BDT_weight/NN_TopFC_main_bkg_analysis.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DNN in TopFC analysis\n",
    "Author: Meisam Ghasemi Bostanabad"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading `pandas` DataFrames\n",
    "Now we load two different `NumPy` arrays. One corresponding to the VV signal and one corresponding to the background."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import uproot\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import h5py\n",
    "import awkward as ak\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# fix random seed for reproducibility\n",
    "seed = 7\n",
    "np.random.seed(seed)\n",
    "\n",
    "filename = {\n",
    "    'ttbar':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ttbar.root',\n",
    "    'WZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/WZ.root',\n",
    "    'ZZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ZZ.root',\n",
    "    'signal_charm':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/signal_charm.root'\n",
    "    }\n",
    "\n",
    "'''\n",
    "filename = {\n",
    "    'ttbar':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ttbar.root',\n",
    "    'WZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/WZ.root',\n",
    "    'ZZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ZZ.root',\n",
    "    'signal_up':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/signal_up.root'}\n",
    "    }\n",
    "'''\n",
    "\n",
    "upfile = {}\n",
    "params = {}\n",
    "df = {}\n",
    "\n",
    "# choose which vars to use (multi dimensional)\n",
    "#VARS = ['jetNo','jetPTLeading','bjetPT','elecPTLeading','met','testnonTopMass', 'weight'] # choose which vars to use (2d)\n",
    "VARS = ['jetNo','jetPTLeading', 'jetETA', 'jetPHI','bjetPT', 'bjetETA',\n",
    "        'elecPTLeading', 'elecETA', 'elecPHI',\n",
    "        'dielecCOS', 'dielecETA', 'met','testnonTopMass'] # choose which vars to use (2d)\n",
    "\n",
    "# define trees and branches to make DataFrame\n",
    "for treename in filename.keys():\n",
    "    upfile[treename] = uproot.open(filename[treename])\n",
    "    params[treename] = upfile[treename][treename].arrays(VARS)\n",
    "    df[treename] = ak.to_dataframe(params[treename])\n",
    "    # add isSignal variable\n",
    "    if treename.startswith(\"signal\"): \n",
    "        df[treename]['isSignal'] = np.ones(len(df[treename])) \n",
    "    else:\n",
    "        df[treename]['isSignal'] = np.zeros(len(df[treename])) "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data exploration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# reduce signal DF size to have resonable Sig/Bkg ratio\n",
    "df['signal_charm'] = df['signal_charm'][:3500]\n",
    "df_all = pd.concat([df[treename] for treename in filename.keys()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3500\n"
     ]
    }
   ],
   "source": [
    "print(len(df['signal_charm']))\n",
    "#print(len(df['signal_charm'][df['isSignal']==1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{0.0}\n",
      "{2, 3, 4, 5, 6}\n"
     ]
    }
   ],
   "source": [
    "print(set(df['ttbar']['isSignal']))\n",
    "print(set(df['ttbar']['jetNo']))\n",
    "#print(set(df['ttbar']['weight']))\n",
    "#print(set(df['signal_charm']['weight']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ttbar : (915, 14)\n",
      "WZ : (2176, 14)\n",
      "ZZ : (174, 14)\n",
      "signal_charm : (3500, 14)\n",
      "signal/tot bkg: 1.1\n"
     ]
    }
   ],
   "source": [
    "# number of entries in the backgrounds and signals\n",
    "tot_bkg = 0\n",
    "for treename in filename.keys():\n",
    "    if not 'signal' in treename: tot_bkg += df[treename].shape[0]\n",
    "    print(treename, \":\", df[treename].shape)\n",
    "\n",
    "print(\"signal/tot bkg:\", round((df['signal_charm'].shape[0]/ tot_bkg),1))\n",
    "#print(\"Null entries:\", df['signal_charm'].isna().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts_signal = df['signal_charm'].shape[0]\n",
    "counts_background = tot_bkg\n",
    "plt.pie([counts_signal, counts_background], shadow=True, labels=['Signal ','Background'], textprops={'color':\"b\"}, autopct='%1.1f%%');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnoAAAIbCAYAAAB8AjPzAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAAsTAAALEwEAmpwYAADbG0lEQVR4nOzdeXgUVdbA4d+pTtiXANkgQVCMgqggIiBBiAEhxAW/cRdEcZw4Ko4rLjCKG+6j4jgyMqCIgDqiImIA2VdFGRQRAQFBk0CShkAW1qT7fn9UE9JZ2FKhQ+e8PP3QVfdUnaqbm87tW5sYY1BKKaWUUsHHCvQGKKWUUkqpqqEdPaWUUkqpIKUdPaWUUkqpIKUdPaWUUkqpIKUdPaWUUkqpIKUdPaWUUkqpIKUdPaWUUkoph4jIuyKSLSI/V1AuIvKmiGwSkZ9EpFOJsltFZKPvdasT26MdPaWUUkop50wAko5Q3h+I871SgDEAItIUGAl0BboAI0WkSWU3Rjt6SimllFIOMcYsBnKOEDIAmGhs3wJhItIc6AfMMcbkGGN2AXM4cofxmIRUdgXqhOkjSZRSStUkclKTiVTV39k7sUfiDhlrjBl7HMvHAGklptN98yqaXyna0QsgkZPa5v0YY/h7h+Npl856bnUK7p4XByR3xOJvKExoH5DcAKEL15LZs0fA8kcvXhrw/D/FXx2w/OcvmxawtgfVo/39dsllAct/xpI5Ad//jT36BSR33NLZ5N/eLCC5ARq+u5Nvu98YsPzdln8UsNxO8nXqAvcH9DhpR08ppZRSQUek2nZxMoCWJaZjffMygIRS8xdWNpmeo6eUUkopdfJMBwb7rr7tBuQaY7YDs4G+ItLEdxFGX9+8Sqm23V2llFJKqRMVqBE9EfkQe2QuXETSsa+kDQUwxvwbSAWSgU3AXmCIryxHRJ4Fvvet6hljzJEu6jgm2tFTSimlVNCxrMB0cYwxNx2l3AD3VFD2LvCuk9ujh26VUkoppYKUjugppZRSKuhU44sxTiod0VNKKaWUClLa3VVKKaVU0NERPZvWglJKKaWCjnb0bHroVimllFIqSGlHrxobP348WVlZrFmzpsKY0aNHs3HjRlavXs0FF1zgaP647rHc98X1PPDlDfS8vUOZ8ouua8fQqddyz8d/4i8TriLijDBH84d26UaTSR/RdMon1B14SzkBoTR86lmaTvmEsH+Pw4qOdjS/dOlByMQZhEyeiXXzHeXmdz35KiGTZ+J6+0OIbuFo/lpduhI+aQrhUz6i/sBBZdN36ECzceOJmr+Q2r0SHM1dHfI36HoBZ3/4L87+eAwRg/5Uprx+h3OIe/cfnLfoUxonOP9Is5rc/up26Uzs5Hdp+eEEGg+8oUx5nQ7nETP+bU5fMIv6CZc4lveQQP/u1evamVZTxtHqo/doMuj6MuV1OpxLy/FvcebCVBokVM3jBGvf/AL1X/ieek8vxjrt/HJjav1pBPVf/YkGb//uaO7GXTvQ4cPX6PjfN2hxy1Vlyht2bMt5771A18WTaXppV0dzO0kkpEpep5oa29ETkeVHKAsTkbtLTLcWESMi95aY95aI3FaV2zhhwgSSkpIqLO/fvz9xcXHExcWRkpLCmDFjHMstlnDl8B5MvHsmb/7fJ5yXdGaZjtxPqZt469qp/OuGz1jy3mr6P+zgH1vLouEDD5E77EFyBt9End6X4WrV2i+kzuVXYvLzybn5Ovb99yPq/7Xc2xKdcH7XfSMoevSvFN16FVZiMrRq4x+SfA2mII+igf3xTp2IK+VBR/M3euBBdg17mB2DB1Gnd58y++/NyiL3+efZP3euc3mrUf6Yh+5ky0PP8OvAewnrcwm1W8f6hRzM2kHaqDfZPWdxleSvse3Psgh/8F4yHx5O2i130KDPpYS2Ps0vpCgrG/fzr1Awd74zOUvlD/TvXsSD95Dx8N/5fdBfaNjnUmqV2X83Wc//g/y5C5zLW4LrvD5YUWew5/GL2P/+g9QZ/Gq5cUU/zmbvsw4/t9gSTn/4dtY/9CKrb36IZn3iqds6xi/kYOZONj83hh1zljmbW1WJGtvRM8Z0P0JxGHB3qXnZwH0iUqvKNqqUJUuWkJNT8U2xBwwYwMSJEwFYsWIFYWFhRDs0qhB7bgQ703LZlZGPp8jLmlmbaZfQ2i/mwJ7C4ve16oaAMY7kBghpdw6ejHS827dBURH7582lVo+efjG1e1zC/lmp9rYsWkCtTp0dyy9tz8NkpMH2dCgqxDs/FSv+Uv+Y+ETMrC8AMIu+Ri7s5lj+0Hbt8GSk4ymx/3V6+I8ceDIzKfptMxivY3mrS/567eI4mL6dg9uyMEVF7J63lEaX+I8cFGZms3/z7xgH290hNbn91W53NoUZ2yjanglFReyZt5D6Pfw/Losyszi4eYujv/OHBPp3r067sylM30bRNnv/8+cupH4P/y+xxfvvdb7tA4Rc0J/C5R8D4P1tJVKvMdI4qkyc97eVmNwsR3M3OOdM9qdncmBbNqbIw865y2lyiX/bPpDpZu/mP8Dr/M/fSZYVUiWvU02N7eiJSIHv/2Ei8r2I/CQiT/uKXwTaiMiPIvKKb54bmAfcWs66OorIt751fO57Rl2Vi4mJIS0trXg6PT2dmJiYIyxx7BpF1ic3c0/xdF72HhpF1S8T1/WGc3hwxo30e6ArX71U4SDpcbPCI/BkZxdPe93ZuCIiysR4s30fch4PZk8B0rixMxsQEQXu7cWTxp1lzytBIiIx7szi/BTkQ+MwR9KX3n+P241Vav+rUqDzh0Y0pTB7R/F0YfZOQiOanrT8Nbn9hUSEU5TtLp4ucu/AFR5e6fUeswD/7oVENCuz/yERJ3H/AatJc0xORvG0N2cb0qT5ScldK6IpB7N2Fk8fdOdQ6yT+7inn1diOHoCI9AXigC5AR+BCEekJPAZsNsZ0NMYMK7HIS8DDIuIqtaqJwKPGmPOBNdjPtSsvX4qIrBSRlWPHjnV4bwJjxce/8NoVHzH7jRUk/KVToDdHKaWUAvQcvUNOvS12Vl/f6wffdAPsjt8f5QUbY34TkRXAzYfmiUhjIMwYs8g3633gkwqWHwsc6uGZO++8s1Ibn5GRQcuWLYunY2NjycjIOMISxy4vew+Now+P4DWKrE9e1p4K49fM2sxVI5w7Kdu7w40rMrJ42oqIxON2l4mxIqPwut3gciH1G2Byc53ZAHcWRBz+Bi0RUfa8Eow7G4mItkccXC5o0BBydzuSvvT+uyIi7P08SQKdv9CdQ2jk4VGU0MhmFLor/WzvY1aT21+RewchkYdHL0MiwvHs2HGEJRwW4N+9IvfOMvtf5K76/Q9N/DOhPe2LfjxbfkCaHj46YzVtgdm1vaJFHXXQnUOtqGbF07UimnLwJP7uOelU7JRVhRo9ogcI8IJv5K6jMeZMY8z4oyzzPPCob9mAmj59OoMHDwaga9eu5ObmkpmZ6ci6M9a6aXZaY5rENMQVYnFeUhvWL/K/sqvZaY2K35/V8zR2/uHQHzmgaP06XLEtsZo3h5AQ6vTuw8FlS/xiDixbSp2kZABq97qUg6v+51h+s+FnJPY0iI6BkFCsxGS8y/1PvDbLFyBJAwCQXn0xq1Y4lr9w/XpcsS1xldj/A8tO3onPgc6/d/1GasU2J7R5JBISQljvHuQt/e6k5a/J7e/A+g2ExsYQ0jwaQkKo3zuBPUu/cWTdxyLQv3v712+gVssYQppHQUgIDfsksGfZt46tvyKF88ez96kE9j6VQNEPqYR2t692ts7ojNmb5/i5eBUpWLeZOrHR1G4egYS4aNanO7uWOte21clX07u7s4FnRWSyMaZARGKAQiAfaFjeAsaY9SLyC3Al8L0xJldEdonIJcaYJcAtwKLylj1eU6ZMISEhgfDwcNLS0hg5ciShoaEAvPPOO6SmppKcnMymTZvYu3cvQ4YMcSItAF6PYcYLy7h1TH8sy+J/0zaQvXkXve++kIy1O1i/6He63tieNt1i8BZ62Zd/kE+fWOhYfjweCt74B41ffQOxLPanzsCzdQv1bv8LRRvWcXDZUvZ/9SWNRoyk6ZRP8ObnkffUE47m94weRcgrY8Gy8M78HLZuxhoyFLNhLWb5Arypn+Ia/iIhk2di8nLxPPOwo/nz3niNJq++BpbFvtSvKNq6hQa3/5nCDes5sGwZIW3b0uS555GGDandPR7P7X9m563l3AbklMzvZdvr/+GM10aCy8WuGXM5sCWNqDtuYt/6TeQt/Z66bc+k1QuPEdKwAY3iOxN1x038OuhvDuWvwe3P42XH628R/Y8XEMsi/6vZFG79nSZ/vpUD639l77JvqN32LKJGPYXVsAH1unejye2DSR/8F4fyB/p3z0v2a/8i5rXnwbLI++prDm75naZ/HsyB9b+yZ9m31G57Fs2ffxJXw4bUj+9G0z8P5o9bUpzbhJ/m4D3/Muq/uBJzcB/73y2+4QP1nlrI3qcSAKh93UhCul4LtepR/9U1FC75gINfvFzJ5F62vvYebV8fjrgssmcsYN+WdGLvuI49639j19L/Ub/dGZz1wkOENKxPWI9OxP75Wn4aNOzo6z7JdETPJlVxxdqpQETyjTENReQ+4NCNmgqAQcaYzSIyBTgfmAn8C5hhjDnXt2wH7MO9txtjJohIR+DfQD3gN2CIMWbXUTbBiARuUNAYw987BO48wedWp+Du6fy9z45FxOJvKExoH5DcAKEL15LZs2ruvXUsohcvDXj+n+KvDlj+85dNC1jbg+rR/n67xOFbchyHM5bMCfj+b+zRLyC545bOJv/2ZkcPrCIN393Jt91vDFj+bss/Oql/9Bo1Oq9KOjh5eWsCfkTveNTI7q6INANyAIwxo4HRpWOMMTeXmnVuibLVlDjsbYz5EXDu+n6llFJKVYqO6NlqXC2ISAtgIVD+HSiVUkopdco7Fe95VxVqXC0YY7YBZwV6O5RSSimlqlqN6+gppZRSKvjpoVtbTb+9ilJKKaVU0NLurlJKKaWCjo7o2bQWlFJKKRV0tKNn00O3SimllFJBSru7SimllAo6OqJn0xE9pZRSSqkgVWMfgVYNaMUrpZSqSU7qo8OiovpXyd/ZrKyZ+gg0dWwC/azZQD9rd0+vCwKSu/6iHwL+rM1V8dcELH+nZZ/ivbRdwPJbC9bxeIdxAcv/wuo7Atb2QNtfp2Wfkterc8DyN1q0knc6PxOQ3HeufJKXL3gtILkBHvnhQfZfGxuw/HWmpgcsd02mHT2llFJKBR09R8+mtaCUUkqpoKMdPZtejKGUUkopFaS0u6uUUkqpoKMjejYd0VNKKaWUClLa3VVKKaVU0NERPZvWglJKKaWCjmVpFwf00K1SSimlVNDS7q5SSimlgo4eurXpiJ5SSimlVJDS7m41F9c9luRHu2NZwv8+X8/id1f7lV90XTu63tAe4/FycF8R055ZjPu33Y7kHj9+PFdccQXZ2dmcd9555caMHj2a5ORk9u7dy2233cYPP/zgSG4AV5fu1Lp3GFgWRV9No3DKe/4BoaHUHv4s1lntMHm5HHj6UUzmdsfyS5ceuIY+Bi4X3q8+xTul1GO7QkNxPf4CcnZ7TO5uPM88BJnbHMvfqGtHYu+/HSyLnV/OI2vS537lDTqcQ+x9Q6jbphVbRr7G7oXfOpYbgIt6IEOHg8vCfDUVPiy7//L4S3DWOZC3G/P0g5Dl3P6f1T2WKx7thmUJ33++gUXv/lRuXPverRn0Wh/eumkaGb/scCx/TW5/gW57ri4XU+fehxHL4uBX0zg45X3/gNBQ6g5/Gpev7vc+/bijdd/y4jZ0f7gfYlmsn/YDP76/zK+83TUX0v66zhiPoXDfQRaPmsHuLc61vdO7t6b3sATEsvhp2hpWvPe9X3nHa8/ngus74vV6KdxbyOzn5rDztxzH8gOE3P4M1gWJcHAfhW89gNnyc9mYmx7B1etaqN+YA7ec7Wh+J+iInq3KR/REZPkRysJE5O4S061FZJ+I/Cgiv4jIv0Wkg2/6RxHJEZEtvvdzffFlW1/lt7nA938LEZnq9PqPeTss4crhPZh490ze/L9POC/pTCLOCPOL+Sl1E29dO5V/3fAZS95bTf+HL3Ys/4QJE0hKSqqwvH///sTFxREXF0dKSgpjxoxxLDeWRa37H2P/I0PZd+s1uHonIa3O8AsJufxqTH4++wYOoPCTydS68z5H87vuG0HRo3+l6NarsBKToVUb/5DkazAFeRQN7I936kRcKQ86mr/lQ39h00OjWDfwfpr06UGd1v7PqDyY5eb3UW+RM2eJc3lL5Jf7nsA8loK57Uqk9+Vl9p/kayE/FzMoCfPJROTOhx1LL5Zw1fDuvHf3bF7/v0/pkNSGyFJtH6BWvVDiB7bnj5+yHcsN1Oz2Vw3aXt37H2XvI3+j4NbrCO3dD6vV6X4hoZcPwOTnUzDw/zjwyRTq3HmvY+nFEuIf7U/q36bw3+ve5sx+7Qk7PdwvZtOsNUy98R0+HTiW1ROX0/2Bvo7m7/NYIp8M/Zzx10ygXVJbmp3R1C/ml5nree/6ibx/4yS+e/97Ln0wwbH8ANYFiUjz0zl4bw8K//0ooSkvlBvnWTmXA49d4WhuJ4mEVMnrVFPlHT1jTPcjFIcBd5eat9kY0xE4HzgHaGOM6eibNx0Y5pvuUwWb68cYs80Yc21V56lI7LkR7EzLZVdGPp4iL2tmbaZdQmu/mAN7Covf16obAsY4ln/JkiXk5FT8LXHAgAFMnDgRgBUrVhAWFkZ0dLQjua125+LNSMNsz4CiIjzzZxPSI8EvxhWfQNHsLwHwLJqLq1MXR3IDSNvzMBlpsD0digrxzk/Fir/UPyY+ETPrCwDMoq+RC7s5lr9+uzM5kJ7JwW1ZmKIids1bSuNLLvKLOZjpZt/m3x39mRdrez5s+6N4/838VIhP9AuR+ETMbHv/WTQbOjm3/y3PjWBnWl5x21896zfaJbQqE9f3ngtZ9N5PFB3wOJYbanb7C3Tbc7Vr71f3hfO/JqRHL7+Y0PheFM6eAUDRonmO1n1k+xjy0naRn7Ebb5GXTV+vpXUv/9Gqwj0Hi9+H1A11tBqanxvN7rTd5Gbk4i3ysm72es5M8O/kHyyRP7RuKODsz8G6qC+ehfYYh9m4Cuo1grDIMnFm4yrY7fCXLOW4kzGid2h0bJiIfC8iP4nI077iF4E2vhG6V0ouZ4wpApYDZ55AzgtFZJGI/E9EZotIc9/8v/i2YbWIfCoi9XzzTxeRb0RkjYg8V2I9xSOGInKbiHwmIrNEZKOIvFwi7s8i8quIfCci/xGRt453m8vTKLI+uZl7iqfzsvfQKKp+mbiuN5zDgzNupN8DXfnqpQoHUB0XExNDWlpa8XR6ejoxMTGOrFvCIzHZWcXTxp2FhEf4xVjhkZjsTHvC48HsKYDGYY7kJyIK3IcPBRl3lj2v5DZGRGLch/NTkO9Y/tCIphzMPnwoqDA7h9CIZo6s+5iER8KhugVwZyHhUaVioiDbV0de3/43CnMkfaPIemXafuOoen4xLdo2o3F0fTYsSSu9eKXV5PYX6LYn4ZF4/eo+Gys8suIYjwf2FCCNGzuSv15kQwqycoun92TnUT+yYZm49td15sZpQ+l2bx+WvTrLkdwADSIbkJ+VXzydn1VAw4iy+S+4vgN/mX47ve7rybyXFziWH0CaRWN2Hj4NwORsR5o58yX+ZNIRPdtJuRhDRPoCcUAXoCNwoYj0BB7DN4JnjBlWapl6QG9gzXHmCgX+CVxrjLkQeBcY5Sv+zBhzkTGmA7AO+LNv/mhgjDHmPOBIJ3p0BG4AzgNuEJGWItICeALoBsQDbY+wbSkislJEVo4dO/Z4duuIVnz8C69d8RGz31hBwl86ObZepaorEbj84W589Y8Vgd4UVUOt/WQlH139Fiv+OY9Of77kpOf/4b+r+c9V77Jo9BIuvqPrSc+vTh0n66rbvr7XD8Aq7M5QXAWxbUTkR2AZ8JUxZuZx5jobOBeY41vP34FDJ5icKyJLRGQNMBBo75sfD3zoe//BEdY9zxiTa4zZD/wCtMLuvC4yxuQYYwqBTypa2Bgz1hjT2RjTOSUl5ag7kpe9h8bRh0fwGkXWJy9rT4Xxa2Ztpt2lrY+6XqdkZGTQsmXL4unY2FgyMjIcWbfZkY1EHh7BkIgozA63X4x3RzYS6fuW6XIh9RtA7m5H8uPOgojmfvlxZ/mFGHc2EnE4Pw0aOpa/0J1DrcjD5wWFRjal0L3TkXUfkx3ZEFniG3xEFGZHVqmYLIj01ZHl2/+83Y6kz8veW6bt52btLZ6uVT+UqDObkDLuch5JvYGW50cwePRlxJwTXt7qjltNbn+BbntmRzaWX91H4t2RXXGMywX1G2Byc3HC3ux8GkQdHh2sH9mIPdn5FcZv+vpnWic4dyFCQXYBDaMOj+A1jGpAvrvi/Otmrycu4bgPfJXhSrqVWq/MptYrszG7spFmLYrLpGlzzM7MIyxdPVlWSJW8TjUnq6MnwAuHzrUzxpxpjBlfQeyhEb4LjDFPnWCutSVynWeMOXSm7ARgqG/k7mmgTonljuUkhwMl3nuo4quWM9a6aXZaY5rENMQVYnFeUhvWL/rdL6bZaY2K35/V8zR2/uHMh92xmD59OoMHDwaga9eu5ObmkpnpzIeBd/1arNjTkOgWEBKCK7EfRcsW+sV4li0ipN+VALh69cHzw/flrOnEmA0/I7GnQXQMhIRiJSbjXe5/eMQsX4AkDQBAevXFrHJudGnP+k3Ujm1OreaRSEgITXr3IHfpSsfWf1Tr10BMq+L9l8RkKG//+9n7T69+8INzV16mr3UTflojmsQ0wBVi0SHpDNaVaPsHCgp5LmESLyd/zMvJH5P2k5uJ981x7Krbmtz+At32POt/wYptWVz3oYl9KVq22C+mcNliQvvZFwGE9OrtaN1n/5JB45ZNadgiDCvE4sy+7fl98a9+MY1aHr44olWPs8j7w7krXrevzaTJaWE0btEIK8SiXb+2bFr4m19Mk9PCit+3ueQMdqXtqnRez6z3OTisHweH9cP73SxcCfbp6RLXCfbm67l4p7CT1TWdDTwrIpONMQUiEgMUAvlA2ZMPKmcDECEiFxtjvvEdyj3LGLPWl2u7b95A4NDw0zLgRmCSb/7x+B54Q0SaYO/PNRzn4eaKeD2GGS8s49Yx/bEsi/9N20D25l30vvtCMtbuYP2i3+l6Y3vadIvBW+hlX/5BPn1ioROpAZgyZQoJCQmEh4eTlpbGyJEjCQ0NBeCdd94hNTWV5ORkNm3axN69exkyZIhjufF4OPjGS9R59W379hapX2C2/kbo7XfhXf8LnuWLKEqdRu0Rz1F38heY/DwOPP2Yo/k9o0cR8spYsCy8Mz+HrZuxhgzFbFiLWb4Ab+qnuIa/SMjkmZi8XDzPOHfVKR4vaa+P48zXnkBcFjtnzGf/ljSa33Eje9dvInfpSuq1bcMZLzyKq2F9Gsd3pvkdN7Ju0P3O5Pd6MG8+h7w8DiwLM/Mz2LoJGXIvZsPPdqfvq6kw/CVk0izIy8U8+5AzubHb/vQXlnP7mP6IJayc9ivZm3fT5+5OZKzdwbpFfziWq1w1uf0Fuu15POx/4xXqvfpPxHJxMHU63q2/Ufv2O/GsX0fR8sUUpn5ByIhnaDD5c0x+HnufHu5MbsB4DEtfmUnyPwciLmHD9B/Z9Zubzncm4F63jd8X/8q5119ETJfT8RZ5OZC/nwVPfeFo/rkvLeC6t69BLGHNFz+z87ed9LirO5m/ZLJp0W9ccENHWnc9DU+RlwN5B/jqidmO5QfwrpqP1SmRWm8thQP7KXz78BXdtV6ZzcFh/QAIGTQC1yVXQ+261H7nezzzPqTov685ui2VcSqeT1cVxFTFFXslE4jkG2Maish9wB2+2QXAIGPMZhGZgn2F7UzgX8AMY8y5Faxrgq98qm+6NbARKHlM4wHfvDeBxtid2TeMMf8RkbuARwA3sAJoaIy5TUROB6YADYAvgPuNMQ18659hjDlXRG4DOhtjhvpyzwBeNcYsFJEUYBiQA6wH0o0xI45SNebvHZw7T+94Pbc6BREJWH5jDHt6XRCQ3PUX/UBhQvujB1aR0IVrWRV/TcDyd1r2Kd5L2wUsv7VgHY93GHf0wCrywuo7Atb2QNtfp2Wfkterc8DyN1q0knc6PxOQ3HeufJKXLwhcR+iRHx5k/7WxRw+sInWmpp/UPzpxcY9XSQdn48YXAvfH8wRUaXdXRJphd34wxozGvujBjzHm5lKzyu3k+WJvKzW9FQitILxnOcuPAcrc7M0YswUoeQO6v5dY/7m+9xOwD/0eWqbkzYOmGGPGiv314XNgWkX7oJRSSqngJSJJ2P0dFzDOGPNiqfLXgUP3S6oHRBpjwnxlHg4fFfzDGHNVZbenyjp6vqtRFwKvVlWOauQpEemDfc7f12hHTymllAqoQBy6FREX9tHJy4B04HsRmW6M+eVQjDHmgRLx9wIlDzHs89032DFVVgvGmG3AWVW1/urEGOPgyVlKKaWUOkV1ATYZY34DEJGPgAHYd+ooz03AyKrcoJN11a1SSiml1ElTVTdMLnlPXN+r5P3SYoCSd3FP980rZ/ukFXA6ML/E7Dq+dX4rIlc7UQ96SYpSSimlgo5VRWNZxpixgBNXU94ITDXGlHyGYytjTIaInAHMF5E1xpjNlUmiI3pKKaWUUs7IAFqWmI7l8K3cSruRww9rAMAYk+H7/zfs6xwqfYsA7egppZRSKugIVpW8juJ7IE5ETheRWtiduelltk2kLdAE+KbEvCYiUtv3Phz7qV0Vndt3zPTQrVJKKaWUA4wxRSIyFPtBES7gXWPMWhF5BlhpjDnU6bsR+Mj438y4HfCOiHixB+JeLHm17onSjp5SSimlgs4xjL5VCWNMKpBaat6TpaafKme55cB5Tm+PdvSUUkopFXQC1dGrbqr8EWiqQlrxSimlapKT+uiwc89+qUr+zv684VF9BJo6Nu6eFx89qIpELP4m4M/7DNSzdo0xAX/W6P6E8wOWv87Cn7iu3dcBy//Jur4Bf9ZuoJ/zHOj2R+LZAcvP/A0Bzx+o+q8OdW/6xAUsvczdeHLz6YgeoFfdKqWUUkoFLR3RU0oppVTQqaobJp9qtBaUUkoppYKUjugppZRSKujoOXo27egppZRSKuhoR8+mtaCUUkopFaR0RE8ppZRSQUdH9GxaC0oppZRSQUpH9JRSSikVdHREz6YdPaWUUkoFHb2Pnk1roZoL7dKNJpM+oumUT6g78JZyAkJp+NSzNJ3yCWH/HocVHe1YbleX7tT94HPqTv6C0JuHlJu79sgXqTv5C+qMmYhEN3csN8D48ePJyspizZo1FcaMHj2ajRs3snr1ai64wNlHukmXHoRMnEHI5JlYN99RNiA0FNeTrxIyeSautz+E6BaO5re6xFNr4nRqTZ6B6+bby80f+uTL1Jo8g1pvT0Yczt+xRzNGp8bzz1k9uPqO1mXKE65uwfhlCbzyWTde+awbidfGOJqfi3og76cik2bBTeXXvzz5GjJpFvL2RxDl3P4Huu1B4NsfAENHwAdfw3+mQ9w55cfEtYdx0+24oSOCIneNr3uAe56A9+fC2C/hzCPk/88MO+6eJ5zNrxxTIzp6IrL8CGVhInJ3ienWIrJPRH4s8RosIit87/8QEXeJstYiEuKb96KjG25ZNHzgIXKHPUjO4Juo0/syXK1a+4XUufxKTH4+OTdfx77/fkT9v97jWO5a9z/G/keGsu/Wa3D1TkJaneEXEnL51Zj8fPYNHEDhJ5Opded9zuT2mTBhAklJSRWW9+/fn7i4OOLi4khJSWHMmDHOJbcsXPeNoOjRv1J061VYicnQqo1/SPI1mII8igb2xzt1Iq6UBx3NH3LfcAofvYuDt16NK7F/mfp3Jf8JU5DHwYFXUDT1A0JS7ncyPX9+oh2jUlbxwJXLiL+8ObFt6peJWz4zk2F/+pZhf/qW+VMzHMuPZSH3PYF5LAVz25VI78vL1D/J10J+LmZQEuaTicidDzuWPqBtDwLf/gC69oSY1nBLX3jtCbj/qfLjHngK/vGEHRfTGrr0PLVz1/S6B+jSC2Jawa194PUn4L5nyo+772l47e92XEwruMih/A4RrCp5nWpOvS0+AcaY7kcoDgPuLjVvszGmY4nXRGNMV2NMR+BJ4OMSZVuBy4BfgevEwaelh7Q7B09GOt7t26CoiP3z5lKrh/8vUu0el7B/VioABxYtoFanzo7kttqdizcjDbM9A4qK8MyfTUiPBL8YV3wCRbO/BMCzaC6uTl0cyX3IkiVLyMnJqbB8wIABTJw4EYAVK1YQFhZGtEMjmtL2PExGGmxPh6JCvPNTseIv9Y+JT8TM+gIAs+hr5MJujuS285+LyfijRP3PKpPfik/AM2s6AN5Fc7Au7OpY/jPPb0zmH3vJTt9HUaFhWWomnRMjHVv/UbU9H7b9UVz/Zn4qxCf6hUh8Ima2Xf8smg2dnKv/QLY9CHz7A6B7b5gzzX6/bjU0aARNI/xjmkZAvQZ2Odjx8b1P6dw1vu4Buvcpkf9HaNDwCPl/LJH/MmfyK0fViI6eiBT4/h8mIt+LyE8i8rSv+EWgjW907pUTTHETMBr4A7i48ltss8Ij8GRnF0973dm4IiLKxHizs+wJjwezpwBp3LjSuSU8EnNovYBxZyHhpXNHYrIz/XLTOKzSuY9VTEwMaWlpxdPp6enExDh0+DAiCtzbiyeNO8ueV4JERGLch/efgnzH9l8iouycJfJLRGTFMR4PpsC5+m8aWYedmfuLp3Oy9tMsqnaZuK59o3h12sU89EYHmkWXLT9h4ZFwqG0BuLOQ8KhSMVGQ7fsZeX313yjMuW04giptexDw9gf46rfkzyDTnlc6xn2UmFMtd02v++J1H66DCvPvqKL8DtERPVuNuRhDRPoCcUAXQIDpItITeAw41zdah4i0xtfxK7H4vcaYJRWstw7QB7gTe3TwJqDcQ8UikgKkALzzzjv8X2V3SqkAWrnQzdKvtlNUaOhzfSxDXziPp4esDPRmKaUUoFfdHlJjOnpAX9/rB990A+yO3x/lxG4+1PE7BlcAC4wx+0TkU+AJEbnfGOMpHWiMGQuMPTTpnvTeEVfs3eHGFXl4FMeKiMTjdpeJsSKj8Lrd4HIh9RtgcnOPcdMrZnZkI5GHv51JRBRmR+nc2UhkNMadXZyb3N2Vzn2sMjIyaNmyZfF0bGwsGRkOnSfmzoKIwxeXSESUPa8E485GIqLtb/wul314w6H9t0fwStW/O7vcmEP5pYFz9Z+TvZ9m0XWKp5tG1WFn1gG/mILdhcXv509N55aH4xzJDcCObIgscSg0IgqzI6tUTBZENrf/t3z1n7fbuW04giptexC49jfgZrj8evv9hjWlfgbRdl2XtCPLnn+kmFMhd0k1se4BrhoIyTfY73/9ya8OKswfXgX1rxxXk7q7ArxQ4ty6M40x4x1Y701AHxHZCvwPaAYkHnGJY1S0fh2u2JZYzZtDSAh1evfh4DL/gcUDy5ZSJykZgNq9LuXgqv85kRrv+rVYsafZV3KGhOBK7EfRsoV+MZ5liwjpdyUArl598PzwvSO5j9X06dMZPHgwAF27diU3N5fMzMyjLHVszIafkdjTIDoGQkKxEpPxLl/gH7N8AZI0AADp1RezaoUjue38a5HYVkh0jK/+k/AuX+gX412+EFfSVQBYvS7Du+o7x/JvWpNH81b1iIypS0ioEJ8czcoF/h3NsIhaxe87J0aS/tsex/Kzfo19crev/iUxGcqr/352/dOrH/zwrXP5j6Iq2x4EsP19MQVSrrZfS+fCZVfb89t1gD35kOP/ZY8cN+wtsMvBjl8+79TLXUKNrHuA6ZPhr1fZr2Ul83c8Sv6OJfLPPfH8VUAP3dpq0ojebOBZEZlsjCkQkRigEMgHGp7ICkWkEXAJ0NIYc8A3bwh2529OpbfY46HgjX/Q+NU3EMtif+oMPFu3UO/2v1C0YR0Hly1l/1df0mjESJpO+QRvfh55Tzl0ibvHw8E3XqLOq2+DZVGU+gVm62+E3n4X3vW/4Fm+iKLUadQe8Rx1J3+Byc/jwNOPOZPbZ8qUKSQkJBAeHk5aWhojR44kNDQUsA99p6amkpyczKZNm9i7dy9DhpRzC5gT5fHgGT2KkFfGgmXhnfk5bN2MNWQoZsNazPIFeFM/xTX8RUImz8Tk5eJ5xrmrPvF4KBr9PKGvjAHLhWfmNMzWzYQMuRvvhl/wLl+IJ/VzQoc/T63JMyAvl8JnHnEsvddjGP/cekaM64RlCQs+yyB90x5uuLcNm3/OY+UCN8mDTqNzYiSeIkNBbiH/evxnx/Lj9WDefA55eRxYFmbmZ7B1EzLkXsyGn+1O31dTYfhL9u1X8nIxzz7kWPqAtj0IfPsDWLEIuvaCSXNg/z54efjhsrHT7A4JwBtPw6MvQO068N1iWLH41M5d0+seYMVC+8rbifPgwD54pcRn+7+n251BgDefgmEv+fIvsl+q2hFjTKC3ocqJSL4xpqGI3AccuilSATDIGLNZRKYA5wMzgX8B64ANJVbxrjHmTd+6bgM6G2OGisitQH9jzI0lcjX1LRt7qPNXAePu6dh1G8ctYvE37Onl/L2/jlX9RT/g4AXKx8UYQ2FC+4DkBghduJb9CecHLH+dhT9xXbuvA5b/k3V98V7aLmD5rQXrAtb2oHq0PxLPDlh+5m8IeP5A1X91qHvTx8FTLI6TzN14Un/x4s9+v0o6OMs23Bq4D5ATEPQjeiLSDMgBMMaMxr461o8x5uZSs+pWtD5jzARggu/9+8D7pcpzgIgyCyqllFJKnWRB3dETkRbAQuDVAG+KUkoppU6iU/F8uqoQ1B09Y8w24KxAb4dSSimlTi7t6Nm0FpRSSimlglRQj+gppZRSqmbSET2b1oJSSimlVJDSET2llFJKBR0d0bNpR08ppZRSQcfSjh6gh26VUkoppYKWjugppZRSKujooVtbjXgEWjWlFa+UUqomOamPDks8+9Mq+Ts7f8M1+gg0dWwC/bzLmpo/dOHagD/rNNDPuwz0z76m739Nb3+BftZxID97auq+g+93/yTSET2bdvSUUkopFXS0o2fTWlBKKaWUClI6oqeUUkqpoKMjejatBaWUUkqpIKUjekoppZQKOnrDZJvWglJKKaVUkNIRPaWUUkoFHT1Hz6YdPaWUUkoFHe3o2bQWlFJKKaWClI7oKaWUUiro6IieTWuhmpMuPQiZOIOQyTOxbr6jbEBoKK4nXyVk8kxcb38I0S2CInd1yD9+/HiysrJYs2ZNhTGjR49m48aNrF69mgsuuMDR/AAMHQEffA3/mQ5x55QfE9cexk2344aOcCx1oOsfqLH7Xy3aHgSu/i/qgbyfikyaBTeVX/fy5GvIpFnI2x9BVHB99tT4/VeOqlEdPRFZfoSyMBG5u8R0axHZJyI/isgvIvJvEbF8838utexTIvKw7/0EEbnWkQ22LFz3jaDo0b9SdOtVWInJ0KqNf0jyNZiCPIoG9sc7dSKulAcdSR3Q3NUhPzBhwgSSkpIqLO/fvz9xcXHExcWRkpLCmDFjHM1P154Q0xpu6QuvPQH3P1V+3ANPwT+esONiWkOXnpXPXQ3qvybvf8DbHgSu/i0Lue8JzGMpmNuuRHpfXqbuSb4W8nMxg5Iwn0xE7ny4cjlL5Q/0Z1+N3n8HCVaVvI6aVyRJRDaIyCYReayc8ttExO3rX/woIneUKLtVRDb6Xrc6UQ81qqNnjOl+hOIw4O5S8zYbYzoC5wPnAFdXyYZVQNqeh8lIg+3pUFSId34qVvyl/jHxiZhZXwBgFn2NXNjtlM9dHfIDLFmyhJycnArLBwwYwMSJEwFYsWIFYWFhREdHO7cB3XvDnGn2+3WroUEjaBrhH9M0Auo1sMvBjo/vXenU1aH+a/L+B7ztQeDqv+35sO2P4ro381MhPtEvROITMbPtumfRbOgURJ89NX3/HWRV0b8jEREX8C+gP3a/4SYRKW84/GNjTEffa5xv2abASKAr0AUYKSJNKl8PNYiIFPj+HyYi34vITyLytK/4RaCNr3f9SsnljDFFwHLgzJO6wRFR4N5+eDvcWfa8EiQiEuPOtCc8HijIh8Zhp3bu6pD/GMTExJCWllY8nZ6eTkxMjHMJwqMgO/PwtDvTnlc6xn2UmBNRHeq/pu//EVR524PA1X94ZKm8WUh5ebN9Px+vr+4bhVUu7yGB/tnX9P0/9XUBNhljfjPGHAQ+AgYc47L9gDnGmBxjzC5gDlDx0P4xqlEdPQAR6QvEYf8wOgIXikhP4DF8I3jGmGGllqkH9AYOnTDTpsSQ64/AX48xd4qIrBSRlWPHjnVmh5RSSilVhlTVvxJ/y32vlBJpY4C0EtPpvnmlXeMbbJoqIi2Pc9njUhOvuu3re/3gm26A3fH7o5zYNr6OnAG+MMbMFJHWHD6kC9jn6B1LYmPMWOBQD88UThl95AXcWRDRvHhSIqLseSXX6c5GIqLtb10uFzRoCLm7j2Vzqm/u6pD/GGRkZNCyZcvi6djYWDIyMiq30gE3w+XX2+83rIHIEofjIqJhh38dsCPLnn+kmBMRqPqv6ft/jKqk7UH1qP8d2aXyRmHKyxvZ3P7f8tV93u7K5T0k0D/7mr7/p4BSf8tPxJfAh8aYAyJyJ/A+kHiUZU5YjRvRAwR4ocSx8TONMeMriD00wneBMeapk7iNAJgNPyOxp0F0DISEYiUm412+wD9m+QIkyR4Vll59MatWnPK5q0P+YzF9+nQGDx4MQNeuXcnNzSUzM/MoSx3FF1Mg5Wr7tXQuXHa1Pb9dB9iTDzlu//gcN+wtsMvBjl8+r3LbQADrv6bv/zGqkrYH1aP+16+BmFbFdS+JyVBe3ffzHQ3r1Q9++LZyOUuuO9A/+5q+/w5yibdKXkeRAbQsMR3rm1fMGLPTGHPANzkOuPBYlz0RNXFEbzbwrIhMNsYUiEgMUAjkAw0Du2mleDx4Ro8i5JWxYFl4Z34OWzdjDRmK2bAWs3wB3tRPcQ1/kZDJMzF5uXiecejqq0Dmrg75gSlTppCQkEB4eDhpaWmMHDmS0NBQAN555x1SU1NJTk5m06ZN7N27lyFDhjianxWLoGsvmDQH9u+Dl4cfLhs7zf5jDPDG0/DoC1C7Dny3GFYsrnzualD/NXn/A972IHD17/Vg3nwOeXkcWBZm5mewdRMy5F7Mhp/tTs9XU2H4S/btR/JyMc8+VLmcJQW67df0/XeQxVE7ZVXheyBORE7H7qTdCNxcMkBEmhtjDp0IeRWwzvd+NvB8iQsw+gKPV3aDxBhT2XWcMkQk3xjTUETuAw5dzlwADDLGbBaRKdhX2M7EvmpmhjHm3FLraF16vu/QbYEx5lURmeArn3qUzTGFCe2d2K0TErpwLTU1f+jCtYhIQHIDGGMg8eyA5Wf+hoD/7Gv6/tf09ue9tF3A0lsL1gX0s6em7jtA6MK1J7XhX9N2bpV0cD5d3+eI+yEiycAbgAt41xgzSkSeAVYaY6aLyAvYHbwiIAe4yxiz3rfs7cChb1WjjDHvVXZ7a8yInog0w65QjDGjgTInyBljbi4169xyYraWnl/ysK4x5rZKb6xSSimlKsU6+mHWKmGMSQVSS817ssT7x6lgpM4Y8y7wrpPbUyPO0RORFsA3wKuB3hallFJKqZOlRozoGWO2AWcFejuUUkopdXK4AnOOXrVTI0b0lFJKKaVqohoxoqeUUkqpmiVAV91WO9rRU0oppVTQCdTFGNWNHrpVSimllApSOqKnlFJKqaCjF2PYdERPKaWUUipI6YieUkoppYKOnqNn046eUkoppYKOHrq11ahn3VYzWvFKKaVqkpP6rNs72n9WJX9nx639U+AeVn0CdEQvgDJ79ghY7ujFS1kVf03A8nda9in7E84PSO46C38K+EPdA/1Q+0A/WH3UBW8GLP+IH/4WsLYH1aP9rY2/MmDp2y/7ksKE9gHLH7pwLfMvHhKQ3InfvMdXXe8KSG6Ay1eMYf+1sQHLX2dq+knNp/fRs+nFGEoppZRSQUpH9JRSSikVdFx6MQagHT2llFJKBSE9dGvTQ7dKKaWUUkFKR/SUUkopFXT0Pno2HdFTSimllApSOqKnlFJKqaCjN0y2aUdPKaWUUkFHL8aw6aFbpZRSSqkgpSN61VitLl1p9Lf7wLLY99UM9kye5Fce2qEDje79GyFntGH3009xYNFCR/M36tqR2PtvB8ti55fzyJr0uV95gw7nEHvfEOq2acWWka+xe+G3jua3usQTMvRRcFl4vvoMz5R3/QNCQwl9fBRy9jmQm0vhM8Mwmdsc3QaGjoCuvWD/fnj5Mdj4S9mYuPbw6AtQuw6sWARvjap02vHjx3PFFVeQnZ3NeeedV27M6NGjSU5OZu/evdx222388MMPlc7r56IeyNDh4LIwX02FD8f5l4eGIo+/BGedA3m7MU8/CFnO1f8Z3VvRd1hPxBJ+nLaWb977n195p2vP5cLrz8d4DQf3FpL63Hx2/JbjWP6a3P4adO1E9P1/Acti95dz2DFpql95vQ7tib7vL9Rp05r0kS+Tt3B5pXOWJF164Br6GLhceL/6FO+Usm3P9fgLyNntMbm78TzzEDhY9027nUvc/TcjLovt0xfz+wepfuUtb+xLi6t6YjxeDu7OZ/2od9mfudOx/BHdzuGcB69HLCFt+jI2T/zar/z0m3rTckA8psjDwd0F/PTcB+zLdK7tA4Tc/gzWBYlwcB+Fbz2A2fJz2ZibHsHV61qo35gDtwTwaS8V0Pvo2WrEiJ6ItBaRMq1URMaJyDlHWfZ+EalXYnqriKwRkZ9E5GsRiRaRFSLyo4j8ISJu3/sfRaT1CW+0ZdHogQfZNexhdgweRJ3efXC18l+dNyuL3OefZ//cuSec5kj5Wz70FzY9NIp1A++nSZ8e1Gnt/+icg1lufh/1FjlzllRJ/pD7hlP46F0cvPVqXIn9kVZn+IW4kv+EKcjj4MArKJr6ASEp9zu7DV17QkxruKUvvPYE3P9U+XEPPAX/eMKOi2kNXXpWOvWECRNISkqqsLx///7ExcURFxdHSkoKY8aMqXROP5aF3PcE5rEUzG1XIr0vh1Zt/GOSr4X8XMygJMwnE5E7H3YsvVhC0mMJfDT0C965ZhLtk84i/IymfjE/z/yV/1w/hXE3fsg37/+PPg9e4lj+Gt3+LIvmD/2V3x96is0D76Fxn57Ubt3SL6Qwy03GqDfInbOocrkqyO+6bwRFj/6VoluvwkpMLtP2rORrMAV5FA3sj3fqRFwpDzqYXzj7oVtY/eDrrLhpBJGXdaVe6xZ+Ifm//sH3Q57hu1uexD1/JW3uud7R/O2H3ch397/FohufoUXfi2hwerRfSN6vaSy99QWWDBrF9vk/0Hbo/zmXH7AuSESan87Be3tQ+O9HCU15odw4z8q5HHjsCkdzK+fViI5eRYwxdxhjyvmK7Od+oF6peZcaY84HVgLDjTFdjTEdgSeBj40xHX2vrSe6baHt2uHJSMezfRsUFbF/3lzq9PB/Nq4nM5Oi3zaDcf5bS/12Z3IgPZOD27IwRUXsmreUxpdc5BdzMNPNvs2/g3H+udHS9lxMxh+Y7RlQVIRn/iys+Ev9Yqz4BDyzpgPgXTQH68Kuzm5E994wZ5r9ft1qaNAImkb4xzSNgHoN7HKw4+N7Vzr1kiVLyMmp+Bv6gAEDmDhxIgArVqwgLCyM6OjoCuOPW9vzYdsfsD0digox81MhPtEvROITMbO/sCcWzYZO3RxL3+LcKHLSdrM7Iw9vkZdfZm/krAT/jtbBPQeL34fWDQWca4c1uf3VbRfHwfTtFPp+93PnLabhJf77VpiZzYHNWzFV8rt/HiYjrbjteeenlql7iU/EzLLbnln0NXKhc22v0TlnsDc9m/3b3JgiD9lzvyOi5wV+MbtXrcd7wG5/uWs3UzuyiWP5w85pzd50N/u27cAUedg2ZyVRPTv4xez83694DxTa2/Lzb9RxMD+AdVFfPAvtUVyzcRXUawRhkWXizMZVsDvb0dxOsvBWyetUU5M6eiEiMllE1onIVBGpJyILRaQzgIj0FZFvRGSViHwiIg1E5G9AC2CBiCwoZ52LgTOrYmOt8Ag82Yd/gTxuN1ZExBGWcFZoRFMOZu8oni7MziE0otlJyy8RURh3VvG0cWchEZEVx3g8mIICaBzm3EaER0F25uFpd6Y9r3SM+ygxVSAmJoa0tLTi6fT0dGJiYpxLEB5Zat+zkPL2PXu7/d7rgYJ8aBTmSPqGkQ3Izyoons7LKqBhRP0ycRdefz53T7+V3vfFM/tl50aXanL7C41oRqHf7/5OQk7i7z4RUeDeXjxp3Fn2vBIkIhJzaL89vrbnUN3XjmjCgezDX7IOZOdQO6LijlSLK3uS880aR3ID1IkMY1/WruLp/dm7qBMRVmF8y6vicX+z1rH8ANIsGrPz8KFwk7MdaebgF0l1UtWkjt7ZwNvGmHZAHnD3oQIRCQf+DvQxxnTCHql70BjzJrANewTv0nLWeQVwzL/hIpIiIitFZOXYsWMrsStKKYD//fcn3r7qfeaPXkaPO7oEenNUDRPV72Iatm3N75NnBiR/TFIXGrdrxW+T5gQkf3VnibdKXqeamnQxRpoxZpnv/STgbyXKugHnAMtEBKAW8M0R1rVARDzAT9gdxGNijBkLHOrhmcxJEyuM9e5w44o8PILgiojA63Yfa6pKK3TnUCsyvHg6NLIphW7nTjY+GnsE5fC3eHv0JLvcGOPOApcLadAAcndXLvGAm+Fy3/k2G9ZAZIlvsRHRsCPLP35Hlj3/SDFVICMjg5YtD583FRsbS0ZGhnMJdmSX2vcoTHn7Htnc/t9yQYOGkLfbkfT52QU0jGpQPN0oqgH57j0Vxq+d/StJw8v7LnZianL7K3TvJNTvd78ZRSfxdx93FkQ0L56UiCh7XgnGnY1ERBfXPQ0aVr7ufQ64d1E78vD5oLUjm3LAvatMXJOLzqH1bVew6u4XMYVFjuQG2J+9m7pRh0cQ60Q2Yb97d5m4Zhe15czbkvjmrtfxOpDflXQrrt43A+DdvBpp1qL4ZAhp2hyzM7PihaspvY+erSaN6JU+maTktABzSpxbd44x5s9HWNelvrjBxpjdjm8pULh+Pa7YlriaN4eQEOr07sOBZcuOvqBD9qzfRO3Y5tRqHomEhNCkdw9yl648afnNhrVIbCskOgZCQnAlJuFdvtAvxrt8Ia6kqwCwel2Gd9V3lU/8xRRIudp+LZ0Ll11tz2/XAfbkQ06pznaOG/YW2OVgxy+fV/ntOIrp06czePBgALp27Upubi6ZmQ5+EK9fAzGtIDoGQkKRxGRY7n/2glm+AOk3wJ7o1Q9+cO6q621rs2h6WhiNWzTCCrE4p18cvy78zS+myWmNi9/HXXI6u9J2O5a/Jre/fes3Uiu2BaHNo5CQEBr37kn+Ugf27RiZDT8jsacVtz0rMRlveW0vyW570qsvZtUKx/Lnr9tCvZaR1GkejoS4iOzThR1L/K9ob3DWabR95FZ+GvYmhbvyHcsNkLvud+q3jKRu82ZIiIsWl3Uma/FPfjGNzorlvMdu5vthYzjoUH7PrPc5OKwfB4f1w/vdLFwJ1wIgcZ1gb361PhdPHVlNGtE7TUQuNsZ8A9wMLAWu9JV9C/xLRM40xmwSkfpAjDHmVyAfaAjsKHetVcXjIe+N12jy6mv27VVSv6Jo6xYa3P5nCjes58CyZYS0bUuT555HGjakdvd4PLf/mZ233uJQfi9pr4/jzNeeQFwWO2fMZ/+WNJrfcSN7128id+lK6rVtwxkvPIqrYX0ax3em+R03sm7Q/Q7l91A0+nlCXxkDlgvPzGmYrZsJGXI33g2/4F2+EE/q54QOf55ak2dAXi6FzzziTO5DViyyb20xaQ7s3wcvDz9cNnaa/ccY4I2nD9/e4rvFsGJxpVNPmTKFhIQEwsPDSUtLY+TIkYSGhgLwzjvvkJqaSnJyMps2bWLv3r0MGTKk0jn9eD2YN59DXh4HloWZ+Rls3YQMuRez4We70/fVVBj+EjJpFuTlYp59yLH0xmOY/dJCbnp7AJZlsfqLtez4LYeed3Vl+y/ZbFy0hc43dOD0ri3xFnnZl3eA6U84ePiqJrc/j5ftr/+bVq89jbgsds2Yy4EtfxBxx0D2r99I/tLvqNM2jtNeGI6rYQMaxl9ExB0D2TzonsrlLc7vwTN6FCGvjAXLwjvzc9i6GWvIUMyGtZjlC/Cmfopr+IuETJ6JycvF84xzV3wbj5df/zGZjm88hFgW22YsYc+WbZz+l6vJX7eVHUt/5Myh1+OqV5tzR9lnAO3P2smaR950LP/Pr35ElzfvRSyL9C+XU7BlO2elXMHudX+QveQn2t17DSH1atPp+b/Y+TN3sXKYc1fee1fNx+qUSK23lsKB/RS+ffiq5lqvzObgsH4AhAwageuSq6F2XWq/8z2eeR9S9N/XHNuOyjoVL5yoClIVV01VN77bnMzCPvfuQuAX4BYgFXjIGPM/EUkEXgJq+xb7uzFmuojcCwwFthljLhWRrUBnY0yZjp+I3OYrG3oMm2Uye/Y4elQViV68lFXx1wQsf6dln7I/4fyA5K6z8CdIDOA9n+ZvwHeKQEAYY/Be2i5g+a0F6xh1gTN/FE/EiB/+FrC2B9Wj/a2Nv/LocVWk/bIvKUxoH7D8oQvXMv9ih78YHaPEb97jq653BSQ3wOUrxrD/2tijB1aROlPTT+oH35Md/l0lHZxnVv81cB/gJ6BGjOj5bnPStvR8EWkG5Phi5gMXlY4xxvwT+GeJ6dZHyDMBmFDJzVVKKaVUJekNk201oqNXHhGZA6wxxmwJ9LYopZRSyll66NZWYzt6xpjLAr0NSimllFJVqcZ29JRSSikVvE7Fe95VhZp0exWllFJKqRpFR/SUUkopFXT0hsk27egppZRSKujoxRg2PXSrlFJKKRWkdERPKaWUUkFHL8aw6YieUkoppVSQ0hE9pZRSSgUdvRjDViOedVtNacUrpZSqSU7qM2Jf7/RylfydfWDVI/qsW3VsMnv2CFju6MVLA/5g++vafR2Q3J+s6xvwh6oHuu5FAvc5ZYwJeP0Hqu1B9Wh/nqQ2AcvvmrWZ/dfGBix/nanpAav/0IVrKUo+IyC5AUJSf8Pb98yA5be+3nRy8+k5eoB29JRSSikVhPTQrU0vxlBKKaWUClLa0VNKKaVU0LHwVsnraEQkSUQ2iMgmEXmsnPIHReQXEflJROaJSKsSZR4R+dH3mu5EPeihW6WUUkopB4iIC/gXcBmQDnwvItONMb+UCPsB6GyM2SsidwEvAzf4yvYZYzo6uU3a0VNKKaVU0HEF5mKMLsAmY8xvACLyETAAKO7oGWMWlIj/FhhUlRukh26VUkopFXQsPFXyEpEUEVlZ4pVSIm0MkFZiOt03ryJ/BmaWmK7jW+e3InK1E/WgI3pKKaWUUsfIGDMWGFvZ9YjIIKAz0KvE7FbGmAwROQOYLyJrjDGbK5NHO3pKKaWUCjoBOnSbAbQsMR3rm+dHRPoAI4BexpgDh+YbYzJ8//8mIguBC4BKdfT00K1SSimllDO+B+JE5HQRqQXcCPhdPSsiFwDvAFcZY7JLzG8iIrV978OBeEqc23eitKNXjdXq0pXwSVMIn/IR9QeWPVcztEMHmo0bT9T8hdTuleD8BlzUA3k/FZk0C266o2x5aCjy5GvIpFnI2x9BVAtH03fs0YzRqfH8c1YPrr6jdZnyhKtbMH5ZAq981o1XPutG4rVHOg3i+EmXHoRMnEHI5JlYN5e//64nXyVk8kxcb38I0c7ufyDrf/z48WRlZbFmzZoKY0aPHs3GjRtZvXo1F1xwgWO5Dwl0/df49gfIXU9ivTsfa8xXcGYFT5M481ysMalY785H7nrSsdwhtz9DrX8updY/5iCnn1t+zE2PUPvf31H7gw2O5YXqUffWnU/iGjcf179SoU3Fde96eyaucfOx7nSu7gHk7ieQ9+Yh/55R8c8+rj3yzld23N1POJrfCSF4quR1JMaYImAoMBtYB/zXGLNWRJ4Rkat8Ya8ADYBPSt1GpR2wUkRWAwuAF0tdrXtCgqajJyKtReTncuaPE5FzjrLs/SJSr8T0VhFZU+JeNm+KyL98738RkX0lyq71LTNNRL51bIcsi0YPPMiuYQ+zY/Ag6vTug6tVa78Qb1YWuc8/z/65cx1LWzK/3PcE5rEUzG1XIr0vh1alHpuUfC3k52IGJWE+mYjc+bCT6fnzE+0YlbKKB65cRvzlzYltU79M3PKZmQz707cM+9O3zJ9aZnS8Uhvgum8ERY/+laJbr8JKTC6z/1byNZiCPIoG9sc7dSKulAcdzR/I+p8wYQJJSUkVlvfv35+4uDji4uJISUlhzJgxjuUGAl7/Nb79AVyUgLRojff2RLyjR2ANfab8Tb33Gbyjh+O9PRFp0Ro69yo37nhYFyQizU/n4L09KPz3o4SmvFBunGflXA48dkWl8/knD3zdS+cEiGmN545EPG8OxzX02XLjXPc8i2f043juSISY1ogDdQ/ARb0gpjVmSG/MG39H/vZ0+dt57zOY10dghvSGmNZwUU9n8jvEhadKXkdjjEk1xpxljGljjBnlm/ekMWa6730fY0yUMaaj73WVb/5yY8x5xpgOvv/HO1EPQdPRq4gx5o5j6BHfD9QrNe/SEj+Evxlj7vHd2yYZ2FyibKqIhAEXAo19J1BWWmi7dngy0vFs3wZFReyfN5c6PfyfjevJzKTot81gquA8hLbnw7Y/YHs6FBVi5qdCfKJfiMQnYmZ/YU8smg2dujmW/szzG5P5x16y0/dRVGhYlppJ58RIx9Z/NNL2PExGWvH+e+enYsVf6h8Tn4iZZe+/WfQ1cqFz+x/o+l+yZAk5OTkVlg8YMICJEycCsGLFCsLCwoiOjnYsf6Drv8a3P0Au7oOZ97k9sf5HaNAImkb4BzWNgHoN7HLAzPsc6X5ZpXNbF/XFs3Cqvc6Nq6BeIwgrW/9m4yrYnV1mfmVUi7rvVqLuN/wI9RtBk1J138RX9xt+tLdj3udIt8rXPYB074OZU+JnX7+Cn339Ej/7Oc787JXzgq2jFyIik0VknYhMFZF6IrJQRDoDiEhfEflGRFaJyCci0kBE/ga0ABaIyIIjr75CfwK+BD7CPh5faVZ4BJ7swx9gHrcbKyLiCEs4LDwSsjMPT7uzkPCoUjFRkL3dfu/1QEE+NApzJH3TyDrszNxfPJ2TtZ9mUbXLxHXtG8Wr0y7moTc60Cy6bPkJi4gC9/biSePOsueVIBGRGLevjjy+/W8c5kz+ANf/0cTExJCWdvgOAunp6cTEOHjoMsD1X+PbHyDNojDubYdnuDOhWanOfLNo2HG4nRr3dqRZqXZ6QrmjMTsP5zY525HSuatKNah7wqMxJbdhRyaER5eNKVn35cWcqGb+dcCOTHtemZjMI8cEmEs8VfI61QRbR+9s4G1jTDsgD7j7UIHvxMa/A32MMZ2AlcCDxpg3gW3YI3glv7YtKHF49oGj5L0J+ND3uqmioJL33hk7ttJXZtd4Kxe6ubv3Yh6++htWL9/J0BfOC/QmqRpE259S6lQQbLdXSTPGLPO9nwT8rURZN+AcYJmIANQCvjnCui41xuw4WkIRiQLigKXGGCMihSJyrjGmzPmCpe69YzInTaxwvd4dblyRhw9VuCIi8LrdR9sc5+zIhsgS3w4jojA7skrFZEFkc/t/ywUNGkLebkfS52Tvp1l0neLpplF12Jl1wC+mYHdh8fv5U9O55eE4R3ID4M6CiObFkxIRZc8rwbizkYho+xu/y7f/ubudyR/g+j+ajIwMWrY8fAeB2NhYMjIcPEctwPVfU9ufXDkISbKfxGR+XYNEtMDwP7swIhp2ZvovsNN/FEkimmN2lmqnx8iVdCuu3jcD4N28GmnWAnNovU2bY0rnriqBqvsrbsHq56v7jT/ZdXmoLNx/5BSAHZn2/CPFHI8rByHJ19vvN6zxqwPCo6H0z3Vnlt0mjhQTYMdyPl1NEGwjeuYI0wLMKXFu3TnGmD87kPN6oAmwRUS2Aq05wqjesSpcvx5XbEtczZtDSAh1evfhwLJlR1/QKevXQEwriI6BkFAkMRmW+x/ZNssXIP0G2BO9+sEPzl2LsmlNHs1b1SMypi4hoUJ8cjQrF/ifixMWUav4fefESNJ/2+NYfrPhZyT2tOL9txKT8Za3/0n2/kuvvphVKxzLH+j6P5rp06czePBgALp27Upubi6Zmc79IQ50/dfU9me+nIT3nivx3nMl5puvkd7/Zxe07Qh78iGn1JfNHDfsLbDLAen9f5hvTuziMM+s9zk4rB8Hh/XD+90sXAnX2uuM6wR78x0/F68iAav7GR/gufcKPPdegflmzuG6P7ujXfe7StX9Ll/dn93R3o7e/4f5thIX5n05CXPXVfZr+RzksmP42e8p8bO/7P8wy6vgwkBVacE2oneaiFxsjPkGuBlYClzpK/sW+JeInGmM2SQi9YEYY8yvQD7QEDjqCF45bgKSfDkRkdOBudg3QjxxHg95b7xGk1dfA8tiX+pXFG3dQoPb/0zhhvUcWLaMkLZtafLc80jDhtTuHo/n9j+z89ZbKpW2mNeDefM55OVxYFmYmZ/B1k3IkHsxG362Ox1fTYXhL9m3/8jLxTz7kDO5Aa/HMP659YwY1wnLEhZ8lkH6pj3ccG8bNv+cx8oFbpIHnUbnxEg8RYaC3EL+9XiZQdQT5/HgGT2KkFfGgmXhnfk5bN2MNWQoZsNazPIFeFM/xTX8RUImz8Tk5eJ5xrmrXgNd/1OmTCEhIYHw8HDS0tIYOXIkoaGhALzzzjukpqaSnJzMpk2b2Lt3L0OGDHEsNxDw+q/x7Q/gu4WYixKw3p0PB/bjfe3R4iLrX1/ivcf+aPW+NRLroZehVm3MykXw/cJKp/aumo/VKZFaby2FA/spfPvwVa21XpnNwWH9AAgZNALXJVdD7brUfud7PPM+pOi/r1UueTWoe/P9AuSiBFzjF8CB/Xhef6S4zPXPGXjuta809rz9JK4HXobadTArF2FWLnRmA75bCF0SkAnz4cA+zKuHf/YyZjrmLvsuIeafI5FhL0OtOvD9IvtVjQTohsnVjhhTehDs1CQirYFZ2OfeXYh9k8FbgFTgIWPM/0QkEXgJOHTW9N+NMdNF5F7s+95sM8Zc6huZy4ficd+fjDGDS+SZYYw51/d+GRBrSlSkiKwC7jLGHOlrnsns2eMIxVUrevFSvJe2C1h+a8E6rmv3dUByf7KuL4UJFdwX6iQIXbg24HXvO30hIIwxAa//QLU9qB7tz5PU5uiBVcQ1azP7r40NWP46U9MDVv+hC9dSlOzIjRlOSEjqb3j7nhmw/NbXm07qB8/0rkOrpINz1Yq3AvcBegKCZkTPGLMVaFt6vog0A3J8MfOBi8pZ9p/AP0tMtz5KnnNLvC9zqaHvYg+llFJKqYAKmo5eeURkDrDGGLMl0NuilFJKqZNHL8awBXVHzxijd29USimlVI0V1B09pZRSStVMp+LNjauCdvSUUkopFXT00K0t2O6jp5RSSimlfHRETymllFJBR++jZ9MRPaWUUkqpIKUjekoppZQKOiF6jh6gHT2llFJKBSG9GMOmh26VUkoppYJU0Dzr9hSkFa+UUqomOanPiP22+41V8ne22/KP9Fm36tj8FH91wHKfv2waj3cYF7D8L6y+A++l7QKS21qwDhLPDkhuAOZvYNQFbwYs/Ygf/hawh7qD/WB3kcB9ThpjAtb2oHq0v0B/9uxPOD9g+ess/InFFw8OSO6e30zk5napAckNMGVdMrkpTQKWv/HYXQHLXZNpR08ppZRSQUfP0bPpOXpKKaWUUkFKR/SUUkopFXT0hsk27egppZRSKujoffRseuhWKaWUUipI6YieUkoppYKOXoxh0xE9pZRSSqkgpSN6SimllAo6LtERPdCOnlJKKaWCkB66tWlHrxpr0PUCYu6/AyyLnC/n4J70mV95/Q7n0OK+P1OnTWv+GPkquQu/cTT/Wd1jueLRbliW8P3nG1j07k/lxrXv3ZpBr/XhrZumkfHLDuc24KIeyNDh4LIwX02FD0s9ySM0FHn8JTjrHMjbjXn6Qcja5lx+gKEjoGsv2L8fXn4MNv5SNiauPTz6AtSuAysWwVujHEl9RvdW9B3WE7GEH6et5Zv3/udX3unac7nw+vMxXsPBvYWkPjefHb/lOJIbQLr0wDX0MXC58H71Kd4pZevf9fgLyNntMbm78TzzEGQ6U//jx4/niiuuIDs7m/POO6/cmNGjR5OcnMzevXu57bbb+OGHHxzJXawGt79Af/ZYXeIJGfoouCw8X32GZ8q7/gGhoYQ+Pgo5+xzIzaXwmWEYh9oeQJNu59Hm/kGIyyJz+iLSPpjhVx5zYxLRV/XCeDwU7s7n11HjOJC507H85/cIZ/Dwc7AsYcHUNL4c95tfec+rY7h5WFtysg4A8PWUrSycmu5YfoA6N7xIyHmXwcF97J1wN94/yn7+177679TqdiNSrzF5f2vpaH7lnCo5R09EtopI+Aksd5uIuEXkRxH5RUT+IiJDfNM/ishBEVnje/+iL/4th7c9QURm+N5fJSKPObn+Y2ZZxDx0J1seeoZfB95LWJ9LqN061i/kYNYO0ka9ye45ix1PL5Zw1fDuvHf3bF7/v0/pkNSGyDPCysTVqhdK/MD2/PFTtrMbYFnIfU9gHkvB3HYl0vtyaNXGPyb5WsjPxQxKwnwyEbnzYWe3oWtPiGkNt/SF156A+58qP+6Bp+AfT9hxMa2hS89KpxZLSHosgY+GfsE710yifdJZhJ/R1C/m55m/8p/rpzDuxg/55v3/0efBSyqdt5hl4bpvBEWP/pWiW6/CSkwuU/9W8jWYgjyKBvbHO3UirpQHHUs/YcIEkpKSKizv378/cXFxxMXFkZKSwpgxYxzLDdTs9hfgzx4si5D7hlP46F0cvPVqXIn9kVZn+IW4kv+EKcjj4MArKJr6ASEp9zuYXzjzocH8/OCrrLzpMSIu60a91i38Qgp+/Z0fhoxk1S1/Z8f87zn9nhsdSy8WDHmiPS+nfM+wKxfT/fIWxLRpUCbu25nbGf6npQz/01LHO3kh516GFdWGgr9fyL4P7qfuwH+UG1e0ehYFL/R2NLeTXOKtkteppjpejPGxMaYjkAA8D6QaYzr65m0DLvVNV3kHzBgz3RjzYlXnKU+9dnEcTN/OwW1ZmKIids9bSqNLuvrFFGZms3/z7xjj/HObW54bwc60PHZl5OMp8rJ61m+0S2hVJq7vPRey6L2fKDrg8BB52/Nh2x+wPR2KCjHzUyE+0S9E4hMxs7+wJxbNhk7dnN2G7r1hzjT7/brV0KARNI3wj2kaAfUa2OVgx8dX/oOvxblR5KTtZndGHt4iL7/M3shZCf5/7A7uOVj8PrRuKOBcO5C252Ey0orr3zs/FSv+Uv+Y+ETMLLv+zaKvkQudq/8lS5aQk1Px6OSAAQOYOHEiACtWrCAsLIzo6GjH8tfk9hfozx5pey4m4w/M9gwoKsIzf1aZtmfFJ+CZNR0A76I5WBd2LW9VJ6ThOW3Yl57N/m1uTJEH99xvadazk19M7qp1eA/Yv395azdTO9K558eeeX4YWX/sJTt9H55Cwzep27kwMcqx9R+LkI7JFH7zEQCeLSuRuo2RxmW3wbNlJSY366Rumzp+le7oicggEfnON8r2joi4jqVcRJJEZJWIrBaReaXXa4zJBjYDZXsXJ7hNIjJGRFaKyFoRebpEfJKIrBeRVcCfSswvHjEUkQki8qaILBeR30TkWt98S0Te9i0/R0RSD5VVRmhEUwqzDx8GLczeSWhE0yMs4axGkfXIzdxTPJ2XvYfGUfX8Ylq0bUbj6PpsWJLm/AaER0J25uFpdxYSXuqDJjwKsrfb770eKMiHRmEObkNUqW3ItOeVjnEfJeYENIxsQH5WQfF0XlYBDSPql4m78PrzuXv6rfS+L57ZLy+qdN5iEVHg3l48adxZ9rwSJCISc2jfPb76bxzm3DYcQUxMDGlph9tdeno6MTExziWowe0v0J89EhFltzcf485CIiIrjvF4MAUFjrW92hFNOJB9+DDsgewcakVU3JGLvrInu74p/7SWE9Eksg47M/cXT+dk7aNpVO0ycRf1jebFaT24740LaBpdx7H8AFZYc7y7Moqnza5tWGHNHc1xMoTgqZLXqaZSHT0RaQfcAMT7Rtw8wMCjlYtIBPAf4BpjTAfgunLWfQZwBrDJwW0aYYzpDJwP9BKR80Wkjm9brgQuBI40LNAc6AFcARwa6fsT0Bo4B7gFuPgI25bi62iuHDt27PHsVrUjApc/3I2v/rEi0JtSo/3vvz/x9lXvM3/0Mnrc0SXQm6NUjRLZrzsN255O2uTUk5p31cJs7uu9kMeuXsqa5Tu464XzT2r+U4ULT5W8TjWVvRijN3bn6HsRAagLZB9DeTdgsTFmC4AxpuQxmhtEpAdwALizVFllt+l6EUnB3u/m2J0zC9hijNkIICKTgJQK1j3NGOMFfhGRQ1+bewCf+OZnisiCijbMGDMWONTDMz+9X/GHQ6E7h9DIw6c5hkY2o9Dt3In2R5OXvZfG0YdHkBpF1ic3a2/xdK36oUSd2YSUcZcD0CC8LoNHX8bE++Y4c0HGjmyILNHnjojC7Ch1iGBHFkQ2t/+3XNCgIeTtrlzeATfD5dfb7zesKbUN0Xau0tsQcZSYE5CfXUDDqMPn5TSKakC+e0+F8Wtn/0rS8EsrLD9u7iyIOPwNXiKi7HklGHc2EhFtj6y4fPWfu9u5bTiCjIwMWrY8fPJ3bGwsGRkZR1jiONXg9hfozx57BO/wqKQ9epddbsyhticNGjjW9g64d1E7slnxdO3Iphx07yoTF3ZRe0677SpW3z0KU1jkSG6AXdn7aVZihK5pVN3iiy4OKdhdWPx+wdQ0bn64baXz1kq4g1qXDAbAs3UVVpOY4i6NNGmBd/f2ihdW1VplD90K8P6hc+iMMWcbY546jvLyfOyL7WqM+dypbRKR04GHgd7GmPOBr4DjHe8u+dsmJ7Btx2zv+o3Uim1OaPNIJCSEsN49yFv6XVWm9JO+1k34aY1oEtMAV4hFh6QzWLfo9+LyAwWFPJcwiZeTP+bl5I9J+8ntXCcPYP0aiGkF0TEQEookJsNy/z60Wb4A6TfAnujVD374tvJ5v5gCKVfbr6Vz4bKr7fntOsCefMhx+8fnuGFvgV0OdvzyMmciHLdta7NoeloYjVs0wgqxOKdfHL8u9L/yrslpjYvfx11yOrvSdlc67yFmw89I7GnF9W8lJuMtr/6T7PqXXn0xq07e6O706dMZPNj+o9S1a1dyc3PJzMw8ylLHoQa3v0B/9pgNa5HYVkh0DISE4EpMwrt8oV+Md/lCXElXAWD1ugzvKue2L3/db9RtGUWd5uFIiIuIPt3YucT/iu76Z7Ui7pHb+HnY6xTuyncsN8DmNblEt6pPRExdXKHCxcnN+d8C/857WMThQ7kXJkaR8VtB6dUct4MLx1HwbE8Knu1J4Y+phF5sX2DiOr0zZl/eKXkunks8VfI61VR2RG8e8IWIvG6MyRaRpkDDYyj/FnhbRE43xmwRkaYnMHJ3vNvUCNgD5PpG4/oDC4H1QGsRaWOM2QzcdJz5lgG3isj7QAT2RSRTKr0XHi/bXv8PZ7w2Elwuds2Yy4EtaUTdcRP71m8ib+n31G17Jq1eeIyQhg1oFN+ZqDtu4tdBf6t0agCvxzD9heXcPqY/Ygkrp/1K9ubd9Lm7Exlrd7Bu0R+O5DnCBmDefA55eRxYFmbmZ7B1EzLkXsyGn+0/ul9NheEvIZNmQV4u5tmHnN2GFYvsW1tMmgP798HLww+XjZ1m/zEGeOPpw7e3+G4xrKj8lYjGY5j90kJuensAlmWx+ou17Pgth553dWX7L9lsXLSFzjd04PSuLfEWedmXd4DpT8ypdN5iHg+e0aMIeWUsWBbemZ/D1s1YQ4ZiNqzFLF+AN/VTXMNfJGTyTExeLp5nnLvqdMqUKSQkJBAeHk5aWhojR44kNDQUgHfeeYfU1FSSk5PZtGkTe/fuZciQIY7lBmp2+wvwZw8eD0Wjnyf0lTFgufDMnIbZupmQIXfj3fAL3uUL8aR+Tujw56k1eQbk5VL4zCPO5AbweNn0j4mc+8YjiCVkzljM3i0ZtPrLn8hft4WcpT9wxtAbcdWrwzmjhgJwIGsnax95w5H0Xo9hwnNreWxcFywLFn6WTsamAq69N47ffs5l1YJs+g1qzYWJkXiKDAW5hbzzuHPnCAIUrfmakHMvo8GoVXBwH/sm3FNc1uCJxRQ8a1/ZXeeapwntcg3UqkfDl37m4NIPOPDlS45ui6o8qexVUyJyA/A49uhgIXAP8BHQ2Rizo7xyY8y3ItIf+6paC8g2xlwmIrf5lhtaQa6th9brm74NeAvYXSKsGxBfQc4JQHcgDcgFphtjJohIEvAGsBdYArQxxlxRcnt8y84wxkz15S4wxjQQEQt4G7uDl4Y90veSMeZof3XNT/FXHyWk6py/bBqPdxh39MAq8sLqO/Be2i4gua0F6yDx7IDkBmD+BkZd8GbA0o/44W8UJrQPWP7QhWvxnVYREMaYgLU9qB7tL9CfPfsTAndOWZ2FP7H44sEByd3zm4nc3O7kns9X0pR1yeSmOHeF8PFqPHbXSf3Fd/e82PnLwoGIxd8E7gPsBFT6hsnGmI+Bj0vNbn2UcowxM4GZpeZNACYcIVfrUtMVxVeU87YK1jsLKHOSQ8n1l17WGNPA979XRB42xhSISDPgO2BNRfuglFJKKXWy6JMxnDFDRMKAWsCzxhgHTxZSSiml1PE6FW9uXBW0o+cAY0xCoLdBKaWUUoedive8qwrV8ckYSimllFLKATqip5RSSqmgcyre3Lgq6IieUkoppVSQ0hE9pZRSSgWdU/HmxlVBO3pKKaWUCjp66Namh26VUkoppRwiIkkiskFENonIY+WU1xaRj33lK0SkdYmyx33zN4hIPye2R0f0lFJKKRV0AnEfPRFxAf8CLgPSge9FZLox5pcSYX8GdhljzhSRG4GXgBtE5BzgRqA90AKYKyJnGWMqNTSpI3pKKaWUUs7oAmwyxvxmjDmI/UjYAaViBgDv+95PBXqL/VzIAcBHxpgDxpgtwCbf+iql0s+6VSdMK14ppVRNcnKfEZt4dpX8nZUFv94JpJSYNdYYMxZARK4Fkowxd/imbwG6GmOGFi8v8rMvJt03vRnoCjwFfGuMmeSbPx6YaYyZWpnt1UO3AeTueXHAckcs/oY9vS4IWP76i34I2IPtjTEUJrQPSG6A0IVrA/5Q9+vafR2w/J+s64v30nYBy28tWBewtgfVo/2ReHbA8jN/Q8DzB6r+q0Pdmz5xAUsvczee1HzGqppDt75O3dgqWXkV0EO3SimllFLOyABalpiO9c0rN0ZEQoDGwM5jXPa4aUdPKaWUUkHHWKZKXkfxPRAnIqeLSC3siyuml4qZDtzqe38tMN/Y59FNB270XZV7OhAHfFfZetBDt0oppZRSDjDGFInIUGA24ALeNcasFZFngJXGmOnAeOADEdkE5GB3BvHF/Rf4BSgC7qnsFbegHT2llFJKBaFjGH2rmrzGpAKppeY9WeL9fuC6CpYdBYxycnv00K1SSimlVJDSET2llFJKBZ1AjehVN9rRU0oppVTw0Y4eoIdulVJKKaWClo7oKaWUUiroVNUNk081OqJXzYV26UaTSR/RdMon1B14SzkBoTR86lmaTvmEsH+Pw4qOdiy3q0t36n7wOXUnf0HozUPKzV175IvUnfwFdcZMRKKbO5YbYPz48WRlZbFmzZoKY0aPHs3GjRtZvXo1F1zg7JM+pEsPQibOIGTyTKyb7ygbEBqK68lXCZk8E9fbH0J0C0fzW13iqTVxOrUmz8B18+3l5g998mVqTZ5BrbcnIw7n79ijGaNT4/nnrB5cfUfrMuUJV7dg/LIEXvmsG6981o3Ea2Mczc9FPZD3U5FJs+Cm8utfnnwNmTQLefsjiHJu/wPd9iDw7Q+AoSPgg6/hP9Mh7pzyY+Law7jpdtzQEUGRu8bXPcA9T8D7c2Hsl3DmEfL/Z4Ydd88TzuZXjgnajp6IbBWR8BNY7jYRcYvIjyVeHUq8zxGRLb73c33LdBQRIyJJju6EZdHwgYfIHfYgOYNvok7vy3C1au0XUufyKzH5+eTcfB37/vsR9f96j2O5a93/GPsfGcq+W6/B1TsJaXWGX0jI5Vdj8vPZN3AAhZ9Mptad9zmT22fChAkkJVVcpf379ycuLo64uDhSUlIYM2aMc8ktC9d9Iyh69K8U3XoVVmIytGrjH5J8DaYgj6KB/fFOnYgr5UFH84fcN5zCR+/i4K1X40rsX6b+Xcl/whTkcXDgFRRN/YCQlPudTM+fn2jHqJRVPHDlMuIvb05sm/pl4pbPzGTYn75l2J++Zf7USt/A3W8D5L4nMI+lYG67Eul9eZn6J/layM/FDErCfDIRufNhx9IHtO1B4NsfQNeeENMabukLrz0B9z9VftwDT8E/nrDjYlpDl56ndu6aXvcAXXpBTCu4tQ+8/gTc90z5cfc9Da/93Y6LaQUXOZTfIQG6YXK1E7QdvUr62BjTscRr9aH32HeuHuab7uOLvwlY6vvfMSHtzsGTkY53+zYoKmL/vLnU6uH/i1S7xyXsn2XfrufAogXU6tTZkdxWu3PxZqRhtmdAURGe+bMJ6ZHgF+OKT6Bo9pcAeBbNxdWpiyO5D1myZAk5OTkVlg8YMICJEycCsGLFCsLCwoh2aERT2p6HyUiD7elQVIh3fipW/KX+MfGJmFlfAGAWfY1c2M2R3Hb+czEZf5So/1ll8lvxCXhm2Tdc9y6ag3VhV8fyn3l+YzL/2Et2+j6KCg3LUjPpnBjp2PqPqu35sO2P4vo381MhPtEvROITMbPt+mfRbOjkXP0Hsu1B4NsfAN17w5xp9vt1q6FBI2ga4R/TNALqNbDLwY6P731K567xdQ/QvU+J/D9Cg4ZHyP9jifyXOZPfIdrRswVFR09EBonId75RtndExHUs5SKSJCKrRGS1iMw7wdyCfePD24DLRKROZffnECs8Ak92dvG0152NKyKiTIw3O8ue8HgwewqQxo0rnVvCIzGH1gsYdxYSXjp3JCY70y83jcMqnftYxcTEkJaWVjydnp5OTIxDhw8josC9vXjSuLPseSVIRCTGfXj/Kch3bP8lIsrOWSK/RERWHOPxYAqcq/+mkXXYmbm/eDonaz/NomqXievaN4pXp13MQ290oFl02fITFh4Jh9oWgDsLCY8qFRMF2b6fkddX/43CnNuGI6jStgcBb3+Ar35L/gwy7XmlY9xHiTnVctf0ui9e9+E6qDD/jirKrxx1ynf0RKQdcAMQ7xtx8wADj1YuIhHAf4BrjDEd8L9L9Q2lDt3WPcImdAe2GGM2AwuBy4+wrSkislJEVo4dO/YE9lap6mPlQjd3917Mw1d/w+rlOxn6wnmB3iSllCqmI3q2YLjqtjdwIfC9PbhGXSD7GMq7AYuNMVsAjDElj9N8bIwZeoz5bwI+8r3/CBgMfFpeoDFmLHCoh2fck9474oq9O9y4Ig+P4lgRkXjc7jIxVmQUXrcbXC6kfgNMbu4xbnrFzI5sJPLwtzOJiMLsKJ07G4mMxrizi3OTu7vSuY9VRkYGLVu2LJ6OjY0lI8Oh88TcWRBx+OISiYiy55Vg3NlIRLT9jd/lsg9vOLT/9gheqfp3Z5cbcyi/NHCu/nOy99Ms+vDgdNOoOuzMOuAXU7C7sPj9/Knp3PJwnCO5AdiRDZElDoVGRGF2ZJWKyYLI5vb/lq/+83Y7tw1HUKVtDwLX/gbcDJdfb7/fsKbUzyDaruuSdmTZ848UcyrkLqkm1j3AVQMh+Qb7/a8/+dVBhfnDq6D+leNO+RE9QID3S5xPd7Yx5qnjKD/xxPYh4GuAJ0VkK/BPIElEGjqx/qL163DFtsRq3hxCQqjTuw8Hly3xizmwbCl1kpIBqN3rUg6u+p8TqfGuX4sVe5p9JWdICK7EfhQtW+gX41m2iJB+VwLg6tUHzw/fO5L7WE2fPp3BgwcD0LVrV3Jzc8nMzDzKUsfGbPgZiT0NomMgJBQrMRnv8gX+McsXIEkDAJBefTGrVjiS286/FolthUTH+Oo/Ce/yhX4x3uULcSVdBYDV6zK8q75zLP+mNXk0b1WPyJi6hIQK8cnRrFzg39EMi6hV/L5zYiTpv+1xLD/r19gnd/vqXxKTobz672fXP736wQ/fOpf/KKqy7UEA298XUyDlavu1dC5cdrU9v10H2JMPOf5f9shxw94Cuxzs+OUndBZMYHOXUCPrHmD6ZPjrVfZrWcn8HY+Sv2OJ/HNPPH9VsEzVvE4xwTCiNw/4QkReN8Zki0hToOExlH8LvC0ipxtjtohI01KjeseiN/CTMabfoRki8j7wf8DESu0VgMdDwRv/oPGrbyCWxf7UGXi2bqHe7X+haMM6Di5byv6vvqTRiJE0nfIJ3vw88p5y6BJ3j4eDb7xEnVffBsuiKPULzNbfCL39Lrzrf8GzfBFFqdOoPeI56k7+ApOfx4GnH3Mmt8+UKVNISEggPDyctLQ0Ro4cSWhoKADvvPMOqampJCcns2nTJvbu3cuQIeXcAuZEeTx4Ro8i5JWxYFl4Z34OWzdjDRmK2bAWs3wB3tRPcQ1/kZDJMzF5uXiece6qTzweikY/T+grY8By4Zk5DbN1MyFD7sa74Re8yxfiSf2c0OHPU2vyDMjLpfCZRxxL7/UYxj+3nhHjOmFZwoLPMkjftIcb7m3D5p/zWLnATfKg0+icGImnyFCQW8i/Hv/Zsfx4PZg3n0NeHgeWhZn5GWzdhAy5F7PhZ7vT99VUGP6SffuVvFzMsw85lj6gbQ8C3/4AViyCrr1g0hzYvw9eHn64bOw0u0MC8MbT8OgLULsOfLcYViw+tXPX9LoHWLHQvvJ24jw4sA9eKfHZ/u/pdmcQ4M2nYNhLvvyL7Fc1ovfRs4kxp17vtDQRuQF4HHuEshC4B/swamdjzI7yyo0x34pIf+B53/xsY8xlInIb8ApQ8jjM3caY5b5cE4AZxpipIvIesMIY8+8S23IVcJcxpv9RNtu4e15c2V0/YRGLv2FPL+fv/XWs6i/6Ad+h9JPOGENhQvuA5AYIXbiW/QnnByx/nYU/cV27rwOW/5N1ffFe2i5g+a0F6wLW9qB6tD8Szw5YfuZvCHj+QNV/dah708fBUyyOk8zdeFJ/8fbd2LxKOjh1P9oeuA+QExAMI3oYYz4GPi41u/VRyjHGzARmlpo3AZhwhFy3lXhf5mu8MWY69i1YlFJKKRUgp+KFE1UhGM7RU0oppZRS5QiKET2llFJKqZJ0RM+mI3pKKaWUUkFKR/SUUkopFXR0RM+mHT2llFJKBR/t6AF66FYppZRSKmjpiJ5SSimlgo7eMNmmI3pKKaWUUkFKR/SUUkopFXT0YgybdvSUUkopFXS0o2cLimfdnqK04pVSStUkJ/UZsbkpTark72zjsbv0Wbfq2AT6weY1NX/owrUBf6h9oB9sHuiffU3f/5re/ryXtgtYemvBuoB+9tTUfQff7/5JpCN6Nr0YQymllFIqSOmInlJKKaWCj47oAdrRU0oppVQQ0vvo2fTQrVJKKaVUkNIRPaWUUkoFHb0Yw6YjekoppZRSQUpH9JRSSikVdHREz6YjekoppZRSQUpH9JRSSikVdHREz6YjetWcdOlByMQZhEyeiXXzHWUDQkNxPfkqIZNn4nr7Q4huERS5q0P+8ePHk5WVxZo1ayqMGT16NBs3bmT16tVccMEFjuYHYOgI+OBr+M90iDun/Ji49jBuuh03dIRjqQNd/0CN3f9q0fYgcPV/UQ/k/VRk0iy4qfy6lydfQybNQt7+CKKC67Onxu+/UyxTNa9TTNB39ERkq4iEn8Byt4mIW0R+FJFfROQvJea/VSp2oYh0rky+clkWrvtGUPToXym69SqsxGRo1cY/JPkaTEEeRQP74506EVfKg46kDmju6pAfmDBhAklJSRWW9+/fn7i4OOLi4khJSWHMmDGO5qdrT4hpDbf0hdeegPufKj/ugafgH0/YcTGtoUvPyueuBvVfk/c/4G0PAlf/loXc9wTmsRTMbVcivS8vU/ckXwv5uZhBSZhPJiJ3Ply5nKXyB/qzr0bvv3Jc0Hf0KuljY0xHIAF4XkSiTmZyaXseJiMNtqdDUSHe+alY8Zf6x8QnYmZ9AYBZ9DVyYbdTPnd1yA+wZMkScnJyKiwfMGAAEydOBGDFihWEhYURHR3t3AZ07w1zptnv162GBo2gaYR/TNMIqNfALgc7Pr53pVNXh/qvyfsf8LYHgav/tufDtj+K697MT4X4RL8QiU/EzLbrnkWzoVMQffbU9P13kLG8VfI61QRVR09EBonId75RuHdExHUs5SKSJCKrRGS1iMwrvV5jTDawGWh1cvbEJyIK3NsPb4c7y55XgkREYtyZ9oTHAwX50Djs1M5dHfIfg5iYGNLS0oqn09PTiYmJcS5BeBRkZx6edmfa80rHuI8ScyKqQ/3X9P0/gipvexC4+g+PLJU3Cykvb7bv5+P11X2jsMrlPSTQP/uavv/KcUHT0RORdsANQLxvFM4DDDxauYhEAP8BrjHGdACuK2fdZwBnAJt8s27wdRZ/FJEfgc7HuI0pIrJSRFaOHTv2xHZUKaWUUkdlLFMlr1NNMF112xu4EPheRADqAtnHUN4NWGyM2QJgjCl5vOQGEekBHADuNMbk+Jb92Bgz9FCQiCw8lg00xowFDvXwTOGU0UdewJ0FEc2LJyUiyp5Xcp3ubCQi2v7W5XJBg4aQu/tYNqf65q4O+Y9BRkYGLVu2LJ6OjY0lIyOjcisdcDNcfr39fsMaiCxxOC4iGnb41wE7suz5R4o5EYGq/5q+/8eoStoeVI/635FdKm8Upry8kc3t/y1f3eftrlzeQwL9s6/p+++gU7FTVhWCZkQPEOB9Y0xH3+tsY8xTx1Feno99sV2NMZ9X1YZXxGz4GYk9DaJjICQUKzEZ7/IF/jHLFyBJAwCQXn0xq1ac8rmrQ/5jMX36dAYPHgxA165dyc3NJTMz8yhLHcUXUyDlavu1dC5cdrU9v10H2JMPOW7/+Bw37C2wy8GOX17m7IPjFrD6r+n7f4yqpO1B9aj/9WsgplVx3UtiMpRX9/3suqdXP/jh28rlLLnuQP/sa/r+K8cF04jePOALEXndGJMtIk2BhsdQ/i3wtoicbozZIiJNS43qBY7Hg2f0KEJeGQuWhXfm57B1M9aQoZgNazHLF+BN/RTX8BcJmTwTk5eL5xmHrr4KZO7qkB+YMmUKCQkJhIeHk5aWxsiRIwkNDQXgnXfeITU1leTkZDZt2sTevXsZMmSIo/lZsQi69oJJc2D/Pnh5+OGysdPsP8YAbzwNj74AtevAd4thxeLK564G9V+T9z/gbQ8CV/9eD+bN55CXx4FlYWZ+Bls3IUPuxWz42e70fDUVhr9k334kLxfz7EOVy1lSoNt+Td9/B+mInk2MCZ6KEJEbgMexRyoLgXuAj4DOxpgd5ZUbY74Vkf7A87752caYy0TkNt9yQ0vlKDPfd+j2YWPMShHZeijfUTbXFCa0r+wun7DQhWupqflDF67Fdwg+IIwxkHh2wPIzf0PAf/Y1ff9revvzXtouYOmtBesC+tlTU/cdIHTh2pPa8DP/HlIlHZzo54oC9wt8AoJpRA9jzMfAx6Vmtz5KOcaYmcDMUvMmABPKiS0z3xiTUOJ9a5RSSikVWNVwRM93NPFj7L7JVuB6Y8yuUjEdgTFAI+wLR0f5+i+IyASgF5DrC7/NGPPjkXIG0zl6SimllFJAtb2P3mPAPGNMHPYpZY+VE7MXGGyMaQ8kAW+ISFiJ8mElrjf48WgJtaOnlFJKKXVyDADe971/H7i6dIAx5ldjzEbf+23YdwiJKB13rLSjp5RSSqngY0mVvEreE9f3SjmOrYoyxhy6I3UmcMQ7jItIF6AW9kMbDhklIj+JyOsiUvtoCYPqHD2llFJKqapU6p64ZYjIXKC8ZxKOKLUeIyIVnkgoIs2BD4BbjTGHjhk/jt1BrOXbhkeBZ460vdrRU0oppVTwCdAxS2NMn4rKRCRLRJobY7b7OnLZFcQ1Ar4CRhhjim+UWGI08ICIvAcc9d42euhWKaWUUsGnig7dVtJ04Fbf+1uBL0oHiEgt4HNgojFmaqmy5r7/Bfv8vp+PllA7ekoppZRSJ8eLwGUishHo45tGRDqLyDhfzPVAT+A2EfnR9+roK5ssImuANUA48NzREuqhW6WUUkoFn8qPvjnOGLMT6F3O/JXAHb73k4BJFSyfeLw5dURPKaWUUipI6YieUkoppYJPNRzRC4SgetbtKUYrXimlVE1yUnte218OrZK/s80fKTylepA6ohdAv11yWcByn7FkDqvirwlY/k7LPg3cg9Xnbwj4Q93Xxl8ZsPTtl32JJ6lNwPK7Zm3mp/irA5b//GXTAv7zD3R++4K9wDDGkNerc8DyN1q0EnfPiwOSO2LxNwH/3A/07/5JpSN6gHb0lFJKKRWM9CoEQKtBKaWUUipo6YieUkoppYKPHroFdERPKaWUUipo6YieUkoppYKPjugB2tFTSimlVDDSjh6gh26VUkoppYKWjugppZRSKvjoiB6gI3pKKaWUUkFLO3rVWN0unYmd/C4tP5xA44E3lCmv0+E8Ysa/zekLZlE/4RLH8zfq2pFzPnyTcz5+i6hB/1emvEGHc2j77itcsOi/hCV0czw/AENHwAdfw3+mQ9w55cfEtYdx0+24oSOCJn+Drp0488MxnPnxO4QPurZMeb0O7Tnj3Tc4Z9E0GiV0dyxvSXLXk1jvzsca8xWc2b78oDPPxRqTivXufOSuJx3L3aDrBZz94b84++MxRAz6U5ny+h3OIe7df3Deok9pnFBFTzqooe1v/PjxZGVlsWbNmgpjRo8ezcaNG1m9ejUXXHCBI3kPcXW5mPoffEqDyZ9T6+ZbywaEhlJ35PM0mPw59cdMQKKbO5o/tEs3mkz6iKZTPqHuwFvKzd/wqWdpOuUTwv49Dis62tH8gf7sh8D+7jvGqqLXKeaU2GQReUpEHhaRZ0Skz1FiJ4hI2b+Kx5bnLBFJFZGNIrJKRP4rIlG+sh4i8p2IrPe9Ukosd7aILBSRH0VknYiMPZH8fiyL8AfvJfPh4aTdcgcN+lxKaOvT/EKKsrJxP/8KBXPnVzpdeflbPvQXNj00inUD76dJnx7UaR3rF3Iwy83vo94iZ84S5/MDdO0JMa3hlr7w2hNw/1Plxz3wFPzjCTsupjV06Xnq57csmj/0V35/6Ck2D7yHxn16Urt1S7+Qwiw3GaPeIHfOosrnK89FCUiL1nhvT8Q7egTW0GfK39R7n8E7ejje2xORFq2hc6/K57YsYh66ky0PPcOvA+8lrM8l1C7T/naQNupNds9ZXPl85anB7W/ChAkkJSVVWN6/f3/i4uKIi4sjJSWFMWPGVDpnMcui7v2PsveRv1Fw63WE9u6H1ep0v5DQywdg8vMpGPh/HPhkCnXuvNfR/A0feIjcYQ+SM/gm6vS+DFer1n4hdS6/EpOfT87N17Hvvx9R/6/3OJo/oJ/9ENjffSdZUjWvU8wp0dE7xBjzpDFmblWsW0TqAF8BY4wxccaYTsDbQISIRANTgL8aY9oCPYA7ReRy3+JvAq8bYzoaY9oB/6zs9tRudzaFGdso2p4JRUXsmbeQ+j38R22KMrM4uHkLGOef21y/3ZkcSM/k4LYsTFERu+YtpfElF/nFHMx0s2/z71WSH4DuvWHONPv9utXQoBE0jfCPaRoB9RrY5WDHx/c+5fPXbRfHwfTtFPrqP3feYhpe0tUvpjAzmwObt2KqqP7l4j6YeZ/bE+t/PPL+r/8RADPvc6R75Z/lWc+3/4fa3+55S2lUzv7v3/x7le1/TW5/S5YsIScnp8LyAQMGMHHiRABWrFhBWFgY0Q6NarnatcebkYbZngFFRRTO/5qQHv4diND4XhTOngFA0aJ5uDp1cSQ3QEi7c/BkpOPdvg2Kitg/by61evh3nmv3uIT9s1IBOLBoAbU6Offs3kB/9kNgf/eV86ptR09ERojIryKyFDjbN694tE5ELhSRRSLyPxGZLSJlxu4rihGRM0Vkrois9o3ctQFuBr4xxnx5aHljzEJjzM/APcAEY8wq3/wdwCPAY77Q5kB6ieUqPt5xjEIiwinKdhdPF7l34AoPr+xqj1loRFMOZu8oni7MziE0otlJyw9AeBRkZx6edmfa80rHuI8ScwrmD41oRqFf/e8k5CTXvzSLwri3HZ7hzoRmpf6YN4uGHYf337i3I82c2P+mZfY/NKJppdd7XGpw+zuamJgY0tLSiqfT09OJiYlxZN0SHok3O6t42rizscIjK47xeGBPAdK4sSP5rfAIPNnZxdNedzauiIgyMSXzGwfzB/qzHwL7u+8oHdEDqmlHT0QuBG4EOgLJwEWlykOxR82uNcZcCLwLjDqOmMnAv4wxHYDuwHbgXOB/FWxS+3LKVvrmA7wOzBeRmSLygIiEVbBfKSKyUkRWjh1b+aO7SimllFJHUl1vr3IJ8LkxZi+AiEwvVX42dsdsjogAuLA7a0eNEZGGQIwx5nMAY8x+X44T3lhjzHsiMhtIAgZgH9btYIw5UCpuLHCoh2d+++CTCtdZ5N5BSOThb5EhEeF4duyoMN5phe4cakUe/hYZGtmUQvfOqk884Ga4/Hr7/YY1EFniW2RENOzI8o/fkWXPP1LMqZTfp9C9k1C/+m9G0Umof7lyEJJkn/xtfl2DRLTAHPqOExENOzP9F9iZCeGH918immN2OrH/OWX2v9Bd8aFExwT65x/o/McoIyODli0PnzMaGxtLRkaGI+s2O7KxIg+PDElEJN4d2eXGeNzZ4HJB/QaY3FxH8nt3uHFFHh5BtCIi8bjdZWKsyCi8bje4XIiD+QP12V9dfvcddQqOvlWFajmidwwEWOs7J66jMeY8Y0zfE4gpaS1wYQVlv5RTdqFvGQCMMduMMe8aYwYARdidzBN2YP0GQmNjCGkeDSEh1O+dwJ6l31Rmlcdlz/pN1I5tTq3mkUhICE169yB36cqqT/zFFEi52n4tnQuXXW3Pb9cB9uRDjv8HLjlu2Ftgl4Mdv3zeqZvfZ9/6jdSKbUFo8ygkJITGvXuSv/S7Sq/3aMyXk/DecyXee67EfPM10tt3tXXbjkfe/7YdAZDe/4f5pvKn0e5dv5Fasc0J9bW/sN49yDsJ+x/wn3+g8x+j6dOnM3jwYAC6du1Kbm4umZmZR1nq2HjW/4IV2xKJbgEhIYQm9qVomf8FN4XLFhPa7woAQnr1xvPD947kBihavw5XbEus5s0hJIQ6vftwcJn/BWcHli2lTlIyALV7XcrBVRUdDDp+gfrsry6/+8p51XVEbzEwQURewN7GK4F3SpRvwL5I4mJjzDe+w7RnGWPWHkuMiKSLyNXGmGkiUht7tG8K8LiIXG6M+QpARHoCOcC/gBUi8pkx5kcRaQa8BDzji0sC5hljCn0XbjQDKvf11uNlx+tvEf2PFxDLIv+r2RRu/Z0mf76VA+t/Ze+yb6jd9iyiRj2F1bAB9bp3o8ntg0kf/JdKpS2ZP+31cZz52hOIy2LnjPns35JG8ztuZO/6TeQuXUm9tm0444VHcTWsT+P4zjS/40bWDbrfmfz/3959h0dVZg8c/55JIUDoBAIBKYqgWFBpAiICAmLDtXdxFXtZ+8oq6q7rigVQVxbUFVFYV/GnIiJKR0BBBBRZQKoCAgk9CCHJzPn9cW+SSUgBptxkcj7PM0/m3vvOnHdu7sy881aABbOh09nw3lTIOgBDHy84NvoT58sQYPjT8OhzUCUJFs6BBWEahellfH+ALcP+RbOXn0bifOyaNI2D638l5ZZryVq5msy5C0lq04pjnnucuBrJ1OjagZRbrmXtdWEc/bdwFtqhB75/z4CDWQRefjT/kO+fnxG460IAAq8NwffgUEisgi6aDd/NCj22P8Bvw96g5ctDIC7Off0baXjL1RxYuYa9c7+japvjaPbcY8TXSKZm1/Y0vOVqfr7u3tBj56nE19/48ePp0aMH9evXZ+PGjQwZMoSEhAQARo0axeTJk+nfvz9r1qxh//79DBw4MOSY+fx+soa/QLUXX0V8cWRPnkhgwzqq3Hwb/pUryJ0/h5zJnxI/+BmSx32MZu5l/9OPl/28RxB/3/CXqPXicMTnI2vyJPwb1lPt5lvJXbWC7Hlzyfr8M2oOHkLd8R8SyNzL3qeeCGN8jz/7wdv3fjhZjR4AErERayESkcHAjUA68CuwGKeWbJKqThCRdjijXWvhFAaHq+obIjLmMNK0wik41gdygMtVdZ2ItAGGA8e6+38E7lPVbW6h7yWgBk5t4XBVHenm9WXgfCDLzf4LqvpeGS9R153l3Qilll9PZXHXSz2Lf/q8j6Bna2+Cz1jlXWw3/vKuF3oWvu28z/D3O9az+HFT1vJj1wGexT9l3iee//+9jh9KV5VQqSp7zw7fKNUjVXP2IjK6R2jexTKkzPkGrz/3PX7vR/XC2zK2ZkQKOI1u2FuhSpDltUYPVX2WIgMsihxfChwyYZSq3nQYaVYDPYvZvxKnn11x8eZQZFBI0LEHgAdKyqsxxhhjjBfKbUHPGGOMMeaoWdMtUHEHYxhjjDHGmDJYjZ4xxhhjYo/V6AFW0DPGGGNMLLKCHmBNt8YYY4wxMctq9IwxxhgTe6wqC7DTYIwxxhgTs6xGzxhjjDGxx/roAVbQM8YYY0wssoIeYE23xhhjjDExq9yudVsJ2Ik3xhhTmUR3rdtP6kdmrdsB2ytUVaE13Xoop0dbz2InzFru+cLini3sPmMVgXNO8CY24Ju5wvP/fdZlTTyLnzRhE1k9TvEu/qwfvbv2oFxcf16/90W8+55UVW8/e/oc501swPfVGs8/e0z0WUHPGGOMMbHH+ugB1kfPGGOMMSZmWY2eMcYYY2KP1egBVtAzxhhjTCyyNkvAToMxxhhjTMyyGj1jjDHGxB5rugWsRs8YY4wxJmZZjZ4xxhhjYo/V6AFW0DPGGGNMLLKCHmBNt+WedOxG/NhJxI/7At81txyaICGBuCdfJH7cF8S9/h9IbRy22HEdz6T6ux+RPO5jEq+5sdjYVYf8neRxH1N95BgktVHYYue7ezC8+xW8MRFanVh8mlZt4c2JTrq7B4cvdoduyDuTkfemwNXFn3t58mXkvSnI6+9Dw/Cde/D2f58n/uZnSHx1LokvTUVanFR8mqsfocq/FlLl3VVhje3r2JXEsRNJHDeJuGtuPjRBQgIJTw4lcdwkEl8fh0Tg9VfW68/r9/5bb73Ftm3bWLZsWYlpRowYwerVq/nhhx847bTTwhof8PZ/D8idTyBvT0f+NQmOK2E1i1ZtkVGfO+nufCJ8scvBZ48JnwpV0BORp0TkIRF5RkR6l5F2jIhcdhQxmovIARFZGnS7QUQWuPd/FZGMoGPNRSTe3fePo391xfD5iLtvMLmP3k7ujRfh69kfmh1bOEn/S9F9e8m99jwCE8YSN+iBsMWuev+j7H/kXvbdeDkJvfria9aiUJKE8y9GMzPZd+0lHPxwPEm33ROe2Hk6dYe05nB9H3j5Cbj/qeLT/ekpeOkJJ11ac+jYPfTYPh9y3xPoY4PQmy5Eep1/yLmn/2WQuQe9rh/64VjktodCjxsU37P/fd7zn9YTadSC7Hu6kfOvR0kY9Fyx6fyLpnHwsQvCGhufj/j7Hifn0TvIvnEAcT3PQ5q1LJQkrv8f0H17yb72AnInvEv8oPvDm4fKev2Vg/f+mDFj6NevX4nHzzvvPFq1akWrVq0YNGgQI0eODGt8T//3AB3OhrTm6MBe6PC/IPc+XWwyuecZdNhgdGAvJ36H8Fx7Xn/2hI1PInMLgYjUFZGpIrLa/VunhHT+oHLGxKD9LdzyyBoR+a+IJJZ5GkLKsUdU9UlVnRbBEGtVtV3QbayqdlLVdsCTwH+Djm0AzgV+Bi6XMC7iKG1ORjdvhC2bIDeHwIzJ+LqeUzhN157olE8B0NlfIWd0DkvsuBPaEti8Ed2yGXJzyZnxFfHdzi6UJqHr2eR8OQmA3NnTiTu9Y1hi5+vSC6Z+4txf8QMk14S6KYXT1E2BasnOcXDSd+0Veuw2p8Bvv+afe50xGbr2LJREuvZEv3TOPbO/hNPDc+7B2/99Hl+HPvhnTXCef/ViqFYTajc4JJ2uXgy708MaW9qchG7+Nf/688+Ycsjr93XtgX+K8/kXmD0V3xmdwpqHynr9lYf3/tdff83OnTtLPH7xxRczduxYABYsWEDt2rVJTU0NXwa8/N8D0qU3OvVjZ2PlUqheQvzqyc5xQKd+jHQ5N/TY5eCzJ8Y9BkxX1VbAdHe7OAeCyhkXBe1/HhimqscBu4A/lhWw3Bf0RGSwiPwsInOB1u6+/No6ETlDRGaLyPci8qWIHNKGUFIaETlORKaJyA8islhEji362MN0NTAC+BU48yif41ApDSFjS/6mZmxz9gWRlAZoxlZnw++HfZlQq3bIoaV+AwLp24Jip+Or36DkNH4//L4PqVUr5Nj56jeE9K0F2xlbnX1F02SUkeaoYjcoEnsbUlzsdPf/E3DPfc3aoccGT//3+c9fLxXd8VtBHnZuQeqF8cu0tNgpDZ3XnBc7YxuS0qDkNH4/um9fWF9/Zb3+ysV7vwxpaWls3Lgxf3vTpk2kpaWFL4CX/3uAeoXf/2zf6uw7JM3W0tMcjXLw2RM2vgjdQnMx8I57/x1gwOE+0K1I6glMOJLHl+uCnoicAVwFtAP6Ax2KHE8AXgUuU9UzgH8Dzx5BmnHAP1X1VKALkHd1H1uk6fasUvKYBPQGPgP+g1PoKyntIBFZJCKLRo8efTinwBhjjDFHQ+Iicgv+Lndvg44gVw1VNa+ssRUoqXSe5D73tyIywN1XD9itqrnu9iagzF845X3U7VnAx6q6HyC4ndrVGjgJmOq2mMZRUFgrNY2I1ADSVPVjAFXNcmOA23R7mHm8AJipqgdE5CPgCRG5X1X9RROq6mggr4SnOeNHlP7MGdsgpaCCUlIaOvuCnzMjHUlJdX51xcVBcg3Ys/sws14y3Z6Or0HB9ScpDQhsTy82jT8j3YldPRndsye0wBdfA+df4dxftQwaBNUgpaTC9sKvn+3bnP2lpTka29OLxG6IFhe7QSPnr88993t3hx4bPPvfx/W7kbhe1wAQWPsDUq8xmpeHuo3QHVtLfnAYOTV4wddfQzQjvdg0ea9fkpNDv/bt+vPuvX8ENm/eTNOmTfO3mzRpwubNm0N7Uq//9xdeh/QPih/0/qd+Kuwo8tw7isQvLs3R8PB7p6Io8l1+CBGZBhTX/FFoxI6qqohoMekAmqnqZhFpCcwQkWXAUb3JynWN3mEQYHlQO/bJqtrnKNKE4mqgt4hsAL7HKXH3LPURh0lX/YQ0OQZS0yA+AV/P/gTmzyycZv5MpN/FAMjZfdDFC8IRGv/K/+Fr0tQZyRgfT0LPPuTOm1MoTc68OST0dTrhx5/dC/+S70IP/Ol4GDTAuc2dBucOcPafcCr8ngk7Mwqn35kB+/c5x8FJP3966PlYuQzSmuWfe+nZH4o7932dc8/ZfWHJt6HHzXtuj/73/invkP1wX7If7ktg4RTiejjjmaTV6bA/M+x98Uqiq5YjTZohqWkQH09cz34E5s8qlCYwfxZx/ZyuK76zzyWweGHoge368+69fwQmTpzIDTfcAECnTp3Ys2cPW7eG+CPE6//9Z++hd1zk3OZPRc69xNnfpl3J8X/f5xwH5NxL0Pmhd1338nsn7CJUo1cWVe2tqicVc/sU2BbUfawRUOyHqqpudv+uA2YBpwE7gNoikldJ1wQo8xdOeS/ozQEGiEhVtwbuwiLHVwEpInImOM20IlJ0HHqxaVQ1E9iUVyUqIlVEpNqRZE5EauLUOh6jqs1VtTlwF6U03x4Rvx//iGeJf2E08e9MJDBrCmxYi2/g3UgXp3NsYPJHSM3azjD4y2/EP3pYWELj95M1/AWqvfgqyWMnkDNzGoEN66hy823Ed3FGduVM/hSpVYvkcR9T5YpryRr1Wnhi51kwG7ZshPemwoN/heFBI89Gf1Jwf/jT8NDfnHRbfoUFcw55qiMW8KOv/A0Z+iYyZhI6cwpsWIMMvAfcc8/nE6BmbWd6i8tvREe/HHrcPF7+712BxTPQbb+Q+NpcEm4fSs6bj+cfS3zhy/z78dcNpsqo76BKVaqM+o74K8IwAs/vJ3fE30l4YSSJ73yKf9ZX6Ia1xA+8E1+XHk6SyR8jNWuTOG4S8ZdfT+7o4aHHDVZZr79y8N4fP34833zzDa1bt2bjxo3cfPPN3Hbbbdx2220ATJ48mXXr1rFmzRreeOMN7rzzzrDG9/R/D7BwFmzZiIyZgfzpWfTVIfmHZGRBw5a+OgR54O/ImBnO4J3vZoceuxx89sS4iUDenEU3Ap8WTSAidUSkinu/PtAV+J+qKjATuKy0xx/yfM7jyi8RGYzzYtJxBjssxmmKnaSqE0SkHfAKUAunKXq4qr4hImMOI00rYBRQH8gBLgcCwAqcAmKef6vqK25+bgLaq+rdInIjcJ6qXhWU37ruY5uo6sFSXprm9ChhbqQoSJi1nL1nt/csfs3Zi6Bna2+Cz1hF4JwTvIkN+GauwOv/fdZlTTyLnzRhE1k9TvEu/qwfvbv2oFxcf16/98M4OcERU1VvP3v6HOdNbMD31RqvP3ui+o/fMvfYiBRwGnVbe9SvQ0TqAR8AxwC/AFeo6k4RaQ/crqq3iEgXnLJJAKdCbriqvuU+viXwPlAXWAJcV0ZZo9z30UNVn6XIAIsix5cCh0wepKo3HUaa1RTfzFq1lHhjgDHu/XcoGD2Td3wnkHLIA40xxhhTqanqDuCQeXhUdRFwi3t/PnByCY9fBxzRfEblvqBnjDHGGHPEDqM/XWVgBT1jjDHGxB6xIg6U/8EYxhhjjDHmKFlx1xhjjDGxx5puAavRM8YYY4yJWVajZ4wxxpjYYzV6gBX0jDHGGBOLfFbQA2u6NcYYY4yJWVajZ4wxxpjYY023gNXoGWOMMcbErHK/1m0MsxNvjDGmMonuWreLukRmrdv2871brPkoWNOth1Z36+tZ7FZzv2RU+2c8i3/boic9W1w7YdZyrxf2ZsaZAz2L3/Obtz1//XPOvMGz+N2/Gev56/c6fkb3Mz2LnzLnG+jZ2rP4zFiFiDff06rK8q4XehIboO28z8i6Is2z+EkfbI5uQGu6Bazp1hhjjDEmZlmNnjHGGGNij9XoAVajZ4wxxhgTs6xGzxhjjDGxxyZMBqxGzxhjjDEmZlmNnjHGGGNij1gRB6ygZ4wxxphYZIMxAGu6NcYYY4yJWVajZ4wxxpjYYzV6gBX0yrVqndqTct/t4Itj76Qv2PXeB4WOJ516Ein33k6VY1uy9am/s2/W3LDGb3rmsXR5qC/i87HykyUsfWdeoeMnXHoGbS9vj/qVnAPZzHl2ErvXbw9bfOnYjbi7H4O4OAKff0Rg/JuFEyQkEPfn55DWbdE9u/E/8yBs/S1m4tftfBKt7r8GifOxZeIcfnl3cqHjTa/qQ+OLuqP+ANm7M1n57L/J2rojbPG9fv11Op/Msfdfh8T52DpxNhvfnVToeNpV/Ui96GzU7ydndyY/P/smB2Po9XsZP6FjZ5LvvR/xxXHg84kcGPfuIbFrDH6ShOPbENi7h71P/YXA1q1hiZ3v7sHQ6WzIyoKhj8Hq/x2aplVbePQ5qJIEC2bDa8+GJfRbb73FBRdcQHp6OieffHKxaUaMGEH//v3Zv38/N910E0uWLAlLbIDkTqeTev+t4POx+7OpbH9vQqHj1U5tS+p9t5J0bHM2DRnK3lnzwxY7T/zAZ/Cd1hMOHiDn9T+h6386NM1VjxLX/TJIrsXBG44Pex5MeFjTbRiJSDsR6R+WJ/P5SHngLjY/9Bd+ue5WavQ+h8TmxxRKkrstg21/f4nMaTPDEjKY+ISuj57H5HvH88Hlr3Nc37bUblG/UJo1U5Yx4apRfHTtaH4YO58uf+oTvgz4fMTdN5jcR28n98aL8PXsD82OLZyk/6Xovr3kXnsegQljiRv0QAzFF1o/eD0/PDCMBVcPpsG5najWvHGhJJk//8p3A59h4fVPkjFjEcfedUUY43v/+o978AZ+euBFFl39GCnndj7k9e/7+ReWDBzC4uv/wvYZ39HirqvCGN/r1+9hfJ+PGn96kD0PP8DOG64mqde5xDVrXihJ0vkXopmZ7Lzmcg588D7Vb78rPLHzdOoOac3h+j7w8hNw/1PFp/vTU/DSE066tObQsXtYwo8ZM4Z+/fqVePy8886jVatWtGrVikGDBjFy5MiwxAXA56PRg7fzy4NPsfbau6jVuztVmjctlCRnWwabnx3Onqmzwxc3OAun9URSW5B9bzdyRj9Kwi3PFZvO//1UDj5+fkTyEBYSF5lbBWMFvfBqB4SloJd0QmtyNv1G7m9bITeXzGmzqN6t8PqUuVu3kb12PQQC4QhZSIO2aezduIvMzbsJ5AZY89Vymp9deH3KnN+z8+/HV01Aw7h8tLQ5Gd28EbZsgtwcAjMm4+t6TuE0XXuiUz4FQGd/hZzROWbi1zyxJfs3pZP1Wwaa6yd92kJSup9WKM3uxSsJHHT+B3uWr6VKgzphi+/1669x4rEcCHr9GdO+pV730wul2bN4Rf7r3xtjr9/L+PEnnIh/8yYCW36D3Fyypk8jsVvhAlSVbmeRNcWpYT44eyaJp7cPS+x8XXrB1E+c+yt+gOSaUDelcJq6KVAt2TkOTvquvcIS/uuvv2bnzp0lHr/44osZO3YsAAsWLKB27dqkpqaGJXbVE1qRvWkLOb9tQ3Nz2TN9DjXO6lQoTc7WdA6u3YCG80M3iK99X/xznFpEXb0YqteC2g0OSaerF8Pu9IjkISx8cZG5VTBW0CtCRJqLyEoRGSMiP4vIOBHpLSLzRGS1iHQUkeoi8m8RWSgiS0TkYhFJBJ4BrhSRpSJyZSj5iE+pR256Rv52bsZ24lPql/KI8KrWoAb7tu3J3/49fS/VG9Q4JF3by9tz1Sd30/me3sx7cUr4MpDSEDK25G9qxjZnXxBJaYBmuM1Ffj/sy4RatWMifpWUOhxML/iiOZi+kyopJRdkGl/YnZ3fLAtLbKCcvP6CZtiD6TtJLOX1p17YnV3f/BiW2IDnr9/L+L76KfjTC768AxnpxKWkHJImkL4tP7b+vg+pVSvk2PnqN4T0oKbgjK3OvqJpMspIEyFpaWls3Lgxf3vTpk2kpaWF5bkTUuqRk17QBSYnfQfxKfXC8tyHS+qmotsLugHoji1I3fAUZE30WUGveMcBLwFt3Ns1QDfgIeBxYDAwQ1U7AucALwAJwJPAf1W1nar+14uMR9vyDxfx/oDXWPDqdE7/41leZ6dSatj3TGq0ac4v477wOiueaNC3CzXatGDjuMllJzbGVB4SH5lbBWMFveKtV9VlqhoAlgPT1akjXwY0B/oAj4nIUmAWkAQcU/xTFRCRQSKySEQWjR49utS0uRk7iG9Q8Cs6PqU+uRnhG+hQlv3pmSQ3LPiFXr1BTX5Pzywx/ZqvfqJ5j9YlHj9iGdsgpVH+pqQ0dPYF0Yx0JMX9lRkXB8k1YM/umIh/MGMXVRrUzd+u0qAuBzN2HZKuTocTaX7TBfz4yAg0JzcssYFy8voLajGqNKhLdjGvv3aHthxz00Usf2RYTL1+L+MHtmcQ16Cgmc6X0gB/RsYhaXwNGubHlurJ6J49hOTia2D0J85tZwY0CKpBSkmF7YVfP9u3OftLSxMhmzdvpmnTgn5zTZo0YfPmzWF57pyMHSQ0KGi9SWhQj9yM8A0yKklc3xtJHPoViUO/QndvQ+oX9ImVeo3QnWEebGOixgp6xTsYdD8QtB3AGakswKVuzV07VT1GVVeU9aSqOlpV26tq+0GDBpWaNmvlKhKbphHfqCHEx1Ojdw9+n/ftUb6cI5f+v83UalqXGo1r44v3cVyftvwy5+dCaWo2LSiINOt2PHt/LblPy5HSVT8hTY6B1DSIT8DXsz+B+YUHnej8mUi/iwGQs/ugixfETPzMFeup1rQBSY3qI/FxNOjdke1fFx7Vl3z8MbR55EZ+fPgVcnaVXAg/Gt6//nVUbdow//Wn9O7MjiKvv/rxzWj1yE389PCwmHv9XsbPXbmCuCZN8TVqBPHxJPXqTfa8rwulOThvLkn9nO7IVc4+h+zF34ce+NPxMGiAc5s7Dc4d4Ow/4VT4PdMp/AXbmQH79znHwUk/f3ro+TgMEydO5IYbbgCgU6dO7Nmzh61hGnV8YOVqEps0JqFRQyQ+nlq9upM5d2FYnrs0/i/fIfuRPmQ/0ofAwi+d0bSAtDod9u8t333xSmKDMQCbXuVofQncIyL3qKqKyGmqugTIBA7tyHY0/AHSX/4naS//HXw+9n7+Fdnrf6HuH2/g4Mqf+X3et1RpczyN/v4kcTVqUL1rZ+r+8QZ+vb70AuThUr8y94Uv6P/qtUicsGriUnaty6D9bT3IWPEbv8z5mZOu6EBaxxYEcgMczMxi5lOfhiU2AH4//hHPEv/CaPD5CHzxMWxYi2/g3eiq5ej8mQQmf0Tc4/8gftwX6N49+J95KGbiqz/Azy+No93wBxGfj98mfc3v63+jxa0DyFyxge1zl3Lc3VcQV60KJz17JwBZ23aw7JFXwpMBz89/gDUvjeWk4Y8gPmHrpDnsX7+ZZrf+gcwV69k5dwkt776KuGpJnPjs3QAc3LaD5Y8MD1N8r1+/h/H9fvYNf4laLw5HfD6yJk/Cv2E91W6+ldxVK8ieN5eszz+j5uAh1B3/IYHMvex96onwxM6zYLYztcp7UyHrAAx9vODY6E+cwiDA8KcLpldZOAcWzAlL+PHjx9OjRw/q16/Pxo0bGTJkCAkJCQCMGjWKyZMn079/f9asWcP+/fsZOHBgWOIC4A+wZdi/aPby00icj12TpnFw/a+k3HItWStXkzl3IUltWnHMc48TVyOZGl07kHLLtay9LnwjnwNLpuM7vSeJr8yD7APkvF4wojtx6FdkP+LMsBB/7WDiul0CiVWpMnIR/hnjyf3w5bDlI2QVsFAWCRKpUTsVlYg0Byap6knu9hh3e0LeMaADMBzoglMrul5VLxCRujiFwATguTL66enqbn0j9TLK1Grul4xq/4xn8W9b9CQ5Pdp6Ejth1nLPYufFn3FmGL8YjlDPb972/PXPOfMGz+J3/2as56/f6/gZ3c8sO2GEpMz5BnqGsZvHkZqxChHxJLSqsrzrhZ7EBmg77zOyrgjPoJGjkfTB5qie+C2rro9IAadR63e9uYCOktXoFaGqG4CTgrZvKuHYbcU8didOIdAYY4wxXrIaPcD66BljjDHGxCyr0TPGGGNM7KmAkxtHgtXoGWOMMcbEKKvRM8YYY0zsqYCTG0eCnQVjjDHGxB4bjAFY060xxhhjTMyyGj1jjDHGxB6r0QOsRs8YY4wxJmZZjZ4xxhhjYo/V6AFW0DPGGGNMLLJ59ABb69ZLduKNMcZUJtFd6/bXByOz1u0xL9lat+bwZN5cz7PYNf69g6GnvexZ/EeWPODdwuYzVhE45wRvYgO+mSv4vNMdnsU/f8FIcvu39Cx+/OR1XHPCZM/ij1/R37trD8rF9bfurHM9i9/y66kE+hznWXzfV2tY3vVCT2K3nfcZIt6VEVSV7Q9V9Sx+/RcPRDegzaMH2GAMY4wxxpiYZcVdY4wxxsQeG4wBWEHPGGOMMbHICnqANd0aY4wxxkSFiNQVkakistr9W6eYNOeIyNKgW5aIDHCPjRGR9UHH2pUV0wp6xhhjjIk9EheZW2geA6araitgurtdiKrOVNV2qtoO6AnsB74KSvJw3nFVXVpWQCvoGWOMMcZEx8XAO+79d4ABZaS/DPhCVfcfbUAr6BljjDEm9kh8RG4iMkhEFgXdBh1Brhqq6hb3/lagYRnprwL+U2TfsyLyo4gME5EqZQW0wRjGGGOMMYdJVUcDo0s6LiLTgNRiDg0u8jwqIiVO6iwijYCTgS+Ddv8Zp4CY6ObhUeCZ0vJrBT1jjDHGxBzxaMJkVe1d0jER2SYijVR1i1uQSy/lqa4APlbVnKDnzqsNPCgibwMPlZUfK+hVAFWueY74k3uj2QfIeutuAr/+eEiaxD8MJqHLlUi1Wuy7s1lY4rbo0pxeD/dAfD5+/GQZC97+rtDxdpedwmlXtCMQCJCzP4cv/zaVHet2hiV2vrsHQ6ezISsLhj4Gq/93aJpWbeHR56BKEiyYDa89G57YHbohdz8OcT708wnwnzcLH09IQP78PBx/IuzdjT79AGz7LTyxgZTOJ3LiA1cgPmHjxHmsHftVoeMtru5F04u7orl+snfv48e/vcuBreE9/77bnkQ69ICDWfhffhjWLj800XEnEffAC5BYBf1uFoFRpf64PGyndKvPDY+fiM8nzJywkc/eXFfoePcBaVzzcBt2bjsIwFfjNzBrwqawxM5XSa+/qh3bU+++OxGfj72TvmDPuP8WOp506snUu/cOElu2JP3pZ/l91tdhiRtM7nwCOvSAgwfQFx+FNcVce63aIg8NhcQk+G4W+vpfwxI7udPppN5/K/h87P5sKtvfm1DoeLVT25J6360kHducTUOGsnfW/LDEzfPWW29xwQUXkJ6ezsknn1xsmhEjRtC/f3/279/PTTfdxJIlS8Kah+oXv0TiCX3R7P1k/ncQ/s1LCydIqErNG8bhq9cSAn6y/zeZ/ZOfCGseQuVVQa8ME4EbgX+4fz8tJe3VODV4+YIKiYLTv++nsgKW2kdPRGqLyJ1lPUkJj71fRKodzWNLec6Tg4YU7wwaYjztCJ9njIjsF5EaQfuGi4iKSP1w5jlUcSf3xtewJb//uQNZ7zxA0g0vFpsud+mX7P9r+JY1Ep/Q+7GefHj3x7x16RhO6NeGei3rFkrzvy9W8vYVY3nnqvdY+M53nPNAj7DFB6BTd0hrDtf3gZefgPufKj7dn56Cl55w0qU1h47dQ4/t8yH3PYE+Ngi96UKk1/nQ7NjCafpfBpl70Ov6oR+ORW4r84fVEcQX2j58FQvvf43ZVz1D4z4dSG5RuCVg788bmXvjc3x93bNsmbGENndfEr74gLTvAWnN8d/SE/8rjxN3d/FfonF3/RX/iD/jv6UnpDVH2p8demwfDHyiLUMHfcfDF86hy/mNSTs2+ZB0336xhcf/MJfH/zA3/IW8ynr9+XzUf+Aetj70OBuvv4Xk3ueQ0PyYQklyt6WT8fcX2DdtRnhiFtXhbEhrjg7shQ7/C3Lv08Umk3ueQYcNRgf2cs59h/Cc+0YP3s4vDz7F2mvvolbv7lRp3rRQkpxtGWx+djh7ps4OPV4xxowZQ79+/Uo8ft5559GqVStatWrFoEGDGDlyZFjjJ7TpS1zKsez6x0nsm3A3yZe+Umy6A7OGs3toO3YP60xC8zNJaNMnrPmIUf8AzhWR1UBvdxsRaS8i+b/mRKQ50BQoepGNE5FlwDKgPvC3sgKWNRijNnBUBT3gfiCsBT1VXRY05HgiBUOMS6wmLcUanNEviIgPZwjz5rBlNkziTzuPnPnOr+nAukVItVpIrUP7bgbWLUL3bAtb3EYnpbJ74272bN5DIDfAii9XclyPwl802b9n599PqJoAhHn96C69YOonzv0VP0ByTaibUjhN3RSoluwcByd9116hx25zCvz2K2zZBLk56IzJ0LVnoSTStSf6pftjbPaXcHrn0OO6ap/YnP2bMjjw23Y0189vUxfRsPuphdLs+P5nAgedGv3dP60jqcEh0zGFRDr3Rqd/7GysWgrVa0KdIue/jnv+Vy0FQKd/jHQO/QfHcafUZtuv+0nfdAB/jvLN5C2c0bOsPsthVkmvvyontCZn82/kbtkKubn8Pn0W1bt1KZQmd+s2steuB43ImvFIl97oVPfaW7nUufaKO/fVk53jgE79GOkS+rVX9YRWZG/aQs5v29DcXPZMn0ONszoVSpOzNZ2DazegEXr9X3/9NTt3llw7f/HFFzN27FgAFixYQO3atUlNLa5L2NFJbHsBWYvGA5D760IkqRZSo8jz5xwgZ+0c574/h9zNS/HVSgtbHsJBJD4it1Co6g5V7aWqrVS1t6rudPcvUtVbgtJtUNU0VQ0UeXxPVT1ZVU9S1etUdV9ZMcsq6P0DONatNXtBRB4Wke/c0R5PA4hIdRH5XER+EJGfRORKEbkXaAzMFJGZbrp9IvKsm+5bEWno7m8uIjPc55wuIse4+8eIyCsiMl9E1onIZSVlUkSuFpFlbvzng/bvc0elLHefO/iT4n3gSvd+D2AekBv02E9E5Hv3sYPcfXFuvn5y4/3J3X+viPzPfQ3vl3XSj4SvTiN0Z0H5M7DzN6ROo3CGKFZyg2Qyt2Xmb2du20eNlBqHpDvtilO5deLNnH1fd6YPnRneTNRvCOlbC7Yztjr7iqbJKCPNUcVuUCT2NqS42Olud4mAH/ZlQs3aoccGkhrU5sC2XfnbWem7SEop+bmbXtSVjG+KadoKRf1UNGNL/qZu3wr1Uw9Ns31r6WmOQp0GSezYmpW/vXPbAeo2PHRwWYc+qfzjk27cN/w06qYmhRy3kEp6/cWn1Cc3PSN/OzdjO3H1o9zQUa8hBF17bN/q7DskzdbS0xyFhJR65KRvz9/OSd9BfEq9kJ83nNLS0ti4cWP+9qZNm0hLC18hK65WYwK7C2rIA3s2E1ercYnpJakWiSf2J2d1mL8DTFiUVdB7DFjr1qBNBVoBHYF2wBki0h3oB/ymqqeq6knAFFV9BfgNOEdVz3GfqzrwraqeCswBbnX3vwq8o6qnAOOA4DriRkA34ALc6s2iRKQx8DxOjVw7oEPeDNJuzEWq2han+nNI0EN/BlLcWamvxin4BbtZVc8A2gP3ikg99/nT3JL0ycDbQefpNPc13F5cPmPVkg9+4I2L/s3sEV9z5i2dyn6ACbu0fh2pdUIz1r031eusRNXiWenc12sWjw2Yy7L527njuVO8zpIxlY8vjhrXvcOBua8T2LnB69wUER+hW8VyJDnu497yenwm4xT8vgZecmvSJqlqSb1ys4FJ7v3vgbw69jOBP7j33wWGBj3mE7fa8n95NYDF6ADMUtUMABEZB3QHPgECQF4v4veA/yvy2P/DmaOmE3BbkWP3ikhep6em7mtdBbQUkVeBzymYqfpHnHbzT9y4xXJrBgcBjBo1iqtLSJfQ848kdL8eAP/6JUjdgl9qvrqN0V1bSnhk+OxL30eNhgU1eDUaJpOZkVli+hVfrqTP470oPAr8KFx8DZx/hXN/1TJoEFQ7lJIK24s0T2/f5uwvLc3R2J5eJHZDtLjYDRo5f31xkFwD9u4OPTaQlb6bqg0LmmKTGtQhK+PQ567XoQ3H3dSPb+4YRiAn95DjR0ouuB5fX6eiW1f/iKQ0ym+Ql/qpTq1JsO1bnf2lpTkKu9KzqBdUQ1e3YdX8QRd59u3OH4jGzAkbueahNiHHtevPqcGLb1DQ+BGfUh//9u2lPCJMLrwO6R907lOCWi7qp8KOIq9/R5FzX1yao5CTsYOEBgU1mAkN6pGbsSPk5w2nzZs307RpQb/BJk2asHlzaD2PkrrcRlKngQDkbvweX+0m+cd8tdLw7yl+oE/yZf/En7GWrK9fCyl+JJTTwRhRdyQTJgvwXNCyG8ep6luq+jNwOk7HwL+JyJMlPD5HCzo0+Dm8QmbwJ7scQV5LUrRDxX+BvwJTg9vBRaQHTifJM90ayCVAkqruAk4FZuHU3OV1nDwf+CfOefhOSri6VHW0qrZX1faDBpU8v2LOjLfY/1QP9j/Vg9wlk0no4nzx+lq2R/fvDWtfvJJsWb6VOsfUplbjmvjifZzQtw1rZhUe9VjnmNr59489qyW7Nu4iZJ+Oh0EDnNvcaXDuAGf/CafC75mwM6Nw+p0ZsH+fcxyc9POnh56PlcsgrRmkpkF8AtKzP8wv3Cyh82cifS92Ns7uC0u+DT2ua8+KX6jetAFVG9VD4uNofG57ts0pPNq65vFNOPmxa/ju4ZFk7yq5EH4kdNK7+O+5AP89F6DfTEV6ub91Wrdzzv+uIud/l3v+W7cDQHpdgn57RGOjirV22R5Sm1UnJa0qcQnCmf0b8f3Mwtd97ZSCptwzejZk87oyu6qUza4/Dq5cRUKTNOIbpUJ8PNV79eD3ud+E5blL9dl76B0XObf5U5Fz3WuvTbuSz/3v+5zjgJx7CTo/9GvvwMrVJDZpTEKjhkh8PLV6dSdz7sKQnzecJk6cyA033ABAp06d2LNnD1u3hvYDK2v+KHYP68zuYZ05uPwzktpfA0D8MR3RrL1o5qHPX63fECSpFr9PDONANBN2ZRW2MoG8ap0vgb+KyDhV3SciaUCO+xw7VfU9EdkN3FLksWX9FJyPU6v2LnAtTg3hkVgIvOKOlt2F0wz7qnvMh7N8yPvANcDc4Aeq6i8iMhgo+ulQC9ilqvtFpA3QGcCNka2qH4nIKuA9dyBHU1WdKSJz3deSDOw+wtdRLP+PUwmcci7V/7HImV7l3/fkH6v21Cz2P9UDgCqXDyG+02WQWI3qLy4j5+t3yf50aAnPWjb1K9Oen8nlr1+K+IRln/7EjnU76HZHF7b+bytrZq/jtCvb0bzTMfhzAxzce5DPnwixNq+oBbOdqS3emwpZB2Do4wXHRn/ifBkDDH+6YHqLhXNgwZzQYwf86Ct/Q4a+CT4f+sX/wYY1yMB70FU/OV+6n0+Ax59H3psCe/egf30w9Lgu9Qf46cX36fjKPYjPx6bP5rNv/RaOH3QBu1f8SvrXP3LCPZcSX60Kp//d6QWRtXUXix4O3+g7/W4m0qEHcW/NdKZXGfZI/rG4Vyfhv+cCAPyvP0ncn4ZClSR00Wx00ayQYwf8ypi/LeexNzvi88Gs/9vE5jX7uOyeVqz7aQ+LZ6bT97rmnNGzAf5cZd+eHEb9+dBph0JSWa8/f4Dtw14j9aXnEJ+PzM+/JGfDL9T5440cXPkz++d9Q5U2x9Pw2afw1UimWpfO1Ln5BjbdcGvZz324Fs6Cjj2QMTMKpldxyciJ6B0XAaCvDkEezpteZbZzC5U/wJZh/6LZy08jcT52TZrGwfW/knLLtWStXE3m3IUktWnFMc89TlyNZGp07UDKLdey9rq7Qo/tGj9+PD169KB+/fps3LiRIUOGkJCQADitQZMnT6Z///6sWbOG/fv3M3DgwLDFBshZMYXENn2p89hyNGc/+/5b0OBV+0/fsntYZ3y10qjW+zFyt62k9v3OD4ED8/7FwYVjwpqXUFiNnkPKGjUkIuOBU4AvgE0UFOT2AdcBxwEv4DST5gB3qOoiEbkHuBun/945IrJPVZPd57wMuEBVbxKRZjh93eoDGcBAVf1VRMbgNAVPcB+T/3h3O/+4iFwNPI5T6/e5qj6a9xicmaP74ExKeKWqZhR97qDn3IDTJy8Tpwm2OU5zbW3gKZyC5NsU1IT+GaeQOBOncCjAe6pabH/CIjTzZu86+Nb49w6GnvayZ/EfWfIA9GztTfAZqwicc4I3sQHfzBV83ukOz+Kfv2Akuf1behY/fvI6rjlhsmfxx6/o7921B+Xi+lt3VvimYjpSLb+eSqDPcZ7F9321huVdL/Qkdtt5n+FMf+YNVWX7Q1U9i1//xQNRffHbt78ZkWHR9evf4t0/8SiUWdxV1WuK7BpRZHstxXTMUtVXKahZI7iQ5hawJrj3f8EZSFH08TcV2U4u6biq/odD14LLO/ZAWc8dtL950OZ5xaXBaZ4tqlsJaY0xxhjjAavRc9hZMMYYY0zMsYKe40gGY1Q4RWsBjTHGGGMqEyvuGmOMMSYGWREHYrxGzxhjjDGmMrPirjHGGGNijvXRc1iNnjHGGGNMjLLirjHGGGNijtXoOewsGGOMMSbmWEHPYU23xhhjjDExyoq7xhhjjIk5VqPnKHOtWxMxduKNMcZUJlFdI3bv3ikR+Z6tWbNfbK11ayLn2y5XeRa78/z3ybqsiWfxkyZsQnu38iS2TFtNTo+2nsQGSJi13PNz7/Wi8nsG1fEsfq3Ruzy79qB8XH/+fsd6Fj9uylrPX3/WFWmexE76YDPbH6rqSWyA+i8eQMS7Mkr0K5asiAN2FowxxhgTg6zp1mGDMYwxxhhjYpQVd40xxhgTc6xGz2E1esYYY4wxMcqKu8YYY4yJOVaj57CzYIwxxpiYYwU9hzXdGmOMMcbEKCvuGmOMMSYGWREHrEbPGGOMMSZmWXHXGGOMMTHH+ug5rEavHKvV6VRO/c/LtPtgOI2vv+iQ4zXateHkt5+j05xx1D2nU0TyEH/zMyS+OpfEl6YiLU4qPs3Vj1DlXwup8u6q8GfgrifgnWkw+jM47sTi07RqC29MctLd9UTYQkvHbsSPnUT8uC/wXXPLoQkSEoh78kXix31B3Ov/gdTGYYudx+vzL3c+gbw9HfnXJDiuhGWrWrVFRn3upLszfOcfIOnKf5D8t+9JfnIuvmNOKTZNlQF/ocY/fqLmKxvDGhuo9Nef3PEkvn/PwDfy85L//8edhG/kZHz/noHc8WR44paD1x4/8BkSX5lL4gulvPeuepQqr39HlbE/hz1+9Ytfos5jP1H7gYXEpbU7NEFCVWr+8f+o/chSaj/0PdX6/zUscd966y22bdvGsmXLSkwzYsQIVq9ezQ8//MBpp50WlriRIBIfkVtFExMFPRGZX8bxm0VkmYj8KCI/icjF7v5nRKR3BPKzQUTqh/QkPqHFQzez8sF/8MM1D1Kvd1eqNi+8PmP21h2s/dtItk+dF1KoErNwWk+kUQuy7+lGzr8eJWHQc8Wm8y+axsHHLgh/BjqeDWnN4MbeMOwJuO+Z4tPd9zS8/BcnXVoz6NA99Ng+H3H3DSb30dvJvfEifD37Q7PC64P6+l+K7ttL7rXnEZgwlrhBD4QeN/j5vT7/Hc6GtObowF7o8L8g9z5dbDK55xl02GB0YC9Iax6e8w/En3QuvobHsu8vZ3Dg3fupeu1LxabL/WEK+57rFZaYhVTy648OPZDGzQnc3JPAiMH47i7+9fvueYbAiMcJ3NwTadwc2p8dWtxy8Np9p/VEUluQfW83ckY/SsItJbz3vp/KwcfPD2tsgIQ2fYlLOZZd/ziJfRPuJvnSV4pNd2DWcHYPbcfuYZ1JaH4mCW36hBx7zJgx9OvXr8Tj5513Hq1ataJVq1YMGjSIkSNHhhzTRFZMFPRUtUtJx0SkCTAY6KaqpwCdgR/dxz2pqtOik8sjk3zicWRt2srB39LRXD87ps2nzlntC6U5uDWD/Wt/hUBkFor2deiDf9YEAHT1YqhWE2o3OCSdrl4Mu9PDn4EuvWHqJ879FUshuQbUTSmcpm4KVEt2joOTvuu5IYeWNiejmzfClk2Qm0NgxmR8Xc8pnKZrT3TKpwDo7K+QMzqHHDeY1+dfuvRGp37sbKxcCtVrFn/+qyc7xwGd+jHSJfTzDxDfrj8537wPgH/9IqRqLaRWw0PS+dcvQvdsC0vMQir59Sdn9kanB/3/k0v4/1cL+v9PD/3/Xx5eu699X/xzgt571WtF9b2X2PYCshaNByD314VIUi2kRmrhRDkHyFk7x7nvzyF381J8tdII1ddff83OnTtLPH7xxRczduxYABYsWEDt2rVJTU0tMb2XrEbPERMFPRHZ5/5tJCJzRGSpW3N3FtAAyAT2AajqPlVd76YfIyKXuff7i8hKEfleRF4RkUnu/qdE5N8iMktE1onIvUFxP3HTLxeRQeF8TYkpdcnetiN/OztjJ4kpdcMZokxSLxXd8Vv+tu7cgtSL4hu6fkPI2FKwnbHV2Vc0zfatpac5GimFY2vGNmdfEElpgGa4sf1+2JcJtWqHHjvv+b0+//WKnP/tW519h6TZWnqao+Sr3YjArs3527rrN3y1G4XluQ9Lpb/+GqIZBdcfGVuh6PVXL7XQ69eMLUio///y8NrrpqLbg957O7YgdaP33our1ZjA7k3524E9m4mrVXLztCTVIvHE/uSsnhnxvKWlpbFxY0E3iU2bNpGWFnoB00ROxSualu4a4EtVfVZE4oBqwH5gG7BeRKYD/6eqnwU/SESSgFFAd1VdLyL/KfK8bYBzgBrAKhEZqao5wM2qulNEqgLfichHqrqDEriFwUEAo0aNovgeR8YYY8xh8sVR47p3ODD3dQI7N3idm3KlIta+RUKsnYXvgH+LSALwiaouBRCRfkAHoBcwTETOUNWngh7XBliXV9MH/Ae3QOb6XFUPAgdFJB1oCGwC7hWRS9w0TYFWQIkFPVUdDYzO2/x2zIwSX0h2xk4SG9bL305MqUt2RsnV6eES1+9G4npdA0Bg7Q9IvcbkNQxL3Ubojq0lPzgcLroW+l/p3P/5R0gJqsFJSYXtRZrotm+D+qmlpzkaGdsKxZaUhs6+IJqRjqSkOjUOcXFO096e3SGF9fz8X3gd0v8K5/6qZYXPf/1U2FHk3O7Y5pzz0tIcgcQet5B41g0A+DcsxlcnDb97TOo0JrB7S8kPDodKfv3Jhdch/ZzXrz8vQ1Iao3zvHExJhaLX346thV6/pDRCQ/j/A9699/reSFyvawEIrF2K1G+MuuObpF4jdGdk33tJXW4jqdNAAHI3fo+vdpP8Y75aafj3/Fbs45Iv+yf+jLVkff1aRPOXZ/PmzTRt2jR/u0mTJmzevLmURxivxUTTbR5VnQN0BzYDY0TkBne/qupCVX0OuAq49Aif+mDQfT8QLyI9gN7Amap6KrAESArtFRTYt2ItSU1SqdIoBYmPo17vLuya+324nr5E/invkP1wX7If7ktg4RTielwGgLQ6HfZnRqYvXrCJ4+D2i5zbvGlw7gBn/wnt4PdM2JlROP3ODNi/zzkOTvr5oXe71FU/IU2OgdQ0iE/A17M/gfmFm0V0/kyk38UAyNl90MULQo7r+fn/7D30jouc2/ypyLnu75g27Uo+/7/vc44Dcu4laAjnP3vWm+z7a3f2/bU7OUsnk3DmVQDEtWiPHtgbmb54wSr59aefvUfgrgsJ3HUh+s1XSK/D+P/vD/r/97oE/Sa01+/Ze+/Ld8h+pA/Zj/QhsPBL4roHv/f2Rvy9lzV/FLuHdWb3sM4cXP4ZSe2dH3zxx3REs/aimYcWNKv1G4Ik1eL3iQ9FNG/BJk6cyA03OD/GOnXqxJ49e9i6NcI/QI9afIRuFUvFy3EpRKQZsElV3xCRKsDpIjINSFXVxW6ydsAvRR66CmgpIs1VdQNw5WGEqwXsUtX9ItIGZ5BH+PgDbHj5bdoMexyJ85E+aSYH1m+iyS2X8/vKdeya+z3VT2jJ8c89SHyN6tTudjpN/ngZP173cNiyEFg8A9/pPUl8bS4czCLn9YKRbYkvfEn2w30BiL9uMHFnDYAqVaky6jv80/9D7gcvh56BBbOckY9jp8PBA/DCYwXH/jXR+TIGeOUpePh5qJIEC2c7t1D5/fhHPEv8C6PB5yPwxcewYS2+gXejq5aj82cSmPwRcY//g/hxX6B79+B/Jrwftp6f/4WzoGMPZMwMOHgAffHR/EMyciJ6h3P+9dUhyMNDITEJvpvt3MIgd9lXxJ90LsnPLobsAxwYc1f+seQn5rDvr87o1qRLnyah46WQWI0az/9E9tx3OfjZ86FnoJJffyychXboge/fM+BgFoGXC/7/vn9+RuCuCwEIvDYE34NDIbEKumg2fDcrtLjl4LUHlkx33nuvzIPsA4Xfe0O/IvsRZ3Rr/LWDiet2CSRWpcrIRfhnjCf3w9DfezkrppDYpi91HluO5uxn339vyz9W+0/fsntYZ3y10qjW+zFyt62k9v3fAHBg3r84uHBMSLHHjx9Pjx49qF+/Phs3bmTIkCEkJCQATpejyZMn079/f9asWcP+/fsZOHBgSPEiyZpuHaIamRGb0SQi+1Q1WURuBB4GcnAGX9wABIC3gcZAFpAB3K6qa0VkDDBJVSeIyIXAC8DvOE3ANVT1WhF5Ctinqi+6sX4CLgC2AJ8AzXEKirWBp1R1lohsANqr6vZSsq3fdrkqbOfgSHWe/z5ZlzUpO2GEJE3YhPZu5UlsmbaanB4lzAkWBQmzlnt+7gN9jvMsvu+rNewZVMez+LVG7/Ls2oPycf35+x1bdsIIiZuy1vPXn3WFN4MHkj7YzPaHqnoSG6D+iwcQEc/iq2pUg+fkbI1IASchIdW7k3gUYqK4q6rJ7t93gHeKSdKzhMfdFLQ5U1XbiPMu+CewyE3zVJHHBM+ceV4Jz9v8MLNujDHGmAiwGj1HTPXRC9GtIrIUWI7TLDvK2+wYY4wxxoTGirsuVR0GDPM6H8YYY4wJndXoOewsGGOMMSbmWEHPYU23xhhjjDExyoq7xhhjjIlBVsQBq9EzxhhjjIlZVtAzxhhjTMwRiY/ILbQ8yeUislxEAiLSvpR0/URklYisEZHHgva3EJEF7v7/ikhiWTGtoGeMMcaYmFMeC3rAT8AfgDkl51vicObzPQ84EbhaRE50Dz8PDFPV44BdwB/LCmgFPWOMMcaYKFDVFaq6qoxkHYE1qrpOVbOB94GL3QUdegIT3HTvAAMOJ6jdKuANGGTxK19si2/xLX7ljV+ZX3t5ugGDcFbPyrsd8XkBZuEslVrcscuAN4O2rwdeA+q7BcC8/U2Bn8qKZTV6Fdcgi18pY1t8i2/xK2/8yvzayw1VHa2q7YNuo4OPi8g0EfmpmNvFXuTXxh4bY4wxxoSJqvYO8Sk249TW5Wni7tsB1BaReFXNDdpfKqvRM8YYY4wpP74DWrkjbBOBq4CJ6rTXzsRp2gW4Efi0rCezgl7FNbrsJBY/BmNbfItv8Stv/Mr82mOCiFwiIpuAM4HPReRLd39jEZkM4NbW3Q18CawAPlDV5e5TPAo8ICJrgHrAW2XGdDv0GWOMMcaYGGM1esYYY4wxMcoKesYYY4wxMcoKeqZCEpEEr/NgjDHGlHc2vUoFISK1gKeAs9xds4FnVHWPZ5mKsqBZwa8BLgAaRjjeH0o7rqr/F8n4bh6OUdVfIx3HmIpERBJUNcfrfBhTEdhgjApCRD7CWSPvHXfX9cCpqlpqYSSM8V8pZvceYJGqljm8O8TYnXEKdwOAusBdOEPNd0U47tulHFZVvTmS8d08LFbV0yMd52iISAdV/S5KsYq7zvcAy1Q1PRp5cPNRHWedyqtU9fwIx6pb2nFV3RnJ+EW5tegnAZujec6D4hf6oaeqEf2hFxT3clX9sKx9UcqLD0hW1b1RiFWurj9z9KygV0GIyFJVbVfWvgjGHw20AfI+3C4F1uMM716nqvdHIObfgcuBX4H/AB/jFCxbhDtWeSUiS1T1NK/zkcddWPtq97ZbVdtHKe7nONMRzHR39QC+B1rg1Gy/G8HYicD5OAWMvsBHwP+p6meRiunGXQ8oIMUcVlVtGeH4/wJeVdXlbovCN4Af58fWQ6r6n0jGD8qHJz/0guIf8mMrmj/ARGQ8cDvOuf8OqAmMUNUXIhzX0+vPhI813VYcB0Skm6rOBRCRrsCBKMY/Beiqqn43/kjga6AbsCxCMW8BfgZGAp+p6kERidovExF5oLTjqvpyFLKRVkJtal4e7o10BkSkOQWFuxygGc4ajRsiHTtIPHCCqm5z89QQGAt0AuYAYS/oiUgfnNfcB6eAORbooKoDwx2rOOXgB81Zqnq7e38g8LOqDhCRVOALnB9fEVPMD72ncX7ovVPqA8MX/zygP4e+B2sCudHIg+tEVd0rItfinPfHcH7kRLSgVw6uPxMmVtCrOO4A3nF/WQuwE7gpivHrAMk4zWUA1YG6quoXkYMRitkIOBfny3a4iMwEqgYt/xJpNYLu3waMikLMog7gfKh7QkS+wfliex+4VFVXi8j6KBfyAJrmFfJc6e6+nSISqb5aU3B/zKjqegARGRGhWIcQkVJrjFR1cYSzkB10/1zc2nxV3eq0okacpz/0gN9wFqy/iMLvwUzgT1HMR4LbbD4AeE1Vc6J8HhCROkArIClvn6rOiWYezNGzgl4FoapLgVNFpKa7HfE+GkUMBZaKyCycgmZ34O9un6VpEYqZpqpTgCkiUgVnAEZVYLOITFfVayIUFwBVfTrvvogMCN6Ooh3RqsEowTYgDWfgSwqwGqc5J9pmicgkCncdmOVef7sjFPN0nKWHponIOpzCblyEYhXnpaD7Z1C4sKE4/dUiabeIXICzlmZX4I8AIhKP8z6MNE9/6KnqD8APbtNpPHCMqq6KdNxijAI2AD8Ac0SkGRC1z38RuQW4D2dd1aVAZ5xm/EhffyZMrI9eOSciN5R2XFXHRjEvjYCO7uZ3qvpbhOMV2w/GLewOiPJr92RQhIh8q6qdox23SB5q4QxAuBrnV31toK+qLoxiHgSncNfV3TUP+Eij9AEmIl1wXv+lOF+4H6tq1JaD8qKvpogcD7yCU+Aapqpj3P19gT6q+mCE4+ePOA/6oXc1zswDEf+hF5SPC4EXgURVbSEi7XD6hV4Ujfgl5ClarRqIyDKgA/CtqrYTkTbA36M1ENCEzgp65ZyIvFrCoYtwaryiVisrImk4/bPyY0ay+r48DUTwsKDXEPgzcBxOX8jnolmbKyJ/CJ5GRkQaAFfgfOEeo6pNo5WXaBORr1S1T5F9PqA3zqjbiI+6DopbbkdfR0p5+aEnIt/j1F7Nyvs8EpFlqnpylOLfB7yN02T8JnAa8JiqfhWl+N+pagcRWQp0cpvQl6tq22jEN6Gzgl4F4tZqXIuzqPH/gGdV9ccoxX4euBJYDgTc3RrJX7Uiko7TXFasSA9EcH/J5o06OxZYk3fICa+nRDK+m4cpOE12c3BqNGqo6k2RjhsUv8QChog0U9VfopSPPwDPAw1wzn/e/6BmBGPaDw1nQMJjQN6X+nLgeVWdHIXY5eL859WqB+dHRH6MxvvfjfWDqp7q1qTeBjwBvBvFUb8f4wzGuR+nwLsLSFDV/tGIb0JnffQqALdPzE3AQ8C3wGUe9BUZALRW1UgNvCiOpwMRcApWXmukqoPd+1+KSKQ74B+2aBXyXEOBC1V1RRRj1pJSJs3WCE+Y7dbm5/3QaFJ09HUUfujcilOweARnUAJAe+AfItIkCk3Xno84dy0XkWuAOBFpBdwLzI9SbCiY3qQ/TgFvuURpNAyAql7i3n3K7SdZC2egkqkgrKBXzonIXTgdYacD/TwY7ZhnHZAARLOg5+lABFX9RUQG4DabquqXXuTDHfGW98EeF7ytkZ+0tI2IFFdrHLVaTde2KBfywPlCu4AS5hEDIr0yyqIS7kfLn3BGHAdfYzPcWr65QKQLel7/0MtzDzAY57NvPPAl8Ncoxv9eRL7CmTPyzyJSg4JWlagQkTicAVnr3V2pONPemArAmm7LOREJ4EwlkUHh0Y5R/aIVZ2WOU3EKnPmFvUj+qvZ6IIKIvI7TZDUf6IUzxUM0P+ARkQ04H+peTZq7HKcmoVhRbLodgfPl8gmFr7+IFbbKQ784EUnB6Re7RlV3Rzn2ClU94UiPhTG+5+ffzUd7nIJecwoqR6L52esD2uFMTL9bROrh9M+OVrede4AhOCPwg7vtROtHngmR1eiVf+Vl0sqJ7i2aXsu7IyJdVXVe0Pbdqvpa8Q8Lm+44y8z5RaQazpxqUS3oqWrzaMYrRnaUm2hLUhPYjzN5cZ5I16pFrXms2ODOtBZ/B9YCLURkkKpG8z24V0ROdacZCc7XqTgDAyItu+wkUTEOp9vMT0S5Jg1AVQPirFJxvIgklfmA8LsPp9vODg9imzCwgl45l/clKyLPq+qjwcfcARKPFvvA8OfDiybUB4D33Puv4sxrludmggqCEZKt7kogqro/mv1iinLnDexV1r4ImFd2ksjTKK1GUcSNeXdEpEpw/1QR6ayq30Y4/v1AW1XNEJGWOAWOaBb0HgQmirPmc14Tanuc83JdFOJ7/UMvT4ZGeLm70pSDeew2UjBRvqmArKBXcZzLoYW684rZF1Yi8oGqXhE0ArWQCFffSwn3i9uOhOD+aQIc625Hc9RtEs4qJPWL9NWriTORcaTlz9UlIvep6oig7TGRHgEsIo+o6tCggQmFRLhD/hgKflx8Q+EfGq8X2Y6EbFXNAFDVde5cclGjqnNFpCPO2rI3ubv/B3RW1a1RyILXP/TyDBGRNzm020qk+2jmuY+CeezOyZvHLkqxwemfPUuc9aaDX380loA0YWAFvXJORO4A7gRaFukUX4PojPy6z/3rxQhULeF+cduRENE+SIfpNpyancZA8IjbvUTni6570P0bgeAlwKLRRydvAIYXgxG8/qFRdKRtoe0ojLpNAeqp6pNF9p8oIv68Qmgks1DC/eK2I2kg0AZnMFp+HzUiPxgnT5aqZolIXs3yShFpHaXY4Ay6+BVIdG+mgrGCXvk3Hmch6+dw5rPKkxmFEZeo6hb3rxf9tNoE1aAdW6R2LaKDEKB8NJu7NWgjROQeVS1p8uxIKu3LNuLymsw86jrg9Q+Nh4tsR3sE6qs4NZdF1QP+AkR6ZQqvz3+eDqoazYJVUZtEpDbOQKSpIrILiNrnsXqz9KMJIxt1W4GISDeglaq+LSL1cSbPXV/W40KMmUkpH6oRnrC2WWnHozji85DRf9GcMNWNVx1nuotjVHWQO59Xa1WdFOG4PwA9AB8ww72fV+CbqaqnRjj+Z5R+/UVjwm7BmSw8b/JuAa5Q1YaRil0eiMgiVW1fwrGfVPWkCMffjzNJeXETlrdU1eqRjB+Uj7eBF1T1f9GIV0Zezsadx05VozJYpYT34B6cWvZRqpoVjXyYo2c1ehWEiAzB6QjdGmc5nESc/itdS3tcqFS1hhv/r8AW4F2cD9prcdbAjGRsT2vUykGzebB/49TodHG3NwMfAhEt6OF8qXxPQeEu2hM2v+j+/QPO9Cp5fbauxpnuIZKCa9SKNh1HrSlZRKYCl+dNr+L21XxfVftGOHSNUo4lRDg2lI+uE+AMfljqjnw9SJT66IpI3WJ2L3P/JgMRb9FxrQNSgP+421fijLo+HngDuD5K+TBHyWr0Kghx1hk8DVisHi7DU9a+CMX2pEZNRGoBdfCo2bxIXhapanspvAxTVM5/eVBc7VJpNU6xRESWqmq7Ivvyr4MIxv0c+KcWWe7MnTD5XlU9L5Lxg+K1ALbk1RyJSFWgoUZp8viSWhYi3aLgzqG6iYIBUcFdJyI+h2ZQPr5T1Q7F7RNb87ZC8HmdAXPYstUplSvkN+VF0+8icq2IxImIT0SuBX6PZEARucMd7dtaRH4Muq2n4JdtxKjqHlXdoKpXA02Bnu6Hu8/98ommbPcLLu//fyxRXKVERKYfzr4Iqu5OMZIXuwXOaOSIE5Gpbh+pvO06IhLNVVL8InJMUPxmRKeP2v3AcBEZIyL3uLd3cAbk3Ff6Q8PqQwrPX+d390WFqv5S3C0KoV/BWVd2Cs5AqJaq2sK9RaWQ50oucv0dg1OjCOVnrkNTCmu6rTg+EJFRQG1x1qD8I/BmFONfg/MBPwLnS2Yeke+M7elAlDxeNZsXMQTnA7+piIxzY98U6aDlYHqXPH/CmeJhnZuHZjgjkqMhRYNWpVDVXSLSIEqxwVmVYa6IzMZ57WcBgyIdVFVXi8jJOO/zvP54s4HbotwvKz64P5qqZotIzI/+VNX7RURw+sVeD7wqzlJoIyPdN7uIB3Guv7U4118L4E63ssGzJSrN4bOm2wpERM6lYGWAL1V1mpf5iSYvBqIExV6Kh83mQfmoh9NfSHDm1NoehZj3UTC9y29Bh/YCb0Rx0lrceeTauJsrgycwjnDc74FLVPVXd7sZ8HHR7gQRzkN9nP89ROl/HxS7BbBVVQ+420lAahSbTqcCr6q7KoiIXIzTdBzpycLLDbdG+SqclXkeV9U3ohw/+L23ygZgVCxW0CvnRGSuqnYLGv0a3E8jgNMh9wVVLW4ahHDmIwmnFrEtkL8Mj6reHMm4buz8GjVVPV5EGgMfqmpUatREZKGqdszrK+j+kv0mGgU9EWnjzptVXKFCgZ3RaEbycHqX4DycBJxI4etvbBTi9gNG49Rm5deoqWpUmm/dWp1rcZrunnGbzlJVdWGU4i8CuuTVqrm1afOK9tuKYPxjcVYFScO55jcBN6jqmlIfWMG5nzMX4wx+SMGZt++DvB8cUYjfU1VniMgfijuu0Zsw2oTICnoVnFvLMz/S8zyJyIfASpxmnGdwvnhWqGrE++p4XaMmIg8BrXBWJ3kOp8A7XlVfKfWB4Yk92p1OZSaH9ssSnDnNflDViI58E4+mdwmKPwSnCetEYDLOqjBzVfWyKMX3skZtJM6Pup6qeoLbhP5VFAtaxQ0GifpAIBFJBlDVfdGM6xUR+R1YjTOtz2qKvP8jXdASkadVdYg408sUpdH4kW/Cwwp6MUBEGqk7sXEEYyxR1dPyClgikgB8raqdy3xw6LE9q1ELyoOnzebuQIw7gW44H/hf4/TVyRKRr1S1T6lPEHr8/+JMs3KDqp4kItVwfmC0i2TcoPjLgFOBJap6qog0BN5T1XOjENvrGrW8696TEddeN526/+u/A41V9TwRORE4U1XfikZ8r4jIGEoedGMFLXPYbDBGDIh0Ic+V4/7d7TahbQWi1SHdk4EopTSb3y7O1AdRaTZ3vYPTLy6vFvEaYCzOxL0RLeS5jlXVK0XkagBV3e8WgKLlgKoGRCRXRGoC6TgjoaPhddwaNZza7EzgI5z1R6MhR0TiKBhxnULhUaiRdjswTkT+SVDTaRTjj8EZBDXY3f4Z+C8Q0wU9jfA60mURkQuBH7VgPtMngUtxVuW4L8oDQkwIrKBnDtdot8noCWAizvD6J0t/SHio6otujdpenEk6/xKNGjVV7eb+LXbi2Lxmc4pfJircTlLVE4O2Z4pINGfq93R6F2CR2yH9DZyaxX3AN1GK3SmvRg3yR91Gc9TnK8DHQAMReRa4DGcJsqhQ1bVAZw+bTuur6gci8mc3fq6I+KOcB8+4nzNDKKjNnws8o6o7Ihz6WdzuCiJyAXAdzkTlpwH/AiI9YbcJEyvomcOiqnk1aLOJwjqzUO5q1A6hqjtEpEeUwi0Wkc6q+i2AiHQiiqsz4NH0LnlU9U737r9EZApQU1V/LO0xYeRpjZqqjnNH/vbCeQ8MUNUV0YpfDppOf3cLO3nnvzPOElyVxfvAHJzaNHC6EfwX6B3huKqq+937fwDeUtXvge9F5M5SHmfKGeujZw5LOfiwLy5PURmI4iW3b5riLDnVGvjV3W6GM8XIiaU8PNx5ifr0LkGxPesnJ87k4FcCp+M0oV+GU6sc0Ul7pfglsPJplOaSFJEvcJtO3f6R8Th9JU+OUvzTgVdx5vL7CWcE6mVRLOh7SopZV1hElkX6/Iuz7GMXYD+wHrhUVRe5x/4Xzc8eExor6JnD4vWHfSn5ivhAFC9JCcsv5Yn01CrlaHoXr0eetqGgRm16NGrUxFkBJq8mO/iDOm+t1agugVVkMMghI3EjnId4nB86gjOPW04ZD4kZIvIysBD4wN11GdBRVR+KcNybgcdxusykq2o/d/9pwIvRGoxjQmcFPXNYysOHvYm+cjS9S9RHnpajGjUfTm1mi6DazEaquiBK8WfhNBtOdf8HnYHnVfXsCMctdv62PJGeXqS8cLuuVKegu4CPguUnVVVrRjB2Gs6gux9UNeDuSwUSNUrz+ZnQWR89c7gqez+ZSklV85ba6k8p07tEISte9JP7nlJq1IhSX1Xgn3g76vcBnAFYx4rIPNym0yjEvbCUY4ozgXDMK2kwWJRibxaR5kBVnO+A63C6MIzwKk/myFmNnjkslb2fTGUnIh/gNOGMc3ddA9RS1SuiFN+TfnJubK9r1DydR8+NV2mbTssDEbkI6O5uztIoTVTuxv4RZw7LU3CmunkTZ1qniNbomvCxGj1zWFR1sYicTdCHPXARYAW9ysHT6V2KG3lK9GqUva5R82TUbylNp8eLSNSaTsvjQLBoEpF/4FxreT+y7hORrqr65yhlIVdVVZyJsl9T1bdE5I9Rim3CwAp65rCpai6wPG9bRIbhfOGZ2Of19C6o6kqcZfhw8/ArcEwUQlfWefTKS9PpGCrhhMlB+gPtgvrIvQMsAaJV0Mt05zC8Duju1nAnRCm2CQMr6JlQRHNlBOOBItO7zHcLV/nTu3iZN6J3/VXKefRUdWCkYxymSj1hsqs2zryhALWiHPtKnK4af1TVrW7XhReinAcTAivomVBYB8/Yd4HXGShFtK4/T1emgENrM6OpHDSdVsqBYFKwhvVzwBJ35Lvg9NV7LFr5UNWtwMtB27/iLL9oKggbjGFKFVSjc8gh4HhVrRLlLJlKRERepeTr78ZITi1RJB9Rn0evvPB6Ds3KOhCsyOCbRhT0CV3oFr4iHb/oykT5h4jwtC4mvKygZ0rl9YS9pnITkRtLO66q70QrL5VVeZhDszKO+hWRdUCJkyJXlnkETeis6daU5Q23+cAYL7RW1ce9zkQl50nTqYj0VNUZxYz+jeqoXw/Vwuk6UVxf1Eozj6AJnRX0TFlSvM6AqdT64SzDZLzj1YTJ3YEZOKN/i5uwOtYLOr+o6s1eZ8JUfFbQM2WpVdpSRJXgV7XxVpy7rm2xI2yjtQxZZVbcHJpRajrNFJEHcPrl5a1QApVnEJjNamDCwgp6pizWfGC81AZnKbKSrr9oLUNW6ZSDptNk929rnIEIn+JcBxcCCyMcuzyI6PrRpvKwwRimVHnLL3mdD1M5BQ8AMNElIk+p6lMi8jbFj7qMSrOiiMwBzlfVTHe7BvC5qnYv/ZGxwS1oPw80wDn3NurVHBGr0TNlseYDYyqn8tJ02hDIDtrOdvdVFkOBCyvTlD4mvKygZ8pyo4gMAI4Dlqnqlx7nx1QuI9yVKJoBa1R1t8f5qUzKS9PpWGChiHzsbg/AWRatsthmhTwTCmu6NaUSkdeBtsB8nAljP1PVv3qbK1NZiMgtOKsyrAVaAINUdaK3uapcykPTqTtp8lnu5hxVXRKt2F4TkRFAKvAJcDBvvw2EM4fLCnqmVCLyE3CqqvpFpBrwtaqe4XW+TOXgXn/nqGqGiLQExqnqmV7nqzIRkVXAKap60N2uAvyoqq29zVnl4PaRLCpqfSRNxWdNt6Ys2arqB1DV/SJiffZMNGWragaAqq5zCxkmuip706mnVHWg13kwFZvV6JlSich+YE3eJnCsu5038usUr/JmYp+IpAPvB+26KnhbVe+NeqYqocrcdOo1EWmCs9ZvV3fX18B9qrrJu1yZisQKeqZUttat8ZKtdWsqOxGZCowH3nV3XQdcq6rnepcrU5FYQc8cFhF5XlUfLWufMcaY8BGRpararqx9xpTE53UGTIVR3K/H86KeC1MpichUEakdtF1HRGyqH1MZ7BCR60Qkzr1dB+zwOlOm4rCCnimViNwhIsuA1iLyY9BtPbDM6/yZSiMleA49Vd2Fs1KAMbHuZuAKYCuwBbgMsAEa5rBZ060plYjUAuoAzwGPBR3KtAXlTbSIyPfAJar6q7vdDPjYluczxpjSWUHPHDYR6Qa0UtW3RaQ+UENV13udLxP7RKQfMBqYjTPi+yycyZOt+dbENHdlmFuB5gRNiWbz6JnDZQU9c1hEZAjQHmitqseLSGPgQ1XtWsZDjQkL98dFZ3fzW1Xd7mV+jIkGEZmPM6XK94A/b7+qfuRZpkyFYhMmm8N1CXAasBhAVX9zl0IyJuLcibr7AS1V9RkROUZEOqpqNNdcNcYL1Wx2AxMKG4xhDle2OtW/CiAi1T3Oj6lcXgfOBK52tzOBf3qXHWOiZpKI9Pc6E6bisoKeOVwfiMgooLaI3ApMB970OE+m8uikqncBWZA/6jbR2ywZExX34RT2DojIXhHJFJG9XmfKVBzWdGsOi6q+KCLnAnuB44G/qOo0j7NlKo8cEYmjoEY5BQh4myVjIk9VrYuMCYnV6JlSichc928m8BFwu3v7WET2iMh6EbnTyzyaSuEV4GOggYg8C8wF/u5tloyJPBGZfjj7jCmJjbo1IRGResB8VW3tdV5MbBORNkAvnOlVpqvqCo+zZEzEiEgSUA2YCfTAue4BagJTVLWNR1kzFYwV9EzIRKSRqm7xOh8m9ohI3dKO26TdJlaJyH3A/UBjYDMFBb29wBuq+ppHWTMVjBX0jDHllrvUnuJ8yQV/WAmgqtrSk4wZEyUico+qvup1PkzFZYMxjDHllqq2ABARH3At0CJvHj2gkaeZMyY6topIDVXNFJG/AKcDf1PVxV5nzFQMNhjDGFMR/BNnVYzgefSs6cpUBk+4hbxuQG/gLWCkx3kyFYgV9IwxFYHNo2cqq7xlz84HRqvq59i1b46AFfSMMRWBzaNnKqvN7mT1VwKTRaQK9t1tjoBdLMaYisDm0TOV1RXAl0BfVd0N1AUe9jRHpkKxUbfGmArB5tEzlZXbP6+Vqr7t1mYnq+p6r/NlKgYr6BljjDHllIgMAdoDrVX1eBFpDHyoql09zpqpIKzp1hhjjCm/LgEuAn4HUNXfAFv/1hw2K+gZY4wx5Ve2Ok1veQORqnucH1PBWEHPGGOMKb8+cEfd1haRW4FpwBse58lUILYyhjHGGFN+pQATcNa4bQ08iTNxsjGHxQZjGGOMMeWUiCxW1dOL7PtRVU/xKk+mYrEaPWOMMaacEZE7gDuBliLyY9ChGsA8b3JlKiKr0TPGGGPKGRGpBdQBngMeCzqUqao7vcmVqYisoGeMMcYYE6Ns1K0xxhhjTIyygp4xxhhjTIyygp4xxhhjTIyygp4xxhhjTIz6f4Bv3C+t9YfxAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 720x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# show the correlation between features\n",
    "plt.figure(figsize=(10,8))\n",
    "#sns.heatmap(df_all.corr(), annot=True, linewidth=.5)\n",
    "sns.heatmap(df_all.corr(), annot=True, linewidth=.5, vmin = -1, vmax = 1,cmap = plt.cm.CMRmap_r, fmt=\".1f\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 6765 entries, (0, 0) to (1418, 0)\n",
      "Data columns (total 14 columns):\n",
      " #   Column          Non-Null Count  Dtype  \n",
      "---  ------          --------------  -----  \n",
      " 0   jetNo           6765 non-null   int32  \n",
      " 1   jetPTLeading    6765 non-null   float64\n",
      " 2   jetETA          6765 non-null   float64\n",
      " 3   jetPHI          6765 non-null   float64\n",
      " 4   bjetPT          6765 non-null   float64\n",
      " 5   bjetETA         6765 non-null   float64\n",
      " 6   elecPTLeading   6765 non-null   float64\n",
      " 7   elecETA         6765 non-null   float64\n",
      " 8   elecPHI         6765 non-null   float64\n",
      " 9   dielecCOS       6765 non-null   float64\n",
      " 10  dielecETA       6765 non-null   float64\n",
      " 11  met             6765 non-null   float64\n",
      " 12  testnonTopMass  6765 non-null   float64\n",
      " 13  isSignal        6765 non-null   float64\n",
      "dtypes: float64(13), int32(1)\n",
      "memory usage: 776.9 KB\n"
     ]
    }
   ],
   "source": [
    "df_all.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x1512 with 15 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Function to plot distributions of the float features\n",
    "import math \n",
    "def hist(df1, df2, cols, bins, ncols = 3):\n",
    "    nrows = math.ceil(len(cols) / ncols)\n",
    "    fig, ax = plt.subplots(nrows, ncols, figsize = (5 * ncols, 4.2 * nrows), sharey = False)\n",
    "    for i in range(len(cols)):\n",
    "        #sns.histplot(data = df, x = cols[i], bins = bins, hue = 'isSignal', palette = ['grey', 'red'], ax = ax[i // ncols, i % ncols], stat=\"probability\")\n",
    "        sns.histplot(data = df1, x = cols[i], bins = bins, color = 'red',\n",
    "                      ax = ax[i // ncols, i % ncols], stat=\"density\", label='bkg')\n",
    "        sns.histplot(data = df2, x = cols[i], bins = bins, color = 'grey',\n",
    "                      ax = ax[i // ncols, i % ncols], stat=\"density\", label='sig')       \n",
    "        ax[i // ncols, i % ncols].set_xlabel(cols[i])\n",
    "        ax[i // ncols, i % ncols].legend()\n",
    "        if i % ncols != 0:\n",
    "            ax[i // ncols, i % ncols].set_ylabel(\" \")\n",
    "        \n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "\n",
    "# Distributions of the float features\n",
    "hist(df_all[df_all['isSignal']==0].replace(-999, np.nan),\n",
    "     df_all[df_all['isSignal']==1].replace(-999, np.nan),\n",
    "     #list(df_all.columns[df_all.dtypes == 'float64']),\n",
    "     list(df_all.columns),\n",
    "     bins = math.floor(len(df_all)**(1/3)),\n",
    "     ncols = 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to make ML output and significant plot\n",
    "samples =['ttbar', 'WZ', 'ZZ', 'signal_charm']\n",
    "\n",
    "def plot_SoverB(data, SoverB_hist_dict):\n",
    "    \n",
    "    signal = r'signal_charm' # which sample is the signal\n",
    "\n",
    "    # *******************\n",
    "    # general definitions (shouldn't need to change)\n",
    "\n",
    "    for x_variable,hist in SoverB_hist_dict.items(): # access the dictionary of histograms defined in the cell above\n",
    "\n",
    "        h_bin_width = hist['bin_width'] # get the bin width defined in the cell above\n",
    "        h_num_bins = hist['num_bins'] # get the number of bins defined in the cell above\n",
    "        h_xrange_min = hist['xrange_min'] # get the x-range minimum defined in the cell above\n",
    "        h_xlabel = hist['xlabel'] # get the x-axis label defined in the cell above\n",
    "    \n",
    "        bin_edges = [ h_xrange_min + x*h_bin_width for x in range(h_num_bins+1) ] # bin limits\n",
    "        bin_centres = [ h_xrange_min+h_bin_width/2 + x*h_bin_width for x in range(h_num_bins) ] # bin centres\n",
    "        \n",
    "        signal_x = data[signal][x_variable] # histogram the signal\n",
    "    \n",
    "        mc_x = [] # define list to hold the Monte Carlo histogram entries\n",
    "\n",
    "        for s in samples: # loop over samples\n",
    "            if s not in ['data', signal]: # if not data nor signal\n",
    "                mc_x = [*mc_x, *data[s][x_variable] ] # append to the list of Monte Carlo histogram entries\n",
    "                #test=[]; test.append(df[s]['NN_output_prob'].values)\n",
    "\n",
    "\n",
    "    \n",
    "    \n",
    "        # *************\n",
    "        # Signal and background distributions\n",
    "        # *************\n",
    "        distributions_axes = plt.gca() # get current axes\n",
    " \n",
    "        mc_heights = distributions_axes.hist(mc_x, bins=bin_edges, color='red', \n",
    "                                             label='Total background',\n",
    "                                             histtype='step', # lineplot that's unfilled\n",
    "                                             density=True ) # normalize to form probability density\n",
    "        signal_heights = distributions_axes.hist(signal_x, bins=bin_edges, color='blue',\n",
    "                                                 label=signal, \n",
    "                                                 histtype='step', # lineplot that's unfilled\n",
    "                                                 density=True, # normalize to form probability density\n",
    "                                                 linestyle='--' ) # dashed line\n",
    "        \n",
    "        distributions_axes.set_xlim( left=bin_edges[0], right=bin_edges[-1] ) # x-limits of the distributions axes\n",
    "        distributions_axes.set_ylabel('Arbitrary units' ) # y-axis label for distributions axes\n",
    "        y_max = max(max(signal_heights[0]), max(mc_heights[0]) )\n",
    "        #distributions_axes.set_ylim( top=max(signal_heights[0])*1.3 ) # set y-axis limits\n",
    "        distributions_axes.set_ylim( top=y_max*1.2 ) # set y-axis limits\n",
    "        #distributions_axes.set_ylim( top=max(signal_heights[0])*1.3 ) # set y-axis limits\n",
    "        plt.title('Signal and background '+x_variable+' distributions') # add title\n",
    "        distributions_axes.legend() # draw the legend\n",
    "        distributions_axes.set_xlabel( h_xlabel ) # x-axis label\n",
    "            \n",
    "        plt.show() # show the Signal and background distributions\n",
    "    \n",
    "    \n",
    "        # *************\n",
    "        # Signal to background ratio\n",
    "        # *************\n",
    "        plt.figure() # start new figure\n",
    "        SoverB = [] # list to hold S/B values\n",
    "        for cut_value in bin_edges: # loop over bins\n",
    "            #signal_weights_passing_cut = sum(data[signal][data[signal][x_variable]>cut_value].totalWeight)\n",
    "            signal_weights_passing_cut = len(data[signal][data[signal][x_variable]>cut_value])\n",
    "            background_weights_passing_cut = 0 # start counter for background weights passing cut\n",
    "            for s in samples: # loop over samples\n",
    "                if s not in ['data', signal]: # if not data nor signal\n",
    "                    #background_weights_passing_cut += sum(data[s][data[s][x_variable]>cut_value].totalWeight)\n",
    "                    background_weights_passing_cut += len(data[s][data[s][x_variable]>cut_value])\n",
    "            if background_weights_passing_cut!=0: # some background passes cut\n",
    "                SoverB_value = signal_weights_passing_cut/background_weights_passing_cut\n",
    "                SoverB_percent = 100*SoverB_value # multiply by 100 for percentage\n",
    "                SoverB.append(SoverB_percent) # append to list of S/B values\n",
    "        \n",
    "        SoverB_axes = plt.gca() # get current axes\n",
    "        SoverB_axes.plot( bin_edges[:len(SoverB)], SoverB ) # plot the data points\n",
    "        SoverB_axes.set_xlim( left=bin_edges[0], right=bin_edges[-1] ) # set the x-limit of the main axes\n",
    "        #SoverB_axes.set_ylim( 100, 1000)\n",
    "        SoverB_axes.set_ylabel( 'S/B (%)' ) # write y-axis label for main axes\n",
    "        plt.title('Signal to background ratio for different '+x_variable+' cut values', family='sans-serif')\n",
    "        SoverB_axes.set_xlabel( h_xlabel ) # x-axis label \n",
    "        \n",
    "        plt.show() # show S/B plot\n",
    "    \n",
    "    return"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define the models"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NN from Tensorflow\n",
    "We'll start with a dense (fully-connected) NN layer.\n",
    "Our model will have a single fully-connected hidden layer with the same number of neurons as input variables. \n",
    "The weights are initialized using a small Gaussian random number. \n",
    "We will switch between linear and tanh activation functions for the hidden layer.\n",
    "The output layer contains a single neuron in order to make predictions. \n",
    "It uses the sigmoid activation function in order to produce a probability output in the range of 0 to 1.\n",
    "\n",
    "We are using the `binary_crossentropy` loss function during training, a standard loss function for binary classification problems. \n",
    "We will optimize the model with the Adam algorithm for stochastic gradient descent and we will collect accuracy metrics while the model is trained."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dividing the data into testing and training dataset\n",
    "\n",
    "We will split the data into two parts (one for training+validation and one for testing). \n",
    "We will also apply \"standard scaling\" preprocessing: http://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html i.e. making the mean = 0 and the RMS = 1 for all input variables (based **only** on the training/validation dataset).\n",
    "We will also define our early stopping criteria to prevent over-fitting and we will save the model based on the best `val_loss`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input (InputLayer)          [(None, 13)]              0         \n",
      "                                                                 \n",
      " hidden1 (Dense)             (None, 20)                280       \n",
      "                                                                 \n",
      " hidden2 (Dense)             (None, 20)                420       \n",
      "                                                                 \n",
      " output (Dense)              (None, 1)                 21        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 721\n",
      "Trainable params: 721\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of batches seen.\n"
     ]
    }
   ],
   "source": [
    "NDIM = len(VARS)\n",
    "dataset = df_all.values\n",
    "X = dataset[:,0:NDIM]\n",
    "Y = dataset[:,NDIM]\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train_val, X_test, Y_train_val, Y_test = train_test_split(X, Y, test_size=0.2, random_state=7)\n",
    "\n",
    "# preprocessing: standard scalar\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler().fit(X_train_val)\n",
    "X_train_val = scaler.transform(X_train_val)\n",
    "X_test = scaler.transform(X_test)\n",
    "\n",
    "# baseline keras model\n",
    "from keras.models import Sequential, Model\n",
    "from tensorflow.keras.optimizers import SGD\n",
    "#from keras.optimizers import SGD\n",
    "from keras.layers import Input, Activation, Dense, Convolution2D, MaxPooling2D, Dropout, Flatten\n",
    "from keras.utils import np_utils\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.pipeline import Pipeline\n",
    "from tensorflow import keras\n",
    "\n",
    "inputs = Input(shape=(NDIM,), name = 'input') \n",
    "hidden1 = Dense(20, name = 'hidden1', kernel_initializer='normal', activation='relu')(inputs)\n",
    "hidden2 = Dense(20, name = 'hidden2', kernel_initializer='normal', activation='relu')(hidden1)\n",
    "outputs = Dense(1, name = 'output', kernel_initializer='normal', activation='sigmoid')(hidden2)\n",
    "\n",
    "# create the model\n",
    "model = Model(inputs=inputs, outputs=outputs)\n",
    "# compile the model\n",
    "optimizer = SGD(learning_rate=1e-3) # define the optimizer\n",
    "model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "#model.compile(optimizer=optimizer, loss='binary_crossentropy', metrics=['accuracy'])\n",
    "# print the model summary\n",
    "model.summary()\n",
    "\n",
    "# early stopping callback\n",
    "from keras.callbacks import EarlyStopping\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=10, mode='min')\n",
    "\n",
    "# model checkpoint callback\n",
    "# this saves our model architecture + parameters into dense_model.h5\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "model_checkpoint = ModelCheckpoint('TopFC_dense_model.h5', monitor='val_loss', \n",
    "                                   verbose=0, save_best_only=True, \n",
    "                                   save_weights_only=False, mode='auto', \n",
    "                                   period=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run training \n",
    "Here, we run the training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "11/11 [==============================] - 1s 31ms/step - loss: 0.6931 - accuracy: 0.5087 - val_loss: 0.6927 - val_accuracy: 0.5499\n",
      "Epoch 2/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6923 - accuracy: 0.5654 - val_loss: 0.6919 - val_accuracy: 0.5536\n",
      "Epoch 3/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6909 - accuracy: 0.5811 - val_loss: 0.6902 - val_accuracy: 0.5713\n",
      "Epoch 4/1000\n",
      "11/11 [==============================] - 0s 9ms/step - loss: 0.6881 - accuracy: 0.5987 - val_loss: 0.6867 - val_accuracy: 0.5898\n",
      "Epoch 5/1000\n",
      "11/11 [==============================] - 0s 9ms/step - loss: 0.6828 - accuracy: 0.6055 - val_loss: 0.6807 - val_accuracy: 0.6016\n",
      "Epoch 6/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6754 - accuracy: 0.6118 - val_loss: 0.6731 - val_accuracy: 0.6112\n",
      "Epoch 7/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6669 - accuracy: 0.6149 - val_loss: 0.6662 - val_accuracy: 0.6090\n",
      "Epoch 8/1000\n",
      "11/11 [==============================] - 0s 9ms/step - loss: 0.6594 - accuracy: 0.6188 - val_loss: 0.6611 - val_accuracy: 0.6149\n",
      "Epoch 9/1000\n",
      "11/11 [==============================] - 0s 9ms/step - loss: 0.6541 - accuracy: 0.6221 - val_loss: 0.6567 - val_accuracy: 0.6238\n",
      "Epoch 10/1000\n",
      "11/11 [==============================] - 0s 24ms/step - loss: 0.6502 - accuracy: 0.6288 - val_loss: 0.6540 - val_accuracy: 0.6260\n",
      "Epoch 11/1000\n",
      "11/11 [==============================] - 0s 10ms/step - loss: 0.6476 - accuracy: 0.6310 - val_loss: 0.6513 - val_accuracy: 0.6312\n",
      "Epoch 12/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6454 - accuracy: 0.6345 - val_loss: 0.6485 - val_accuracy: 0.6378\n",
      "Epoch 13/1000\n",
      "11/11 [==============================] - 0s 9ms/step - loss: 0.6440 - accuracy: 0.6329 - val_loss: 0.6468 - val_accuracy: 0.6386\n",
      "Epoch 14/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6423 - accuracy: 0.6378 - val_loss: 0.6462 - val_accuracy: 0.6378\n",
      "Epoch 15/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6408 - accuracy: 0.6410 - val_loss: 0.6440 - val_accuracy: 0.6393\n",
      "Epoch 16/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6393 - accuracy: 0.6395 - val_loss: 0.6426 - val_accuracy: 0.6401\n",
      "Epoch 17/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6380 - accuracy: 0.6402 - val_loss: 0.6410 - val_accuracy: 0.6423\n",
      "Epoch 18/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6366 - accuracy: 0.6430 - val_loss: 0.6392 - val_accuracy: 0.6482\n",
      "Epoch 19/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6353 - accuracy: 0.6482 - val_loss: 0.6381 - val_accuracy: 0.6482\n",
      "Epoch 20/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6338 - accuracy: 0.6493 - val_loss: 0.6366 - val_accuracy: 0.6445\n",
      "Epoch 21/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6327 - accuracy: 0.6482 - val_loss: 0.6356 - val_accuracy: 0.6445\n",
      "Epoch 22/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6311 - accuracy: 0.6504 - val_loss: 0.6342 - val_accuracy: 0.6452\n",
      "Epoch 23/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.6298 - accuracy: 0.6541 - val_loss: 0.6320 - val_accuracy: 0.6519\n",
      "Epoch 24/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6284 - accuracy: 0.6548 - val_loss: 0.6309 - val_accuracy: 0.6489\n",
      "Epoch 25/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6271 - accuracy: 0.6572 - val_loss: 0.6290 - val_accuracy: 0.6526\n",
      "Epoch 26/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6259 - accuracy: 0.6547 - val_loss: 0.6275 - val_accuracy: 0.6541\n",
      "Epoch 27/1000\n",
      "11/11 [==============================] - 0s 11ms/step - loss: 0.6244 - accuracy: 0.6565 - val_loss: 0.6267 - val_accuracy: 0.6504\n",
      "Epoch 28/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6230 - accuracy: 0.6613 - val_loss: 0.6250 - val_accuracy: 0.6556\n",
      "Epoch 29/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6219 - accuracy: 0.6624 - val_loss: 0.6239 - val_accuracy: 0.6563\n",
      "Epoch 30/1000\n",
      "11/11 [==============================] - 0s 21ms/step - loss: 0.6202 - accuracy: 0.6580 - val_loss: 0.6230 - val_accuracy: 0.6511\n",
      "Epoch 31/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6193 - accuracy: 0.6585 - val_loss: 0.6214 - val_accuracy: 0.6556\n",
      "Epoch 32/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6184 - accuracy: 0.6580 - val_loss: 0.6207 - val_accuracy: 0.6526\n",
      "Epoch 33/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6171 - accuracy: 0.6608 - val_loss: 0.6193 - val_accuracy: 0.6600\n",
      "Epoch 34/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6160 - accuracy: 0.6637 - val_loss: 0.6183 - val_accuracy: 0.6593\n",
      "Epoch 35/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6147 - accuracy: 0.6613 - val_loss: 0.6178 - val_accuracy: 0.6615\n",
      "Epoch 36/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6139 - accuracy: 0.6595 - val_loss: 0.6167 - val_accuracy: 0.6622\n",
      "Epoch 37/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6127 - accuracy: 0.6633 - val_loss: 0.6157 - val_accuracy: 0.6622\n",
      "Epoch 38/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.6119 - accuracy: 0.6620 - val_loss: 0.6150 - val_accuracy: 0.6630\n",
      "Epoch 39/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6108 - accuracy: 0.6635 - val_loss: 0.6140 - val_accuracy: 0.6644\n",
      "Epoch 40/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6099 - accuracy: 0.6661 - val_loss: 0.6138 - val_accuracy: 0.6644\n",
      "Epoch 41/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.6090 - accuracy: 0.6628 - val_loss: 0.6130 - val_accuracy: 0.6637\n",
      "Epoch 42/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.6081 - accuracy: 0.6615 - val_loss: 0.6124 - val_accuracy: 0.6637\n",
      "Epoch 43/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6078 - accuracy: 0.6641 - val_loss: 0.6114 - val_accuracy: 0.6696\n",
      "Epoch 44/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.6067 - accuracy: 0.6635 - val_loss: 0.6117 - val_accuracy: 0.6600\n",
      "Epoch 45/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.6058 - accuracy: 0.6665 - val_loss: 0.6117 - val_accuracy: 0.6630\n",
      "Epoch 46/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.6048 - accuracy: 0.6680 - val_loss: 0.6095 - val_accuracy: 0.6637\n",
      "Epoch 47/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.6036 - accuracy: 0.6659 - val_loss: 0.6098 - val_accuracy: 0.6667\n",
      "Epoch 48/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.6031 - accuracy: 0.6663 - val_loss: 0.6093 - val_accuracy: 0.6637\n",
      "Epoch 49/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.6021 - accuracy: 0.6669 - val_loss: 0.6086 - val_accuracy: 0.6593\n",
      "Epoch 50/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.6013 - accuracy: 0.6672 - val_loss: 0.6082 - val_accuracy: 0.6622\n",
      "Epoch 51/1000\n",
      "11/11 [==============================] - 0s 19ms/step - loss: 0.6010 - accuracy: 0.6676 - val_loss: 0.6072 - val_accuracy: 0.6615\n",
      "Epoch 52/1000\n",
      "11/11 [==============================] - 0s 16ms/step - loss: 0.6006 - accuracy: 0.6667 - val_loss: 0.6060 - val_accuracy: 0.6637\n",
      "Epoch 53/1000\n",
      "11/11 [==============================] - 0s 11ms/step - loss: 0.5997 - accuracy: 0.6689 - val_loss: 0.6057 - val_accuracy: 0.6622\n",
      "Epoch 54/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5986 - accuracy: 0.6680 - val_loss: 0.6056 - val_accuracy: 0.6637\n",
      "Epoch 55/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5976 - accuracy: 0.6705 - val_loss: 0.6045 - val_accuracy: 0.6622\n",
      "Epoch 56/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5974 - accuracy: 0.6687 - val_loss: 0.6057 - val_accuracy: 0.6585\n",
      "Epoch 57/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5960 - accuracy: 0.6704 - val_loss: 0.6037 - val_accuracy: 0.6615\n",
      "Epoch 58/1000\n",
      "11/11 [==============================] - 0s 19ms/step - loss: 0.5958 - accuracy: 0.6694 - val_loss: 0.6031 - val_accuracy: 0.6571\n",
      "Epoch 59/1000\n",
      "11/11 [==============================] - 0s 9ms/step - loss: 0.5949 - accuracy: 0.6698 - val_loss: 0.6039 - val_accuracy: 0.6585\n",
      "Epoch 60/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.5942 - accuracy: 0.6709 - val_loss: 0.6024 - val_accuracy: 0.6608\n",
      "Epoch 61/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5936 - accuracy: 0.6726 - val_loss: 0.6029 - val_accuracy: 0.6704\n",
      "Epoch 62/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5935 - accuracy: 0.6733 - val_loss: 0.6013 - val_accuracy: 0.6659\n",
      "Epoch 63/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5924 - accuracy: 0.6735 - val_loss: 0.6020 - val_accuracy: 0.6615\n",
      "Epoch 64/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5919 - accuracy: 0.6735 - val_loss: 0.6014 - val_accuracy: 0.6652\n",
      "Epoch 65/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5912 - accuracy: 0.6781 - val_loss: 0.6003 - val_accuracy: 0.6630\n",
      "Epoch 66/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5913 - accuracy: 0.6754 - val_loss: 0.6022 - val_accuracy: 0.6644\n",
      "Epoch 67/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5903 - accuracy: 0.6766 - val_loss: 0.6004 - val_accuracy: 0.6667\n",
      "Epoch 68/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5895 - accuracy: 0.6754 - val_loss: 0.6007 - val_accuracy: 0.6644\n",
      "Epoch 69/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.5887 - accuracy: 0.6774 - val_loss: 0.6003 - val_accuracy: 0.6571\n",
      "Epoch 70/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5883 - accuracy: 0.6789 - val_loss: 0.6007 - val_accuracy: 0.6593\n",
      "Epoch 71/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5874 - accuracy: 0.6778 - val_loss: 0.6000 - val_accuracy: 0.6644\n",
      "Epoch 72/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5869 - accuracy: 0.6766 - val_loss: 0.5996 - val_accuracy: 0.6630\n",
      "Epoch 73/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5869 - accuracy: 0.6805 - val_loss: 0.5993 - val_accuracy: 0.6630\n",
      "Epoch 74/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5858 - accuracy: 0.6787 - val_loss: 0.5990 - val_accuracy: 0.6674\n",
      "Epoch 75/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5855 - accuracy: 0.6816 - val_loss: 0.5983 - val_accuracy: 0.6674\n",
      "Epoch 76/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5848 - accuracy: 0.6790 - val_loss: 0.5981 - val_accuracy: 0.6652\n",
      "Epoch 77/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5845 - accuracy: 0.6785 - val_loss: 0.5983 - val_accuracy: 0.6637\n",
      "Epoch 78/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5843 - accuracy: 0.6839 - val_loss: 0.5975 - val_accuracy: 0.6615\n",
      "Epoch 79/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5835 - accuracy: 0.6833 - val_loss: 0.5978 - val_accuracy: 0.6652\n",
      "Epoch 80/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5830 - accuracy: 0.6824 - val_loss: 0.5967 - val_accuracy: 0.6667\n",
      "Epoch 81/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5822 - accuracy: 0.6831 - val_loss: 0.5957 - val_accuracy: 0.6718\n",
      "Epoch 82/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5823 - accuracy: 0.6835 - val_loss: 0.5957 - val_accuracy: 0.6681\n",
      "Epoch 83/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5818 - accuracy: 0.6848 - val_loss: 0.5963 - val_accuracy: 0.6689\n",
      "Epoch 84/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5809 - accuracy: 0.6853 - val_loss: 0.5950 - val_accuracy: 0.6681\n",
      "Epoch 85/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5810 - accuracy: 0.6861 - val_loss: 0.5963 - val_accuracy: 0.6681\n",
      "Epoch 86/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5807 - accuracy: 0.6866 - val_loss: 0.5949 - val_accuracy: 0.6748\n",
      "Epoch 87/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5800 - accuracy: 0.6857 - val_loss: 0.5956 - val_accuracy: 0.6659\n",
      "Epoch 88/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5792 - accuracy: 0.6888 - val_loss: 0.5940 - val_accuracy: 0.6748\n",
      "Epoch 89/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5786 - accuracy: 0.6881 - val_loss: 0.5962 - val_accuracy: 0.6689\n",
      "Epoch 90/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5784 - accuracy: 0.6864 - val_loss: 0.5946 - val_accuracy: 0.6681\n",
      "Epoch 91/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5784 - accuracy: 0.6903 - val_loss: 0.5955 - val_accuracy: 0.6659\n",
      "Epoch 92/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5780 - accuracy: 0.6866 - val_loss: 0.5948 - val_accuracy: 0.6689\n",
      "Epoch 93/1000\n",
      "11/11 [==============================] - 0s 13ms/step - loss: 0.5773 - accuracy: 0.6883 - val_loss: 0.5942 - val_accuracy: 0.6696\n",
      "Epoch 94/1000\n",
      "11/11 [==============================] - 0s 8ms/step - loss: 0.5774 - accuracy: 0.6894 - val_loss: 0.5958 - val_accuracy: 0.6644\n",
      "Epoch 95/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5764 - accuracy: 0.6894 - val_loss: 0.5939 - val_accuracy: 0.6711\n",
      "Epoch 96/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5763 - accuracy: 0.6872 - val_loss: 0.5952 - val_accuracy: 0.6659\n",
      "Epoch 97/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5762 - accuracy: 0.6909 - val_loss: 0.5947 - val_accuracy: 0.6689\n",
      "Epoch 98/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5754 - accuracy: 0.6911 - val_loss: 0.5945 - val_accuracy: 0.6696\n",
      "Epoch 99/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5751 - accuracy: 0.6938 - val_loss: 0.5936 - val_accuracy: 0.6667\n",
      "Epoch 100/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5750 - accuracy: 0.6905 - val_loss: 0.5949 - val_accuracy: 0.6622\n",
      "Epoch 101/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5746 - accuracy: 0.6918 - val_loss: 0.5939 - val_accuracy: 0.6674\n",
      "Epoch 102/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5747 - accuracy: 0.6896 - val_loss: 0.5936 - val_accuracy: 0.6667\n",
      "Epoch 103/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5742 - accuracy: 0.6922 - val_loss: 0.5941 - val_accuracy: 0.6674\n",
      "Epoch 104/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5736 - accuracy: 0.6901 - val_loss: 0.5929 - val_accuracy: 0.6667\n",
      "Epoch 105/1000\n",
      "11/11 [==============================] - 0s 15ms/step - loss: 0.5741 - accuracy: 0.6929 - val_loss: 0.5937 - val_accuracy: 0.6659\n",
      "Epoch 106/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.5742 - accuracy: 0.6892 - val_loss: 0.5927 - val_accuracy: 0.6696\n",
      "Epoch 107/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.5728 - accuracy: 0.6916 - val_loss: 0.5937 - val_accuracy: 0.6667\n",
      "Epoch 108/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5726 - accuracy: 0.6912 - val_loss: 0.5924 - val_accuracy: 0.6718\n",
      "Epoch 109/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5725 - accuracy: 0.6922 - val_loss: 0.5932 - val_accuracy: 0.6748\n",
      "Epoch 110/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5722 - accuracy: 0.6953 - val_loss: 0.5935 - val_accuracy: 0.6696\n",
      "Epoch 111/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5724 - accuracy: 0.6940 - val_loss: 0.5917 - val_accuracy: 0.6814\n",
      "Epoch 112/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5717 - accuracy: 0.6951 - val_loss: 0.5921 - val_accuracy: 0.6718\n",
      "Epoch 113/1000\n",
      "11/11 [==============================] - 1s 8ms/step - loss: 0.5719 - accuracy: 0.6938 - val_loss: 0.5923 - val_accuracy: 0.6711\n",
      "Epoch 114/1000\n",
      "11/11 [==============================] - 0s 14ms/step - loss: 0.5712 - accuracy: 0.6948 - val_loss: 0.5923 - val_accuracy: 0.6778\n",
      "Epoch 115/1000\n",
      "11/11 [==============================] - 0s 12ms/step - loss: 0.5717 - accuracy: 0.6935 - val_loss: 0.5925 - val_accuracy: 0.6741\n",
      "Epoch 116/1000\n",
      "11/11 [==============================] - 0s 16ms/step - loss: 0.5713 - accuracy: 0.6968 - val_loss: 0.5917 - val_accuracy: 0.6785\n",
      "Epoch 117/1000\n",
      "11/11 [==============================] - 0s 15ms/step - loss: 0.5715 - accuracy: 0.6905 - val_loss: 0.5927 - val_accuracy: 0.6711\n",
      "Epoch 118/1000\n",
      "11/11 [==============================] - 0s 34ms/step - loss: 0.5706 - accuracy: 0.6953 - val_loss: 0.5920 - val_accuracy: 0.6822\n",
      "Epoch 119/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5709 - accuracy: 0.6935 - val_loss: 0.5917 - val_accuracy: 0.6763\n",
      "Epoch 120/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5705 - accuracy: 0.6942 - val_loss: 0.5919 - val_accuracy: 0.6733\n",
      "Epoch 121/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5698 - accuracy: 0.6960 - val_loss: 0.5918 - val_accuracy: 0.6778\n",
      "Epoch 122/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5700 - accuracy: 0.6981 - val_loss: 0.5904 - val_accuracy: 0.6748\n",
      "Epoch 123/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5700 - accuracy: 0.6949 - val_loss: 0.5919 - val_accuracy: 0.6748\n",
      "Epoch 124/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5695 - accuracy: 0.6955 - val_loss: 0.5917 - val_accuracy: 0.6807\n",
      "Epoch 125/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5695 - accuracy: 0.6959 - val_loss: 0.5909 - val_accuracy: 0.6785\n",
      "Epoch 126/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5683 - accuracy: 0.6979 - val_loss: 0.5914 - val_accuracy: 0.6770\n",
      "Epoch 127/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.5683 - accuracy: 0.6957 - val_loss: 0.5903 - val_accuracy: 0.6785\n",
      "Epoch 128/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5680 - accuracy: 0.6972 - val_loss: 0.5904 - val_accuracy: 0.6741\n",
      "Epoch 129/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5678 - accuracy: 0.6948 - val_loss: 0.5905 - val_accuracy: 0.6792\n",
      "Epoch 130/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5678 - accuracy: 0.6986 - val_loss: 0.5912 - val_accuracy: 0.6755\n",
      "Epoch 131/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5673 - accuracy: 0.6986 - val_loss: 0.5905 - val_accuracy: 0.6785\n",
      "Epoch 132/1000\n",
      "11/11 [==============================] - 0s 6ms/step - loss: 0.5665 - accuracy: 0.6992 - val_loss: 0.5896 - val_accuracy: 0.6763\n",
      "Epoch 133/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.5668 - accuracy: 0.6949 - val_loss: 0.5887 - val_accuracy: 0.6829\n",
      "Epoch 134/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5664 - accuracy: 0.6997 - val_loss: 0.5912 - val_accuracy: 0.6763\n",
      "Epoch 135/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5660 - accuracy: 0.6984 - val_loss: 0.5910 - val_accuracy: 0.6807\n",
      "Epoch 136/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5658 - accuracy: 0.6977 - val_loss: 0.5896 - val_accuracy: 0.6800\n",
      "Epoch 137/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5653 - accuracy: 0.6997 - val_loss: 0.5896 - val_accuracy: 0.6770\n",
      "Epoch 138/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5654 - accuracy: 0.7001 - val_loss: 0.5893 - val_accuracy: 0.6770\n",
      "Epoch 139/1000\n",
      "11/11 [==============================] - 0s 4ms/step - loss: 0.5650 - accuracy: 0.6960 - val_loss: 0.5906 - val_accuracy: 0.6800\n",
      "Epoch 140/1000\n",
      "11/11 [==============================] - 0s 14ms/step - loss: 0.5655 - accuracy: 0.6968 - val_loss: 0.5899 - val_accuracy: 0.6785\n",
      "Epoch 141/1000\n",
      "11/11 [==============================] - 0s 24ms/step - loss: 0.5649 - accuracy: 0.6996 - val_loss: 0.5900 - val_accuracy: 0.6800\n",
      "Epoch 142/1000\n",
      "11/11 [==============================] - 0s 7ms/step - loss: 0.5643 - accuracy: 0.7042 - val_loss: 0.5891 - val_accuracy: 0.6763\n",
      "Epoch 143/1000\n",
      "11/11 [==============================] - 0s 5ms/step - loss: 0.5639 - accuracy: 0.6977 - val_loss: 0.5893 - val_accuracy: 0.6800\n",
      "Time taken: 13.1s\n"
     ]
    }
   ],
   "source": [
    "# Train classifier\n",
    "import time\n",
    "start = time.time() # time at start of whole processing\n",
    "\n",
    "history = model.fit(X_train_val, \n",
    "                    Y_train_val, \n",
    "                    epochs=1000, \n",
    "                    batch_size=512, \n",
    "                    verbose=1, # switch to 1 for more verbosity \n",
    "                    callbacks=[early_stopping, model_checkpoint], \n",
    "                    validation_data=(X_test,Y_test))\n",
    "\n",
    "elapsed = time.time() - start # time after whole processing\n",
    "print(\"Time taken: \"+str(round(elapsed,1))+\"s\") # print total time taken to process every file"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot performance and reports\n",
    "Here, we plot the history of the training and the peNNormance in a ROC curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 835us/step\n",
      "170/170 [==============================] - 0s 712us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 842us/step\n",
      "170/170 [==============================] - 0s 732us/step\n",
      "212/212 [==============================] - 0s 938us/step\n",
      "Test acc score: 0.679970436067997\n",
      "Train acc score: 0.6995565410199557\n",
      "Classification report for the test set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.70      0.61      0.65       666\n",
      "      signal       0.66      0.75      0.70       687\n",
      "\n",
      "    accuracy                           0.68      1353\n",
      "   macro avg       0.68      0.68      0.68      1353\n",
      "weighted avg       0.68      0.68      0.68      1353\n",
      "\n",
      "Classification report for the total set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.70      0.64      0.67      3265\n",
      "      signal       0.69      0.75      0.72      3500\n",
      "\n",
      "    accuracy                           0.70      6765\n",
      "   macro avg       0.70      0.69      0.69      6765\n",
      "weighted avg       0.70      0.70      0.69      6765\n",
      "\n",
      "Confusion matrix for total set\n",
      "[[2075 1190]\n",
      " [ 869 2631]]\n",
      "212/212 [==============================] - 0s 729us/step\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve, auc, classification_report, accuracy_score, confusion_matrix\n",
    "\n",
    "%matplotlib inline\n",
    "def loss_accuracy_plotter(history):\n",
    "    # plot loss vs epoch\n",
    "    plt.figure(figsize=(15,10))\n",
    "    ax = plt.subplot(2, 2, 1)\n",
    "    ax.plot(history.history['loss'], label='loss')\n",
    "    ax.plot(history.history['val_loss'], label='val_loss')\n",
    "    ax.legend(loc=\"upper right\")\n",
    "    ax.set_title('Loss over epochs')\n",
    "    ax.set_xlabel('epoch')\n",
    "    ax.set_ylabel('loss')\n",
    "\n",
    "    # plot accuracy vs epoch\n",
    "    ax = plt.subplot(2, 2, 2)\n",
    "    ax.plot(history.history['accuracy'], label='acc')\n",
    "    ax.plot(history.history['val_accuracy'], label='val_acc')\n",
    "    ax.legend(loc=\"lower right\")\n",
    "    ax.set_title('Accuracy over epochs')\n",
    "    ax.set_xlabel('epoch')\n",
    "    ax.set_ylabel('acc')\n",
    "\n",
    "    # Plot ROC\n",
    "    Y_predict = model.predict(X_test)\n",
    "    Y_predict_train = model.predict(X_train_val)\n",
    "    fpr, tpr, thresholds = roc_curve(Y_test, Y_predict)\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    ax = plt.subplot(2, 2, 3)\n",
    "    ax.plot(fpr, tpr, lw=2, color='cyan', label='auc = %.3f' % (roc_auc))\n",
    "    ax.plot([0, 1], [0, 1], linestyle='--', lw=2, color='k', label='random chance')\n",
    "    ax.set_xlim([0, 1.0])\n",
    "    ax.set_ylim([0, 1.0])\n",
    "    ax.set_xlabel('false positive rate')\n",
    "    ax.set_ylabel('true positive rate')\n",
    "    ax.set_title('receiver operating curve')\n",
    "    ax.legend(loc=\"lower right\")\n",
    "    plt.show()\n",
    "\n",
    "loss_accuracy_plotter(history)\n",
    "\n",
    "\n",
    "from sklearn.metrics import accuracy_score, classification_report\n",
    "\n",
    "Y_pred_test = model.predict(X_test)\n",
    "Y_pred_test = Y_pred_test.reshape(Y_pred_test.shape[0])\n",
    "Y_pred_train = model.predict(X_train_val)\n",
    "Y_pred_train = Y_pred_train.reshape(Y_pred_train.shape[0])\n",
    "Y_pred_tot = model.predict(scaler.transform(X)).round()\n",
    "Y_pred_train = Y_pred_train.reshape(Y_pred_train.shape[0])\n",
    "\n",
    "# accuracy score\n",
    "print(\"Test acc score:\", accuracy_score(Y_test.round(), Y_pred_test.round()))\n",
    "print(\"Train acc score:\", accuracy_score(Y_train_val.round(), Y_pred_train.round()))\n",
    "\n",
    "# classification report for the test and total sets\n",
    "print(\"Classification report for the test set\")\n",
    "print(classification_report(Y_test.round(), Y_pred_test.round(), target_names=['background', 'signal']))\n",
    "print(\"Classification report for the total set\")\n",
    "print(classification_report(Y.round(), Y_pred_tot.round(), target_names=['background', 'signal']))\n",
    "\n",
    "print(\"Confusion matrix for total set\")\n",
    "print(confusion_matrix(Y.round(), Y_pred_tot.round()))\n",
    "\n",
    "fpr_nn, tpr_nn, thresholds_nn = roc_curve(Y_test, Y_pred_test) # get FPRs, TPRs and thresholds for neural network\n",
    "\n",
    "df_all['dense_NN'] = model.predict(scaler.transform(X)).round() # add prediction to array\n",
    "#df_all['dense_NN'] = model.predict_proba(X)[:, 1]\n",
    "#print (df_all.iloc[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "212/212 [==============================] - 0s 973us/step\n",
      "212/212 [==============================] - 0s 970us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# add TF NN probability\n",
    "y_tot_prob_NN = model.predict(scaler.transform(X))\n",
    "y_tot_NN = model.predict(scaler.transform(X)).round()\n",
    "\n",
    "cumulative_events = 0 # start counter for total number of events for which output is saved\n",
    "for key in filename.keys(): # loop over samples\n",
    "    df[key]['NN_output_prob'] = y_tot_prob_NN[cumulative_events:cumulative_events+len(df[key])]\n",
    "    df[key]['NN_output'] = y_tot_NN[cumulative_events:cumulative_events+len(df[key])]\n",
    "    cumulative_events += len(df[key]) # increment counter for total number of events\n",
    "    #print(df[key]['NN_output']) # print the dataframe column NN_output\n",
    "\n",
    "# probability and significance plots\n",
    "NN_output_prob = { # dictionary containing plotting parameters for the mllll histogram\n",
    "    # change plotting parameters\n",
    "    'bin_width':0.1, # width of each histogram bin\n",
    "    'num_bins':8, # number of histogram bins\n",
    "    'xrange_min':0, # minimum on x-axis\n",
    "    'xlabel':'Neural Network output', # x-axis label\n",
    "}\n",
    "\n",
    "SoverB_hist_dict = {'NN_output_prob':NN_output_prob}\n",
    "plot_SoverB(df, SoverB_hist_dict)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NN Hyperparameter optimization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-18-7b86a5d1fe7c>:4: DeprecationWarning: KerasClassifier is deprecated, use Sci-Keras (https://github.com/adriangb/scikeras) instead. See https://www.adriangb.com/scikeras/stable/migration.html for help migrating.\n",
      "  (\"clf\", keras.wrappers.scikit_learn.KerasClassifier(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "34/34 [==============================] - 0s 1000us/step\n",
      "34/34 [==============================] - 0s 807us/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 1s 2ms/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 2ms/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 861us/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 925us/step\n",
      "34/34 [==============================] - 0s 844us/step\n",
      "34/34 [==============================] - 0s 947us/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 845us/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 819us/step\n",
      "34/34 [==============================] - 0s 1ms/step\n",
      "34/34 [==============================] - 0s 938us/step\n",
      "34/34 [==============================] - 0s 846us/step\n",
      "34/34 [==============================] - 0s 899us/step\n",
      "34/34 [==============================] - 0s 793us/step\n",
      "34/34 [==============================] - 0s 936us/step\n",
      "34/34 [==============================] - 0s 796us/step\n",
      "Best score obtained: 0.7023300785283572\n",
      "Parameters:\n",
      "\tclf__verbose: 0\n",
      "\tclf__optimizer: Nadam\n",
      "\tclf__network_layers: (128, 128, 128)\n",
      "\tclf__k_initializer: normal\n",
      "\tclf__epochs: 200\n",
      "\tclf__dropout_rate: 0.1\n",
      "\tclf__batch_size: 256\n",
      "\tclf__activation: selu\n"
     ]
    }
   ],
   "source": [
    "def main():\n",
    "        \n",
    "    keras_pipeline = Pipeline([(\"scaler\", StandardScaler()),\n",
    "                               (\"clf\", keras.wrappers.scikit_learn.KerasClassifier(\n",
    "                                   build_fn=make_model))\n",
    "    ])\n",
    "    \n",
    "    param_grid = {'clf__network_layers': [(32, 32), (64, 64), (128, 128, 128)],\n",
    "     'clf__batch_size': [64, 128, 256, 512],\n",
    "     'clf__epochs': [5, 10, 15, 50, 100, 200],\n",
    "     'clf__dropout_rate': [0.1, 0.01],\n",
    "     'clf__optimizer': ['Nadam', 'Adam', 'SGD'],\n",
    "     'clf__activation': ['selu', 'relu', 'tanh'],\n",
    "     'clf__k_initializer': ['lecun_normal', 'normal'],\n",
    "     'clf__verbose': [0]\n",
    "    }\n",
    "    \n",
    "    rs_keras = RandomizedSearchCV(keras_pipeline,\n",
    "                                  param_distributions=param_grid,\n",
    "                                  cv=5, refit=True,\n",
    "                                  verbose=0,\n",
    "                                  n_iter=5,\n",
    "                                  scoring=\"accuracy\")\n",
    "    \n",
    "    history = rs_keras.fit(np.array(X_train_val),\n",
    "                 np.array(Y_train_val))\n",
    "    \n",
    "    print('Best score obtained: {0}'.format(rs_keras.best_score_))\n",
    "    print('Parameters:')\n",
    "    for param, value in rs_keras.best_params_.items():\n",
    "        print('\\t{}: {}'.format(param, value))\n",
    "    return  rs_keras  \n",
    "\n",
    "def make_model(network_layers=[(32, 32)],\n",
    "               dropout_rate=0,\n",
    "               optimizer=\"Nadam\",\n",
    "               activation=\"selu\",\n",
    "               k_initializer='lecun_normal',\n",
    "               n_input=NDIM,\n",
    "               n_class=1):\n",
    "    \n",
    "    model = keras.models.Sequential()\n",
    "\n",
    "    for index, layers in enumerate(network_layers):\n",
    "        if not index:\n",
    "            model.add(keras.layers.Dense(layers,\n",
    "                                         input_dim=n_input,\n",
    "                                         activation=activation,\n",
    "                                         kernel_initializer=k_initializer))\n",
    "        else:\n",
    "            model.add(keras.layers.Dense(layers,\n",
    "                                         kernel_initializer=k_initializer,\n",
    "                                         activation=activation))\n",
    "        if dropout_rate and index:\n",
    "            model.add(keras.layers.AlphaDropout(dropout_rate))\n",
    "\n",
    "    model.add(keras.layers.Dense(n_class,\n",
    "                                 activation=\"sigmoid\"))\n",
    "    \n",
    "    model.compile(loss='binary_crossentropy',\n",
    "                  optimizer=optimizer,\n",
    "                  metrics=['accuracy'])\n",
    "    #print(model.summary())\n",
    "    return model\n",
    "\n",
    "if __name__ == '__main__':\n",
    "    opt_model = main()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nBest score obtained: 0.7023300785283572\\nParameters:\\n\\tclf__verbose: 0\\n\\tclf__optimizer: Nadam\\n\\tclf__network_layers: (128, 128, 128)\\n\\tclf__k_initializer: normal\\n\\tclf__epochs: 200\\n\\tclf__dropout_rate: 0.1\\n\\tclf__batch_size: 256\\n\\tclf__activation: selu\\n}\\n'"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''\n",
    "Best score obtained: 0.7023300785283572\n",
    "Parameters:\n",
    "\tclf__verbose: 0\n",
    "\tclf__optimizer: Nadam\n",
    "\tclf__network_layers: (128, 128, 128)\n",
    "\tclf__k_initializer: normal\n",
    "\tclf__epochs: 200\n",
    "\tclf__dropout_rate: 0.1\n",
    "\tclf__batch_size: 256\n",
    "\tclf__activation: selu\n",
    "}\n",
    "'''"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## NN with dropout layer for overfit checking"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_27\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_88 (Dense)            (None, 128)               1792      \n",
      "                                                                 \n",
      " dropout_3 (Dropout)         (None, 128)               0         \n",
      "                                                                 \n",
      " dense_89 (Dense)            (None, 128)               16512     \n",
      "                                                                 \n",
      " dropout_4 (Dropout)         (None, 128)               0         \n",
      "                                                                 \n",
      " dense_90 (Dense)            (None, 128)               16512     \n",
      "                                                                 \n",
      " dropout_5 (Dropout)         (None, 128)               0         \n",
      "                                                                 \n",
      " dense_91 (Dense)            (None, 1)                 129       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 34,945\n",
      "Trainable params: 34,945\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/500\n",
      "22/22 [==============================] - 2s 18ms/step - loss: 0.7079 - accuracy: 0.5687 - val_loss: 0.6624 - val_accuracy: 0.6149\n",
      "Epoch 2/500\n",
      "22/22 [==============================] - 0s 14ms/step - loss: 0.6700 - accuracy: 0.5944 - val_loss: 0.6454 - val_accuracy: 0.6171\n",
      "Epoch 3/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.6560 - accuracy: 0.6171 - val_loss: 0.6456 - val_accuracy: 0.6186\n",
      "Epoch 4/500\n",
      "22/22 [==============================] - 0s 12ms/step - loss: 0.6491 - accuracy: 0.6164 - val_loss: 0.6412 - val_accuracy: 0.6061\n",
      "Epoch 5/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.6470 - accuracy: 0.6232 - val_loss: 0.6492 - val_accuracy: 0.6075\n",
      "Epoch 6/500\n",
      "22/22 [==============================] - 0s 11ms/step - loss: 0.6431 - accuracy: 0.6256 - val_loss: 0.6241 - val_accuracy: 0.6356\n",
      "Epoch 7/500\n",
      "22/22 [==============================] - 0s 7ms/step - loss: 0.6394 - accuracy: 0.6299 - val_loss: 0.6286 - val_accuracy: 0.6275\n",
      "Epoch 8/500\n",
      "22/22 [==============================] - 0s 13ms/step - loss: 0.6312 - accuracy: 0.6378 - val_loss: 0.6283 - val_accuracy: 0.6504\n",
      "Epoch 9/500\n",
      "22/22 [==============================] - 0s 8ms/step - loss: 0.6322 - accuracy: 0.6465 - val_loss: 0.6293 - val_accuracy: 0.6356\n",
      "Epoch 10/500\n",
      "22/22 [==============================] - 0s 10ms/step - loss: 0.6265 - accuracy: 0.6506 - val_loss: 0.6154 - val_accuracy: 0.6548\n",
      "Epoch 11/500\n",
      "22/22 [==============================] - 0s 8ms/step - loss: 0.6261 - accuracy: 0.6443 - val_loss: 0.6167 - val_accuracy: 0.6445\n",
      "Epoch 12/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.6237 - accuracy: 0.6434 - val_loss: 0.6119 - val_accuracy: 0.6482\n",
      "Epoch 13/500\n",
      "22/22 [==============================] - 0s 10ms/step - loss: 0.6168 - accuracy: 0.6532 - val_loss: 0.6079 - val_accuracy: 0.6622\n",
      "Epoch 14/500\n",
      "22/22 [==============================] - 0s 7ms/step - loss: 0.6229 - accuracy: 0.6611 - val_loss: 0.6084 - val_accuracy: 0.6630\n",
      "Epoch 15/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.6152 - accuracy: 0.6580 - val_loss: 0.6162 - val_accuracy: 0.6489\n",
      "Epoch 16/500\n",
      "22/22 [==============================] - 0s 8ms/step - loss: 0.6122 - accuracy: 0.6591 - val_loss: 0.6065 - val_accuracy: 0.6615\n",
      "Epoch 17/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.6195 - accuracy: 0.6493 - val_loss: 0.6077 - val_accuracy: 0.6556\n",
      "Epoch 18/500\n",
      "22/22 [==============================] - 0s 7ms/step - loss: 0.6233 - accuracy: 0.6574 - val_loss: 0.6125 - val_accuracy: 0.6541\n",
      "Epoch 19/500\n",
      "22/22 [==============================] - 0s 7ms/step - loss: 0.6161 - accuracy: 0.6534 - val_loss: 0.6100 - val_accuracy: 0.6659\n",
      "Epoch 20/500\n",
      "22/22 [==============================] - 0s 8ms/step - loss: 0.6061 - accuracy: 0.6709 - val_loss: 0.6069 - val_accuracy: 0.6578\n",
      "Epoch 21/500\n",
      "22/22 [==============================] - 0s 11ms/step - loss: 0.6119 - accuracy: 0.6648 - val_loss: 0.6062 - val_accuracy: 0.6718\n",
      "Epoch 22/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.6128 - accuracy: 0.6617 - val_loss: 0.6015 - val_accuracy: 0.6792\n",
      "Epoch 23/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.6019 - accuracy: 0.6657 - val_loss: 0.5959 - val_accuracy: 0.6785\n",
      "Epoch 24/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.6071 - accuracy: 0.6672 - val_loss: 0.6031 - val_accuracy: 0.6807\n",
      "Epoch 25/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.6059 - accuracy: 0.6676 - val_loss: 0.6006 - val_accuracy: 0.6644\n",
      "Epoch 26/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.6075 - accuracy: 0.6678 - val_loss: 0.6058 - val_accuracy: 0.6556\n",
      "Epoch 27/500\n",
      "22/22 [==============================] - 0s 14ms/step - loss: 0.6059 - accuracy: 0.6728 - val_loss: 0.5966 - val_accuracy: 0.6622\n",
      "Epoch 28/500\n",
      "22/22 [==============================] - 0s 17ms/step - loss: 0.5990 - accuracy: 0.6720 - val_loss: 0.6038 - val_accuracy: 0.6674\n",
      "Epoch 29/500\n",
      "22/22 [==============================] - 0s 12ms/step - loss: 0.5966 - accuracy: 0.6754 - val_loss: 0.5981 - val_accuracy: 0.6681\n",
      "Epoch 30/500\n",
      "22/22 [==============================] - 0s 8ms/step - loss: 0.5959 - accuracy: 0.6807 - val_loss: 0.5970 - val_accuracy: 0.6644\n",
      "Epoch 31/500\n",
      "22/22 [==============================] - 0s 8ms/step - loss: 0.5956 - accuracy: 0.6781 - val_loss: 0.5965 - val_accuracy: 0.6718\n",
      "Epoch 32/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5995 - accuracy: 0.6691 - val_loss: 0.6034 - val_accuracy: 0.6659\n",
      "Epoch 33/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5990 - accuracy: 0.6729 - val_loss: 0.5932 - val_accuracy: 0.6792\n",
      "Epoch 34/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5922 - accuracy: 0.6748 - val_loss: 0.5919 - val_accuracy: 0.6807\n",
      "Epoch 35/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5906 - accuracy: 0.6803 - val_loss: 0.5900 - val_accuracy: 0.6807\n",
      "Epoch 36/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5897 - accuracy: 0.6768 - val_loss: 0.5939 - val_accuracy: 0.6696\n",
      "Epoch 37/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5933 - accuracy: 0.6757 - val_loss: 0.5944 - val_accuracy: 0.6696\n",
      "Epoch 38/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5967 - accuracy: 0.6735 - val_loss: 0.5876 - val_accuracy: 0.6755\n",
      "Epoch 39/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5925 - accuracy: 0.6811 - val_loss: 0.5885 - val_accuracy: 0.6800\n",
      "Epoch 40/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5902 - accuracy: 0.6833 - val_loss: 0.5838 - val_accuracy: 0.6800\n",
      "Epoch 41/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5867 - accuracy: 0.6877 - val_loss: 0.5913 - val_accuracy: 0.6711\n",
      "Epoch 42/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5875 - accuracy: 0.6870 - val_loss: 0.5895 - val_accuracy: 0.6659\n",
      "Epoch 43/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5880 - accuracy: 0.6774 - val_loss: 0.5907 - val_accuracy: 0.6755\n",
      "Epoch 44/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.5817 - accuracy: 0.6914 - val_loss: 0.5881 - val_accuracy: 0.6741\n",
      "Epoch 45/500\n",
      "22/22 [==============================] - 0s 13ms/step - loss: 0.5855 - accuracy: 0.6853 - val_loss: 0.5839 - val_accuracy: 0.6896\n",
      "Epoch 46/500\n",
      "22/22 [==============================] - 0s 10ms/step - loss: 0.5869 - accuracy: 0.6903 - val_loss: 0.5802 - val_accuracy: 0.6837\n",
      "Epoch 47/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5810 - accuracy: 0.6859 - val_loss: 0.5864 - val_accuracy: 0.6785\n",
      "Epoch 48/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5865 - accuracy: 0.6814 - val_loss: 0.5925 - val_accuracy: 0.6755\n",
      "Epoch 49/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5788 - accuracy: 0.6929 - val_loss: 0.5875 - val_accuracy: 0.6866\n",
      "Epoch 50/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5807 - accuracy: 0.6872 - val_loss: 0.5914 - val_accuracy: 0.6755\n",
      "Epoch 51/500\n",
      "22/22 [==============================] - 0s 15ms/step - loss: 0.5815 - accuracy: 0.6864 - val_loss: 0.5793 - val_accuracy: 0.6859\n",
      "Epoch 52/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5742 - accuracy: 0.6909 - val_loss: 0.5818 - val_accuracy: 0.6807\n",
      "Epoch 53/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5823 - accuracy: 0.6887 - val_loss: 0.5813 - val_accuracy: 0.6851\n",
      "Epoch 54/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5759 - accuracy: 0.6933 - val_loss: 0.5875 - val_accuracy: 0.6800\n",
      "Epoch 55/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5778 - accuracy: 0.6935 - val_loss: 0.5804 - val_accuracy: 0.6844\n",
      "Epoch 56/500\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.5700 - accuracy: 0.6942 - val_loss: 0.5793 - val_accuracy: 0.6792\n",
      "Epoch 57/500\n",
      "22/22 [==============================] - 0s 14ms/step - loss: 0.5719 - accuracy: 0.6990 - val_loss: 0.5881 - val_accuracy: 0.6748\n",
      "Epoch 58/500\n",
      "22/22 [==============================] - 0s 11ms/step - loss: 0.5718 - accuracy: 0.6964 - val_loss: 0.5875 - val_accuracy: 0.6718\n",
      "Epoch 59/500\n",
      "22/22 [==============================] - 0s 18ms/step - loss: 0.5734 - accuracy: 0.6979 - val_loss: 0.5782 - val_accuracy: 0.6829\n",
      "Epoch 60/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5689 - accuracy: 0.6936 - val_loss: 0.5785 - val_accuracy: 0.6822\n",
      "Epoch 61/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5709 - accuracy: 0.6983 - val_loss: 0.5797 - val_accuracy: 0.6822\n",
      "Epoch 62/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5684 - accuracy: 0.6968 - val_loss: 0.5753 - val_accuracy: 0.6874\n",
      "Epoch 63/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5717 - accuracy: 0.6959 - val_loss: 0.5833 - val_accuracy: 0.6755\n",
      "Epoch 64/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5707 - accuracy: 0.6966 - val_loss: 0.5781 - val_accuracy: 0.6807\n",
      "Epoch 65/500\n",
      "22/22 [==============================] - 0s 7ms/step - loss: 0.5652 - accuracy: 0.7016 - val_loss: 0.5859 - val_accuracy: 0.6733\n",
      "Epoch 66/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.5662 - accuracy: 0.7060 - val_loss: 0.5722 - val_accuracy: 0.6881\n",
      "Epoch 67/500\n",
      "22/22 [==============================] - 0s 10ms/step - loss: 0.5640 - accuracy: 0.7034 - val_loss: 0.5749 - val_accuracy: 0.6962\n",
      "Epoch 68/500\n",
      "22/22 [==============================] - 0s 9ms/step - loss: 0.5641 - accuracy: 0.7101 - val_loss: 0.5754 - val_accuracy: 0.6770\n",
      "Epoch 69/500\n",
      "22/22 [==============================] - 0s 12ms/step - loss: 0.5623 - accuracy: 0.7001 - val_loss: 0.5735 - val_accuracy: 0.6896\n",
      "Epoch 70/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5641 - accuracy: 0.6981 - val_loss: 0.5749 - val_accuracy: 0.6940\n",
      "Epoch 71/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5583 - accuracy: 0.7068 - val_loss: 0.5780 - val_accuracy: 0.6829\n",
      "Epoch 72/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5573 - accuracy: 0.7082 - val_loss: 0.5709 - val_accuracy: 0.6844\n",
      "Epoch 73/500\n",
      "22/22 [==============================] - 0s 19ms/step - loss: 0.5648 - accuracy: 0.7014 - val_loss: 0.5691 - val_accuracy: 0.7007\n",
      "Epoch 74/500\n",
      "22/22 [==============================] - 0s 10ms/step - loss: 0.5562 - accuracy: 0.7036 - val_loss: 0.5738 - val_accuracy: 0.6837\n",
      "Epoch 75/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5542 - accuracy: 0.7081 - val_loss: 0.5723 - val_accuracy: 0.6962\n",
      "Epoch 76/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5539 - accuracy: 0.7071 - val_loss: 0.5711 - val_accuracy: 0.6911\n",
      "Epoch 77/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5541 - accuracy: 0.7093 - val_loss: 0.5692 - val_accuracy: 0.6881\n",
      "Epoch 78/500\n",
      "22/22 [==============================] - 0s 13ms/step - loss: 0.5588 - accuracy: 0.7075 - val_loss: 0.5795 - val_accuracy: 0.6999\n",
      "Epoch 79/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5497 - accuracy: 0.7127 - val_loss: 0.5684 - val_accuracy: 0.7007\n",
      "Epoch 80/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5536 - accuracy: 0.7116 - val_loss: 0.5738 - val_accuracy: 0.6940\n",
      "Epoch 81/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5541 - accuracy: 0.7123 - val_loss: 0.5666 - val_accuracy: 0.6984\n",
      "Epoch 82/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5510 - accuracy: 0.7142 - val_loss: 0.5683 - val_accuracy: 0.6881\n",
      "Epoch 83/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5515 - accuracy: 0.7105 - val_loss: 0.5694 - val_accuracy: 0.6888\n",
      "Epoch 84/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5586 - accuracy: 0.7068 - val_loss: 0.5659 - val_accuracy: 0.6829\n",
      "Epoch 85/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5499 - accuracy: 0.7143 - val_loss: 0.5720 - val_accuracy: 0.6896\n",
      "Epoch 86/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5476 - accuracy: 0.7206 - val_loss: 0.5687 - val_accuracy: 0.6866\n",
      "Epoch 87/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5450 - accuracy: 0.7184 - val_loss: 0.5643 - val_accuracy: 0.6933\n",
      "Epoch 88/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5483 - accuracy: 0.7103 - val_loss: 0.5708 - val_accuracy: 0.6903\n",
      "Epoch 89/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5431 - accuracy: 0.7193 - val_loss: 0.5607 - val_accuracy: 0.7007\n",
      "Epoch 90/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5498 - accuracy: 0.7123 - val_loss: 0.5795 - val_accuracy: 0.6822\n",
      "Epoch 91/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5425 - accuracy: 0.7184 - val_loss: 0.5626 - val_accuracy: 0.6948\n",
      "Epoch 92/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5428 - accuracy: 0.7217 - val_loss: 0.5717 - val_accuracy: 0.6881\n",
      "Epoch 93/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5484 - accuracy: 0.7160 - val_loss: 0.5626 - val_accuracy: 0.6984\n",
      "Epoch 94/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5451 - accuracy: 0.7203 - val_loss: 0.5705 - val_accuracy: 0.6940\n",
      "Epoch 95/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5441 - accuracy: 0.7178 - val_loss: 0.5628 - val_accuracy: 0.7014\n",
      "Epoch 96/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5381 - accuracy: 0.7313 - val_loss: 0.5704 - val_accuracy: 0.6984\n",
      "Epoch 97/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5361 - accuracy: 0.7245 - val_loss: 0.5631 - val_accuracy: 0.6999\n",
      "Epoch 98/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5360 - accuracy: 0.7295 - val_loss: 0.5643 - val_accuracy: 0.6933\n",
      "Epoch 99/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5377 - accuracy: 0.7214 - val_loss: 0.5621 - val_accuracy: 0.6903\n",
      "Time taken: 19.2s\n"
     ]
    }
   ],
   "source": [
    "# Packages for modeling\n",
    "from keras import models\n",
    "from keras import layers\n",
    "from keras import regularizers\n",
    "\n",
    "drop_model = models.Sequential()\n",
    "drop_model.add(layers.Dense(128,  kernel_initializer='lecun_normal', activation='selu', input_shape=(NDIM,)))\n",
    "drop_model.add(layers.Dropout(0.1))\n",
    "drop_model.add(layers.Dense(128,  kernel_initializer='lecun_normal', activation='selu'))\n",
    "drop_model.add(layers.Dropout(0.1))\n",
    "drop_model.add(layers.Dense(128,  kernel_initializer='lecun_normal', activation='selu'))\n",
    "drop_model.add(layers.Dropout(0.1))\n",
    "drop_model.add(layers.Dense(1, kernel_initializer='lecun_normal', activation='sigmoid'))\n",
    "\n",
    "# compile the model\n",
    "drop_model.compile(optimizer='Adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "# print the model summary\n",
    "drop_model.summary()\n",
    "\n",
    "# early stopping callback\n",
    "from keras.callbacks import EarlyStopping\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=10, mode='min')\n",
    "\n",
    "# Train classifier\n",
    "import time\n",
    "start = time.time() # time at start of whole processing\n",
    "\n",
    "drop_history = drop_model.fit(X_train_val, \n",
    "                    Y_train_val, \n",
    "                    epochs=500, \n",
    "                    batch_size=256, \n",
    "                    verbose=1, # switch to 1 for more verbosity \n",
    "                    callbacks=[early_stopping], \n",
    "                    validation_data=(X_test,Y_test))\n",
    "\n",
    "elapsed = time.time() - start # time after whole processing\n",
    "print(\"Time taken: \"+str(round(elapsed,1))+\"s\") # print total time taken to process every file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 1ms/step\n",
      "170/170 [==============================] - 0s 942us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 1ms/step\n",
      "170/170 [==============================] - 0s 911us/step\n",
      "3060/3060 [==============================] - 4s 1ms/step\n",
      "Test acc score: 0.6903178122690318\n",
      "Total acc score: 0.695142378559464\n",
      "\n",
      "\n",
      "\n",
      "Classification report for the test set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.70      0.65      0.67       666\n",
      "      signal       0.68      0.73      0.71       687\n",
      "\n",
      "    accuracy                           0.69      1353\n",
      "   macro avg       0.69      0.69      0.69      1353\n",
      "weighted avg       0.69      0.69      0.69      1353\n",
      "\n",
      "Classification report for the total set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.07      0.70      0.13      3265\n",
      "      signal       0.99      0.70      0.82     94643\n",
      "\n",
      "    accuracy                           0.70     97908\n",
      "   macro avg       0.53      0.70      0.47     97908\n",
      "weighted avg       0.95      0.70      0.79     97908\n",
      "\n",
      "Confusion matrix for training set\n",
      "[[ 2276   989]\n",
      " [28859 65784]]\n",
      "3060/3060 [==============================] - 3s 855us/step\n"
     ]
    }
   ],
   "source": [
    "loss_accuracy_plotter(drop_history)\n",
    "\n",
    "drop_Y_pred_test = drop_model.predict(X_test)\n",
    "drop_Y_pred_test = drop_Y_pred_test.reshape(drop_Y_pred_test.shape[0])\n",
    "drop_Y_pred_train = drop_model.predict(X_train_val)\n",
    "drop_Y_pred_train = drop_Y_pred_train.reshape(drop_Y_pred_train.shape[0])\n",
    "drop_Y_pred_tot = drop_model.predict(scaler.transform(X)).round()\n",
    "drop_Y_pred_tot = drop_Y_pred_tot.reshape(drop_Y_pred_tot.shape[0])\n",
    "\n",
    "# accuracy score\n",
    "print(\"Test acc score:\", accuracy_score(Y_test.round(), drop_Y_pred_test.round()))\n",
    "#print(\"Train acc score:\", accuracy_score(Y_train_val.round(), drop_Y_pred_train.round()))\n",
    "print(\"Total acc score:\", accuracy_score(Y.round(), drop_Y_pred_tot.round()))\n",
    "print(\"\\n\\n\")\n",
    "\n",
    "# classification report for the test and total sets\n",
    "print(\"Classification report for the test set\")\n",
    "print(classification_report(Y_test.round(), drop_Y_pred_test.round(), target_names=['background', 'signal']))\n",
    "print(\"Classification report for the total set\")\n",
    "print(classification_report(Y.round(), drop_Y_pred_tot.round(), target_names=['background', 'signal']))\n",
    "\n",
    "print(\"Confusion matrix for training set\")\n",
    "print(confusion_matrix(Y.round(), drop_Y_pred_tot.round()))\n",
    "\n",
    "fpr_nn_opt_drop, tpr_nn_opt_drop, thresholds_nn_opt_drop = roc_curve(Y_test, drop_Y_pred_test) # get FPRs, TPRs and thresholds for neural network\n",
    "\n",
    "df_all['dense_NN_opt_drop'] = model.predict(scaler.transform(X)).round() # add prediction to array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3060/3060 [==============================] - 3s 933us/step\n",
      "3060/3060 [==============================] - 3s 852us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# add TF NN probability\n",
    "y_tot_prob_NN_opt_drop = drop_model.predict(scaler.transform(X))\n",
    "y_tot_NN_opt_drop = drop_model.predict(scaler.transform(X)).round()\n",
    "\n",
    "cumulative_events = 0 # start counter for total number of events for which output is saved\n",
    "for key in filename.keys(): # loop over samples\n",
    "    df[key]['NN_opt_drop_output_prob'] = y_tot_prob_NN_opt_drop[cumulative_events:cumulative_events+len(df[key])]\n",
    "    df[key]['NN_opt_drop_output'] = y_tot_NN_opt_drop[cumulative_events:cumulative_events+len(df[key])]\n",
    "    cumulative_events += len(df[key]) # increment counter for total number of events\n",
    "    #print(df[key]['NN_output']) # print the dataframe column NN_output\n",
    "\n",
    "# probability and significance plots\n",
    "NN_opt_drop_output_prob = { # dictionary containing plotting parameters for the mllll histogram\n",
    "    # change plotting parameters\n",
    "    'bin_width':0.1, # width of each histogram bin\n",
    "    'num_bins':8, # number of histogram bins\n",
    "    'xrange_min':0, # minimum on x-axis\n",
    "    'xlabel':'Neural Network optimized dropout', # x-axis label\n",
    "}\n",
    "\n",
    "SoverB_hist_dict = {'NN_opt_drop_output_prob':NN_opt_drop_output_prob}\n",
    "plot_SoverB(df, SoverB_hist_dict)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Optimal model with early stopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_4\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input (InputLayer)          [(None, 13)]              0         \n",
      "                                                                 \n",
      " hidden1 (Dense)             (None, 128)               1792      \n",
      "                                                                 \n",
      " hidden2 (Dense)             (None, 128)               16512     \n",
      "                                                                 \n",
      " hidden3 (Dense)             (None, 128)               16512     \n",
      "                                                                 \n",
      " output (Dense)              (None, 1)                 129       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 34,945\n",
      "Trainable params: 34,945\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of batches seen.\n"
     ]
    }
   ],
   "source": [
    "inputs = Input(shape=(NDIM,), name = 'input') \n",
    "hidden1 = Dense(128, name = 'hidden1', kernel_initializer='lecun_normal', activation='selu')(inputs)\n",
    "hidden2 = Dense(128, name = 'hidden2', kernel_initializer='lecun_normal', activation='selu')(hidden1)\n",
    "hidden3 = Dense(128, name = 'hidden3', kernel_initializer='lecun_normal', activation='selu')(hidden2)\n",
    "outputs = Dense(1, name = 'output', kernel_initializer='lecun_normal', activation='sigmoid')(hidden3)\n",
    "\n",
    "# create the model\n",
    "opt_model = Model(inputs=inputs, outputs=outputs)\n",
    "# compile the model\n",
    "opt_model.compile(optimizer='Adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "# print the model summary\n",
    "opt_model.summary()\n",
    "\n",
    "# early stopping callback\n",
    "from keras.callbacks import EarlyStopping\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=10, mode='min')\n",
    "\n",
    "# model checkpoint callback\n",
    "# this saves our model architecture + parameters into dense_model.h5\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "model_checkpoint = ModelCheckpoint('TopFC_opt_dense_model.h5', monitor='val_loss', \n",
    "                                   verbose=0, save_best_only=True, \n",
    "                                   save_weights_only=False, mode='auto', \n",
    "                                   period=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/500\n",
      "22/22 [==============================] - 1s 13ms/step - loss: 0.6940 - accuracy: 0.5811 - val_loss: 0.6528 - val_accuracy: 0.5987\n",
      "Epoch 2/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.6553 - accuracy: 0.6135 - val_loss: 0.6719 - val_accuracy: 0.5772\n",
      "Epoch 3/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.6417 - accuracy: 0.6269 - val_loss: 0.6323 - val_accuracy: 0.6371\n",
      "Epoch 4/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.6234 - accuracy: 0.6473 - val_loss: 0.6327 - val_accuracy: 0.6356\n",
      "Epoch 5/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.6179 - accuracy: 0.6491 - val_loss: 0.6155 - val_accuracy: 0.6571\n",
      "Epoch 6/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.6091 - accuracy: 0.6676 - val_loss: 0.6195 - val_accuracy: 0.6489\n",
      "Epoch 7/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.6026 - accuracy: 0.6778 - val_loss: 0.6133 - val_accuracy: 0.6305\n",
      "Epoch 8/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5951 - accuracy: 0.6779 - val_loss: 0.6117 - val_accuracy: 0.6644\n",
      "Epoch 9/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5934 - accuracy: 0.6776 - val_loss: 0.6113 - val_accuracy: 0.6556\n",
      "Epoch 10/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5900 - accuracy: 0.6774 - val_loss: 0.6039 - val_accuracy: 0.6475\n",
      "Epoch 11/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5808 - accuracy: 0.6842 - val_loss: 0.5979 - val_accuracy: 0.6667\n",
      "Epoch 12/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5777 - accuracy: 0.6959 - val_loss: 0.5997 - val_accuracy: 0.6696\n",
      "Epoch 13/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5842 - accuracy: 0.6905 - val_loss: 0.5930 - val_accuracy: 0.6792\n",
      "Epoch 14/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5735 - accuracy: 0.6881 - val_loss: 0.5972 - val_accuracy: 0.6696\n",
      "Epoch 15/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5649 - accuracy: 0.7086 - val_loss: 0.6159 - val_accuracy: 0.6548\n",
      "Epoch 16/500\n",
      "22/22 [==============================] - 0s 12ms/step - loss: 0.5653 - accuracy: 0.6997 - val_loss: 0.5940 - val_accuracy: 0.6800\n",
      "Epoch 17/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5600 - accuracy: 0.7099 - val_loss: 0.5937 - val_accuracy: 0.6822\n",
      "Epoch 18/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5577 - accuracy: 0.7099 - val_loss: 0.6130 - val_accuracy: 0.6681\n",
      "Epoch 19/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5646 - accuracy: 0.7018 - val_loss: 0.6044 - val_accuracy: 0.6711\n",
      "Epoch 20/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5645 - accuracy: 0.7051 - val_loss: 0.5926 - val_accuracy: 0.6704\n",
      "Epoch 21/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5616 - accuracy: 0.7040 - val_loss: 0.5971 - val_accuracy: 0.6644\n",
      "Epoch 22/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5456 - accuracy: 0.7217 - val_loss: 0.6047 - val_accuracy: 0.6741\n",
      "Epoch 23/500\n",
      "22/22 [==============================] - 0s 10ms/step - loss: 0.5475 - accuracy: 0.7110 - val_loss: 0.5937 - val_accuracy: 0.6770\n",
      "Epoch 24/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5448 - accuracy: 0.7197 - val_loss: 0.5902 - val_accuracy: 0.6741\n",
      "Epoch 25/500\n",
      "22/22 [==============================] - 0s 6ms/step - loss: 0.5354 - accuracy: 0.7317 - val_loss: 0.5801 - val_accuracy: 0.6785\n",
      "Epoch 26/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5347 - accuracy: 0.7299 - val_loss: 0.5900 - val_accuracy: 0.6755\n",
      "Epoch 27/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5310 - accuracy: 0.7324 - val_loss: 0.5980 - val_accuracy: 0.6733\n",
      "Epoch 28/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5336 - accuracy: 0.7271 - val_loss: 0.5973 - val_accuracy: 0.6807\n",
      "Epoch 29/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5304 - accuracy: 0.7360 - val_loss: 0.5879 - val_accuracy: 0.6704\n",
      "Epoch 30/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5186 - accuracy: 0.7461 - val_loss: 0.5919 - val_accuracy: 0.6755\n",
      "Epoch 31/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5251 - accuracy: 0.7334 - val_loss: 0.5951 - val_accuracy: 0.6630\n",
      "Epoch 32/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5259 - accuracy: 0.7373 - val_loss: 0.5865 - val_accuracy: 0.6851\n",
      "Epoch 33/500\n",
      "22/22 [==============================] - 0s 5ms/step - loss: 0.5208 - accuracy: 0.7323 - val_loss: 0.5905 - val_accuracy: 0.6814\n",
      "Epoch 34/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5096 - accuracy: 0.7480 - val_loss: 0.5859 - val_accuracy: 0.6851\n",
      "Epoch 35/500\n",
      "22/22 [==============================] - 0s 4ms/step - loss: 0.5048 - accuracy: 0.7478 - val_loss: 0.5918 - val_accuracy: 0.6807\n",
      "Time taken: 5.4s\n"
     ]
    }
   ],
   "source": [
    "# Train classifier\n",
    "import time\n",
    "start = time.time() # time at start of whole processing\n",
    "\n",
    "opt_history = opt_model.fit(X_train_val, \n",
    "                    Y_train_val, \n",
    "                    epochs=500, \n",
    "                    batch_size=256, \n",
    "                    verbose=1, # switch to 1 for more verbosity \n",
    "                    callbacks=[early_stopping, model_checkpoint], \n",
    "                    validation_data=(X_test,Y_test))\n",
    "\n",
    "elapsed = time.time() - start # time after whole processing\n",
    "print(\"Time taken: \"+str(round(elapsed,1))+\"s\") # print total time taken to process every file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 746us/step\n",
      "170/170 [==============================] - 0s 765us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 1ms/step\n",
      "170/170 [==============================] - 0s 941us/step\n",
      "3060/3060 [==============================] - 4s 1ms/step\n",
      "Test acc score: 0.6807095343680709\n",
      "Total acc score: 0.695142378559464\n",
      "\n",
      "\n",
      "\n",
      "Classification report for the test set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.70      0.62      0.66       666\n",
      "      signal       0.67      0.74      0.70       687\n",
      "\n",
      "    accuracy                           0.68      1353\n",
      "   macro avg       0.68      0.68      0.68      1353\n",
      "weighted avg       0.68      0.68      0.68      1353\n",
      "\n",
      "Classification report for the total set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.07      0.69      0.13      3265\n",
      "      signal       0.98      0.69      0.81     94643\n",
      "\n",
      "    accuracy                           0.69     97908\n",
      "   macro avg       0.53      0.69      0.47     97908\n",
      "weighted avg       0.95      0.69      0.79     97908\n",
      "\n",
      "Confusion matrix for training set\n",
      "[[ 2252  1013]\n",
      " [29436 65207]]\n",
      "3060/3060 [==============================] - 3s 850us/step\n"
     ]
    }
   ],
   "source": [
    "loss_accuracy_plotter(opt_history)\n",
    "\n",
    "opt_Y_pred_test = opt_model.predict(X_test)\n",
    "opt_Y_pred_test = opt_Y_pred_test.reshape(opt_Y_pred_test.shape[0])\n",
    "opt_Y_pred_train = opt_model.predict(X_train_val)\n",
    "opt_Y_pred_train = opt_Y_pred_train.reshape(opt_Y_pred_train.shape[0])\n",
    "opt_Y_pred_tot = opt_model.predict(scaler.transform(X)).round()\n",
    "opt_Y_pred_tot = opt_Y_pred_tot.reshape(opt_Y_pred_tot.shape[0])\n",
    "\n",
    "# accuracy score\n",
    "print(\"Test acc score:\", accuracy_score(Y_test.round(), opt_Y_pred_test.round()))\n",
    "#print(\"Train acc score:\", accuracy_score(Y_train_val.round(), opt_Y_pred_train.round()))\n",
    "print(\"Total acc score:\", accuracy_score(Y.round(), drop_Y_pred_tot.round()))\n",
    "\n",
    "print(\"\\n\\n\")\n",
    "\n",
    "# classification report for the test and total sets\n",
    "print(\"Classification report for the test set\")\n",
    "print(classification_report(Y_test.round(), opt_Y_pred_test.round(), target_names=['background', 'signal']))\n",
    "print(\"Classification report for the total set\")\n",
    "print(classification_report(Y.round(), opt_Y_pred_tot.round(), target_names=['background', 'signal']))\n",
    "\n",
    "print(\"Confusion matrix for training set\")\n",
    "print(confusion_matrix(Y.round(), opt_Y_pred_tot.round()))\n",
    "\n",
    "fpr_nn_opt, tpr_nn_opt, thresholds_nn_opt = roc_curve(Y_test, opt_Y_pred_test) # get FPRs, TPRs and thresholds for neural network\n",
    "\n",
    "df_all['dense_NN_opt'] = model.predict(scaler.transform(X)).round() # add prediction to array\n",
    "#df_all['dense_NN'] = model.predict_proba(X)[:, 1]\n",
    "#print (df_all.iloc[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3060/3060 [==============================] - 3s 971us/step\n",
      "3060/3060 [==============================] - 3s 1ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# add TF NN probability\n",
    "y_tot_prob_NN_opt = opt_model.predict(scaler.transform(X))\n",
    "y_tot_NN_opt = opt_model.predict(scaler.transform(X)).round()\n",
    "\n",
    "cumulative_events = 0 # start counter for total number of events for which output is saved\n",
    "for key in filename.keys(): # loop over samples\n",
    "    df[key]['NN_opt_output_prob'] = y_tot_prob_NN_opt[cumulative_events:cumulative_events+len(df[key])]\n",
    "    df[key]['NN_opt_output'] = y_tot_NN_opt[cumulative_events:cumulative_events+len(df[key])]\n",
    "    cumulative_events += len(df[key]) # increment counter for total number of events\n",
    "    #print(df[key]['NN_output']) # print the dataframe column NN_output\n",
    "\n",
    "# probability and significance plots\n",
    "NN_opt_output_prob = { # dictionary containing plotting parameters for the mllll histogram\n",
    "    # change plotting parameters\n",
    "    'bin_width':0.1, # width of each histogram bin\n",
    "    'num_bins':8, # number of histogram bins\n",
    "    'xrange_min':0, # minimum on x-axis\n",
    "    'xlabel':'Neural Network optimized output', # x-axis label\n",
    "}\n",
    "\n",
    "SoverB_hist_dict = {'NN_opt_output_prob':NN_opt_output_prob}\n",
    "plot_SoverB(df, SoverB_hist_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8efb2e7fa0>"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot ROC curve\n",
    "plt.plot(fpr_nn, tpr_nn, label='Neural Network') # plot neural network ROC\n",
    "plt.plot(fpr_nn_opt, tpr_nn_opt, linestyle='dashed', label='Neural Network optimized') # plot neural network optimized ROC\n",
    "plt.plot(fpr_nn_opt_drop, tpr_nn_opt_drop, linestyle='dashed', label='Neural Network dropout optimized') # plot neural network optimized and dropout ROC\n",
    "plt.plot([0, 1], [0, 1], linestyle='dotted', color='grey', label='Luck') # plot diagonal line to indicate luck\n",
    "plt.xlabel('False Positive Rate') # x-axis label\n",
    "plt.ylabel('True Positive Rate') # y-axis label\n",
    "plt.grid() # add a grid to the plot\n",
    "plt.legend() # add a legend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8eecc51a90>"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot AMS\n",
    "def AMS(tpr, fpr, b_reg): # define function to calculate AMS\n",
    "    return np.sqrt(2*((tpr+fpr+b_reg)*np.log(1+tpr/(fpr+b_reg))-tpr)) # equation for AMS\n",
    "\n",
    "ams_nn = AMS(tpr_nn, fpr_nn, 0.001) # get AMS for NN classifier\n",
    "ams_nn_opt = AMS(tpr_nn_opt, fpr_nn_opt, 0.001) # get AMS for NN-opt classifier\n",
    "ams_nn_opt_drop = AMS(tpr_nn_opt_drop, fpr_nn_opt_drop, 0.001) # get AMS for NN-opt classifier\n",
    "\n",
    "\n",
    "plt.plot(thresholds_nn, ams_nn, label=\"Neural Netwrok\")  # plot NN AMS\n",
    "plt.plot(thresholds_nn_opt, ams_nn_opt, linestyle='dashed', label=\"Neural Netwrok optimized\")  # plot NN-opt AMS\n",
    "plt.plot(thresholds_nn_opt_drop, ams_nn_opt_drop, linestyle='dashed', label=\"Neural Netwrok dropout optimized\")  # plot NN-opt AMS\n",
    "plt.xlabel(\"Threshold\")  # x-axis label\n",
    "plt.ylabel(\"AMS\")  # y-axis label\n",
    "plt.title(\"AMS with $b_r=0.001$\")  # add plot title\n",
    "plt.legend()  # add legend"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extend to the whole dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = {\n",
    "    'ttbar':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ttbar.root',\n",
    "    'WZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/WZ.root',\n",
    "    'ZZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ZZ.root',\n",
    "    'signal_charm':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/signal_charm.root'\n",
    "    }\n",
    "\n",
    "'''\n",
    "filename = {\n",
    "    'ttbar':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ttbar.root',\n",
    "    'WZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/WZ.root',\n",
    "    'ZZ':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/ZZ.root',\n",
    "    'signal_up':'/Users/mghasemi/Desktop/IPM/plotter/newtrees/signal_up.root'}\n",
    "    }\n",
    "'''\n",
    "VARS = ['jetNo','jetPTLeading', 'jetETA', 'jetPHI','bjetPT', 'bjetETA',\n",
    "        'elecPTLeading', 'elecETA', 'elecPHI',\n",
    "        'dielecCOS', 'dielecETA', 'met','testnonTopMass'] # choose which vars to use (2d)\n",
    "\n",
    "upfile = {}\n",
    "params = {}\n",
    "df = {}\n",
    "\n",
    "# define trees and branches to make DataFrame\n",
    "for treename in filename.keys():\n",
    "    upfile[treename] = uproot.open(filename[treename])\n",
    "    params[treename] = upfile[treename][treename].arrays(VARS)\n",
    "    df[treename] = ak.to_dataframe(params[treename])\n",
    "    # add isSignal variable\n",
    "    if treename.startswith(\"signal\"): \n",
    "        df[treename]['isSignal'] = np.ones(len(df[treename])) \n",
    "    else:\n",
    "        df[treename]['isSignal'] = np.zeros(len(df[treename])) \n",
    "\n",
    "NDIM = len(VARS)\n",
    "df_all = pd.concat([df[treename] for treename in filename.keys()])\n",
    "dataset = df_all.values\n",
    "X = dataset[:,0:NDIM]\n",
    "Y = dataset[:,NDIM]\n",
    "\n",
    "#X_train_val = scaler.transform(X_train_val)\n",
    "#X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 1ms/step\n",
      "170/170 [==============================] - 0s 2ms/step\n",
      "3060/3060 [==============================] - 3s 856us/step\n",
      "Total accuracy 0.667218204845365\n",
      "Classification report for the total set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.07      0.71      0.12      3265\n",
      "      signal       0.99      0.67      0.79     94643\n",
      "\n",
      "    accuracy                           0.67     97908\n",
      "   macro avg       0.53      0.69      0.46     97908\n",
      "weighted avg       0.95      0.67      0.77     97908\n",
      "\n",
      "Confusion matrix:\n",
      " [[ 2325   940]\n",
      " [31642 63001]]\n",
      "3060/3060 [==============================] - 5s 2ms/step\n"
     ]
    }
   ],
   "source": [
    "y_pred_NN_opt = opt_model.predict(X_test).round() # make predictions on the test data\n",
    "y_pred_NN_train_opt = opt_model.predict(X_train_val).round()\n",
    "y_pred_NN_tot_opt = opt_model.predict(scaler.transform(X)).round()\n",
    "\n",
    "# See how well the classifier does, Acc = correct data / tot data\n",
    "# The accuracy score is defined as the ratio of correctly identified data points to all data points.\n",
    "#print(\"Test accuracy:\", accuracy_score(Y_test, y_pred_NN_opt))\n",
    "#print(\"Train accuracy\", accuracy_score(Y_train_val, y_pred_NN_train_opt))\n",
    "print(\"Total accuracy\", accuracy_score(Y, y_pred_NN_tot_opt))\n",
    "\n",
    "#print(\"Classification report for the test set\")\n",
    "#print(classification_report(Y_test.round(), y_pred_NN_opt.round(), target_names=['background', 'signal']))\n",
    "#print(\"Classification report for the train set\")\n",
    "#print(classification_report(Y_train_val.round(), y_pred_NN_train.round(), target_names=['background', 'signal']))\n",
    "print(\"Classification report for the total set\")\n",
    "print(classification_report(Y.round(), y_pred_NN_tot_opt.round(), target_names=['background', 'signal']))\n",
    "\n",
    "print(\"Confusion matrix:\\n\", confusion_matrix(Y.round(), y_pred_NN_tot_opt.round()))\n",
    "\n",
    "fpr_nn_opt_tot, tpr_nn_opt_tot, thresholds_nn_opt_tot = roc_curve(Y, y_pred_NN_tot_opt.round()) # get FPRs, TPRs and thresholds for neural network\n",
    "\n",
    "df_all['dense_NN_opt'] = opt_model.predict(scaler.transform(X)).round() # add prediction to array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43/43 [==============================] - 0s 945us/step\n",
      "170/170 [==============================] - 0s 1ms/step\n",
      "3060/3060 [==============================] - 3s 837us/step\n",
      "Total accuracy 0.695142378559464\n",
      "Classification report for the total set\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  background       0.07      0.70      0.13      3265\n",
      "      signal       0.99      0.70      0.82     94643\n",
      "\n",
      "    accuracy                           0.70     97908\n",
      "   macro avg       0.53      0.70      0.47     97908\n",
      "weighted avg       0.95      0.70      0.79     97908\n",
      "\n",
      "Confusion matrix:\n",
      " [[ 2276   989]\n",
      " [28859 65784]]\n",
      "3060/3060 [==============================] - 3s 857us/step\n"
     ]
    }
   ],
   "source": [
    "y_pred_NN_opt_fit = drop_model.predict(X_test).round() # make predictions on the test data\n",
    "y_pred_NN_train_opt_fit = drop_model.predict(X_train_val).round()\n",
    "y_pred_NN_tot_opt_fit = drop_model.predict(scaler.transform(X)).round()\n",
    "\n",
    "# See how well the classifier does, Acc = correct data / tot data\n",
    "# The accuracy score is defined as the ratio of correctly identified data points to all data points.\n",
    "#print(\"Test accuracy:\", accuracy_score(Y_test, y_pred_NN_opt))\n",
    "#print(\"Train accuracy\", accuracy_score(Y_train_val, y_pred_NN_train_opt))\n",
    "print(\"Total accuracy\", accuracy_score(Y, y_pred_NN_tot_opt_fit))\n",
    "\n",
    "#print(\"Classification report for the test set\")\n",
    "#print(classification_report(Y_test.round(), y_pred_NN_opt.round(), target_names=['background', 'signal']))\n",
    "#print(\"Classification report for the train set\")\n",
    "#print(classification_report(Y_train_val.round(), y_pred_NN_train.round(), target_names=['background', 'signal']))\n",
    "print(\"Classification report for the total set\")\n",
    "print(classification_report(Y.round(), y_pred_NN_tot_opt_fit.round(), target_names=['background', 'signal']))\n",
    "\n",
    "print(\"Confusion matrix:\\n\", confusion_matrix(Y.round(), y_pred_NN_tot_opt_fit.round()))\n",
    "\n",
    "fpr_nn_opt_tot_fit, tpr_nn_opt_tot_fit, thresholds_nn_opt_tot_fit = roc_curve(Y, y_pred_NN_tot_opt_fit.round()) # get FPRs, TPRs and thresholds for neural network\n",
    "\n",
    "df_all['dense_NN_opt_overfit'] = opt_model.predict(scaler.transform(X)).round() # add prediction to array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3060/3060 [==============================] - 3s 1ms/step\n",
      "3060/3060 [==============================] - 3s 939us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# add TF NN probability\n",
    "y_TOT_prob_NN_opt = opt_model.predict(scaler.transform(X))\n",
    "y_TOT_NN_opt = opt_model.predict(scaler.transform(X)).round()\n",
    "\n",
    "cumulative_events = 0 # start counter for total number of events for which output is saved\n",
    "for key in filename.keys(): # loop over samples\n",
    "    df[key]['NN_opt_tot_output_prob'] = y_TOT_prob_NN_opt[cumulative_events:cumulative_events+len(df[key])]\n",
    "    df[key]['NN_opt_tot_output'] = y_TOT_NN_opt[cumulative_events:cumulative_events+len(df[key])]\n",
    "    cumulative_events += len(df[key]) # increment counter for total number of events\n",
    "    #print(df[key]['NN_output']) # print the dataframe column NN_output\n",
    "\n",
    "# probability and significance plots\n",
    "NN_opt_tot_output_prob = { # dictionary containing plotting parameters for the mllll histogram\n",
    "    # change plotting parameters\n",
    "    'bin_width':0.1, # width of each histogram bin\n",
    "    'num_bins':8, # number of histogram bins\n",
    "    'xrange_min':0, # minimum on x-axis\n",
    "    'xlabel':'Neural Network optimized output', # x-axis label\n",
    "}\n",
    "\n",
    "SoverB_hist_dict = {'NN_opt_tot_output_prob':NN_opt_tot_output_prob}\n",
    "plot_SoverB(df, SoverB_hist_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3060/3060 [==============================] - 4s 1ms/step\n",
      "3060/3060 [==============================] - 3s 850us/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# add TF NN probability\n",
    "y_TOT_prob_NN_opt_fit = drop_model.predict(scaler.transform(X))\n",
    "y_TOT_NN_opt_fit = drop_model.predict(scaler.transform(X)).round()\n",
    "\n",
    "cumulative_events = 0 # start counter for total number of events for which output is saved\n",
    "for key in filename.keys(): # loop over samples\n",
    "    df[key]['NN_opt_tot_output_prob_overfit'] = y_TOT_prob_NN_opt_fit[cumulative_events:cumulative_events+len(df[key])]\n",
    "    df[key]['NN_opt_tot_output_overfit'] = y_TOT_NN_opt_fit[cumulative_events:cumulative_events+len(df[key])]\n",
    "    cumulative_events += len(df[key]) # increment counter for total number of events\n",
    "    #print(df[key]['NN_output']) # print the dataframe column NN_output\n",
    "\n",
    "# probability and significance plots\n",
    "NN_opt_tot_output_prob_fit = { # dictionary containing plotting parameters for the mllll histogram\n",
    "    # change plotting parameters\n",
    "    'bin_width':0.1, # width of each histogram bin\n",
    "    'num_bins':8, # number of histogram bins\n",
    "    'xrange_min':0, # minimum on x-axis\n",
    "    'xlabel':'Neural Network optimized output', # x-axis label\n",
    "}\n",
    "\n",
    "SoverB_hist_dict = {'NN_opt_tot_output_prob_overfit':NN_opt_tot_output_prob_fit}\n",
    "plot_SoverB(df, SoverB_hist_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8ecad0cbb0>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot ROC curve\n",
    "plt.plot(fpr_nn_opt_tot, tpr_nn_opt_tot, label='Neural Network total') # plot Neural net-opt ROC\n",
    "plt.plot(fpr_nn_opt_tot_fit, tpr_nn_opt_tot_fit, label='Neural Network overfit total') # plot Neural net-opt w overfit ROC\n",
    "plt.plot([0, 1], [0, 1], linestyle='dotted', color='grey', label='Luck') # plot diagonal line to indicate luck\n",
    "plt.xlabel('False Positive Rate') # x-axis label\n",
    "plt.ylabel('True Positive Rate') # y-axis label\n",
    "plt.grid() # add a grid to the plot\n",
    "plt.legend() # add a legend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8eda3eba30>"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#ams_rf = AMS(tpr_rf, fpr_rf, 0.001) # get AMS for random forest classifier\n",
    "ams_rf_opt_tot = AMS(tpr_nn_opt_tot, fpr_nn_opt_tot, 0.001) # get AMS for random forest classifier\n",
    "ams_rf_opt_tot_fit = AMS(tpr_nn_opt_tot_fit, fpr_nn_opt_tot_fit, 0.001) # get AMS for random forest classifier\n",
    "\n",
    "plt.plot(thresholds_nn_opt_tot, ams_rf_opt_tot, label=\"Neural Netwrok\")  # plot Random forest AMS\n",
    "plt.plot(thresholds_nn_opt_tot_fit, ams_rf_opt_tot_fit, label=\"Neural Netwrok overfit\")  # plot Random forest AMS\n",
    "plt.xlabel(\"Threshold\")  # x-axis label\n",
    "plt.ylabel(\"AMS\")  # y-axis label\n",
    "plt.title(\"AMS with $b_r=0.001$\")  # add plot title\n",
    "plt.legend()  # add legend"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Overfit checking"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "82/82 [==============================] - 0s 2ms/step\n",
      "88/88 [==============================] - 0s 1ms/step\n",
      "21/21 [==============================] - 0s 1ms/step\n",
      "22/22 [==============================] - 0s 1ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def new_compare_train_test(clf, X_train, y_train, X_test, y_test, xlabel):\n",
    "    decisions = [] # list to hold decisions of classifier\n",
    "    for X,y in ((X_train, y_train), (X_test, y_test)): # train and test\n",
    "        d1 = clf.predict(X[y<0.5]) # background\n",
    "        d2 = clf.predict(X[y>0.5]) # signal\n",
    "        decisions += [d1, d2] # add to list of classifier decision\n",
    "\n",
    "    highest_decision = max(np.max(d) for d in decisions) # get maximum score\n",
    "    bin_edges = [] # list to hold bin edges\n",
    "    bin_edge = -0.1 # start counter for bin_edges\n",
    "    while bin_edge < highest_decision: # up to highest score\n",
    "        bin_edge += 0.1 # increment\n",
    "        bin_edges.append(bin_edge)\n",
    "\n",
    "    # Training plots\n",
    "    plt.hist(decisions[0], # background in train set\n",
    "            bins=bin_edges, # lower and upper range of the bins\n",
    "            density=True, # area under the histogram will sum to 1\n",
    "            histtype='stepfilled', # lineplot that's filled\n",
    "            color='blue', label='Background (train)', # Background (train)\n",
    "            alpha=0.5 ) # half transparency\n",
    "    plt.hist(decisions[1], # signal in train set\n",
    "            bins=bin_edges, # lower and upper range of the bins\n",
    "            density=True, # area under the histogram will sum to 1\n",
    "            histtype='stepfilled', # lineplot that's filled\n",
    "            color='orange', label='Signal (train)', # Signal (train)\n",
    "            alpha=0.5 ) # half transparency\n",
    "\n",
    "    # Testing plots\n",
    "    hist_background, bin_edges = np.histogram(decisions[2], # background test\n",
    "                                            bins=bin_edges, # number of bins in function definition\n",
    "                                            density=True ) # area under the histogram will sum to 1\n",
    "\n",
    "    scale = len(decisions[2]) / sum(hist_background) # between raw and normalised\n",
    "    err_background = np.sqrt(hist_background * scale) / scale # error on test background\n",
    "\n",
    "    width = 0.1 # histogram bin width\n",
    "    center = (bin_edges[:-1] + bin_edges[1:]) / 2 # bin centres\n",
    "\n",
    "    plt.errorbar(x=center, y=hist_background, yerr=err_background, fmt='o', # circles\n",
    "                c='blue', label='Background (test)' ) # Background (test)\n",
    "\n",
    "    hist_signal, bin_edges = np.histogram(decisions[3], # siganl test\n",
    "                                        bins=bin_edges, # number of bins in function definition\n",
    "                                        density=True ) # area under the histogram will sum to 1\n",
    "\n",
    "    scale = len(decisions[3]) / sum(hist_signal) # between raw and normalised\n",
    "    err_signal = np.sqrt(hist_signal * scale) / scale # error on test background\n",
    "\n",
    "    plt.errorbar(x=center, y=hist_signal, yerr=err_signal, fmt='*', # circles\n",
    "                c='orange', label='Signal (test)' ) # Signal (test)\n",
    "\n",
    "    plt.xlabel(xlabel) # write x-axis label\n",
    "    plt.ylabel(\"Arbitrary units\") # write y-axis label\n",
    "    plt.legend() # add legend\n",
    "\n",
    "new_compare_train_test(model, X_train_val, Y_train_val, X_test, Y_test, 'Neural Netwrok output')    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "82/82 [==============================] - 0s 2ms/step\n",
      "88/88 [==============================] - 0s 2ms/step\n",
      "21/21 [==============================] - 0s 1ms/step\n",
      "22/22 [==============================] - 0s 1ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "new_compare_train_test(opt_model, X_train_val, Y_train_val, X_test, Y_test, 'Neural Netwrok optimized output')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "82/82 [==============================] - 0s 2ms/step\n",
      "88/88 [==============================] - 0s 4ms/step\n",
      "21/21 [==============================] - 0s 1ms/step\n",
      "22/22 [==============================] - 0s 2ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "new_compare_train_test(drop_model, X_train_val, Y_train_val, X_test, Y_test, 'Neural Netwrok optimized dropoutoutput')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Baysian Optimization for NN model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import packages\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, BatchNormalization, Dropout\n",
    "from keras.optimizers import Adam, SGD, RMSprop, Adadelta, Adagrad, Adamax, Nadam, Ftrl\n",
    "from keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from math import floor\n",
    "from sklearn.metrics import make_scorer, accuracy_score\n",
    "#from bayes_opt import BayesianOptimization\n",
    "from kerastuner.tuners import BayesianOptimization\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from keras.layers import LeakyReLU\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "LeakyReLU = LeakyReLU(alpha=0.1)\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "pd.set_option(\"display.max_columns\", None)\n",
    "\n",
    "# Load dataset\n",
    "trainSet = pd.read_csv('/Users/mghasemi/Desktop/Code-factory/Python/ML/Kaggle/competitions/Titanic/input/train.csv')\n",
    "# Feature generation: training data\n",
    "train = trainSet.drop(columns=['Name', 'Ticket', 'Cabin'])\n",
    "train = train.dropna(axis=0)\n",
    "train = pd.get_dummies(train)\n",
    "# train validation split\n",
    "X_train, X_val, y_train, y_val = train_test_split(train.drop(columns=['PassengerId','Survived'], axis=0),\n",
    "                                                  train['Survived'],\n",
    "                                                  test_size=0.2, random_state=111,\n",
    "                                                  stratify=train['Survived'])                                                  \n",
    "\n",
    "# Create function\n",
    "def nn_cl_bo(neurons, activation, optimizer, learning_rate,  batch_size, epochs ):\n",
    "    optimizerL = ['SGD', 'Adam', 'RMSprop', 'Adadelta', 'Adagrad', 'Adamax', 'Nadam', 'Ftrl','SGD']\n",
    "    optimizerD= {'Adam':Adam(lr=learning_rate), 'SGD':SGD(lr=learning_rate),\n",
    "                 'RMSprop':RMSprop(lr=learning_rate), 'Adadelta':Adadelta(lr=learning_rate),\n",
    "                 'Adagrad':Adagrad(lr=learning_rate), 'Adamax':Adamax(lr=learning_rate),\n",
    "                 'Nadam':Nadam(lr=learning_rate), 'Ftrl':Ftrl(lr=learning_rate)}\n",
    "    activationL = ['relu', 'sigmoid', 'softplus', 'softsign', 'tanh', 'selu',\n",
    "                   'elu', 'exponential', LeakyReLU,'relu']\n",
    "    neurons = round(neurons)\n",
    "    activation = activationL[round(activation)]\n",
    "    batch_size = round(batch_size)\n",
    "    epochs = round(epochs)\n",
    "    def nn_cl_fun():\n",
    "        opt = Adam(lr = learning_rate)\n",
    "        nn = Sequential()\n",
    "        nn.add(Dense(neurons, input_dim=10, activation=activation))\n",
    "        nn.add(Dense(neurons, activation=activation))\n",
    "        nn.add(Dense(1, activation='sigmoid'))\n",
    "        nn.compile(loss='binary_crossentropy', optimizer=opt, metrics=['accuracy'])\n",
    "        return nn\n",
    "    es = EarlyStopping(monitor='accuracy', mode='max', verbose=0, patience=20)\n",
    "    nn = KerasClassifier(build_fn=nn_cl_fun, epochs=epochs, batch_size=batch_size,\n",
    "                         verbose=0)\n",
    "    kfold = StratifiedKFold(n_splits=5, shuffle=True, random_state=123)\n",
    "    score = cross_val_score(nn, X_train, y_train, scoring=score_acc, cv=kfold, fit_params={'callbacks':[es]}).mean()\n",
    "    return score\n",
    "\n",
    "\n",
    "# Set paramaters\n",
    "params_nn ={\n",
    "    'neurons': (10, 100),\n",
    "    'activation':(0, 9),\n",
    "    'optimizer':(0,7),\n",
    "    'learning_rate':(0.01, 1),\n",
    "    'batch_size':(200, 1000),\n",
    "    'epochs':(20, 100)\n",
    "}\n",
    "# Run Bayesian Optimization\n",
    "nn_bo = BayesianOptimization(nn_cl_bo, params_nn, random_state=111)\n",
    "nn_bo.maximize(init_points=25, n_iter=4)"
   ]
  }
 ],
 "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": "3.9.1"
  },
  "vscode": {
   "interpreter": {
    "hash": "aee8b7b246df8f9039afb4144a1f6fd8d2ca17a180786b69acc140d282b71a49"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}