{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chapter 4 - Classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[4.1 An Overview of Classification](#4.1-An-Overview-of-Classification)\n",
    "\n",
    "[4.3 Logistic regression](#4.3-Logistic-regression)\n",
    "\n",
    "[4.4 Linear Discriminant Analysis](#4.4-Linear-Discriminant-Analysis)\n",
    "> [4.4.4 Quadratic Discriminant Analysis](#4.4.4-Quadratic-Discriminant-Analysis)\n",
    "\n",
    "[4.5 A Comparison of Classification Methods](#4.5-A-Comparison-of-Classification-Methods)\n",
    "\n",
    "[4.6 Lab: Logistic Regression, LDA, QDA, and KNN](#4.6-Lab:-Logistic-Regression,-LDA,-QDA,-and-KNN)\n",
    "> [4.6.1 The Stock Market Data](#4.6.1-The-Stock-Market-Data)<br>\n",
    "> [4.6.2 Logistic regression](#4.6.2-Logistic-regression)<br>\n",
    "> [4.6.3 Linear Discriminant Analysis (LDA)](#4.6.3-Linear-Discriminant-Analysis-%28LDA%29)<br>\n",
    "> [4.6.4 Quadratic Discriminant Analysis (QDA)](#4.6.4-Quadratic-Discriminant-Analysis-%28QDA%29)<br>\n",
    "> [4.6.5 K-Nearest Neighbors](#4.6.5-K-Nearest-Neighbors)<br>\n",
    "> [4.6.6 An Application to Caravan Insurance Data](#4.6.6-An-Application-to-Caravan-Insurance-Data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import sklearn.linear_model as skl_lm\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis\n",
    "from sklearn.metrics import confusion_matrix, classification_report, precision_score, roc_curve, auc, log_loss\n",
    "from sklearn import preprocessing\n",
    "from sklearn import neighbors\n",
    "\n",
    "from scipy import stats\n",
    "\n",
    "import scikitplot as skplt\n",
    "\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.formula.api as smf\n",
    "\n",
    "from ipywidgets import widgets\n",
    "\n",
    "from classification_helper import print_classification_statistics, plot_ROC, print_OLS_error_table, plot_classification\n",
    "\n",
    "%matplotlib inline\n",
    "plt.style.use('seaborn-white')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4.1 An Overview of Classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>default</th>\n",
       "      <th>student</th>\n",
       "      <th>balance</th>\n",
       "      <th>income</th>\n",
       "      <th>default2</th>\n",
       "      <th>student2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>729.526495</td>\n",
       "      <td>44361.625074</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>817.180407</td>\n",
       "      <td>12106.134700</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>1073.549164</td>\n",
       "      <td>31767.138947</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  default student      balance        income  default2  student2\n",
       "1      No      No   729.526495  44361.625074         0         0\n",
       "2      No     Yes   817.180407  12106.134700         0         1\n",
       "3      No      No  1073.549164  31767.138947         0         0"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# In R, I exported the dataset from package 'ISLR' to an Excel file\n",
    "df_default = pd.read_excel('Data/Default.xlsx')\n",
    "\n",
    "# Note: factorize() returns two objects: a label array and an array with the unique values.\n",
    "# We are only interested in the first object. \n",
    "df_default['default2'] = df_default.default.factorize()[0]\n",
    "df_default['student2'] = df_default.student.factorize()[0]\n",
    "df_default.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1gAAAFgCAYAAACmKdhBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3XmQo/d93/n3c+Luuzk3Z4ai9HA4liXTckTq4pCirNiKJW9lXWbFdtmbtV32yhvJ62xsKnJZqvKuXI6sRKlI67WyKVVtKcXYziqWrejgqdDUQUaWZJIzeihq7hnOTN84H+C59o8HwABoAA2ggQZ6+vuqmuI08KCfAxjp+eD7+31/ShiGCCGEEEIIIYTYPnXcByCEEEIIIYQQtwoJWEIIIYQQQggxJBKwhBBCCCGEEGJIJGAJIYQQQgghxJDo4z6AYbEsKwb8BPAq4I/5cIQQQgxOAw4Az9u2XR73wQyT/H+VGBP5NyXE8HX8d3XLBCyif1zPjPsghBBCDM3bgb8d90EMmfx/lRgn+TclxPBt+nd1KwWsVwE+97nPsX///nEfixBCiAFdu3aNX/iFX4Dq/67fYuT/q8SOk39TQgxft39Xt1LA8gH279/P4cOHx30sQgghtu9WHO4j/18lxkn+TQkxfJv+XUmTCyGEEEIIIYQYEglYQgghhBBCCDEkErCEEEIIIYQQYkgkYAkhhBBCCCHEkEjAEkIIIYQQQoghkYAlhBBCCCGEEEMiAUsIIYQQQgghhkQClhBCCCGEEEIMiQQsIYQQQgghhBgSfRS/1LKsXwF+pfpjHHgjcAr4JOABX7Vt+6OWZanAp4E3AGXgV23bfsWyrHt73XYUxy+EEEIIIYQQgxhJBcu27c/atn3Ktu1TwLeBfwb8KfBPgLcBb7Ys6x7gZ4G4bdv3Ab8H/En1V/SzrRBCCCGEEEJMhJFUsGosy3oTcBJ4BPht27Z/WH38K8A7gQPAlwFs2/6mZVlvsixrCoj1su0oj30384MQNwgwVBVNVcZ9OLckucZCCHFrefLJJ3nssccGfv36+joAMzMzA/+Od73rXTz44IMDv14Mn3wuxCBGGrCADwEfBaaAbMPjOeCO6uMbDY/7/WxrWZZu27Y3guPelcIw5GreYaXkoqkKfhAynzA4mI6jKLdOCBhnuNkr11gIIUR/VldXge3dSItbj3wu9qaRBSzLsmaAu2zbfqpalco0PJ0B1oFky+MqUbjqaVsJV82u5h1KbsCJ+TSGpuL6ARc2SlzNOxzKJMZ9eNs2CeHmVr/GQgixVz344IPbqhI88sgjAHzsYx8b1iGJCSCfCzGIUXYRfAfwOIBt21mgYlnWayzLUoB3A88AzwI/DVBtbPFCP9uO8Nh3HT8IWSm5HJ1OYGjR22poKkenE6yUXPwgHPMRbl9juDm5kOHEfJqSG3A17+zI/vfCNRZCCCGEENszyoBlAWcbfv4N4HPAc8B3bNv+FvB5wLEs6+vAvwZ+e4BtBeAGAZqq1G/8awwtGkbnBsGYjmw4JiHc3OrXWAghhBBCbN/Ihgjatv2vWn7+JnBvy2MBUZhqfW3P24qIoarR3CQ/aAoArh/gByGGuruXPOsl3GiqNtJjuNWvsRBCCCGE2D65I7xFaKrCfMLgwkYJ148qKbX5QfMJY9d3umsMN412Mtzc6tdYCCGEEEJs36i7CIoddDAd52re4cxKflMTiN2uMdzUhgmOI9zcytdYCCGEEEJsnwSsW4iiKBzKJNifit+SazRNQri51a+xEEIIIYTYHglYtyBNVUY+H2kcJinc3KrXWAghhBBCbI8ELNHVOBf17UTCjRBCCCGEmFQSsERbk7CorxBCCCGEELuNBKxdaCeqSo2L+jY2lLiadziUSYxkn0IIIYTYvSzL+hXgV6o/xoE3AqeATwIe8FXbtj9qWZYKfBp4A1AGftW27Vcsy7q312137KSEGIAErF1kp6pKtUV9a+EKbi7qe2Ylz/5UfGKGC4oJ9fip6L8PPT3OoxBCCLGDbNv+LPBZAMuyPgX8B+BPgX8MnAW+aFnWPcAxIG7b9n3VUPUnwPv63FaIiSUBaxfZqarSJCzqK4QQ42BZlkF0U3gMiAF/CFwG/hr4QXWz/8u27f9kWdYfAO8h+rb9g7ZtP2dZ1p1EN5gh8CLwftu2mxfwE+IWZ1nWm4CTwCPAb9u2/cPq418B3gkcAL4MYNv2Ny3LepNlWVNArJdtd/p8hOiXLDS8S9SqSrU1oOBmVWml5OIH4dD2NQmL+opd6vFT0Z8bX4v+1H4WYvf4RWDFtu23Az8F/DvgHuATtm2fqv75T9Vv1u8H3gw8DHyq+vpPAB+uvl5BvmkXe9OHgI8CU0C24fEcMF19fKPhcb+fbS3LkgKBmGjyAd0ldrKqNCmL+gohxBj8BfCXDT97wI8DlmVZ7yOqYn0QeBvRHJEQuGhZlm5Z1mJ1269VX/sl4CeBz+/UwQsxbpZlzQB32bb9VLUqlWl4OgOsA8mWx1WicNXTtrZte6M4diGGRQLWLtFYVWoMWaOqKk3Cor5iF6rNuZI5WGKXsm07D2BZVoYoaH2YaKjgv7dt+9uWZf1L4A+IbvxWGl5a+7ZdqYauxseE2EveATwOYNt21rKsimVZryGaV/VuosrWYeBngD+vzqt6oZ9td/yMhOiTjPXaJRqrSrWhe6OsKtUW9T25kOE1M0lOLmQ4lEmMrEW7H4Q4nj/UoY5CCDEIy7KOAE8B/69t2/8R+Lxt29+uPv154Mfo/G170OYxIfYSiygg1fwG8DngOeA7tm1/i+jfkWNZ1teBfw389gDbCjGxpIK1i4yjqjTqRX1lva1blFSuxC5lWdY+4KvAb9m2/UT14a9YlvW/2rb9HNGk+28DzwJ/bFnWx4m+YVdt2162LOs7lmWdsm37aaI5XE/t/FkIMT62bf+rlp+/Cdzb8lhAFKZaX9vztkJMMglYu0itqrQ/FR/5Olg7Zbevt7UTa5IJIXbUh4BZ4Pcty/r96mP/G/BvLMuqANeAX68OZ3oG+AbRaJD3V7f9HeAzlmWZwBma53MJIYTYAyRg7UKjrirtlN283pZU3oS4Ndm2/QHgA22eekubbT8CfKTlsZeJugsKIYTYoyRgibHZzett7fbKmxBCCCGEGA1pciHGZreut7WTa5KJHSbrdgkhhBBimybzDlbsCTvdGXFYeqm8CSGEEEKIvUkClhirg+k4CUPlzEqel5ZznFnJkzDUiV5va7dW3kQXtcrVja9Ff6SSJfaos2fP8vM///OcO3du3IcihBC71p6+E5S1l8Zvp9fbGoZJqrzJZ1gIMUwf//jHKRaLfPzjHx/3oQghxK61J5tcSAe4ybPbOiOOY02yRvIZHrLaul21qpWs4yX2oLNnz3Lp0iUALl68yLlz5zh+/PiYj0oIsVs9+eSTPPbYYwO/fn09Wqd9ZmZm4N/xrne9iwcffHDg1w9qT1awGjvAnVzIcGI+TckNuJp3xn1oYpcYd+VNPsNCiGFrrVpJFUsIMU6rq6usrq6O+zAGsucqWLt57SUxecZReZPP8AhJ5UrsYbXqVc3FixfHdCRCiFvBgw8+uK3q0SOPPALAxz72sWEd0o7ZcxUs6QAnYHfPXZLPsBBiFFKpVNPP6XR6TEcihBC7256rYDV2gGu8Qd2rHeD8IMQNAgxVnfiqxzCOdTfOXWo9b/kMCyFGwfO8pp9d1x3TkQghxO625wJWYwe42kKxu2HtpWHbTUFjmMfaOHep8b2/mnc4lEmM6AwG0+285TMshBi2Bx98kC996Uv1n9/5zneO8WiEEGL32nMBC8bfAW4STFLQ2KoydTlXIl/xed1ciriuDXysu23uUrf3SD7DQohhe/jhh3nsscfwPA9d13n44YfHfUhCCLEr7cmAVesAtz8V3zXD44ZpUoLGVpWpMAy5nHU4ny0xGzd4ebVQf36QY+1l7tKktIrv5T2auM+wtDgXYlebm5vjHe94B08++ST3338/s7Oz4z4kIYTYlfZkwKrZbWsvtRp0TtKkBI2tqmhX8w5Fz8OaS3FbKr7p+X6PdTfNXer1Pdqxz7CEJyH2hDDcfY1/hBBi0kzOHaXoWRiGXMmVeGk5xw/Xi7y0nONKrtTz/zE2Bo1GOxk0ahWa2hwiuFmhWSm5VLyAlZLLsekkWvV4G5+vdQDs51gb59/Vzn3QuUuj7kI4Ce9Rzx4/Ff258bXoT+3ncR2HEGIgq6urPPvsswA888wzrK2tjfmIhBBid9rTFazdarvzpyah0cdWFZqS56OpCjFdwwtC8hWPtKljaCqqAufWiwMd63bnLu1Uc5BJeI+Am4Hlxteaf5ZKlhC3nEcffZSgusxDEAQ8+uij/OZv/uaYj0oIIXYfCVi7zLDmT427ScJWw/USulZ/PmloFF2f9bKLHwSsOS7HphMDHet259/tZHOQcb9HPauFrXGFLwmBQgzF008/XW/V7nkeTz31lAQsIYQYgASsXWZY86fG3ehjqwqNqatNz6dMHd3zOb9R5thUkiNT2wszg8xd2unmION+j4CO4Wk3rZ8mhOjNqVOnmroIPvDAA+M+JCGE2JUkYO0yw27UMM5GH1tVaCatgjOu5iCT1Iyl6xDJcVWMxl1BE+IW8fDDD/PEE08AoKqqtGkXQogBjSxgWZb1CPBewAQ+DXwN+CwQAi8C77dtO7As6w+A9wAe8EHbtp+zLOvOXrcd1fEPQ7tv+bf7zf/EzM0Zgq0qNBNRwWmwm7oQDl01tFzNlSZm/TQhxHDNzc3xzne+ky9/+cs89NBD0qZdCCEGNJKAZVnWKeAtwFuBJPDPgU8AH7Zt+2nLsv4UeJ9lWReA+4E3A0eA/wz8RJ/bTpx23/LPxXVAYdXZfnOESavsbNdWFZqtnt+p4Wq3UrgdxKSsn9aRVK6E2LaHH36YixcvSvVKCCG2YVQVrHcDLwCfB6aA/x34NaIqFsCXgJ8EbOCrtm2HwEXLsnTLshaBH+91W9u2l0Z0DgNr1wjh729kiWnqUL75n7TKzrj03dFvCEPIbrVw249JWT9NCDE6c3Nz/NEf/dG4D0MIIXa1UY1pWgDeBPwc8BvA5wC1Go4AcsA0UfjaaHhd7XGlj20nSrv1nVRFQVMUFlNmPQg1ruk06FpKmqoQ17WJClejXh+qUWOQPbmQ4cR8mpIbcDXvjGyftXB7ciHDa2aSnFzIcCiTGGqL9km1q9bmAlkXS4gBrK6u8nu/93uyBpYQQmzDqO6IVoCv2LZdsW3bBhyaw1AGWAey1b+3Ph70se1EafctvxsEGLqKoaoEDYsBN37zv9ttd/Hjfm21UHFTwBvBQriTGG5HbZgLNQshJtOjjz7K6dOnefTRR8d9KEIIsWuNKmD9LfAPLctSLMs6CKSAJ6pzswB+CngGeBZ4t2VZqmVZtxNVuZaB7/Sx7URp9y2/oaq4XoAbBKgNlY5Bv/nfiSpRv/vY6WpSL8PVRBvbDJcH03EShsqZlTwvLec4s5InYaiTNURyGIFaql9iD1pdXeWJJ54gDEMef/xxqWIJIcSARjIHy7btv7Es6x3Ac0Qh7v3AOeAzlmWZwBngL23b9i3Legb4RsN2AL/Tx7YTJ2NqnNsocnw6iaFFVSs/DFkqVJiLm6iaMtA3/33PORrAIPsYR/ODvjr6SRvvTQZtDCLz/4S4dT366KME1S+ngiDg0UcflYWGhRBiACNr027b9r9o8/D9bbb7CPCRlsde7nXbSdEYTFQFsmWPtVKFtKnjh7AvZQLKtpojtGueMYwW2Y0329cK/e9jHM0PhtHRb08tllsNl+GN/8bVufexcvqLaJqJP3/vQCF9ktbm2mQ7gbr2mhtfG+x3SIgXu9jTTz+N53kAeJ7HU089JQFLCCEGIAsND0m78HNuvYipKRyZStZv4A+kB/vmfxRVotZqlecHOJ7Pj+2b7msf41ofqu+OftWb3p2oBE6qq3Pvo3TklzgxpWAoDu58WtaxEkIAcOrUKR577DE8z0PXdR544IFxH5IQQuxKErCGoFP4OT6T5MxKvmnbXr75b1dZGUWVqDUU5ssuZ1byuEGI2cc+xrU+1KDD1UZVCYQJroo99HT0OT39xShcHXkPAAZMzjpWwzZIFWnQ6td2K19CTICHH36YJ554AgBVVWUtLCGEGJAErCEYVvjpVlkZdpWoXShMGDqGqpJ3PRKGVm/I0cs+xrk+VD/D1UY1X2w3VMXcIEDTTAylufGIrGMlhIBoDay3vvWtPPXUU7z97W9ndnZ23IckdhnLsh4B3guYwKeJ1jT9LBACLwLvt207sCzrD4D3AB7wQdu2n7Ms685et93ZsxKifxKwhmBY4WerysqwqkR+EJKveCgKTcerqdFaXVfzZaZMnYSh97yP3dL8YFTzxUZZFRsWQ1Xx5+/FnU9jNDw+setYbcd2K0j9vk4aqYhbRD4fjbrI5XJjPhKx21S7P78FeCuQBP458Angw7ZtP21Z1p8C77Ms6wLRPPs3A0eA/wz8RJ/bCjHRbqE7qvEZxvpAvazrtN0W2Y1rVV3KlVh3XC5sFJvWqlpMmJRcH3u1MNA+Jn19qFEsltvXmlxjJOtYCSG6WV1d5fnnnwfgueeekzbtol/vBl4APg/8NfA3wI8TVbEAvgQ8BLwN+Kpt26Ft2xcB3bKsxT63FWKiSQVrSLY7RK6Xykpc17ZVJWqtsqw7FS5slAA4Op3E9QMuZh2OTk9+JWpQo5gvNo4uioPq93M6sXPKOhn3XCipXIld7M/+7M82/fy7v/u7YzoasQstAEeBfwQcB75AtGZp7VvGHDANTAErDa+rPa70se3SiM5BiKGQgDUk2x0i188ww0FaZLebezQdMzg6DS8s5chWPMKQpnlDkxIKhm3Y88XG1UVxEL1+TnfDnDIhxHB9/etfb/r52WefHdORiF1qBfi+bdsVwLYsyyEa1leTAdaBbPXvrY8HfWwrxESbnDu/W8SgQ+RGPXyrXZVFURRm4iazcYPbMwlOLmQ4lEncUjfQfhDieH7TML1ayLhrLs3hdJy75tIDn3etwjMb13fV0LutPqdX8w6la9/kRPkZTi5kODGfpuQGXM07bbefGA89Hf257f7oT+1nIcSWGoeLt/tZiC38LfAPLctSLMs6CKSAJ6pzswB+CngGeBZ4t2VZqmVZtxNVuZaB7/SxrRATTSpYE2SUnfi6VVmCENKmPpFBYFDdKjBA++e+9dMohD3dkLdbQ0xV4PRyDl1Td7SL4rDVq53xIobiA8PptNhkJ4burX032o8ELLGHPPnkkzz22GND+32PPPJIX9u/613v4sEHHxza/sXuYdv231iW9Q7gOaIv8N8PnAM+Y1mWCZwB/tK2bd+yrGeAbzRsB/A7fWwrxETb0wFr0uaXjLIT37jWqhqXbl39gPbPJd/CoWJvQ2I6/f6YprKYMifmMzUI9+mfRTvwfgz9cvTA5S8AYBx+78TNKevooadvhjghRE8ymUxT98BMJtNlayE2s237X7R5+P42230E+EjLYy/3uq0Qk25PBqxJn18yyByrXoxzraqd1G2tq9PLOULg5ELm5nOv/g1HQ40zsdey/8LH0baormy1ltbBzO5esNcICvh+BVfTMRSv/vhQ5pQNswlFp9c+fiqqXrkb29+HELvMgw8+OHAFaXV1lV/+5V+u//ypT31K1sISQogB7MmAtRvWLBqF3bJW1XZ16+qHAhptnlN8NM3E1abYKtrupq6Bg9Aeepz5XIkL177J0XgR4/B7eqt2SpARYlebm5urV7He9ra3SbgSQogB7bmAtVX1YSjzS7bY/7jDzbAqZJNwLu10m29GCD4tzx1+b1SdWf+vGHNvgIceH/j3T1rXwEEdTMe5uvQ4Z6buQ0vkhlftHMaCvN2qYLW/16pXxvTg+xFiDzp48CCXLl3i13/918d9KEIIsWvtuYA1rurDpA1L3E44mrRzadVtvtlC0gRoPxct+w00bg6J63SNJmE+26jDraIoHHrL/8H+XvYz7rWnhBBDYxgGd9xxh1SvhBBiG/ZcwBpX9WFShiUOEo5ab+Yn5Vy62Wq+Wdvn7vtDUJSertG45rNNerjt2XbCV7cq2DAqZEIIIYQQ27DnAtYwqg/9Vg/GPSyxUT/hqN3N/GxcZ7Xkcndjk4gxnUs3W8036/ZcL9doXPPZuh1bp2MZpNrVV5CTUCOEEEIIUbfnAhYMXn0YtHqw08MSO91Q9xv02t3Mv7JWwAvDXdPgodt8s8bnatdMRWl7jQ5PxTm9nGcxEcPU1Z5+/7B1ev9un4rz369tsFSsYDSswXUgFePVQnmgatduqFJ2DXIS8oQQQggxJnsyYA1afRj0prPbsESvOjTRD8JtV0C2CoD9BL1ON/PHZ5I8/+o6Zc8npt8MFuNo8DCMeUit16zi+ZT9EL36+xqfr/gBLy1nuS0VG8uQvE7v31KpQsLQsOZSJAy9/rl8aTlHQtf7/rwOXHGVUCOEEEIIwe5vd7YNmqoQ17W+hvnVhhXCzZvOlZKLH4Rd9zOfMDi3XiRfjrateD5/fyOL4/mcz0Y3w1dyJcKw8+/ZSmMAPLmQ4cR8mpIb1BfXbQx6jdqFo04383Fdw1BVzm8Ucf2AIAwpuR7nNoo71uAhDEOu5KJr9sP14rauXes1u3shQ0xTuZgtNT3/utkUr51LYc2nyZY9LudKbX+fH4Q4nt/18zCodu+fH4QsFSocTMfqgdfQVI5MxVkvexyeivf9ee0liAshhBBCiPb2ZAVrENsZ5heGIWEIa06FVaeCV628LMRNXr84jalr2x6C1WvVodf5Z92qbgldJaFr/N21dYLqvlVFITkVJwzDkVd2Bq0ktla82l2zmK5xfCbB6eU8C0mTlZLL62ZTOJ7PuuNyxfVRFTi/UYJQ4fBUVMnaieYT2pMPMJ98Kxf0D9Xfv5Lr4QYBaUNHbdiPpiptq4kdP68N86cmsg29zO8SO8SyLAP4D8AxIAb8IXAa+CwQAi8C77dtO7As6w+A9wAe8EHbtp+zLOvOdtvu8GkIIYQYoz1dwepHP9WfVlfzDo4XcM/+Gf7BwVnu2TfNdMxgKm5gNlQdeqkudNJr1eFgOk7CUDmzkuel5RxnVvIkDHXT/LPGMFY758ZW55qqMBs3ef1ChvsOzfETB2ZwvLBeLRuVQSqJnSpeFd9ve82mYwYxXeX0Up6i61PwfFYclyAMOTGf5kcWp7DmUhQ9r36+W1UPh+Vg8etN798P14sEYYjRUjmshclWvXxeu733O1WlHNjjp26GMSEG84vAim3bbwd+Cvh3wCeAD1cfU4D3WZZ1D3A/8GbgYeBT1ddv2naHj18IIcSYSQWrR4N2H2xXJTF1lcOZONcKlaa5V9tpFNFr1aGf+WedmoHsS8Y4vZJvOicNZUc6CQ5SSexU8VoOK22vmVe9Xq+dT/H9lRwJXeWK69dfH71nKsemk9irBRYTsdF2iWxYZ0oBDn3rp9iPjnvqKxiqyrWCw8Ws0/S5vJR1mInpXG55fNPntcMaVgff+dRY2tBvImtsiZ33F8BfNvzsAT8OVD+EfAn4ScAGvmrbdghctCxLtyxrscO2n9+JAxdCCDEZJGD1YZDug+0CgaooaKqKqtAUCLYzBKvfANhL97tOYczx2ld+dqKTYL/D17YaOjkb63zNEobGbakYFzZKqAr1cJWveMQ1lVh1/l5pBNejaThjm+c1PLRq9bPT5/KOmQyvFsoDhaRxtaEfmAQxMSS2becBLMvKEAWtDwMfrwYpgBwwDUwBKw0vrT2utNlW7IDPfOYznD17dmz7r+37kUceGcv+77jjDn7t135tLPsWQjSTgNWHbjednTratQsEqqKgKZAte/XthjEEa1SL37aGsU4hx/F8yl7U6nxU+g2SW1W8FpIGq47b8ZodTMcJwhLnN0rcKDhoqkpcU0kaWj3UJXRtaHOW2s7levOXorlcTzwQbdQSGrp9LrcMSVusYbWTbejbGucaWxLS9izLso4QVZ0+bdv2f7Qs648bns4A60C2+vfWx4M2j4kdcPbsWX74/e9xaHq0Q9U7SavRLZXz6rd2fN9XNnZ4dIEQoisJWANovOncqrlBp0CwVKyQMTVeXi0MLQztVNWh9Zx0VeFitsSVnENMV/n+an7oDR4a9RIkG9e16tYiPwhhfyre8ZopisKRqSSECmuOy7HpJLGGpiTzCQNTV7e9eHVN1wYeW7y2UxgaWkia5MAx6iA2yecuhsqyrH3AV4Hfsm37ierD37Es65Rt208Tzct6CngF+GPLsj4OHAZU27aXLctqt63YIYemHT7w9nPjPowd98lnjo/7EIQQDSRgbVMvHe06BYI7Z1MEIUMPQ9u9oe5lfanGcyq6PjFN5e6FNNMxAy8IR7oobbcg2S7wKoScXy9ybCaJoalUPJ8XlnJ4QcD5bKmnjn+Hp6LztTsE4mFUD7fsBPngU6MbpjfpwWEclavG4YZr34XZN+78MUz6+3Jr+hAwC/y+ZVm/X33sA8C/tSzLBM4Af2nbtm9Z1jPAN4gaRr2/uu3vAJ9p3HZHj14IIcTY7emAtd2Fav0gZKlYwZpLNTWqaG1u0C0QaArjHYLVoJ9W47VzWkzEeGk5y90L6YZ1mHam4UW7INku8J7fKOJ4fj38bDguU6bO6xczPbfI36o6OIzq4XaWAhipUc9vGubvG0UgWfsuuBvR+UvwueXZtv0BokDV6v42234E+EjLYy+321YIIcTesScDVi1ILBUraIqCH4YsJs2+hrSFYcilbHTjXvICil5Qn5vT6YZ47PNZtjDI+lIBIaau1cNVTa+hYFPI7XIDu1Ug7lQBOjad5MxKnrvm0rhhwA/8AKuhI6AfhhyeivPyamHLQLjVe7id93hi158aZfVmHNWhrTQON6wdWy1cjpo06xBCCCF2vT0ZsK7kHNYdlwPpGIaq4gYBS4UKYQiHp3ob0nY171DxQ+KaRsaMFnnNVzyKro9ZvXEf1g3xIJW2fl/T60LFrQYNBR2rZSgohE3Hryr0VFnbqgIUEKIp0fO6qnAlV2r6nUXXp+R56Ko6lq55gy4F0LM+btbr1x8dbfaN0WtGUbmatOpQ6zGM6twnya18bkLsS7epAAAgAElEQVQIIcQY7LmA5QfRorN3zqbImAYBISo6uqLwylqBA+l4/Ya7U0BpDCM3iuX6DXHa1FkuldnIe0O5Ie5nyF7ra5aLlWiJyxAWeqjODTo8rZ9Q0HhNrxVaqmWXvsiFjSRXvFkUQlZOfxFNM/Hn70UhJK5rnJhPo6kKZc/nUs7ZVFnrNez5QcjFbAnXD+v7d1yPF5ZzfO96lpmE2dO1HoVRdYLsVf0zd+GJ6PrP/iLzS3/FwccfQBlmpakxXNVMaiVrHPvsFnokEAkhhBATbc8FLMfzCcKQohtwOZerh5C5uIkfhpRcj42K1xRqZuM6CwkTU9MawkYURlpviNdKFQ5l4kO5IR5kyN6VXInrhUq9UuN6AdfyZcIw5PBUsuO+tjM8batQ0BoUXT+g6Pq8af/0zWqZ4nM0XuTvDv0Gs2qBE1MKhuLgzKX4u2sbHJ6KU/EDnEqAAkyZGufWSywkzPrwxF7D3mxc51LW4Z590/VhgrmKz9GpBFdzZe6aSxOE22/UMUjlsT6X61s/g6umME79l+FVrnoYdlb/zE1H19/VfC7Ef4mrG9/mEOFwb+obh94Z0zerRePQeo3+Yib6789VO2zfimFGhiMKIYQQI7HnAhYKFF2f6wUnGgpWDSHXCw4l1+dGsUwQKpyYT9fbj1/cKPFqvkxMU1lImuxLxprCSK25Qcn18PyA26eS2656DDJkL6rMOCwkTI5XO+a5fsC59SIXsw4H0omON+vbGZ62VYOH1qBYcj2+v1pgqVS5GV4OvxctDAncH3JQ/yHGkXdHj3s+UzGdshei6CEzMb3aeVElpMRzr65zx0yyXmnqpQK0kDC5UaxQ8HyKnk8YhgSELCRMlkoubhAQ17WBG3UMUnnc9H7goQUbsIPDFJs+c3PvAcC4/AWOhhpnFn+X/QuZtoseD6TdPCe5sb+pW+VKApEQQggx0fZcwDIUFT+EpKlxx3QKXVPw/JCzGwVWnQrZssfJxSkMTeVytshysULS0FAUhSAM69Wg1jAShCHXChUWkuZQ5u4MMmTP8aPqXC1c1bY/PpPkRrGM4/uk1M5v+XaHp7Vr8NAuKMZ0jYPpGFdzDjNxg3i1Mlj2/GjhXqVy83xVFS8IcYOQBUMjCCEIo7lUSV0jDEMKFb9eaeqlm5+paRiqStqI9huGkKt4BCFN1bpBu/cNUnmsG8VNdI9rRHX8zCn+aLsY7mS46nQNaj/XKle1oYu3cogZ5yLOQgghxC1szwUsNwyIaSqLCZP1sksQhqiKwmLC5EYhmrdUGzbWWA3ywxAFuLBR4mLW4R8cmOF6sTyyuTIDDdkLQVcV1JYqiVpd8Jiw+z5HsVBxu5t2heg4C67PD1YLqIrCdEzH8XzU+DzBgffUKyWaqpAxNK4XHG5LmoTV11/KOiymTNbLHgfTMc5ulJoqTd26+dWqdZeyDkenE/XwcG692FStG6R7nx+ErFx4IhpiV6sCVSuPLy3nmgLltozgprjtZ+7we6PrsJLvr2lLr8cnN/W9k0AkhBBC7AojC1iWZX0HqM1gPwf838AnAQ/4qm3bH7UsSwU+DbwBKAO/atv2K5Zl3dvrtn0fWBiNulp1XIquX68wJA0NVQE/iG6sK0FQrwZpqoLvh5gN1aByEAw9jDQaZMheXNdQgHWnwkzcrAe/daeCWn2+130Pq1LR7qa96PoQQlxXOZSOERByJVfGUBVun4pvOmcvCClUPE4v51CUKEDOJwwWEyYrJZeEofddYTmYjnM5V+KFpRyGplCsRN0Dj01HFaZBu/e5QYCmmRiKU38sDEMqfkDJ83llrQAhTMd0DmcS6C3VopHeRG/xu0bexXCculUGG/9em3O1l0LMXjhHIYQQYgeNJGBZlhUHsG37VMNj3wX+MXAW+KJlWfcAx4C4bdv3VUPVnwDvA/60j237Yqgqjh8QhnByIVO/iTy7XqTsBxydjm4wF5MmuqqgEN1sa4pSvbmnqRo0yrWt+h2yp6lRBerVfBkvDJta0B/KdJ5/NUqtN+2aqpB3PTYcl9m4iaIo6IrK/pTJ+fUSr51NseK4m875NWaavOuRMjT2JeOEhJzbKDIbMwjCsK9KUxiGXM46LJcqmGr0/h9Ix1EV+P5qYfCK5OOnMNDxZ38RV/MxLn8BgOJtP43j+cRVlbmYyUqpwlKxwvVCmcNTif46FQ5zCGGb1267i+Egx7eXwswwyHUSQgghJtqoKlhvAJKWZX21uo+PADHbtn8IYFnWV4B3AgeALwPYtv1Ny7LeZFnWVK/bDnJgbhhVm+aTBijghyEoMJ80uFYoM58w2Sh7nN8o4ngBy6Uys/EobA1SDdqOQYbsHcrEURR4NV/etIjyuDTetCvARtnlQDrOfNwgbepcL5ZZKrqECnznepbbpxPcPZ+uh8RonlTUOOJStsSlrIMXRBXFoho1JtmX6m3uWxiGvLiUxQuia6WpKpoCS9W5dicXMk3XOqq++T1XKDU85pf+igvxX+JoqKEpYTVQeuiaQtkPOLmYQVWUqKV/2eNq3tn8Ho/pJnoUw0S3rVMA6yeYtasMPn4q+tMuDEqIEUIIIcSARhWwisDHgX8PvBb4ErDe8HwOuAOY4uYwQgC/+li2l20ty9Jt2/b6OrIQYrpKXNPIlT0UJbp5j2saMU2l4ofsT8XZn4pzcaPI9UIFNwhJ6Bp+GG6rGjRI227or0o2iTfIQQjzCZPFRAw3CHh5NU/a1MjEjPp6WHfNRRUqPwjYqPhcL5abGkLUzssPQlZLLvvSJjFNq1fooplZndWu/fWCQ67ic8++aeKGhh+E5Csei0mTi9ko6MSrzTNaFyLeshNg9ab84OMPcHXj25xZ/N0oUK4WODqdoOT59aF3EFVTj6Tj/P1SluViBb0696/rfoYxhLCHKlP9M9fvflq7A3Z7rXTFE0IIIcQtaFQB62XgFdu2Q+Bly7I2gLmG5zNEgStZ/XuNShSuMr1s23e4Iqo8qUDFD5iOG9GDISwVy5T9gCt5h8s5h7m4ga4qOJ7P5ZyPF4SowNHpBIcy/VWDhtG2u1+jHLrYq07nPRMzuJovMx0zWCm5vG42RdkPSOgaZV/hSMbgB2vFTe3R/SBkvexxcjGDpir1BiVzcZMzK/n6ItG1bd0gQFcUrhfLrJRcVAVWShVShkZMj0KOpiqkTZ31cvR8bR7X5axD0fOw5lLEdK17J8CWYKAQcqj4LPsXMjjVeVezMYNsxa+HKz8ICQGfEEVReM1MknTM6K/jYL9qx7lbdApgNYMEs8ZtpGmEEEIIIUZgVAHrnwKvB/4Xy7IOEoWjgmVZryGaV/Vu4KPAYeBngD+vzqt6wbbtrGVZlV62HeTAWucp6apK0fVZLlY4Pp3g6EwK1w+wV/KU/YB79s/UW4hfyjko1blY/dhW2+5drNN5mxqUXJ8Xb+TwwpCC62FqKjEtmh8X0xsXdL4ZEls7Eta6JaqaUt9eVdSmULfhuEyZOnfNpVBVheVimbPrJTbKLjNxE6Aa/gJcPwpkl7JFzmdLWHMp8m4UrpNGH+tiVW/UNSBl6iwkTa7my7hetI8whPWyS1xTKbg+pqqSMKJ/ilutdda6j4H0EizahZte16uqvdbdiF7fa2v07ZzT2nej/UhIEkIIIcSYjSpg/T/AZy3L+luidhD/FAiAzxHdd37Vtu1vWZb1PPAuy7K+TjTG63+qvv43+ti2b7UK1Nm1IkEYUvYD9qVizCdMwjBEUxUWUyav5sqo1a51CUPn+HSy74VnB10weKeH9w17n1ud95FMgnzFZa3s4QYBiqJQKFUwNAWvQ3v0XlrXN4Y6VVF44UaWA+lYNMxTUzE0jQOpGOfWS9y9oBHTNRzX50quzEIixvVimXzFZzZucFsqXh9CWHR9UmZLt8Ieh7jV5qAtFct86+oapqYS11V8P6Tg+Rybbh5yOuj6Wx1tVQnqcuwT4bb7o/92m4O1nercJJ6zEEIIIXatkQQs27YrwD9p89S9LdsFRGGq9fXf7HXbQYRhyKVskYofktBVvEAhZWh4QUjR9YnpKoaqYuhq003uIDe+/SwYPI6hhP3us9cgttV5LyQN1ssVgjAKYrdP6aQMg3zF4+XVQtu24Fu1EQeaQp3j+Ri6ykw8WvMsYWjENRUtbnC9WOb0cg5DU8mWPTKmxoF0jNMreV43l+Ll1UI9yNWGEOrVhZD7Wg+K5vlj62WX25ImpqZR8aOqqB80L1C21fpbQwvD3YJFu7lUW1Wk2r2223aNla7Gn/tRq1zJPC4hhJgIE7tMjxA7aM8tNAzw4lIOXVN5420ZYobGjYLDuuOxjksmZhDXo+YJrhc03eQOsvBsPwsGD2MoYb83373us9cgVtu/itJ03rXHa9toikqIwpv2T3OjWOHsehFdU3G9gKLnc9d8uu3xdmsjXvabQ13t2gfVRaKDMBrqtxEEUUXLUKj4IUem4hzOJOqvj+vapiDnBwHnN8rNwa+POTy1+WMn5hvnj0WftdPLefalo+Ya3dadGjiAdzvOSWw00Rjo4GYFq9V2K1dCCCGGapKX6RFiJ+25gFXxAtbKHifmUyTM6PQzpoGuqlzYKJE2dSq+z1Khgh9GN+caStsb317CTLeqy2xcr4aOKBD0O5Sw0SA33/0MX9wqiLXbv0LIubUCMUNjo+yhKtSrRX5YDTOGzu3TOocyN6/l91fz0fy4NsfcrUtia5itXftz60Wm4zqqouAFIctFl2NTSRZTZsfXNwY5VYE1J3pNL+3u230uOs0fm44ZxHSV08t5YnpzF8FWPYXhYQak2u8adPHdrbZrDX417kbrlr3/DqlcCbFnra2tsbIR55PPHB/3oey4yxtx5uNr4z4MmOBleoTYSXsuYJU8n5imoFVvpjVVIWlEQ/Qc3+fMSp6MqbOQMIGwbaWkFiaWi5VoNlgIC9W1ptqFmdaqi+cHqAoUXMhWomFnGVNDVehpKGE7g1S/eh2+2EsQq7Vbb9z/+Y0iy8UycU/nUCaGpqoczkRrTi0X3U1hSFO1rlXC1uDSej3ahdnFhMm1fI41p8Jyye0aPFtfvz8VJ23oXM6VODad4MhUhypi9aa+W8jtVMn0qud611yagBAVhYCQIASt4fD6ncvXNvy3Cx+THFCM6ei/PRyTj46rpjCq/6aFEEKMxeQu0yPEDtpzASuha7jV6kqu4pE2oo51KuD5IYfScY7Ppuo3aQfSm29UL2ej9bE0JQonrhdwLV8mDEMOTyU37bO16rJUqFD2g6aK1rmNItmy19NQwla1m29rLoWiREPheql+9Tp8casg5vh+2/0fmUpwrVDmxHwaXVPqDUNShs6ZlTyzcb3jfKrG4+2nOtduCOH+dIx9yVjTwsWdHEzHuZIr8fyr6wRhiK4qKESVpjAMuw7F2yrkdps/pipwNeewUfbaroe1ZRh++mfRgg3CG/+Nq3PvY+X0F9E0E3/+3v7n8e30sMFOzSt6UP9s/Oj/F73fy7mRz1sUQkym2dlZEs7LfODt58Z9KDvuk88cJz47O+7DgAlepkeInbTnApapq0ybOufXixzKxKn4AX4QcCnnYKgKr5lLNd2YtVZK/CDkYtZhIWFyfCZ5MyCtF7mYdTiQ7rwIcfS4ylp5cyXi+HSStVKFc+vFpt/bOpSw3e+u+D5uEJB3/VpBjbimkjTatztvPJ5uN/2dht7V1IJYWK2WtO5fVaLX6pqC3hAQ640uEiarjtu2Sli71m4QRGuUeWFP1bluQwh7+bDX2vAvJE2OZOLE9KiCt1U1sJcKU7vwNxePgtu3r62jKgqGqjIT01lMRAsf1/a5ZRgOCgBcnXsfpSO/xIkpBUNxcMvPcGEjydX999YbbbQd1jpJlauaHo5pVy2BMIlVQiGEGK6JXaZHiJ205wIWQBgGbJQ98m4BU1Mpe1HImk8Y9XDV6UbU8X2CMKyHIKgGpJkkN4plHN8npXa+rN0qEWlTx9SUrkMJ2307v1yqUPYCUrpG3NDqrcU3yu6W1a9uTSNqtgpia2W37f79MMANogCgNxxCLRSYmsahjL4pDIVhyJVcqb448JrjcvtUAr0W+HqoztXCbLdg2k67oKRqypb762W4ZVzXNoW/awWHghtwbCbJQiJGEEZhbonKpn12DcMPPR4d++kvRuHqyHuiY7/01+yPVXilUCYIYK3cpgr4xAPRwdZu/Mc9bLDH/Q+yBIIQQoiRmuhleoTYKXsuYFW8gKwbcN+hObwwZN1xmYkb6IrC89fWKbs+y06l83C0EHRVqTcpqFGVaLFbwg47rupaiQjhSHWIYaehhK3fzvtByJrjcSgT53LOqW8b01TOrORZSJhdbzK7VXwadQpi+5JRa/OD6RgXcyWOTSUwdY2YpnJ6pUBK17iQLXF8OtmxQtZaJWysSgRhyFKxTNENms6729y07bS7d7yoCtd6DbaaC7fpfb38heh9PPCPNoXc2vnWAsJrZ5OUvCigaWwOc7V9bhWG3SBA00wMxSG89AWuBvOs+EfQVJ3V3BpqKcddR05iNnQrvJp3ONT1ivRhhwNZP0sgjNUkdmoUQogRmPRleoTYKXsuYJU8Hw34wVqedcfD0BTOrReZietowMVsCUVROg45iusaCrDuVJiJm/Ub3XWnggrE9e43dL0Oy+s0lLD9zbfC7VOJTTffZT9gIWn0dF3aNY1o1CmIldxooeC0GWOlVOF7S1liqkrR8ym6PvNxg+VihbVShbSp44d07JIHN6sSd82luFEss1ysEBKFpuVShX3JWNTOvcPctGgIZxE3aD+ksFOQbGxckqt4LBXLpA2dpKFFc6C2mAu36X0F3FDr2HIdbgaEmK5R9IJ605VaQCi5XtM+FUVhfyrOTMyoN8SI61o9NBqqij9/L+58mhuXv0HJPMiJeIkQlUI+zsHqgssm1c9S4WnObITsX/oGWljZfOM/rgDQYyDpZwkEIYQQQoidsucCVkLXyFU8kqbOPfuniOkaZc/nB2tFshWPhKHzI4uZrqHmUCbBq/lyvWlCrdp0KNM8/6rTMMN2lYjZuM5c3KjfZPfSWEILlPp6U14QNgUggJdXC5ha98DX77pZrUGsNjwxbegsJGNUvIBX1qJOjMdnkiwmY/hByLn1IqamcGQq2XU/tfNeKlUouQF3L0TzWmvv0aVcicOZxKbg0hqQXjuXouIH6NVrePtUnO9c32C5WGnbRKJWNbt7IcONYpkNx0OvBhdTU7sGpZqD6ThXv/Fhzkzdh+bl8f0K8+X/k4PFr8NDT23aXkWh7AVU/IC4ppKveKRNnSAMcb2Aq/mb6241np8XRgtiG6pKQlfrHSwb29IXYz/CycUMaukSKxWdeCzJ7PRUtOByGEbzvRQfTYvhalNo3vKW731Hw6jQDPCa3r+sGLNxD7kUQgghxI7acwELoiF+x2YSJHQdVVVQ9ejn5WIZTd26VfqhTBxFgVfzZTRFwQ9DFqs3ubD1ELXGalDF91kuuqw5btM8q33VYNL67XzF89lwXH7QsICvQsjZtQKHMnESho6mKFveZG5nGF3NpuGJ+hKaAmV/hgPpGCk9WntK1RSOzyQ5s5Lf8ncaqornR4H1ZDXohmFIoEVB4ny12+JcwmgKpLWA9JqZJBdyDguJGPmKR9H1SZl61DlSUXjNTJJ0zNhU1Wqcy1MLwJeyDiXfJ6FrTe9vJ4qicKj4LPuL38Jd/R6Gn0VbvK/rtQ8J+btrGxxKx5hPmCyXylzNlyl5Poupm/usnd/R6QQKCjFN5XLOwdAUSg3DJw+m41zYKLLq+OQqHmEFEuUrhKrVtOCyqijR8EUzjzH3BsCbnBv/PgLJpi8rrn2N+ew3OHjfH476KIUQQggh2tpzASvvesR0jYSmUfaDaOUFIKFpxHWNshtsOeRoq3lLvXY201SF1YJL2d+87fViuf7t/JGpeD0EvbiUY8rUsarbVzyfl1cLXCs4rJVdvCC6eT6cjjcFkNZKVesxOp7PufUiQViqzwPbyubhiQ6hqlPWAhRFJd7Q2aLXeTGaqjAd01kqVurz3IIQHC9gNmawHtNJ6TprjlcPpLNxndWSy90LGVQlOtcgDEmbOutll3gQknc9TFUlYej146lVJmdiRlO1sPH9fXE5y52zKVJGj/9UHno6OuCnfxaCAjz0+KZNGq+9ripslF3OrZd4tVAhrkddBF+/kEGvHk9jG/686zMT09HUm3O1XjeX4uXVQr3CemQqyXrZI6GrxGZvR1WOspArNS243FTpYZvdbrdTodlm9WvTv8Xv/pvofCaxRfukBFghhBBCjNSeC1iGplDxA8IQYrpaX9uoUh2qtS8VazvkaDZmbOpI127eUj+dzbba9sRciu+X8jz36jqGqlLxfRRF4fUNQxjdIORAOoYXhNy1kMZxfS5kS1zOO+Q8v96JMAipD42bjRmsOhXuXsigq0pTx77zGyUIFQ5PbV3Jqs2B8S7/Vw4pPvsr13FCk5dj90Eph5q8nSCMwo4fhD3PizmcSXC9UGa5VMZQ1Xrbd0NVKLo+MU1rCqSvrBWi4ZrVa9I4bEwBSq7H1XyZ21ImEDWyMFS1HvpQaFstjCo+CvEthlnWhGHI5awTHfeB3yLwy8znShxMxwnCKJCqKJve85m4yd0LGqeXc9w9n8HUm69RLchq1TW56u3ztebPYi28aqrCYtLkSq7M0ekEqqZ0XXB5Ym/8+xku+OQDaCCNJIQQQggxdnsuYCV1HUUBezWPNZcmbmg4ro+9mkdR4Ph0kuvFcvtW6dWmA92G0vXT2Wyrba/kHRK6zmvn0gCUvYCXV/P1RgVBGOL4AdMxgx/4Rc4s59DUWoAMsWZTXMk7rDsuR6cTzMRNXD/g7HqBkheFiSu5UlMl60bBYc1xe15HaMrUOLee4niigKEEmHgEbo6lbIFY+gB+CH4QcCVXRlehl2kxuqZyeCrBRtlrWovq3EYRP6Rti/znX12n7PnE9Jvd9l5aylHyfeKaRtH18GI6L1WvkR+ETMd0PD8grmnbnssThiEv3MjihSGHM3G06fvRFLhRKPPiUpaQKCCVvYCQsN5yviama5i6RtCmDWUtyPpBSAg35+lVK6tUH2sMr9tdcHkgg4QZmZ8khBBCiFvMngtYmqpw50yS8xsOz726jqmpVPwABbhzJomuqU1DjmqL3HZrld6on85m3bZ1/YANP2q6UA8TqoqhquRdj4Sh1efUXMk7xDSVE/NpiAoynF7JcyFbouD6WHNpCl60fpehqRybTnKjuE7B9ZqqKdGNe/S83TDkrFXT/C0F8qkfZTmEKe95At9l3+0nWCu5nF7OMxXTCUKYSxhUvKDtdWvXaKMWEH6wVqwHhIypMRXTNwXSuK5hqCrnN4rcMZPC0FRuS8YoVHxm4jpHppKcXs6yXvY2hWpViT4TvawH1kkYhvz9jQ02yj5v3JfB1KJOk44XNRvxArh7IUVM13A8n7+7tsFG2WUmbm56z/1qta/xuteaOVzKOiwkDfIVrz4HazqmcznrbAqC211weVD9Nk0ZmmEGNQl7QgghhNiGPRewAA5PJVFVlaVCGR+IaxqLqdimxXVBZc3xmoZzRV0EY00BxA9CHM8Hhb6qId26oE3HdIpec/DSVIXFlMnVfJkpU486IPo+l3MOJxfSmLpa7ZyncjAd40rOwdBU4oZGsRqwVCVqC540NM6uFaJuctVwla94xDWVmK51nS/Vbo7ZuY0ixuXnuT3/NNz9HlYdj3v2TwPUb7ZdP2gaJrlVo41DmQSLiRglL2o0oakKLy3n2gbShK6S1PW2ASkIoyGSR6cTFDyfoucTEv18YaNUDzRtA0kPN9uXcyWyFZ/ZuMFUzIy6APoBhqawUfE4OnWzu2S02HCcc+sl7l7QiOlafR5d0fU5Xz2e1ippLQBe2Ch17CLYzlbt97vpJyxtu2nKuFvCS5gSQgghxJDsyYDV6+K6jUP4wupNrVOtdpU8n4sbBXRV5WreIYCowQRw+1SCuK70VA3ZagHf1jCxmDA5v17kzEoeU1PJOi6KopAx9XqHuFzFI23o1epciONGgaLWNML1A3RFYSqmczHrcKPgoKkqcU0laWhd1xHqOG9sKsHpY7/FoYUPEVSvW+uaYKqiEBLi+D4pVW8KatHwOZ9LOYeread+XVpv2GdMnbPrBY5NJ4k1LJi7kDSjLnrB5vfU9X10TWUmbtbnhKlKtFj0lXy5KUj2G0j8IGSp6DId0wmr19bQovldTjlq9V7bV02tKciLSzliukqh4qGrKm/aP715EeBqta/1M6uiEDCaoX6NLeFRgJB6iOsUlnpt7NKzQYPPMCpXW83jklAmhBBCiC72ZMDqpPUb+8YhfBU/wA1CpqvrFCU0rX7zf3wmGS2iG4RczJa4knM4PJXg5EJmywpAu0pNrclBa3WrVulQqlWnih9yMBPjWqHCcqlSX5NLrTZmCEKYjevYq/mo2UFD97haIFFRWXPcTYFlq8Vxa+GqMXg6XsD3bmS5LWni+Te7MdZu2JcKFRzf5werBeYTBqvVoFb2fUrlAFWJOgieWy/ieD5+SP2GvXbu2bKLqqrcKK6TNDR0RWmq4LQLSK1DMRuDZsfGGy032/7jD+GqKYxT/6XpurhBVKkKQ5iO6U3vV+iXyDoltOlkfZ9hGH1GvCAkZSg4XoAXhPzYvgxmNZB2aorS6fw6fX4HdSVX4nqhglb9nLlewLV8mTAMOdymw6T/+EOs7P8gJ0789JaNXSZGuzC19l2YfePgv+uhpyV8CSGEEGJvBqz6DX8xuomMhog1d9qrVZxqC7emTI3ZuEGpWmVJmRr5SshcwiRjGsT06Mb9zlmVv7+R5XK2yIF0fFMVp9OxtBta1VrdypfdtpWOlKHVG0LMaAbZisuZlQLlarMLVYELGyWu5MubqmmHp6J92KuFnuYetYaVouvj+gFZx8ULQ9K6ypW8g0LI+fUix2aS3CiWKVR8jmZrM60AACAASURBVExF18PU1KjzXxCQcz1cP8SsBrFY9fpf3HC4ezFdXyy4EgTsS5m4QcCP3jaNGwScWy+SNrUtKyStQzFVRal3FtyqiUWIwtW597Gy/39G00z85VzT0DdDVevhquIFmLrKmZU8CrBWLGBsvMBS5i3E9Wjo5aVsiXXH5eRCmpm4Sdnz+f5qgaVS5eZ5XP4CBqDFH9iyrf1Wn6FaxanX8BV9SeCwkDDrzURcP7rWF7MOB9KJTa931RSaZvbU2GVLnapINaMMLrNv7B6Stjo2IYQQQgj2aMC6kos66+1LmShEXfdWSi5zcZ2jM6mm4U0H03G+d32Dq3mHhaQZVYViBjFNIV/xMDWVWENVJKZrJAyNdSfA8aJFbrvZamhVbViY4/m8stbS9KJaJTi9nCOhaU0NIRYSJgtJA1PT2q6DVROEMJ8wWUzEOg45a31t4/pcTjVcZSser1/MMBM3cVyf0ys5Sp7P6eUcuYrHa+dSxKtzvxQlqvp94/IqFT9gMRFDr577ulOJ2rIbKvMJk0LFZ7VUoRJE4cvxAy5lixydTnLnbIozK3kOpqNOet0CxMF0nCu5Es+/uk4QhmiKgqJA0kjUW/U3qd5cX/36v6S0+BAn7ri/7ftTux7Fio+pq2yUPXBusO4qTL3658zn/zuXjBQ31m5Djc3h+AGvm0sDCmuOS0xTOJgyeTVfaar0uKHWc1v7bp+hTkMtOw33c57+Hwlu/yOOH3jNpk6NN4rl+vBOoB4wjKVv4E//D7iXvoih+HD4vdE5dKsQjtswmmKsfTf6r7sR/fcvZm7+XSpZQgghxJ615wKWH4RczhaZiZtcL1TQVYWSGzUoWC5WODyVbBretJiIoSgK06bOsakECSNa5LXiBfxwoxi1RCckmqwS3VR6fkgYhrWHuh5Lt3WwWkOPXp3b08jQVHRNZTFlcjDTeU5Z69CyblWPrbY5kIrxaqHMS0s5yr6PF8Drb8swHTMAiBsahzNxruQc7pxNcX6jxGIy1jQPyVBVAmC15LKQiBGGIV4QsOq4KETz2dzaGl7ATExHV1USmkbFD+sBR1XgYrZIruJ3DRDR36O5arXg6QUBS4UKAIenNlfB/CBkZeo+TsSLXYe+NYYYRQE3+wrHss+i5F/COfKL/ERmHZVVXk7cz7rjsZg0iTmX8FHIKwfRNZVKEFC6/FXSahm3uMQFd5b5q7+Jtv5F+Ln1+vG0vr9bfYaCgLYLWXeeGxW1kFdbwpeqRG3ma13k/SDEVacxggJaWGF+6a+4MPXPOBovYkDfbe7rWoNPzU6ub9Xud7cbBlhTOzYhhBBCCPZgwHK8aL6QqircPRM1Vyi6PtcKZRw/oOz7JFW9Pryp5FUbJCR1rhUqHJ3W0IiqH34Qcj3vkDI0ZqsNFM6tF/GCoNo0ovuwqE7rYOnVYVUvLkdNEGqLAzfOa6r/joYqQT8NGnppStB5mzKKEq1p5Xg+mZhB0ri531q799p8p9raTap280a75HokdBVNiToD6tUug7elYlT8gI2yx6Wsw3RcJ2PquEHIpVyRxZTJbckYZ1byzCdMsuWoirhVgPCDkCu5EnfOppiJm/UwpisKr6wVOJDePE/IDQK0/fdjLGSammO0Dn3b1DRl/t3Au3np9Bd5XUbFP/BuQhQqKzni1U6P/z97bx5kWXqWd/6+76x3z7UqK7Oqq3qr05tkTSPQwqKW1AMjpBDY4ZhwOGDM2CFjYzPGMTNgMDH2OGAYvAXEMBgGw2BkxuMZthZIGE8DElhqJKy1u6r69FJr1prb3e/Zv/njnHvzZua9mTerKqs6VecXoajMrHPP+c69p9Tfk+/7Po+JQENRNnVWux6JUrzZjNA1i7irMbvycRZbnwZ2F8O7ZakJAas9n2fmq3vPRmWiwV55CTH/t6g3ClT0iKRwHImgFYRIwMqy09Z6Idrb/226ljd+mmPdl7i+8G7O9UK01da+bO73xe3OSY3jToTaqCpYXrnKycnJycl54HngBFaCIlGwWLJIgCRRSClYKKW25n36wqWQhdzOF0xWCAbzUGGUuglamuTMSgtTk8RZLpWlSR6q7j3YPy4Hq+GH+FHCsws17KFZq/4s1e2G4fbZq+qxUEo3xeOO+dKNOlOWwXTBxIti2kHMajegZOqUDY12EKOJtP3Q1kfb1l9r+2hCsFC2UApMTYAQSJXe40NVm1jB6+sdyqZOkqi0+mSbxEoBivMbHbQsHHovAeFl1ux9cQVpVW/KNknojGznNKQkytoW46wgGau0OhlG8Y7Wt2GB2wsjuvo8ryU1tHqXIIrxY8Vxo8WVesSjpoclY5L2Ja41BEcrczx07DtTgfY7C2gqHLSbXXvpx8e2KS6U7F2y1NLZtv3MRmkqYPHqL/J68SeQ5hRmt00QJySJYqlic7PrjxDdz3N9hYmcOYHJRMio+ae+uLqX4mVSZ8GcnJycnJycnIwHTmBJBLoU+HGCrWtIKYiShEYYow9svTeFi6lLZgsGl5seJ2vpBrJvkDBfSt3rlls9bnUCNAkCsWsu0TCjcrD8KOZCvcdSZdMgY3jWyprQ/n03dqt69Dfe/fVtP0aTgiT7048Snp6rcqPrs5G19vXCGFuXrHSDgfAbZ0VfNAqsdAPmiyZerAjiiOVWWhE8US3ixwnXWh6LZYtpy+Byy+Nrq010kc4vTVs6JUObTECItO0wUQoNMahIxUn6v1HtnJoUA1F7errEuh9yK6t0RoniRscbO8u02g2xpp7g9GwabOxHMWdX2wSJQKoer2546FIS4tPrrvK2xRMDgRYDnj6LEfUA9mxTHJelNl802PCineLryieJGwpj5sObCx6qxojiE5QrS8wVDYxwnTCOWU2miJVipRPy1Nw20f3I+zhX+wYWsjyx283dGktfXIWNVOgctMjZ7/mHj8uFV05OTk5OzgPPAyew+lUjP05Y6QZIKUiyjbcXxby+0UEgtgiXcQKhv7k+US2yWC7clkX29nP3Kx0PbZsJGsxaFa3bvtb292C3dkNg5DF+FBMlipsdn5Khc7ntEWXZYJebXWIlKBoa80Mic1zuWL/17XLTQ4r0+paUKODsWps4UVQtnYYf0YsSlIK3zVXw4oRTClZ6AavdYM/7gDQAWgrBhXqHIyWLWEGcJCy3PMIkve524iStdk7ZBl++1aRgaCxWbIq6RjeMaATxyFmmOFFs+CEPTxXws7VZusbJms2ZlYin58vUoq/QUybXSt/M/OxD6JmL4nKrx8q3vJFav9/8E6qtr6Ad/TaM+eqO56EvInd7PqXwdoovr8hs80U0+ZGd94zOevU9g2yy5NrLiCTCr36AS40upiZ5LbPZHzgpTuoWeCfVoOl33J9Zp7thhpGTk5OTk5PzQPHACawEhRBwqxvw2JAN9Rv1LroUnKqmmVbDwmWSYOLb/c39qADZV9fbRInCGJpZut1Zq1GMqpyNajccdcyVlkcYJ0zbJk/PVzZtvBtphcWPEx6fKVEydj5a29c9zsFw2MxBClhu9bjY6OHMlOhEySAQuWzqbPQCLtS7WyzFR7VNalLwUNXmQr3HSi8NBo4SRVHXMITgZtffMbPVDiKkFByvFNjwQpyZEpauITuXCEON4+VjuOupGUk/uww2K4Q1y6ATRqx7QWoaocDQBBfqPey2SRyHzE5rLJZtlFK8stKkFcRUsky1qfIsvvEttIMRVaih52G353OL+Fp7Kb3m8i+zuP4CdD+bnmxINITP/QFavYtx/fcAkL3rXOU4oX6TJ0o6onCEiqGznAVCL1UKB+8WeK9ETt4OmJOTk5OTk3MXeOAEliR1SKuZOm/WN23Na6ZO0w8pZG15XjZjM2nI650wzgb9TmetdmO3qsdux0xbBgVdMl80B111mhQslixeXW9jTGDusZeD4fb3eb5o0fAjZgomUmw63BmaoGzqmNpkbZMLpdTZ8KFaAUHqilfUNQwpeHW9w0LJRgoGa5MC6l7IxXoHXQoKmWiMlSDAIMmqja+stjiatYv2s7GiOKHhh8QKNCEG1vAFXePJ2QrJ/PNbnq8rjR5honj7fGVgWX9Fewa9AnE34EKjO5g1201Ebn8+t4ivV34+df1b/8Oxn82guqk0DBETK8mavsTpkqQTJVh6KqKPV2xe2+gwWzBZbnqTPZu3I5RuR+QchDC6k3PlQi0nJyfnvvFLv/RLnD9//r5dv3/tH/3RH70v13/kkUf42Mc+dl+u/SDzwAmsBEXJ0KnZBvMlkzgBTUIQK9a9gGvtHs09LL/vFnvZoN/prNVuTFKVG3VMmCQ0g5CCIan7IaUs18rQJEkWuLu9CgVbM6omcTAcxpDpuZUCKbdV9RScqBZ3XGMUkVKULYOjJXvgCNgXa/0Wt5VOQDeKcGZKGJqk7gVcbnp0goiweREpBeuRiSkUxegGxbDH6aOPs9zcrOgMz245M+kMlhfGuOttpGBLtQtSgb3a8zlWtjAyUwpb13hUk7yy0sLQBIa8s9k7TQq0D7yQfrPLhl/7o/czW/xmLj3xY5ysFYiv/QckZXxzEdtO2z+7YUwniumG6VxZX1xOTD8/ajt7CZG7LVC2X+92K2W5gMrJycl5y3L+/Hlee+NNZo8t3ZfrG8UyAGsdb48j7z5r16/e82vmpDxwAqsvEvwoJlGpjXgYpbNG3TCmYqp9ZAbdGeOExvWOP7kj2xDjwoR3Y5Kq3NZjJLGCiqETxIpOGKfvYZKglGIpC/TtV4CafoQmoGzqxAqmbZ31XjgyMHmHdfjQ9YfDjfsi40pWOYG9xRVsVmfiRO1ot4vihJuddB7MmSmlph0ifc180WS16/NyXbBUBERERXRY7lnMijVs/YkdDoz92a3XNjqD9U7ZBnUvzGzsh4TiQIgKtCHRZ+kapiZZ9yIWSjZHSgoyZ8a7VckcxWL3c1wzJOfW2oi2Tt2IOY6iaOgIISiZOnoUY2mCp+cqOwTjnkxis3477Xp3qcUvRieUJYxtn9O+yVsOc3Jyct4SzB5b4ru+/+/d72Xcc174xZ+930t4YHngBBakm99WEHOytpmHdLHRJUjUoC0P9t743wnjrNKPV23Orm7O9UzSkrhby93drrz1xU7fVbFopg55q62A49UCt3rBQDTe6vqYMnVhtDOx8Pp6ezCjNMxeRgnHShZnVlt84Xp9IJKnTB1L0zmz2provse1X16od0mUohOkgdMlI83dslNbSGxdUjR01sMCvqcjwjpFAbOzCyyWT+1YP4CuSU7WijtEbzts7bjH/v0kKhWs5SzU1wtjmn6IUopX11qYWWTAHX+2uwXp3voMAlj6/IdYyITGiv12Vmf+ASVDx9DSvLIrTY8jJWusuBop9scJjj7DP7/beVfDjFmH+uAfp/+O3v5b6fO02tr9vc4FVE5OTk7O1zEPensl3H6L5QMnsMIkoWrplEwNd32zulAxNYojKgMTO6TdxjqGbdCHRVIQJ7yy2mTaNjhRKaBru1cI9ttyd6eMm986WrQ4u9bmydkyUoiBgJRCsOEFrHtpvlc7iLnV8aiYaUCxEGJPo4TrHZ+CrvP4THnwszfWO9zqhrz9SHXi+x5ee7/CJlP/CVDgJwntIGSuaAOp26Qm0lmtgiaxdEm9tcLp0hp25YnBeXdzYOw/N+PuUZOCuYLFzY5PoZa2XsZJwtWWjwCqlsFTcxWsoUy0wT0e4KZeI0JLGhzv/qdBRWuvFsUoTrjS6tHwIwxNbhWE+7l4P+9qP/d3h2YYd/3fUe5AmJOTk5NziDl//jxvvvpVlmr3vr0RoCxTmeJd//x9uf7Vxu2P5jxwAqs/z3OkaO2YLVpuecSJYviX8gflkLbdKr2/uTs9XaTuR4DieifgervOyVph7G/RJwkN3j4TtVclbq9jhRAslGymLCOt8GipMPWieCAah78GWOsFCCF45kiV8/VOugHP3lNTk7uaeIy6x0Qp5ksm11v+kOnF3hXH4bmyy83UdvxExaYXJVRMna/cbHCl5TFtm8jsfVvueMyXTOp+xEMVm7r/OJeLT/Nw9tlN6sC42z0er9rU/YBX1zpUTJ0o+0VAL4p5ZKqItS0TbXCPt/H5jWS7GOizvaL13B+MPG//FwSXGr3Uzr5sUTZ0DCm43J9P20tw3A0hMqr6NcGsVZwo1lZbe/472kIuoHJycnJyvs5Zqnn8vW+9cL+XcV/42T99+LZf+8AJrO1tYnZWFVhuekxZOley1reDcu8btY7jVZu1Xsjp6RKtIMLUJDXLYLZgcWalRWdM3hLsHRocxDHrnXDP9sE4UQRxzGo3ZMPfPHba1pkrmJiZiNqtHXGLaBz6OkoUG17IXzhSRQrBbMGkF6VtZr04pqBvzc2a5B4TlYpeQ5dbqov7qTi2gniQ99SNEqQQvG2+wmevbvDKagtDCrw4YaFkM18wWeuFFAydqqXvaTpxtGhxpdXjzGprUMmZtgxmbGPHDBakwu/J2QqXm102vAhTlzT8GEtPn4Udn+3aS4Sv/DzaUHuaQnDtXZ+auFV0v0JMI0LTR7+n19oenSCmqGs8PVdBCkE7iADurM32dkRLv/q1D/YM3/70d6MljdtbTy68Dh2O47wL+GnXdZ9zHOdZ4HeB17O//leu6/57x3H+EfBhIAJ+yHXdLziO8xjwq6RF8VeAv+O6bnLv7+DB5WrDvqNN0Z3Q9NMtVdWK7vm1rzZsHj12zy+bk5MzhgdOYMH4FrdHpioH7t43ah1nV9sEcUIrCPHihKOWlRpLIDB0yWLZ4nyjN3KDuldo8GovwI/GG3cMC6ZuGGNpkoenClRNnSstjytNj1udtEo0V7AQArxofBvVsHidsXW+erNBrBQA7nqHoi45VrKZq1oslm1eWW3y2PTo3Ky+AJCIHffYN9YIo2RLdbF/3xKxxWp/u5jYvqG2NUk7iCibOlO2gSFTR8SHLIOioQ+EdqLS8OGHxjgXbhegAihqEs2ADT+kGUY7hM/212hSUDY0Fqdt3PXO6Ey0OMRIOlver2vF907U4rbrzN5tVJX6FcZHagUutbzBe1o2dep+SMHQtoreg3AJHDUP1a9kjZuRGrrenuHb297rO1p3XvF6S+M4zg8D3wv0P/RngX/puu6/GDrmWeB9wLuAE8BvAt8I/Evgx13X/bTjOL8AfBfw2/dw+Q80jzzyyH29/rVsXuTIsXu/jkeP3f/7z8nJ2eSBFFi7WZTfjnvfna5jzjb58xt1rjQ9pBRseBGzBYP5gkmcKAqGPrYqs5vD3rRlsOHt3j54o5O1Js6UeG29w+npEn6ccKHRI1GKJ2ZL+LFCE7Dc9OhEMe86NjX2fMPitRtGaFJwvFxACCgYGtdaHlfbHidkaj0uEDtys0YJAIHifL3DUtmmYOgkSrHSCYiVSjOmSOe4Lta7CBSvrqciOYwSVDphtcUkYq5gEkQxfhRj6drAfnx1/TK9doBnzVD3YywtwJSS+ZLJfMHcUtGME7XjMx01x/PqWhsvinl6vjJoZxy2dR83+7PaC8a3Gp78INozHxls1uMP/DFrZz/Jk51PY8x8eORn03+W7/asUV+sFgx92+xZKjD9KD7YIOK7wNj8ufOfYXblRbRbWXZYLo4eBN4E/hLw8ez7bwAcx3G+i7SK9UPAtwD/0XVdBVx2HEd3HGc+OzZT9Pw+8O3kAuuecb+zfvqD+D/1Uz91X9eRk5Nz/3kgBVafdMMpM+e3TTF1UIHC41j1AqYsg4WyRdHQ0ITgUqPHjXaLI0WLXhgRxcnYDepIhz1LZ7FcpBlGY9uevCgezDbFKm1bsw0NKQTrXsBj0yUkgilLYOkaltR4ebVJkCSYQ9M/w215tq6xVCkwX7A4s9rkqbkKYZLQDRME8OhUiVfXU/F1sxNsab/sV5lWuv6WqlsQxbx8q0ndj9jwQqIkzbA6UbERgi0VR4HC1jVOVguEWWvira6PpUlOVQsYUvDaeodLjR5SCM6utnl4qkDNMjA1SWPjAsfWPsfCN/wwuhBcb3s0/Ii6H7HWCwc5ZX0r+u0mH8OzYkop/DimZulseAFnVlpYukRlFu6r3YD5grXrDN1Ts2VudveuqoZJgqaZGGLrIOr2lskd82zLn8AATh77yOg2vkxI7NZO2K/+JEptESn9KuNqK7j9NttJBc1u81ATnmNkZXvlRRa7n9vPiseTuw4eClzX/U3HcU4N/egLwL92XfeLjuP8Q+AfAXVgbeiYFlADRCa6hn+Wk5OTk/OAcWACy3GcI8AXgf+StEf9V9nWl76fHvZRx97J+pRSLDc9Vnv+ICT3IEOFx9Hf8D4xUyJMFK0gQiCoWho3uz4rHY8bnXTm6EbHG7m+7Q57iVJcafZY6YZbKgr9TXL/ugiyUFxBkGVBhXFCgkIXmbuipRMphQDKVvq4tIKIoqEPqjFh9rpYqcF8UYLC1DUMTdIOY2YLBl6U0IlivCjhUsMjSBKemC1vqVhJARteyEPVAnq2IV/pBVQsg+O2zlzBIohjrrTSat9SpcCx8mYr4avrbU7VigRxklqt65KnZiu8up62YXqx4ljZImoqnpmvcDVr0bS6b6IHt5D1r5AUlrj4xh8RxwGzJz/I03MVomzmS5OCq63eyArQlVZvS9thN4wJY0UviimbOg/Xili6BAXX2j69KKG3zQykT18YRUrtXlXNnPbkpz+MN/PXaMqY0pXfRRMKjn90h0nLnrNG26qkk0QADFd/HqrarJCKySDLRjteLRxYm+3dZGRl+8hPpn+Zi6EHmd92Xbfe/xr434AXgMrQMRVS0ZWM+FlOTk5OzgPGgQgsx3EM4BeBXvajHX3pjuNcYsIe9l2OvS2UUryy0iRKYKlio0mJJmClG+xok7otR7Z90N/wmrqGCRR0jU4Y0QkVmhDMly1mbHOrG9u29fUrEroUdMMYL06Ytg3c9Q41U+dCvYOlazQyS/KmH1ExNUwhaHghL99qYuip899XbzU4WbXxooQoSUVKX88lKq0c3WinFbeCoRNEMa+td+iGMRcbvS3VnDhR+FGMADQpKZlyKKC2yuv1DpFS3Gz7A8EiBKz3App+auyxULIH99cKIhRpy+TDteKWiosmtYFzoSYFXpBQNTVagcLUJboUGJqgFSQcKZrc6AZESnGyVuRoyeLsq2tMNb9AOP0XOGlsYBQ9QqVxKUy42fUH7/mWCtD13wPAOP5RTtYKnFltIUgFp5YZZJR0yc2Oj61LrrZTl8qqpbNUNrne8TC1PWZ/MmE0XFUdfialgKvFb+bq9EeIrKN8SVgYoc9JeZ0jUczlLIy5/+wOZo2ufBJDxNC7nl7vyieJG2rQXthn0nbCfvXn1Sz6QAg4UjQ5PkHMwEhut9oz6u/3KYoOrIKduw4eVv7AcZwfzH6p90HSXxx+FvinjuP8c+A4IF3XXXUc58uO4zznuu6ngQ8Bf3zfVp2Tk5OTc984qArWPwd+Aegng43qS3eZvId95LGu667czuKWWz1aQcyzR2vYRto21Q4i5osml5vppl4K7kl47/bheikFJVOnE0YUdMlCyR5svEfN0wxXJDpBRJQopqx0Q32rGyBJLdJtXWepYqFJyfFKKiZfXW9TNXWOlS2mbAM/ijm71uart1qYmuByw+PJOR1Tk/TCmGttn8WyxbW2z6trHXQp6ISp3fo7F2qYQzlNN7s+swWDKy2PmqUPWsguNXpM2wYJamBGMdyyliiVrdHgtY0OVVNHACLLqhq2ZN9ecTFkKgC6QWpz78cJfpzmWvWiOG0tBKJEbREvtq5hHH0v63Pv5pnwsxhiHo5/NG2di5Mt7/luFSBDk5T01HJ+qWIhgKstj6KhcXqmTNlM34c36x0uNHoUdIkQk1u6j5tN49Ef5rGKzVRyEx+dC70KF70al280Bhb/fTarTUVO2l0MIFQ6l7wis80X0eRHBsdOGgEAu8813hU2vnL3znU75GLo0OI4TgX4EeAY8Enga67rvrGPU/xt4OccxwmAG8DfdF236TjOnwIvARL4O9mx/z3wS47jmMA54Dfu0m3k5OTk5BwiJhZYjuM8DjwGvAxcHeoz337c9wErruv+geM4fYE1qi+9yuQ97OOO3bfAihPFSjekaunYRrox16QYOJ5JkYqWtV5wW0YA+7a+HjFc70cx1zsBR0qpm2CUpBbi40RFv1LkxclAXIVxQqLg4ekiX7rRSCtcWtoOKIWgoGt84Xqddy7USBSs9gIgzQfzowRDCtpByBdvNKhmAmm+aBLECRVTI1EQKUUQp21+4+aHrnc83lzvcEF0iBXoUlAydL50o0HZ0OiGaVVt+D5tTeLHCUGcVsd6UcyNtk/BkIOw2u0Vnn7AbSeIeH2jy0zBINHB0gQX6j3mbBMAL0642OhuES9pi6PC1CRGFG/5fLa/54aUxDc+Q9gTGF5a/WH5E/hKJzDey2NTRW51fc6stvGjiDiBR2dKWJrEjxL8OKZi6ry50YXMqGOpYk3kXrm9muRHMWdWW1iaZMo20boJRXo8Pj3HmdUWAAulnb8QWCzbXFt4N+d6IVr7JeI4ZPbku1l87Lktx40Vk9d/D61lEc48j/ZHH0x/mAmQu1b9GZfL9fVALtbuNb9C+su69wG/nP3vfbu9wHXdi8C7s6+/BLx3xDH/GPjH23722l7nzsn5euetPiKSk3MvmEhgOY7zd4G/CMwA/4ZUaP3dMYf/dUA5jvM88A7g14AjQ3/f70tvMnkP+7hj902YJBiaIFEQRDG6JrMWNkGcJARRkhktBDw9V5k4dHSSWZVxbB+uD+OEXhBRrNpseCGC9P9tNMEOs4u+QLvY6DJtb4qrQQVECAwp0TWBvu11hpQIIShk7YElQ6dBiKFJpi2dMBNuTT/MRFxC0ZCUTJ1HpkrESvHGRgeBoBPE2IbcFIIirRQ2g5iqbbDeC5BCcKpWQJeSeWVwteVzdrWFHyuu2x4FXSNRCkuTeFFEN0w4PZMGL7eCiIJuX20/6wAAIABJREFU0g3jLcHEUsDVVm8QcPvodJFOkHCjnZo9KKBkaNT9kIYfsu6FVLN5KGDovdJZ60V4Cx/GHsp6GtWqN9t8iUvW85xUOrqIaWjHuBDO4UUJX7zRQBNQMnW8SKBrUDV1vChJq3AqbZ1b7QYcK1s0/ZjrHX/P6s+oapImBUsVmyvNrBO3fAoAk9QxMW3x3Ok8uaXa9MrPYySd1JFwG2Oty5WW2sT3n6eNr6RC6CCEQ79yFTbSP++0ve52X5+39R1mZl3X/RXHcb7Hdd3POY5z74Zsc3IOOY7jPAP8K2AK+HXgFdd1f2+X49/SIyI5OfeKSStYfwX4VuCPXNf9Gcdx/nzcga7rflv/a8dxPg38LeCfjehLf4PJe9hHHru/W00xpCRJFAVd8ka9w4lKuuGPk4TLzR5BlHC+3sWLYoI4QZdiIJB2C7G9E+vrUe1VZ1ebXG56ODNlbEPDC2Pc9fag2jPMYtlmuenhrne41Q1IFEzbehpqq9KKWpwo9KFCxCjDC10Kgjhtn2v4EfNFK3ODg5udgOstj7ofsVQtEMQJVjY/pAlohRF+km38RWqEYUg5CPJd6frUvRA/VpQMDV1Jnpo1+dLNBmVTstELKdc0CjIVQ5cbPU5UbOaKFvNFixsdf2Qw8aiAW1sPKZoSW9N4faPNsbKdCVSBlD2qpj6YFYriBCmgE4JC8aUbDZYqNg9VC0SJ4kK9S9Xc+lkvvucnUkF8qUNoHsWPH2WpZqOJdAZutmBg6xpyCr54o0HDjzAyw4opOzX7iBPFjG0ybbNjlmwUo6pJUoj0npIAr3WZUu1kemyc5oMJwa7W6JoUaB94Yde/ny0YXDr/mbSdUMTpXFrTZ3b542jnvmdT+ByUyJp+R/pnfxYrJ+c2cBzniezP40C8x+E5OTmb/Czw3wK/RFr9/X1grMDiLT4ikpNzr5hUYPV3af3WPX+f19nRl+66bjxpD/sux+4bTQok0A5iKpbGGxsdpBDU/ZAoTvjWE7NoUvLyrSZe1i5WMtO3aXs1o89+ZlX2WlvfSjvJrLxf2+gMKmJTtkHdCwdOfX2EEJyoFUAoWn5EydTY8CIafkTTj9AFXG72OFUrDsTflabHlKnzxnqHuaKBEKkIut72CJMYWzc4VSugSUmUJEwXDDQBb9S7TFsmvSgGEspZIPFswaBm6USxwl1v40cxD0+l14sy0fjIVImzqy0KhsZM1rKXKMUjU0XWeiFvrndTW/Y4phVETNsGktSR8FjZZqFk87WVJierBaqWQaIYGXBbMXV6YUzR0CgaOrYuKWYZTSA4MRQSvNIJ8OOEk7XUtbDhh1yo97je9khUKhbLps6Z1daWiuRSpcD8jZ/jlYUf4tlHnsWQkjOrLZ6cLQ+ep2nb4HjF5lbH57GZElGSujNe7ngcLVsD84dxon2YUdUkKURqbBL4dGMdO5tzu1DvEivFQtZieicslm2urbzIuep70DQrbSdc/jiL6y+AGPp3EDYORmTdLWOI2zXNOKzW6odlnfeG/w74P4EnSWeifuD+Licn53Dhuu4bjuMo13VXHMdpjTvurT4ikpNzL5lUYP1fwJ8AJx3H+RTwO5O8yHXd54a+3dGXvp8e9lHH3g5xouiEEVO2SSdM0LPZmBlbpxXEqaOgFMyXTNZ6ITVbUciMMEYZD8D+ra/3IkzSdZ2sFXfMdLXD1tjzHa8UeMVrUveiLYYWtzo+XhTvmPGxNJ1b3ZAb7QApwY8SvChGFxJdpu1+iUrFnikllp5mdAVxaju+4QXUTI1WAO56h2k7rZ6ls1WbbW5SiMygIv26LxDbftpOqRA8PJUGHAsUlqbxtZUmq92AiuVhaWnr4HovoO1HXG17XGl5VEwNKdgScKtLgRclKNK5svVeSM3SMaXc4agHkg1/qzCesk2emtP40o0GMwWDR6dKYyuSyXOfxKp3sfVNB8P+eQTQDiLKps7FRo8v32igS0HR0DhSsgbzVeNE+3ZGBuFe+SQrXonazc9xvv1O3qinLo7SnuWhqn3n1ugvPocAlm59hgVhEh79r9J2Qj2CI1mheuMrm1WsfrVp3Ob+xefS46ffcVs5VTvOtd/XDNNfR86DwOvAD7iu+2XHcb6bdI44JydnMtYdx/l+oOQ4zl9h9/GMt/SISE7OvWQigeW67s85jvOHwDPAq67rHtr/QHlRDELw+EyZOEkzitLwV8VXsqpVydQH7Wevr3e41QlQMNJ4AHaZVRnaPI8yvxhniLH9fH0xtddmPFGgEDw5W0II0LPzlgydc2ttnpgpk7D5+jOrLd5+pJrlYKUuexteyOvrbUw9qzwN30+SZO6KPg9PaQghiFRCrOBUrcB80Rqc+3rHx49iCoaeGVMoVro+vSgmVoq2H7Lc8jIBl/7CShNpoHGcKOL+z6Rg2ja4WO/SDmNOzxY5WioQxgkXGl2afrQl4HauaACCkqlxvZ06Ga72Qq40vR2OersJYwQcL9u7ViS3fE7Z10EUD3LFpNAoGzozBZNHp4ucudVCE4IjRQshxFi3wHEMz+oJAWFDY775hzx2+SdJhIF39MOAwH7uN+56pICmArSksTkT1RdJo0TTQRhT3GklZnhto0TeJK+7G+s4aA5rxe1g+XXgReDLwGngvwb+6n1dUU7O4eFvAD8GrALvzL4fyVt9RCQn514yqcnFN5HOYdnA+x3HwXXdw9lmIVKhstzq0fCjtMIUJVi6JIxi+jZ1ItsIr3YDHpkqYuva2E3ryOpCtnmesXVudLaaX8zYBqBY96KRhhi7nW+3zXgQx4RJGuabGmPE2JqkaGiD8N++gcP2iktB6tk8leCCFPTChDfrXR6qFpAibZ1b6QQcrxQQAs6stAbW5w/VCtnPN135pBBcaXmcrBbw4zRwdqUbEiWK622P1/wIXaYzRBfrPYyZVFx1grRCVdQlzURxqd7lRtun4Uc8PVdGqbSt0NAkD9eKbPQCLtS7nKoVuJkEfO1Wi2rmcni0bLFQsgjiBHe9s8NRb5ww9qOYOEnztobZXpHsf04XGl1OdF+ipBZ5feNE6mCoQBdwpZV+ZiVD5x1Hq/znGw3OrLYGwc/jRPvIR1cIFss2SQKrPR/z6HvZOPJeJAmL3c9R+sBvTXSeYXZ1vdwuSmCz6tP/u77I6h8HOzf3sLXSdesz8P9ObZ5vP2LgTgRE/z7CRvr6XHw8KCy5rvsLAK7r/lPHcfJsqpycyemQdjH1/0P1GLAfF7+3zIhITs69ZNIWwX8D/DSwcYBruSfYmkakFHUv5PR0iVApumHEcsujFyfc6vicqmlEWUvgXNEczGDtxnYnwP7mWSl2mF+4a238OOHtR6pj28/GnW+3zfhqL8CPEkq6tiXfq5G5AA5XvobFhS7FwAExHbMTlHXJRi8YuBhGieJUrcBSxSZKFJ0gZsrW0aTAjxRRojC0zarMiYrFhhfxZ1c3kAIsXWO2YPLETIkrzR66EEwXTEwpuFDv8rVbLWq2DggsKehFCcfLBWxDoguBFB4zBYv1XjAIPDY0SdnUMTXBq+sdFKmRx/FqgZppIDMxZGpa+j5va60cJ2SvtLxBe6TGpujYXkFUSqGUYrUbcM1/HEO3SHoBTT9CCLghJVP25mdm6ho1O5ttEwLj09+BRrRlg7+r4HnxOa4Vvxn/iR/jmfnhZ+d5rl28yNK2+afdzrWn6+V28RG1t4qk4b8/TJbqw4JuPxwWEXbYKm73CMdxTruu+5rjOI8CB5AinZPzdcunSM1pN2BgavyX9nrRW3FEJCfnXjKpwHrddd1fPciF3EsKemqysOaFWJqkbBg8Np3O+Wz0Qm526hQNbd/Vhe1OgMDA+GDYWnu+ZHK95W8Jzd3efrbf4NY4UWx4EUsVm+WWNxAMliY5t9ZmrmBuef2wuDA0QRAlVAyNdS9kyjbohDFVy+CR6SIo2PBDNryIs2vtHWJvuxCcsXU2eiHtMGbK0mkEEUeLFjVLx4sT6n7E6Zl05srUJP/FQo3X1tusdgPmi1Z6jBdxsioo6hq2rnGl5eGF8Zaw4TBO2xNPVIsopbjU7NLwIuIk4c1Gh26Qtn+GsaIbxegj7PLHCdli1d6zgnit7eHd/DzPWj1cFjlVKpDoZWT3Km9KB2emhBcnmWQdEmhCkgz8YlImsfmP0VmrvocnszUNnp1H3sc5v8PCjZ9BA9SL7+da8b2sPfYjY881setlf6Ou4k1xNY69Nvd3YwbrTgTE7bx2vy2FOW9Ffgj4fxzHOQpcJW1bysnJmQzbdd082y0nZ59MKrB+03Gc/xs42/+B67r/5GCWdLD0xUrRkMRKUbN1BAJdCWxd41S1yJv1Dk/MlDGHfc0nZNhqe3sbHmTtbVJi6FsrKqkRhMKLY0pSH3m+UfSrFLFKjSMeqhZ2CAY/TrLZpK0slm2WWz0uNnrM2gahEjw5W6ZgaPhRwrm1Nhu9iBO1AmXLYLE8uiKyXQhea3lECp49WsPUJStdn4YXZSYipPbhuqQTxtT9gJYfZXlkglgpjldsGn7IjY7P6WkDhaJqarjr7cw2fuf80tWWh1KCk7UCyy2fsqlxeiY1qGgHEdfbPje7/g7L/ETBbMFkvmAN5tM0KQaCZ1wFMU4Ua5f+kCftDWIxi67p1IyYWIuo20eZ102ut32qVirc40RxsdFFoHj11U+haSbx9Pcwu/ICiy++n+Xit9A+/cOcnilh69pWwfP5DwEQrp9Bm29hXM8cco9/FJY/gQFoUZtw/atoLz7HNbVEb/75seJph+tldo6Txz7CuXOfSoVav8pj1LY+NP3vc7Fx5xxklSn/fAa4rvt50oH7nJyc/fMnjuN8B2nLHgCu616+j+vJyTkUTCqwfgD4Lb4OnFskgm4Up+G68Wa7VxKnQqRgaJi6tqPCcDsYMt3c9sIIS9eQQqTiIElzigwpBxv5lU6AF8e8vt4ZZDwNKhcjWr22VzzCOKEbxoRxskXwALy23sHUdoo0IQTzRYsNLyRIFE/NljCzGS3b0FiqWFxteSxW+hlN48XesMX8as9nqWJjG+mxJUNDAsstj/miSRAntPyIdpjOOj0xW0KTkppl0Akj2kFM1dRRCr5wo44hU+t2QwouNXpcbftbBM+wYBDAjbbHlG0TJIpWGFIxdE7PlHg1m8MaFlCjKkb992a3CmKYJGiaiWHXkIsfIr7sEgobo3wS0QtYMDWutf0t7ooCha1pnKoJDOERajEX7e/lZa9No/YuppXitfXOYB39quY8Nom0kIlHHPmEqoohNqN80uDfACNuEtfPsPbw9/GkOo9xPXXTNY5/dEuFdFfXS80glKXNHqrbcdobt7nf789v5xp367XD81qwOTeWV7IOHY7j/DfAP2BzhgTXdR+5fyvKyTlUHAV+hs39nwLee/+Wk5NzOJhUYK27rvvTB7qSe0S/SrHc9qlZOklmmnCl6VE1daJkMtvsvVBKcaPj0Q1jXl3vsFi2KBs6hhSsdALizD3vRsenE8ScqNrYuoapyUG1od++NkoEjGrxctfavLzSGsx2JUrsaYxhSEmUgKmJgbgCMit1OXJ2aTfCpO98mIY3e1GCHyUgBO0gYrWbRqhdzRwEn5qroAlBM0jDeE9WC3z5ZpOSkWZXnZ4tD8693PSwdDFwK+zfUxhvVgq9KKZsGcwXLRKVBibbhhw4KvbvZdIWuZGi8sXnMNCJp7+HUIsxrv0us5HJpY7D8ULaxpgoCGM1cFeUCM6ttThWtpDT3wlSYCx/AjOaozX7FKcrNkdLNmHzIpfaimsssFi26YYxZ479fUwpiI98DOGvcbH0vjTTDAiPfST9jP3/BU0v4M28F023MERz6+c8ZNAxmL+78slUqPWup+/jlU8Sxwrjud+BP3p/+sJJ2v5G/fwwca/vIXf6u9f8CPBR4Mr9XkhOziHEcV33yfu9iJycw8akAmvVcZxfBL5EFjbsuu7/cWCrOkAMKSno6WzSxXqXS40eppTMFkwMTQ5Cc/eyUt+L/gb+nQs1VnoB11s+QdJDKZVt4BVnV1u0gojHs7awopFan/erDVE2O/TYdBFTkwSZCIiSLg0/3hFsfHqmxJdvNji72kKf0KVOk4L5osGVZjrj1BcpQZwgyRwBmfy+DSkHAb3rvRBDk0zZJlGSYOsati4xs+yxWCk6WRUrShKEgG6UYGmCKGEw/9Sn/74slgvjbe2zr+NEDWa1+i2Fw5b5dxoMrRExu/ICl+zv5aTSWLQCLheP8qWbDSxdYki55b2/2OjSCiIutbzB53I0ETSY4XglrVam64OTJcG5XkgQJ1ia5KlqjCUiQqVxsffQyEyzxe7nYPodGM/9DvHZTxLaVYzjHx6sd/j+B/N3azEnjUYq1JTOJa/IbPNFNPmRnTf8oG3+b9fWPeetyHnXdd+434vIyTmkvOw4zrtJYw76+7/g/i4pJ+etz6QCq/8fp4WDWsi9QpOCuaJJL0xd/G52fDZ6Idc76cb3eCXNSprEdGAc2zfw/VazXhjxZr3LsbKd5TtFnK930wrH0Dl1KeiEEXUvoGYZnFlJW71UFsJ7q+Nj6RJ9mwgwdY2yZXCqWhjkNE0iCo9XCmz0Av7zjTqWJjGkIEgSgji1bX91vT3xvfc377c6ASVTo2RqdMKIy80eU7ZO1dS51vY5Wja53PAIkxgNiRDp6xKVCiJdkxMHN293A5wtGFyod5ktGNj6zpDoUbNx484/UmBnG+3FF9/PtcYXOTefmUn4ESeqNnMFE1PbtPW/2uoRJorHZ0rMFdLK2qVGj+XqB5BhGm5tNb5KO1ykrDwMkSC8iOVGzNvKTSwv/cW7UTjGqUKXMwoenS4gEZvxAc+nztMaMNt8iUvW85zM7OdHGXQslm2ubXyec9X3oGlHiOOQ2ZPvZvGx57bc41juRRXmoCs796uSlDv93Wu6juP8PvAVNjeIP3Z/l5STc2j4NuDDQ98rIG+xzcnZg0mDhv9nx3E+DDydfuu+cLDLOlj6LXavrrUB6EURMYKapbPhh8g2KAVeNIHL2ghGzbhoUlC2DPShljtb11Ckm3ipiUHo7nKzh6VJTlULzBRMumHM5WaPmqVzrGyz3gu43vG53OxxslbcvG5Wpdgtswt2igYhBDMFCyFC5osGmpD4ccKGFzJtGxwtWRPfe//9vdjocqHe5Yau0QljTE0QaRJIQ3hnbROJZKMXMVc0qRipRfuVZo+5gsWGH+4a3DzuMz231kYKaPoRG15A2dSJ1daQ6EmCoZVSLLd6rHRDDE2ghs7RF5kCxVL3syzMVcZWOYfFdhAntIOIsqlzslbglVtNunHC8QqUoht0e4q6ViNSFutBQim6SS3edO5TCAJjnl4Uc77eRSBGCt/F9/zE7hb/Lz6HAJZufYYFYeIVTgEC+6lzewroB45c/Hw98Kn7vYCcnMOK67pvdxxHAPPAmuu68V6vycnJmTxo+KeAx4H/BPw1x3G+1XXd/+FAV3aAKKVY7wW0gxgpIYgVMwWDZ+YqxAou1LtseAHPLkzdVgvZJBt42Ky8XGx0mS+axCqdJ7rc9HBmiiQqNcVIAGemzGsbHY4BCMGpWoFzq22mLIOyqQ+qIru1N46ryh0tWqx7qQiQQrDuBSyULY6WLM6ttTk2ZLgwSfucEIKHqkWutTwKuuSpuTK6TFsPLzfTgOdXVpqUTZ1mENPwQ0qmThgnzBUsjldtZJstNuleFHOh3mU6y94adc1RNvmjhM9eQc5SwCsrTVpBTDWb06tZOt0g3ioys823BmNn1IbFti4F3TCm7qfZYl6SUNIlK92A4sJ3QqIIm9e47oFmlYgKs3jzVaxb/wFJQvfIh/A6Nyk0Vnjm2On0Mz//Ga6tvMjSe39y1/di1HumENyY/k7Wjv+N1NVwtTW+Urm92nKQVZh7VVm6Xdv2u7WWXLzdK34d+EbAIE1MWLy/y8nJOTw4jvMc8CtAA5h2HOdjruv+f/d3VTk5b30mbRH8Ntd1vxnAcZyfBf7s4JZ08JxZbZEo+MZjNRAClShe2+hwZrXF247UWCxbrHvBjk3puBa17ey1gR8+72LZ5pWVJmdX21QtnSBWFA2JqWn0oph2ECFIXf2kgIafft8OYiKlsntJ540eqtq7tjeOq8ott3oDERAlCZoQ6EMiMMzmp0bd+24zaroUHC1Zg9Degq6xULJoBhFTloGtS56YrXC942WVIjmoIB4rWVzv+JxdbdGLEsIkoZjZx0dJhxOVAkKILdcetZZxn9NuQc7LTY8oSW3mbWPTMt3UJWu9cCKROXhmtontkqlTUAo/iinoGs/MVbjZ9fnyzQZCCMygx7xY5dSRd3Jmtc2Zmzc5YTqUVJ1OENLohcyLVTTpoCE4aXc5V30PC4nasaaxro/Zxv7a5/5hauf+yPv2XaW9r6HCk85F5S14OSm/RRqUukT6+5BrwL+7ryvKyTk8/ATwLa7rXnMcZ4n031MusHJy9mBSgWU4jiNd103YTPI+lARRGnT7jQtTmHpqHFEwdZyZMn9+o04QJRQMnSjJNsHG5lu0W4vadvob+LOrrcE7NpfZrw+TqLSS8OxCmi8kSWeeCnpatZFZ1aPpB/SiGEsTXGsFSAlPzJaZK1gEccyVlocQabvf1VZvh0PeblW5s6stVHZ/mhSDtsV+flPfbn743veaUQuThLKpY2lpeLGlSRRg6xpVU+d4xeZ8o8f1jocfKd42X9myyb/eSTOrkgS6UcSpapFIKdphxNWWx8VGl6KuDSpfmkjXvd3cY1zL27gqzyib+eHqpRDsy1VxlNiOE8XVls980UTXJAslm9VuwKNTRQrGFF70KFGi+IajNb68XOdarAiYx+vc5LHkDItqGd58LV1bEqBFFcJPfzda0phYTMSJSkOL7e7uVdpt1aT4xecJZQkDHY3oYDOc7uVM1IvPpf8bd63c+e8wU3Nd932O4/xr4AfJN4c5Ofshdl33GoDrulcdx/Hu94Jycg4Dkwqsfw981nGcPwPelX1/KOlFMYaU2IaGUqlOTJTCNjQMKelFMUUjzay60vJ4uFbctQK1FwrQEMRjNGm/hcweskifLRgst3xKpiRJFJYmOF/vMW0Z+HFCMwg5WSugS4EUUDB0Hq4V08ykgjXSIW+3qpyuSYq6HIgAW5PUvYC1XpgZT+xsP9zL5tyQkliBJsHUJFVLR2ZOebFK1ywFrHTDgbjqr2eQ/5TNYjkzJfw4IVEwV7DwojTfa6lsobJw4psdn7Kh8fhMmSgztZikEpPej8wyw+Q2m/nNqpChSaRI20n3a+HfF9tnVltoIl3v/JDYDpMEXZOUrfS99uKEqfg6Goqa7HLK/xKRKPBV9RRzcoVhzRgqPc3ASjr7WlOYJGgL78OYq2z5uaFJhGAwK9Z/KhWCazPfxdrMX0bTTeKwMwhJFqjbExr7FSmTZlONE0N9clH0oBFmf5Zc1+05jmPe19Xk5Bwumo7j/CDwJ6SGF+v3eT05OYeCSU0u/oXjOH8APAH8iuu6Lx/ssg6Ogq4RJgleGGMbGlpm4Z1k7WW6lmZHPVRNqx9jjQL2oC9Anp6r7Np+NWpea7Fsc7nZ4431LkVD0g0TgiRhvQslUyMBSkY6G9QNY0qmPmhf7I1xyNurKneiUuBm10/vN9tgxwqCRB8IrX6orxfFrHYDnpobLYz61Y/ZgsGVlkfN0hGILW5+idq853Fuft0wIkxSY4ggUdQsnTBJqHshSxWLsmGkghnBkzNlzq61aQcRFcvYshYYPYs1qgpXM3WCKEYKtSkypMALY5qZS+AWkboPkSBIc9i2y/PhZ0CI9DgNRZhAnMTYIkDTEszeKhftx3jEamCc+HD6TJ3/DLP+i2jPv7jn9cdds//+K6W43OxR99IZsUTB7Lt+PxWIL/142k5odzFETNhtcsn+Xq41vshS97P7uvbE3AshNGllaj9Vtby69Vbjtx3H+Z+Ar2a/JGzd7wXl5Bwivgf4ceAngbPAX7+/y8nJORxManLxMeAp13X/vuM4/9FxnI+7rvvxA17bgWDqkilLx11v48yUsXRJJ4h5baNDnCje3OhuaS/byyhgFPvJWRrVQhYlKg2pnUpDapWCc2ttTs8USZTi/EYPW0+rbHU/pJC18sVJOuc0ymCjP6c1riqnD9nJjzKJkIKBGBFAK4i41fVZKFkoUmv17TNq/crNcN7YfMlkvmByqbG7W2AUJ6z0ArwooaBLRJy6I7b9KBMhElvX8OIEUBRMHUuXtIcEpxRwudmlFcQj2xiHq3C6FDT8kAv1Hs0g5kK9x1LFIvLSvK6rLZ+KqXG8Utg66zXB89C/zlNjxPbwM3CiaqMAzzrBcstjVruKZh4lVBqFXpvirc9zrvoutEIrvZ+VF9MMrH0y6rnri6u3zVeYss3NGb2mx0a/nfBE6tZrLH+Ck0rj3PyPsDBXmeh9GHC77XZ7ZVNtP8/2ytWdtvcNXzfn0OC67v/e/9pxnE8Cr9/H5eTkHDbmgC+5rvs/Oo7zvwI1YOM+rykn5y3PpC2Cfxt4b/b1h0lLxYdSYAE8PVfhzGqLP79RT+eLkoSaqfNNx2pY+laXurFGAbswyqYdxptk7Ga4IISgF0YoFH6sMsMLyZv1Do/UigjAj2KutnxmCwamLgc5UItli4Kx6TA4SVVu+/32vx6e69KkYKXjse5FtIKII8VUZGkConhTnPXnnI4WLZZbqXtg3Y+2VMS2uwUGUczLKy3COKYVREzbBpebHtO2ThSn99LwIxbLaculUul74oUxiQJDisHsWNOPMDU5so1xoWRvEcGdIEIgeHK2jLueGo68uZHa5UuRtiYuVazNitfanxHHAbPRNIvrLyDGbNonFdv9Z8Bd7xAmCX6UsFSxWQxfIwx1Lvkl5pY/zpK+wWL3JcInPpQKvCM/ueVa+/llwPBzJwQDcVWzjC3rfGWliX702zDmq1ufZxFPZPqy/f0IZQ0j6exPlN1NNr6ka9RYAAAgAElEQVSy+fV+5712M9bI57TeUjiO8+8YPy/8V+/lWnJyDjG/BvzD7OtPAb8MfPD+LScn53AwqcCKXdf1AFzXDR3HObQmFwBSSt52pEYQJfSiGDObO9lPlWo3JrVp77OXrfZqL8CPEkq6hm2kJhEX6l2+ttIkzKpW/ZkepRRKwYYXsO4FRMmmw+BS5ry336rcKJGga5IZOw0NrlkGYZzgrqcZVDvcBDXJqanSSAGwXVw2vJCqqXN6usTltoczXeL8Roeza20EHawsR+xK00Opfutjeu2qqSPEZiuiJhhU62CrsJmyjIEIHsw9ZT/TNcmxss2xss3Z1TZPz1UwdbnVPCTwCJXGJfG9XAOWxvxCb1KxPfwMBHHMai9gw4s4Gy6mIcDLv8zRjd+nc/Q7AYE99B5OEoo96r0fvmbfrXLKNnes09AkQbzteT7+0fR5XmtPNJO2ZY1v/7fpGt/4aRa7n0NkIckTs12wTCpsbtf5MBdOh5VfuN8LyMn5esB13U9nf/6J4zj7G0LOyXlAmVRgveA4zp8CXwCeBT5xcEu6dxiaYKUXsdYcvzG9HSa1ad++6R1VLYsTxYYXsVSxWW55g/OdqBZorUYcLZicnCoOznm11cOLEp5dmEKTAj/a6jDYX99+qnLbRUKiFCgoGjpe3OOV1SYCwZRtUPfCLeYQ29+X7dcd3uR7cczrcYKT5XHFiUKRiqgjRYtjZYs4IcvE6vDVXsBMwQSVVs+utT3W/RBDSqqmRjlrFRymL2wQ7Jx7kmKLCNakwNIlCWqnyDz+UQzg+OVPcfb49zP/xLdj6jv/uzOJ2N7+HBSkzglDZ7GsCGeeRxeCG82X+OLJH0ZVnVQ0X9sYiOZRhiMXGl0uNrqD2brdxJcmRZalxsh1Jgrmi5PFDoxjtCnK81xbSb2z7wn9ylXfJGNYKN3NXKtcgL0lcF33MwCO48wA38HWHKzP3Mel5eQcJuqO4/xN4CXgmziEM4wbGxusra3xwi/+7P1eyj1n7foycnb2fi/jgWRSk4ufcBzn9wAH+DXXdb96sMu6N+zlhHcn7Nb2N0nFAVIB0A7SmaOHqoUd5wsSxbGKtUWwba80DTsM7ie/aZjtIiFRKs1s0iSWlDwyVaSYtVa2w9a+Wsb6aFKgJWKL6UW/1bEbxjw6XaSgazT8EEvX+MZj03z1VoMkVsQoQuChWoG5gomppdc+s9oaK2xsTRuI4OMVm1gpemE0aLXcLra2i8zBZ5g8QlAyOLPa5EjJGnyGw6JpnNiesQ1udMY/B31BerXVo77wl3ls+ghTtWkSpbhQ73KzE6CADS8afOZKKYI4oWbpvL7e4XrLo2oZPDFTwtS1sc/4Xr8U2O153ouxbZKPvI9ztW8YmeG1LyYVNv3ZqVv73Fvnwumw8xvAa8DbAA/o3t/l5OQcKr6PtEXwL5KbXOTkTMykJhcngG8H7PRb57tc1/0nB7qyA2Y/RhT7OedwJWJc29+onKrhTe+wAJPZbMzlZo+HqpvnA3htvTMQE7D/2a9J2b751kSazXWx0SVIFBfq6SZ8vmBOnBM26r2LE0UU///svXmMZNl1p/fdt8eae61ZW7PZ0WxugpYhKUHDZrO1kkNqDI/RBiyMDY8sCzIgAQOboKCBZEOyYEEYmAY8GJnGLDZgENYyoiBRpNjNbomWKIoakRK7qzqql9qzllxje/G2e6//eBFREZERmRGZWZVVne8DGlUZ8ZZ774vsur845/yO6omiU0WPazWf9bbETyShVCQqtTg3DYOS53C+nK7XW1s+x/Meqq/kouSYXKn5Y632TxbcXi2eKQRhopjP2TwxWxoRoRkUmV1x/tTpd9NKJEXb5EY94FajjRBiQDTNezaetb32TWtNO9Y7CnypNKt+xMnj72E2l4ppE8GF2TyvrTa414pwrfui1I8lidIs5lzueRFxojhZdImVxmHnz/hutYA7pbHuxL4+l7ul+00jdh6WUMoE2CNHtVr9byuVyr8B/hlpDXFGRsYEVKvV1Uql8j+T7v8A8oc5nr0wNzeHcnJ8+md/4bCH8tD54m9/jrnCZO7XGQfLpCmCvwO8CNx4gGN5qBykGNkpIjWcFjeJsLvTGoysbQUR12ptAM7N5FFajEzPmrb2axpThP7Nd6wU7ViymHP4wFIBDVzZ8rnTbHCi6E4lTIfXLkgkf3+vzvuXSjiWyYmix61miG0ISo5NI0ow+xofp26KkGjNq2sNHFNQDxNMAQXbpB5JNtsRRcdCagaiLrdbITnL4t3zxVTsJJJrtYC/vl0jb5vbrOkLtsGVms/Zco71dsxTc2l/Ls80cC2TczM5vnV7i8W8s0005WyD9y6WBhwaX1tr7CrwY5U2UR5+RrZpYFsGuiNIu02iu7VkqQ2+Tl0zPafnNjnK7bHLJCJqL6Yv034u98zDaEic8VhSqVQ8oEBqelE85OFkZDw2VCqVfwX8BHCbNMVWc9/0LCMjYwyTCqxGtVr95Qc6koeM3dmkt+MEt2N5Dnvb9E2TaribsAsSuU2Azbg252bgu6sN6lGCHhIKXSat/Zo0RbGf7uZ7Kefy6lqD7zsxg9JQ65gjzHgWm0HE8bw7cF7XSCRnmSNrlIbXLkoklzda/M2dGjOejVSasmOy3o4p2FZqYR5LbjYC5lybWCn8OCFMFN97Yob1doTTScvzLBPHNLiy5eOYgjPl+7Vq49IpC7bFxbUGT88XsU3BrUYaldKQ2ucrxWorwDQMWonEMw3ydio4DJE6GJ4peWNFU7ehdNDpV2YagkQpDCFGip+0YXN63/7atlgq4kRhCJjPuVyrpbbyaf+q1ORjKW+zGSRpSif3rfp3+4zvRUTtxKSfywF2axa8H8OJTCgdNf534BeBbwHXgQfUuC0j4x3JPwDeVa1W1WEPJCPjcWJSgfVqpVJ5Afg2HdvbarV6+YGN6gGjteZOK8CPJa9vtDhVdCnaFrYhuF4PdjSiGGZcRGq5nDrQLeXcAWGx27f5dFz4+t8TQjDrOcx5NmdKuV7z21FMUiuzn9ozhca1jF6z4pzWvY37Wjsm0RoLUErx2lqDrTDp1TDNuhbvXSyhSQWEgdi2do5lUlko8tpag/MzOTzT7PXg6rcwL9gmrRi2wpjNIGbGtbDE/ev1eoTZJhdm0xq0fsYJXdcycSwThWalGbIVxGmNmZ2u+fVam1oQkWgo2iaudV+ItOMEyxADr3U/D8MRI0sImmHMqh9iGwYa8EwDu/PMuuLHNARLeYfVVoQVrTNrJ6jcGa5s+UitOdGp++quTzuR5EyTpULqKmmIgCtbPjOe1RNX05hTHBT7qeHKyNgnbeC/AWpADPz24Q4nI+Ox4i3S9MCsdjEjYwomFVjfA3xw6LXnDngsD42uwPj+EzOstiNuN0Ii1UZrzXI5N5URxVjzg3ZMJNU284Nx3+ZfqfmUHBNbjBdgSrOjuILd07z2W3s2LBC7kZfhqMhraw2Uhh84MYtnmwSxpLrR5JsrmxQcu+NwqNBorKH7dY0uzM56AQMW5pc3migNp0suhhDM52z8WHKz0R54Ft2ozSiBs5vQNRDca4U4ppGmKJqpO+KMa2EZBnEiuVZvD9R3rTRDIO1LtltU9K4fYhkGtSDhwmw+FYRBxJVmuE38pJ9HePPWdZS7hLQ3e9b7w/VRN+o+kdQcy7sIIVjKOdxpNtgMItba8aEJm6lruHarl8oMJzIm51eAD3VqSU4AfwB8+JDHlJHxuHAGuFapVN7k/hfsWYpgRsYu7CiwKpXKN0h/oYZ3Qo9tH6xhgdHd9LXjhLe2fE4U0g3rbkYUXYY36j3zg7nCgPlB/3kD3+YLaEYJUkPZtXh9o4lAc7Xmc36MOcMkdNO8uvVD3Q3tfmvPJkn3ihLFVpj0xBWAZ5u8azbPt27X+MDxPJ5p0owSXl2tUwvjgf5L41LYTEPgYCKEwTMLBUxDIEgjVQueQ3Wjie6cbwiBhrEpcbvNQ6EJEoVrmbx3cfAzECpN3jGxjfvGFYlM0/WU0hNFRdfbaVPf1XbUu0acKPxE8vTCYImIeOljLAMnV79BYC3CwofxVA3z+Re3rc+5mfy2SNGJosvxvEui9UT1dg+Sfacf7rWX1bTXz4TbO4lGtVpdBahWq3cqlUrrsAeUkfGoU6lU/lm1Wv0/gWud/7o8tvu/jIyHyW4RrBceyigeIv0CQ/WltxVdG6vTxBaMiaM8/Rv15bI3YH7gdjazZ8oe1Y1W77zhiINtGFyYvS+mrm75BIncczqVVDptVuvHbIaDEbjjeXffhgO7pXu1O4KuK64gjexZpoFnCVbqAe2OKYNhCF5dbfD9J2bwbGtXMdl9fv1peJ5pEHauV7BNrmz5vRqsbtPhUdfbaR6x1MRKc7rPBt82DZZL6TmeaXK2nO+NadUPCRNNZb44Nio6PAfHMrdFdV7faPbSLIcxdUQhXgG1PvbZ7BQpmjRc/UgxTuhkAihjFzrOZwBWp83I/0daTxIe3qgyMh4buqZmXz7UUWRkPKbsuOeqVqvXdnr/ccQ2DBKp2AqiXs0Tml4UYlTPo965Y6I83Y36xbUmkVS04gQEJAoilRqHx0oRSUnOGFzyeiS3CbnznZqhp+eLKCaPOvSnJ/qxxDUNLszmmHFtko7QuOuH0xsODLFbulfOMjtOg0maRujfAGEQWcfxk3Q9+k0t/n61zt/cqTHXsXnvCsH+yFv/8xs2J8nbaX+szSBGaU0zkmwGo50DJ52HQlFwTKRKjTocy0BrCKTCMtKmyvfHZQz0ohoXFe2fQ7/I7UZ1xorcPdiLH7RRxUEzjYMlLz6bNgme+56dzS32E30aZ6qRCbnHmerQnwBfPIyBZGQ8blSr1a90/vz3hz2WjIzHkcfyS+39YBoCA3hr0+d0ycUxTSIluVoLcYxuzc90ttL9DnuvrdUxDDAwKLppvVQQS8JEsdaOOGPfX/LdhJxC91znJqGXnjhf4PJGqxdJ82NJwbF6EbhnForc9UMurTcxRDqvxZw7dV3OuE28bQosAa+uNThXzmHqMkKGXKm1cPqidek1DM6Uc6w0Qs7P5HANg7t+yMUx0bFx5iRrfsz5cp6lgtN7PpNu4EfNwzYMLCEQAmpRDBEonQoDQ8ByX5roqOdoGoNR0f7r78lVb4+MFDIPUjzscu29OFhmZOyFbGOYkZGRkXFYHDmBJZWmHiW4psGdZoRtGcSJQmtNPZJpVAsoO2mqWX/q3m4bYMcyWMy5XN0KeM9CMY12ScXNRsDpksdmkHCqeN9q+yD7A/XXlkmd3sOzzbSPVqcHUle4JVpzquihtGbVj7FNg80wxmhyILU6K82AhZwDretc25zHEtCIFEnrTYrFRQwx3xtzM0oo2haOFWMKwV0/HKh9CxLJlS0fpdsYQuxqTtK/Sd9PBMc0BIt5h3U/5kzZ69Vy3WwGnCnnsPqe116e46mix81Gm++uNrBNMdZ6f4ApBNGOQmbiqxw8UzlY/s5s+mdcu/+aPZNGskZFrg7Cuj2LXGUcMp///Od5++23D+3+3Xt/9rOfPZT7P/HEE/zMz/zModw7IyMj46A4cgIrkBINfPD4TLpp7ny7r7Tmr25tcLXm04plz3xi7XZE2bVQk2yAgcW8za1mwOXN1raN7cVOk97uxv8gIxn9URRD6YENf9dNr/ua1JqbjTZhonn/Ummg/9S1WrvXf2ovkQWpNKt+RGW+gNu+gtRXCYM6hoq4zBJBW7LWPjnSmrzftt0yUqOR9XaMIeDKVhtLwPefnB2oXRqXhncQHM+73Gi0ubTexDaNsbVw0z7HrvjZDBIcUwxEEA9qDiOFzNt/xsrqi5x+EGlwE4ic/TpYZmQcBd5++20uv/kWCydPH8r97XxqsrPeCh76vddv33ro98zIyBjP5uYm6zWPz339wmEP5VC4WfNY8Db3dO6RE1hosIzUWnwgNUzRa+javym9UvOxDcHZvia1O+GYJnnb5Kn5AsB9974x0YyD6g80HEXpGW+UPDQgpeaNzRZ+nEaENoOYs+VczyI9VpqTRZekrtPar06z2kl6Y3XRWnOjnhp0tBOFv/CjeKZB4d6fINDY8z/MjGVQCxPOlDzcIRMKhe6JxGEXx7utgFuNgNV21BuPaYi0H5dIhXPBOJiP83D0RwA5U3Cs5JGzrZEiaJrnuJc+ZNPULI0VMp7PpfJHOCEcTB1NvzD7ZOLaxq4460au7Jn0z90MLw5CMI46N4tqZTxkFk6e5tM/+wuHPYyHzhd/+3OHPYSMjIyMA+HICSzPMhHAVhBRctJNvYGgFkaA6PU1gnTjd2Fme5PanehGM27WA87N5Hrialw0Y+r+QIzebA9HUU4VPa7X2/zt3RqulRp7WIbB95+YwTAEG+2IeihZaQacLHoEUjHrOdzxI2Kl8Cxz6sjCSjMgkhrPNCk5aWPbZpTgW8dx4lWk0pwp5bjrh7yx6W8TIt0apyCRAwJBKo0pBKeLLrebEScKaZ+ym402m+2YUCne2GixlHcOJArUL4AsQ3C93uZWI2Cz0zR5VGRv0uc4bRRnLzVLY4XMmU9g5hrE934cU9UOVjBM0LfKxkJ+4PcPJCU2IyMjIyMj48EyNzdHLrjML/zwlcMeyqHwua9fwJub29O5R05gmYbgVNHjjY0WhgGOaRBJRSw1RdscsP+GyftD9bOXqNQkrm+7bbZH3fdM2WPOtXlry+eZxVLPnt40DJZLNpc3WyzmnIE0wu5Gd3juXfGDAM80B4RAv3C454c9oVd0LNZKH6AWJizYJlafy942kSjSNMwrWz6GoCeumlFCzjIBk1C2ubrVYq0dIUQaXVrKu5wpeVwf6je2F4YF0K1Gm1hqvvf4zNi+Ztue49c+nv4wQsBM61C5l2jXrjVh6nDaAJkkk6VS9ok1iUX87FfSz8luNzjoCFPmLJiRkZGRkZGxB46cwOridgwpHEsQJZrVdkgQywP5dv2golLD73V7LY3bbI+7b5BILNPozcsQotc7yhD3UyO7/aO69+/O3RKCm/U2txptFJAojQGcm8n17tsvHIaF3mY74nTJGxCY4xohp5GsNldrbe61AkzDwDMN8nZqYx4kklqUcKqUYyFng4ZbzTR18CBqefrnMSy2/ER2mvnmuLjWYNZNe21Nei+pdLrWUk30OdtrzdKuNWFDDYoH2K+IGI5kdemIlFPf/ElW8j/IpSc/s+OXD1prVvI/xHr5I5hb/uCXCS99bH9jzMjIyMh4IFQqFRP4PFABJPBfkTbE+XekTYpfBX6+Wq2qSqXyK8AngAT4xWq1+teVSuXJSY99qBPLyJiSByKwHuVfMKnSKNCTcwXKrp02GvYEedvg4npzaufAndhvVArovWcIttVN7dT8uP++oyIa/b2j3try8aMEyzA4P5NGRfrnftcP2QpiLszm095SSnO93uZmPUBpOFZwMRAD9+g3oUik4mw5P5DSttO8z5TzoAWbQcz5mTyulYqrq7U2nmXy9HyRtlTkLBMhBOc7aZzdNZgm2jhM/1p13Ri7YkuTfojv+iGNKOHtLR+pNTOuxZlSDuvl59KLDEU89MdfHpirH0uq602emi/gdOY26nM2bbSrn4Oq7TtoBJrT/l9wYrG045cPK82A9tO/xHuGBOJKM+Chlf5nzoIZGRkZ0/KPAKrV6g9VKpVngX9J+k/nL1er1Vcqlcq/Bj5dqVSuAR8FPgScAX4P+IHO8ZMem5HxyPKgIliP7C9YkEgUaWrgVhAhdZqa5lkmlhBYQjzUTelOKWBA7z0hGKib6qaHTbLZHhXRSJQe6B1ldSzSX98YdD88nnd5da3BUt6h0Onhlbct3jWb569vb/H2Zot6lJD2bNZc3fI53xGoSmvutCIW8862TfRuqW/L5VQgVPvGU3ZMio6FZ5u0pUJ1nl13DdpxMjbaOKlJRP9aLZe9NMIWS0Kp8EyDO62QViR591yegmPRiiUrzZDbzS3O5X+IU/5fbrNBH55rlEi+u9rg23drFF177OfMQBAlkjCRA6mrk0RVp4qiTtrIdxKG0+qOfXTwz871TMbb6I+N3LVe4VJNc2L1G6lJRyZ8MjIyMh4pqtXqH1QqlT/q/HgOuEv6xXjnHwX+BPhR0gbgf1qtVjVwvVKpWJVKZQn4vkmPrVarqw9nVhkZ0/NABNYj/QsmIIglNxttgkT1NveelfZcOl5yOWeaE6f27YedUsAurjXQwHvH1E31ojUTpjCOi2h0+17tlF4o0GwGEXdbIY5p9MRU0bZYKricKLgorbla82kncleBOmrehhCcKDg9y3XT2D4egNfWGkil8Uwj7aHlWCitiRPFSjPcFgXai0lEd60ub6Sui5fWm1yYzeGaJqutiDNlDwRIBYs5l3nP4bXVBq3KZ1lxTE5/8yfSCz3/SjrXtcbAXB3L5APHylxca3C+nNuWZtg/5lBqLq6l959xbZI+18VJPpuTRFEfBhKL2ChgK73ruMdG7oTENF1is4yZrO19MNMKs0zAZWRkZExMtVpNKpXKvwf+MfCfAp/s7N0AGsAMUAbW+07rvi6mODYTWBmPLA+sButR/QWzRVp7FMq0lsmzLYI44fWNFqFU2MJ4aJvSnVLAEGAiBgRIf91UrBRKi4k328MCqhuxujhCDHl90RJLpL2pyq7F2XIO0zAwhObKVhvXElgC2okkVpo5z6a60eJsKcexgoNjjq5P6s47bRycsNaO2AwSTEPQiBKu133Oz6QphcPPohtdOltORdtaO2SlGdJOJEsFZ5uY24tJRP9aRVKy5sdcrwdoNIGUuKZJKCVF10rHh8C2DE4VXd6utTmBhUmy6zO2zO5nbXx0r+tieHGtiWsZAy6G+2Y42gSjG/lOw1Ba3XB6pFxr7Cpwe2maN/4YW0hY/hQA8clPIp0m9vwHgSQTPhkZGRmPKNVq9Z9WKpXPAN8E+v+xLQFbQL3z9+HX1RTHZmQ8sjxQX+RqtfpPgadI67EeiV+wUEoc02C55NGMJRvtiGYsWS55OKZBKOXA8V0TBqn0mCuOZpLz+ut9+omlQnVS2sLk/njytolGsxnEvLHZ4tJ6k5xtTLXZNg2BZ5nc9cPeJv69iyXes1CkHateamKX260A2zR4cq7AvOcw61qgBcfyLn6c1ikpDbOuzbGCx6xr04xj1vx4rOizhKAWxHz3Xp2L601u1ANmXIvKXIF3zxeIO3VyozhV9MjZBq9vtLhab3O7GTLrWvyDk7OcLuVQmt66dyNl3bTIrsnEctljvR3v+ky7fbbOzOR472KJJ+cKnWgTPfHXfV5SaXJ2KrjiZ7/S2/zv9IxHRR6HxyyE4NxMnu89MYNA8PR8MTUWeelj240kHkH6xeJOn7N+emmaQZ5Yp+J6oE6tI16n5sVn0//u/Vn6X/fnjIyMjIwDoVKp/HSlUvls50efdD/3N51yEYCfAL4O/AXwY5VKxahUKmcBo1qtrgHfnuLYjIxHlgdlcvHTwHK1Wv0Nhn7BqtXqK6S/NC8DbwK/WalUfgtYpvNLU6lUvj3psdOOLVYa2zR6Zg1Ka4xO1Mo2DeLOprs/TcsQ6QZvMeeyXN65z9I0KWmjaqOiRHJ5o0U7URhCbEsP66+b2msK46TudFJpVv2YOdfGM01CqRCCnslGui6anGVgCPCjhEBKzpQ83tj0QWiWOy6D/dz1Q8qOxYmiC0DRtrjeaHNly+dUyWPec8a65I1LZdRac6vRHlj3kmNidMY7/J4fSyIpyU3YnNg0BAXDYinvcKORCsLu56fXKHnI5n7cM97JPGVcxMuzTFzLQDGd0N+RB2niMCY9chIXRF58llMIVpJZLi19GrP5FaSMWTj38fTLhCxylZGRkfGo8vvAv61UKn8O2MAvApeAz1cqFafz99+tVquyUql8HfgG6Zf9P985/59PcWxGxiPLg0oRfGR/wYq2RZRI1tsRrmlgCIHSilAqokRS7Bg5rDQD/FhytuwhNUiluNUI2Qoj3rdUHiuypk1JG66NaoZxryGwbRpj08P200x3Une6WClsU6B0KqbS99IaGqk1fqw4VUrH0YwSbjYCTuQ9ThRzrLVjmpHcNu+uuHt6vkAoFX6nr9bJgssbGy3cvtS53Yw7dusXdaXmUw8TrtfTXlbd94JOXdWaH3NmZrpfge7zurrlc63WxjEMlgoOSzlnrGiaxtFv1x5Wr/wYkDwWvZn244Io0Jze+ANObH6J+PiPY6sW5vs+ub8BZa6AGRNSqVQ+BPwv1Wr12WlcbccdexhzyMg4DKrVagv4z0a89dERx/4q8KtDr12e9NiMjEeZB2Vy8cj+gnXFwZ1mwJPzhV49zbVa0LPl7oqAs2UPgWC2U2tTdmz+9m6Nm412aiU+xF76FvVHZIJE8uam6jUEBjg3k+d4weXiWpOn54s41vaUsmkNOXbdxHcbDRsGWsOMa3G9HnC27GGQCtL1dkTZsXhzs82sG+MnipOFNMIXdxz+npzLc3mjNTDv7qbbsdKmw6FM68Fcx8K1TRKt0VP2Hhu37hdm8mz4ITcbAd93fKaXJhhKxYXZHNfrAafU7j2zhtf4dCnH8bzLzUabWpiwFSast+OxomkaR7/hiFe3Tq1n4LHX9LideEBCY9LP2djxvPhs6jb43Bcnu+E0wmnzO+nxBzX3wxZth33/dxCVSuV/AH4a6HbknsbVdtuxwH942HPIyMjIyDhcjlyj4UBKcpZJybG4vN7CNgWxTHsZ+bEkkBJTiE4DXnriCsCzTcquxaofc6q43Q1tP9/Yd80O+hsCdxmVHrYXd7z+e02SttY9zo8kjpX2CdNa004UJcfkfUtlbjbaNKKEU0WXk0VvwOWu647XP+/hTXfOMmnFErcjfoCpe4/ttO452yJWilYi8ROJhk7jYgvTCHd8JjutsWUanJ8tTCVwx5mnDF/jVNHjVqPNt25vobTGMgQCyNs59MdfTp/vY7ChnjY98qHx/CtZ7RBzL/wAACAASURBVFXGTrwF/CfA/935eWJX2zHHZgIrIyMj44hx5AQWOt14v2uuiFSaSMqe2916EPfeT79lV4ONXzuRGdscLZb2/I39Hs7fizteP5OmrXWP6zU7DhPOlnM9l7/lUo5rNZ83NlpsBQkaetcZNe7hTXe34fGl9SahVFzeaE3tkrfTugFYhkHRNjHvfhVDJxjLn5zomUyyxvtxnNxNJC/mHc6UPFzLRHaE66TPF3gkRNi+Gh5POu5hN8Sd5j3NsQd97wfBYd//HUi1Wv29SqVyvu+laVxtRx2bkZGRkXHEOHICq2tBvt4OKToWlmmQaMVWO0F03jcNwWLO5VYjpOzYeLbZ21zPuBa1MBm5Md/vN/aTnj9JKiKwLbIyKtVtVNraTset+iFhokmUxjYFidIkKhVVliG4MJvHs8wd5z1q072Yc1jM22Ot3SddtzPl+wYdN+oBi3kHgBv1gHPtVQyRTPRM9pLuOS39Aq4/FfBmPWAzHOoTZorBez8mG+ipGh4fJTIh9DgxjavtqGMzMjIyMo4YR05gmYYgbxlcq7U5UfBwLIMoUdxpBeSt+5u/5bLHRhDyrTtbzLgWIJhxLaJE7bgx39c39hOev1NKnCHget2nEcne+fOeDWg2Or2mpNLMeRaLubRXVVd07hRR6UZqlku5keN7YrbM7VbI5Y3WrvM+6E23VJo512Ld9/nr21vYhkGsFLOuxROzJdTXfowbxY/ymn0W23SRl77EQv0bnPrIr4295n7SPScdc9fs454f9tY8ThT3/JCya+393o9gVMP82scwH9QYpjGv6KYH7rff117u/SA47PsfDSZ2tR3jgJuRkZGRccQ4cgJLKk0jSmt+1tohlmGQKIUAGlHaP8kQdPr0pGJsM0hwzNQKfDG/vZltP/sVD5Ocv1NKXD1McExjIK2t2km/+8Cxcq9x7fV6wN1WOv+lzpwmSYkTQnCi0+8KAV5fxGnaee+3oXO/IEx7himWSx4nCi5CCG7U27y21kDP/5eYloOIfQr3vsxy/ctYOgLx62OvPW2657RmI10Bt9qOtq35q6sN6mGy51TTjEeYR1D8ZuzKNK622449jAFn7I2vfe1rfPWrX93z+W+//TYAn/3sZ3c5cjw/8iM/wnPPPbfn8zMyMh4NjpzACmRqdPDB4zNoDe0kNb0QAr65skkgJVtBTDu+7+YXJpKrNZ+8ZU1c/7Jf8bDT+eNSCa9s+ZgCLszkextz0xAsFRxuN0IMIbjVCGiECe+aTftTaZ321pJKsxUmO6bEdYXnTsYa+533NHQFYWW+QDOWFCyTm42AjSDmdCnHUt7h4lqT712cwzMSYr/ONe+nuFs8w2n/L3a9ftkxubLlc2E2PzZdc69mI7ZhkEjFaivivUv3XSMNIThVcrmyJXe991imjGrsxYlyYl58FolFvPEatqxjPkhBsds1h8VN97WDGMthC6TDvv87jGq1ehX4cOfvE7vajjs242gwPz9/2EPIyMh4RDhyAgsNpoA7zZDN8P6meM61MQSoEbU3rmXyxGyBS+vNiWy9HwajUgnLjknRGUwtUzqNeNiWQZBIbjXaPDlXYNZziFVqkW4bBtX1JmXPxjYNlO42YBYDaWnrI6ItUxsvHBD9NVJCkNbP2WZPEB7Lu0gNZdeCuR8Dy8R+699yzmhx6cnPcGKxxCgZOCCYRNrfa+12RNm1UJptaY97NRsxjTTldNWPMMT92rdmlFC0LUqOhWOKPaeaTsJ+nCgnvn7+h1gvfwRzqYFMQhbCNzjl/yWH/xt0SGQpfRkZjyzPPfdcFj3KyMg4EI6cwPIsk0gqGnFMZb6Aa5mEieStrRaxVBhCPNDam4NiVCohwGtrjYHUMkOkY44ThdIaDcx6Dt39s2kIZj0HBESJYiuIkDoVLJpUjCZSYSAeuOnDNPTXSHXnJZXuPadQSmIpiaTG6G7n3QVs2PE5jmtYbBuCs+X8wBz3a4SxXMpxtxWytnEd2zTQ3nE808A2BFLT67W25+jSiM17f7TqTmsXcbhPEbDSDGg//Uu8ZyaHffuPiHWZa4V/woptcPowBEYmbjIyMjIyMjIeAkdOYEEakVrMOUitCRKJ0qmLXb3jDrgfq/XDZFTqoFSa1VaE1Bqtdcf1LzW6MoVACIFSqf08pD2oKvNFPNskiCXVjWYa2UM/UsJz+Dl5pkEzSnDNNPXuXiviZiMgbxu8vtFMIzOn/xGJ0sj15sjnuFPD4kvrzW3H79cIwzINlss5anf+njO5CHd2uWfHPtiPbP/rOhytiqXCjyXff2IG+/YfpeNe/tSgONzpgruIlJFrKeT962M9mKbJjwuZuMvIyMjIyHjHcuQEVqwUJceiYKeNhbt1SIVOWpZCP5rNUYcYl951suByuxUOpJZ1XQSv1NpEUvHmZosLM3kc+37tlgCEgFnP5vLmfSfAWc9mK4gxEI+U8Nypn1YtjFEaKgt5PNPCNQ1uNgKu19vEUo99jtMKpn2L8Ref5RSClWSW6tI/xmy+jEwCls597EBTAWF7ZK4dJ7y+0WK1HXF6eK7r3yB+9V9h7sOIYdtaLn8qvf7NP8RsuMQbf4eZrB1ONCkTNxkZGRkZGRkPkCMnsGzDQGpwTIOcbfZqjaTSSJ2+v1+r9QdN2uPJJ0gUT8zkyNkWSqeRj9utcKyb38mi5lqtxUYQc3mjhW0ZxIlCas2poosvNedm8ttMD5pxY0B4DveaOizhOeo5dcXkM4tFHNPAjyWtRFJ2TV5fb3KunB/7HKcVTPvte9aPQKN0+udETCFM+qNJpiFIlMIxTU6ZW9xeDzgR3cUUCm7+IbE2kVJjq9bO9x0nvjo/28+9PHottYmUMbasTzbPB8WYcWfiaweyNcrIyMjIyJiIIyewJt0UP4rNUbtRq9VWRDNOUBq2woScZbCYdzhb9nh9o9Wr/RlOLTMNwYXZAm4zYM2PUnt6AScKLsfzLhfXm70NcffcfnFxsuDy2lpjZK+pw2BUHVqsFM1E4nZ6e+VtE42mrQWOabAexFimGGnksBfBtC8x/vwrrDTatO/8Fc94PvaZ5w7WOKSzIY6ffQlTQCQVQaR69XWWDohwaGuPovCJtcm1IM/CuQ9jvu+T+9pQj13LwrMszBqYdz6y52tnZGRkZGRkZDzKHDmBBZNvih+m5fgkdNO8zs54KA1F2+JmI8A2Be1YsUq0a+3PTn22dhMXtxoBOcvi3fPFntPgSjPsRc362Y/197TnDj6nwSiUH0ukgpJtkbctnpovcLMejBUw0wqm/fQ960WWPB9bSGACk4w99FGyjbQ+LUgks57Tm9eWPk/QblCNj6V93twPszC7izgcF/EZHtfvzKbpjz++8mhFgkeME4C4Nvh+Jvzuk/XuysjIyMjImIojKbC6m+KlnNvrg+VYB1dDtJtA2Mv73c14Zb5AI0rI2+mYu5vxp+YLvL7WRAgmqocaJR53Ehfd+z89XyBWmkCmkZAZ1+Lqls/xvItlGvuy/j4I2/D+yMmZskcgVa8/1kLOxrPMAQEDgy59exVMexHjvTqlM58YeH3fxiHDG+Kv/Qjy1P/IuvdByq6NiUCqtPZOKnBNk1hJFt2htd7nBlqgx69ltjnPyMjIyMjIeIdyJAVWdyO/5kc9P/LFvLPv/j+7CYT9vN/djJuGSPtTGUZfOl96bKQUx/LOntMZdxIXsZSdTX/qRDjr2r37Xqu1udloc362sOe+ULD3nlLDdIXixbUGodTkbXMgcpLWBGne2mzSThRWx22x/1k8jOjlnkwy9lAzFBsFyqJFwTF74rkWxJQdi7MzOeZzz5EozZUtn5uNds8efkeG79v9eUREyATMByGo9hJJ2S0Cd1DjfCdFebI6tYyMjIyMjKk4kgLrVqPN3VaE2WmkGyeKO80QrTXLk2wux7CbQNjP+ycKaRRJqrTnU6p7BK0ooR0n1KM4Hf8BNPwdJS66gq4ZJyzm3J7wUlrjGAa1MCFK1J77Qu23p1Q//RHKV9caPDVfwOvUZGmtuV5vs+5H5B0LxxDMuhZLOYfrO6QOPggO0iRjgGHDiWf/ALXW4Fje5UTBI5CSN6TiybkCzTjhdjNkM4gxBFyttUELlssH02w4IyMjIyMjI+OoceQEllSa6/WAxZzDhdn8/WayWz7X6wEni7kdRcC4tLHdBMJSzt3X+ycKHgs5mxv1gMW8TTOSuKbBnVbIgufQiCTL5RyWuXN64DT1TcPHzrgWK82Qec/BRPTEwFLBYStMaCdyz32h9ttTahSOZXC84HCzHvQEzPWaz0YQ88RcgWN5h0RpbjUDtOZQmibv2SSjL4qw2zMdFnKmEFiGIPTvUKtvogsXep+7e62AzSDeu9D8J1vpnwcR7Rh3jYOoCRoXgdsv7+R6pXfCHDIyMjIyMh4CR05gBTJtLNwVV9BpJjub554fEkhJwRhclklqg3YTCLuJj0nESXczfq3WJtEaP5bYhtFzEdxpUz5NfdO4Y5eLHrebNV5bbWBb99PqlnIO6+2YnGXuuS9UN10uSGTv525D3P302eoXMELAuh/1Ilpde/vlosfljRYni95Db5q8H5OMXZ9p34Z4eB22gphT8QpNWeYZaxW7DTJ/DtMwOD+Tp9rnRrlf9mN4kjGGd5Jwy8jIyMjIeIdx5AQWGqxOHVM/RqfuZlQboklqg3arp9lNfEwiToY34wYChZ5o4zpNfdO4Y+8RcW4mRyuSnCq6A/23FnI2jmVMnPI2vOk2RGqK8Ld3apRdi0RpSrZJosY3Bp6E/jWrhTHNKOFYwaUWJkilMQ1B3rEwjDTdcloxN86QZFpBsWPN15jN9DTPdOCz88pPsep9gNszH0LnCxgqRGpBM0rwTAPXMvcvNJ9/JRWAjfb0piXjnP660bFHqSZoeAyP0tgyMjIyMjIyDoUjJ7A8y0QAW0E0aFkdRBid9/uZtDZot3qa3cTHNOJkWgOGaeqbdjv2mYUidwl5u9Ye6TQ479ms6Whsytu4qIvWGs8yOV3yCGVqpHG3FdCKEt7lFNFa9zblexUwbifqpjV4ZmpdXnRSkRhLxa3m5LVPo+bRbXK8ESR7dkHcif55A3uqWTMNgaHqCBQN5RKLGf62ISgZAcv+t8nLu8QnP7GvqGGXgzIteWx4GKLqnZyCmJGRkZGR8Q7hyAks00i/yb/dDEm07jWnXW1FnC5tr7+apjZot3qa/b6/V6aZw27HJnq79bYh2CY25lybxbyNY5oDazpq032l5rPWCnlmqUSQKGxDUHYtFjyL6kYLP5asNIPe+kwaERkWYp6Zius3NppcmM1jGYK1dshKI6QeJZwoeBOv9ah5VNebhFLxgWPlgxEUnc2zvvfnrMx/mvWLf4xpOsiFD1NyTAyRPpduTzKjY9qyW/Rp5UNfwo8lH9z6Eon9BqF3gfV2Qi2KcQ1j/yYb7NO0ZAdHwoH3H4XIVb/Q2fwOzH1P+nMmeDIyjhwbGxv85m/+Jp/5zGeYm5s77OFkZGQcIkdOYAGcLnkIAbebIaYQSK1ZGlPDNI2V9m71NPt9f69MM4fdjjUQBEla+9WN9t1qtEdGKjaCmNOl+x+x7qb7qfkCUmsMpbFNgzMlj9vNANVJ35zzHIRI72kaglNFl7drbZTWhIneNSIyOrpkAWmkcStM+NbtGo5p4BiQKHj3XJ5TJY9Q7r7u/T3JhEidFE1DsFRwuN0Ie+mntmmwXPa4uJaamOy119rK/Kdpn/lp3lMW2CIgXihypeZTD2K2ggipe90GMAXEiew5Tg7PI5GKm/U252fztOc+gtYKIQNmPY832k+z5lgs2sZDFfWjkEoT42HLOo9Oq+8d2PxOKgTv/dmDjSplKYjveDY3N1lfX+eLv/25wx7KQ2f99k2MhYXDHsae+cIXvsDFixf5whe+wM/93M8d9nAyMjIOkSMpsKYRMnux0t4thW+a9w/CIGCaOYw79uqWj0Dz+sZgdO14fnf3w+71IynxY8nljda2axhCIND3a+E6hFLhWmm0ZtWPef9Sadf7jIou/f29Oq5p8L3HZ4iVphnF3GqGBLHk7EwOELy21pgoMhZJSawUzVj2hI1tpL3JbCuNiBrC6Im8SCpeW6tzrOBOnS4osVhf+jTvsTexxVJv3hdm8txthVzd8nl6oYRnm7SjhG/frSGBq/X2yHncbLQxhEit9gtnkErTbKzgtleYced4YjZPwdn//xb21OOLIXH8vv8LKSMW6t/glP+XiOdf3ve4Dox+odONXHWjWRkZGUeOjY0NXnrpJbTWvPjii7zwwgtZFCsj4whzJAVWl0lrmR5U6t5OTOP6N06E9b8+zRxGHStIa6TOz+QHBNqNRnviSMWaH+OaBk/NFfBs8/41OmLgrh8x59lIlaa83agHlB2LUEqCWGJbxq73GZWaZgiBKdIIk2UaOJYgZ5uUXZvqRgsNhMnktUJr7YgwURQsE89O79mIEvxYEifpendF3lNzBVqJpGib3NhDn63YKGBaLraoD7xuGgJLCIqOxeXNVLDWgpiSa3Msb7PU6ZvWPw+pNLUwwTbStEKzE9Erlk6xZi0gm+G2GsS9stceXyvNgPadv+I9no9t3SQ2La7NfYiV1lVO73TDw4zozH1Pet+HOYYscvWOZW5uDuXk+PTP/sJhD+Wh88Xf/hxzhQf37+qD5Atf+AJKKQCUUlkUKyPjiHOkBdakPKjUvZ2YxCBgnAg7WXC53QpHirNJ5jDKrfD1jWZPXMH96NFraw0E7BqpkEqzGcZcmM2laXhmKpaWSx7/8W4NW0DOMrm65XOt1sYxDOZzFvVQ8nf3GnimQT1MuFbzOVvO9URmkEjCJB0jpDb8uhMJ641FqVScdYSFIVIXyZxtYQi404x4/7HdI2O9eQQJp0seNxv3+2t5psGbrQB0er/1dsxTcwVCqXrOfFP32Xr+FWylkRf/mNgrYy9/ovdWmEik1lyYLfTm/YZUPL1QpBElKK23zSNWCss0mHWtVPhYq9gGqNxZVpohs651oJ/riUR9nyjpiWPPxxYdu36RcK5ocunkf86JESmPe6H3xcMrP4ZJsj+xkgmdjIwM4JVXXiFJEgCSJOHll1/OBFZGxhEmE1hTMK17316Z1CBgnAh7ba1BzrK2GUlcrfmcLecnjlJ05xvs0KPLNg0KlnG/ia0hCBPJjUYwEKno1uTMuDZ+LNkK4156nWcZzLsOsdR84FiZO82QtXbIW1sRBcviXDnHrGcRK831WhuAs+Uc1+ttbjYCXFPw2lodyxBIlYqN796rs1RI6+pMIYgSSSQlhrCBVJxer7fZCmIc0+DyRmsgQjiuVqg7j7Pl3DbxkCjN8bzDxbUmkVS0EolnGuRts7de09qfm4Zgof4NrrnPc64jTGOpuNEIMIToCSlTpWM2hEDDQB1Y957dtL2lnMMqEZfWA0zTJG42aCeS9y+WJhrTpIz9YmJMpKdXt3WmIyRv/mE6h+VPYOYao9dtCle9bV9InPjFNP2wz6FyX4wSW1mtVEbGkeDZZ5/lq1/9KkmSYFkWH/vYxw57SBkZGYdIJrAeQSYxCACjJ8JMQ5B0zlkue3zr9hZPzafiSmtNJBUzrsUbGy3qYdJrSjzppnK3eprlUo47rZC/vbOFIhWIhhDky17PXr17jURpCo5FruN8141ynSnnueuHvLHpEytFmChc0+SDx8pYHUt12xCcm8nx3dUGtxptbNPg/EyOnGVyuxkQSc352RyOYRAkkvV2xNtbPkXHpBWnos+zTGZcuyeu3r9URgF5y0gFWz3gzExubK1Q/zxO177KCS2IT/4kAJc3WpydKXBaaV5bq1O0Tdw+MbvXpsmnPvJrIyNB+bLHtVqb5XKaDhjFks0gwh0TSeym7V2/8uec83xORCu0tceKOMPS2lexln99qnGNZUhUjPxiYvM76XF9wsjGQn7g9wc/Z+E68Y0/Rro/fHC28eHXsYUkdla5NvN9rHzjlznt/0Umgg6KTFRmHEFeeOEFXnrpJQAMw+CFF1445BFl9LN++9ahGcf4jTTFP18qP/R7r9++xcKT73ro983IBNYjySQGAbFSmAIiqQgi1YsGCVLDhW7kyI8lidIs5lzWcjHnSh53WtFUtUC71dNYpoHWUHYtThdzFJz7DYi79zENwZxn8eZmiwuzaRRNqrTOqnuN06UcSzmXV9cavP/YDDebAW4n+lN0LLbCmDnPZsa1aEQJF2byzHkOANfrbZ6cy+MnkhknjVIVnFRUzjgW52fSGqSLa00cU9BOFO9fKjHj2mwEEbUwYc5La7ISrUjk6AbHA2uhTWwhUUIM1BaZhuBYweVGPZiq/mgc2yJBndQ29dzXeG2twbdub2Ebglac0FxLeGo+TU00hWCtHQ3c81TRY2X1RS6VP4KZFFITifCrnPL/cqox7Ylh2/XN7wy8bZIMfs6WP0V844+5FuRZmB2zbhO66g1Ehf2+9EN7k0vlj3DC/+bBuhVm/aoyMo4U8/PzfPzjH+fLX/4yzz//fGZw8QjxxBNPHOr9a3dXAFg4ceyh33vhyXcd+vyPKpnAegSZzCAgjeoEiRxomLzRDgmSjqW60ARSMevaKK1JpEKINMp1eaM1eS0Q4+tpThZcrm35XKv7PL1QJNaaIJHk7fs1R8fzLnf9kI12TKJ1RxAY5CyjF03rotC4lkHRsQZEZk8wRglRkkbI5jrzDhKZpio6Fn4iSZTCMQ1c02Des3nXXKHnjHe86PHdew3mXJtZz6EVJVjCoOgZaKBom6z7Ma4leLJYGL0W3/xJVvI/yCX33Wlfqq0vpalmH/m1gfW62Wjz3dUGtinQmn0bo9yPBKV5/rdbITnL4qn5IqFURInkVjOkutGi5Fg0IoltwIdO3f+HXgjB6R/8dU4oTfzKT2GrFubzLw7eaIQYmMjNclpRMfc99x34Osec6qTxXbr0JUzTRvqrLKx+kVPX/idA7yxQulGxEccMRIWXP5W+ePMPsQFz/qPET//Eo20H/zgItExUZhxxXnjhBa5fv55Frx4xfuZnfuZQ7//Zz34WgN/4jd841HFkPFwygfWIMolBgNSw3o4puzYmaT3OZpCg0VyrtzlT8hCAVIrvrjZIlOrZd/txWpOUMyb7CIyqpzEEvLpaJ0w0Jcei7NoIIEgUfiwpOKlpwo1GG6ngmcXUSCJMJFdrPnnL2hZF60bvlNY9kXm2nPan8iPJ9XaIHydokc7LNMzeOVGn/1MtTDCMNErVTuRAypxnmXh2KljDRPYEqGmkdVpCCJ5ZKvDmpo/SaV+p3np3RQYmp/2/4MS13yI2y9jzH0zNEjopl91an80gwTEEYaJYzE2XljmSvg2sFA7r117iPTMCc+4nacaSkmvzLsvkjU2fpxYKoOHv7tW51Qw4U84PzsEw8FRt11tO42a567i7kSt7Jv2z33mvQ+9zdud/JTYK2He/jKkjOPbRne8x4lr9jI0Ka3NPaZvAzgIi61eVkZGRkZFxZMkE1iPKbs6FsVKUXYuCY24TYYs5B9sQVDdatBOJUpoZ1+b9SyUcyySIJZfWm6z5MWdmpvsI9NfT3Kj7NCLJB4+VeWvLpzs6zzJoxhIrkSRSUZOK9y7ed+lzO3bvF9caHC8MNuDtj96dLXusEvEf79R6TXRnPZvjBYfX11u8ueXz7rkCtmkw41pcWm+xmHc4nkut3u80A0whuOuHPSEXS4XWsJhzuVrzmfNScRVLxfV6wFLeIW9bA2YU20TGB34/FRnf/Ek8NAxFgFaaAX4sOVv2kDoVgrcaIfUo5n1L5R2FyaR9z2KzjGk62CIg0bqXIpp3LBzLwBQCzzYpuxarfszJguKuP+Qs+aE/SYVS96IjIhAr+R+i/fQvTWZhP9wbqvta57pSOKkgFWDqeEfxYT7/YhpRmkSgTBA5GRkVPvnJNCpsP3hX0G1j3U1wDYvFxyEqlInKjCPOUW80XKlUbODfAOcBF/g14CLw70j/iXoV+PlqtaoqlcqvAJ8gTcn4xWq1+teVSuXJSY99mPPKyNgLmcDaJ1JpAilBp9GRaTZqk2ymxzkXppbjcCzvDogwpdNak7OdiMW1WovbrYgn5wo4nbqnUCouzOa4Xg84pSZPExwe+6ofd0SexULO5kY94EzZQ5OKiqu1kBnXopXcjxponUbPAqkIpebVtQbHC4PRnW707vWNFkJAO5Y8tVDgWN7tzaEWJKz5IWEicSyTME5YCyKkVtSjBKk0ectgzrW42ww5UfB6dWHdCMzNekB1o8U9P0L1pfANm1GMtczP/2BqjjC0Lun6ewhEz/q87Nj87d0aNxvtXjSpn4kiRX0bWBsLufBh4oUiZsdNUHdMQ3p1elKhNNjm/SjipL2+oNPkuPwR3tMRJLCzhf049MdfTud27aU0pXLhwz2BmnoeHjDdtMMhDqyfXZYKl5GRMUTWaBiA/wJYr1arP12pVBaAbwPfAX65Wq2+UqlU/jXw6Uqlcg34KPAh4Azwe8APAP9yimMzMh5pMoG1R7TW3Gq0uV4PUFpjGQIBnC7lOF0anz4llSaSkjU/ZjPce9rV8DfynmWONFI4WfTYDBNaicRPZGqLbhqdKE040KB3t2bFwxE02xQonUaFupvX6kYLqVTHhKLAyaLLxfVmLzWra7pRsEzytslT8wVuDjXg7Y/e1cKYZpRwspjr3d80BO+eL3L35jpKaRKl0AgWci4fXCoTK8VaO2KjHeMnkq0o4ZsrGxRsq1fzpTQsFRwUCj9WPDmXH7mGO1rmP/kZTiyWBmp3YqUwRJq+OeumfbaUTuvKutGkU8Xt/Zwm6Xs28PyHDCE806QWxtTC9PWumJxxLbaC9PX+KOJIoTQUgYif/Qrmlj9RE+ke/emA9/5sMAo2I7BFQBx+nWu1PCvJLKc3/mCyVLudGI6c9dV09XMY/ex6TCrKho/rpkZ2/3wcRNzjMMaMjAMmazQMwO8Av9v3cwJ8H9D5Uv7t0wAAIABJREFUHxp/AvwoUAX+tFqtauB6pVKxKpXK0jTHVqvV1Qc/nYyMvZMJrD2y0gy424pYzDlcmM1jCMFWEHG7GSIE2zbF/REKP07rgi7M5phxbRKld40mjGK3b+Sl0ijAFIKinUbXuk12u1EaSwhuNdoTNSuec20W8zaOmdY9aQ0z3aa1MzlOl3LMujaX1pqcKeVYKjgIcV8Inil7BFJRsExudvpkeTs04E1NCQRJR0T1b+Y1YBkGTy0W03qwTjNkDWwEMWGieWaxhCEEq37ARpBQsMzemnXnlchUEFXXm1imsW0NJ7HM7x9XN3IklUJpTazSCE0sFVGSitLhcybte9Zj2BBivYkp0nknKq33Wm/HzLgWUaK2RRHv95f61I59uSZxs9yNCI+75R/mmbKHPZ/2t7Jv/iHnPJ9LS5/mxOaXBs0l9hoN6oqrPmE37jr77mc3KhXuxWcHDTaGhVK37iwjI+MdSdZoGKrVahOgUqmUSIXWLwO/1RFHAA1gBigD632ndl8XUxybCayMR5oDF1hHIQdXKs2aH2EKwYXZfG/zOes5JFpzu5OS1r8p7kYonpovcHmjxVNzqZV21wxi2rQroJPW5rCUc1Hc73XUja51RYQfS97YaPHUfAHLGnQkvOuHuzYrtgxBLYy5stXmVjMgb5ss5GzmPYt2rHAsg0vrTQwBtSDGEFCPBP6Wj1Saec/CswwurjUIpe6d3xUxo8RKV5Cu+hFSKV5ba/R6XgkhuF5rp722LKu3Xgs5mytbPn4see9SKR1PGAOCxVxqwb4VRuSHmjBfq7VxTYOlgrMtqjGtyDANwWLO5Xq9jWcZFJ20HuxqM8AxRaef1+A504q4LsMRmYoQ3G4FrPoxtkkvmnU8PxhF3G0Ovf5VnTXd2c2SkefqFz/GSv4HuffEf08k9WAj5+VPpe59za8QH/9xzOe+uP06e2Hue+4LmgfBflIBh1MWx0WusjqmjIzHkmeffZY//dM/RUqJaZpHttFwpVI5A/wH4F9Vq9X/p1Kp/Gbf2yVgC6h3/j78upri2IyMR5oHEcF6x+fgxkqBSDfA/RtW00gb6ppicFPcH6GQOk0P82wT2zTYCmNyWu+6me5np3od2J5uFiWS7642+PbdGkXX7h3f3XgPR05Ollz+5nbAh06mr7eiBIHgPQtFLm+2eml9nmWQd0zW2zFCQCRTN8GcnZpY9G/Ic7bgvYtlXl1r8NR8AW+XBrw3G22akUyt3xPJ6+st/u5uA9cSxFJjCDg/kxvY5J8qelyr+WwEkkaUpOl6CJbyDqZhcLcVESYKxzNGRopOlbaL227/rjc2miyXPAqOPVDLNXy87tRCrbcjvnNXkbcEfqywDIFlGLQTzUojYLl8Px10v5Gi/ojMmXKeU8XtaZ0LOZtrb/8Z5zwfO7hNrC2uvf1nLKy+iHlsfIPhvdYtreR/kPbS8zyzWKIZy17UshuljaVCyhhbtdITdkmh2zWF9bmX09cfpjDpj1yNS/0bjnKNqQ/LyMh4vHnhhRf4yle+AqT/DhxFq/ZKpXIc+FPgv6tWqy91Xv52pVJ5tlqtvgL8BPAy8Cbwm5VK5beAZcCoVqtrlUpl4mMf6sQyMvbAgxBY7/gcXNswQEOcqIFNcXezJ/Xgprg/QmF0TAi65wnSGp1+c4Ld2Kle50TB25Zu5lgmHzhW5uJag/PlXM+MI0jkQOSkK9zutiJMAc04IdGKMFG9XlvdzW1XlLx3sdSLoHTT9LriCranuh0vONzcoQFvIhXX6m1WGgFzOZvLGy2E1hQdk3OzHrECNNxuBWwFCctl3RMqQgjOlPNshQmeaaC17o07lulzWS673PNjpLpfB2WbBub6Nwhe/T8w/+H/OxQJDLjVCJBKd8wwNG5nTsMiQyrNtVqLVpzw/sUia+2EtXaE13H1W8g7nLMNtsKk96y6Bilz7u6RokkdBkelwA02GC51Ggy/uGuDYSEEJwoes549sZGLVJr1Jz/T+wwmHWOV5ZLH5c0WC7n0M7Bw7uOY7/vkjtca92XCqBTWhZzNKcTBG2fsVj811DB5R4brwzLDjIyMjHcOvwTMAf+iUqn8i85rvwD8b5VKxQEuAb9brVZlpVL5OvANwAB+vnPsPwc+P+GxGQ+JWzWPz339wqHcux6mMqXsJody/1s1j3ed3Nu5By6wjkIOrmkIFvMOd5ohV7b8gRqs1VbUiZiMTzPrpl0tl7qOe5ob9WB82lUfu9XrdHs6jUo3szpRsp6wGBpXV7hV5gtcXG9QtE3aiUrruAwxEFHpXidWqrPpNgkSiRCpoFSdqJwhxEB0bqeGxbca7TTaZZmcn80z79lYQnBxvYlrCFzTougIXNNkIeeMdOUzjTRidaPRHrBgT8WKg22aGCIeiBRGiaSmi7xx7L/G7qQ1LuRstNZsBQnvnisw6zlIpbhS82lECRrwk1QcWYZgpRFwpxWS6PSje7cVcTzv8sxikYWcQ5AobrdCNsOEC+U8375X4+qW36klE6A1hbt/xEV3EevEPxyIFB1EL6qJGgwPsdf7Dqc85m0TP5ap0UosubjW7DlH9hiTGrfSaNO+81e8x/Oxz3xiZArrwJcMH/rSVHWM+6I71hENk3tMEuXayz0zAZaR8UjxhS98AcMwUEphGMaRNLmoVqu/QCqohtnWyLBarf4q8KtDr12e9NiMh8MTT/z/7d15mGR5Xef799ninFhzqayuqqyt99PdBTIsCo2DFE0BAl7wKuPDM6IPMHIvyr0zjN6BaR981Ps4A94Hvegz6riMMjpzZwQdFgXZaUagbVrWXqpOd9HVVZWVteQee8RZfvePcyIyMjMyMyIzcqv8vnh4OjPWc05EZJ1PfH+/7+/2HX3+yWeeAeCWIzuzHXcc2fgx2JImF/thDG7rxPdSsc6NaiNuIEHc3GJ5ZWN5x7/xnMOlYo1vXV/ANnUsXe+5XfR683XQ4qFl1WbQvqyzqUVnhaxzu44V4spXa27YqJNiotTgaN5mvhFQ9wMmSo12CFz+eEopblQazFSbBEm1KJcyOZZzSCXNIyxdXzJvqLO9/dVKnUozJGMa3DeWI1CKZqiINMXRvM3FYo2TQzoK0DWWrPG0vCvfWi3YFxo+xcbiNyH+5U/xWP0wBVPhchWr/mV8ZXDBfDlTlQZ3jmY7qncGtw1l+Mdr83gzZRzTiBdWDkNSus4dIxmGnBQ68K3rC4ymUxiajmUYhAqOZG2enq3QjOL26SOOxV2jucUGKc0fYPTa/+DgPT+6pEp1pVTrq8PgWgxdw1hlgeHlFbLJh99H7eCZONxoYXvdqPWed3lw1zSNbMrEDEJsQ+PUWH5x7bM1AkMYqXgxZWcOS4ubRFiGzrGCw6NX57l7tMemIJu12ryojw7H/20toDz3naWNLjbz2EKIPUWaXIib0Tve8Y4dff4HH3wQgPe///07uh0bsRVNLvbFGFxN0zhWyHAkl+5pHaxulZvjBYexdIqU0fv6WWvN1wnCiJlqk4V6kydnIg5lHSxdwzY1pqt+1wpZa7senyoRRBFlPyBtGtw2lOZqpcHTc1WKjXiO1bEkPDaCkAsLVbLW4jC0yXKdYiPAHc2StkwcQ+dSqcZEqU6QVD5az62U4lplsTLiJ80+njuW51K5Tso0IAxRCqp+iIaGqWk0goi0FTe5aK3xZOpQbgbkUovNLjRN4/hQGjRFqRFwohDPnQoixVS1ST5l8NRsJR6SuACB0eS5xlNYWvyPo6WFHM87XK802mtM1ZoBtmHgR4ogUhzOOtw2nEbXNGZqPvN1nyCemkekFEMpi5Sh40cRNT9ItlVH02C62sQPI24dzsTt/SsXGUYjyGa4at/F+EOvwSCAMw8tViwbf49VDeHYG/oLEx0n7ivmKyW6VapGbIvZwv3cl4Qr6D3EdF3UN4y4XKxzy7KFpVfoCBj+Qz+OMfYWLFWFWhW+/2fxFSd/tuswyXXnMa7XGr3XcNPZir7TanOrBhGiZCihELva6dOn+fznP08QBJimuW+bXAghYltRwdozY3B7ndOyFkPXyOrrH8ZBrcGz2snrxYVa3L0vWe+oFihmqk2aUUTNDxlNW9w5kl31cU0NapEijCLAaG/vgXSKJ6ZKjKVTzNd9pqtNSs0AUBiaxvVKg2N5m9lawJG8zVgyHK4ShAzbJudmKugobh1Kt+c9LZ9DVvMDzs1WmG/6BGHEXL1BEMaVqmYUEUSKhUb8nGYSyJ5dqOIYGterPpdKNVRHlUrT4vlTGhqlZsCTM2WCSKFrGicKDneOZOP1u6KIsPAani3WSNWT0abH3gCArRQGcHmhxgVVbR/nlK6hFJwYcjC0+DVMmzr5IYdn52vx2lpKI1SK+H9Q9uNjARoomKk1CFW8PY0wwtAzmFEtDghWFl/PtqtM7YplEnJa+mqKgsZkl1b8rWPVrlTd/vJ4Py9/imeKQ9RIY9XPLz7QxCfj7n/OK9Z93q5DQc//Zjzn68yX1w8MXziNNf8E4dD/iu+YcfhVPmhW+7iEkaIzq/XTPn6FXqpPy4NSS6sF+2bCzmZDnxBiR735zW/mi1+Mv1PWdX1fNrkQQizaijlYu34M7iDmtGzUptfgAQ5lbC6X4nkoVjL8bsS2KDUV43mbkbTNmK4xnneo+QFPz1aIVNzW3Vi2e62wc+pggRvVBguNEDM5QU0ZOhPFOodzNkfzafRilZmqz6mxHKNpm0gpnpmvMFlqLHZQ1HWyKZ20UkSRImvVqfgBFxZqKGDEMZmt+dzXseitbRqM52wmi3VqQcj5uSonC2kUGpYOU1UfpRSPT5cp2CaRAseI28+fKKQ5OZRZMWxuslynHkS84HA8jKvY9LlarqOgHa4sXcfSk6CtjCUhJowUtTCeX3bfWI6MZVL3Qx6fLsVDFE0jPp563FKhFbZQxO3jLYNLxRoH0zZZS2O21mSy0qTihwzZJkbyXClDx08doAn49QlCPY11+uNxugSsh15DePjd+KmpOGQka1j5R35s7TDREWAmR3+86zymdlOUwv3xda3hdlrIrdpFblg/QD26hKM32w/rK6OnENP1C4XvfW3N+yxnDJ/iwNQnuOj8DCdT81gqnjs3ceF/Mly8xOXsm3prH79amGtpXd7rEL/VgtZ6BhHAJHgJsSuNjo7yyle+ks985jOcOXOGkZGRnd4kIcQO2pcLDa/VhW/bJshvwPJgqAFZU+dYPk2gFPMNf0llzNA1crZFyjIIomjdBW4PZWwuhzUuFWvUw4i0aXAwEzcjaN32SN5mNG2ja3FQGc85LNR9ik2fWhASRhGGHje3KPo+9TDirtEsh7LxY5yfqxAkDTBadE0jZ5kUmz45y2TUScULNgMlPyBnmYw4FiNpi5lagKnD9arPiUKaE4X49eocvnYwbTNT87lnNMv1SoMrpRoR4IeKyWKdi/NVhtOp9tysUcfionk6PlknroR8f76CDtw6nKbqR9SCeC7XsbzNkzMVwlChJcfZNnSKDZ+aH6JrGs1koeFyM6DSDBkLbPwwYihlcjRnkzYNSlMhz8xVuGs0h6FpTNcaTBXnOFj8Gob+uvaxMQg4UHyYi0Mv5KQ1F2+fMtZei6pDqKWYOfjG9jwqAOvq33JSGTy5AOnpj6CPvA6rPt8Obxx7A/bEJ8k0prmgHeLOjFoyB+uA1Xv11dA1jC+9Mv6lW8CxhlZvDgGMf3SEyctw9pY3YehOu/vh7dWHuXr3T3H27KcxDIvwwP09z2NcorMDoL+wuXlUnTYShDrbuMtQQCH2nDe/+c1cunRJqldCiP0XsNbrwjfwCfIDtFowvF6NFzYOlWoPnWrtgx9G+EGEprHqAremrrUXJm7tehQp7hzJkEvFQ7LqfoAG7QAXz4FSZFMmWdvCjxRXy3UMDcYyNmGkeGa+ygHHopCy0DUN3YgXZn706jyNIMTuWAurFdgO5x0OpFPcohQq2Y+n56rousahrMOxvE65GXCpVONYPk0jXBxu2Ro2V0vaz0/VmszX4yBmGTqhiucBVZMhe4eyDhcXajimRtrSlwxncwyNtGW0K3VhpJJttNBmK1woVrm1kKGpIhpByPVKg4of8OjV+Xab/tuHM8zWfE7mHdKWiUJRbATYho6uwVzd55HJOQw9nl92cuQuxm/7jaUv+pmHGFeKyYffF7dYH315XHG19LXDRHJC7n/sDgwri3X89e2rlIIb0TAle4TLB36SYuo25u1rDFFHa0zDxCfxq1OYtVly03/H2YNvwEiPEabKGwsxm6AFJY7O/Q2HX/0/VnQ/PAocvvYh/NIFrPxtq3dFXKv60wo1rflUnfOoVrt967JBhJ5+A5QELSF2rdHRUT7wgQ/s9GYIIXaBfRew1uvC163Ks9l5WoPQSzA8mEkxVWliahrDTopIKS7MVwmV4nDWXtkUIGmYcalYoxlG3DmcQdc1mmGIN1PhmbkqzzmY52qlwXS1SbEZUPFDslbcgCJlxvdvBiFKKepBxJPTZSy9ShApNE1x90iWTEczDMc0sHSNZ+ar3D6cwTaNZE5VjYJtYhs6pqZhJ93nIJmLFS62h89aBqVGwGM3iljJNhxIWxxMpwgjRdo0CMKIG5UGBzM2acsknzJQCvKWxbdvLHBxocaRXLrrel6t43K9OkfdD3EsAz0ZW+n7AZqCuZrPbG0BjXjb0pbOsG1RsE1uydqkTTMZDlrjWqXJySGj/f56dqHJ8cJiJ8UgVDy7UOXEUKa9z8vfd0erX+Nw9RH8e17b23sxOXG36hOEfgX/8qewmjfAPsBk3aCSHuOugs5BY47L+hgX9Xs4OXaI4enP4NcXuOiPMDb1Fxyd/TjjpS/hj74Uy3119+ddLyR0BJwQE1/PxkHpxhcXb9NZNeqs5Ki46mZ86RUYnYv0Js9p3PgKBkAws7Ly1G27OitUvbRZX8taAaxVgfro8PqP29qm1n1g9cqeEEIIIXa9fRew1urCt7zt+E7N0+qml2AYt46H83MVIipJ1SVu7NCt6mDoGiOOyeVinR84mI8DgabRUBp3jGQ4N1Ph8akSGcvkvrE8N6oNZms+lg75lEUYxQGu3AwZcVK4B3IsNHwMTWOyXKPUjKtUreOllGKh4dMMI+YbPt+4Ok/GMjA1jdG0RTOMMHWNih/GXQE1qPshxUbA8cJiZfF6tUEhZXIkZy8JktfKJQ7n4i51Q7bJtUo8NyyfMtG0+Pkdy2DItpitNamHIVnLXLGeV+vYDNsm3mwZdzSHYxnU/RBvtsLYzKe5b+4vqL/8M6Di91SE6hp8ljd7KDd8TF3n1qE0hq6RUjpXSzXGkrXTVn3fvfLLGJpGv7P3DNWM5zEV/iUnNRNdaUzpxznupHBqT6GHdU5EHvgVHpvKMOKcJpj/KkM3Psmhub8j1FJxuIoqm/qCQSnFZOaHmSncj2Gk4qF+QZ7x2U/Q9VGD8tLhe62fl4elll6G93WGqrUu7zZnq9vQvc7A16/lz9HPYsX9PL4ENCGEEGLb7buAtVYXvs45LbttnlYvwTBuHZ/mSM6hHoSggbNOC/ixdIoblSYLzSA+gdc1MqZBwUqRS9WZqwe4yTFoBbgL8zWaUQU9OeE3iDictZOqlcaQY5FLmTx6dZ4LC1VuG8pgGToLDT857hlODmWoByEX5qvkUgbHCxl0rcZUNV6oeb7hE0YRV0oN8imDY8kxb1Xy7hnN4kfxvDMNGHJM5urNpIthyJGcw9VygyBSKKVQChQQhHFXQjNpSLFW57lTY3memC7x6LV5LD1uuT5sm5ya+8/oRGSt/rtHmprG9WqDc0mb+M7gDgN833VUg8aZZ/LwSzhb81Eo6o1JnPozZOoX4210DnCSGsWUia1rhAd/mIph8o2xN2FYWXJjpwgVHCjVln7B0Efr8Mlyndo9v8y9nZ85O8tk4T6OvvTfdX+8TsuDTCsUdRvet1pAgsXbfnR48efl291L2Gk9742vrBw2+IXTcQXKX1j6HGuFnY1W0YQQQgix6+y7gAWrtJDuOMltBvHwss5Odzs9T6vXYNi6bTbV20tr6Tq1IORKqU7GijvEjWVSWLpGEIFtaEvWl2oFuH+8OkdKh0hpLDQjzs9XaAQR+ZTJyKEClqFTsE0sXePsTLk95+h43mE8F89rckyDO0eynJ0pM55T7dflUrGOniyYPJa2OVZYPKlvVfJSpkEK4m6FSqGjMVGq88R0GTsZNpi1DK6VawzZBrmURTOIuFisYhs6jSDer7WaRei6zn1jBarNgEApsl97AynqG2pA0Nk9crV2/Vs1P1BDLVnc+ekbF0jZCi19OL7BsTfEa5Fdm8c2jLhamXkZqdJ1DqQtnGyGlKFv+AuGVdfycqqcLdzP4Uh136+RfwJTXwUz11voaAWjjVaVOp+3U7c5W53VrI3oNi+s366E3ch6WUIIIcSO25cBa7U1qZRSXCnVuF5p0gwVZT8kiBSZZHHbftYe6jSoeVzrBcONuF5tMGRbHMxYjKVtFLSH242mLSZKwYr1hlrhqNIMcCyd5x8eImeZNMO4xfoT0yXuOZAnUnCikAGg1PCpByF526LUDFGEOIZOxlqcm+SYxrprhS2v5Omahq5pzNebNIKIFxwewmnP66pSaQZ870aJjGUQKYVtxIFSAedmygzZJocy9orj0nWoXuaFjFe/3n1IG/29zt3a9fc7P7Cnbeg4sTZ0DQeD4WMv50KouK36lbg7YBhxYaFKqOC24Ux78eR7D52Ij23DJ20ZK4Nej63DV13L6/jrMdKlxf3aSOjoFnRaj/PR4e63a81vWn775eGktb5Vt+fs3LZVOiBKAwshhBBif9qXAatl+UnuRLFONQi450CW83NVsqZBI4yo+iHZlNn3QqaDnsc1qMWKWzqH25X8gOlak5ShM+SYzNSbNIN4SNzlYn1F1WzYMSk2A+4bzaG01sm/zsnhNN+6uoBOiQPpVHv7Sn5AM1RkTQMnqZSVmwELDX/FMY3vEw/JA7197Fr7vLyS1whCLszXOJp3cJLOhJahc+tQJu4Y6JhMVeMW9qEi7rgYRRSbIdUg4smZ8orXZaJYp+z73FpIk02ZREpx0fxlJi2do4+8Nt7Q5IR4UK9zr/MDu1lvG5Zcr0G5GTCt3R+vKzZTJp8y4oqjoVNPujC2tkEjDtUb/YJhw2t5QW8L//635M9Y5/C/zt83Or+pWyVs0CFoqx5PKldCCCHEjtnXAatFKcVEqcazxRruaJZaEJG1DC6XahzPp6kEIWYQcrlY72ntoZatmsc1iMWKYelwu1FDp+IH1IKovdCwbercVchytdJoD/NrDdvLpwwsvYlh6ERKYWg61SDuiGcZOlO1JkNOCqUUkYK5esDRvMNEqc6RvE0Qxm3Pz8/VGOsIYsuDQhBG7RbuZrKo8qhj4piLbdWbQUgjVO01sVosQ8c09KS9e6Yd0K5V6jRDrT0EtPN1Gc85XC5WeXquSiFl8sxClShS3JK1OVFwODdb4TAmBkH7eQb1Oq81DHTEMZcEzuXW24bW9e5oth3ALhZrWLrWrjI+MV3CDyMMLW7D3wxCDF1H1a6izzyGf+R13QPROif1fa/lNehQsNaQv26X9xpONnu9EEIIIW5KErCIT07jTngWtyQL4hrATN3nqbkKVT/ENjRuydo9D8fbC+ttLa+Y5FIWGUvRCELSpsF4Lk0zUhzKxHOmpqo+lqEzW2/SCOPKUTWIFwKu+CGRUmRMHcfQOTqcpuyHTJbr7UrWsZzNN67Oc6lYJWXoNMKISEXcPbIYQpYHhYsLVebrPieH0gw7qXZwsI2402Gri9+52TJBpLCMxWPaWflphdL1XpcogmvlJqNOiucknRWrzYCJch2tFocg//RnMZJK2aBf5+XDQFsBs+JDsRl2rY4t2Yarfxtvw7E3LFl4ebra5ORQ/JpoxA0/bsmkuLhQA+L9GnUsvJkyB7Mp0qbO03MVDmZSZJrXCdXa89XW1OtaXmsEm67DL1u3V0uHHq4aqAYxx2mvkHAnhBBC7Jh9H7BaJ6d3j2Z5arbSDht52yIEjuRszs1UODWWJ9UxEWm9+TYbnU8zqPlazSCilgSlzu3u1K1iEkaKiVK8YPC52aWtxZ8zlmO67jNVaTJXD4iU4ns3Sjz/YB4/UmRMg2eLNYZtC8swOO6keHquysF0vPDw49NlMpbJcw9mMHSdZhTy9GyF794ocf+xUSLFkrASRoqFRoA7mqMSxAHO1DXGMhaPTZWYb1ooBQfSFqNObw1A1npdNA2mqnUsQ2M8b7fnd2VSJkdzDufn4s6JnRWcQcyb6rR8GOhUpUkjjFbsV2d1bL1tqAUhgVJoxK3nWxWscjMgWHgK/4nfwHjgE4CiEUZcLTUwDY2F0jTTJYNcvYJVm+JA498zXv06nPny0o3uobGCpiVreU3+Ef6Nl2Kd/nhP7+81hz72fFRZsT0DuZ3YV1zX/TaQjDvlAvCHwO8AAfA5z/N+3XVdHfh94HlAA/g5z/PO78T2CiGE2Dn7PmC1Tk4d0+gSNiIuFxscyqbaIaXX+Tb9zqfpfNzVOuj1IooinpguMd8IlrYWH8ujd5nr0q1xhobCMQxuHc5g6Bo3KnUWGgHebIWMZXLqYB5diy9/dqHGP1ydxzENbFNn1ElxIu8wXfdJJS3iIxRDtsEzczVedGQIy9CxdJ2UoXHfgRyPTC5wqVjlUFLtaR2v9mtjGVSTgNUIIjQ0RhyLO0eyGFpcWXFMjbSlr9sAZK3XxQ8Vuq7H1TzLpNwM4vW4dA1T1/AjxS2ZpYGt39e51wDdmoc211i/OmbpOuG1r+DXNKz61fgBJj6JrwxC+2WkDD2pwi4+p6HHizlX9QJ61CCMFLP1gB+4pYBSUAtC7q49SYTOk/V57jr/i2gHX0aE0fdaXG1nHsL4wmmMaCFePbpljYC25tDH1SpTqwUkmZckNsh1XQfA87zTHZd9B/hJ4BngU67rvgC4FXA8z7t/ZMF2AAAgAElEQVTfdd2XAL8FvHHbN1gIIcSO2vcBq/MEuTNstNqK31rILDlJ73W+TT9t1VuPW/VDThQcQkV7Daj5RpPnHCz0HLKemC4RKfjBw8Mdi+OWeWK6xHNvWdkVbXnFREfj3GyZW4czmLpGuRmgaRq3ZFJ890aJ25PLNU0jZeqcKKS5sFAlpWlxlc/QKTUCdKAZhu2QMZSysM06erKYcURcUUlbJmlL50bF50g2vSSstF6buh93/dPQqIcRWdMgSoYGGrrWDhynxvLrNgBZ63U5mLGYrfn4QYSV3Le1zlZrCOSxZXOqen2dN9IIo9fqmKFr8Rwn+wwnlYmlBfEcp3qGA8MWmhYfq4lSR7OS4rNMVBRWbYJo9pv4D72J6Mi7uOa8hLl6EG+j8yOMOibN6mOce85/I3XopfF2L18Pq5e5SxtoH74XhtkutyJAS6i7WTwPyLiu+znifzd/DbA9z/s+gOu6nwVeCRwBPgPged4/uK77op3ZXCGEEDtp3wes5SfIR/NpDqRTXJivcutQmuMdjRP6PeHrta1663FPFJwlw7gKKYtvXV9golTjeNKIYC3NIGK+EbTDFYBjGbijOR69Nk8ziNYcLmjoxpIOcpVmQBApZms+FT/E1DWqQURIQM4yQMWNMBxTp+aHPDld4lDWwTF0mqHi/FyVUScOGbZp0Awj/DAkk8xf0jSNuh/iR4q0HYeuEcfk/FyF24YzOKbBkG3izZY5OZRGoQijiImSvyS8dAYOxzTWHZK31uuiaxrXyg2eXahx23AG29SZqzWZr/scL6Qxjd6qgMtf5400wuinOjZ+/2/E23CxgmFYhPb9HBiOtyFSkDZ1LENb3MZynSFmcZqXmRo6zcwtb6Ps3EGlWOdY3uFEIU0QKb57fQEje5T7mMEeyw9uwe3WWlKwakDzl3UzbB+X5cMvzzy0NNgsf65tWBtq1QCNhoYa2POIHVMFPgj8CXAX8HfAfMf1JeB2oMDiMEKA0HVd0/O8ACGEEPvGvg9Y0HsQ6ne+Ta9t1f0obmQQKtrhCuJwVLDjFuPjuVUWY+1QC0IsXW+HqxbHMtoLCq8WsNr7kpzUN4KQehhRbYZoGtw1kuGp2Qp5y6AWRszWfDKWgR9FNEPFkZzNXD1gutpAT9q2a2goFXGlVGOm5uOYBt+9UeJEPs2tw5l2da2QMkHBVKXJbN0nUIpHr87H+2IsLghs6BpzdZ8ThfSS16af9vmtE/HDWWfF6xJ3KLSSENfgRrWBoWvowJGck1y38nVY73XuGsyv/i0nlcFZ+2WrVmL6qYK2t+Hah/D1LJb76sXhgBqMZVLU/Ii7R7PxHYbvYqJYx5h5msbtv8C9t/0I8w2fjGlwtdJgslznYLIu2vjwAayFCCY+uaR5Rud2h5HCP/3F7gEHloaoVrhaJ+D0EjDVF17BZOalzNz53oEshbBRKwL05U9xcSHDZDDM0dmP7+5K1m7ett3jKeC853kKeMp13QVgtOP6PHHgyiQ/t+gSroQQYv+RgEXvQWij6xSt11bd0vXkMaKlDRnCKB4KZ/TWLCFtxoGn7odLQlZcJYpIm+vPnmmd1D+7UGXINplr+Nw9kqEaRGQsgwvFGkdycVdBU9OYrDbJWQZXK03c0SyaFg8BLKRMFPDo1XnGMinuPZDD1DUmS3WulOpcLFYxdJ1CyiRt6DSiiEYYtVunN4KQZ+Yr2LrByeG4eudHEVPVBo1AtTsGrjXsstNaQ/SAdghsXXc8n2YkHQe/uYbPXD2g5Hfv4td57Lq9RqsGcy1ctxFGv4tLG2e+0DXgtB7nqdlK+3GGbIOSc5jjTq09l8sxDY5kbbzZMhBXKNOmQaSZ6KoZb3fHFwq6pvc+9LEVrvyFuJq0/MR+eXv3zoBZeSheFPnIjy15va9kXkrt4Jm1K4NrDGEcRFOZrgFaCznpVDl78I0cnvv0xuetid3i7cBzgV9wXXecOEhVXNe9g3gO1muAXweOAf8L8JFkDtZjG3mymatX+MQf/s5ANrxf1VIRgEy+sO3PPXP1CgfuvGPbn1cIIQZNAlaH9YJQv/Oq+nnesbTNlVKDQsrCsYz24w7ZJgtJw4r1pEyd4WRInTuaWzIHa9g2161etYznHCaKdbzZCrahUwkiHEPnntEcE+Ua52crhEphGzq3ZG2O5R2+P19lNJ1qd96D+MQzUorjead94jmed0hbOt+7USKfMtA0jXTKoFZb7JQXhBGT5Tr1IGIhDCj6AQczKcZzDseSNZ16DRwtaw3RA7pet9CIv3huBGpTa1wtCeZJG3VqV/GVSTj/FazvfAjOfKHrfQe1uHTn4zTDkOmqz1StSXP4hyiPZrH9CKUUuqaRS5nx8MyUydTMJXzdQa9NxO3QO5pnWLre/9DHkX8CN75CqKXw9SGsqILRrYKSXDb+yi/Hr/eCwjBswlSyKPQjryPEYGbkLdyrnsG6WoqP9SoVtuXWnBP3xVes3J41dA3Qx96ABRjlz+If+tG4S+MXTsf/3y2Vom0YOnkT+U/Ah13X/SrxKgdvByLgvwIGcRfBR1zXfRR4leu6Xydeo/tt/T7R7bffPrit3oCF65MAHDh8y7Y/94E779jx/RdCiEGQgNWnfisKvTpWcJhvNPnW9QUKtkmkYMg2aQZRX+Ht1FieJ6ZLPHptfkUXweVW+/Ze0zSOD6WJiLhcrJMxddKWSRgphu0U+ZTJswv1duv6OEiBUqAnjxMpRanht+dfdT72aNpmLNPgeD5NLhUvoFtsxnO8rpRqXFyokbYMxvMOoFGwDK6UG+2T9n4Dx1pz556cLqGS42YliyZrGhwvOJydKaNBu6rWeb9+miwsCebKiCsxyuSiP8KB4sNLFi1e6zEGsbi0oWvMVvykWpjjqdkKWdOgEUagQbkZYBv64sLOfpmpYpXRSIu7W3Y0zwD6a0Jx5qE42Dz8PmYK92McTtbDOv+bjFe/3rXtuvbFV3AUODz1ML5RwBp9Xvt4+XoWw7SxtOKS+6zaIr8jOKwZDPs8pmtWtkMfK6r0+Yhit/E8rwn88y5XvWTZ7SLgnZt5rne84x2bufumPfjggwC8//3v39HtEEKIvUwCVp8GUVHoFmw0TeM5BwtMlGrJgr4aC42g7/Cm6zrPvWVozXWwOr+91zTwQ8XBjMWxfHrJsK4ThQzFRsC5mQpH8zaGrmNo8Ul1Z+v6zgBxouDgR4qyH3ClVKfuRxQbPkO21X7s1tDHVgt0iE9QLxVrNIKIjGlwaiyPRtzFL1SsOGnvJ3CsNXcODQziNuyVZkA9jNoL8bZeo0GscdUO5vbL4mA+/xUOFB9m/P7fgG2cK7Q8bB5IW0xMX+JYVicwDxNpEWdnqjTCiKdmKxw6dC+gcfbSlRXNMxrh2nMS62GIEWlL3ueT5Xo8pM+pYtW/HAe2oRcyWXySo53zs1pVFSvufGmoJkYwDa0weuYhrEgRPvkpfKeAdez17edfb8juqoG78hBnFxSHpx7GUM2eKzprVrZPvhLj2gfjx9ptlaJeuj8KIYQQom8SsDZoIxWF9Vp1a5rG8UKG8dzm54WkTH3VIYGT5TrVZtgefpgyNC4X68zX/SUt4duhrxjPm7KSyka30NcKEN++vhC3cNd1DmVtDrQXAIZhJ9V1SKWha4w4JpeLdU4dyDNRqccBJozIp0zKfkgmaUfe78K9sHaFAQUhioWGn3RwjLer7oc0Q0UzCPqec9fNimD+nQ/FlZhu4aqPE95+5xAtD5vjOYfJ2UmeahygalWxjXi46ljGaq9jBnCkS/OM1Y5rMwhZqPs8nVzeep8fythxsHGqWFoYP0bzBietKJmr9F9WzlVqdRps6TgmS9rTJ8/Vy5DdNefEGTa+UYjD3Nx31j2eLVtV2RZCCCHE3iMBa8DWOuHtZw2tQQwHW237Zmo+Q7aJHy7OLar7IWdnykwU6xwfWtyW1nDB8fzaFTtN0zicdZiuNrljOEPaiqtTSiko1nhsqsSIY60a0MbSKW5UmzSikKofUG4GpE0jXnPLD2kEYd+hpmWtCsNYJkUUwYX5GvceyLWD3UQpbld+rVznwnyV24YzA5lz135tV5lz1auNrKsFK+eDacDR5lUORCmeLB/h1LUPkTrzmZXb3aV5xmrH9bGpEoWUibvsfX65lDTTSMIVx94QdyZszGBkjuA/77MYreGky0Pm8sWEE+329KsFmy5hddXAfeTHCFNlrNHnQQ/DNjutWdne7ZWi3bY9QgghxB4nAWtA1jvh3S2LpvpRhKbBQiNYsi2OZXA0b3OlVOdQ1iZCLTlJ7CX0+VGEaejkbKt9maZpnBzKUGwG7TlXq3VoDMK40UI+ZTJRqnFrIYOuxfPIpkvNTTUSWavCUA9CrpTrPDVXWXFduRlX+PqtTGyoO10fTQc2sq4WrD4fbCIY4lDx70lR721bE8uPaxBGBFHEcw+unLf2xNnPoKHwU1NYWgDf/zOImnGzj9o01kOvWT14rhICNjJkd91mNXP/GN/QT5Yz6iMYbeWXI0IIIYTYGyRgDch6J7z9rqG1VeKW8IqUsXRbwkhhaBqRgsenS9im3nNVpPOxW5WB1lpYlh43jlBL5lytdL3awNR1Fhohtw9luF5p8sR0iVApNODYsrWvOvUSZtY6EU8ZBhnLaK8R1brODyNCRXuR515O4DdaWerHZsP6mvPBvviKvjrdLT+uYaR4tlgjtWxJAMvQsQyD7PyjXCzcx0lrDgsWm33Uv7m02UefVZUVwWZ5WP3o8JK1t9Yc0tcalti676BIpUgIIYTYFyRgbVDnST2s301to2toDZqhaxzMWFwu1tvrZYWRotwMCFS8FtULDg/hmEbPVZHOxx51TL53o4ihaVimjh9EhEpxKJtqn8guDyqtwPDcg3mmak3OJWs1mZpGPYz4oSPDSzoRtvQaZpY/5/Ig26poTBTrnBxKt8PVyrli6wfgjVaWgJ6Hkm02rPc1H6xHreMaRmr19/mB+zl296u4/sivcLbVRfDazjT7GOiQvt069E8IIYQQO0ICVp+6ndTnUwa6xponvI5pbMkaWhtxLJ9mvu5zdqbc7g6oa4orpQZH8w5OEmY2NoRRwzZ0DmZTmLreXnNJqZWL+bbCUCswpExjxUnvudkyquPRO8PStcraYaafatIgmhT0VVnaxEn5oML6kvlgrTWaNtnpbr3hd6ahc7T6NQ5XH8G/57UDCXdtnV0IzzwU/24NxUP9Oof7LWuUIUP6hBBCCDFIErD61K1CcWGhSrGxfre5fk/iNzSPpwfdugP6YUQzVJwoLK2y9DOEMYwUs3Wfe0azNENFPQyxDYNbshrnpkscyaW7hqHDWYcgjCg3/HZzDEOPK2jxMD1FEEZcrzbaYckPI6p+yIsOD60aZtYLYMuPyWbb7w9sGOhm2oJvc1jvZt33+ZmHMIhXZ91ss48t00vAlIV6hRBCCNGFBKw+rFahuG0ow1ytuW63uV5P4rdjHs/y7oA6GudmywSRwjIWn6OfqkgrYPiRIlSKESceFtgMIhQaYxmr60K/UQT1IO5iaOlx9WvMsXh6rkrVD3l2ocZC3aeQMrlnNEvKNKj5AedmK0zVmkuOZ+caTBuZp7SZikZPlaUBnJSHkWLUsZiO/JUh5pHXAarvk/zwgS/Hx/Ch18QVpU2EhG7vc4BGOPgvC4DFylWrSnXjK4tzrv7Z/NYGnz5auQshhBBif5CA1Ye1KhS5lNlzt7n1TuI3NY+nT53bstmqSNxAI6LsB4yl7fZ9mmGIbegoBZFS6ElItAydWhBRDQKef2iovUDxZLnB+dkyuZTFiw4PYeg6j90ociRn40eKFGCbBkeyKb4/V2Wq2myvtzRkmwTJ+lbb3VRkqytL3YL3iGMylk51rFml1n2cNR/z8LvjOVFKbTrMG7qGrulb/mXBQPUTgFvDEDuHJW7Htm3meaTKJoTYYq7rvhj4Tc/zTruueyfwYeJ/nB4H3uV5XuS67q8CrydeE+Pdnud9o5/bbvtOCdEnCVh9WLNC0We3udXsZDv38ZzDRCles8oyNNQqa1atxtA1hmyTyXKDUSeFQTyUb7LcQNcgVGpJwKr6Ac0w5GShQMo0SAFpyyBvmXzzWsCpg3lSpkE9CLFMnWEnxXzDJ508RsWPsE2du0cyZFIWdT/Emy2ja+CYccOFepAsaNvZGXALm4r0MjwOaJ/otitHkVr3dV0teM/WfY4+cia+UZ+VsRWPOfI6Li68YmBhftNfFvSyH53HtFvY2YowsTyIzX2nr+6LQghxM3Jd9z3AzwCV5KLfBt7ned5Druv+R+CNruteBF4OvBg4Dvw18IN93laIXU0CVh96rVBspjKyU+3cW5WMuXq87pMfRoyl7b4rDcfzaa6W53liqoTV0erdNmyulBoM2RaGplho+JyfrcZBKQgJlSJjGeiaRsrUl+x/K9hGScv2Vtv3YsPnSM6hFkTUwyYKODmU5uJCDaUUGopvXVugYJtECoZsk2YQbek8pZ6HgaIxmXkpM9OlVddN67z/WsH7iekSw/oBnGhhxWLAa9nqML9b1n7ry0YXBW61dt8Kg5jrJfPFhBDb4/vATwB/kfz+QqC15sXfAa8GPOBznucp4JLruqbrugf7ua3neVPbsztCbIwErD4NotvcWnaqnfughiWayQl0pRkynrNJWyaRUjw7X8XUwZut4EcRjSDiSM6m2NDImgaNpGlFNmW2w0VLK9hemK8y5Jjomka1GRBEimHbIm0Z7cqYrmlcKTe4XKrhmAbHC2lCBUEUcaVUx9Q07kzWu9pK6w4DffGn4+O9LKhfKdXRNFYMqRt1rBXBWylFM4yoBSHnn/OnaGgcOP+bjFe/jnbmy+tuo//Qj2MceRfWLa9ecvmgwvymvizYSCDYzrCw0SAmhBA3Mc/z/tp13Vs7LtKScARQAoaAAjDTcZvW5f3cVgLWHvClL32Jz3/+8xu+/zPPPAPAgw8+uOHHeNWrXsUDDzyw4ftvlASsPg2i29xatmIez3rdCAddaWiF0GcWaktCwp25LH6oeHy61F5r60qpxkSpzrG8QyUIMYOQy8U6w7bZXpfKMnQOplNcK5eYqzeZrvkEYUSkFJautYMVxEE0CCMWwohTY3lMXaPSDFFKcSzv4M1WmCjVOJZP79gcoLWO97euzTPipFYE3WnVXBG8q35IPQhJGwbPGSsQKcXFhTNMTsHRHrbDiiqEYXPLhlHulrXfNmQ3BaZBhDkJhEKInRF1/JwH5oFi8vPyy/u5rdgHRkdHd3oTNkwC1gZt5fo5g6qS9dqNcNDDEtcKoRHxvKnWWlutfX1qrkLVD7ENjVuyNrcP57laaSw5BodzNocyNoFS7XWwLnWEsFYYGbJNKkF8Ul9pBks6Gt6oNik3wy1pGNKr1Y63oWtEwHjO7hp0R+zF4G3oGmU/YKEecDBZxNlA4+TtL+fs0As5vNacri+cJsSkOfs9cB7nW5fupWApIvuWgQ6jXPFlwdW/xVcGF7On13/8JACEXziDr2exTn989w0nBAkqQgixtm+7rnva87yHgNcCXwbOA/+P67ofBI4Buud5067r9nzbndgR0b8HHnhgR6pHu4EErF1oUFWyXof96Wg0g5BGEGKbi0Fqs5WGbiF0eVWjta8H0imenC5zaixPyoyfb7Vj0HrTrhZED2Vsnpwp0whC6mHEsG21KzORgjtHMjw1WxnIHKCNrFXWrbITRopSI67Mpa2lH8tW0B3LWMwmC0RrwELD57bhzJLgvV4oVkoxmflhZgr3Uz2cJpU7zh3pANNyCO0UV0oNTB3uzA1mGOWS16hkE4Y+B4b1db8saH85cPjdGEaKcLq0c90Hd0vVZxDPv9P7IITYb34J+GPXdVPAWeCvPM8LXdf9e+BhQAfetYHbCrGrScDaxTZTJetl2J+u0a5wNULFk9NlbhtOM2RbBJHakoVrDV1jxDE5P1fhtuEMjhkvKDxRrHMom2qHq87br3YM1gqiB9IWzy5UGXEWw1VrfxzT2PQco82sVdZZ2TlRcJiqNZmqNGlGEUEEE6UaJwqLQxhbQTdlGBzNmxzOOtSDkPNzFW7J2EurkeuE4slyndo9v8zdBYenJp7m7lGbRmocQ9NwLJ0h28KbrRApMAbwsmtffAVH577D4ZEX4c9+FyssYly7P75yjZP99pcD975uy5cqEEIIMTie5z0LvCT5+SniLoDLb/NrwK8tu6zn2wqx221ZwJJ1EHZWL8P+ZmrNdoXL1DUuFWs8OV3GNnUsXR9o8w5YDCWzNZ9AKR69Oo+l66RNnbFMasPP1S2EjeccJop1vNkKN6pNoo6W84OYA7TZpiCtys4/XlsgbRmM521ylkkzCrm0UAfg5FBmlS6VGtmUyVgm1ddcvc7QHSqFYRg4usJKmcw3fDKagWluTbdKgwAj6G1Ux67pPiid94QQQgixAVsSsGQdhJ23XoMBHW3FSezJoQyHsjZPTpe5ZzS3opq0Wa1Qct9YHsvQaQQhzy5UyZjmwKsSmqZxfCgNmqLUDLh1KE3GMgdSmRtEANA0jcNZh6lqE3c0i23GLeozykAb0nhsqkSxGay5Flm/c/U6Q7ceKUL7EH46h6Vr7fb3YaQG04BieTgBsIZ6WpB3p5YqACRECSGEEGLTtqqCJesg7LD1uhFGqK4nsY5pYJs6EWqVR96YbqHENg1uH85ydqbMeDT4qoRK1s2arwfM1ooYuoZOPLdrM5W5QQUAP4rDb+ecK03TGHZSjDgWx/Npcilz1ePS71y95aG79f44lndQxK/R5WJ9S9cJ68Wu6T4onfeEEEIIsQFbErBkHYTdYa0KR6TY1pPYnahKTJbr1APFDx4ZRtc0an7AZLmBptF1nlSvDSsGFQDWepxIsWa46tTrXL3loXs853CpWONb1xcGPyx0E+FkK5YqWJcMBxRCCCHEgGxXkwtZB2EHrFXhMDS29SR2u6sS3SpmOdviNtNYMYyv34YVgwoA2xUkOoNjt9B9vOAwlk6RMoxd0wp9qxf07ouELCGEEEL0YbsClqyDsINWq3Bs50nsdlcl+qmYbaRhxaCO3Va+BmsFx61aKHuFDYaTrV7QewUZDiiEEEKIAdmugCXrIOxCy09idTQi1MBadC+3nYGu14rZRhtWDCoAbGWQWC84blmjiAHaygW9hRBCCCG2wpYFLFkHYe/QNZipNTe0plM/trMq0WvFbLNzwwYVAAYdJHZNq/O9RipXQgghhNgkWWhYbHpNp35tV1Wil4rZrulYN2A72up8mV6bh4jNkeMshBBC7A4SsPa5m7nS0UvFbEc61q1jECfKuyE49ts8RGyMHGchhBBid5GAtc/tRKVju79pX69itls61imlmCjVmKr6WIa2ZJHhfk+Ud0Nw3O7K6H4lx1kIIYTYXSRg7XPbWenYrd+0b3vHui6UUjw+VaTUDCnYJpGCIduk2gw3fKK8k8Fxs5VRGe7Wm5u5Ai2EEELsVRKw9rntrHTs9m/ad7Jj3USxThDBCw4N4VhG+9ikTJ2Zmr+hE+WdDI4brYzu1hC+W+2muXZCCCGEiO3NGfxioMZzDmlL5+xMmSemS5ydKZO29IFWOlrftLdCHCx+0z5T8wkjNbDn2mvCSDFda3A0b+NY8clw69gsNAI0LT6R3ihD13DM7V1EuLMy2mm9ymhnCD81lufeAzlqfsRkub4dm73nbPQ4CyGEEGLryL++ol3pODWW547hDKfG8hzNpwdaMejlm/b9yo+i5DjoS4KmZejoGvjh3jtR7qyMtk7+16uMSgjv30aOsxBCCCG2lgwRFG1bOURuN3S1260sXW8v7lxuBuRSJoauUfdDio2A44W9OY+m3zlgMtxtY3ZLkxYhhBBCxCRgiW2xG7ra7VatYzNVbXIwk2K+4RNGEVdKDfIpg2O7YH7aRvQ7B0xC+MbshiYtQgghhFgkAUtsG/mmfXWtY3OpWE+GBUaMpW2OFfZ+c4deK6MSwjdnJ5u0CCGEEGKRBCyxbeSb9tXJsYlJCBdCCCHEXicBa4+4mdYFkm/aV7ffj40ETSGEEELsdRKwdjlZF0jsR/s9aAohhBBi75KAtcvt9sV5hRBCCCGEEIukLdcuJusCCSGEEEIIsbdIwNrFZHFeIYQQQggh9hYJWLtY57pAnWRdICGEEEIIIXYnOUPfxTrXBWqFLFkXSAghhBBCiN1LmlzscrIukBBCCCGEEHuHBKxdTtYFEkIIIYQQYu+QgLVHyLpAQgix+7muqwO/DzwPaAA/53ne+Z3dKiGEENtJ5mAJIYQQg/PjgON53v3AvwV+a4e3RwghxDaTCpYQQggxOP8U+AyA53n/4Lrui7bzyb/0pS/x+c9/fsP3f+aZZwB48MEHN3T/V73qVTzwwAMbfn4hdpud/kyBfK72IglYQgghxOAUgIWO30PXdU3P84Kd2qB+jI6O7vQmCLHCXh56K5+p/UkClhBCCDE4RSDf8bu+neHqgQcekG+6xc2oPfTWdd2XEA+9feN2PLF8psRGyBwsIYQQYnC+BrwOIDkRfGxnN0eIm8KSobfAtg69FaJfErCEEEKIwfkYUHdd9+vA/wv86x3eHiFuBl2H3u7UxgixHnlzCiGEEAPieV4EvHOnt0OIm8yODr0Vol9SwRJCCCGEELuZDL0Ve4pUsIQQQgghxG72MeBVydBbDXjbDm+PEGuSgCWEEEIIIXYtGXor9hoZIiiEEEIIIYQQAyIBSwghhBBCCCEGRAKWEEIIIYQQQgyIBCwhhBBCCCGEGJA91eTCdV0d+H3geUAD+DnP887v7FYJIYQQQgghRGxPBSzgxwHH87z7k3UQfgt4Y3KdAXDt2rWd2jYhhBAD0PF33NjJ7dgi8m+V2HbymRJi8Nb6XO21gPVPgc8AeJ73D67rvqjjuiMAP/3TP70T2yWEEGLwjgDf3+mNGDD5t0rsJPlMCTF4Kz5Xey1gFYCFjt9D12V1KkwAAAjlSURBVHVNz/MC4FHgZcBVINyJjRNCCDEQBvE/WI/u9IZsAfm3SuwE+UwJMXirfq40pdT2b84Gua7728A/eJ73keT3Cc/zju3wZgkhhBBCCCEEsPe6CH4NeB1AMgfrsZ3dHCGEEEIIIYRYtNeGCH4MeJXrul8HNOBtO7w9QgghhBBCCNG2p4YIDtJ+a/nuuu63WZy/dgH4Q+B3gAD4nOd5v34zHhPXdV8M/Kbneadd170T+DCggMeBd3meF7mu+6vA64mPxbs9z/vGarfdiX3YjGX7/wLgb4Cnk6v/wPO8v7xZ9991XQv4U+BWwAZ+A3iSffIeWGX/J9hH7wGxNtd1TwMfB57red7l5LIPAOc8z/vwDm7aruW67l8B/+h53geS33PAN4Gf8jzvuzu6cWLHyWeqfzfrZ2qvDREcpHbLd+DfErd8vym5rusAeJ53Ovn/24D/CPxz4s6ML05Ovm+qY+K67nuAPwGc5KLfBt7ned7LiCugb0z2++XAi4E3A7+32m23c9sHocv+vwD47Y73wV/ezPsPvAWYSfbhtcB/YH+9B7rt/357D4j1NYE/c11X2+kN2SPeCfy867r3Jb9/EPijvXwiKAZOPlP9uSk/U3ttiOAgrdXy/WbzPCDjuu7niF/zXwNsz/O+D+C67meBVxJ3QrmZjsn3gZ8A/iL5/YXAV5Kf/w54NeARV/AUcMl1XdN13YOr3PZj27XhA9Jt/13Xdd9IXMF4N/Hn4Gbd/48Cf9Xxe8D+eg+stv/76T0g1vcl4i9b30UcwgFwXfeXiAN3APxPz/PeuzObt7t4njftuu7/AfyJ67oPAncQnxw+F/hd4i8jZoC3AyngL4mPrwW80/M8mTt+85PPVB9u1s/Ufq5gdW35vlMbs8WqxN8IvIb4m4I/Sy5rKQFD3GTHxPO8vwb8jou05CQSVt/n1uXdbrundNn/bwD/xvO8HwGeAX6Vm3v/y57nlVzXzRMHjfexj94Dq+z/vnoPiJ79PPCvXde9K/k9D/wU8NLk/3e5rvtjO7Vxu43neX8DnCMeQvvW5HPyx8TDaE8DnwbeA/wQ8WfrtcC/JP6sif1BPlN9uBk/U/s5YBWJ3/AterKe1s3oKeC/eJ6nPM97ivjNOdpxfR6Y5+Y/Jp3zR1bb59bl3W67133M87xvtn4Gns9Nvv+u6x4Hvgz8hed5/x/77D3QZf/33XtArM/zvBniauaHic8LHOIlUfzkROfvgVM7t4W70p8Dj3iedyX5/V7g913XfYj4m/Zx4srvV4BPAP83Sz9T4iYmn6kNuak+U/s5YO2nlu9vJ5lP5bruOJABKq7r3pGMEX4N8Yf9Zj8m304moEL87Udrn1/juq7uuu4J4lA5vcpt97rPuq77Q8nPrySeRHrT7r/ruoeAzwHv9TzvT5OL9817YJX931fvAdG75BtkD3grUCeem2sm/0b8CPEXdWJ1HvCzybft7wE+BZwGrnqe92riJjP/fse2Tmw7+Uxt2p7+TO3Z4V8DsJ9avv8n4MOu636VuCPY24lT/38lXoX6c57nPeK67qPc3Mfkl4A/dl03BZwF/srzvNB13b8HHmZxzHTX2+7EBg/YzwP/wXXdJnAN+N88zyvexPv/y8AI8Cuu6/5Kctm/An53n7wHuu3/LwIf2kfvAdGfdxMH7xLwEeLwrQNfJe6MJlb388Cfu65rJL//C+J5I3/puu67gZD4G3exv8hnauP29Gdq37ZpF0IIIYQQQohB289DBIUQQgghhBBioCRgCSGEEEIIIcSASMASQgghhBBCiAGRgCWEEEIIIYQQAyIBSwghhBBCCCEGZD+3aRdi2yXrCn0EeJK4Fb4FvN3zvHNdbnsr8N89z3vJdm6jEEJsF9d1HeCc53m3rnL9nwN3A2/t9ndylfvcSvK303XdHwHmPc/73oA2WYhdTz5XO08ClhDb70ue570ZwHXdVwMfBH5sZzdJCCF2pdd4nndoE/d/O/DfATkRFGKRfK62mAQsIXbWCPCs67ovB341uSwD/CzQbN3Idd03ES8AqyUXvQl4DvDe5Ha3AX/ped6/c133LuBPgBRQBd4MOMAfJf+tEy8we3lrd00IIVZyXTdHvND9CHA+uey5wO8S/42bIT6Bez8w4rruJ4CfIf67NgyMAX/sed4fuK77EPBOz/POua77TuAw8OHkMV8I/CjwAtd1n/Q879K27aQQ20w+V7uLzMESYvs94LruQ67rPgz8KfBXwCngLZ7nPQB8Evhny+5zN/B6z/NOAx7wmuTyk8BPAvcD70ku+yDwfs/z7gf+EHh+ctnvep73iuTnD2zRvgkhxHreCjzued6PEP+NAvhj4F3J37hPA+/xPO8XgFnP894I3Ek8POnVxBX/X1zvSTzP+ybwmeSx5CRQ3Ozeinyudg2pYAmx/TqHCLrAw8DbgN91XbcMHAW+tuw+N4D/nFx/T3IfgMc8zwuAwHXdWnJZ6zHxPO8jyfN8CPhl13XfS/xNVhMhhNgZp4hP0PA87xHXdX3gXuD34z+JWMBTy+5zDXi367o/ARST2yyndblMiP1CPle7iFSwhNhZ15P//gnwNs/z3gpM0vEHzXXdIeDXiYf6/RxQ67hedXnMs8APJvf9add1/0/gHPDe5Fus/524aiaEEDvhHHHVHdd1n098UucBP5v8jXoP8Kll9/m/gIc9z3sL8FEW/wbWgSPJzy/o8lwRcq4j9gf5XO0iUsESYvs9kIxvDoE8cUn+B4BHXNedIw5d4x23LxJXtL4FVIC55PoLqzz+vwH+0HXd9xHPwXoL8R/VP0g6C6WBfzXgfRJCiF79HvBnrut+lfiksAH8PPDnrusayW3+xbL7/A3x37CfJp5LEriuaxPPL/k913UvA1e6PNcjwAdc173ged7ZLdgXIXYL+VztIppS3b4AF0IIIYQQQgjRLynvCSGEEEIIIcSASMASQgghhBBCiAGRgCWEEEIIIYQQAyIBSwghhBBCCCEGRAKWEEIIIYQQQgyIBCwhhBBCCCGEGBAJWEIIIYQQQggxIP8/b+cAOcv/tWkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 864x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(12,5))\n",
    "gs = mpl.gridspec.GridSpec(1, 4)\n",
    "ax1 = plt.subplot(gs[0,:-2])\n",
    "ax2 = plt.subplot(gs[0,-2])\n",
    "ax3 = plt.subplot(gs[0,-1])\n",
    "\n",
    "# Take a fraction of the samples where target value (default) is 'no'\n",
    "df_no = df_default[df_default.default2 == 0].sample(frac=.08)\n",
    "# Take all samples  where target value is 'yes'\n",
    "df_yes = df_default[df_default.default2 == 1]\n",
    "\n",
    "ax1.scatter(df_yes.balance, df_yes.income, s=40, c='orange', marker='+', linewidths=1)\n",
    "ax1.scatter(df_no.balance, df_no.income, s=40, marker='o', linewidths='1',\n",
    "            edgecolors='lightblue', facecolors='white', alpha=.6)\n",
    "\n",
    "ax1.set_ylim(ymin=0)\n",
    "ax1.set_ylabel('Income')\n",
    "ax1.set_xlim(xmin=-100)\n",
    "ax1.set_xlabel('Balance')\n",
    "\n",
    "c_palette = {'No':'lightblue', 'Yes':'orange'}\n",
    "sns.boxplot('default', 'balance', data=df_default, orient='v', ax=ax2, palette=c_palette)\n",
    "sns.boxplot('default', 'income', data=df_default, orient='v', ax=ax3, palette=c_palette)\n",
    "gs.tight_layout(plt.gcf())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4.3 Logistic regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### scikit-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='newton-cg', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n",
      "classes:  [0 1]\n",
      "coefficients:  [-10.651330005794106, 0.0054989165568046445]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train = df_default.balance.values.reshape(-1,1) \n",
    "y = df_default.default2\n",
    "\n",
    "# Create array of test data. Calculate the classification probability\n",
    "# and predicted classification.\n",
    "X_plot = np.arange(df_default.balance.min(), df_default.balance.max()).reshape(-1,1)\n",
    "\n",
    "clf = skl_lm.LogisticRegression(solver='newton-cg')\n",
    "clf.fit(X_train,y)\n",
    "prob = clf.predict_proba(X_plot)\n",
    "\n",
    "fig = plt.figure()\n",
    "ax = plt.axes()\n",
    "# Right plot\n",
    "ax.scatter(X_train, y, color='orange')\n",
    "ax.plot(X_plot, prob[:,1], color='lightblue')\n",
    "\n",
    "ax.hlines(1, xmin=ax.xaxis.get_data_interval()[0],\n",
    "              xmax=ax.xaxis.get_data_interval()[1], linestyles='dashed', lw=1)\n",
    "ax.hlines(0, xmin=ax.xaxis.get_data_interval()[0],\n",
    "              xmax=ax.xaxis.get_data_interval()[1], linestyles='dashed', lw=1)\n",
    "ax.set_ylabel('Probability of default')\n",
    "ax.set_xlabel('Balance')\n",
    "ax.set_yticks([0, 0.25, 0.5, 0.75, 1.])\n",
    "ax.set_xlim(xmin=-100)\n",
    "print(clf)\n",
    "print('classes: ', clf.classes_)\n",
    "print('coefficients: ', [*clf.intercept_, *clf.coef_.tolist()[0]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### statsmodels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.079823\n",
      "         Iterations 10\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Coef.</th>\n",
       "      <th>Std.Err.</th>\n",
       "      <th>z</th>\n",
       "      <th>P&gt;|z|</th>\n",
       "      <th>[0.025</th>\n",
       "      <th>0.975]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>const</th>\n",
       "      <td>-10.651331</td>\n",
       "      <td>0.361169</td>\n",
       "      <td>-29.491287</td>\n",
       "      <td>3.723665e-191</td>\n",
       "      <td>-11.359208</td>\n",
       "      <td>-9.943453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>balance</th>\n",
       "      <td>0.005499</td>\n",
       "      <td>0.000220</td>\n",
       "      <td>24.952404</td>\n",
       "      <td>2.010855e-137</td>\n",
       "      <td>0.005067</td>\n",
       "      <td>0.005931</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Coef.  Std.Err.          z          P>|z|     [0.025    0.975]\n",
       "const   -10.651331  0.361169 -29.491287  3.723665e-191 -11.359208 -9.943453\n",
       "balance   0.005499  0.000220  24.952404  2.010855e-137   0.005067  0.005931"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = sm.add_constant(df_default.balance)\n",
    "y = df_default.default2\n",
    "est = smf.Logit(y.ravel(), X_train).fit()\n",
    "est.summary2().tables[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.145434\n",
      "         Iterations 7\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Coef.</th>\n",
       "      <th>Std.Err.</th>\n",
       "      <th>z</th>\n",
       "      <th>P&gt;|z|</th>\n",
       "      <th>[0.025</th>\n",
       "      <th>0.975]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>const</th>\n",
       "      <td>-3.504128</td>\n",
       "      <td>0.070713</td>\n",
       "      <td>-49.554094</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-3.642723</td>\n",
       "      <td>-3.365532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>student2</th>\n",
       "      <td>0.404887</td>\n",
       "      <td>0.115019</td>\n",
       "      <td>3.520177</td>\n",
       "      <td>0.000431</td>\n",
       "      <td>0.179454</td>\n",
       "      <td>0.630320</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Coef.  Std.Err.          z     P>|z|    [0.025    0.975]\n",
       "const    -3.504128  0.070713 -49.554094  0.000000 -3.642723 -3.365532\n",
       "student2  0.404887  0.115019   3.520177  0.000431  0.179454  0.630320"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = sm.add_constant(df_default.student2)\n",
    "y = df_default.default2\n",
    "est = smf.Logit(y, X_train).fit()\n",
    "est.summary2().tables[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.078577\n",
      "         Iterations 10\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "        <td>Model:</td>              <td>Logit</td>      <td>Pseudo R-squared:</td>    <td>0.462</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <td>Dependent Variable:</td>     <td>default2</td>           <td>AIC:</td>         <td>1579.5448</td> \n",
       "</tr>\n",
       "<tr>\n",
       "         <td>Date:</td>        <td>2018-06-23 13:11</td>       <td>BIC:</td>         <td>1608.3862</td> \n",
       "</tr>\n",
       "<tr>\n",
       "   <td>No. Observations:</td>        <td>10000</td>       <td>Log-Likelihood:</td>    <td>-785.77</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "       <td>Df Model:</td>              <td>3</td>            <td>LL-Null:</td>        <td>-1460.3</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "     <td>Df Residuals:</td>          <td>9996</td>         <td>LLR p-value:</td>    <td>3.2575e-292</td>\n",
       "</tr>\n",
       "<tr>\n",
       "      <td>Converged:</td>           <td>1.0000</td>           <td>Scale:</td>         <td>1.0000</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "    <td>No. Iterations:</td>        <td>10.0000</td>             <td></td>               <td></td>      \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "      <td></td>       <th>Coef.</th>  <th>Std.Err.</th>     <th>z</th>     <th>P>|z|</th>  <th>[0.025</th>  <th>0.975]</th> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>const</th>    <td>-10.8690</td>  <td>0.4923</td>  <td>-22.0793</td> <td>0.0000</td> <td>-11.8339</td> <td>-9.9042</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>balance</th>   <td>0.0057</td>   <td>0.0002</td>   <td>24.7365</td> <td>0.0000</td>  <td>0.0053</td>  <td>0.0062</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>income</th>    <td>0.0000</td>   <td>0.0000</td>   <td>0.3698</td>  <td>0.7115</td>  <td>-0.0000</td> <td>0.0000</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>student2</th>  <td>-0.6468</td>  <td>0.2363</td>   <td>-2.7376</td> <td>0.0062</td>  <td>-1.1098</td> <td>-0.1837</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary2.Summary'>\n",
       "\"\"\"\n",
       "                          Results: Logit\n",
       "==================================================================\n",
       "Model:              Logit            Pseudo R-squared: 0.462      \n",
       "Dependent Variable: default2         AIC:              1579.5448  \n",
       "Date:               2018-06-23 13:11 BIC:              1608.3862  \n",
       "No. Observations:   10000            Log-Likelihood:   -785.77    \n",
       "Df Model:           3                LL-Null:          -1460.3    \n",
       "Df Residuals:       9996             LLR p-value:      3.2575e-292\n",
       "Converged:          1.0000           Scale:            1.0000     \n",
       "No. Iterations:     10.0000                                       \n",
       "-------------------------------------------------------------------\n",
       "             Coef.    Std.Err.     z      P>|z|    [0.025    0.975]\n",
       "-------------------------------------------------------------------\n",
       "const       -10.8690    0.4923  -22.0793  0.0000  -11.8339  -9.9042\n",
       "balance       0.0057    0.0002   24.7365  0.0000    0.0053   0.0062\n",
       "income        0.0000    0.0000    0.3698  0.7115   -0.0000   0.0000\n",
       "student2     -0.6468    0.2363   -2.7376  0.0062   -1.1098  -0.1837\n",
       "==================================================================\n",
       "\n",
       "\"\"\""
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = sm.add_constant(df_default[['balance', 'income', 'student2']])\n",
    "est = smf.Logit(y, X_train).fit()\n",
    "est.summary2()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4.4 Linear Discriminant Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_LDA(mean1=-2, mean2=1, mean3=2, sigma1=0.5, sigma2=0.5):\n",
    "    mean1 = mean1*np.array([1, 1])\n",
    "    mean2 = mean2*np.array([1, 1])\n",
    "    mean3 = mean3*np.array([1, -1])\n",
    "    cov = np.array([[sigma1, 0], [0, sigma2]])\n",
    "    N = 500\n",
    "    K = 3\n",
    "    # if you sample from a t-distribution, the LDA results are really bad\n",
    "    def multivariate_t(means, S, N):\n",
    "        df = 1\n",
    "        m = np.asarray(means)\n",
    "        d = len(means)\n",
    "        x = np.random.chisquare(df, N[0])/df\n",
    "        z = np.random.multivariate_normal(np.zeros(d), S, N)\n",
    "        return m + z/np.sqrt(x)[:,None]\n",
    "    sample1 = np.random.multivariate_normal(mean1, cov, (N,)) \n",
    "    sample2 = np.random.multivariate_normal(mean2, cov, (N,)) \n",
    "    sample3 = np.random.multivariate_normal(mean3, cov, (N,))\n",
    "    \n",
    "    maxX = np.max([sample1[:,0], sample2[:,0], sample3[:,0]])\n",
    "    minX = np.min([sample1[:,0], sample2[:,0], sample3[:,0]])\n",
    "    maxY = np.max([sample1[:,1], sample2[:,1], sample3[:,1]])\n",
    "    minY = np.min([sample1[:,1], sample2[:,1], sample3[:,1]])\n",
    "\n",
    "    # priors\n",
    "    pi1 = pi2 = pi3 = N/K\n",
    "\n",
    "    # grid of points to plot the bayes and LDA regions/lines\n",
    "    N_points_grid = 200\n",
    "    xx, yy = np.meshgrid(np.linspace(minX, maxX, N_points_grid), np.linspace(minY, maxY, N_points_grid))\n",
    "    X = np.c_[xx.ravel(), yy.ravel()]\n",
    "\n",
    "    # Bayes regions\n",
    "    inv_cov = np.linalg.inv(cov)\n",
    "    delta1_fun = lambda X: np.dot(X, np.dot(inv_cov, mean1)) - 1/2*np.dot(mean1.T, np.dot(inv_cov, mean1)) + np.log(pi1)\n",
    "    delta2_fun = lambda X: np.dot(X, np.dot(inv_cov, mean2)) - 1/2*np.dot(mean2.T, np.dot(inv_cov, mean2)) + np.log(pi2)\n",
    "    delta3_fun = lambda X: np.dot(X, np.dot(inv_cov, mean3)) - 1/2*np.dot(mean3.T, np.dot(inv_cov, mean3)) + np.log(pi3)\n",
    "    region1 = np.logical_and(delta1_fun(X) > delta2_fun(X), delta1_fun(X) > delta3_fun(X))\n",
    "    region2 = np.logical_and(delta2_fun(X) > delta1_fun(X), delta2_fun(X) > delta3_fun(X))\n",
    "    region3 = np.logical_and(delta3_fun(X) > delta1_fun(X), delta3_fun(X) > delta2_fun(X))\n",
    "\n",
    "    # LDA prediction\n",
    "    est_mean1 = 1/N*np.sum(sample1, axis=0)\n",
    "    est_mean2 = 1/N*np.sum(sample2, axis=0)\n",
    "    est_mean3 = 1/N*np.sum(sample3, axis=0)\n",
    "    est_cov = (np.cov(sample1, rowvar=False) + np.cov(sample2, rowvar=False) + np.cov(sample3, rowvar=False))/K\n",
    "    inv_est_cov = np.linalg.inv(est_cov)\n",
    "    est_delta1_fun = lambda X: np.dot(X, np.dot(inv_est_cov, est_mean1)) - 1/2*np.dot(est_mean1.T, np.dot(inv_est_cov, est_mean1)) + np.log(pi1)\n",
    "    est_delta2_fun = lambda X: np.dot(X, np.dot(inv_est_cov, est_mean2)) - 1/2*np.dot(est_mean2.T, np.dot(inv_est_cov, est_mean2)) + np.log(pi2)\n",
    "    est_delta3_fun = lambda X: np.dot(X, np.dot(inv_est_cov, est_mean3)) - 1/2*np.dot(est_mean3.T, np.dot(inv_est_cov, est_mean3)) + np.log(pi3)\n",
    "    est_region1 = np.logical_and(est_delta1_fun(X) > est_delta2_fun(X), est_delta1_fun(X) > est_delta3_fun(X))\n",
    "    est_region3 = np.logical_and(est_delta3_fun(X) > est_delta2_fun(X), est_delta3_fun(X) > est_delta1_fun(X))\n",
    "\n",
    "    ### Plot Bayes regions and LDA lines\n",
    "    fig = plt.figure(figsize=(8,8))\n",
    "    ax = plt.subplot(1,1,1)\n",
    "\n",
    "    # Bayes regions\n",
    "    plt.contourf(xx, yy, region1.reshape(xx.shape), alpha=0.5, colors='g', levels=[0.5, 1.0])\n",
    "    plt.contourf(xx, yy, region2.reshape(xx.shape), alpha=0.5, colors='orange', levels=[0.5, 1.0])\n",
    "    plt.contourf(xx, yy, region3.reshape(xx.shape), alpha=0.5, colors='b', levels=[0.5, 1.0])\n",
    "\n",
    "    # Samples\n",
    "    ax.scatter(sample1[:,0], sample1[:,1], s=20, c='green', marker='o', label='1')\n",
    "    ax.scatter(sample2[:,0], sample2[:,1], s=20, c='orange', marker='o', label='2')\n",
    "    ax.scatter(sample3[:,0], sample3[:,1], s=20, c='blue', marker='o', label='3')\n",
    "    ax.set_xlabel('X1');\n",
    "    ax.set_ylabel('X2');\n",
    "\n",
    "    # LDA lines\n",
    "    plt.contour(xx, yy, est_region1.reshape(xx.shape), alpha=0.5, colors='k')\n",
    "    plt.contour(xx, yy, est_region3.reshape(xx.shape), alpha=0.5, colors='k');\n",
    "    \n",
    "    # statistics\n",
    "    pred_green = sum([np.logical_and(delta1_fun(x) > delta2_fun(x), delta1_fun(x) > delta3_fun(x)) for x in sample1])/N*100\n",
    "    pred_orange = sum([np.logical_and(delta2_fun(x) > delta1_fun(x), delta2_fun(x) > delta3_fun(x)) for x in sample2])/N*100\n",
    "    pred_blue = sum([np.logical_and(delta3_fun(x) > delta2_fun(x), delta3_fun(x) > delta1_fun(x)) for x in sample3])/N*100\n",
    "    print('Bayes accuracy: ', np.round(pred_green, 1), np.round(pred_orange, 1), np.round(pred_blue, 1))\n",
    "    est_pred_green = sum([np.logical_and(est_delta1_fun(x) > est_delta2_fun(x), est_delta1_fun(x) > est_delta3_fun(x)) for x in sample1])/N*100\n",
    "    est_pred_orange = sum([np.logical_and(est_delta2_fun(x) > est_delta1_fun(x), est_delta2_fun(x) > est_delta3_fun(x)) for x in sample2])/N*100\n",
    "    est_pred_blue = sum([np.logical_and(est_delta3_fun(x) > est_delta2_fun(x), est_delta3_fun(x) > est_delta1_fun(x)) for x in sample3])/N*100\n",
    "    print('LDA accuracy: ', np.round(est_pred_green, 1), np.round(est_pred_orange, 1), np.round(est_pred_blue, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a68581678eee484bbd7bff34f6b73a4b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(FloatSlider(value=-2.0, description='mean1', max=2.0, min=-2.0, step=0.5), FloatSlider(v…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "interactive_plot = widgets.interactive(generate_LDA, \n",
    "                 mean1=(-2,2,0.5), mean2=(-2,2,0.5), mean3=(-2,2,0.5),\n",
    "                 sigma1=(0.1,5,0.1), sigma2=(0.1,5,0.1),\n",
    "                 continuous_update=False);\n",
    "output = interactive_plot.children[-1]\n",
    "output.layout.height = '15cm'\n",
    "interactive_plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Default dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearDiscriminantAnalysis(n_components=None, priors=None, shrinkage=None,\n",
       "              solver='svd', store_covariance=False, tol=0.0001)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = df_default[['balance', 'income', 'student2']].values\n",
    "y = df_default.default2.values\n",
    "\n",
    "lda = LinearDiscriminantAnalysis(solver='svd')\n",
    "lda.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.974     0.998     0.986      9667\n",
      "         Up      0.782     0.237     0.364       333\n",
      "\n",
      "avg / total      0.968     0.972     0.965     10000\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.997724  0.002276\n",
      "     False  0.762763  0.237237\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_classification_statistics(lda, X, y, labels=['Down', 'Up'])\n",
    "plot_ROC(lda, X, y, label='LDA Classification')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4.4.4 Quadratic Discriminant Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_QDA(mean1=-2, mean2=1, sigma1=1, sigma2=0.5):\n",
    "    mean1 = mean1*np.array([1, 1])\n",
    "    mean2 = mean2*np.array([1, 1])\n",
    "    cov1 = np.array([[sigma1, 0], [0, sigma1]])\n",
    "    cov2 = np.array([[sigma2, 0], [0, sigma2]])\n",
    "    inv_cov1 = np.linalg.inv(cov1)\n",
    "    inv_cov2 = np.linalg.inv(cov2)\n",
    "\n",
    "    N = 500\n",
    "    K = 2\n",
    "    sample1 = np.random.multivariate_normal(mean1, cov1, (N,)) \n",
    "    sample2 = np.random.multivariate_normal(mean2, cov2, (N,)) \n",
    "    maxX = np.max([sample1[:,0], sample2[:,0]])\n",
    "    minX = np.min([sample1[:,0], sample2[:,0]])\n",
    "    maxY = np.max([sample1[:,1], sample2[:,1]])\n",
    "    minY = np.min([sample1[:,1], sample2[:,1]])\n",
    "\n",
    "    pi1 = pi2 = N/K\n",
    "\n",
    "    # grid of points to plot the bayes and LDA regions/lines\n",
    "    N_points_grid = 150\n",
    "    xx, yy = np.meshgrid(np.linspace(minX, maxX, N_points_grid), np.linspace(minY, maxY, N_points_grid))\n",
    "    X = np.c_[xx.ravel(), yy.ravel()]\n",
    "\n",
    "    delta1_fun = lambda Xin: -1/2*((Xin-mean1).dot(inv_cov1)*(Xin-mean1)).sum(axis=1) + np.log(pi1)\n",
    "    delta2_fun = lambda Xin: -1/2*((Xin-mean2).dot(inv_cov2)*(Xin-mean2)).sum(axis=1) + np.log(pi2)\n",
    "    region1 = delta1_fun(X) > delta2_fun(X)\n",
    "    region2 = delta2_fun(X) > delta1_fun(X)\n",
    "\n",
    "    # prediction\n",
    "    est_mean1 = 1/N*np.sum(sample1, axis=0)\n",
    "    est_mean2 = 1/N*np.sum(sample2, axis=0)\n",
    "    est_cov1 = np.cov(sample1, rowvar=False) \n",
    "    est_cov2 = np.cov(sample2, rowvar=False)\n",
    "    inv_est_cov1 = np.linalg.inv(est_cov1)\n",
    "    inv_est_cov2 = np.linalg.inv(est_cov2)\n",
    "    est_delta1_fun = lambda Xin: -1/2*((Xin-est_mean1).dot(inv_est_cov1)*(Xin-est_mean1)).sum(axis=1) + np.log(pi1)\n",
    "    est_delta2_fun = lambda Xin: -1/2*((Xin-est_mean2).dot(inv_est_cov2)*(Xin-est_mean2)).sum(axis=1) + np.log(pi2)\n",
    "    est_region1 = est_delta1_fun(X) > est_delta2_fun(X)\n",
    "\n",
    "    fig = plt.figure(figsize=(8,8))\n",
    "    ax = plt.subplot(1,1,1)\n",
    "\n",
    "    # Bayes regions\n",
    "    plt.contourf(xx, yy, region1.reshape(xx.shape), alpha=0.5, colors='g', levels=[0.5, 1.0])\n",
    "    plt.contourf(xx, yy, region2.reshape(xx.shape), alpha=0.5, colors='orange', levels=[0.5, 1.0])\n",
    "\n",
    "    # Samples\n",
    "    ax.scatter(sample1[:,0], sample1[:,1], s=20, c='green', marker='o',)\n",
    "    ax.scatter(sample2[:,0], sample2[:,1], s=20, c='orange', marker='o',)\n",
    "    ax.set_xlabel('X1');\n",
    "    ax.set_ylabel('X2');\n",
    "\n",
    "    # LDA lines\n",
    "    plt.contour(xx, yy, est_region1.reshape(xx.shape), alpha=0.5, colors='k');\n",
    "    \n",
    "    pred_green = sum(delta1_fun(sample1) > delta2_fun(sample1))/N*100\n",
    "    pred_orange = sum(delta2_fun(sample2) > delta1_fun(sample2))/N*100\n",
    "    print('Bayes accuracy: ', np.round(pred_green, 1), np.round(pred_orange, 1))\n",
    "    est_pred_green = sum(est_delta1_fun(sample1) > est_delta2_fun(sample1))/N*100\n",
    "    est_pred_orange = sum(est_delta2_fun(sample2) > est_delta1_fun(sample2))/N*100\n",
    "    print('LDA accuracy: ', np.round(est_pred_green, 1), np.round(est_pred_orange, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6e3ffe5ce9b04013a5b8426437b0f6a5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(FloatSlider(value=-2.0, description='mean1', max=2.0, min=-2.0, step=0.5), FloatSlider(v…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "interactive_plot = widgets.interactive(plot_QDA, \n",
    "                                 mean1=(-2,2,0.5), mean2=(-2,2,0.5),\n",
    "                                 sigma1=(0.1,5,0.1), sigma2=(0.1,5,0.1),\n",
    "                                 continuous_update=False);\n",
    "output = interactive_plot.children[-1]\n",
    "output.layout.height = '15cm'\n",
    "interactive_plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4.5 A Comparison of Classification Methods"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compare Logistic regression, LDA, QDA and KNN under different conditions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4.6 Lab: Logistic Regression, LDA, QDA, and KNN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.6.1 The Stock Market Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Lag1</th>\n",
       "      <th>Lag2</th>\n",
       "      <th>Lag3</th>\n",
       "      <th>Lag4</th>\n",
       "      <th>Lag5</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Today</th>\n",
       "      <th>Direction</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2001-01-01</th>\n",
       "      <td>0.381</td>\n",
       "      <td>-0.192</td>\n",
       "      <td>-2.624</td>\n",
       "      <td>-1.055</td>\n",
       "      <td>5.010</td>\n",
       "      <td>1.1913</td>\n",
       "      <td>0.959</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-01-01</th>\n",
       "      <td>0.959</td>\n",
       "      <td>0.381</td>\n",
       "      <td>-0.192</td>\n",
       "      <td>-2.624</td>\n",
       "      <td>-1.055</td>\n",
       "      <td>1.2965</td>\n",
       "      <td>1.032</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-01-01</th>\n",
       "      <td>1.032</td>\n",
       "      <td>0.959</td>\n",
       "      <td>0.381</td>\n",
       "      <td>-0.192</td>\n",
       "      <td>-2.624</td>\n",
       "      <td>1.4112</td>\n",
       "      <td>-0.623</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-01-01</th>\n",
       "      <td>-0.623</td>\n",
       "      <td>1.032</td>\n",
       "      <td>0.959</td>\n",
       "      <td>0.381</td>\n",
       "      <td>-0.192</td>\n",
       "      <td>1.2760</td>\n",
       "      <td>0.614</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-01-01</th>\n",
       "      <td>0.614</td>\n",
       "      <td>-0.623</td>\n",
       "      <td>1.032</td>\n",
       "      <td>0.959</td>\n",
       "      <td>0.381</td>\n",
       "      <td>1.2057</td>\n",
       "      <td>0.213</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Lag1   Lag2   Lag3   Lag4   Lag5  Volume  Today  Direction\n",
       "Year                                                                   \n",
       "2001-01-01  0.381 -0.192 -2.624 -1.055  5.010  1.1913  0.959          1\n",
       "2001-01-01  0.959  0.381 -0.192 -2.624 -1.055  1.2965  1.032          1\n",
       "2001-01-01  1.032  0.959  0.381 -0.192 -2.624  1.4112 -0.623          0\n",
       "2001-01-01 -0.623  1.032  0.959  0.381 -0.192  1.2760  0.614          1\n",
       "2001-01-01  0.614 -0.623  1.032  0.959  0.381  1.2057  0.213          1"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_stock = pd.read_csv('Data/Smarket.csv', usecols=range(1,10), index_col=0, parse_dates=True)\n",
    "# convert direction to binary. Up is 1, Down is 0\n",
    "df_stock.replace({'Up': 1, 'Down': 0}, inplace=True)\n",
    "df_stock.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Lag1</th>\n",
       "      <th>Lag2</th>\n",
       "      <th>Lag3</th>\n",
       "      <th>Lag4</th>\n",
       "      <th>Lag5</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Today</th>\n",
       "      <th>Direction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1250.000000</td>\n",
       "      <td>1250.000000</td>\n",
       "      <td>1250.000000</td>\n",
       "      <td>1250.000000</td>\n",
       "      <td>1250.00000</td>\n",
       "      <td>1250.000000</td>\n",
       "      <td>1250.000000</td>\n",
       "      <td>1250.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.003834</td>\n",
       "      <td>0.003919</td>\n",
       "      <td>0.001716</td>\n",
       "      <td>0.001636</td>\n",
       "      <td>0.00561</td>\n",
       "      <td>1.478305</td>\n",
       "      <td>0.003138</td>\n",
       "      <td>0.518400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.136299</td>\n",
       "      <td>1.136280</td>\n",
       "      <td>1.138703</td>\n",
       "      <td>1.138774</td>\n",
       "      <td>1.14755</td>\n",
       "      <td>0.360357</td>\n",
       "      <td>1.136334</td>\n",
       "      <td>0.499861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-4.922000</td>\n",
       "      <td>-4.922000</td>\n",
       "      <td>-4.922000</td>\n",
       "      <td>-4.922000</td>\n",
       "      <td>-4.92200</td>\n",
       "      <td>0.356070</td>\n",
       "      <td>-4.922000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.639500</td>\n",
       "      <td>-0.639500</td>\n",
       "      <td>-0.640000</td>\n",
       "      <td>-0.640000</td>\n",
       "      <td>-0.64000</td>\n",
       "      <td>1.257400</td>\n",
       "      <td>-0.639500</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.039000</td>\n",
       "      <td>0.039000</td>\n",
       "      <td>0.038500</td>\n",
       "      <td>0.038500</td>\n",
       "      <td>0.03850</td>\n",
       "      <td>1.422950</td>\n",
       "      <td>0.038500</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.596750</td>\n",
       "      <td>0.596750</td>\n",
       "      <td>0.596750</td>\n",
       "      <td>0.596750</td>\n",
       "      <td>0.59700</td>\n",
       "      <td>1.641675</td>\n",
       "      <td>0.596750</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.733000</td>\n",
       "      <td>5.733000</td>\n",
       "      <td>5.733000</td>\n",
       "      <td>5.733000</td>\n",
       "      <td>5.73300</td>\n",
       "      <td>3.152470</td>\n",
       "      <td>5.733000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Lag1         Lag2         Lag3         Lag4        Lag5  \\\n",
       "count  1250.000000  1250.000000  1250.000000  1250.000000  1250.00000   \n",
       "mean      0.003834     0.003919     0.001716     0.001636     0.00561   \n",
       "std       1.136299     1.136280     1.138703     1.138774     1.14755   \n",
       "min      -4.922000    -4.922000    -4.922000    -4.922000    -4.92200   \n",
       "25%      -0.639500    -0.639500    -0.640000    -0.640000    -0.64000   \n",
       "50%       0.039000     0.039000     0.038500     0.038500     0.03850   \n",
       "75%       0.596750     0.596750     0.596750     0.596750     0.59700   \n",
       "max       5.733000     5.733000     5.733000     5.733000     5.73300   \n",
       "\n",
       "            Volume        Today    Direction  \n",
       "count  1250.000000  1250.000000  1250.000000  \n",
       "mean      1.478305     0.003138     0.518400  \n",
       "std       0.360357     1.136334     0.499861  \n",
       "min       0.356070    -4.922000     0.000000  \n",
       "25%       1.257400    -0.639500     0.000000  \n",
       "50%       1.422950     0.038500     1.000000  \n",
       "75%       1.641675     0.596750     1.000000  \n",
       "max       3.152470     5.733000     1.000000  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_stock.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Lag1</th>\n",
       "      <th>Lag2</th>\n",
       "      <th>Lag3</th>\n",
       "      <th>Lag4</th>\n",
       "      <th>Lag5</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Today</th>\n",
       "      <th>Direction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Lag1</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.026294</td>\n",
       "      <td>-0.010803</td>\n",
       "      <td>-0.002986</td>\n",
       "      <td>-0.005675</td>\n",
       "      <td>0.040910</td>\n",
       "      <td>-0.026155</td>\n",
       "      <td>-0.039757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lag2</th>\n",
       "      <td>-0.026294</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.025897</td>\n",
       "      <td>-0.010854</td>\n",
       "      <td>-0.003558</td>\n",
       "      <td>-0.043383</td>\n",
       "      <td>-0.010250</td>\n",
       "      <td>-0.024081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lag3</th>\n",
       "      <td>-0.010803</td>\n",
       "      <td>-0.025897</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.024051</td>\n",
       "      <td>-0.018808</td>\n",
       "      <td>-0.041824</td>\n",
       "      <td>-0.002448</td>\n",
       "      <td>0.006132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lag4</th>\n",
       "      <td>-0.002986</td>\n",
       "      <td>-0.010854</td>\n",
       "      <td>-0.024051</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.027084</td>\n",
       "      <td>-0.048414</td>\n",
       "      <td>-0.006900</td>\n",
       "      <td>0.004215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lag5</th>\n",
       "      <td>-0.005675</td>\n",
       "      <td>-0.003558</td>\n",
       "      <td>-0.018808</td>\n",
       "      <td>-0.027084</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.022002</td>\n",
       "      <td>-0.034860</td>\n",
       "      <td>0.005423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Volume</th>\n",
       "      <td>0.040910</td>\n",
       "      <td>-0.043383</td>\n",
       "      <td>-0.041824</td>\n",
       "      <td>-0.048414</td>\n",
       "      <td>-0.022002</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.014592</td>\n",
       "      <td>0.022951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Today</th>\n",
       "      <td>-0.026155</td>\n",
       "      <td>-0.010250</td>\n",
       "      <td>-0.002448</td>\n",
       "      <td>-0.006900</td>\n",
       "      <td>-0.034860</td>\n",
       "      <td>0.014592</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.730563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Direction</th>\n",
       "      <td>-0.039757</td>\n",
       "      <td>-0.024081</td>\n",
       "      <td>0.006132</td>\n",
       "      <td>0.004215</td>\n",
       "      <td>0.005423</td>\n",
       "      <td>0.022951</td>\n",
       "      <td>0.730563</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Lag1      Lag2      Lag3      Lag4      Lag5    Volume  \\\n",
       "Lag1       1.000000 -0.026294 -0.010803 -0.002986 -0.005675  0.040910   \n",
       "Lag2      -0.026294  1.000000 -0.025897 -0.010854 -0.003558 -0.043383   \n",
       "Lag3      -0.010803 -0.025897  1.000000 -0.024051 -0.018808 -0.041824   \n",
       "Lag4      -0.002986 -0.010854 -0.024051  1.000000 -0.027084 -0.048414   \n",
       "Lag5      -0.005675 -0.003558 -0.018808 -0.027084  1.000000 -0.022002   \n",
       "Volume     0.040910 -0.043383 -0.041824 -0.048414 -0.022002  1.000000   \n",
       "Today     -0.026155 -0.010250 -0.002448 -0.006900 -0.034860  0.014592   \n",
       "Direction -0.039757 -0.024081  0.006132  0.004215  0.005423  0.022951   \n",
       "\n",
       "              Today  Direction  \n",
       "Lag1      -0.026155  -0.039757  \n",
       "Lag2      -0.010250  -0.024081  \n",
       "Lag3      -0.002448   0.006132  \n",
       "Lag4      -0.006900   0.004215  \n",
       "Lag5      -0.034860   0.005423  \n",
       "Volume     0.014592   0.022951  \n",
       "Today      1.000000   0.730563  \n",
       "Direction  0.730563   1.000000  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_stock.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.04090991,  0.04338321,  0.04182369,  0.04841425,  0.03486008,\n",
       "        0.02295096,  0.7305629 ,  0.        ])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# very small correlations (today and direction are obiously correlated)\n",
    "corr = df_stock.corr().values\n",
    "np.max(np.abs(np.triu(corr, k=1)), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# volume increases with year\n",
    "plot = sns.boxplot(df_stock.index, df_stock['Volume'],)\n",
    "plot.set_xticklabels([str(date.year) for date in df_stock.index.unique()]);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# volumne by year and direction\n",
    "ax = sns.boxplot(df_stock.index, df_stock['Volume'], hue=df_stock['Direction'])\n",
    "ax.set_xticklabels([str(date.year) for date in df_stock.index.unique()])\n",
    "handles, _ = ax.get_legend_handles_labels()\n",
    "ax.legend(handles, [\"Down\", \"Up\"]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_stock[df_stock.columns.difference(['Today', 'Direction'])]\n",
    "y = df_stock['Direction']\n",
    "X_train = X[:'2004']\n",
    "y_train = y[:'2004']\n",
    "X_test = X['2005':]\n",
    "y_test = y['2005':]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.6.2 Logistic regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic regression, not test/train split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No. Observations: 1250\n",
      "Df Residuals: 1243\n",
      "Df Model: 6\n",
      "Log-Likelihood: -863.79\n",
      "AIC: 1741.58\n",
      "           Coefficients  Standard Errors  t values  p values\n",
      "Intercept       -0.1259            0.241    -0.523     0.601\n",
      "Lag1            -0.0731            0.050    -1.457     0.145\n",
      "Lag2            -0.0423            0.050    -0.845     0.399\n",
      "Lag3             0.0111            0.050     0.222     0.824\n",
      "Lag4             0.0094            0.050     0.187     0.851\n",
      "Lag5             0.0103            0.050     0.208     0.835\n",
      "Volume           0.1354            0.158     0.855     0.393\n",
      "\n",
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.507     0.241     0.327       602\n",
      "         Up      0.526     0.782     0.629       648\n",
      "\n",
      "avg / total      0.517     0.522     0.483      1250\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.240864  0.759136\n",
      "     False  0.217593  0.782407\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "logistic = skl_lm.LogisticRegression(C=1e10)\n",
    "logistic.fit(X, y)\n",
    "print_OLS_error_table(logistic, X, y)\n",
    "print_classification_statistics(logistic, X, y, labels=['Down', 'Up'])\n",
    "plot_ROC(logistic, X, y, label='Logistic Classification')\n",
    "\n",
    "# same results as statsmodels\n",
    "#smLogistic = sm.Logit(y, sm.add_constant(X)).fit()\n",
    "#print(smLogistic.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic regression, with test/train split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No. Observations: 998\n",
      "Df Residuals: 991\n",
      "Df Model: 6\n",
      "Log-Likelihood: -690.55\n",
      "AIC: 1395.11\n",
      "           Coefficients  Standard Errors  t values  p values\n",
      "Intercept        0.1912            0.334     0.573     0.567\n",
      "Lag1            -0.0542            0.052    -1.046     0.296\n",
      "Lag2            -0.0458            0.052    -0.884     0.377\n",
      "Lag3             0.0072            0.052     0.139     0.889\n",
      "Lag4             0.0064            0.052     0.125     0.901\n",
      "Lag5            -0.0042            0.051    -0.083     0.934\n",
      "Volume          -0.1162            0.240    -0.485     0.628\n",
      "\n",
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.443     0.694     0.540       111\n",
      "         Up      0.564     0.312     0.402       141\n",
      "\n",
      "avg / total      0.511     0.480     0.463       252\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.693694  0.306306\n",
      "     False  0.687943  0.312057\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEPCAYAAABP1MOPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3XucTPX/wPHXLnY3t7AUheSyH3ehr3tUKHYRK/V12bLIPfdLWEouuRZyT5IvERUlqxQhcieXtB+RW+Qnt5bFsrvn98eZHbO7s7Oza2b29n4+HvswM+fMnLex5j2f2/vjZRgGQgghsi/v9A5ACCFE+pJEIIQQ2ZwkAiGEyOYkEQghRDYniUAIIbI5SQRCCJHN5UzvAIRIDaWUARwFYgEDyA1EAr201vss5+QBxgKtgGjLeeuA8Vrr2zav9TrQE3gI8AG2A8O01teTuXaqzhcis5AWgciMntNaP6W1rq61VsDnwIcASqmcwI+Yv9tPaa2rAHWAvMD3luMopUYC3YDWWuungGrAPcyEkURqzxciM/GSBWUiM7G0CIporS9b7ucE3gfKaK2DlFLtgQFa69qJnucFHAQmAOHARaCG1voPm3NyA22A1VrruzaP50npfGAkUFhr3ddy7J34+0qpLcBVoDywABgNPKa1vquUygGcBZoAF4CZQBUgF7AJGKq1jlFKjbVc6y5wBeistf77gd5MISykRSAyo5+UUoeVUheA45bHQi1/1gO2JX6C1trA/GBtgPmBfNv2Q91yzi2t9XLbJGCR2vPtuaa1rqi1ngn8htltBfACcEpr/TvwAbBfa10TqA4UBgYppUoAA4D/aK2fBjYCtZNcQYg0kkQgMqPntNZVgRaYYwQ/aa0v2RzPlczzfDHHC+JI3e9+as+352eb24uAzpbbocBHltstgB5KqV+B/UAtzNbBeeAQcEApNQ34VWu99gHjEcJKEoHItLTWB4CBwBKlVCnLwzuAhkqpBL/blvsNgV+AY0AupVS5ROf4KaXClVKPJbqUM+cbgJfNYZ9Er3HT5vZqoLZSqgLQyHIfIAfQzjL+8RTmt/6+Wus4y3mdMbuFPlBKTUnufREitSQRiExNa70C2IPZrQLwBRAFzFBKPQRg+fNDzA/jNVrraGAy8LFS6lHLOb6W18ijtb6Q6BrOnP8PUFMp5aWUyof57T65mO8AK4ElwJda61uWQ98DAy2v4Qt8A/RVSlXDnCn1u9b6Pct1/5OmN0wIOyQRiKygLxColHpRax2D2e9+E9ivlDoKHLDcb6q1vgegtZ4IfIk5k+hXzK4XL+Alexdw4vzlmMngD+BbYGsKMX+E2fWzyOaxfkAe4Ahw2PLnFK31IWAVsE8ptQ/oAgxy4n0Rwikya0gIIbI5aREIIUQ2J4lACCGyOUkEQgiRzUkiEEKIbC5DFp2zTJ37D/A3ZnExIYQQjuUAigF7LVOenZYhEwFmEvg5xbOEEEIk9gxmZVynZdRE8DfA8uXLKVq0aHrHIoQQGd7Fixfp2LEjWD4/UyOjJoJYgKJFi1K8ePH0jkUIITKTVHeny2CxEEJkc5IIhBAim5NEIIQQ2ZwkAiGEyObclgiUUrUtW/QlfrylUmqvUmqnUuoNd11fCCGEc9ySCJRSwzDL6/olejwXZi31FzA32uiulJL5oUIIkY7cNX30JBAM/C/R4xWAE1rrawBKqe2Yix9WI4QQIkVBvwYRfiU8yeM5r+SkNKXT9JpuaRForb8E7tk5lB/41+b+DeBhd8QghBBZUZIkcALY/WCv6enB4kggn839fMB1D8cghBCZ3o3aNxi8YTA5+uTA/31/6Jj21/L0yuLfgXJKqUKYWwc2BKZ5OAYhhMjwkusCAmAXVOpSibNnz+Lt7U1ISAizZ89O87U80iJQSnVQSnW37Bc7CHOT7p3AYq31eU/EIIQQmYndJHAZGAuEwdmzZ6lRowZ79uzhgw8+eKBrua1FoLU+DdSx3P7M5vF1wDp3XVcIITIre60Ao7G5r7xhGNStW5fdu3eTN29exo8fT58+fciZ88E/xmVBmRBCZBCJk0CgfyCGYSYCLy8vpkyZQuvWrTl27Bj9+/d3SRIASQRCCJHhGI0NbtS+Qfn/ladHjx7Wxxs2bMiaNWsoUaKES68niUAIITKYdevWUbFiRd5//30WL17Mn3/+meScoKAgvLy8rD8PQhKBEEJkFJeBd6BVq1acO3eOGjVqsHv3bkqXTrpQLDw8mRlFaSCJQAghMoA5c+ZAF2A75M2blxkzZrB7925q1qyZ5Nu/bQvAMAwMw+DUqVNpvnZG3aFMCCGylQMHDsAtoD4cW3EswThAct/+AwMDXXJtSQRCCOEmDheF3cbsCor/vA8CngDqk+xgcPwMIleTriEhhHCTZJPATsxuoLeBu5bHHgbqA7tJthvIXaRFIIQQbha/KMzuh7qTvTuu6gayR1oEQgjhYkG/BuG16f6HfmxsLLNmzbLejx8MvnfvnnWwN6Wf9evXuy1eaREIIYSL2XYJBfoH0qpVqwQDvseOHXP5orAHIYlACCHcJL5LaFn7ZRw+fJi//voLSH4wOL1I15AQQrjaL8DX9+927NiRiIiIdAsnJdIiEEKINEoyPfQfYA6wHcgFXh+6f8aPK0iLQAgh0siaBGKBNVhXBvMQ8IL957hz9k9aSYtACCHscLgYzNYfQK/7d9u0acOsWbMoXry422JzNUkEQghhh1NJACj8v8Jc5jIAX3/9Na1atXJnWG4hiUAIIRyIn/lj69atW+TOnRuAP0r+QUBAAECmTAIgYwRCCOG0c+fO0aZNG1q2bGmt+1OuXLl0jurBSSIQQogUxMbGMnPmTCpWrMjatWvZu3cvx48fT++wXEa6hoQQwoH9+/fTo0cP9u/fD0BwcDDXrl2jfPny6RyZ60giEEJkaw5nBy2CWqtqERcXR4kSJZgzZw4tW7a0WzwuI04LdZYkAiFEtuZodlBAoQBOcIJBgwYxduxY8ubNm+C4u/YH8DRJBEIIgTk76Ny5c/zxxx88//zzAEQ3iCaiewTVqlVL5+jcSxKBECLLcnpRWCzMnDmTsLAwfH19iYiIoHDhwvj6+lqTQFBQkEs3jM9IJBEIIbIsp5LAccj/YX4G/D4AgBdeeIHY2Nikr5UoCWTmMYHEJBEIIbI8e4vCbty4QdWqVTl9+jSRRFof/+qrr/jqq6+Sf60sMi5gSxKBECJTcrrbJxlt27bl9OnTqXpOVmoF2JJEIITINNLy4R/ob//DOywsjB9++AHImt/yU0MSgRAi00icBAL9A1n/VMp7+cbExDB79mxOnTrFzJkzAWjYsKFbYsyMJBEIITIde33+ydm/fz/du3fnwIEDALzxxhtUrlzZXaFlSlJrSAiRoQX9GoTXJi+8NqVut68bN24wYMAAatWqxYEDByhZsiTffPONJAE7JBEIITI0e91BKVm7di0VKlRg5syZeHl5MXjwYH777TdatmxJUFAQXl5edstEZFdu6RpSSnkDc4FqQDTQTWt9wub4EKA9EAdM1FqvcUccQoisIzXdQd9++y3nz5/nP//5Dz4+PkyfPp3p06fbPTerzgRKDXeNEbQG/LTWdZVSdYDpwEsASqkCQD+gLJAH+BVzt08hhABSPzsoJiaGCxcuULJkSQCmTJlCzZo16d69OzlzJv2YCwwMZP36lAeZswt3JYIGwHcAWutdSqmnbY5FAWcwk0AezFaBECIbcvYD31F30L59++jevTu3bt3i0KFD+Pr6EhISQnh4OL1797ael92niDrirjGC/MC/NvdjlVK2SecccAw4AMxyUwxCiAzOURII9A/EaGxgNDbsThGNjIykf//+1K5dm4MHD3L79m1OnTplvm4WLgfhDu5qEUQC+Wzue2utYyy3mwPFgCct979XSu3QWu9xUyxCCBd40JW8jqSm/98wDNauXUuHDh24c+eO9fGzZ89SoUKFJOeKlLmrRbADCASwjBEcsTl2DbgNRGut7wDXgQJuikMI4SLuSgLOzAKy9cYbbxAcHJwgCdh9XWkFOM1dLYI1QFOl1C+AFxCqlBoEnNBaf6OUagLsUkrFAduBH9wUhxDCxVLz7d0d9u3bl+C+fOt/cG5JBFrrOKBnoocjbI6/DbztjmsLIbKWffv2ceTIEUJDQwE4dOiQ9Zh863cNKTEhhMiQIiMjGT16NLNnzwagS5cuCY5LS8B1JBEIITKU+MHgN998k/Pnz5MjR44kG8VIS8C1JBEIIZLlzplC9pw9e5Y333yTb775xvqYbRKQVoB7SCIQQiTg6MM/tTN8UitxEkhwbWkFuI0kAiFEAmmt+Z9WMTEx1jIQ06ZNI0+ePKxYsQKQFoCnSCIQQtjl7mmikZGRhIWFERERwffff0+LFi2SrAgWniGJQAjhUYZhsGbNGt58800uXLhAjhw5OHjwoJSFSEeSCIQQHnP27Fn69u3LunXrAKhduzYLFiygWrVq1nOkO8jzZGMaIYRHzJ49m4oVK7Ju3Try58/PnDlz2LFjR4IkINKHtAiEEB5x7do1oqKiaNeuHTNmzOCxxx5L75CEhSQCIYRbREZG8vvvv1O7dm0Ahg0bRp06dWjatGk6RyYSk64hIYRLGYbBV199RYUKFWjRogVXrlwBwNfX124SiN9DWKQfpxOBUqqgOwMRQmR+Z86coVWrVrRt25YLFy5QpkwZrl+/7vA5trOFZKZQ+kixa0gp1QiYA+RQSq0GzmitP3Z7ZEKITCMmJoYqVaoQERGR4PHdu3dTtmxZp15DZgulH2daBOOAhsBFYCLQ2/HpQojsplOnTkmSQGpISyB9OTNYHKe1vqqUMrTWd5RSN9welRDC7dJSUC4oKCjF1b/yzT7zcSYRnFBKvQf4K6XeAs64OSYhhAul5QM/ueJyKSUB+WafOTmTCHoC3TC3lIyy3BZCZBKOksCDFpTbtWuXdXqoyLycSQQztNZ94+8opZYCr7kvJCGEOzxIEbmYmBhmzpxpvZ8/f34mTZrE008/7YrQRDpLNhEopfoAYUAhpVSw5WEv4JgnAhNCpA9nxgEiIiIoVqyYhyIS7pZsItBazwHmKKVGaq0nejAmIUQ6cmYcQJJA1uJM19B8pVR7IBdmi+AxrfV77g1LCJHe4mf/XL16lZkzZzJ8+HBy586dzlEJd3AmEXwBHAeqAreBW26NSAjhMY66gaKjo/H19aVQoUKMHTvWw5EJT3KqxITWuicQATQFpNSEEFlEckkgZ86cHDlyxMPRiPTiVCJQSvkBeQADyOvWiIQQDyzo1yC8Nnnhtcl+MbfEhd5s9wR49dVXOXv2rMwIykac6RqaAwwANgLnMNcTCCEyGEcLxxIvEEvcEjh06BClSpVi7ty5NG/e3G0xiowpxUSgtf4y/ral6Fwet0YkhEiTxEnAmcViS5cupUuXLgwePJgxY8bIYHA25WgdQQ1gLHAVGKK1/gfoBIwCHvdMeEKI1HK0cOz06dPs2bPHer9Tp07UqVOHcuXKeSI0kUE5ahF8BIwAngAmKKXyYCaAhp4ITAjhGo5mBnl5eUkSEA4Hi6O01hu11h8BzYFTwHNa65OeCU0I4QrJJQEpECfiOUoEMTa3L2itw7TWUl9WiAzE0eygf//9l759rWXCePLJJ9mwYQOGYWAYBuvXp73YnMhaHHUNeSulcmEmi9uW214AWuu7nghOCOGYvQHieL169WLFihXW+0ePHpXBYGGXo0TwBKAtt70st70w1xKUdnNcQohUiB8gDgwMxGuD/bUDkgREchwVnXvSk4EIIdKuefPmfPfdd8kel/EA4YgzC8pSTSnlDcwFqgHRQDet9Qmb482Bty13DwB9ZPxBiLTtJsYxEiSB2rVrs2vXLhdHJrIyp0pMpEFrwE9rXRd4C5gef0AplQ+YCrTQWtcBTgOF3RSHEJlKqpLATWAm0P/+Qxs2bJAkIFLNqRaBUio/5pjBn1rrKCee0gD4DkBrvUspZVu0pB5wBJiulCoNLLIsVhNCWNhbFGZbG8ieF198kWbNmrkrJJGFpdgiUEq9DGwFPgMGKaXCnHjd/MC/NvdjlVLxSacw8BwwHHN9wgClVECqohYiC0mpQFxihmHw7rvvAlC3bl0OHz6MYRgOxwiEcMSZFsFAoA7mN/zxwD7Ln45EAvls7ntrrePXJVwB9mqtLwIopbYBT2HueSBEtuNoCmi8e/fuJbg/bNgwSpUqRceOHfH2dlcPr8gunPkNitNaRwOGZUDXma6hHUAggFKqDmZXULz9QGWlVGFLK6EOsg+yyMTiSzqn9ceqifkTXj08yTk+Pj4Jrunr60tISIgkAeESzrQIflZKrQCKK6XmA3udeM4aoKlS6hfMtQehSqlBwAmt9TdKqRHA95ZzV2mtj6YleCEygvDwcJgA1Hb/tRo0aOD+i4hsx5ky1COVUs0wp3lGaK3XOfGcOKBnoocjbI6vBFamMlYhMpRHPnqEf0r/Az8++GsF+gey3jBLPhiGwerVq+nfvz8XL14kZ86cDB06lLCwMFkUJtwixUSglNoHLAYWaK0j3R+SEJnDP6UTTnZzpv6/M/r378+HH34IQL169ViwYAGVK1d+4NcVIjnOdDAGAbmBTUqpJUqp+m6OSYgMy3Y8IJ7R2MBobLgkCQC8/PLLFCxYkAULFvDzzz9LEhBul2Ii0Fr/n9Z6GtAW8ANS7BoSIqtKrqTzg9i5cyfjx9+fiNewYUPOnDlD9+7dZTBYeIQz6wheU0ptApYC4cjuZCIbSrzZu2E8eEWU69ev07t3b+rXr8/o0aPZvv3+duD58uVz8EwhXMuZWUPVgN5aa53imUJkUbYtgQct4GZvMHjYsGHUrFnzQcMUIk0c7VncQmv9LeZCr0ZKqUbxx7TWCz0RnBAZzYO2BE6dOkWfPn3YsGEDYA4GL1y4kEqVKrkiPCHSxFHXkL/lz6JAMZufou4OSoiMInGX0IOaPHkyGzZsoECBAixcuJCff/5ZkoBId472I/jUcjNWa20dyVJKvef2qIRIB3ZLQA+y/Fg4Ww/I1u3bt3nooYcAmDhxIgBjx47l0UcfTWuoQriUo66hrkA3oIJSKr5T1BvwAUZ4IDYhPCIoKMgcA0jDwjB7dYHiXb9+nREjRrB161YOHjyIr68vhQoVYv78+Q8QrRCu52iweBmwCRiJuYAeIA645O6ghPCkJFNCmyS8GxgYmKqN3g3DYNWqVQwYMMA6GLxjxw6ef/55F0QrhOs5SgRVtNb7lFJfAsrm8QrARveGJUT6eZAB4VOnTtG7d29rSej69euzYMECGQcQGZqjRNAYs+T0fxM9biCJQGQRQb8GuaRWEMC8efMYPHgwt2/fpkCBAkyZMoWuXbvKojCR4TkaLJ5s+TNUKZUDs4poXWC3h2ITwu2c2QvAWXnz5uX27dt07NiR6dOny2CwyDScKTo3GfgTc6vKGsBFoLN7wxLCtVLcFL5J6ruErl+/zi+//GJdYNapUycCAgKoXdsD9aiFcCFn2qwNtNYLgLpa62ZACTfHJIRL2G4B6TAJpLKNaxgGn3/+OeXLlyc4OJjjx83N9by8vCQJiEzJmRITOZRStYDTSikfoIibYxIizRx980+uTLRXE+fXBvz555/07t2b778391WqX1+K8YrMz5kWwVLgQ2AaMAWY6daIhHgASZLAbhxuAensquF79+4xadIkKlWqxPfff29dGbxt2zYCAgJc/vcQwpOc2aFsrlLqc6A0MF5rfdn9YQnhPLutgCb2z01OSoXk+vXrZ10IJoPBIqtxpgz1K8AvwChgl1Kqk9ujEiIV7LYCMPvynf1JacHYoEGDqFSpEhs3bmTZsmWSBESW4swYwUCgptb6plIqH7AZc9WxEBmLTSvgQUpFG4bBypUrWbduHcuXL8fLy4ty5cpx5MgRlxagEyKjcCYRxGmtbwJorW8ope64OSYhHsiDrAw+efIkvXv3ZuNGc81khw4daNGiBYAkAZFlOZMITiqlpgPbgIbASfeGJITn3b17l+nTp/Puu+9y584dChQowNSpUx94ExohMgNnEkEXoAfQFPgdeMutEQnhYb/88gs9evTg6NGjgAwGi+zHURnqPEAocBOYp7WO81hUQnjQpk2bOHr0KGXKlGHevHk0bdo0vUMSwqMctQg+BU4ABYAAzHLUQmR6hmFw9uxZnnjiCQCGDRtGnjx56NWrl3UDGSGyE0fTRwtrrd8CegO1PBSPEG518uRJmjVrxtNPP83ly+aSGF9fXwYNGiRJQGRbjhJBHIClS0jq6IpM7e7du7z33ntUrlyZjRs3Ehsby2+//ZbeYQmRITj6gPdWSuVSSvna3Pax1BsSIl3FbyrvzJTOHTt2UKNGDUaOHMmdO3fo1KkTERERNGrUyAORCpHxORojeALQlttelttemBvTlHZzXEI4lGR7SYvE0z0nTpzIqFGjAChbtizz5s2jSZNU1p8QIotztDHNk54MRIjkWDeXt8MwDLw2eVlvJ9agQQN8fHwYNmwYI0eOlHEAIexwZh2BEOnK2W//YA4Gr1+/nn79+gHQsGFDzpw5Q9GiRd0aoxCZmSQCkWk4LB1xz+wGGjduHHfu3KFatWrWMQBJAkI45lQiUErlxxwz+FNrHeXekIRIpSPADBh1xhwLCAkJoUKFCukbkxCZiDN7Fr+MWYI6J7BKKWVorcen8BxvYC5QDYgGummtT9g5Zz3wtdZ6fhrjF9lIkn0HbgAfAZaHZDBYiLRxZn3AQKAOcBkYD7Rx4jmtAT+tdV3M2kTT7ZwzHijkZJxCJN13YClmEsgJZd8oy5EjRyQJCJEGziSCOK11NGBorQ3Ama6hBsB3AFrrXcDTtgctrYw4YEPqwhXZXiwYjQ2MxgaXP7pM69at+e3Qb/yx8A/8/PzSOzohMiVnEsHPSqkVQHGl1HxgrxPPyQ/8a3M/VimVE0ApVRnoAIxJbbAi67JdIJbsnsKfAX0gOjoaAH9/f9asWUPFihXTJ2ghsogUE4HWeiRmAbqPgG+11oOdeN1IIJ/tdbTWMZbbrwGPY+501hkYpJRqlpqgRdaT3BTReHnz5oXFwImUzxVCpI4zexa/BjwC/B9QyHI/JTuAQMvz62DO6wBAaz1Ma11ba/0ssAR4X2v9XepDF5mNM9/6bfcRvnLlCiWCSwBw8+ZN8+vDFGjTxplhKiGEs5yZPho/D88LeAq4ijlM58gaoKlS6hfL80KVUoOAE1rrb9IarMjcUvomH79ALOjXIMLXhsNM4Drmb+l/gQ4QWEx2DBPC1VJMBFrrEfG3lVJewLdOPCcO6Jno4Qg7572Tcogis0tcIiKlPYXDr4TDHcwkUAUavtOQrcFb3RukENmYM+sIbKuNFgOkBpFIFdskkNwewHfv3mX37t0888wz5gNN4Jv63xAUFIS3t1RBF8KdnOka0pgVR72A28BUt0YkMjxHReAcSdwSsC4QOwJ8APyNOSWhOOAFLVu2dEG0QoiUOJMIRmutl7k9EpFppCUJ2GsJhJ8KNz/441eTPI65QzYQ6C9jAUJ4ijOJ4A1AEoFIIqW+fkfP++yzz6AvcB1y5crFiBEjGDFihCwKEyIdOJMIfJVSBzG7iOK3r+zg1qhElpKkRtBizMVhAFXh0MpDUiROiHTkTCIY7vYoRKaQmrGBJB/+tl4AvgdCoXmH5pIEhEhnySYCpdTnWutXtdYyb08Azs3+sZ5rmwSOQIntJTiz+ox18djd9nfx8ZHtr4XICBzNyyvisShEpmIYBuvXr0/5xEjotrIbDIRzX55j9erV1kOSBITIOBx1DZVRSk20d8BSf0hkAmmd6un069vrAjKATcB8WHR9ET4+PowYMYJWrVq5LQ4hRNo5SgS3MAeIRSbm6iSQuEsoSRL4C7M0xEHzbqNGjZg/fz7ly5d3aRxCCNdxlAguaq0/9VgkwmXstQLSOtXTWUZj8/XHjx/P6IOj8ff3Z9q0abz++usJS0kLITIcR4lgv8eiEC6VOAmkNLD7wCLv3xw6dCg3b95kyJAhFC5c2L3XFUK4RLKJQGs9xJOBCOekps/f3a2Aq1evmpuQ7oQrDa/g7++Pr68vkyZNcut1hRCuJdW8Mhlnk4CrWwFBvwbhtcnL/PnRC68RXviX8TfLQ0TB9u3bXXo9IYTnOLOgTGRA7v62n5h1UDjRYDBVzTLRL730kkfjEUK4jrQIhPPCwbeHLxw09wv+5JNPiPs1jq1tZM2hEJmZtAiE84qbG8e//vrrTJs2TQaDhcgiJBGIZDXd2pQfw380awMBVIXff/9d1gQIkcVIIhBJGIbBsmXL+LHfj+Z2kY8C1cw9AiQJCJH1SCIQCRw/fpxevXqxefNm84FqoFtoAgIC0jcwIYTbyGCxAMy+/3HjxlG1alU2b96Mv78/DAOmIUlAiCxOEoEA4N1332XMmDFER0fTuXNnIiIizLEBqQ4hRJYniSAbs12LMHjwYJ555hl++uknPvnkE5kRJEQ2IokgGzIMg6VLl9KoUSOio6MBKFSoENu2bePZZ59N3+CEEB4niSCbOX78OE2aNOH111/n559/NjeRF0Jka5IIMomgoKAHKudsbzB4yZIldO7c2XVBCiEyJZk+mkmkZr/gxLZv384bb7xhDgADnTt3ZurUqXbHARxuOi+EyJIkEWQyaSk2d+rUKSIiIlBKUWBIAZY8uYQlh5Y49dxAfzfvZSCESHeSCLIgwzA4evQoVapUAaBTp07ExsbSvn17/Lb7pfj8QP9A1j/lxOb0QogsQRJBBpaWjeePHz9Oz5492bpjK3EL46C45UAJCN0eaj0vfmtJIYSQweIMLDVbTkZHR/Puu+9SpUoVfvrpJ+L84uCi/XOlu0cIYUtaBBlIci2AlMYFtm7dSo8ePdBaAxAaGsonLT6Bh+WbvxAiZZIIMhB7SSClGUIzZsxg4MCBAOQplYeoN6P4pNonbolPCJE1SSJIJ476/1MzM2jtk2shD/AyRL0aBT73j0kXkBDCGW5JBEopb2AuUA2IBrpprU/YHB8I/NdyN1xrPdYdcWREKQ0Ap9QCOH78OIsXL+a9997Dy8uLrXm3wmeYyQCZ8SOESD13tQhaA35a67pKqTrAdOAlAKUo6Iq3AAAbJklEQVRUaaAjUBswgJ+VUmu01ofdFEuGknhh2Pr1zn1oR0dHM2nSJCZOnMjdu3eZ7DUZmlgO5pGxACFE2rkrETQAvgPQWu9SSj1tc+wc0ExrHQuglMoF3HFTHBlWarp/tmzZQs+ePa2DwTQD/nP/uHQBCSEehLsSQX7gX5v7sUqpnFrrGK31PeCyUsoLmAoc1Fofd1Mcmdrly5cZOnQoS5YsAaB8+fLMnz+fZ2OeBaQVIIRwDXetI4gE8tleR2sdE39HKeUHLLec09tNMWR6ixYtYsmSJfj6+vLuu+/y66+/0qhRo/QOSwiRxbgrEewAAgEsYwRH4g9YWgJfA4e01j3iu4iE6c6d+71kAwcOpFu3bhw+fJhdLXfht90Pr02yZZgQwrXclQjWAHeUUr8AHwADlVKDlFKtMAeSGwHNlVJbLD913RRHuoovHW37k5zo6GjGjh1L2bJluXLlCgC+vr589NFHBAQEJKkIKuMCQghXccsYgdY6DuiZ6OEIm9spVz7LApKbJpp4iuiWLVvo0aMHx4+bQyXr1q2jc+fOdktCy7iAEMLVZEGZByQ3Q+jy5csMGTKETz/9FLg/GBw/DiCtACGEJ0giSCdff/01Xbt25cqVK/j6+hIWFsbQoUPx9fVNcq60AoQQ7iSJIJ0ULFiQK1eu0LhxY+bNm0e5cuUA2SFMCOF5kgg85M6dO9SfV58DVQ/cf3AWbKqwiYCzAXDW/vOkO0gI4W6yH4GLOJoh9NNPP1GtWjUODDoAh2yeVBGwM5Eo0D8Qo7GB0diQukFCCLeTFoGLJDdD6PHHH+f5558375QEfKTPPyvZvXs3AwYMoGzZsgBERUVRvHhxpk2bho+PD1evXmXy5MlcuHCB2NhYihUrxltvvUWRIkUA2LdvH3PmzCEmJoZbt24RHBxMx44d0/OvZNfhw4cZMWIEzz//PIMHD072vL/++otBgwaxatUql15/9uzZbNmyhZw5czJy5EiqVq2a5Jzx48fTrVs3ihYt6tJrp9aqVatYuXIlOXPmpFevXjz33HMJjm/cuJEpU6ZQrFgxAN58802qV6/OyJEjOX/+PHfv3qVXr140btyYFStWUKpUKerWdfMMe8MwMtxPQEBAqYCAAOPcuXNGZoFZQM8wDMOIi4szPvnkE8Pf398ADF9fX2PcuHEGGzD4kXSOVLjSrl27jAEDBiR4bNCgQcaGDRuMuLg4o3379sYPP/xgPbZjxw6jTZs2RkxMjHH27FnjpZdeMv755x/DMAzj9u3bRrt27YytW7d69O/gjDlz5hhLly5N8bxz584Z7dq1c+m1jx49aoSEhBhxcXHG+fPnjeDg4CTnHDx40Hjvvfdcet20uHTpktGiRQsjOjraiIyMtN629f777xvfffddgse++OILY/z48YZhGMbVq1eNRo0aGYZhGPfu3TM6depkxMTEpHjtc+fOGQEBAUZAQEApI5WfudIicEJq9w6eOnUqw4cPB0gwGDx602h3hShI2x7PKUlNhViAu3fvcunSJR5++GGOHj1Kvnz5aNKkifV4vXr1KFmyJHv37mXfvn20bt2awoULA+Dn58fHH39M7ty5E7zm6dOnCQsL4969e/j5+fHBBx8wZcoUAgMDadiwIdu2bSM8PJxJkybx3HPPUbp0aUqWLMn27dv5+uuvyZ07N4sWLSJnzpy8+OKLjB49mujoaHx9fRk3bpz1mynAvXv3GDlyJOfOnSM2NpbQ0FCKFy/OF198Qa5cuShatChNmza1nj937lx+/PFHYmNjad++PQ0aNLAe++6771i+fLn1/syZMwEYMGAAhmFw7949xo4dS6lSpejfvz83b97kzp07DB06lNq1a1uft3//fho0aICXlxePPfYYsbGxXL16lUKFClnP+d///kdoqLkn9/Hjx5k0aRJxcXFERkYSFhZGjRo1rO9N6dKl6dKli933Yfr06Rw9epSoqCjKlCnDe++9l+DfYtSoUZw9e39A7+GHH2b27NnW+4cPH6Z69er4+Pjg4+NDyZIliYiISNCC+e233/j999/59NNPqVq1KkOGDKFZs2a8+OKL1nNy5MgBQM6cOalUqRJbtmyhcePGdn7jXEMSgROc/XCJXyjWtWtXPv30U0aMGEHHjh0drigWmd+uXbsICQnhypUreHt788orr1C3bl3Cw8MpUaJEkvNLlCjBhQsXuHTpEuXLl09wLF++fEnOnzx5Mt27d6dhw4aEh4dz7NixZGP5+++/+eqrryhYsCBTp05l48aNtG7dmvDwcD7++GPGjh1LSEgIjRo1YufOnUybNo3p06dbn//5559bn3vz5k2Cg4NZuXIlbdq0oXDhwgmSwLFjx9i2bRurV6/m7t27TJ8+nfr161uPnz59moULF/LQQw8xZswYtm/fTv78+cmXLx/Tp0/nxIkT3Lx5k7Nnz3L58mWWLFnClStXOH36dIK/082bNylQoID1fp48ebhx40aCRLBnzx7rh/aJEycYPnw4SinWrVvHV199RY0aNRK8NwMGDEjyPowdO5b8+fPzySefEBcXR1BQEP/3f//Ho48+ar3OhAkTkn3v42O1/TfMkycPN2/eTHBO/fr1adKkCcWLF+ftt99m5cqVdOrUyfr8fv36MWDAAOv5Sin27NkjiSCjMJJZGLZ582ZmzJjB6tWrAfD39+fIkSN4e8tYvCel5pu7K9WpU4cPPviAa9eu0aVLF4oXLw7Ao48+yvnz55Ocf+bMGerVq8elS5e4ePFigmMREREYhkGFChWsj506dYrq1asD979sfPvtt9bjtr+XBQsWpGDBggC0a9eOd955h9KlS1OqVCkKFizI8ePHWbBgAYsWLcIwDHLlypXg+idPnqRevXoA5M2blzJlynDu3Dm7f+9Tp05RtWpVcuTIwUMPPURYWBh//fWX9bi/vz/Dhw8nT548/Pnnnzz11FM0bNiQ06dP07t3b2sferly5ejYsSODBg0iJiaGkJCQBNfJmzcvUVFR1vtRUVFJEmZcXBw+Pub2fI888ghz587Fz8+PqKgo8ubNm+S9sfc++Pr6cvXqVQYNGkTu3Lm5desW9+7dS3CdlFoEzsTatm1b8ufPD5g9Bt9//z1gJvE+ffrQoUMHWrZsaT2/SJEi7Nq1y+6/gatIIngA//zzD0OGDGHp0qUALFiwgH79+gFYk4CsC8g+4r9Jv/baa6xdu5YaNWpw+fJlNm/ebJ0wsG3bNs6cOUOtWrUoUaIEffr0ITAwkEKFChEVFcWYMWPo06dPgkRQpkwZjhw5Qr169fjmm2/4999/8fHx4Z9//gFI0EKw/fJRqlQpDMNg0aJFtG/fHsDaLVKjRg1OnjzJ3r17E/wdypQpw759+2jatCk3b97k+PHj1sSWWOnSpVmxYgVxcXHExsbSvXt3Ro82uz9v3LjBrFmz2LJlCwChoaEYhsHu3bt55JFHWLx4MQcPHuT9998nLCyMqKgoFi5cyKVLl/jvf/+bYIC1Ro0aTJ06la5du3Lx4kXi4uIStAbArMsVGxtLjhw5mDBhAtOmTaNMmTLMmjXLmoxt3xt778O2bdv4+++/mTFjBlevXuWHH35I8uUvpRZB1apVmTFjBtHR0dy9e5eTJ08SEBBgPW4YBq1atWLlypUULVqUnTt3UqlSJS5fvkyXLl0YM2ZMkoHhyMjIJH9fV5NEkAaGYbBkyRKGDBnC1atX8fX1ZfTo0fTsmbi8kpSJyG7Kli1LSEgI48ePZ9asWcyfP5+JEyeyYMECAIoWLcrChQvJkSMHxYsXZ+jQofTt25ccOXIQFRXFyy+/nKTU+LBhwxgzZgzz5s3Dz8+PqVOncu7cOUaOHMm6desoVapUsvG8/PLLzJw5kzp16gAwfPhw3nnnHaKjo7lz5w6jRo1KcP4rr7zC6NGjad++PdHR0fTt2xd/f3+7r12hQgWeeeYZ2rdvT1xcHO3bt7d+K8+bNy81atSgTZs25M6dm/z583Pp0iWef/55Bg4cyKeffoq3tzd9+vShVKlSzJkzh7Vr15IrVy7rl6l4lStX5umnn+bVV18lLi6OMWPGJImlRo0a/Pbbb1StWpVWrVrRu3dv/P39KVq0KNeuXUtyvr33oXjx4sydO5dXXnkFHx8fSpQowaVLl+x27yWnSJEihISE0KFDBwzDYODAgfj6+rJz5072799P3759GT9+PH379sXPz48yZcrwyiuvMHnyZCIjI5k7dy5z584F4KOPPsLPz49Dhw4l6HJzB6/kujvSk1KqFHBq06ZNyX4b8aT4Pn7DMIiIiKBnz55s3boVgCZNmjBv3jz63+zv8Ju/TBkVwn0OHjzI+vXrCQsLS+9QXComJobQ0FCWLFliHUBOzl9//RU/jvCk1vp0aq4jndjJsF0gZmvv3r1s3bqVIkWKsGzZMjZu3EjZsmUdJgFpBQjhXtWrVyc2NjbJmEtm9/nnn9OjR48Uk8CDkq6hZCSeKRQ/SNepUycuXbpEaGio3X47+eYvRPp4++230zsEl/PU4kJJBMmZAChgAbAFwl8Jv7872FMw5OCQ9ItNCCFcSBKBHYZhwFUgFLgB5AKOAykMV0gXkBAiM8rWicCZlajxg8HxtWSEECKrydaJwGESKAD0go0TNsrKYCFElpatZg0lLhUdL77wktYaPz8/unXrBouBxkgSEA7t3r2bunXrEhISQkhICMHBwfTr14+7d+8CcPXqVYYPH26dWz548GDrQjAwq4+GhoYSEhJC27ZtE9TmyUgOHz5MUFBQgnIU9vz111+88sorbonhzJkztGjRItnj8+fP5+jRo265dmps3ryZtm3b8uqrr9qtwvrbb7/xzDPPWH9n4r+QTp48mVdffZW2bdtan7d161a++OIL9wed2ip1nvhxV/VRLBVCbX+UUkZcXJz1nPPnz5vn/iiVQkXKpPpoQu6oPmoYhrFmzRqjTZs2Rr169ewev3DhgjFw4ECXXze17t69azRp0sS4fv26ER0dbQQHBxuXLl1KcM6qVauMjz/+OMFjO3fuNHr37m0YhmFER0dbX8MwDKNr165GZGRkiteW6qOp1PxAczYs3wALQWuN91ve8ILNCb+nW2jiAbijnEegf2CqNgeS6qOurz4KZk2fZcuWJbi2rRUrVlird168eNG6avj69ev06dOHJk2a0KJFC0qVKoWPjw9jx45l1KhR1lXHYWFhKKWsa4NiYmLIly8fH374oXW1NMAHH3zAgQMHElz7448/tp5z8uRJSpYsycMPPwxAzZo12bdvH82bN7eef/ToUU6dOsWmTZt44oknGDlyJNWrV09QViQ2NpacOc2P50aNGrFmzRpee+01u393V8iWiWBD5w1w2HKnJlDJ/nkyC0g4Q6qPurf6KJBkc5fE9uzZQ3BwMAB//vknoaGh1K5dmwMHDvDhhx/SpEkTbt26Re/evalYsSJTp06lTp06dOjQgdOnTzNixAiWL1/O9evXWbJkCd7e3nTt2pUjR45Qs2ZN63UGDhzoMA5nqo9WrVqVdu3aUblyZebNm8ecOXMYPnw4vr6+3Lt3j7feeotXX32VPHnyAGb10aVLl0oicIU7d+7cv3MYKADL5yynffv2Mg6QRaTXtp5SfdS91Uedce3aNWvLqkiRIsybN48vvvgCLy8vYmJirOc9+eSTgFl9dNeuXWzYsAEwC7t5e3uTK1cua/XRixcvJngupNwicKb6aNOmTa3VR5s2bcq4ceMA+Pfff+nXrx+1atWiR48e1vOLFCnC9evXU/2epEa2SQTxRb8ACAS6QYc2HdItHpH1SPVR91QfdUahQoWIjIwkb968zJw5k3bt2tGoUSO+/PJL1qxZk+T9KV26NK1ataJly5ZcuXKF1atXExERwY8//sjq1au5ffs2wcHBSaqPptQiKFOmDGfOnOH69evkzp2bffv20bVr1wTndO3aldGjR1O1alVr9dE7d+7QuXNnQkNDadWqVYLzpfroA4ovSwvQs2fP+5s9DErHoESWJtVHXV991Bm1atXi0KFDPPbYYzRr1owJEyawYMECihUrZrf6aM+ePRk1ahSrVq3i5s2b9O3blyeeeIKHHnqI4OBgfHx8KFKkCJcuXUpVHLly5eKtt96ia9euGIZB27ZtefTRRzlx4gTLli3jnXfe4Z133mHcuHHkypWLwoULM27cOFauXMm5c+dYvXq1dV+TiRMnUqJECQ4dOuT2PYszdfXRNG9N+KP5h9QFEiJrOH/+PJMnT2bWrFnpHYrLde3alZkzZ1o32ElOtq0+mpYkEN/HKoTIOh5//HGUUhw5ciS9Q3GpLVu28OKLL6aYBB5UpkwE8QvD4hmGwe3btwkLC7MOfj3yyCN89tlnxMXFJZgvm17bGQoh3KtPnz5UqVIlvcNwqWeffdZtC/RsZcpEYNsSCAwMZMeOHVSpUoXx48dz7949unfvTkREBMsqLMN7szdem7wS/AghhLgvUw8Wx49vbNu2jRMnTlCpUiUWLFhgncscfkA2ixFCiJRk6kQQr2HDhnz77bc0bdo0wSrAeDIoLIQQycvwXUOJC8XZjg3E7xscf56Pjw9BvwZJF5AQQqSCW1oESilvYC5QDYgGummtT9gcfwPoAcQA47XW39p9IZKfGeTj45Nk6TaQpNaMdAEJIYRj7uoaag34aa3rKqXqANOBlwCUUkWBfsDTgB+wXSn1g9Y62t4LBQYGcvnyZa5evcqJE2Yu6d69O5MmTbIupbdHuoOEEMI57koEDYDvALTWu5RST9scqwXssHzwRyulTgBVAdu17jnArCJ4tvxZImZFmMGWygldYLFazOIfFtu9cE7LX8m25okQQmR1NnWrcqT2ue5KBPmBf23uxyqlcmqtY+wcuwE8nOj5xQA6duwImHVBrJY5F0BjGqcuYiGEyBqKASdT8wR3JYJIwLbknrclCdg7lg9IXFpvL/AM8DcQ66YYhRAiK8mBmQT2pnRiYu5KBDuAlsAqyxiB7brvPcAEpZQf4AtUABLsL2fpNtruptiEECKrSlVLIJ5bis7ZzBqqCngBoZjFn09orb+xzBrqjjl9daLW+kuXByGEEMIpGab6qCunnGZ2TrwXA4H/Wu6Ga63Hej5K90vpfbA5Zz3wtdZ6vuejdD8nfh+aA29b7h4A+mitM8Z/bBdz4r0YArQH4jC/ZK6x+0JZhFKqNjBZa/1sosdbAmMwPy8Xa60/cvQ6GWlBmXXKKfAW5pRTIMGU0/rAi8B7SinfdInSMxy9F6WBjkA9oC7wglKqarpE6X7Jvg82xgPu3bUj/Tn6fcgHTAVaaK3rAKeBwukRpIc4ei8KYH5O1MXchXxGukToIUqpYcAizGn4to/nAj7AfA8aAd0tn6HJykiJIMGUU8x1BvGsU0611v8C8VNOsypH78U5oJnWOlZrHQfkAu4kfYkswdH7gFLqZcxvfhs8H5pHOXof6mGOwU1XSv0M/J/W+h/Ph+gxjt6LKOAMkMfyE+fx6DzrJBBs5/EKmN3w17TWdzHHW59x9EIZKRHYnXKazDF7U06zkmTfC631Pa31ZaWUl1JqGnBQa308XaJ0v2TfB6VUZaADZvM3q3P0f6Mw8BwwHGgODFBKBXg4Pk9y9F6A+UXpGGYXWdbbpcaGZWz1np1Dqf68zEiJ4EGnnGYljt4LLDOullvO6e3h2DzJ0fvwGvA4sBnoDAxSSjXzbHge4+h9uALs1Vpf1FrfBLYBT3k6QA9y9F40x5w++SRQEmitlKrl4fgyglR/XmakRLADc2YRyUw5fUYp5aeUehg7U06zmGTfC6WUF/A1cEhr3UNrnZXXWST7Pmith2mta1sGyZYA72utv0uPID3A0f+N/UBlpVRhyzfjOpjfiLMqR+/FNeA2EK21voP54VfA4xGmv9+BckqpQkopH6AhsNPREzJSGeo1QFOl1C9YppwqpQZxf8rpLOBnzOQ1yvIPnVUl+15gLhppBPhaZosAjNBaO/yHzqQc/k6kb2geldL/jRHA95ZzV2mts/KXpJTeiybALqVUHGbf+A/pGKtHKaU6AHm11gst78n3mJ+Xi7XW5x09N8NMHxVCCJE+MlLXkBBCiHQgiUAIIbI5SQRCCJHNSSIQQohsThKBEEJkcxlp+qjIxpRSpYDDmCtC423WWr+bzPlLgJVpXTuglDoNnMXc78Ibc2HW61rrG6l4jbcwF7QdBjpprRcppToDV9M6vdUmrjjMqcJ5gTe01vscPKev1np2Wq4nBEgiEBnLscRVFN3shfj1KEqpyZjl0p0uS6C1nmR5bimgG7BIa73ExXG9CLwDtHBwfhggiUCkmSQCkaEppXIAC4ASgD+wQWs92uZ4AObK4nuYJXdf01qfV0q9h7mi0htz1fFqB9fwxlyBqi2VGxcDZTC/kb+vtf5cKdUbeB3zm/p2rfXQ+FYJ0BaoqJQaY7neRSAAc/X3p5bKj+u11jVTE5fFE5grZuOL7PXBXEgF8DJmafZCSqm5QH9gPlDO8vphWustKby+EDJGIDKUikqpLTY/j2MmgF1a6xcxK0/2SvScpphlFpoAE4CClhXXT2qt62MWZBtlKVGc2Eal1E/Aj5gftksxP1gva63rWV5zvFKqMGZrob+l/PGfiQqdTcBszdh2Y32EmTgAQoBPUhnXHqXUX5iVd4dYHg8AgiytJg28qLWegNkV1RuzVXJZa90QeAmYY+e1hUhCWgQiI0nSNaSUyg/8Ryn1HGYxrcT7UHyMWXnzO8yKiyOBKkBNpdQWyzm5ML9ZJy68Ze2CsbleBczEgNb6hlLqGGbrIBQYYulC2sn9b+V2aa1/V0rlVEo9AbyKmVS6pyYupdREzAJqlyyPXwI+VUrdBMqTtH5MFcyaXLUt93Mqpfy11lccxSqEtAhERtcZuK617oi5CUluS+G9eC8BP2utGwOrMZNCBPCTJak8D6wC/nTyer9jqd1u2fSlCnAKeAPoqbVuBFTH3AcgXhz2/y99DEzBTHDX0xBXGPAY0NtSbHEs5s503TCLq8W/D/F/RgArLK/fHPP9uObk31tkY5IIREa3CQi0FBmbB/yB+eEYbx8wwbIpS0/gQ2AdcNPy2H7ASMVsoIWAv1JqO7AFGKu1voRZ5XKvUmoz5jfz3TbPuQT4WFoLtlZj7qi3yHI/VXFZNh7qipkQ8mJW3jyAWXzxts37cEwptQxzLKW8Umor8AtwxvIaQjgkReeEECKbkxaBEEJkc5IIhBAim5NEIIQQ2ZwkAiGEyOYkEQghRDYniUAIIbI5SQRCCJHNSSIQQohs7v8B0bEP1S/V0a0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "logistic_test = skl_lm.LogisticRegression(C=1e10)\n",
    "logistic_test.fit(X_train, y_train)\n",
    "print_OLS_error_table(logistic_test, X_train, y_train)\n",
    "print_classification_statistics(logistic_test, X_test, y_test, labels=['Down', 'Up'])\n",
    "plot_ROC(logistic_test, X_test, y_test, label='Logistic Classification Train/Test')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.6.3 Linear Discriminant Analysis (LDA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# use only Lag1 and Lag2\n",
    "X_train2 = X_train[['Lag1','Lag2']]\n",
    "X_test2 = X_test[['Lag1','Lag2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prior probabilities of groups: \n",
      "      Down        Up\n",
      "  0.491984  0.508016\n",
      "\n",
      "Group means: \n",
      "          Lag1      Lag2\n",
      "Down  0.042790  0.033894\n",
      "Up   -0.039546 -0.031325\n",
      "\n",
      "Coefficients of linear discriminant: \n",
      "           LDA\n",
      "Lag1 -0.642019\n",
      "Lag2 -0.513529\n",
      "\n"
     ]
    }
   ],
   "source": [
    "lda = LinearDiscriminantAnalysis()\n",
    "lda.fit(X_train2, y_train)\n",
    "print('Prior probabilities of groups: ')\n",
    "print(pd.DataFrame(data=lda.priors_.reshape((1,2)), columns=['Down', 'Up'], index=['']))\n",
    "print()\n",
    "print('Group means: ')\n",
    "print(pd.DataFrame(data=lda.means_, columns=X_train2.columns, index=['Down', 'Up']))\n",
    "print()\n",
    "print('Coefficients of linear discriminant: ')\n",
    "print(pd.DataFrame(data=lda.scalings_, columns=['LDA'], index=X_train2.columns))\n",
    "print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.500     0.315     0.387       111\n",
      "         Up      0.582     0.752     0.656       141\n",
      "\n",
      "avg / total      0.546     0.560     0.538       252\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.315315  0.684685\n",
      "     False  0.248227  0.751773\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEPCAYAAABP1MOPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3XmcjeX7wPHPwczIlgxFUSJzU5KlbIWy/Koh2fsKRWRPdmGI7HuKGKkQEZVKRvki2bLLEm5fsoykEQZjmfX5/fGcOZ2ZOXPmzPKc2a7363VezvKc81xOOtdzb9dtMwwDIYQQuVeezA5ACCFE5pJEIIQQuZwkAiGEyOUkEQghRC4niUAIIXI5SQRCCJHL5cvsAIRIDaWUARwBYgEDKABcB3pprffajykIjAWaA5H249YA47XWt50+63WgJ3AX4AtsA4ZqrcOTOXeqjhciu5AWgciOntNaV9VaV9NaK+BL4EMApVQ+YAPmv+2qWuvHgdpAIeAn++sopUYA3YAWWuuqwBNANGbCSCK1xwuRndhkQZnITuwtghJa63/sj/MBM4HyWuumSqn2QH+tda1E77MBB4AJQAhwEaiutf6f0zEFgJbAKq11lNPzBVM6HhgBFNda97W/Nib+sVJqM3AFqAgEA6OA+7XWUUqpvMA5oDFwAZgNPA74ABuBIVrrGKXUWPu5ooDLQGet9V/p+jKFsJMWgciOflZKHVJKXQBO2J/rYv+zLrAl8Ru01gbmD+szmD/It51/1O3H3NJaL3NOAnapPd6Vq1rrR7XWs4HfMbutAP4POK21PgbMAvZprWsA1YDiwEClVBmgP/CU1vpJYD1QK8kZhEgjSQQiO3pOa10FaIY5RvCz1jrM6XWfZN7nhzleEEfq/u2n9nhXtjrdXwh0tt/vAnxsv98M6KGU+g3YB9TEbB38CRwE9iulpgO/aa2/TWc8QjhIIhDZltZ6PzAAWKSUKmt/ejtQXymV4N+2/XF9YAdwFPBRSlVIdEx+pVSIUur+RKfy5HgDsDm97JvoMyKc7q8CaimlKgEN7I8B8gJt7eMfVTGv+vtqrePsx3XG7BaapZSamtz3IkRqSSIQ2ZrWejmwG7NbBeAr4CbwvlLqLgD7nx9i/hiv1lpHAlOAT5RS99mP8bN/RkGt9YVE5/Dk+EtADaWUTSlVGPPqPrmY7wArgEXA11rrW/aXfgIG2D/DD/ge6KuUegJzptQxrfUk+3mfStMXJoQLkghETtAXCFRKPa+1jsHsd48A9imljgD77Y+baK2jAbTWE4GvMWcS/YbZ9WIDXnZ1Ag+OX4aZDP4H/AD8kkLMH2N2/Sx0eq4fUBA4DByy/zlVa30QWAnsVUrtBd4ABnrwvQjhEZk1JIQQuZy0CIQQIpeTRCCEELmcJAIhhMjlJBEIIUQulyWLztmnzj0F/IVZXEwIIYR7eYFSwB77lGePZclEgJkEtqZ4lBBCiMTqYVbG9VhWTQR/ASxbtoySJUtmdixCCJHlXbx4kQ4dOoD99zM1smoiiAUoWbIkpUuXzuxYhBAiO0l1d7oMFgshRC4niUAIIXI5SQRCCJHLSSIQQohczrJEoJSqZd+iL/HzLyml9iilflVKvWnV+YUQQnjGkkSglBqKWV43f6LnfTBrqf8f5kYb3ZVSMj9UCCEykVXTR08BrYDPEz1fCTiptb4KoJTahrn4YRVCCCFcavpbU0Iuh7g9Jt/lfJSjXJo+35IWgdb6ayDaxUtFgGtOj28Ad1sRgxBC5BRuk8BJYFf6Pt/bg8XXgcJOjwsD4V6OQQghsiWjkeG43ah1g0HrBkFPYCTsr7Q/zZ/r7ZXFx4AKSqlimFsH1gemezkGIYTI1n744Qf69OnDuXPnHM/5+vqm+fO80iJQSr2qlOpu3y92IOYm3b8Cn2qt//RGDEIIkd39+eeftGnThpdeeolz585RvXp1x2sFCxZM8+da1iLQWp8Batvvf+H0/BpgjVXnFUKI7MrtoLABrVu3ZteuXRQqVIjx48fTp08ffHx80n1eWVAmhBBZhMskYJh/BBYPZOrUqbRo0YKjR4/y9ttvky9fxlzLZ9Xqo0IIkWsZjQwiIiJ49913uXHjBgsWLHC8Vr9+/Qw/n7QIhBDCy5r+1hTbRluSW7w1a9bw6KOPMnPmTD799FP++OMP831Nm2Kz2RLcMoIkAiGE8LJkxwH+gfsm3kfz5s0JDQ2levXq7Nq1i3LlzIViISGu3xcYGJiueKRrSAghMonRyKBp06YJfuD/5m/H/f379/Pkk08mfZ9hJHnu/PnzaY5DEoEQQmSStHTtpPfq3xVJBEIIkcE8qQ3E7X/vBgYGsmTJErZt28bLL79sbXAuSCIQQog08ugH35VfgelQsWJFfvvtN/z8/AAyJQmAJAIhhEgzd0kg0D+QtVXXJngucVdQgfIFCAsLo0yZMpbE5ylJBEIIkU5Go6SDt85iY2OZO3eu47HzyuCMWhSWHpkfgRBCZHFp7gKya968eYKZQUePHs30VoAzWUcghBApSKkLKCXt27endOnSjsdZKQmAtAiEEMIlV62AlLqA4n3//feEhobSp08fADp06EDLli0pVKhQhseZESQRCCGEC4mTgPOVf+JFYMnp27dvhsdlBUkEQgjhhqtWgCdJIDlWLAhLLxkjEEIIO+dicElecyr4Fs8wDPbt20eNGjUcz7Vs2ZLQ0FAMw3B5W7t2bZLPzmzSIhBC5DqezgJy7g5K3AqIv7IfMWIE+/bto0yZMsyZM4fmzZtnbLBeIIlACJHruE0Cu4CR9uMIwUbC1oFhGNy6dYsCBQoA8OGHHxIcHMyYMWOy7GBwSiQRCCFylFTN+W+cus9+7rnnaNmyJdevX2fDhg3YbDYqVKjA9OnTUx9oFiKJQAiR5aR3AZcnAv0DCcE8h6uyzs5iY2OZM2cOQUFBREREULhwYU6cOIFSytIYvUUSgRAiy0lvEnBV58eVxN0+ruzbt48ePXqwb98+AFq1asXs2bMTLBDL7iQRCCGyLE8XcFll+PDhTJ06lbi4OMqUKcPcuXN56aWXMjUmK8j0USFEluBu6maGnsfFNNDkFCxYEICBAwdy9OjRHJkEQFoEQogswt1K3gw9TzLTQAFCQ0P53//+R8OGDQEYMmQIL730Ek888YQlsWQVkgiEEF6V0kCwVd1BictCOA8QOw8G+/n5cfz4cYoXL46fn1+OTwIgiUAI4WXpreSZ5vM6JQHnVkDiweD/+7//IzY21rI4siJJBEKITJFZA8HxLYEbN24wevRoPvjgA+Li4njwwQeZM2dOjh0HcEcSgRDCct5YF5BarVu35r///S958+Zl0KBB2XplcHpJIhBCWMLdj79VXUCelocGCAoK4tq1a8yfP59q1apZEk92IYlACGEJV7OAPFnkla5zppAEypYt67hfv359du7c6dE00pxO1hEIISxlNDIwGhmWJ4EE53Qq+7x3716qV68OwJkzZzhy5IjjOEkCJkkEQogc6caNG/Tv35+aNWuyf/9+HnzwQb7//nsqV66c2aFlOZIIhBA5zrfffkulSpWYPXs2NpuNQYMG8fvvv+fKGUGesGSMQCmVB/gIeAKIBLpprU86vT4YaA/EARO11qutiEMIkXO5Gxj+4Ycf+PPPP3nqqacIDg7O9YPBKbFqsLgFkF9rXUcpVRuYAbwMoJQqCvQDHgEKAr8BkgiEyAG8OU00uSQQGBjI1KlTqVGjBt27dydv3rxeiSc7s6pr6BngRwCt9U7gSafXbgJnMZNAQcxWgRAiB/BWvSBne/bsoVq1aiiluHPnDmvXrqVYsWL06tVLkoCHrGoRFAGuOT2OVUrl01rH2B+HAkeBvMAki2IQQmQSb64arlWrlmNl8OnTp6lYsaLXzp1TWNUiuA4Udj6PUxJ4ESgFPAw8CLRQStW0KA4hhMW8VT46nmEYrF79b2+yzWZj8ODB/P7775IE0siqRLAdCASwjxEcdnrtKnAbiNRa3wHCgaIWxSGEsJi3u4PefPNNWrVq5Xi8d+9epk2blmvLQ2QEq7qGVgNNlFI7ABvQRSk1EDiptf5eKdUY2KmUigO2Af+1KA4hhJdY0R3kScmIqlWrZvh5cxtLEoHWOg7omejp406vvwu8a8W5hRA5R0pJwLmctEg7WVAmhMgUzltGJneL5+vri9Y6QekIwzBYu9Z7ZStyMkkEQohM4WmVUJvNRr9+/XjggQcsjij3kkQghEg1dzOFPLnSd77ad77CP3v2LM2bN3e8Fl8naNq0aY6N5EXGkzLUQgiPeLq/gKdX+pC0j/+tt97i+++/p0iRIkyaNIkePXrIojAvkEQghPBIcvsLxM/ssZGwdeC8Obw7MTEx5Mtn/hRNnz6dggULMn36dO6///6MCVykSBKBELlUWusCJZ4m6qoF4MlsnuvXrxMUFMTx48f56aefsNlsVKhQgS+++CLVMYn0kUQgRA5kVfE3d4vFPG0BxK8Mfuutt7hw4QJ58+blwIEDjs1jhPdJIhAiB/I0CaR1+8jU7A3s7Ny5c/Tt25c1a9YAZp2g4OBgnnjiiVR/lsg4kgiEyCFctQKsKv6WOAl40hU0Z84c3nnnHW7evCmDwVmMJAIhcghv1PxJ3BLwtDsI4OrVq9y8eZO2bdvy/vvvy2BwFiKJQIgcxsoS0M5JIKVWwPXr1zl27Bi1atUCYOjQodSuXZsmTZpYFp9IG0kEQohUc9cScB4MjoqK4vjx4/j7++Pn5ydJIIvyeGWxUuoeKwMRQmR/8SuDW7duzYULFyhfvjzh4eGZHZZIQYqJQCnVQCl1BNihlHpPKdXVC3EJITyQ1k1hPC0DkVxZiMRiYmKYMWMGjz76KD/88ANFihRh7ty5bN++nfLly6f3ryks5knX0DigPvA1MBFz05lPrAxKCOGZtA4Qp2Xqp+McLsYGOnbsyJdffglAu3btmDVrlgwGZyOeJII4rfUVpZShtb6jlLpheVRCiFRxN0Dsbs5/amb9uNO7d292797NnDlzZI+AbMiTRHBSKTUJ8FdKvQOctTgmIUQGSi4JpPUH2zAMvv76a/bu3cvkyZMBqF+/PlprfHx80hynyDyeJIKeQDfMLSVv2u8LIbKZjLj6P3v2LH369HFsCNOyZUvH9FBJAtmXJ4ngfa113/gHSqklwGvWhSSEcMeqOkLuxMTEMHv2bEaPHs2tW7coUqQIkydP5sknn/RqHMIaySYCpVQfIAgoppRqZX/aBhz1RmBCCNc8HSBOaz2gxHbv3k337t05ePAgYA4Gv//++5QqVSrdny2yhmQTgdZ6LjBXKTVCaz3RizEJkWt59OO9wf5nY/OPEJLuBZBYegZwFyxYwMGDBylbtiwfffQRL774Ypo/S2RNnnQNzVdKtQd8MFsE92utJ1kblhC5R4KunoH2WwYJDAxM9QbvhmFw+fJlihcvDsDUqVN54IEHGDZsGAUKFMi44ESW4Uki+Ao4AVQBbgO3LI1IiFwmLf39gf6BrDVSXz46JWfOnKFv376cPHmSgwcP4ufnR7FixRg7dmyGn0tkHR7VGtJa91RKfYo5Y2iLtSEJkTsZjQzH6t2Mmt/vqejoaGbPns27777LrVu3uPvuuzl8+LAMBucSHtUaUkrlBwoCBlDI0oiEyAXSWhrCCrt27eKpp55iyJAh3Lp1i1deeYVjx45JEshFPEkEc4H+wHogFDhuaURC5ALe2DvAE6NHj6ZOnTqOweCQkBBWrFghM4JymRS7hrTWX8ffV0qtwmwZCCEyQipm/lihQoUK5M2bl0GDBjF69GgZDM6l3K0jqA6MBa4Ag7XWl4COwEjgAe+EJ0TuY2WtnjNnzrB7927atWsHmMXiateuTYUKFSw7p8j63LUIPgaGAw8BE5RSBTETQH1vBCZETtW0adMEU0S9MTDsPBgcGxtL1apVCQgIwGazSRIQbscIbmqt12utPwZeBE4Dz2mtT3knNCFyptRs95gREg8Gt2jRgiJFilh+XpF9uEsEMU73L2itg7TW3p3TJkQ25XLjl4n2WUIb/j0utYu9UuPatWv07dvXMRj88MMPs27dOlasWEHJkiUtO6/Iftx1DeVRSvlgJovb9vs2AK11lDeCEyK7clz1TwBquT7G6plCvXr1Yvny5eTLl4/BgwczatQoGQwWLrlLBA8B2n7fZr9vw1xLUM7iuITIlpLUCkqUBAL9A1lb1bpWgGH8uyht7NixXLx4kdmzZ/P4449bdk6R/bkrOvewNwMRIrtKrlBcYGAgIZjPu9tBLCNER0fz/vvvs3XrVr777jvHIPCmTZssPa/IGTwqMZFaSqk8wEfAE0Ak0E1rfdLp9ReBd+0P9wN9ZPxBZFchISEuu4Dik4DVdu7cSY8ePTh06BAAW7ZsoUGDBl45t8gZPCoxkQYtgPxa6zrAO8CM+BeUUoWBaUAzrXVt4AxQ3KI4hPCOZMYBwLqxgPDwcHr37k3dunU5dOiQYzBYkoBILY9aBEqpIphjBn9orW968JZngB8BtNY7lVLORUvqAoeBGUqpcsBC+2I1IbI9q7uA4n399df07duXixcvymCwSLcUE4FSqg3mauJ8wEqllKG1Hp/C24oA15wexyql8mmtYzCv/p8DqgIRwFal1K9a6xNp+hsI4UUut4nc4PpYKx09epSLFy9Sp04dgoODZTBYpIsnXUMDgNrAP8B4oKUH77kOFHY+jz0JAFwG9mitL2qtIzDLWlf1PGQhMo+7vQOsnA4aHR3NkSNHHI+HDh3KkiVL2LZtmyQBkW6eJII4rXUkYNgHdD3pGtoOBAIopWpjdgXF2wdUVkoVV0rlw0wysg+yyFaMRgaBMwPNonH2m1XTQnfu3EmNGjV47rnnuHz5MgB+fn506tSJPHmsGuYTuYkn/4q2KqWWA6WVUvOBPR68ZzVwRym1A5gFDFBKDVRKNbePBwwHfgJ2Ad9orY+4+SwhsiSrS0U4DwYfPnyYwoULExoamuHnEcKTMtQjlFIvYE7zPK61XuPBe+KAnomePu70+gpgRSpjFcJSCdYDuFkRDDgWbUHGF40zDINVq1bx9ttvOwaDhwwZQlBQkAwGC0t4Mli8F/gUCNZaX7c+JCEyh7sVwQns+veuFS2Bt99+mw8//BCAunXrEhwcTOXKlTP8PELE86RrqClQANiolFqklHra4piEsJyronDxnK/wjUZG0tsIA8Mwb1YUjWvTpg333HMPwcHBbN26VZKAsJwnXUN/A9OVUiuBqcAaoJjVgQlhJXddQN7eR/jXX39l48aNBAUFAVC/fn3Onj1L4cKFU3inEBnDk66h14DXgbyYXURdrA5KCK/JhBXB8cLDwxkxYgTz58/HMAyeffZZnnnmGQBJAsKrPFlZ/ATQW2utUzxSiCyuadOmhDwdkmQRmLdWBIPrweChQ4dSo0YNr8UghDN3exY301r/AJwAGiilHAVMtNYLvBGcEOnhchXwwKTHWX3l7+z06dP06dOHdevWAeZg8IIFC3jssce8FoMQiblrEfjb/0y8lZFUCRXZQkqrgK3cFyA5U6ZMYd26dRQtWpSpU6fStWtXWRQmMp27/QgW2+/GOtcWUkpNsjwqITKQ0chIMCsoMDDQ0i0iE7t9+zZ33XUXABMnTgTMTWPuu+8+r8UghDvuuoa6At2ASkqp+LZzHsAXc2WwEFmOy+4gJxm9+Mud8PBwhg8fzi+//MKBAwfw8/OjWLFizJ8/32sxCOEJd11DS4GNwAjMSXYAcUCY1UEJkVaJk4A3+//jGYbBypUr6d+/v2MwePv27TRs2NDrsQjhCXedk49rrc8AXwPKfqsEyK4XIuuzF4ILqRaSoFvIaqdPnyYwMJD//Oc/XLx4kaeffprffvtNkoDI0ty1CBoBe4H/JHreANZbFpEQFrGiHISzefPmMWjQIG7fvi2DwSJbcTdYPMX+ZxelVF7ABtQhQaUVIbIub44HABQqVIjbt2/ToUMHZsyYIYPBItvwZGXxFOAPzK0qqwMXgc7WhiVE1hceHs6OHTscLY2OHTsSEBBArVruKtYJkfV40mZ9RmsdDNTRWr8AlLE4JiFS5KponLfGAgzD4Msvv6RixYq0atWKEyfMXVZtNpskAZEteVJiIq9SqiZwRinlC5SwOCYhUpSgZLSLwnFWjQf88ccf9O7dm59++gmAp5+WYrwi+/OkRbAE+BCYjll9dLalEQmRCoZhJE0C/hm/YCw6OprJkyfz2GOP8dNPP1G0aFEWLFjAli1bCAgIyNBzCeFtnpSh/kgp9SVQDhivtf7H+rCEMCW7QMxeNM65ZLSVheP69evnWAgmg8Eip0mxRaCUagfsAEYCO5VSHS2PSuRqTX9rim2jDdtGm9tVws6sXjg2cOBAHnvsMdavX8/SpUslCYgcxZMxggFADa11hFKqMLAJc9WxEJZI8uO/C/MyJBGragYZhsGKFStYs2YNy5Ytw2azUaFCBQ4fPuzVxWlCeIsniSBOax0BoLW+oZS6Y3FMIhdy1QUU39Vja2z++HpjXcCpU6fo3bs369ebayZfffVVmjVrZsYhSUDkUJ4kglNKqRnAFqA+cMrakERu5KoVEJ8AvCEqKooZM2bw3nvvcefOHYoWLcq0adMsX40sRFbgSSJ4A+gBNAGOAe9YGpHI3Rq7ftrKH+QdO3bQo0cPjhw5AshgsMh93JWhLoi5P3EEME9rHee1qESu583yEBs3buTIkSOUL1+eefPm0aRJE6+dW4iswF2LYDFwEigKBGCWoxYiwyQ3NdTq7hjDMDh37hwPPfQQAEOHDqVgwYL06tXLsYGMELmJu+mjxbXW7wC9gZpeikfkQMmVg3A1LmAYhqW7h506dYoXXniBJ598kn/+MZfE+Pn5MXDgQEkCItdylwjiAOxdQlJHV6RZknIQG3AsCAMcewcEbreuJRAVFcWkSZOoXLky69evJzY2lt9//92y8wmRnbjrGsqjlPLBTALx920AWusobwQnsp+mTZsm/OF3YhhGgpXAYC8HYVi7f/D27dvp0aOH44e/Y8eOzJgxg3vvvdfS8wqRXbhLBA8B2n7fZr9vw9yYppzFcYlsKrkkkLjf38pyEM4mTpzIyJHmarRHHnmEefPm0bhxMlOThMil3G1M87A3AxHZX9Pfmibs8nESQkiS1oA3PPPMM/j6+jJ06FBGjBgh4wBCuODJOgIhPJIV6gKdOnWKtWvX0q9fPwDq16/P2bNnKVmypGXnFCK7k0QgMl5j728TGRUVxfTp0xk3bhx37tzhiSeeoEGDBgCSBIRIgUeJQClVBHPM4A+t9U1rQxIidbZt20aPHj04evQoAJ06daJSpUqZHJUQ2Ycnexa3waz9mA9YqZQytNbjU3hPHuAj4AkgEuimtT7p4pi1wHda6/lpjF/kYlevXmXYsGF8/PHHgAwGC5FWnqwPGADUBv4BxgMtPXhPCyC/1roOZm2iGS6OGQ8U8zBOkQUlXijmbWPGjOHjjz/Gx8eHUaNGcfjwYUkCQqSBJ4kgTmsdCRhaawPwpGvoGeBHAK31TuBJ5xftrYw4YF3qwhVZiadTRTNSbGys4/7o0aNp0aIFv/32G++99x758+e37LxC5GSeJIKtSqnlQGml1HxgjwfvKQJcc3ocq5TKB6CUqgy8CoxObbDC+5IrD+HcAjAMI8HgsBUlIqKiopg4cSJPPfUUkZGRAPj7+7N69WoeffTRDD+fELmJJ3sWj1BKvQDsB45prX/w4HOvA4WdHufRWsfY778GPIC501lZIEopdUZr/WOqIhdekdxVfzxv1OtPPBgcEhJCy5ae9FAKITzhyZ7FrwH3An8DxeyPU7IdCLS/vzZwOP4FrfVQrXUtrfWzwCJgpiSBzOHuaj+5q/7ENysLxF25coXu3btTr149jh49yiOPPMKGDRskCQiRwTyZPho/D88GVAWuAEtSeM9qoIlSaof9fV2UUgOBk1rr79MarMhYKV3tx8uMXbq++eYbevXqRVhYGD4+PrzzzjuMGDFCxgGEsIAnXUPD4+8rpWxAil1D9oqlPRM9fdzFcWNSDlFktMSF4dKz+Cu5PQXS6+bNm4SFhVGvXj2Cg4NlXYAQFvJkHYGv08NSgNQgyuack0B6r/YTJ4G0lo+Iiopi165d1KtXDzArhBYtWpSmTZuSJ49UQRfCSp50DWnMiqM24DYwzdKIhNekpiWQ0pV/eqqJbt26lR49evDHH39w6NAhAgICsNlsvPTSS2n+TCGE5zxJBKO01kstj0Rkae6SQFpbAVeuXGHYsGEsXLgQMFcGh4eHp+mzhBBp50kieBOQRCCAjNlHwDAMvvjiCwYMGMClS5fw8fFh+PDhDB8+XAaDhcgEniQCP6XUAcwuovjtK1+1NCqRJVg1EBwUFMTEiRMBs0z0/PnzZTBYiEzkySjcMKA/MA8Itt9ENuO8ZsBTGTUQnFjnzp25//77+fTTT9m8ebMkASEyWbItAqXUl1rrV7TWv3gzIGGNxGsGUjNbKL3dQVu3bmXZsmXMmzcPm81GhQoVOH36NL6+vim/WQhhOXddQyW8FoXwGm9uGJN4MLhhw4a0a9cOQJKAEFmIu0RQXik10dULWusRFsUjcgDDMFi2bBkDBw7k0qVL+Pr6Mnz4cJo3b57ZoQkhXHCXCG5hDhAL4bH//e9/9OrVi40bNwLQoEED5s+fT8WKFTM5MiFEctwlgota68Vei0TkCF9++SUbN27E39+f6dOn8/rrr2fKpjVCCM+5SwT7vBaFyNauXLlCsWLmZnNDhgwhIiKCwYMHU7x48UyOTAjhiWSnj2qtB3szEGGN+GmjVrhy5QrdunWjYsWKXL58GQA/Pz8mT54sSUCIbESqeeVwnhSYa/pbU2wbbUluyTEMg6VLl1KxYkU++eQTrl27xrZt2zI8diGEd3iyslhkM4nLTEPSaaOerhpOvIhMBoOFyHkkEeRAniwec7VqeG1V97uNLVy4kL59+xIZGSmDwULkIJIIshlXV/vJiW8FxHf9uDwmFauGAwICiIyM5PXXX2f69OkyDiBEDiFjBNlMWraXTK4LKKXaQVeuXGHx4n9nENevX59jx46xaNEiSQJC5CDSIsjC3F39p6VUhKdX//G62U9VAAAeeklEQVSDwYMGDeLSpUuULVuWBg0aAMhYgBA5kCSCTJKaLp7EPCkYl9YS0idOnKBXr15s2rQJgGeffZZSpUql+nOEENmHdA1lktR08RiGkeC2dq37QV1IfQnpyMhIxo0bR5UqVdi0aRP+/v4sWrSITZs2ERAQ4FGsQojsSVoEXpa4JWB1NVBPu4Pee+89x2YxnTt3Ztq0aTIOIEQuIS0CL/NkgZe3OCehQYMGUa9ePX7++Wc+++wzSQJC5CKSCCzkvCtY4t3BPO3isYJhGCxZsoQGDRoQGRkJQLFixdiyZQvPPvtspsQkhMg8kggslNw4gFUtAedSEck5ceIEjRs35vXXX2fr1q188cUXlsQihMg+ZIzAC7y1K5i7AeLIyEimTp3KhAkTHCuDZ8yYwWuvveaV2IQQWZckghwo8QDxtm3bePPNNzl+/Dggg8FCiISka8gCVpZ+TovTp09z/PhxlFIyGCyESEJaBBbI7JlBhmFw5MgRHn/8cQA6duxIbGws7du3x8/Pz+vxCCGyNmkRpFOWmxl0Hho1asRTTz3FiRMnALDZbHTu3FmSgBDCJUkE6eTtmUGJOWYKrbPB58Cb8PPPP1OoUCHOnDnjlRiEENmbdA1lEG/NDEos5HIIHATeB0LN57p06cLUqVNlHEAI4RFJBNnd18A8865SiuDgYEelUCGE8IR0DWV3tYCCMHbsWA4ePChJQAiRapa0CJRSeYCPgCeASKCb1vqk0+sDgP/YH4ZorcdaEUdOdOLECT799FMmTZpkDkyXBr6A0c1HZ3ZoQohsyqquoRZAfq11HaVUbWAG8DKAUqoc0AHzWtYAtiqlVmutD1kUS7aVYE+BKGAFsByIhim2KdDYfmDBzIlPCJEzWNU19AzwI4DWeifwpNNrocALWutYrXUc4APcsSiObM2RBH4DegBLgGjgBeCpf49Laa8BIYRwx6oWQRHgmtPjWKVUPq11jNY6GvhHKWUDpgEHtNYnLIoje7sGLAB+Mh9WrFiR+fPnyziAECJDWZUIrgOFnR7n0VrHxD9QSuUHPgVuAL0tisFS6dlqEjzcSjIE+An8/PwYOXIkQ4cOlUVhQogMZ1Ui2A68BKy0jxEcjn/B3hL4DtiktZ5i0fktl94yEskmgSjA136/NZS5UoYNkzfIdpFCCMtYlQhWA02UUjsAG9BFKTUQOAnkBRoAfkqpF+3HD9da/2pRLJZK70Ky+EqhkZGRTJ48mY8//piDBw/i7+9vHvCimzcLIUQGsCQR2AeBeyZ6+rjT/fxWnDe72rx5Mz169HDUBlqzZg2dO3fO3KCEELmGrCy2mNuxgGvm3gCLFy8GZDBYCJE5JBFYLNkksB18ZvmwOHwxfn5+BAUFMWTIEBkMFkJ4nZSYSAXnktNuj3Oxd7DRyEhw++XZX4gOj6ZRo0YcPnyYoKAgSQJCiEwhLYJUCAkJgQmYa6Lt3G0UHy/QP5A7d+7w448/0qJFCwDq16/Pjh07qF27dpbazUwIkftIiyC1aqV8CJg//vFX/4OvDuaJJ56gZcuW/PLLL45j6tSpI0lACJHppEXgAVeLxxJvEO/KP//8w+DBgx2DwZUqVSJ/fpkwlZPs2rWL/v3788gjjwBw8+ZNSpcuzfTp0/H19eXKlStMmTKFCxcuEBsbS6lSpXjnnXcoUaIEAHv37mXu3LnExMRw69YtWrVqRYcOHTLzr+TSoUOHGD58OA0bNmTQoEHJHnf+/HkGDhzIypUrM/T8c+bMYfPmzeTLl48RI0ZQpUqVJMeMHz+ebt26UbJkyQw9d2qtXLmSFStWkC9fPnr16sVzzz2X4PX169czdepUSpUqBcBbb71FzZo1CQ4OZtOmTURHR9O+fXvatm3L8uXLKVu2LHXq1LE2aMMwstwtICCgbEBAgBEaGmpkBZjF8czbBvPmTlxcnPHZZ58Z/v7+BmD4+fkZ48aNMyIjI70UsfCWnTt3Gv3790/w3MCBA41169YZcXFxRvv27Y3//ve/jte2b99utGzZ0oiJiTHOnTtnvPzyy8alS5cMwzCM27dvG23btjV++eUXr/4dPDF37lxjyZIlKR4XGhpqtG3bNkPPfeTIEaNTp05GXFyc8eeffxqtWrVKcsyBAweMSZMmZeh50yIsLMxo1qyZERkZaVy/ft1x39nMmTONH3/8McFzO3fuNHr06GHExsYaERERxgcffGAYhmFER0cbHTt2NGJiYlI8d2hoqBEQEGAEBASUNVL5mystgkTclY4wDMOjMYFp06YxbNgwwNw/eN68eVSoUCFD4xRJpbfshyuBgYGp2nc6KiqKsLAw7r77bo4cOULhwoVp3Lix4/W6devy4IMPsmfPHvbu3UuLFi0cO8nlz5+fTz75hAIFCiT4zDNnzhAUFER0dDT58+dn1qxZTJ06lcDAQOrXr8+WLVsICQlh8uTJPPfcc5QrV44HH3yQbdu28d1331GgQAEWLlxIvnz5eP755xk1ahSRkZH4+fkxbtw4x5UpQHR0NCNGjCA0NJTY2Fi6dOlC6dKl+eqrr/Dx8aFkyZI0adLEcfxHH33Ehg0biI2NpX379jzzzDOO13788UeWLVvmeDx79mwA+vfvj2EYREdHM3bsWMqWLcvbb79NREQEd+7cYciQIdSq9W8f7L59+3jmmWew2Wzcf//9xMbGcuXKFYoVK+Y45vPPP6dLly6AWap98uTJxMXFcf36dYKCgqhevbrjuylXrhxvvPGGy+9hxowZHDlyhJs3b1K+fHkmTZqU4L/FyJEjOXfunOPx3XffzZw5cxyPDx06RLVq1fD19cXX15cHH3yQ48ePJ2jB/P777xw7dozFixdTpUoVBg8ezLZt2wgICKBPnz5EREQwdOhQAPLly8djjz3G5s2badSoUfL/8NJJEkEiGbEHcdeuXVm8eDHDhw+nQ4cOMg6Qw+3cuZNOnTpx+fJl8uTJQ7t27ahTpw4hISGUKVMmyfFlypThwoULhIWFUbFixQSvFS5cOMnxU6ZMoXv37tSvX5+QkBCOHj2abCx//fUX33zzDffccw/Tpk1j/fr1tGjRgpCQED755BPGjh1Lp06daNCgAb/++ivTp09nxowZjvd/+eWXjvdGRETQqlUrVqxYQcuWLSlevHiCJHD06FG2bNnCqlWriIqKYsaMGTz99NOO18+cOcOCBQu46667GD16NNu2baNIkSIULlyYGTNmcPLkSSIiIjh37hz//PMPixYt4vLly0n22o6IiKBo0aKOxwULFuTGjRsJEsHu3bsdP9onT55k2LBhKKVYs2YN33zzDdWrV0/w3fTv3z/J9zB27FiKFCnCZ599RlxcHE2bNuXvv//mvvvuc5xnwoQJyX738bE6/zcsWLAgERERCY55+umnady4MaVLl+bdd99lxYoVXL16lQsXLjB//nzOnz9Pr169+PHHH7HZbCil2L17tyQCb0h8NWmkonTEpk2beP/991m1ahV+fn74+/tz+PBh8uSRsXhvSs2Ve0aqXbs2s2bN4urVq7zxxhuULl0agPvuu48///wzyfFnz56lbt26hIWFcfHixQSvHT9+HMMwqFSpkuO506dPU61aNeDfC5IffvjB8brzv9V77rmHe+65B4C2bdsyZswYypUrR9myZbnnnns4ceIEwcHBLFy4EMMw8PHxSXD+U6dOUbduXQAKFSpE+fLlCQ0Ndfn3Pn36NFWqVCFv3rzcddddBAUFcf78ecfr/v7+DBs2jIIFC/LHH39QtWpV6tevz5kzZ+jdu7ejD71ChQp06NCBgQMHEhMTQ6dOnRKcp1ChQty8edPx+ObNm0kSZlxcHL6+ZpGue++9l48++oj8+fNz8+ZNChUqlOS7cfU9+Pn5ceXKFQYOHEiBAgW4desW0dHRCc6TUovAk1hbt25NkSJFALPH4KeffqJYsWKUK1cOX19fypUr54jF39+fEiVKsHPnTpf/DTKK/FLZuSsi52pdAMClS5d4/fXXadSoEWvWrCE4ONjxmiSB3Cf+SjooKIiwsDCqV6/OP//8w6ZNmxzHbNmyhbNnz1KzZk2aNWvGqlWruHLlCmD+aIwePZqwsLAEn1u+fHkOHzbrNn7//fd8/vnn+Pr6cunSJYAELQTnf3dly5bFMAwWLlxI27ZtAShXrhyDBw/m888/Z+zYsTz//PNJzrV3717AvLo9ceKEI7ElVq5cOY4ePUpcXBzR0dF06dKFqKgoAG7cuMEHH3zArFmzGD9+PH5+fhiGwa5du7j33nv59NNP6dWrFzNnzkRrzc2bN1mwYAGTJ09m3LhxCc5TvXp1tm3bRlxcHBcuXCAuLi5BawDMCr2xsbGAedXer18/pkyZQkBAgCNROn83rr6HLVu28NdffzFz5kwGDhzInTt3klwQTpgwgc8//9xxc04CAFWqVGHfvn1ERkZy48YNTp06laBgpGEYNG/e3HEB8Ouvv/LYY49Ro0YNtm7dimEY/P3339y+fdvRCrp+/XqSv29GkxZBIq5aAolXB79Y7EU+++wzBg8ezJUrV/Dz82PUqFH07Jm4vJLIbR555BE6derE+PHj+eCDD5g/fz4TJ050XCSULFmSBQsWkDdvXkqXLs2QIUPo27cvefPm5ebNm7Rp0yZJiZGhQ4cyevRo5s2bR/78+Zk2bRqhoaGMGDGCNWvWULZs2WTjadOmDbNnz6Z27doADBs2jDFjxhAZGcmdO3cYOXJkguPbtWvHqFGjaN++PZGRkfTt2/ffAoiJVKpUiXr16tG+fXvi4uJo376946q8UKFCVK9enZYtW1KgQAGKFClCWFgYDRs2ZMCAASxevJg8efLQp08fypYty9y5c/n222/x8fGhX79+Cc5TuXJlnnzySV555RXi4uIYPTrptqzVq1fn999/p0qVKjRv3pzevXvj7+9PyZIluXr1apLjXX0PpUuX5qOPPqJdu3b4+vpSpkwZwsLCXHbvJadEiRJ06tSJV199FcMwGDBgAH5+fvz666/s27ePvn37Mn78ePr27Uv+/PkpX7487dq1w8fHhz179tCmTRsMw2D06NHkzZsXgIMHDybocrOCLTVdIN6ilCoLnN64cWOyVyMZLb4fP/77cFUjyGhkcPz4cXr27OlYD9C4cWPmzZvnmD4ohPC+AwcOsHbtWoKCgjI7lAwVExNDly5dWLRokSMxJOf8+fPx4wgPa63PpOY8ubL/wrlURHIlIxIngfjtIPfs2cMvv/xCiRIlWLp0KevXr5ckIEQmq1atGrGxsUnGXLK7L7/8kh49eqSYBNIrV3YNpWZmkNHIIDQ01NE87NixI2FhYXTp0sXyfjshhOfefffdzA4hw3lrcWGubBHES7yoIsmsk3B4/fXXqVChgmOvAJvNxqBBgyQJCCFyjFydCJJjGAasA7rAkiVLAHNRixBC5ES5KhHEjw24c+zYMZ599lmYAdwwB4OPHDlC+/btvRKjEEJ4W64aI0hpw/nPP/+crl27motIigK9YP2E9bIyWAiRo+WqFkE8l+MBQK1atcibNy/dunWDT4FGSBIQbu3atYs6derQqVMnOnXqRKtWrejXr59jYdWVK1cYNmyYY275oEGDHAvBwKw+2qVLFzp16kTr1q0T1ObJSg4dOkTTpk0TlKNw5fz587Rr186SGM6ePUuzZs2SfX3+/PkcOXLEknOnxqZNm2jdujWvvPKKyyqsv//+O/Xq1XP8m4m/QP3mm29o27YtrVq1Yu7cuQD88ssvfPXVV9YHndoqdd64WVV9FHsF0XhhYWHGpEmTjLi4OMdzf/75p3msB1VGhZDqowlZUX3UMAxj9erVRsuWLY26deu6fP3ChQvGgAEDMvy8qRUVFWU0btzYCA8PNyIjI41WrVoZYWFhCY5ZuXKl8cknnyR47uzZs0abNm2M27dvG7GxscasWbOMqKgowzAMo2vXrsb169dTPLdUH02lwAOBrFu2DhYAN2D41eHwf04HHMusyER6uFoEmF6B/oGsrSrVRzOz+iiYNX2WLl2a4NzOli9f7iiXcfHiRceq4fDwcPr06UPjxo1p1qwZZcuWxdfXl7FjxzJy5EjHquOgoCCUUo61QTExMRQuXJgPP/zQsVoaYNasWezfvz/BuT/55BPHMadOneLBBx/k7rvvBqBGjRrs3buXF1980XH8kSNHOH36NBs3buShhx5ixIgR7Nixg8qVKzNs2DAuXbpEz549HXWgGjRowOrVq3nttddc/t0zQo5LBMmWIp4AbADOwrrO6+CQ/fkawGOuPyt+EZkQ7kj1UWurjwJJNndJbPfu3bRq1QqAP/74gy5dulCrVi3279/Phx9+SOPGjbl16xa9e/fm0UcfZdq0adSuXZtXX32VM2fOMHz4cJYtW0Z4eDiLFi0iT548dO3alcOHD1OjRg3HeQYMGOA2Dk+qj1apUoW2bdtSuXJl5s2bx9y5cylSpAh79+5l+fLlREZG0r59e7766iuKFCmCUoolS5ZIIkiNZOvRVwMWASuAGPAt5stnH35G+/btZRwgh0jNlXtGkuqj1lYf9cTVq1cdLasSJUowb948vvrqK2w2GzExMY7jHn74YcCsPrpz507WrVsHmIXd8uTJg4+Pj6P66MWLFxO8F1JuEXhSfbRJkyaO6qNNmjRh3LhxvPDCC9SsWZNChQo5vvczZ85QpUoVSpQoQXh4eKq/k9TIcYkgnmEYCbsKvgGWmnfffPNNJk+eLIvCRIaKv5J+7bXX+PbbbxNUH23YsCGQsPpomTJl6NOnD4GBgRQrVsxRfbRPnz4JEkF89dG6devy/fffc+3atVRXH42f/hy/KUv16tU5deoUe/bsSfB3iK8+2qRJE4+qjy5fvpy4uDhiY2Pp3r07o0aNAv6tPrp582YAunTpkqT66IEDB5g5cyZBQUGO6qNhYWH85z//SbEFkFixYsW4fv06hQoVYvbs2bRt25YGDRrw9ddfs3r16iTfT7ly5WjevDkvvfQSly9fZtWqVRw/fpwNGzawatUqbt++TatWrZIUoUypRVC+fHnOnj1LeHg4BQoUYO/evXTt2jXBMV27dmXUqFFUqVLFUX20evXqfPHFF0RGRhIbG+voYgKpPppuIWEhEF+ioxncd/w+vn7va8sr+YncS6qPZnz1UU/UrFmTgwcPcv/99/PCCy8wYcIEgoODKVWqlMvqoz179mTkyJGsXLmSiIgI+vbty0MPPcRdd91Fq1at8PX1pUSJEklKgqfEx8eHd955h65du2IYBq1bt+a+++7j5MmTLF26lDFjxjBmzBjGjRuHj48PxYsXZ9y4cRQqVIjWrVvTvn17DMOgd+/ejjLUBw8elD2L3QkMDEy4n7DTbeHChQalMfhaZv4IkdOdP3/eeOuttzI7DEu88cYbxo0bN1I8Lj2zhrL1OoLkxgPuuececy3AeWC9d2MSQnjfAw88gFLKsYFPTrF582aef/55xy5rVskRXUOGYXDnzh0mTJjAlClTuHr1Kvfeey9h3cIgdV2NQohsqk+fPpkdQoZ79tlnvXKebN0iiLd9+3Yef/xxxo8fT3R0NN27d+f48ePQEJAJQUII4VaWTwSuNpFJvJlMbGwsJ0+e5LHHHmPbtm0EBwc7ptAJIYRwL8sngmTXBdjFr6784Ycf2L9/PxMLTnS50bwQQgjXsnwiiBc/uh1fsAnMgZT44nFNmzbF19c32S0mhRBCuGbJYLFSKg/wEfAEEAl001qfdHr9TaAHEAOM11r/4PKDnNy+fZuJEycyZcoUoqOjuffee5Ms3XZmNDKSfU0IIcS/rJo11ALIr7Wuo5SqjbnNy8sASqmSQD/gSSA/sE0p9V+tdaSrDyqxoASXrl6iQIUCEL9avymEdQujWf5msNGiv4EQQuQSViWCZ4AfAbTWO5VSTzq9VhPYbv/hj1RKnQSqAM5r3fOCWUXwvuP3cfUD+8rAssAbgAKigMuuT/5c0ecS1DwRQoiczqluVV53x7liVSIoAlxzehyrlMqntY5x8doN4O5E7y8F0KFDB8CsC+KwNOWTn+UsjWiU+qiFECL7KwWcSs0brEoE1wHnknt57EnA1WuFgcSl9fYA9YC/gFiLYhRCiJwkL2YS2JPSgYlZlQi2Ay8BK+1jBM7rvncDE5RS+QE/oBKQYH85e7fRNotiE0KInCpVLYF4NsPI+Nk1TrOGqmCu7e0CBAIntdbf22cNdcecvjpRa/11hgchhBDCI5YkgrSwYsppduXBdzEA+I/9YYjWeqz3o7ReSt+D0zFrge+01vO9H6X1PPj38CLwrv3hfqCP1jpr/I+dwTz4LgYD7YE4zIvM1S4/KIdQStUCpmitn030/EvAaMzfy0+11h+7+5ystKDMMeUUeAdzyimQYMrp08DzwCSllF+mROkd7r6LckAHoC5QB/g/pVSVTInSesl+D07GAzl9hyF3/x4KA9OAZlrr2sAZoHhmBOkl7r6Lopi/E3UwdyF/P1Mi9BKl1FBgIeY0fOfnfYBZmN9BA6C7/Tc0WVkpESSYcoq5ziCeY8qp1voaED/lNKdy912EAi9orWO11nGAD3DH+yF6hbvvAaVUG8wrv3XeD82r3H0PdTHH4GYopbYCf2utL3k/RK9x913cBM4CBe23OK9H512ngFYunq+E2Q1/VWsdhTneWs/dB2WlROByymkyr7macpqTJPtdaK2jtdb/KKVsSqnpwAGt9YlMidJ6yX4PSqnKwKuYzd+czt3/G8Uxi60PA14E+iulArwcnze5+y7AvFA6itlF9oE3A/M2+9hqtIuXUv17mZUSQXqnnOYk7r4L7DOultmP6e3l2LzJ3ffwGvAAsAnoDAxUSr3g3fC8xt33cBnYo7W+qLWOALYAVb0doBe5+y5exJw++TDwINBCKVXTy/FlBan+vcxKiWA75swikplyWk8plV8pdTcuppzmMMl+F0opG/AdcFBr3UNrnZPXWST7PWith2qta9kHyRYBM7XWP2ZGkF7g7v+NfUBlpVRx+5Vxbcwr4pzK3XdxFbgNRGqt72D++BX1eoSZ7xhQQSlVTCnlC9QHfnX3hqy0Q9lqoIlSagf2KadKqYH8O+X0A2ArZvIaaf8PnVMl+11gLhppAPjZZ4sADNdau/0PnU25/TeRuaF5VUr/bwwHfrIfu1JrnZMvklL6LhoDO5VScZh94//NxFi9Sin1KlBIa73A/p38hPl7+anW+k93780y00eFEEJkjqzUNSSEECITSCIQQohcThKBEELkcpIIhBAil5NEIIQQuVxWmj4qcjGlVFngEOaK0HibtNbvJXP8ImBFWtcOKKXOAOcw97vIg7kw63Wt9Y1UfMY7mAvaDgEdtdYLlVKdgStpnd7qFFcc5lThQsCbWuu9bt7TV2s9Jy3nEwIkEYis5WjiKooW+7/49ShKqSmY5dI9LkugtZ5sf29ZoBuwUGu9KIPjeh4YAzRzc3wQIIlApJkkApGlKaXyAsFAGcAfWKe1HuX0egDmyuJozJK7r2mt/1RKTcJcUZkHc9XxKjfnyIO5AlXbKzd+CpTHvCKfqbX+UinVG3gd80p9m9Z6SHyrBGgNPKqUGm0/30UgAHP192J75ce1WusaqYnL7iHMFbPxRfb6YC6kAmiDWZq9mFLqI+BtYD5Qwf75QVrrzSl8vhAyRiCylEeVUpudbg9gJoCdWuvnMStP9kr0niaYZRYaAxOAe+wrrh/WWj+NWZBtpL1EcWLrlVI/Axswf2yXYP6w/qO1rmv/zPFKqeKYrYW37eWP/0hU6GwCZmvGuRvrY8zEAdAJ+CyVce1WSp3HrLw72P58ANDU3mrSwPNa6wmYXVG9MVsl/2it6wMvA3NdfLYQSUiLQGQlSbqGlFJFgKeUUs9hFtNKvA/FJ5iVN3/ErLg4AngcqKGU2mw/xgfzyjpx4S1HF4zT+SphJga01jeUUkcxWwddgMH2LqRf+feq3CWt9TGlVD6l1EPAK5hJpXtq4lJKTcQsoBZmfz4MWKyUigAqkrR+zOOYNblq2R/nU0r5a60vu4tVCGkRiKyuMxCute6AuQlJAXvhvXgvA1u11o2AVZhJ4Tjwsz2pNARWAn94eL5j2Gu32zd9eRw4DbwJ9NRaNwCqYe4DEC8O1/8vfQJMxUxw4WmIKwi4H+htL7Y4FnNnum6YxdXiv4f4P48Dy+2f/yLm93HVw7+3yMUkEYisbiMQaC8yNg/4H+aPY7y9wAT7piw9gQ+BNUCE/bl9gJGK2UALAH+l1DZgMzBWax2GWeVyj1JqE+aV+S6n94QBvvbWgrNVmDvqLbQ/TlVc9o2HumImhEKYlTf3YxZfvO30PRxVSi3FHEupqJT6BdgBnLV/hhBuSdE5IYTI5aRFIIQQuZwkAiGEyOUkEQghRC4niUAIIXI5SQRCCJHLSSIQQohcThKBEELkcpIIhBAil/t/yoVeyHYxryMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_classification_statistics(lda, X_test2, y_test, labels=['Down', 'Up'])\n",
    "plot_ROC(lda, X_test2, y_test, label='LDA Train/Test, only Lag1 and Lag2')\n",
    "plot_classification(lda, X_test2, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.6.4 Quadratic Discriminant Analysis (QDA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prior probabilities of groups: \n",
      "      Down        Up\n",
      "  0.491984  0.508016\n",
      "\n",
      "Group means: \n",
      "          Lag1      Lag2\n",
      "Down  0.042790  0.033894\n",
      "Up   -0.039546 -0.031325\n",
      "\n"
     ]
    }
   ],
   "source": [
    "qda = QuadraticDiscriminantAnalysis()\n",
    "qda.fit(X_train2, y_train)\n",
    "print('Prior probabilities of groups: ')\n",
    "print(pd.DataFrame(data=qda.priors_.reshape((1,2)), columns=['Down', 'Up'], index=['']))\n",
    "print()\n",
    "print('Group means: ')\n",
    "print(pd.DataFrame(data=qda.means_, columns=X_train2.columns, index=['Down', 'Up']))\n",
    "print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.600     0.270     0.373       111\n",
      "         Up      0.599     0.858     0.706       141\n",
      "\n",
      "avg / total      0.599     0.599     0.559       252\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.270270  0.729730\n",
      "     False  0.141844  0.858156\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_classification_statistics(qda, X_test2, y_test, labels=['Down', 'Up'])\n",
    "plot_ROC(qda, X_test2, y_test, label='QDA Train/Test, only Lag1 and Lag2')\n",
    "plot_classification(qda, X_test2, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.6.5 K-Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "           metric_params=None, n_jobs=1, n_neighbors=3, p=2,\n",
       "           weights='uniform')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_neighbors = 3\n",
    "knn = neighbors.KNeighborsClassifier(n_neighbors)\n",
    "knn.fit(X_train2, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.466     0.432     0.449       111\n",
      "         Up      0.577     0.610     0.593       141\n",
      "\n",
      "avg / total      0.528     0.532     0.529       252\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.432432  0.567568\n",
      "     False  0.390071  0.609929\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_classification_statistics(knn, X_test2, y_test, labels=['Down', 'Up'])\n",
    "plot_ROC(knn, X_test2, y_test, label='KNN Train/Test, only Lag1 and Lag2')\n",
    "plot_classification(knn, X_test2, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.6.6 An Application to Caravan Insurance Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>MOSTYPE</th>\n",
       "      <th>MAANTHUI</th>\n",
       "      <th>MGEMOMV</th>\n",
       "      <th>MGEMLEEF</th>\n",
       "      <th>MOSHOOFD</th>\n",
       "      <th>MGODRK</th>\n",
       "      <th>MGODPR</th>\n",
       "      <th>MGODOV</th>\n",
       "      <th>MGODGE</th>\n",
       "      <th>...</th>\n",
       "      <th>APERSONG</th>\n",
       "      <th>AGEZONG</th>\n",
       "      <th>AWAOREG</th>\n",
       "      <th>ABRAND</th>\n",
       "      <th>AZEILPL</th>\n",
       "      <th>APLEZIER</th>\n",
       "      <th>AFIETS</th>\n",
       "      <th>AINBOED</th>\n",
       "      <th>ABYSTAND</th>\n",
       "      <th>Purchase</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>33</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>37</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>37</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>40</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 87 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  MOSTYPE  MAANTHUI  MGEMOMV  MGEMLEEF  MOSHOOFD  MGODRK  MGODPR  \\\n",
       "0           1       33         1        3         2         8       0       5   \n",
       "1           2       37         1        2         2         8       1       4   \n",
       "2           3       37         1        2         2         8       0       4   \n",
       "3           4        9         1        3         3         3       2       3   \n",
       "4           5       40         1        4         2        10       1       4   \n",
       "\n",
       "   MGODOV  MGODGE    ...     APERSONG  AGEZONG  AWAOREG  ABRAND  AZEILPL  \\\n",
       "0       1       3    ...            0        0        0       1        0   \n",
       "1       1       4    ...            0        0        0       1        0   \n",
       "2       2       4    ...            0        0        0       1        0   \n",
       "3       2       4    ...            0        0        0       1        0   \n",
       "4       1       4    ...            0        0        0       1        0   \n",
       "\n",
       "   APLEZIER  AFIETS  AINBOED  ABYSTAND  Purchase  \n",
       "0         0       0        0         0        No  \n",
       "1         0       0        0         0        No  \n",
       "2         0       0        0         0        No  \n",
       "3         0       0        0         0        No  \n",
       "4         0       0        0         0        No  \n",
       "\n",
       "[5 rows x 87 columns]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_caravan = pd.read_csv('Data/Caravan.csv')\n",
    "df_caravan['Purchase'] = df_caravan['Purchase'].astype('category')\n",
    "df_caravan.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "No     5474\n",
       "Yes     348\n",
       "Name: Purchase, dtype: int64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_caravan['Purchase'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = df_caravan.Purchase\n",
    "X = df_caravan.drop('Purchase', axis=1).astype('float64')\n",
    "X_scaled = preprocessing.scale(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = X_scaled[:1000]\n",
    "y_test = y[:1000]\n",
    "X_train = X_scaled[1000:]\n",
    "y_train = y[1000:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 neighbors\n",
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No      0.948     0.937     0.943       941\n",
      "        Yes      0.157     0.186     0.171        59\n",
      "\n",
      "avg / total      0.902     0.893     0.897      1000\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.937301  0.062699\n",
      "     False  0.813559  0.186441\n",
      "\n",
      "Using 3 neighbors\n",
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No      0.946     0.979     0.962       941\n",
      "        Yes      0.231     0.102     0.141        59\n",
      "\n",
      "avg / total      0.903     0.927     0.913      1000\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.978746  0.021254\n",
      "     False  0.898305  0.101695\n",
      "\n",
      "Using 5 neighbors\n",
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No      0.944     0.993     0.968       941\n",
      "        Yes      0.364     0.068     0.114        59\n",
      "\n",
      "avg / total      0.910     0.938     0.918      1000\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.992561  0.007439\n",
      "     False  0.932203  0.067797\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for i in [1, 3, 5]:\n",
    "    print(f'Using {i} neighbors')\n",
    "    knn = neighbors.KNeighborsClassifier(n_neighbors=i)\n",
    "    knn.fit(X_train, y_train)\n",
    "    print_classification_statistics(knn, X_test, y_test, labels=['No', 'Yes'])\n",
    "    #plot_ROC(knn, X_test, y_test, label='KNN')\n",
    "    #skplt.metrics.plot_confusion_matrix(y_test, knn.predict(X_test), normalize=False)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No      0.941     0.994     0.966       941\n",
      "        Yes      0.000     0.000     0.000        59\n",
      "\n",
      "avg / total      0.885     0.935     0.909      1000\n",
      "\n",
      "Confusion Matrix:\n",
      "           Predicted          \n",
      "                True     False\n",
      "Real True   0.993624  0.006376\n",
      "     False  1.000000  0.000000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "logistic = skl_lm.LogisticRegression(C=1e10)\n",
    "logistic.fit(X_train, y_train)\n",
    "print_classification_statistics(logistic, X_test, y_test, labels=['No', 'Yes'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No       0.95      0.98      0.96       941\n",
      "        Yes       0.34      0.19      0.24        59\n",
      "\n",
      "avg / total       0.91      0.93      0.92      1000\n",
      "\n",
      "Pred   No  Yes\n",
      "True          \n",
      "No    920   21\n",
      "Yes    48   11\n"
     ]
    }
   ],
   "source": [
    "# using 25% changes of buying instead of 50%\n",
    "pred_p = logistic.predict_proba(X_test)\n",
    "cm_df = pd.DataFrame({'True': y_test, 'Pred': pred_p[:,1] > .25})\n",
    "cm_df.Pred.replace(to_replace={True:'Yes', False:'No'}, inplace=True)\n",
    "print(classification_report(y_test, cm_df.Pred))\n",
    "print(cm_df.groupby(['True', 'Pred']).size().unstack('True').T)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "013d5d7b1aa446d190005ec1533089c0": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "04d85ef6d6bf40b5bc00c8405e54764b": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "layout": "IPY_MODEL_2c8bb972c2e5454bbe1cc7b674889caa",
       "outputs": [
        {
         "name": "stdout",
         "output_type": "stream",
         "text": "Bayes accuracy:  93.4 82.2\nLDA accuracy:  94.0 81.6\n"
        },
        {
         "data": {
          "image/png": "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\n",
          "text/plain": "<Figure size 576x576 with 1 Axes>"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ]
      }
     },
     "060dd90dce484d8488b21ac24a8689a2": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "127f59fb10e6405f9c032090244584ac": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "130a9ffcdb2f4e2a8af156c2b1e5d19d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "VBoxModel",
      "state": {
       "_dom_classes": [
        "widget-interact"
       ],
       "children": [
        "IPY_MODEL_57cd4b3d181f471fab0b9e136ea4beaa",
        "IPY_MODEL_3092abfa779b4e01862e66f0e28b7064",
        "IPY_MODEL_734cdfcb36a74bf9947f2e04a7b69d2f",
        "IPY_MODEL_9185d46ef72f460ba90d7fad794836c6",
        "IPY_MODEL_1505547a028a4959aa8f8225b611d507",
        "IPY_MODEL_36bba20749084534bec436b74418a25f"
       ],
       "layout": "IPY_MODEL_867eb8d642324162ab76f0ec96bc4462"
      }
     },
     "1505547a028a4959aa8f8225b611d507": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "sigma2",
       "layout": "IPY_MODEL_45f98f7f180b47728ee09dc6b66eadee",
       "max": 5,
       "min": 0.1,
       "step": 0.1,
       "style": "IPY_MODEL_5267ded6b1884918af7d6043cc3b9560",
       "value": 0.5
      }
     },
     "270c25a8c2764db39c5c18e0874bbdaa": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "2c8bb972c2e5454bbe1cc7b674889caa": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {
       "height": "15cm"
      }
     },
     "2d488c83ab604ab681ce6414d1a89628": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {
       "height": "15cm"
      }
     },
     "3092abfa779b4e01862e66f0e28b7064": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "mean2",
       "layout": "IPY_MODEL_e9e7ffa7d08b4bbda69a30e02348813e",
       "max": 2,
       "min": -2,
       "step": 0.5,
       "style": "IPY_MODEL_f50eebfa501248babd0ec54efb2931a5",
       "value": 1
      }
     },
     "36bba20749084534bec436b74418a25f": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "layout": "IPY_MODEL_2d488c83ab604ab681ce6414d1a89628",
       "outputs": [
        {
         "name": "stdout",
         "output_type": "stream",
         "text": "Bayes accuracy:  87.8 96.8 89.4\nLDA accuracy:  87.6 96.8 89.8\n"
        },
        {
         "data": {
          "image/png": "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\n",
          "text/plain": "<Figure size 576x576 with 1 Axes>"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ]
      }
     },
     "3bd6103280d2445891fb4c6f003303ca": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "45f98f7f180b47728ee09dc6b66eadee": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "489f535e20da4b1da4108d8a552f05dd": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "5267ded6b1884918af7d6043cc3b9560": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "57cd4b3d181f471fab0b9e136ea4beaa": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "mean1",
       "layout": "IPY_MODEL_7bbb0964c5c84a778ad66c5b2824b0e6",
       "max": 2,
       "min": -2,
       "step": 0.5,
       "style": "IPY_MODEL_b659be192f754c87b171c849edac40ea",
       "value": -2
      }
     },
     "69ea6971801a422a90211f4a0df9751f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "70776e09c27d4f9f8c18c6259b80ca33": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "sigma2",
       "layout": "IPY_MODEL_69ea6971801a422a90211f4a0df9751f",
       "max": 5,
       "min": 0.1,
       "step": 0.1,
       "style": "IPY_MODEL_729dbd1b05f84ff082e5c80bfcebab20",
       "value": 0.9
      }
     },
     "729dbd1b05f84ff082e5c80bfcebab20": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "734cdfcb36a74bf9947f2e04a7b69d2f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "mean3",
       "layout": "IPY_MODEL_127f59fb10e6405f9c032090244584ac",
       "max": 2,
       "min": -2,
       "step": 0.5,
       "style": "IPY_MODEL_060dd90dce484d8488b21ac24a8689a2",
       "value": 2
      }
     },
     "7bbb0964c5c84a778ad66c5b2824b0e6": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "8017aecdb1124b2ba4934d9668b5ece2": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "867eb8d642324162ab76f0ec96bc4462": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "9185d46ef72f460ba90d7fad794836c6": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "sigma1",
       "layout": "IPY_MODEL_013d5d7b1aa446d190005ec1533089c0",
       "max": 5,
       "min": 0.1,
       "step": 0.1,
       "style": "IPY_MODEL_3bd6103280d2445891fb4c6f003303ca",
       "value": 2.3
      }
     },
     "935faf5295db42a69a040fdddb5b8cbd": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "9576e45393bd4da986afcaadb9ea01b3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "962ce5ec503f483884c8b1401be529a0": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "VBoxModel",
      "state": {
       "_dom_classes": [
        "widget-interact"
       ],
       "children": [
        "IPY_MODEL_d24d66952038418b93b6c54715023112",
        "IPY_MODEL_e7d07c9ceb3c479bb156d2b5f1ea8a67",
        "IPY_MODEL_cbea21d7a40a40be986b9d0e3c66e8b8",
        "IPY_MODEL_70776e09c27d4f9f8c18c6259b80ca33",
        "IPY_MODEL_04d85ef6d6bf40b5bc00c8405e54764b"
       ],
       "layout": "IPY_MODEL_c419dea5bd784b7fa242a308b465b6f4"
      }
     },
     "9be199aed5b040c5867c67e4a560a69f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "b659be192f754c87b171c849edac40ea": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "c419dea5bd784b7fa242a308b465b6f4": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "cbea21d7a40a40be986b9d0e3c66e8b8": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "sigma1",
       "layout": "IPY_MODEL_935faf5295db42a69a040fdddb5b8cbd",
       "max": 5,
       "min": 0.1,
       "step": 0.1,
       "style": "IPY_MODEL_9be199aed5b040c5867c67e4a560a69f",
       "value": 3.3
      }
     },
     "d24d66952038418b93b6c54715023112": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "mean1",
       "layout": "IPY_MODEL_489f535e20da4b1da4108d8a552f05dd",
       "max": 2,
       "min": -2,
       "step": 0.5,
       "style": "IPY_MODEL_9576e45393bd4da986afcaadb9ea01b3",
       "value": -2
      }
     },
     "e7d07c9ceb3c479bb156d2b5f1ea8a67": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "FloatSliderModel",
      "state": {
       "description": "mean2",
       "layout": "IPY_MODEL_270c25a8c2764db39c5c18e0874bbdaa",
       "max": 2,
       "min": -2,
       "step": 0.5,
       "style": "IPY_MODEL_8017aecdb1124b2ba4934d9668b5ece2",
       "value": 0.5
      }
     },
     "e9e7ffa7d08b4bbda69a30e02348813e": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.0.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "f50eebfa501248babd0ec54efb2931a5": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.2.0",
      "model_name": "SliderStyleModel",
      "state": {
       "description_width": ""
      }
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}