOecaW3nHkdaUN+3CEOCZrH1PJ/68M2F80WmVuLmkXsCjhAmShxFWIlgQWB/oQ+iFBqE79snAWELXbueca1xlXI+01jyRwjNmdkPxYhVzBrAisJ6ZefOsc65zaICaZlqdtmk3SVIvSXPl2P84MAjommz6zVFub0lvSZoh6TVJW6W87/zA74FhZjZW0lyS5lilQNI+hDFNANcnYtknq9zykh6UNFXSZ5LOluS/Y+dc/ZnVkn6rc/4hC5cD3wAzJP0nKwmeATwJtBIG6Wa2pF2A04FrCcvuzAeMiLNrFLMJYXmftyXdTFiJ/TtJYyRtnig3Cjgz/nlIIo5RiTLzASOB94AjCM3TRxMStXPO1RdrTb/Vuc6cSGcSZvs/HNiRkHwWBx6QtAOAmT1MSExmZjdktqzrLA+sb2ZnmdlFhGeqPYE9U8SwQnw9K15nX8JKBPMBD0paN8bxHvBwLPtMIpbkOIiFgbPN7I9mdoWZ7QK8Auyf7+bJxXLvmDouRbjOOVcm3tmo8ZnZ0/ywUxGSrgPeBC4C7kl5qdvNbGLiui9L+hpYNsW5vTK3BjY3s29iHA8TOhUdT5gvMo2ZwJVZ+54gayLmpORiuS8uuXP9/2t1zjWNRhjWklZnrpHOwcy+AP4JLCtpuZSnfZBj35eEHrcASFo0a8sk0Mxis3dnkmiM4yNCs+3G7Qj/EzPLXuDvB3E451zdaKIaqSfSOX0YXxdKWT7fk3Al/jwhazsy7s8sNjuROU0kDINJq/6fyDvnXEZLS/qtznXapt0CMjXRSfG1HF+HsnvxZp5tZta5WzLHOUsmYihXHM45Vx8aoKaZVqdNpJIWNrPPsvYtRejs85aZvR93TyUMf+llZt+Wci8zG5ln/zuSRgM7JeORtDKhWff6RPGp8XWBUmJwzrl6Yp5Im8LNkmYQOhxNJNREBxOGo/wpUS5Ta7xE0kigxcxuLmMchxGGrTwj6QqgO/Bn4Cvg1ES5NwnPVP8kaRohsf4nkfCdc65xeCJtCncShqgcCvQmdMx5FDjDzMYkyt0A/JTQe3YQ4dln2RKpmY2StCXwV0LinEWYfOFoMxuXKDdV0iDCVIVXEH53+wJlSaQ//eyFclzGVcF83eeYs6PhffP9tOKFGlAz/q7Kpol67cp8mR8HdOu+hP9DaBDN+OHcrIm037y9ax1CRUyY8oaKlyrsmz9tm/ozZ77L7+/w/SqpM9dInXPO1Yi1NE+N1BOpc8656uvsz0gldQW2Jsze04cfjpmEMKXe6R2MzTnnXLPqzIlU0trAHcCPmDOBZhhhInfnnHNuDs00/KWUmY0uJ8wRuyvQx8y65Ni6ljdM55xzTaWJpggspWl3LeAkM7ur3ME455zrHGxW/SfItEpJpBMJYx2dc8650jRATTOtUpp2LwH2kdS93ME455zrJFrbsdW5UmqknwLfA29IupqwWsoc0/Ob2fAOxuacc65JNVNno1ISaXIi9Xw9cw3wROqccy63BqhpplVKIt287FE455zrVDp1jdTMnqhEIM6V2+p9+tc6hIp479tPax2CS2mW1f+i1LViTdRltZTORgBI6iJpPUm7xW09SSVfzznnXCdS5s5GkrpLOl3Sh5KmS3pV0p5pw5HUR9LfJX0gaYakTyTdLmn+YueWOkXgr4GLgMVpm93IgE8kHWpmd5RyXVcaSQZcZWb71zoW55xLw8r/jPQqYC/CpEGvEiYNGi6pi5ndWOhESYsBTwE9gCuBD4B+wEbAvMDXhc4vZYrAXwG3AOOAo4E34qEfA38EbpG0g5nd395rd1aSlgXOBLYi/NL+S1gX9c5Emc2AzYALzKzgL9U55+peGROppHWBAcCpZnZK3HclMAo4T9ItZjazwCWuALoCa5nZpPbev5Sm2BOA14A1zew8M7svbucBawKvAyeWcN1OSdKSwLOEJPp34C/ANOAOSbslim4GnAwUbWZwzrl6Z63ptxR2J7SKXjb7+mGx7cuBRYFN850oaQVgR+BcM5skaW5Jc7fnvZSSSNcArjGzb7MPxH1XExKqS+cYoC+whZmdZmaXEpLmi8DfJc1Vy+Ccc64SypxI1wHG5ahNPp84ns/W8XW8pAcJFZlpkp6UtFaam5eSSL8H5itwfP5YxqWzCfCGmb2S2WFmrYTm88WAn0u6hlAbBfhIksWtf/JCkraVNCY+aB8raa+qvAPnnGsna1HqTdJgSS8mtsFZl1sMmJDjNpl9ixcIZYX4OpSQE/cEDgZWBB6TtFSx91JKZ6NRwMGSbjWzt5IHJK0EHAT4EJn05ga+yrH/u/i6HvAPwheUXYA/A1/GY8lvXz8BdiC09V8F7A9cL2mMmb1Zgbidc65k7elsZGZDCYkun3mAz3Lsn544nk+v+DoJ+KVZGLMk6UXgOeAw4PBC8ZWSSI8lPNN7VdJ9QCaZrgxsS0gAx5Zw3c7qbeAXkvqY2eTE/k3i6xJm9qykVwmJdISZfZzjOqsSHpT/F0DSrYTpG/cjPHedQ/xWNxhAXXvTpUvPsrwh55wrxlrzLWddkmmESkm2Honjhc4FuDmTRAHM7D+SxgIbF7t5KRMyvCFpPeAsYBvCQ1qAqcBdwPFm9r/2XrcTu4zwd3ibpKOBycBvCEkTCn+TShqVSaIAZjZR0lvAsvlOSH7L69Z9ieaZZsQ5V/fKPPxlArB8jv2LxdfxBc7NHJuY49hEYJFiNy9pHGlMlP8XJ2DoF3dPis/2XDuY2UOSDgDOpe3B+ATgUGAI8E3KS32QY9+XQJ8OB+mcc2VmVtYa6WhCy16/rA5HGySO5/NSfF0yx7ElyP3s9Qc6NBORmbWa2cS4eRItkZn9g/Ct56dx6w+8Hw+/k/Iy+eYiK+u/VuecK4fWWUq9pXAb4bPuwMwOSQIOINQqR8V9vSWtLKl34tzHCauaDZA0T+L8rQifxQ8Wu3nRGqmkveMfrzczS/xckJldl6acC8xsGvCfzM+SMl2yR2aKVD0o55yrECvjJ5qZvSBpOHCipD60zWy0CTAoMRnDLoQhmvsC18Rzv5d0OGHFsqckXUtoaT2UUKH5e7H7p2navYbwIX4zYVjLNWneF+CJtESx9/Ng4K7E8+ap8XUBIFdnI+ecaxhl7mwEoWPlOGBv4A+E1rwBxaYHBDCzmyRNB44HziF83t4JHG1mU4qdnyaRLhNv9H3yZ1cecWajOwkdtT4hjGnKNEf8MVE0045/ZuyROxO4x8ymUgbr9M31nL6xLT3XArUOoSImTJ9cvFCDWbbXorUOoSKWmXuhWodQt8qdSM1sBiERHl+gzDXkqQya2QhgRCn3LppIzeyDQj+7DvuaMEzlj4QZjj4lLJ5+qpl9kSlkZo9JOoPwretXhOfby9BWU3XOuYZRzqbdWitl0vr3gEPN7O48x7cHLjazvMMuXJs4Af2uKcueQJjrOHt/zq92ZrZZh4JzzrkKqUDTbs2UMvylP20zQeTSC1i6pGicc851Cq0tnTuRQuEepOsCRR/OOuec67xayzuOtKZSJVJJBxMm8c04X9KpOYouACxEmHDdOeecy6nMEzLUVNoa6deEHqUQpmGawpzTKRlh3tiXgAvKEp1zzrmm1OmekZrZtcC1AJLeB47J19nIOeecK6ZT99o1Mx9H6pxzrkM6XY00H0nzAb3JMWevmX3YkWs755xrXi2tHZrqva6UlEgl/Q44itzL1mR0LSki55xzTa+Zmnbb/ZVA0iBgGGE2nhMIM+5fBPyNsEL5GMLsO84551xOrabUW70rpW59KPC4mW1FXBQa+LeZHQesBiwMzFem+JxzzjUhM6Xe6l0pTbsrAcfEP2fWIJ0LwMy+kDQMOAi4tOPhuWoZ/fnYWodQdj0X/nGtQ6iIpebtV+sQKqJnl+61DqHsuvpywHk1U9NuKYl0GmHlEYBvCMl0kcTxicCPOhiXc64TacYk6gprps5GpbyTsYRaKWY2C3gd+DXMXpF8Z3y9TOeccwV09mekDwB7SJor/nwBsIOkscD/gF8Cl5Upvtkk7SrJJH2euLdzzrkGZO3Y6l0pifSvhE5Fs2D2rEd7Aq8CLwODzOzvZYuwzUDgfcJcvttW4PrOOeeqpJlqpKXMbDQT+CJr37+Af5UrqGyS+gDbAYcD+xKSqk9R6JxzDaoReuOmVco40sUkbVLg+CaSFu1YWHPYgxDrLcBwQlPyAln3XUvSvZImSpou6RNJt0tasj1lYrn9JL0Sy0ySdH12mVhuuXhsgqQZkt6XNCzO+JSTpLkkfSHptjzH/yfpycTPPSSdKWmcpO/j65mS5s5x7uaSHpI0RdJUSa9JOjb/X6tzztVGazu2eldKr93zCIt7b5Tn+BmEJthBJcaUy0BgpJlNknQzcC6wG2FiCCT1A0YCXxKe2X4OLA5sAywBfJymTLzWMcBZwFOE2ZuWAP4MbCppbTObHMutAjxN+DscCrwDLAnsQmh+/ibXGzGzmZJuBwZIms/MZpeT9BPCbFEXxJ8F3EFoyr4BeBbYEDgWWB3YIXHuXsD1wHvAhYTe0ysDO8X345xzdaOliWqkpSTSTYEhBY7fDxxQWjhzkrQ88DNCMsXMxkt6PP48LBbbkPjs1MxeSJx+euLPRctI6gucAjwJbBF7JSNpFHAvYfzsUbH4ZcA8wNpm9lbieifFBFjIcOD3hB7O1yf270l49nxr/PlXhCR6tpllapaXS5oAHClpOzO7L9aALwfeAH5mZt8m3lPeWCQNBgYDqGtvunTpWSRs55wrj9YmGmNbSmejfsCkAse/4IfjSjtqAPAdcGdi33BgY0n9489T4uuOuZo821FmS2Bu4KJMEgUws38TktT2MDvhbg5cn5VEM+WLdTQbRVjfdc/MDkldgN2Bh83s87h7+/h6Xtb552Qd35qweMBZySRaLBYzG2pm65nZep5EnXPVZCj1Vu9KSaQTgTUKHF+T0GxaLgMIzayLSlo+1lBfITSdD4hlRgE3E+b+/ULSA5IOlrRQ4jppyvSPr3MkR0IizSwht1x8fa1Q4JIWzdp6AZhZK6Fz1paJ+29CaEYenhXPJDPL7tw1ifB3nIkns3hAwXicc65eNNMz0lIS6b3A7yVtmn1A0mbA/rFMh0nakJC0tiaMUc1sLxBWl8k095qZ7QmsS5g8fx7Cc8K3JK2WtkwFTMjajkwcG06YWvH/4s97EmaNSta8nXOuKTVTjbSUZ6SnEp7dPSbpYdpqQasDWxGaLE8uT3gMBL4lDHnJtiZwgqT1zex5ADMbDYwGTpe0BvAScETy/CJlxsViKxNqoEmrEDpRAbwbX1cvEv9WWT+/l4jjJUnvAHtKuoqQUO/JapodB2wjaaFkrTQ2LfdNxJOZKHd1vFbqnGsAjVDTTKvdNVIz+wxYH7gW2ICQhI4AfgpcA/zEzCZ2NDBJ3QnPDB8ws9uyN0KtcgYwUNKCOTrVvEmo4S0Qr1e0DPBwvOYhkmZ/yZC0LbAqcE/8O/gceCzee+UcsSuWG5m1vZdV9CZCk+4+hI5Qw7OO3xNfD8/a/5es4w8BXwHHZpqPs2Nxzrl60oJSb/WupIW9YzLdT2GB78xSFJNSdLJpj18Bfcgz8YKZfSvpUeA3hLVRD5A0glA76xb3z0dIVhCG4xxcqExcveYUwnCRR+JYz8zwlw8JyTvjYMLwl+clDQXeBhYjzDu8E22120JuItTeLyB0hro/6/h9hCkZj4vjWP9D+MIyELjXzO6PcX8j6WDCl5sxkq4HPgVWJPRW3rBYINsuunaKcBtLtwb4H7AU3VXS/7Z1ba2uCxQv1IDes+9qHULdam2i/z079H9kTJyflSmWbAOBFuDfBcrcRRge0oMwxvLXwKKEXr6vAzub2V2x7BPAekXKYGZnS5oEHELoLfsNYSznMZkxpLHc65LWJzR1DyIk5PGEWm2qzlZm9rak0cA6wFVm9n3WcZO0K3ASsBfhOeoEQqI/Lavs9ZI+JYwx/QuhteE9wvhT55yrK800/EXtrURKOilFMTOz04sXc/Vihx9t3whzQ7dLs9ZIv2sbmdU0vEbaWG794K4O/89156J7pf7M2fnT4XX9P3MpNdJTChwzQPHVE6lzzrmcOntnoy7ZGyEhLw9cTBia0q/gRZxzznVqLVLqLQ1J3SWdLunDOE/6q5L2LH7mHNdZMM6xbpIGFD+jtHGkczCzVjN7z8wOI3Syuagc13XOOdecKjAhw1XAcYS+MwcThmIOl/TbdoZ2JmGegdTKkkizPEZiMnXnnHMuW6vSb8VIWpcw093pZnawmQ0jLL35FHCepLnSxBSv83tCMk2tEol0pQpd1znnXJNoRam3FHYn9M25LLMjjiq5nDBKY46Z+LLFMfeXEaaSfaY976XdnY1yTQ0YLUCYyP0g2lYvcc455+ZQ5mEC6wDj4jzkSc8njj9S5Br7A6sRhkiu0J6bl9Jr93Fy/x2IMO7zJsIYTOeccy6n9kzIkFzyMRpqZkMTPy9GGGOfLbNv8SLX70MYn3+GmX0iqeKJdPMc+4ywYPa45ELVzjnnXC4t7Sgbk+bQAkXmIffkQNMTxws5izC73AXtCGu2didSM3uilBs555xzGWWeInAaYS3pbD0Sx3OS9BNCs+7OZjajlJs336Sdzjnn6l6ZJ2SYQNu6zEmLxdfxBc49H3gReE1S/7hv0fjaN+772Cz/lGJFE2mcGL69zMx+UcJ5zjnnOoEyJ9LRwC8k9cvqcLRB4ng+PwKWpm1ZyqQL47YMBRYiSVMj7cKcnYuWApYlLN2VWRpsGULP3XeBj1Jc19WRk1ubr3Hiqm7dax1CRazZ0nzv66Dxj9U6hIrYe/Gf1TqEumXlbdq9DTgKOJA4jW0cznIAMBEYFff1JnZMMrOv4rmDgXmzrrcaYZrbiwgLnhRcnKXop6eZbZb8WdLPgHvjza/JVHfj+p37ER7aDix2Xeecc51XOZdeMLMXJA0HTow9cF8FdiWs9zzIzGbGorsAVwP7EtbPxsweyr6epCnxjy+Z2Z3F7l/KxAnnAdeZ2ZXJNmMzmxV7Vl1HaHMuG0n7xHkPM9t0SRMkPSLpqPgXVzGSDpW0dyXvkTKOnVOuvuOcc3XN2rGltB9wNiFZXkZoOR1gZteVMeycSmnPWxu4scDxsfxwvE85nQ68Q4i7H7AxYSqnIyT9n5k9WaH7Hkp4XxX/hRSxM2EarNOKFXTOuXpW7oW9Y4/b4+OWr8w1xJpokWs9DunXYSwlkX4G7CRpiGUtZhrbpHcm5cLWJXjIzJ5K/HyupHWAB4E7Jf3YzCZW6N5lJakL0N3Mphct7JxzTaZTL6NGmLtwK+Cl22FLAAAgAElEQVQhSTtKWjluOwEPA1uQmO+w0sxsNHAY0IcwPSEAklaVdJekKZK+k/ScpO2zz5c0l6RjJL0em4wnSXpY0ibxuBF6dP0i0bT8eOL8pSTdEM+bLukVSftk3aN/PO8ESX+Q9BYwA/hlPH6EpCfjNWZIekvSkfGLSeYajwODgK7JZu7EcUn6Y7z/dElfSLpZ0tId/1t2zrnyqsDqLzVTyoQM50iaBziWkDQzBHwP/NXMzilTfGn9CxgGbEN42LwiYdLhmYReV18D+wB3S9rDzG6F2bXCOwmrBNwN/AOYC9iQMMnxk4SOUxcSen6dHe83MZ7fN95nIeASwrI9uwNXS+prZudlxfkboHe8z2TaulMfDvyb0PNsFuGLyrnAgrQ1U5xB+OKzESGhZruE0EPtRmAIsDBhKaGnJa1lZpVqJXDOuXYr81y7NVXSmAczO1XSpYQP/EyN5wPgYTP7olzBtSOemZLeoW1A7plAT2BNM3sdQNIw4BXgIkkjYkepAYQkeqqZnZK45PmZ2qCZ3SDpr8CnZnZD1q2PAZYEfmlmD8b7DCF0lz5d0tVZfx/LAiua2cdZ11nBzL5L/HyZpCuBQySdZmYzzOzhuK7ehtlxxJ7UBwKD4/JBmf23A2MINfY5nhsk5688doE12bVn/+wizjlXEbPK/Iy0llI17UraT9IyWbt7AyPM7G9xu9nMvpC0tqRq10gBvgHmk9SV0GR6XyaJAsQ5gIcQJi9eJ+7ejVBbPTvrWpkleIrZHvhvJonG82YSarA9gOxJKe7OkUTJJFFJ3RRWZ+9LWBygJ2FZumL2IEyBdY+kvpmNUHN+mx+2HCTvO9TM1jOz9TyJOueqqQK9dmsm7TPSYcDskcWSFgL+R2hmzPZj4IiOh9Zu8xGSaT9CAnorR5k34mvmS8HywNgOdPjpD7yZ4j4Z7+a6iKTtJD1HSIaTgUnA9fHwAiniWJEwKfOEeG5yW43QzOucc3WjFUu91bu0Tbu5KuF1UzGX1J2QTF6rdSxFzDFxsqQNgXuAZ4E/EZ6zfk+oNf+NdF92uhBWLtgt7X2dc66WGqETUVrNMi/c7oSm1AcItbCpwMo5yq0SXzNzKo4Ffi6pR5Faab6vRONS3qeQ3QiJc8tkDJKWbUccY4GtgRcS014551zdqv96ZnqlDH+pK3Ec6YWEJtHLzKwFuB/YVtIqiXK9gD8SVgHITGB8K6FJ+Jgc103WuKeSu4n1HmB1SVslzutGmMBhOjAyxVtoJfyb6pq4Rg9Cj9tsUwnDX3pl7b+Z0EKQc6KG+LzUOefqRqce/lJjWyssadOVtpmNtge+AHZNTMZwPKGGNir2Ls4Mf1kG2CMxteENwF7AyZLWAh6J194QeJnQ+xfgJWCApOMJzzk/M7NHCU2vvyFMBpEZ/rIb4dnxX8xscor3dDehV+1ISdcTEvsg2hakTXopvl4iaSTQEjt5PSXpYuDPklYnfJH4Jr7fnQiJ9pRCQew09X8pQm0si/RYsNYhVMSIaVXvGO9K9Mb3Puosn1lqnjppexLpCvF5HoQeuwCrSsr+wF+x42HldWJ8/Z7wTPB14DjgquQwEzN7J8Z6FmGMZnfC0JcdzezeRLnWOJHEUYShMNsSku5owhCWjBOARQg1117x2KNm9rmkjeJ99ickwbeB/czs6jRvyMyekDQwvo8LCDNHXUMYw5o9mfINwE8JkzEPItRCb47XOUTSS4TnrCfH8h8RvhzckiYW55yrluZJo6A0ozwkZZoff7A7vubab2bWFdcwllhw1Wb6dw00b410fBPWSD//7utah1AR6/VdodYhVMRz4x/vcGfTY/vvlfoz56xxw+umc2suaWuk+1Y0Cuecc51KIwxrSStVIjWzaysdiHPOuc6jedJo43U2cs451wRmNVEq9UTqnHOu6ponjXoidc45VwONMD40LU+kzjnnqs6aqE7qidQ551zVeY3UOeec64BON/zFOeecK6cWT6TOOedc6bxp1znnnOsA72zkms7oNfvVOoSyO2Fc870ngKXmyrWiX2Nrnb/WEVRGD/mU4/l4jdQ555zrgGaqkTb8wt5pSdpHksVtuzxl7ojHZ+U6XoYY1pJ0iqQfVeL6zjnXKJppYe9Ok0gTphPWHv0BSX2AX5F7Qe1yWYuwVqgnUudcp9Zilnqrd50xkd4L7CSpV9b+3eNr9mLadU9Sz1rH4Jxz7dGKpd7qXWdMpDcBcwO7Zu0fCPwb+DK5U9KOku6W9LGkGfF1iKQFssr1lPQ3Se9Kmi7pC0nPSfq/ePwU4OpY/MlEM/NmiWv8QtKjkr6RNFXSE5I2ybrPKfG81SX9U9LnwMdpYnDOuXph7fiv3nXGRPoZodY5MLND0rLAhsD1OcrvB7QAlwIHAXcRFjq/N6vc5cBhwD2x3BnAO8AG8fgdwND459Pj/QcCb8YYdqetNnwicAKwAPCIpE1zxHUTsDihqfjslDE451xdKPczUkndJZ0u6cNYkXhV0p4pzltS0omSno2Vjy/jn/dI+146a6/dG4DrJS1mZhMIz0y/JNRId8oqu5eZfZfcIenZeP5GZvZ03L0jMMzMDs11QzN7NZ43GHjIzJ5KXK8nMAS42cx+m9h/BfAacBawUdYl3wV2NvvBA4SCMWSTNDjGw7krrMDAxRdPc5pzznVYBZpsrwL2IlQoXiW0Og6X1MXMbixw3s7AcYRKUqbcr4GbJa1lZscWu3FnrJEC3Al8R/hLh5BIbzWz77MLZpKogvkl9QUyyXPdRNEpwAaSliohnq2APsANkvpmNqAnMBL4qaR5s84ZkpVE2x2DmQ01s/XMbD1Pos65amrBUm/FSFqX8Dl+upkdbGbDgO2Ap4DzJM1V4PTHgB+Z2W/M7FIzuxTYgvDZe2TsiFpQp0ykMTneAQyQ9FNgBXI36yJpZUl3Ad8CXwGTgPfi4eRz0iOAVYAPJL0s6dz4y01jxfh6X7x+cvsD4fe0UNY57+a4TkdicM65qjGz1FsKuxPWCr8scX0j1E4XBXI9HsuUe93MJmXtM0KO6Ebb53NenbVpF0Lz7kPAmcD7tNUyZ5M0P/AEYUjMScD/CDXZrsADJL6ImNkdkp4CdgC2JDxbPULS8WZ2VpFYMtf5HfBhnjKTsn6ell2ggzE451zVlLlpdx1gXHZCBJ5PHH+kndfMNNN9UaxgZ06kjwATgM2BM3I0kxKPLQxsZmZPZHZKyvkNxcw+I7TTXyVpHkIN8xRJ55nZTMj7L2dsfP3czEaW9G7Sx+CcczXXnokWkv05oqFmNjTx82KEz/NsmX3tenYVm3MHA6PN7H/FynfaRGpmrZIOBNakbVhKtszvOrsJ/C/JHyR1BXqZ2VeJ60+T9DawGeFZ5xRgajycPVnqg/H4CZIeNLMZWdfvl+ObFlll0sbgnHM1155hLTFpDi1QZB7CiIxs0xPHU4mfpTcDCwLbpzmn0yZSADMbAYwoUORp4HPgOkmXEJp1tyfUUpPmAz6RNAJ4BZgMrA3sD9xvZpkENppQKz1W0kLADOBRM/ssfuO6CXhN0o3AeGAJ4Ofx3M2LvJ20MXQaV49/ptYhuE6uR7futQ6hbpW5aXcaYX6AbD0Sx9MaRng0to+ZvZDmhE6dSIsxs8mStgXOJ4zt/B64H9gbmJgo+h1hnOmWhGkG5yY86zwTOCdxvfck/ZnQKegqwrPWzYHPzOxWSZ8QumEfAswLfEpo4/9ninBTxZDPIo8/nuIWDab7ErWOwDmXR5mn/psALJ9j/2LxdXyai0g6nzBPwJFmdl3amytljyjX/JruH0I3T6Suxpq1Rvrtd++ro9fYeqlfpv7MeeijBwreT9LfCI/cFkk+BpO0F2Fs6JZmVrCzkaSTgVOAs8zsuLSxQScd/uKcc662yjzX7m2AgAMzOyQJOIDQejgq7usdhzT2Tp4s6RBCEr2ivUkUvGnXOedcDZSzNdTMXpA0HDgx9rjNzGy0CTAoMWJhF0Ln0n2BawAk7QxcSHgU9qyk7NXBnjGz9yjAE6lzzrmqq8AUgfsB4wh9WP5AmGd8QJHpASEsbynC8pbX5ji+L22T8OTkz0hdRtP9Q/BnpK7W/BlpfpstuWXqz5zHPx7Z4ftVktdInXPOVV0jLNidlidS55xzVdcIC3an5YnUOedc1Xkidc455zqgmfrneCJ1zjlXdV4jdc455zqg1dqz/kt980TqnHOu6rxG6ppOM4653HbRtWsdQkUs2WXeWodQdjPatTpl45jQMrV4oU6qmZ6Rdpq5diV1kfRXSeMktUh6uUr3PUWSSVqyGvfLunf/eO99qn1v55wrpMxz7dZUTROppH3iB31ma5H0qaSbJa1Y5tvtDRwPPECY8qndExM755wrD2vHf/WuXpp2TyfMizg3sC7wO2BLSaub2YQy3WNLYArwR2umNgXnnGtArU30MVwvifQhM3sq/vkqSW8DFwH7AGdlF5Y0F2Ge4O/bcY+Fga88iTrnXO21NFGv3Xp9Rjoyvi4jabPY7Lu3pOMljQOmAz8GkNRH0uWSxkuaIektSUdK6hKP95dkwFbA0olm5H0KBSBpWUk3SfokXvdTSfdJWj2r3NqSRkj6XNJ0Se9IuijHJXtJ+oekyZK+lXSbpIVy3HcHSc9J+k7SFEl3SVolR7lV47EpsexzkrYv/lfrnHO15027lbd8fP08se8oQuK/DJgFTJY0N/AosBrwD+ANYFvgXKA/cBAwCRgIHAMsAhwWr/dMvpvHGu9DQE9gCPBxPPfnwErAa7HcZsD9wFfAFYT17JYFdgcOzbrsDYQFZk+M7+9gYCawZ+K+exJWc38NOAGYP5Z7RtJPzGxsLLdijH8moeb+NaH2frekPczs1nzvzTnn6oE37ZZfb0l9ge6EZ6R/B1oJq57PH8ssCKxsZt9kTpJ0ELAmcICZ/SPuuxz4F3CgpCFm9jpwQ6yB9jKzG1LE82NgOWD3rKQ0u5k51niHEZLYmmY2MXHs2BzXHGtmv0mUEXCQpAPM7KuYvC8A3gU2MrNvY7k7gDHAmYQETfxzz3jf12O5YcArwEWSRpjZrGJvUtJgYDCAuvamS5eexU5xzrmyaISaZlr10rR7L6Hm+AlwN9ADGGhmoxNlrk8m0Wh74Evgn5kd8RnoufHHX5UYz1fx9ZeSeuUpszahZvn3ZBJNxJDt8qyfnwC6EhaThfAFYlFgSCaJxmu9SuhpvG0cwtMV+CVwXyaJxnLfEGrPiwPrFH+LYGZDzWw9M1vPk6hzrppazVJv9a5eaqSHAf8FWggJ9U0za8kq826O8/oTanozs/a/EV+XKXRTSb2BeRK7WsxskpmNk3QOoTl5gKRnCU24N5rZx7Fspvn5tUL3SPgg6+cv42uf+No/vr6V49w3gO2AfoSV3HsWKAfhfT+fMi7nnKs6r5GW34tmNtLMHjOz/+ZIogDTKnDfvwMTEtsLmQNmdjSwMuFZ5QzgVOAtSb8o8V653hOExOicc51Ki7Wk3updvSTSUo0DlpOUXbPO9HJ9v8j55xB682a23yYPmtnbZnaumW0DrEDoLXxCPDw2vv6gF28HjIuvK+c4tgrwLaG2PgmYWqAcFH/fzjlXU2aWeqt3jZ5I7yE0je6btf/I+HpvoZPN7I1YE85sTwNImj87OZvZR4QktkDcNYbQ3HyIpEWSZWNHovZ6EfgUOEDS7AeWklaj7Zloa6yt3094ZrpKolwv4I/AeGA0zjlXx5ppisB6eUZaqiuB3wNDJK0BvEkY/rI9cJmZvVHo5AK2iNe8jTDj0sx4zZWBowHMrDX2er0feEXSlYTnoEsDv6HtGWoqZjZL0uGE4S9PS7qWtuEv3xCmN8w4HtgaGCXpUtqGvywD7JGmx65zztVSI9Q002roRGpmMyRtAZwB/B+wEKGJ9Cjg/A5c+hVCbXYbYD/CuNV3gP3M7OrE/R+VtDFwEmHM6tyEsaR3lXJTM7tJ0lRCojwT+B54HDg2M4Y0lntH0oaE4TiHE4YNvQLsaGYFa+H5LDTPfKWcVtcmzfq2eKEG9PbMT2sdQtkt0r13rUOoiG5q9Ea/ymmE3rhpqZm+FbjSLdJ75ab7h7BMz0VrHUJFTJ7ZfF8QPJE2lsc+frjDnSTb85kz8au36rpTZkPXSJ1zzjWmRnj2mZYnUuecc1XXTK2hnkidc85VXTM9I/VE6pxzruq8Ruqcc851gD8jdc455zqgpbV5Fvb2ROqcc67qfNJ655xzrgPKvYyapO6STpf0oaTpkl6VtGfaeCQNjOdMl/SBpFPjOtFFeSJ1zjlXdRWYtP4q4DjCzHIHE9a3Hi7ptwXPAiT9DrgO+Cieew9wInOuI52TN+0655yrunI27UpaFxgAnGpmp8R9VwKjgPMk3ZJj3erMuT2AswnTsW5vMXNLmgIcJ+liMyu47rTXSJ1zzlVda2tr6i2F3QEDLsvsiAnxcmBRYNMC524O9CUsdJLM7pcT1ovevdjNPZE655yrOmvHlsI6wDgzm5S1//nE8ULnJsuG+MzGAx8XORfwpl0XVXNSaEmDzWxote5XLc34vprxPUFzvq9Ge0+zvv8k9WdOXLJycGLX0Kz3uhgwIcepmX2LF7j8Yllls88vdC7gNVJXG4OLF2lIzfi+mvE9QXO+r2Z8TwCY2VAzWy+xZX9hmAeYkePU6Ynj+cwTbpHzGer0IucCnkidc841vmmE9aCz9UgcL3Su8gx16VHkXMATqXPOucaXrwk202w7vsi5FDi/0LmAJ1JXGw3zHKedmvF9NeN7guZ8X834ntIaDSwtqV/W/g0SxwudC/CT5E5JiwNLFjk3lG2mGfidc851PpJ+Quh1mxxHKuAJYEVgKTObKak3sWOSmX0Vy/Ug9M59DdgiMY70r8DxwJpm9mqh+3uvXeeccw3NzF6QNBw4UVIf4FVgV2ATYFCiI9EuwNXAvsA18dzpko4l1OjvkXQnsBbwJ+CfxZIoeCJ1zjnXHPYDxgF7A38A3gEGmNmNxU40s2GSvgf+QpjU4TPgr8DpaW7sTbvOOedcB3hnI+faQdI8ki6QtEOtY3HFSZpL0hqSlqx1LK55eSJ1VSFpQUlrx23BWsdTKjObBhwALFzrWFwqBrwE/LrWgVSCpPkkbSXpt5IWqXU8nZUnUldRktaX9BTwOfBi3D6XNCr2tGtEY4CVax1EJUhaPUWZ31QjlnIws1mEHplN1x9E0lGEMY4PEpYAWzXu7yvpO0kH1DK+zsQTqasYSesTup+vAQwDDovbMEKvuCcaNJkeA+wnaadaB1IBD0laLt9BSfsC11cxnnK4EthX0ry1DqRcJO1PWPrrbmAQYZUSAMzs87i/KWvh9ajpvqW5uvJXYCKwYVxJYTZJpwHPxDLb1CC2jjgR+BK4Q9IE4H3mnEbMzKzR3hfAh8BISZuY2cfJA7GGcxlwU00iK92HhCbetyVdB7xHjmnfzGx4tQPrgEOAu83st5IWynF8DGGBalcFnkhdJf0UODM7iUJYokjSFYQV7RvNioQP5g/jz83UkWUbwmLImWQ6CUDSYcD5hDF4+9cwvlJcm/jzsXnKGNBIiXQF4NICxycT1th0VeCJ1FWSgJYCx1tJNEk1CjPrX+sYKsXMpkjaCngSeFjSZsCBhPF0l5vZQbWMr0Sb1zqACvgaKNRpb0XCWEhXBT6O1FWMpMcItbUNzGxy1rE+wH+AD83sF7WIz+Un6UeEZNoNWBS40MyOrG1ULiPO4rMusCbQE5gEbGlmj8bf3SvArWbWtEur1RNPpK5iJG0CjCR8e74GeDseWpkw+8j8hP/5n6pJgB0gqTswENiCMBTmKDMbE4f27AA8mv2MsR7FibnzWQm4E7gVOCl5IFdzfb2TNB/hccPCwEgzm1jjkEomaVnC3LKTCL+f44GLCS1AvwO+B9ZthH+DzcATqasoSZsDFxJ67ia9DBxuZo9XPagOisnyEULP468IXwi2irWBLoTORzeZ2TE1DDMVSa2E54N5i+Q6bmZdKxZUBcShIicSam9G2++rL+FZ9+FmdkUtY2wvSSsClwBb8sNHJI8AfzKz/9UksE7In5G6ijKzx4C1JC0B9I+7x5nZJ7WLqsPOJjyD2prwhWD2sygza5V0B/BLwjCZencahRNpw0sMFbkJeIBE5yMz+1xSZqhIQyVSM3sH2EbSAoTOR12Ad+PwF1dFnkhdVcTE2cjJM2lH4GIzG5ln6MFYYJ/qhlSazJJTTa7phopIOgm4w8z+a2ZTgBeyjq8K/NrMTqtJgJ2MJ1JXUXFNwM2BZQm9DLN76ZqZnVv1wDpmQcJYxHy6Ad2rFIsrrhmHipxC+ML23zzHVwNOJrQ4uArzROoqRtLahI4Qy5B/mIsBjZZI3wcKTaX3c9o6VjUUSRcCO5jZ8nmO/w8YYWZHVTeyDumMQ0UWJHQ4clXgidRV0lBCD8lDgGeBKbUNp2xuAI6TdBdhmAHE54xx4oKdCFMhNqLtgFsKHP8X4XliIyXSkYQpAi/KPhCHiuxP+MJX1yRtDGya2LWTpP45ii4A7AW8XoWwHJ5IXWWtBpxqZoWa1RrR34ANgIcJTbwGXBp7gPYjDBm5pHbhdciPKNxsPS6WaSQnEIaKjCEkTAN2kLQdbUNFGqEJ9BeE5loI72G3uOXyIXB4NYJynkhdZY2jCZuX4ooiO0raE9id8B67ED6sbzazG2sZXwd9CyxV4PjSwIwqxVIWZvaepA0JX26OJzxmOCQezgwVaYTxlhcRxmOL8GXnUOCurDIGTDWzL6obWufm40hdxUjahzAEZH0z+7rG4bgUJI0g1LbXyB5GIWlhQlP282bWkCvfNMtQEUk/B940s2Z7ttuQPJG6ipL0e8IKL7cDHzHn3LuN2Gu3acUOYk8TJpq4iLZeoasTanELABub2Uu1idC5+uOJ1FVMHMv2bwo/U7N6nyVH0qMlnGaNOodwnKh+GLAcbZM1CHgXGBwn2Wg4klYhDMPqQ45e5GZ2XdWDSknSPwm/i8Fm1hJ/LsbM7HcVDs3hidRVkKSnCFMDHkeBXrtm9m4142ovSY8z5+w/SxE+lL+irXPOMoQa27vAR2a2RbViLLc4/ndtIDMMZiwwxhrwAyP2bL0B+BkFhmHV8xc6SeMIqyWtZGYz48/FfhdmZstWOjbnidRVkKTvgNPN7Kxax1JOkn4G3AscDVwTOx8hqRuwH3AW8Csze652UboMSQ8BGxG+0D1BWJR9Dmb2QTXjcs3De+26SvqIBuvhmdJ5wHVmdmVyZ0yoQ2MT4vmED++GJGkuwuovCxA65vyAmY2qelCl2xg4x8z+XutAXHPyROoq6W/AUZKuMrOvah1MGa0NFBriMhZo2HUgJZ1O6FjUs0Cxum0GzeEroCF757rG4InUVVJfYCowVtKtNE+v3c8Is8oMyX5mGJ8t7kyDfnBLOoIw1vIqQjPodYQm7K8IE7vPiD83kmuA3SVd1ojPePOJPeL3p60DVTYzM/+MrwJ/RuoqJq51WUxdd/LIJa5teTZhMP8lwDvx0EqEZLM5cKyZnVObCEsn6U3C+MRd40opkwiLrz8qqQfwEnCLmZ1a00ALiJMvJM1FaI6fBvyDMOtP9hc6zOyZykdXHpLOIkzT+CqFn/vW7e+pmXgidRUjabk05eq9124ukk4GjiV8SM/eTZjl6G9mdnLOE+ucpOnAYWY2JE5eMBnYzsweiMePBv5oZv1rGGZBeRYrz/TWzfWBJxrsC52kicBTZvbrWsfivGnXVVAjJsi0zOxUSZcCWxGmzQP4AHi4wadn+5q2z4VvgFnAElnHF652UO20b60DqIKewIO1DsIFXiN1zs0maRTwhpkdEH9+llDL3pKQYB8B5jOzQsvIuQqT9CAw1swOrHUszmukrozieD0jNAW2xJ+LMTPbpsKhVYyknuSfKefD6kfUYXcBh0nqYWbTgdOBuwmTaRgwDzCghvG1W5yZ6gwzeyTP8c2BExtsAo0/ASNjh6OrzCxNfwRXIV4jdWUTZzIyYHMzm5X4uSAz26TiwZWRpLmBkwg9JvvmK9dIz9wKiRNQ7EaYWefuBhtDmnlmOsDMhuc5vgcwvJ5/X5LeybF7fsKyfTOBT8jdI36lSsfmvEbqysjMNi70cxO5lDCD0d0U6DHZLMzsWcIUj42s0Be65QjLx9Wz8cz5Hj4B3qxBLC6L10hdxUiaz8y+KVJmJTN7u1oxlYOkycAdZrZ/rWNxuUkaCAyMP25JWMXm0xxFFyBMsPGwmW1XpfBKImlpn8awPnmN1FXSvyVtHZ+1zUHSGsBDwKLVDavDugIv1DqIcmjilW36ENYdhVCTWxSYL6uMESYMuRE4oXqhlex9SR8CT2Y2M/MaaR3wGqmrGEmfEQbw72hmM7OOrQM8DEwws9VqEV+pJN0CzDCzgUUL17k8K9sUZWablz+ayij2jLRRSLqYMG/wGoT5j40wzvcp2pLrS97xqPo8kbqKkbQuYbjESGD3zP/gsfPK/cA4wqw5DTWdnqSFgQcI72GImX1c45BcJyJpPmBDQlLdBFgf6EFbDfs52mqsj9cozE7FE6mrKEkbE5LO7WY2KC4afQ+hk8Q2ZtZwHXUkZWrXmVVRcs2kY2Y2d/Wiqow4TeD6QC/gFTPL1Xu0rknqTliJ58eEnq7fAK8Dz5hZw69OFFfqWY+2xLohsCA+127V+F+yqygze0rSr4G7JC0IbAG8TBhr+nVtoyvZjZTQHFrPJB1CGB86i7DG6j8k/ZmwtmqPRLkbgH0bpflQ0qGEdUgX4odjfQ34QtKZZnZRTYIrk7jQ98eEXrzjCYsq5Bzb7CrDa6SuKiTtAtxCGC6yo5l9V+OQXCRpAGGVl08IQ3lWJXS+OYOwgPkjhC/dOxBqPEeZ2fm1iTY9Sf8E9gHeI3z5eZkwxeH8wFrAb4FlCF8cflejMEsi6ceE38UmhJroUoTxpKOBpzObmU2qWZCdiCdSVzaS3ihSZGnCEIrhDsMAABSwSURBVIRkc5qZ2aqVi8oVE6cBbCFMpDFT0qmEWty/zGxAVtlHgH5mtkYNQk1N0m7Av4DLgMOzO7vFMt0IC7AfRHiGf3t1o2wfSUcSkuZGhBr2l8AztCXOF/L1kHeV5YnUlU3amYyyNdrMRhmSFgfWIYxF7JJ93Myuq3pQJZD0OXCamV0cf14BeBv4jZndklX2z8DZZjZv9SNNT9JIoKeZ/SxF2WeBqWa2ZeUjK13sfTwTuBm4yMzG1DgkF/kzUlc2TTyT0Q/EzitXAXvSNgwh1zJdDZFICc/Tkj2nJ8fXXL2pvwAaoRPV2sBfU5a9BTixgrGUy82EGulAwkLlL9I29OUZM5tSy+A6szm+RTvnijqFkERPBjYjJNFBwDaEoT5j4P/bu/NoO6vyjuPfHyEgYYaoCFYCYdCgQrGMYhAsiiUCklKUmQpEtIIKUitiLdQiOKHWBSpSTZFBlHlQhKAQhyJoQYWAERVlUlACSAjTr3/s9zTHy81wOec9b845v89ad937Dufy3MXNfc5+97P3Q7qjNGsCMG8p751HW0HVssr2vrZfQpnXPRy4FdiDUgX/gKSfSjpN0n6S1l/c94ruyog0aiPpZcAmti9uOzcVOI7yOPRs259uKr4O7AOcZfsj1fIQgLttz6oeKX4PmAEc1ViEY7expO2rr1evPm9WNfput0kPY+rEb4HNl/LeLYC+WQtcbRP439UHkiZSRqqt5S//CCwv6e4q8UbNMkcatZF0GTDO9hur43WBOZR5ngeAjYCDbJ/VXJRjJ2kB8C7bX5C0OqXoY5rtK6rr76ZUtq7bZJxLq5p7G/mHYLRH1a3zXpY7pQBIOhV4G7CV7TmLuW9Tyu5bZ9h+d6/i6zZJE4DtgKmUJyNb0wf/nwZFRqRRp1cB7Wv09gPGA5vZ/q2kKygVk32VSCnzhKtUXz9MqUJuf+f/DGWJRb84pOkAavBRyrrY6yS9DzjH9hOti9U891uBk4HHq/v7hqTns3AEugNlVD2O8kbnQUpf2esbC3DIZEQatakeC77d9per41nAE7Z3rY5nACfZXqu5KMeualh+n+0Dq+NZlOUIu1H+mF1BqQLdurkoQ9JWwAXAupQ3O7ezcB3pppSiqXuBvWzf0FScS0vSwSx8hNvakF+Ux9itvXavy0b2vZcRadTpQWA9+ItHT+3VkctTRqj95mzgHZKeV63b+wBl04JWi6snKJsXRINs/0jSZpT56t1ZuEXgw5SNCy4BPm97aYuSmnZm9XkOcAYLE+ddzYUUkBFp1EjSuZTF4+8E/g44DJjS6j8q6dOU/XZf2lyU3SFpEuWP9TPAN23PbTSgGDiS9gRm91uTh2GQRBq1kTSZ0iptUnXq47aPra6NA+4CLrd9eDMRRkR0Lok0alUVdbwcmGf7l23nVwN2AX5i+86m4ouI6FQSacQSSPoVY9/60LYn1xFPRCxbUmwUXSXpQ8AFtn9WHS9P6Y9488iiDkmvBt5pe9/eRzom32XA2qZFRPdkRBpdVS3u39/22dXx2pT+iLvYnjXi3v2AmVk0HhH9LHvtRi+kwXBEDKwk0ogxkrSZpL1GnNtJ0jWSbpJ0dFOxRUTvZY40YuxOoYyyLwCQtB5lcf/jwB+AUyT9oV/6kUZEZzIijRi7LSkFSC37UbYG3ML2FOBKyiYUETEEMiKNOmwj6anq61UpFa87Vu2e2m3b27C6Zk3g/rbjNwLftX13dXwpfbYJekQ8d0mkUYd3VR/tjh/tRvpzWckDwIsBJK1M2UP4uLbr48m/rYihkX/s0W27NB1AD1wPHCHpVspodDxljrRlE+Du0V4YEYMn60gjxkjSBpQ9hDesTp1s+1+qa+Moba0utT2joRAjooeSSKNnql2OtgZ+avuRpuPpRPWzTKHsIfybtvOrAjtTdnL6dUPhRUQPJZFGz0h6IXAPo+xyFBHRrzJHGr3Wl7scSVp3rK+xfU8dsUTEsiWJNGLp/I6xVxhnD+GIIZBEGrF0TqA/l+pERM0yRxo9I2k8MBX4se0/NR1PREQ3JJFGRER0II92o3aSNqasuVyLUYqNWr1L+1WqkSOGWxJp1EbS+sBMYAcWXa1roK8TaaUvq5EjonNJpFGnzwNbAcdSuqVkXjQiBk4SadRpKvAx259oOpCIiLqkH2nUaR6l0fWgexT4N+DOpgOJiN5L1W7URtLJwLa2d2w6loiIuiSRRm0k7QB8CniEMl96F/D0yPts39Dj0DoiaXfg9bb/aRHXPwt80/blvY0sIpqQRBq1kfRM2+Fov2gCbLuvttKTdD1wp+2DFnH9TGByRuIRwyHFRlGnwxnMbfU2A85bzPWbgN17FEtENCyJNGpj+4ymY6jJisD4JVyf0KNYIqJhqdqNGLvbgGmLuT4NuL1HsUREwzIijVpJWgHYA3gVsAbPfvNm2zN6HlhnvgicJukM4IO27wOQtA7wEWBHYNRCpIgYPCk2itpIehEwC9iUUrm7CmVt6eqUQqM/Ao/ZfkljQT5Hkk5n4RzwH6vTrb2Ez7B9eFOxRURvJZFGbSR9BdgTeBNwK/B74G+BH1K2DTwA2Nn2bxoLsgOSpgJ7A5OrU3OB821f31xUEdFrSaRRG0n3ATNtHytpbcouR7vYvqa6fh6wvO3pTcYZEdGJFBtFndYA5lRfL6g+r9x2/Vpgp55GFBHRZSk2ijrdB7wQwPajkh4GXgZcUl2f2FRgnZI0BTiKxRdRTX7WCyNi4GREGnX6IaUDTMvlwNGS3iJpP+DdwPcbiawDkrYDbqTM/95LaVp+Z/X1+pTCqusaCzAieipzpFGbqhhnH+Bo249LWg+4BtikumUusJvtXzQV43Mh6dvAJEqv1eWpiqhsz5L0GuAyYLrtq5uLMiJ6JYk0ekrScsAWwDPAz20/2XBIYyZpHnCS7Y9KWgt4gLKJ/dXV9U8C29nersk4I6I3MkcaPWX7GeDHTcfRoXGU5AnwWPV5zbbrt1LWmEbEEEgija6RtC6A7Xvaj5ekdX8fuYsyF0r1yPq3wPbA+dX1LSgbT0TEEEgijW76HWBJK9l+onW8FK/rqzZqlN2a9gSOr47PAo6VtCrlZ9mf0n81IoZAEml0U2vLvCdHHA+aU4BrJa1oewHwYcr2gPtQ5n5nAv/cXHgR0UspNoqIiOhARqQRYyBpG2ADSrHR9dWINCKGWEak0TWSPvAcXmbbJ3U9mC6TtBplfeir207fQ1kHe0szUUXEsiCJNLpG0jOjnG79gmmU86Ik0mW+2EjSx4H3AhcBVwMbA0cAt9ressnYIqJZebQb3TR+xPE6lFHc7cCp1WeAl1K2B9wYmNaz6DqzJ3Cx7b1aJyTdAfynpPX7tRVcRHQuI9KoTdUmbQXbb17E9YuABbb36W1kYydpAXCU7dPbzk2i7LE71fbshkKLiIZl0/qo067AVYu5/i3gDT2KpVPjgYdHnHuk+rxCj2OJiGVIEmnU6Wlg88Vcb+252y8W9fgmj3UihlgSadTpQuBQSR+odv0BQNKqko4D3lbd0y9mSnqi9UFpmwZwVfv56iPLYiKGROZIozbVkpFLgddQRqf3V5deSNlKbzbwJtvL/L60kv5rrK+xfUgdsUTEsiWJNGon6c3AblQbvQO/oTT5vsj5BYyIPpdEGhER0YHMkUZ0SNJESXdKSiPviCGUDRmiVpJ2Ag4FNqR0SHnWDke2N+15YN01DpgErNRwHBHRgCTSqI2kI4FPAQ8CNwBzm40oIqL7kkijTscA1wO72n686WAiIuqQOdKo00TgnCFIovOBr1C6wUTEkMmINOr0E8rc4UCz/TCQNaMRQyoj0qjTe4GDq4KjiIiBlHWkURtJV1CqdTcGfkHZiOHpEbfZ9m69jq1Tkg7jL6uRR7LtPPGJGAL5hx51eiVlQ/d7gJWBKaPc03fv5CSdBBwL3AJ8FfhTsxFFRJMyIo0YI0n3A7NtT286lohoXuZII8ZuZUov1YiIJNKI5+B7LL7PakQMkTzaja6SNJ+xzXva9sp1xVMHSZOBq4H/AL5ku5+ak0dElyWRRldJOosxFhDZPqCmcGoh6Q5gNeD5wJPA3YxejdzvewhHxFJIIo0YI0nfYSneLNjO+tmIIZBEGhER0YEUG0XPSFpb0h2Stm06loiIbsmGDNFLywMbAROaDqQbJI0HNgXWYJQ3pbav63lQEdFzSaQRz4GkE4GjKGtKF2Vcj8KJiAbl0W7EGEk6GjgOOA84CBDwfuAI4FZK15vXNxZgRPRUEmn00nzK3rT3Nh1Ihw4FLrJ9GHBlde4m218AtgJWAnZoKriI6K0k0qiFpJUkHSlpauuc7YdtH2D7tiZj64INgG9XX7fWj64AUDUxn0n6k0YMjSTSqIXt+cApwCZNx1KDh1lYX/AI8BSw3ojrL+h1UBHRjCTSqNNtwF81HUQN5gCbAdh+GrgJOFDSeEkrAQcAv2wwvojooSTSqNMJwDskbdF0IF12MTBN0vOq4xOBVwMPAX8AtqHswxsRQyA7G0VtJH2BkmA2BW6gjNLmj7jNtmf0OrZuk7QdsDfwDHBJ1pBGDI8k0qiNpKXpimLbWW8ZEX0riTQiIqID2dkoYgkkXUt5ZPsG209JmrUUL7Pt19UcWkQsA5JIo3aSXgfsTFkS8gnbcyStAmwJ3GL7oUYDXDLxl4V5y7HkNmqqL5yIWJbk0W7Upqpq/QawKyWxGNjF9ixJKwK/Az5j+8QGw4yI6EiWv0SdTgB2AWZQur78/yjN9gLgfOBNzYQWEdEdSaRRp32A022fAcwb5frtlO32+oqkdSVtNeLcFElfkvQNSdObii0iei9zpFGndYBbFnP9SWCVHsXSTZ+hzPdOBZC0FvAdSl/S+cCekvawfVljEUZEz2REGnW6G5i8mOtbA7/qUSzdtC3wrbbjfYE1KcVTE4EfAMc0EFdENCCJNOr0dWCGpClt5wwgaU9gP+DcJgLr0NrAPW3HuwGzbf/M9pPAOVR78UbE4EsijTqdQNkW8Ebga5QkerykGynVvDdSOsT0m3mUkSeSVqD0Hr267fpTwIQG4oqIBiSRRm1sP0pJMicCawGPA9sBKwIfAl5b9e/sNzcCh0raEvggJWle3nZ9I+C+JgKLiN7LOtKIMZK0OWUEuhZlSc85tvdruz4X+L7tAxsKMSJ6KFW7EWNk+2ZJLwO2Bx5q7/QiaXXgdMqoOyKGQEakUStJGwOHABtSKltHbp1n22/oeWA1kPQ3lKbebwXWTlebiOGQEWnURtKBwJmUufh5lMbXA0XSS4D9KQl0E+AJYBal+XdEDIGMSKM2kn4N/BmYbntOw+F0jaTVKE28D6AUU5nyZuHfgVNs/7nB8CKix1K1G3V6PvC5QUiiksZJmibpPEpF7ucpy1wOp1QiC7g5STRi+OTRbtTpJso2gYPgXspGDD8GjgPOtX0vgKTF7d4UEQMuI9Ko0zHAYZJ2bDqQLphI2c7wTGBmK4lGRGREGrWxfYOkI4FrJP0KuAt4+tm39UXV7nTKnOingFMlXQ2cRYqKIoZeEmnURtLfA2dTnnxMpI9/32xfCFwoaU3gLZRK3a9SiqlmUwqOUrkXMYRStRu1kfQLynKQgarabanmRg+gbL4/GXgEuBK4BLjC9sAt94mIZ0sijdpI+jNwrO3PNR1L3SRtBxxIWRazFvCk7exuFDEEUmwUdfpfBqdqd7Fs/8D2EcCLKPOpaeodMSQyIo3aSNoGuADY3/a1TccTEVGHJNKojaSrgEmU+cM76e+q3YiIUSWRRm2qLQKX9Atm2xv2IJyIiFr07XKEWPbZnrS465KWBzIajYi+lkQaPdfWbuwtlPWlaTcWEX0riTR6Iu3GImJQJZFGbdJuLCKGQdaRRlel3VhEDJuMSKPb0m4sIoZKEml020TKmtEzga/ZfrDheCIiapVHu9Ft04FbKO3G7pF0uaS3SprQcFwREbXIiDS6Ku3GImLYZGejqF3ajUXEIEsijZ5Ku7GIGDRJpNEISeOBaZTOMNObjici4rlKIo2IiOhAqnYjIiI6kEQaERHRgSTSiIiIDiSRRgwYSQdLctvH05Luk3SepE2bji9i0GRDhojBdSJwB7ACsDmlccDOkl5u+/5GI4sYIEmkEYPrKtuzWweSbgNOo6zj/VhjUUUMmDzajRge11efN2qdkPQaSedK+rWkBZJ+L+ksSS8e+WJJq0o6SdLc6t57JV0sabO2eyTpCEk3S3pc0oPV91+/Bz9fRCMyIo0YHpOqz39sO7c3pe3dGcD9wMbADGAbSa+0PR+gajrwHeCvgbOAHwKrATsBrwJ+Xn2/zwJvp+yvfBrwAuBdwPckbWH7gZp+tojGJJFGDK7VJU1k4RzpqcAzwNfb7nm/7cfaXyTpEsrodS9KQgQ4BtgSOMT2l9tu/6gkVa/bDngncLjtL7Z9v28APwHeQ+lRGzFQkkgjBtdlI47vBfa1fVPrRHsSlbQqsCIwB3iIMtJsJdK9gTtGJNHW92htj7YPMB+4tErgLfcDtwM7d/LDRCyrkkgjBtd7gJ8BqwL/AOwJPNl+g6R1gVMo+x6vPuL1a7R9vRFwxRL+e5sAK1ES9mjuXKqoI/pMEmnE4LqxrWr3QkkXA1+W9D+275a0HPBtYB1KFe+twKOUfrHnMvZixOUoI9m9F3F9/lh/gIh+kEQaMTyOpSTL4ykFQa8ApgAH2/5K6yZJKwFrjnjtXODlkuRFd7qYC7we+JHted0OPmJZleUvEUPC9u3AhcAhktajFB7Bs/8OHD3KufMpj24PGvl9W8VGlFGsgBNG+++PmDeNGBgZkUYMl1OA6ZQq3PdRdj76hKSXUIqCXgtsDzw44nUfp1TxnilpZ+AHwMqUAqJzgZm2Z0v6DHCkpFcAVwKPABsAe1T3fbjOHy6iCRmRRgwR2zdQ1oMeTikmmkZJiu8BTgYmUNaGPjridY8BOwKfBKYCn6aMXBcAN7bddxRl1DoB+FdKAt4duAb4Wm0/WESD0tg7IiKiAxmRRkREdCCJNCIiogNJpBERER1IIo2IiOhAEmlEREQHkkgjIiI6kEQaERHRgSTSiIiIDiSRRkREdOD/ALOdv6LpAaLGAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113d3da58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_edu_race = pd.crosstab(adult_mis['Education'], adult_mis['Race']).divide(\n",
    "    pd.crosstab(adult_data2['Education'],adult_data2['Race']))\n",
    "sns.heatmap(df_edu_race, vmin=0, vmax=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113d37fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_edu_sex = pd.crosstab(adult_mis['Education'], adult_mis['Sex']).divide(\n",
    "    pd.crosstab(adult_data2['Education'],adult_data2['Sex']))\n",
    "sns.heatmap(df_edu_sex, vmin=0, vmax=1)\n",
    "plt.savefig('heat_edu_sex.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11543bda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_male = y[X['Sex']==1]\n",
    "y_hat_male = lr.predict(X[X['Sex'] == 1])\n",
    "\n",
    "adult_male = adult_data2[adult_data2['Sex'] == 'Male']\n",
    "adult_male['Age'] = pd.cut(adult_male.Age, 10)\n",
    "adult_male_edu_age = pd.crosstab(adult_male['Education'],adult_male['Age'])\n",
    "\n",
    "adult_male_mis = adult_male[y_male != y_hat_male]\n",
    "adult_male_mis_edu_age = pd.crosstab(adult_male_mis['Education'],adult_male_mis['Age'])\n",
    "\n",
    "df_male_edu_age = adult_male_mis_edu_age.divide(adult_male_edu_age)\n",
    "sns.heatmap(df_male_edu_age, vmin=0, vmax=1)\n",
    "plt.savefig('heat_male_edu_age.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11359a8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_female = y[X['Sex']==0]\n",
    "y_hat_female = lr.predict(X[X['Sex'] == 0])\n",
    "\n",
    "adult_female = adult_data2[adult_data2['Sex'] == 'Female']\n",
    "adult_female['Age'] = pd.cut(adult_female.Age, 10)\n",
    "adult_female_edu_age = pd.crosstab(adult_male['Education'],adult_male['Age'])\n",
    "\n",
    "adult_female_mis = adult_female[y_female != y_hat_female]\n",
    "adult_female_mis_edu_age = pd.crosstab(adult_female_mis['Education'],adult_female_mis['Age'])\n",
    "\n",
    "df_female_edu_age = adult_female_mis_edu_age.divide(adult_female_edu_age)\n",
    "sns.heatmap(df_female_edu_age, vmin=0, vmax=1)\n",
    "plt.savefig('heat_female_edu_age.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f666160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1,3,figsize=(10,7),gridspec_kw={'width_ratios':[1,1,0.08]})\n",
    "g1 = sns.heatmap(df_edu_sex, vmin=0, vmax=1, cbar=False, ax=axes[0])\n",
    "g2 = sns.heatmap(df_male_edu_age, vmin=0, vmax=1, ax=axes[1], cbar_ax=axes[2])\n",
    "g2.set_ylabel('')\n",
    "g2.set_yticks([])\n",
    "g2.set_title('Sex=Male')\n",
    "#g3 = sns.heatmap(df_female_edu_age, vmin=0, vmax=1, ax=axes[2], cbar_ax=axes[3])\n",
    "#g3.set_ylabel('')\n",
    "#g3.set_yticks([])\n",
    "#plt.xticks(rotation=75)\n",
    "plt.tight_layout()\n",
    "plt.savefig('heat_male_edu_age2.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e245a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_male = y[X['Sex']==1]\n",
    "y_hat_male = lr.predict(X[X['Sex'] == 1])\n",
    "\n",
    "adult_male = adult_data2[adult_data2['Sex'] == 'Male']\n",
    "adult_male_edu_work = pd.crosstab(adult_male['Education'],adult_male['Workclass'])\n",
    "\n",
    "adult_male_mis = adult_male[y_male != y_hat_male]\n",
    "adult_male_mis_edu_work = pd.crosstab(adult_male_mis['Education'],adult_male_mis['Workclass'])\n",
    "\n",
    "df_male_edu_work = adult_male_mis_edu_work.divide(adult_male_edu_work)\n",
    "sns.heatmap(df_male_edu_work, vmin=0, vmax=1)\n",
    "plt.savefig('heat_male_edu_work.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e5047b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1,3,figsize=(10,7),gridspec_kw={'width_ratios':[1,1,0.08]})\n",
    "g1 = sns.heatmap(df_edu_sex, vmin=0, vmax=1, cbar=False, ax=axes[0])\n",
    "g2 = sns.heatmap(df_male_edu_work, vmin=0, vmax=1, ax=axes[1], cbar_ax=axes[2])\n",
    "g2.set_ylabel('')\n",
    "g2.set_yticks([])\n",
    "g2.set_title('Sex=Male')\n",
    "#g3 = sns.heatmap(df_female_edu_age, vmin=0, vmax=1, ax=axes[2], cbar_ax=axes[3])\n",
    "#g3.set_ylabel('')\n",
    "#g3.set_yticks([])\n",
    "#plt.xticks(rotation=75)\n",
    "plt.tight_layout()\n",
    "plt.savefig('heat_male_edu_work2.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Recall of Injected Ground Truth\n",
    "Compare the effectiveness of different search strategies based on recall and the injected ground truth."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ground Truth Injection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "adult_data = pd.read_csv(\n",
    "    \"data/adult.data\",\n",
    "    names=[\n",
    "        \"Age\", \"Workclass\", \"fnlwgt\", \"Education\", \"Education-Num\", \"Martial Status\",\n",
    "        \"Occupation\", \"Relationship\", \"Race\", \"Sex\", \"Capital Gain\", \"Capital Loss\",\n",
    "        \"Hours per week\", \"Country\", \"Target\"],\n",
    "        sep=r'\\s*,\\s*',\n",
    "        engine='python',\n",
    "        na_values=\"?\")\n",
    "\n",
    "# drop nan values\n",
    "adult_data = adult_data.dropna()\n",
    "\n",
    "# Encode categorical features\n",
    "encoders = {}\n",
    "for column in adult_data.columns:\n",
    "    if adult_data.dtypes[column] == np.object:\n",
    "        le = LabelEncoder()\n",
    "        adult_data[column] = le.fit_transform(adult_data[column])\n",
    "        encoders[column] = le\n",
    "\n",
    "# Original dataset (X, y)\n",
    "X, y = adult_data[adult_data.columns.difference([\"Target\"])], adult_data[\"Target\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original dataset size: 30162\n",
      "Unmarried white demographic size: 2341\n",
      "Unmarried white female demographic size: 1732\n"
     ]
    }
   ],
   "source": [
    "unmarried_white = X[(X['Relationship']==4) & (X['Race']==4)]\n",
    "unmarried_white_female = unmarried_white[unmarried_white['Sex']==0]\n",
    "print('Original dataset size: %s'%len(X))\n",
    "print('Unmarried white demographic size: %s'%len(unmarried_white))\n",
    "print('Unmarried white female demographic size: %s'%len(unmarried_white_female))\n",
    "\n",
    "# Corrupted labels\n",
    "import copy\n",
    "y_c = copy.copy(y)\n",
    "y_c.loc[unmarried_white_female.index] = (y_c.loc[unmarried_white_female.index] + 1)%2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=5, 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=10, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Train a model\n",
    "lr_c = RandomForestClassifier(max_depth=5, n_estimators=10)\n",
    "lr_c.fit(X, y_c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate Slice Finder\n",
    "sf = SliceFinder(lr_c, (X, y))\n",
    "metrics_all = sf.evaluate_model((X,y))\n",
    "reference = (np.mean(metrics_all), np.std(metrics_all), len(metrics_all))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multiple-Comparison Problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = 'UCI'\n",
    "encoders = {}\n",
    "\n",
    "main_df = pd.read_csv(\"data/creditcard.csv\")\n",
    "main_df = main_df.dropna()\n",
    "main_df.head()\n",
    "feature_size = len(main_df.columns)\n",
    "class_index = feature_size -1 \n",
    "\n",
    "fraud_raw = main_df[main_df['Class'] == 1]\n",
    "normal_raw = main_df[main_df['Class'] == 0]\n",
    "\n",
    "# Undersample the normal transactions\n",
    "percentage = len(fraud_raw)/float(len(normal_raw))\n",
    "normal = normal_raw.sample(frac=percentage)\n",
    "fraud = fraud_raw\n",
    "cc_data = fraud.append(normal)\n",
    "\n",
    "X, y = cc_data[cc_data.columns.difference([\"Class\"])], cc_data[\"Class\"]\n",
    "\n",
    "reg_model = RandomForestClassifier(criterion='entropy',n_estimators=100)\n",
    "lr = reg_model.fit(X,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "sf = SliceFinder(lr, (X, y))\n",
    "metrics_all = sf.evaluate_model((X,y))\n",
    "reference = (np.mean(metrics_all), np.std(metrics_all), len(metrics_all))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Workclass ['Federal-gov' 'Local-gov' 'Private' 'Self-emp-inc' 'Self-emp-not-inc'\n",
      " 'State-gov' 'Without-pay'] [0 1 2 3 4 5 6]\n",
      "Education ['10th' '11th' '12th' '1st-4th' '5th-6th' '7th-8th' '9th' 'Assoc-acdm'\n",
      " 'Assoc-voc' 'Bachelors' 'Doctorate' 'HS-grad' 'Masters' 'Preschool'\n",
      " 'Prof-school' 'Some-college'] [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15]\n",
      "Martial Status ['Divorced' 'Married-AF-spouse' 'Married-civ-spouse'\n",
      " 'Married-spouse-absent' 'Never-married' 'Separated' 'Widowed'] [0 1 2 3 4 5 6]\n",
      "Occupation ['Adm-clerical' 'Armed-Forces' 'Craft-repair' 'Exec-managerial'\n",
      " 'Farming-fishing' 'Handlers-cleaners' 'Machine-op-inspct' 'Other-service'\n",
      " 'Priv-house-serv' 'Prof-specialty' 'Protective-serv' 'Sales'\n",
      " 'Tech-support' 'Transport-moving'] [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13]\n",
      "Relationship ['Husband' 'Not-in-family' 'Other-relative' 'Own-child' 'Unmarried' 'Wife'] [0 1 2 3 4 5]\n",
      "Race ['Amer-Indian-Eskimo' 'Asian-Pac-Islander' 'Black' 'Other' 'White'] [0 1 2 3 4]\n",
      "Sex ['Female' 'Male'] [0 1]\n",
      "Country ['Cambodia' 'Canada' 'China' 'Columbia' 'Cuba' 'Dominican-Republic'\n",
      " 'Ecuador' 'El-Salvador' 'England' 'France' 'Germany' 'Greece' 'Guatemala'\n",
      " 'Haiti' 'Holand-Netherlands' 'Honduras' 'Hong' 'Hungary' 'India' 'Iran'\n",
      " 'Ireland' 'Italy' 'Jamaica' 'Japan' 'Laos' 'Mexico' 'Nicaragua'\n",
      " 'Outlying-US(Guam-USVI-etc)' 'Peru' 'Philippines' 'Poland' 'Portugal'\n",
      " 'Puerto-Rico' 'Scotland' 'South' 'Taiwan' 'Thailand' 'Trinadad&Tobago'\n",
      " 'United-States' 'Vietnam' 'Yugoslavia'] [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24\n",
      " 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40]\n",
      "Target ['<=50K' '>50K'] [0 1]\n"
     ]
    }
   ],
   "source": [
    "adult_data = pd.read_csv(\n",
    "    \"data/adult.data\",\n",
    "    names=[\n",
    "        \"Age\", \"Workclass\", \"fnlwgt\", \"Education\", \"Education-Num\", \"Martial Status\",\n",
    "        \"Occupation\", \"Relationship\", \"Race\", \"Sex\", \"Capital Gain\", \"Capital Loss\",\n",
    "        \"Hours per week\", \"Country\", \"Target\"],\n",
    "        sep=r'\\s*,\\s*',\n",
    "        engine='python',\n",
    "        na_values=\"?\")\n",
    "\n",
    "# drop nan values\n",
    "adult_data = adult_data.dropna()\n",
    "\n",
    "# Encode categorical features\n",
    "encoders = {}\n",
    "for column in adult_data.columns:\n",
    "    if adult_data.dtypes[column] == np.object:\n",
    "        le = LabelEncoder()\n",
    "        adult_data[column] = le.fit_transform(adult_data[column])\n",
    "        encoders[column] = le\n",
    "        print(column, le.classes_, le.transform(le.classes_))\n",
    "\n",
    "X, y = adult_data[adult_data.columns.difference([\"Target\"])], adult_data[\"Target\"]\n",
    "\n",
    "# Train a model\n",
    "#lr = LogisticRegression()\n",
    "#lr.fit(X, y)\n",
    "lr = RandomForestClassifier(max_depth=5, n_estimators=10)\n",
    "lr.fit(X, y)\n",
    "\n",
    "sf = SliceFinder(lr, (X, y))\n",
    "metrics_all = sf.evaluate_model((X,y))\n",
    "reference = (np.mean(metrics_all), np.std(metrics_all), len(metrics_all))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "degree 1\n",
      "crossing\n",
      "done\n",
      "effect size filtering\n",
      "done\n",
      "degree 2\n",
      "crossing\n",
      "done\n",
      "effect size filtering\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m-------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                 Traceback (most recent call last)",
      "\u001b[0;32m~/git/slicefinder/slice_finder.py\u001b[0m in \u001b[0;36mfilter_by_effect_size\u001b[0;34m(self, slices, reference, epsilon, max_workers, alpha, risk_control)\u001b[0m\n\u001b[1;32m    210\u001b[0m                 \u001b[0mbatch_jobs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mexecutor\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubmit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0meff_size_job\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreference\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 211\u001b[0;31m             \u001b[0;32mfor\u001b[0m \u001b[0mjob\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mconcurrent\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfutures\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mas_completed\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_jobs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    212\u001b[0m                 \u001b[0;32mif\u001b[0m \u001b[0mjob\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcancelled\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/concurrent/futures/_base.py\u001b[0m in \u001b[0;36mas_completed\u001b[0;34m(fs, timeout)\u001b[0m\n\u001b[1;32m    214\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 215\u001b[0;31m             \u001b[0mwaiter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mevent\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwait\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mwait_timeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    216\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/threading.py\u001b[0m in \u001b[0;36mwait\u001b[0;34m(self, timeout)\u001b[0m\n\u001b[1;32m    548\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0msignaled\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 549\u001b[0;31m                 \u001b[0msignaled\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_cond\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwait\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtimeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    550\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0msignaled\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/threading.py\u001b[0m in \u001b[0;36mwait\u001b[0;34m(self, timeout)\u001b[0m\n\u001b[1;32m    292\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mtimeout\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 293\u001b[0;31m                 \u001b[0mwaiter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0macquire\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    294\u001b[0m                 \u001b[0mgotit\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: ",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-9-78f297cb6e14>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mrecommendations\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfind_slice\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m100\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepsilon\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.4\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdegree\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m3\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrisk_control\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mslices\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0muninteresting\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/slicefinder/slice_finder.py\u001b[0m in \u001b[0;36mfind_slice\u001b[0;34m(self, k, epsilon, alpha, degree, risk_control, max_workers)\u001b[0m\n\u001b[1;32m    108\u001b[0m             interesting, uninteresting_ = self.filter_by_effect_size(candidates, reference, epsilon, \n\u001b[1;32m    109\u001b[0m                                                                     \u001b[0mmax_workers\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmax_workers\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 110\u001b[0;31m                                                                     risk_control=risk_control)\n\u001b[0m\u001b[1;32m    111\u001b[0m             \u001b[0muninteresting\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0muninteresting_\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    112\u001b[0m             \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'done'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/slicefinder/slice_finder.py\u001b[0m in \u001b[0;36mfilter_by_effect_size\u001b[0;34m(self, slices, reference, epsilon, max_workers, alpha, risk_control)\u001b[0m\n\u001b[1;32m    220\u001b[0m                         \u001b[0mfiltered_slices\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    221\u001b[0m                     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 222\u001b[0;31m                         \u001b[0mrejected\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    223\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mfiltered_slices\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrejected\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    224\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/concurrent/futures/_base.py\u001b[0m in \u001b[0;36m__exit__\u001b[0;34m(self, exc_type, exc_val, exc_tb)\u001b[0m\n\u001b[1;32m    579\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    580\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__exit__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexc_type\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexc_val\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexc_tb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 581\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshutdown\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mwait\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    582\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/concurrent/futures/process.py\u001b[0m in \u001b[0;36mshutdown\u001b[0;34m(self, wait)\u001b[0m\n\u001b[1;32m    492\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_result_queue\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mput\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    493\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mwait\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 494\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_queue_management_thread\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mjoin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    495\u001b[0m         \u001b[0;31m# To reduce the risk of opening too many files, remove references to\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    496\u001b[0m         \u001b[0;31m# objects that use file descriptors.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/threading.py\u001b[0m in \u001b[0;36mjoin\u001b[0;34m(self, timeout)\u001b[0m\n\u001b[1;32m   1052\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1053\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mtimeout\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1054\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_wait_for_tstate_lock\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1055\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1056\u001b[0m             \u001b[0;31m# the behavior of a negative timeout isn't documented, but\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/threading.py\u001b[0m in \u001b[0;36m_wait_for_tstate_lock\u001b[0;34m(self, block, timeout)\u001b[0m\n\u001b[1;32m   1068\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mlock\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;31m# already determined that the C code is done\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1069\u001b[0m             \u001b[0;32massert\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_is_stopped\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1070\u001b[0;31m         \u001b[0;32melif\u001b[0m \u001b[0mlock\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0macquire\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mblock\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1071\u001b[0m             \u001b[0mlock\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrelease\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1072\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_stop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Process Process-2:\n",
      "Traceback (most recent call last):\n",
      "  File \"/usr/local/Cellar/python3/3.5.2_3/Frameworks/Python.framework/Versions/3.5/lib/python3.5/multiprocessing/process.py\", line 249, in _bootstrap\n",
      "    self.run()\n"
     ]
    }
   ],
   "source": [
    "recommendations = sf.find_slice(k=100, epsilon=0.4, degree=3, risk_control=False)\n",
    "slices, uninteresting = list(), list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "39358\n",
      "\n",
      "============t-Testing:\n",
      "\n",
      "============Bon Ferroni:\n",
      "\n",
      "============Alpha-Investing:\n"
     ]
    }
   ],
   "source": [
    "# slice finder results before risk control\n",
    "with open('slices.p','rb') as handle:\n",
    "    slices = pickle.load(handle)\n",
    "with open('uninteresting.p', 'rb') as handle:\n",
    "    uninteresting = pickle.load(handle) \n",
    "\n",
    "print(len(slices))    \n",
    "    \n",
    "slices, rejected, bf_slices, bf_rejected, ai_slices, ai_rejected = sf.filter_by_significance(\n",
    "    slices, reference, 0.1, max_workers=10)\n",
    "print('\\n============t-Testing:')\n",
    "'''\n",
    "for s in rejected:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    for k, v in list(s.filters.items()):\n",
    "        values = ''\n",
    "        if k in encoders:\n",
    "            le = encoders[k]\n",
    "            for v_ in v:\n",
    "                values += '%s '%(le.inverse_transform(v_)[0])\n",
    "        else:\n",
    "            for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                if len(v_) > 1:\n",
    "                    values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                else:\n",
    "                    values += '%s '%(v_[0])\n",
    "        print ('%s:%s'%(k, values))\n",
    "    print ('---------------------\\neffect_size: %s'%(s.effect_size))\n",
    "    print ('---------------------\\nmetric: %s'%(s.metric))\n",
    "    print ('size: %s'%(s.size))\n",
    "'''\n",
    "with open('accepted_%s.p'%dataset,'wb') as handle:\n",
    "    pickle.dump(slices, handle)\n",
    "with open('rejected_%s.p'%dataset, 'wb') as handle:\n",
    "    pickle.dump(rejected, handle)\n",
    "    \n",
    "print('\\n============Bon Ferroni:')\n",
    "'''\n",
    "for s in bf_rejected:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    for k, v in list(s.filters.items()):\n",
    "        values = ''\n",
    "        if k in encoders:\n",
    "            le = encoders[k]\n",
    "            for v_ in v:\n",
    "                values += '%s '%(le.inverse_transform(v_)[0])\n",
    "        else:\n",
    "            for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                if len(v_) > 1:\n",
    "                    values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                else:\n",
    "                    values += '%s '%(v_[0])\n",
    "        print ('%s:%s'%(k, values))\n",
    "    print ('---------------------\\neffect_size: %s'%(s.effect_size))\n",
    "    print ('---------------------\\nmetric: %s'%(s.metric))\n",
    "    print ('size: %s'%(s.size))\n",
    "'''\n",
    "with open('bf_accepted_%s.p'%dataset,'wb') as handle:\n",
    "    pickle.dump(bf_slices, handle)\n",
    "with open('bf_rejected_%s.p'%dataset, 'wb') as handle:\n",
    "    pickle.dump(bf_rejected, handle)   \n",
    "    \n",
    "print('\\n============Alpha-Investing:')\n",
    "'''\n",
    "for s in ai_rejected:\n",
    "    print ('\\n=====================\\nSlice description:')\n",
    "    for k, v in list(s.filters.items()):\n",
    "        values = ''\n",
    "        if k in encoders:\n",
    "            le = encoders[k]\n",
    "            for v_ in v:\n",
    "                values += '%s '%(le.inverse_transform(v_)[0])\n",
    "        else:\n",
    "            for v_ in sorted(v, key=lambda x: x[0]):\n",
    "                if len(v_) > 1:\n",
    "                    values += '%s ~ %s'%(v_[0], v_[1])\n",
    "                else:\n",
    "                    values += '%s '%(v_[0])\n",
    "        print ('%s:%s'%(k, values))\n",
    "    print ('---------------------\\neffect_size: %s'%(s.effect_size))\n",
    "    print ('---------------------\\nmetric: %s'%(s.metric))\n",
    "    print ('size: %s'%(s.size))\n",
    "'''\n",
    "with open('ai_accepted_%s.p'%dataset,'wb') as handle:\n",
    "    pickle.dump(ai_slices, handle)\n",
    "with open('ai_rejected_%s.p'%dataset, 'wb') as handle:\n",
    "    pickle.dump(ai_rejected, handle) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#print (len(slices), len(rejected)) # fraud (1544, 4182), census (2530, 876)\n",
    "#print (np.mean([s.size for s in slices]), np.mean([s.size for s in rejected])) # fraud (7.86, 8.36), census (44.19, 5.18)\n",
    "#print (np.mean([s.effect_size for s in slices]), np.mean([s.effect_size for s in rejected])) # fraud (1.54, 0.60), census (1.28, 0.65)       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "39358 0\n",
      "39358 0\n",
      "39358 0\n"
     ]
    },
    {
     "ename": "ValueError",
     "evalue": "max() arg is an empty sequence",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m-------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                        Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-10-c398fe1571e3>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     21\u001b[0m \u001b[0msignificant\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mslices\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     22\u001b[0m \u001b[0minsignificant\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrejected\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 23\u001b[0;31m \u001b[0mbins\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlinspace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmax\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msignificant\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m50\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     24\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mhist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msignificant\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbins\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.5\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'Accepted'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     25\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mhist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minsignificant\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbins\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.5\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'Rejected'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: max() arg is an empty sequence"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12115e198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(len(rejected), len(slices))\n",
    "print(len(bf_rejected), len(bf_slices))\n",
    "print(len(ai_rejected), len(ai_slices))\n",
    "\n",
    "dataset='fraud'\n",
    "with open('accepted_%s.p'%dataset,'rb') as handle:\n",
    "    slices = pickle.load(handle)\n",
    "with open('rejected_%s.p'%dataset, 'rb') as handle:\n",
    "    rejected = pickle.load(handle)\n",
    "with open('bf_accepted_%s.p'%dataset,'rb') as handle:\n",
    "    bf_slices = pickle.load(handle)\n",
    "with open('bf_rejected_%s.p'%dataset, 'rb') as handle:\n",
    "    bf_rejected = pickle.load(handle)\n",
    "with open('ai_accepted_%s.p'%dataset,'rb') as handle:\n",
    "    ai_slices = pickle.load(handle)\n",
    "with open('ai_rejected_%s.p'%dataset, 'rb') as handle:\n",
    "    ai_rejected = pickle.load(handle)\n",
    "\n",
    "    \n",
    "plt.figure(1, figsize=(6,5))\n",
    "significant = [s.size for s in slices if s.size]\n",
    "insignificant = [s.size for s in rejected if s.size]\n",
    "bins = np.linspace(0, max(significant), 50)\n",
    "plt.hist(significant, bins, alpha=0.5, label='Accepted')\n",
    "plt.hist(insignificant, bins, alpha=0.5, label='Rejected')\n",
    "plt.xlim([0,100])\n",
    "plt.legend(loc='upper right')\n",
    "plt.xlabel('Slice Size',fontsize=17)\n",
    "plt.tight_layout()\n",
    "#plt.savefig('figures/longtail_%s.pdf'%dataset)\n",
    "\n",
    "plt.figure(2, figsize=(6,5))\n",
    "significant = [s.size for s in slices if s.size]\n",
    "bf_significant = [s.size for s in bf_slices if s.size]\n",
    "ai_significant = [s.size for s in ai_slices if s.size]\n",
    "#insignificant = [s.size for s in rejected if s.size]\n",
    "bins = np.linspace(0, max(significant), 50)\n",
    "bf_bins = np.linspace(0, max(bf_significant), 50)\n",
    "ai_bins = np.linspace(0, max(ai_significant), 50)\n",
    "plt.hist(significant, bins, alpha=0.5, label='t-Testing')\n",
    "plt.hist(bf_significant, bf_bins, alpha=0.5, label='Bon Ferroni')\n",
    "plt.hist(ai_significant, ai_bins, alpha=0.5, label='Alpha-Investing')\n",
    "#plt.hist(insignificant, bins, alpha=0.5, label='Rejected')\n",
    "plt.xlim([0,100])\n",
    "plt.legend(loc='upper right')\n",
    "plt.xlabel('Slice Size',fontsize=17)\n",
    "plt.tight_layout()\n",
    "#plt.savefig('figures/longtail_%s.pdf'%dataset)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}