{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1><center> Predictive statistics on the I-SPY1 Clinical Trial</center></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 0. Load modules and clean data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "# import custom modules wrote by julio\n",
    "from ispy1 import predictive_statistics\n",
    "\n",
    "# reload modules without restartign the kernel (makes development easier)\n",
    "import importlib\n",
    "importlib.reload(predictive_statistics);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SUBJECTID</th>\n",
       "      <th>age</th>\n",
       "      <th>White</th>\n",
       "      <th>ER+</th>\n",
       "      <th>PR+</th>\n",
       "      <th>HR+</th>\n",
       "      <th>Bilateral</th>\n",
       "      <th>Right_Breast</th>\n",
       "      <th>MRI_LD_Baseline</th>\n",
       "      <th>MRI_LD_1_3dAC</th>\n",
       "      <th>MRI_LD_Int_Reg</th>\n",
       "      <th>MRI_LD_PreSurg</th>\n",
       "      <th>Alive</th>\n",
       "      <th>Survival_length</th>\n",
       "      <th>RFS</th>\n",
       "      <th>RFS_code</th>\n",
       "      <th>PCR</th>\n",
       "      <th>RCB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1001</td>\n",
       "      <td>38.73</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>88.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>No</td>\n",
       "      <td>1264</td>\n",
       "      <td>751</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1002</td>\n",
       "      <td>37.79</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>29.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>No</td>\n",
       "      <td>1155</td>\n",
       "      <td>1043</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SUBJECTID    age White  ER+  PR+  HR+ Bilateral Right_Breast  \\\n",
       "0       1001  38.73   Yes  Yes   No  Yes        No           No   \n",
       "1       1002  37.79   Yes  Yes  Yes  Yes        No          Yes   \n",
       "\n",
       "   MRI_LD_Baseline  MRI_LD_1_3dAC  MRI_LD_Int_Reg  MRI_LD_PreSurg Alive  \\\n",
       "0             88.0           78.0            30.0            14.0    No   \n",
       "1             29.0           26.0            66.0            16.0    No   \n",
       "\n",
       "   Survival_length   RFS  RFS_code PCR  RCB  \n",
       "0             1264   751         1  No  2.0  \n",
       "1             1155  1043         1  No  3.0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('./data/I-SPY_1_clean_data.csv')\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1><center> Prediction of categorical outcomes</center></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.0 Pathological Complete Response (PCR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAAEWCAYAAACDoeeyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSeqYIS9JHTPkJaljhrwkdcyQl6SO/T++vL1YKAmi+QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113d4dba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# allocate  outcome \n",
    "outcome = 'PCR'\n",
    "y = predictive_statistics.labels_to_numbers(df, outcome);\n",
    "\n",
    "# check how unbalanced the data are\n",
    "df[outcome].value_counts(normalize = True).plot.barh();\n",
    "plt.title('Normalized Sample size for PCR')\n",
    "plt.xlabel('group size (%)');\n",
    "plt.savefig('Sample_Size_PCR.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# allocate continous predictors\n",
    "cont_predictors = ['age','MRI_LD_Baseline', 'MRI_LD_1_3dAC', 'MRI_LD_Int_Reg', 'MRI_LD_PreSurg']\n",
    "X_cont = df[cont_predictors].values\n",
    "\n",
    "# allocate clinical predictors\n",
    "cat_predictors = ['White', 'ER+', 'PR+', 'HR+'];\n",
    "X_cat = pd.pandas.get_dummies(df[cat_predictors], drop_first=True).values\n",
    "\n",
    "# allocate a single predictors matrix X\n",
    "X = np.concatenate( (X_cont, X_cat), axis=1)\n",
    "\n",
    "# allocate  outcome \n",
    "outcome = 'PCR'\n",
    "y = predictive_statistics.labels_to_numbers(df, outcome);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Why Kappa is a good metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       1.00      0.74      0.85       167\n",
      "          1       0.02      1.00      0.04         1\n",
      "\n",
      "avg / total       0.99      0.74      0.84       168\n",
      "\n",
      "Kappa = 0.032\n",
      "AUC = 0.868\n",
      "Accuracy = 0.738\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "def mymetrics(ypredicted, yexpected):\n",
    "    print(metrics.classification_report(ypredicted, yexpected))\n",
    "    k = metrics.cohen_kappa_score(ypredicted, yexpected); k = np.round(k,3);\n",
    "    auc = metrics.roc_auc_score(ypredicted, yexpected);   auc  = np.round(auc,3);\n",
    "    accuracy = metrics.accuracy_score(ypredicted, yexpected); accuracy = np.round(accuracy,3);\n",
    "\n",
    "    print(\"Kappa = \" + str(k))\n",
    "    print(\"AUC = \" + str(auc))\n",
    "    print(\"Accuracy = \" + str(accuracy))\n",
    "\n",
    "# make at least one observation positive\n",
    "y_crazy = np.zeros_like(y)\n",
    "y_crazy[np.argwhere(y>0)[0]] = 1\n",
    "mymetrics(y_crazy, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- ### logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.77      0.87      0.82        39\n",
      "          1       0.29      0.17      0.21        12\n",
      "\n",
      "avg / total       0.66      0.71      0.68        51\n",
      "\n",
      "The estimated Cohen kappa is 0.0449438202247\n",
      "The estimated AUC is 0.519\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.82      0.72      0.77        39\n",
      "          1       0.35      0.50      0.41        12\n",
      "\n",
      "avg / total       0.71      0.67      0.68        51\n",
      "\n",
      "The estimated Cohen kappa is 0.190476190476\n",
      "The estimated AUC is 0.609\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZmaW4aRgZmYZTgpmZpbx/02Nj+L/47EsAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1076f9208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# standard\n",
    "auc1, kappa1, fpr1, tpr1 = predictive_statistics.Logistic_Regression(X, y, oversample = False)\n",
    "\n",
    "# oversampled\n",
    "auc2, kappa2, fpr2, tpr2 = predictive_statistics.Logistic_Regression(X, y, oversample = True, K_neighbors = 4)\n",
    "\n",
    "title ='Effect of oversampling on Logistic Regression for PCR'\n",
    "predictive_statistics.plot_compare_roc(fpr1, tpr1,fpr2, tpr2, auc1, auc2, title = title)\n",
    "plt.savefig('PCR_Logistic.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- ### random forests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.80      0.92      0.86        39\n",
      "          1       0.50      0.25      0.33        12\n",
      "\n",
      "avg / total       0.73      0.76      0.73        51\n",
      "\n",
      "The estimated Cohen kappa is 0.209302325581\n",
      "The estimated AUC is 0.587\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAckAAAEWCAYAAAANV2yLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x108c5abe0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.79      0.87      0.83        39\n",
      "          1       0.38      0.25      0.30        12\n",
      "\n",
      "avg / total       0.69      0.73      0.70        51\n",
      "\n",
      "The estimated Cohen kappa is 0.13768115942\n",
      "The estimated AUC is 0.561\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113db4d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAckAAAEWCAYAAAANV2yLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJuEfCAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113f553c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113db4588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# standard\n",
    "auc1, kappa1, fpr1, tpr1, forest = predictive_statistics.RandomForest_Classifier(X, y,oversample=False)\n",
    "t ='Feature Importance for RFC unbalanced'\n",
    "predictive_statistics.plot_forest_feature_importances_(forest, cont_predictors + cat_predictors, title = t)\n",
    "plt.savefig('RFC_std_PCR.png')\n",
    "\n",
    "# unbalanced learning\n",
    "auc2, kappa2, fpr2, tpr2, Forest = predictive_statistics.RandomForest_Classifier(X, y, oversample=True, K_neighbors = 5)\n",
    "t ='Feature Importance for RFC ADASYN'\n",
    "predictive_statistics.plot_forest_feature_importances_(Forest, cont_predictors + cat_predictors, title = t)\n",
    "plt.savefig('RFC_Adasyn_PCR.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113f10ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare\n",
    "title ='Effect of oversampling on RFC for PCR'\n",
    "predictive_statistics.plot_compare_roc(fpr1, tpr1,fpr2, tpr2, auc1, auc2, title = title)\n",
    "plt.savefig('RFC_std_PCR.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.82      0.72      0.77        39\n",
      "          1       0.35      0.50      0.41        12\n",
      "\n",
      "avg / total       0.71      0.67      0.68        51\n",
      "\n",
      "The estimated Cohen kappa is 0.190476190476\n",
      "The estimated AUC is 0.609\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.84      0.69      0.76        39\n",
      "          1       0.37      0.58      0.45        12\n",
      "\n",
      "avg / total       0.73      0.67      0.69        51\n",
      "\n",
      "The estimated Cohen kappa is 0.229333333333\n",
      "The estimated AUC is 0.638\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.82      0.72      0.77        39\n",
      "          1       0.35      0.50      0.41        12\n",
      "\n",
      "avg / total       0.71      0.67      0.68        51\n",
      "\n",
      "The estimated Cohen kappa is 0.190476190476\n",
      "The estimated AUC is 0.609\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "K = [4,5,6]\n",
    "for k in K:\n",
    "    predictive_statistics.Logistic_Regression(X, y, oversample=True, K_neighbors = k)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.0 Survival (`Alive`) using Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAAEWCAYAAACDoeeyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAFolJREFUeJzt3XmYZXV95/H3B5oGWQSlcQFtOjouEfcAiUQNUWZUEAmB\nGYmIW4ySx4UxIYHER4JbQmZ0NDgjDmpEjbhvEUUQSauAjDQGWdUBZRHjArIrSMN3/ji/hkulqvtW\ndS2X37xfz3OfOts953tO3frc3/mde0+lqpAk9WmTpS5AkrRwDHlJ6pghL0kdM+QlqWOGvCR1zJCX\npI4Z8pq1JEcn+ac2vDLJzUk2nedtXJ5kr/lc53xLsmeSH83zOp+W5Hvzuc623rckuSbJT+Z73TNs\nb3WSl7fhg5Ocuhjb1b9nyE+gFnA/S7LVyLSXJ1m9hGVNq6qurKqtq+qOxdpmkock+XQLrRuSXJjk\nJYu1/YVUVd+oqkfN5zqTrAT+HHhMVT1oHtebJD9IcvH6lquqj1TVf5qv7Wp2DPnJtSlw2MaupP0h\n9vZ7/jBwFbAzsD1wCPDTJa1osq0Erq2qn832iUmWrWf204EHAA9Lsttci9PC6u2Pvyf/HTg8yXbT\nzUyyR5JzWkv2nCR7jMxbneStSc4EfsnwR7i6nbKf1bpXvpBk+yQfSXJjW8eqkXX8Q5Kr2rxzkzxt\nhjpWJakky5I8pa173ePWJJe35TZJcmSSy5Jcm+QTSe4/sp5DklzR5r1+A8dmN+CEqrqlqtZW1b9W\n1ckj6/pkkp+0Y/P1JLuMzDshybuTnNxqPDPJg5K8M8l1Sb6b5Ekjy1+e5K+SXNzmfyDJFjMcix3b\nGcbPk/wwyWtn2oEke7d13pTk6iSHt+l3dQElef6U43nburO5JJsneVuSK5P8NMl7ktxnmu3sBXwF\n2LGt44Q2/XlJLkpyfXtt/OaUfT4iyfnALesJ+hcDnwe+1IZn2teXJDmjDR+X5G1T5n8+yZ/N9hhq\nTFXlY8IewOXAXsBngLe0aS8HVrfh+wPXMbRglwF/1Ma3b/NXA1cCu7T5m7VplwIPB7YFLga+37az\nDPgQ8IGRGl7I0EpexnCq/xNgizbvaOCf2vAqoIBlU/ZhM+BrwN+18cOAs4GHAJsD/xv4aJv3GOBm\nhpbh5sD/ANYCe81wfE4DzgQOAlZOM/9lwDZtXe8EzhuZdwJwDfBbwBbA6cAPgRcxnD29BfiXKb+L\nC4GHtuN+5sjvZE/gR214E+Bc4ChgOfAw4AfAs2bYh38DntaG7wc8eeo6pyx/X+AS4JVt/B3AP7ea\ntgG+sO5YT/Pce6wTeCRwC/Af2+/pL9trY/nIPp/X9vk+M6xzS+BGYG/ggHZMl4/MXw28vA2/BDij\nDT+d4SwsI/v+K2DH2R5DH2PmyVIX4GOaX8rdIf9Y4AZgB+4Z8ocA35rynG8CL2nDq4E3TZm/Gnj9\nyPjbgZNHxvdlJAynqek64Alt+Gg2HPLHAScBm7TxS4Bnjsx/MHA7w5vIUcDHRuZtBfyamUP+fsAx\nwEXAHS2Qdpth2e1afdu28ROA947Mfw1wycj444Drp/wuDh0Z3xu4rA3vyd0h/9vAlVO2/VeMvHFO\nmXcl8ErgvlOm37XOkWmbtGN5XBsPQ0g/fGSZpwA/nGFb91gn8AbgE1PWfzWw58g+v2wDr9EXAj9v\nv78tGF6n+095vU0X8mn7/vQ2/ifA6XM5hj7Ge9hdM8Gq6kKGP+4jp8zaEbhiyrQrgJ1Gxq+aZpWj\n/da/mmZ863UjSQ5Pcknr8rieofW/Ypy6k7ySIVheUFV3tsk7A59t3QPXM4T+HcAD2/7cVW9V3QJc\nO9P6q+q6qjqyqnZpzz8P+Fy7/rBpkmNat9CNDIHFlNrHPg7N6LG8otU71c4MXSLXj+zjX7f6pnMA\nwxvGFUm+luQpM+0v8FaG1vq6rosdGFrS545s68tt+jju8fppv6Or2PDrZ9SLGd4o1lbVrcCnWU+X\nzci2CvgYw9knwAuAj7Th2R5DjWF9F1U0Gf4G+DZDy3udHzP8QYxayfCHvs6cby/a+t//EngmcFFV\n3ZnkOoZW2DjPfTPw1Kq6cWTWVQytwzOnec6/AaN9wlsydBVtUFVd0/p4X8zQdbE3sB/DmdDlDG9O\nY9W+Hg8dGV7JcPynuoqhJf2IcVZYVecA+yXZDHg18Ikp2wEgyUEMgbhbVd3eJl/D8Ga0S1VdPfZe\n3O3HDGcs67aRtu3Rdc34+knyEOAZwO5JDmiTtwS2SLKiqq7ZwPY/Cpya5BiG1vv+bfqsjqHGY0t+\nwlXVpcDHubsVB8OFrkcmeUG74Pl8hn7tk+Zps9sw9In/HFiW5CiGPuH1SvJQhrB6UVV9f8rs9wBv\nTbJzW3aHJPu1eZ8CnpvkqUmWA29iPa/NJH+f5LFt37cB/hS4tKqubbXfxnAmsCXwt2Pv9cxeleFj\nm/cHXs/w+5jqW8BN7YLlfdoZxWMzzadOkizP8NnxbVtw3wjcOc1yTwLeBfxBVf183fTW8n4v8I4k\nD2jL7pTkWWPuzyeAfZI8s73J/DnDMTtrzOcfwnA951HAE9vjkcCPuLuFPqOq+leGN6r3AadU1fVt\n1tjHUOMz5O8d3sTQTw1AC7PnMvxxXsvQ6n7uGC2ocZ3CcFbwfYbT+lvZ8Ok7DC3/BwKfGvlEyEVt\n3j8wXCg8NclNDBdhf7vtz0XAq4ATGS5IXscQGDPZEvgscD3Dhbmdgee1eR9qNV/NcHH57DHq3pAT\ngVPbti5juDh7DzV8T+C5DIH3Q+4OsW1nWOchwOWtS+lQ4OBpltmP4frDGSPHc92niI5guFh6dlvH\naQyhu0FV9T2GPvV3tTr3Bfatql+P83yGs6Z3V9VPRh8Mb+Qb7LJpTmQ42zpxpK7ZHkONYd0VbknT\nyPAR0JdX1WlLXYs0F7bkJaljhrwkdczuGknqmC15SerYkn9OfsWKFbVq1aqlLkOS7jXOPffca6pq\nrC+/LXnIr1q1ijVr1ix1GZJ0r5Fk6jfeZ2R3jSR1zJCXpI4Z8pLUMUNekjpmyEtSxwx5SeqYIS9J\nHTPkJaljhrwkdcyQl6SOGfKS1DFDXpI6ZshLUscMeUnqmCEvSR0z5CWpY4a8JHXMkJekjhnyktQx\nQ16SOmbIS1LHDHlJ6pghL0kdM+QlqWOGvCR1zJCXpI4tW+oCLrj6BlYd+cWlLuNe4fJj9lnqEiTd\ny9iSl6SOGfKS1DFDXpI6ZshLUscMeUnqmCEvSR0z5CWpY4a8JHXMkJekjhnyktQxQ16SOmbIS1LH\nDHlJ6pghL0kdM+QlqWOGvCR1zJCXpI4Z8pLUMUNekjpmyEtSxwx5SeqYIS9JHTPkJaljsw75JJXk\n7SPjhyc5el6rkiTNi7m05G8D/jDJivkuRpI0v+YS8muB44HXTZ2RZFWS05Ocn+SrSVZudIWSpDmb\na5/8/wIOTrLtlOnvAj5YVY8HPgIcO92Tk7wiyZoka+745Q1zLEGStCFzCvmquhH4EPDaKbOeApzY\nhj8MPHWG5x9fVbtW1a6bbjn1fUKSNF825tM17wT+GNhqnmqRJM2zOYd8Vf0C+ARD0K9zFnBQGz4Y\n+MbcS5MkbayN/Zz824HRT9m8BnhpkvOBQ4DDNnL9kqSNsGy2T6iqrUeGfwpsOTJ+BfCM+SlNkrSx\n/MarJHXMkJekjhnyktQxQ16SOmbIS1LHDHlJ6pghL0kdM+QlqWOGvCR1zJCXpI4Z8pLUMUNekjpm\nyEtSxwx5SeqYIS9JHTPkJaljhrwkdcyQl6SOGfKS1DFDXpI6ZshLUseWLXUBj9tpW9Ycs89SlyFJ\nXbIlL0kdM+QlqWOGvCR1zJCXpI4Z8pLUMUNekjpmyEtSxwx5SeqYIS9JHTPkJaljhrwkdcyQl6SO\nGfKS1DFDXpI6ZshLUscMeUnqmCEvSR0z5CWpY4a8JHXMkJekjhnyktQxQ16SOmbIS1LHDHlJ6pgh\nL0kdM+QlqWOGvCR1zJCXpI4Z8pLUMUNekjpmyEtSxwx5SeqYIS9JHTPkJaljhrwkdcyQl6SOGfKS\n1DFDXpI6ZshLUscMeUnqmCEvSR0z5CWpY4a8JHXMkJekjhnyktQxQ16SOmbIS1LHDHlJ6pghL0kd\nM+QlqWOGvCR1zJCXpI4Z8pLUsWVLXcAFV9/AqiO/uNRlSNKiufyYfRZtW7bkJaljhrwkdcyQl6SO\nGfKS1DFDXpI6ZshLUscMeUnqmCEvSR0z5CWpY4a8JHXMkJekjhnyktQxQ16SOmbIS1LHDHlJ6pgh\nL0kdM+QlqWOGvCR1zJCXpI4Z8pLUMUNekjpmyEtSxwx5SerYekM+gzOSPGdk2n9O8uWFL02StLGW\nrW9mVVWSQ4FPJvmXtvzfAs9ejOIkSRtnvSEPUFUXJvkCcASwFfChqrosyYuBVwHLgbOAVzOcGXwA\neCIQ4PiqOnahipckrd8GQ755I/Bt4NfArkkeC+wP7FFVa5McDxwEXAasqKrHASTZbrqVJXkF8AqA\nTe+7w8btgSRpRmOFfFXdkuTjwM1VdVuSvYDdgDVJAO4DXAWcAjwqybHAF4FTZ1jf8cDxAJs/+BG1\n0XshSZrWuC15gDvbA4aumH+sqjdMXSjJ44HnMHTlHEBrsUuSFt9cP0J5GvBfkqwASLJ9kpVJdgBS\nVZ8EjgKePE91SpLmYDYt+btU1QVJ3giclmQT4HbgUOAO4P0Z+nCK4WKtJGmJjB3yVXX0lPETgROn\nWfRJG1mTJGme+I1XSeqYIS9JHTPkJaljhrwkdcyQl6SOGfKS1DFDXpI6ZshLUscMeUnqmCEvSR0z\n5CWpY4a8JHXMkJekjhnyktQxQ16SOmbIS1LHDHlJ6pghL0kdM+QlqWOGvCR1zJCXpI4Z8pLUsWVL\nXcDjdtqWNcfss9RlSFKXbMlLUscMeUnqmCEvSR0z5CWpY4a8JHXMkJekjhnyktQxQ16SOmbIS1LH\nDHlJ6pghL0kdM+QlqWOGvCR1zJCXpI4Z8pLUMUNekjpmyEtSxwx5SeqYIS9JHTPkJaljhrwkdcyQ\nl6SOGfKS1DFDXpI6ZshLUscMeUnqWKpqaQtIbgK+t6RFzGwFcM1SFzGDSa4NJru+Sa4NJrs+a5u7\n+axv56raYZwFl83TBjfG96pq16UuYjpJ1ljb3ExyfZNcG0x2fdY2d0tVn901ktQxQ16SOjYJIX/8\nUhewHtY2d5Nc3yTXBpNdn7XN3ZLUt+QXXiVJC2cSWvKSpAViyEtSxxYl5JM8O8n3klya5Mhp5ifJ\nsW3++UmevBh1zaK+Ryf5ZpLbkhw+YbUd3I7ZBUnOSvKECaptv1bbeUnWJHnqYtU2Tn0jy+2WZG2S\nAyeltiR7JrmhHbvzkhy1WLWNU99IjecluSjJ1yaltiR/MXLcLkxyR5L7T0ht2yb5QpLvtOP20gUv\nqqoW9AFsClwGPAxYDnwHeMyUZfYGTgYC/A7wfxa6rlnW9wBgN+CtwOETVtsewP3a8HMW69iNWdvW\n3H3d5/HAdyfp2I0sdzrwJeDASakN2BM4abGO1xzq2w64GFjZxh8wKbVNWX5f4PRJqQ34a+Dv2/AO\nwC+A5QtZ12K05HcHLq2qH1TVr4GPAftNWWY/4EM1OBvYLsmDF6G2seqrqp9V1TnA7YtU02xqO6uq\nrmujZwMPmaDabq72aga2AhbzKv84rzuA1wCfBn42gbUtlXHqewHwmaq6Eoa/kQmqbdQfAR9dlMrG\nq62AbZKEoRH0C2DtQha1GCG/E3DVyPiP2rTZLrNQlnLbGzLb2v6Y4YxoMYxVW5L9k3wX+CLwskWq\nDcaoL8lOwP7AcYtYF4z/e92jdXednGSXxSkNGK++RwL3S7I6yblJXjRBtQGQZEvg2Qxv4othnNr+\nJ/CbwI+BC4DDqurOhSxqEm5roHmQ5PcZQn5R+703pKo+C3w2ydOBNwN7LXFJo94JHFFVdw4Nq4ny\nbYaukJuT7A18DnjEEtc0ahnwW8AzgfsA30xydlV9f2nLuod9gTOr6hdLXciIZwHnAc8AHg58Jck3\nqurGhdrgYrTkrwYeOjL+kDZttssslKXc9oaMVVuSxwPvA/arqmsnqbZ1qurrwMOSrFjowppx6tsV\n+FiSy4EDgXcn+YNJqK2qbqyqm9vwl4DNJuzY/Qg4papuqaprgK8Di3HRfzavu4NYvK4aGK+2lzJ0\nc1VVXQr8EHj0gla1CBcjlgE/AH6Duy9G7DJlmX2454XXby10XbOpb2TZo1ncC6/jHLuVwKXAHotV\n1yxq+w/cfeH1yQwv+ExKfVOWP4HFu/A6zrF70Mix2x24cpKOHUOXw1fbslsCFwKPnYTa2nLbMvR3\nb7UYx2wWx+044Og2/MD2N7FiIeta8O6aqlqb5NXAKQxXn/+xqi5Kcmib/x6GTzbszRBWv2R4t1sU\n49SX5EHAGuC+wJ1J/ivDVfMFO8UatzbgKGB7hlYowNpahDvdjVnbAcCLktwO/Ap4frVX94TUtyTG\nrO1A4E+TrGU4dgdN0rGrqkuSfBk4H7gTeF9VXTgJtbVF9wdOrapbFrqmWdb2ZuCEJBcwNGqPqOFM\naMF4WwNJ6pjfeJWkjhnyktQxQ16SOmbIS1LHDHlJ6pghL21AkkPn42v7SZ6U5P1t+IB2F8JvJNm+\nTXt4ko+PLL88ydeT+M10zZkfodS9TpJlVbWgN3VaCEk+Cbylqr6TZDXDd0P+kOEuou9K8lHgqKr6\nvyPP+RuGm159ZEmK1r2eLXlNlCRvaPfjPiPJR9Pu399uhPXOJGuAw5KsSnJ6u4HXV5OsbMudMHpf\n+CQ3t597tlbxF9v635Pk373+kxyT5OK23re1aUcnOTzJjiP3KT+v3ad85yQ7JPl0knPa43enWe82\nwOOr6jtt0p3A5gzfFr09ydOAn4wGfPM54OCNPKz6/5ingZoYSXZj+JbsE4DNGG7Sde7IIsvXfZs3\nyReAD1bVB5O8DDgW2NB9Z3YHHgNcAXyZoRX9qZHtb8/wTclHV1Ul2W70yVX1Y+CJbdlXAb9XVVck\nORF4R1Wd0d5sTmH42v+oXRm++r/O3wGnMdyN8IXAJxnutTLVhQz/y0CaE0Nek+R3gc9X1a3ArS3I\nR318ZPgpDCEN8GHgv42x/m9V1Q8AWtfIUxkJeeAG4Fbg/UlOAk6abiWtpf4n3H3Hz72Ax4zcyfK+\nSbaudoOx5sHAz9eNVNVXgK+09b2I4dYej2xnLtcx3IL2l1V1R5JfJ9mmqm4aYx+lezDkdW8yzn1I\n1tK6IVt3zPKReVMvQN1jvN17ZHeG2+ceCLya4Zawd8nwz2zeDzxvJMQ3AX6nvTnN5FfAFlMntnue\nv4ThFrQnMbxxHcjQRfPettjmDG8+0qzZJ69Jciawb5ItkmwNPHc9y57F3d0bBwPfaMOXM9znHOB5\nDN0+6+ye5Dda+D8fOGN0hW2b29Zwa9/XMeXWuUk2Y+hWOaLued/0Uxn+w9S65Z44Tb2XMNyVc6q/\nAI6tqtsZ7steDP31W7Z1bQ9c0+ZLs2bIa2LU8C8W/5nhzoYnM/znnBtmWPw1wEuTnA8cAhzWpr8X\n+L0k32Ho0hlt/Z/D8J95LmG4j/dnp6xzG+Ckts4zgD+bMn8Phr71N45cfN0ReC2wa7tYezFw6DT7\n9l1g23YBFoD23N2r6nNt0rtajYcCJ7Zpv8/wX7WkOfEjlJoo6/qyWzfG14FXVNW352G9ezL8L4D1\nnR0sqCSvA26qqvfN4jmfAY6syfqPS7oXsSWvSXN8kvMYPlnz6fkI+AlyHHDbuAsnWQ58zoDXxrAl\nL0kdsyUvSR0z5CWpY4a8JHXMkJekjhnyktSx/wdCZD++/mthWwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113db40b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# allocate  outcome \n",
    "outcome = 'Alive'\n",
    "y = predictive_statistics.labels_to_numbers(df, outcome);\n",
    "\n",
    "# check how unbalanced the data are\n",
    "df[outcome].value_counts(normalize = True).plot.barh();\n",
    "plt.title('Normalized Sample size for Alive')\n",
    "plt.xlabel('group size (%)');\n",
    "plt.savefig('Sample_Size_Alive.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- ### Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.50      0.27      0.35        11\n",
      "          1       0.82      0.93      0.87        40\n",
      "\n",
      "avg / total       0.75      0.78      0.76        51\n",
      "\n",
      "The estimated Cohen kappa is 0.236734693878\n",
      "The estimated AUC is 0.599\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.36      0.45      0.40        11\n",
      "          1       0.84      0.78      0.81        40\n",
      "\n",
      "avg / total       0.73      0.71      0.72        51\n",
      "\n",
      "The estimated Cohen kappa is 0.208893485005\n",
      "The estimated AUC is 0.615\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# standard\n",
    "auc1, kappa1, fpr1, tpr1 = predictive_statistics.Logistic_Regression(X, y)\n",
    "\n",
    "# unbalanced learning\n",
    "auc2, kappa2, fpr2, tpr2 = predictive_statistics.Logistic_Regression(X, y, oversample=True, K_neighbors = 4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYPJjVGnDAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113db48d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "title ='Effect of oversampling on Logistic Regression for Alive'\n",
    "predictive_statistics.plot_compare_roc(fpr1, tpr1,fpr2, tpr2, auc1, auc2, title = title)\n",
    "plt.savefig('Alive_Logistic.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- ### random forests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.38      0.27      0.32        11\n",
      "          1       0.81      0.88      0.84        40\n",
      "\n",
      "avg / total       0.72      0.75      0.73        51\n",
      "\n",
      "The estimated Cohen kappa is 0.16393442623\n",
      "The estimated AUC is 0.574\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.67      0.36      0.47        11\n",
      "          1       0.84      0.95      0.89        40\n",
      "\n",
      "avg / total       0.81      0.82      0.80        51\n",
      "\n",
      "The estimated Cohen kappa is 0.375510204082\n",
      "The estimated AUC is 0.657\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGkkKERHRSFKIiIjGb51lcE5r3lPhAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x107c5d828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# standard\n",
    "auc1, kappa1, fpr1, tpr1, _= predictive_statistics.RandomForest_Classifier(X, y)\n",
    "\n",
    "# unbalanced learning\n",
    "auc2, kappa2, fpr2, tpr2, _= predictive_statistics.RandomForest_Classifier(X, y, oversample=True, K_neighbors = 6)\n",
    "\n",
    "title ='Effect of oversampling on RFC for PCR'\n",
    "predictive_statistics.plot_compare_roc(fpr1, tpr1,fpr2, tpr2, auc1, auc2, title = title)\n",
    "plt.savefig('Alive_RFC.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.0 Survival (`Alive`) including PCR as predictor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## - Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# allocate new predictor variable\n",
    "pcr = predictive_statistics.labels_to_numbers(df, 'PCR').reshape(168,1)\n",
    "newX = np.concatenate((X,pcr), axis  = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.43      0.27      0.33        11\n",
      "          1       0.82      0.90      0.86        40\n",
      "\n",
      "avg / total       0.73      0.76      0.74        51\n",
      "\n",
      "The estimated Cohen kappa is 0.198952879581\n",
      "The estimated AUC is 0.586\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.38      0.45      0.42        11\n",
      "          1       0.84      0.80      0.82        40\n",
      "\n",
      "avg / total       0.74      0.73      0.73        51\n",
      "\n",
      "The estimated Cohen kappa is 0.238805970149\n",
      "The estimated AUC is 0.627\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# standard\n",
    "auc1, kappa1, fpr1, tpr1 = predictive_statistics.Logistic_Regression(newX, y)\n",
    "\n",
    "# unbalanced learning\n",
    "auc2, kappa2, fpr2, tpr2 = predictive_statistics.Logistic_Regression(newX, y, oversample=True, K_neighbors = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbIAAAEWCAYAAAAD/hLkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlMhERCSnKZGJiEhOUyITEZGc9v8BWtdLyqkY3OAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113f22128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "title ='Effect of oversampling on Logistic Regression for Alive (PCR included)'\n",
    "predictive_statistics.plot_compare_roc(fpr1, tpr1,fpr2, tpr2, auc1, auc2, title = title)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## - Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.33      0.09      0.14        11\n",
      "          1       0.79      0.95      0.86        40\n",
      "\n",
      "avg / total       0.69      0.76      0.71        51\n",
      "\n",
      "The estimated Cohen kappa is 0.0555555555556\n",
      "The estimated AUC is 0.52\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# standard\n",
    "auc1, kappa1, fpr1, tpr1 , _ = predictive_statistics.RandomForest_Classifier(newX, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.38      0.45      0.42        11\n",
      "          1       0.84      0.80      0.82        40\n",
      "\n",
      "avg / total       0.74      0.73      0.73        51\n",
      "\n",
      "The estimated Cohen kappa is 0.238805970149\n",
      "The estimated AUC is 0.627\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# unbalanced learning\n",
    "auc2, kappa2, fpr2, tpr2, _ = predictive_statistics.RandomForest_Classifier(newX, y, oversample=True, K_neighbors = 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.0 Survival (`Alive`) including `RCB` as predictor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.50      0.36      0.42        11\n",
      "          1       0.84      0.90      0.87        40\n",
      "\n",
      "avg / total       0.76      0.78      0.77        51\n",
      "\n",
      "The estimated Cohen kappa is 0.292559899117\n",
      "The estimated AUC is 0.632\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.50      0.45      0.48        11\n",
      "          1       0.85      0.88      0.86        40\n",
      "\n",
      "avg / total       0.78      0.78      0.78        51\n",
      "\n",
      "The estimated Cohen kappa is 0.340775558167\n",
      "The estimated AUC is 0.665\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "rcb = pd.get_dummies(df['RCB']).values\n",
    "newX = np.concatenate((X,rcb), axis  = 1)\n",
    "\n",
    "# standard\n",
    "auc1, kappa1, fpr1, tpr1 = predictive_statistics.Logistic_Regression(newX, y)\n",
    "auc2, kappa2, fpr2, tpr2, _= predictive_statistics.RandomForest_Classifier(newX, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113dc5cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare\n",
    "title ='Predicting Survival including RCB as a predictor'\n",
    "predictive_statistics.plot_compare_roc(fpr1, tpr1,fpr2, tpr2, auc1, auc2, title = title)\n",
    "plt.legend(['Logistic Regression','Random Forest Classifier']);\n",
    "plt.savefig('LG_vs_RFC_Alive_RCB.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.33      0.45      0.38        11\n",
      "          1       0.83      0.75      0.79        40\n",
      "\n",
      "avg / total       0.73      0.69      0.70        51\n",
      "\n",
      "The estimated Cohen kappa is 0.180722891566\n",
      "The estimated AUC is 0.602\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n",
      "Data was oversampled using the ADASYN method\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.30      0.27      0.29        11\n",
      "          1       0.80      0.82      0.81        40\n",
      "\n",
      "avg / total       0.70      0.71      0.70        51\n",
      "\n",
      "The estimated Cohen kappa is 0.101057579318\n",
      "The estimated AUC is 0.549\n",
      "============================================================\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# unbalanced learning\n",
    "auc3, kappa3, fpr3, tpr3 = predictive_statistics.Logistic_Regression(newX, y, oversample=True, K_neighbors = 4)\n",
    "auc4, kappa4, fpr4, tpr4, _= predictive_statistics.RandomForest_Classifier(newX, y, oversample=True, K_neighbors = 4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEWCAYAAACT7WsrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xd4FOX2wPHvSQiGjhSRIgQUlBaCdBFQkaqAelVUpFlR\nASsXC1fRq7+Lol6vgqIiYkFQsaCCgiCKhSJopEuTEkSpoYqQcH5/vJOwLNlkA9ndbHI+z7NPsjOz\nM2dmdufM+87M+4qqYowxxmQlJtIBGGOMyb8sSRhjjAnIkoQxxpiALEkYY4wJyJKEMcaYgCxJGGOM\nCajQJQkRSRARFZEi3vvPRaTvCcynuojsE5HYvI8yfESkl4jMyIP59BOR7wKMy5NtJSLjReTxvIjP\ni6fWycRjQic/7S8ROUVElotI5UgsPxREpJKIrBCRU3KaNl8mCRFZLyJ/eV+MP72DQ8lQLEtVu6jq\nG0HGdLHP5zaqaklVTc/rmESkmoh8ICLbRWS3iCwVkX55vRwAVZ2gqh1DMW+fZYRsW50oL551eT1f\nERkuIoe9726qiPwgIq38piktIs+JyEZvurXe+wreeN/v/y4RmSoiZ+R1rNEkmP3lfwKYh24B5qjq\nFr/lDfeW18JveD8RSff23z4R+U1EXheROlnEXNKb5vMsxp3vfX92i8hOEfleRJqJSH1vWB2/6WeJ\nyAjvfxWRJSIS4zP+cREZD6CqfwKzvXXLVr5MEp5uqloSOBdoCgzzn0Cc/LwOJ+otYBNQAygP9Ab+\nPJEZheAHY3L2rvfdrYD7Ib6fMUJEigKzgPpAZ6A00ArYDjT3mUfG978ybt+/EJ7Q8160fAeziXMA\n7jfpO60AfYCd3l9/c739Vwa4GPgLWCQiDfym+wfwN9BBRE73mX9p4DPcfi8HVAUeBf5W1WXA08Br\nXhyIyI3eNMN95l0FuCabVZ4A3JrNeEdV890LWA9c7PN+JPCZ9//XwBPA996GP8vbEa8BW4DNwONA\nrDd9rLdBtwPrgDsABYr4zO8mn2XdDKwA9gLLcUnqLeCIt7x9wD+BhCzm828vrr3ADKCCz3z7ABuA\nHcC//NfRb/33AUkBxl0ApATaXrgvyWTgbWAP8LAXdzmf6Rt72yMO6Ad85w1/CXjab95TgHu8/+8H\n1vpsm8t9psucTxYx53ZbnQ/8AKTikmU/b/h44PFAy/OWcZb3f3ngE28bLPCW912AaccDo4GpXjzz\ngTN9pu0I/ArsBl4EvsHnO+MXw3DgbZ/39bxlVfTe34Q76JfMxfe/K7Aqm+n7c/Q7uw641WdcBdzB\nJhV3QPsWiAkwn/9523sPsAhok80yxwNjgC+95X4D1PDbvncAq4HfvGHneNPv9Lbn1T7T52Z/FQOe\nwf2edgPfecM2etPt816tcCfCw7xptwJvAmX8vpc3ep+dk8V6Vsf9for4DW/rDe+F+00Xzem34O2H\nyX7DvsIdz34C7vMZ3hRIzWb7xwHJ3jauhPs9n+e3vYZ62z/jd/c4MN5nmiLAAd/9ltUr35+Fe8Xs\nrsDPPoN744pJpXA7fzyQhksYjXE/6pu8aW8GLvWGNwWuzGZZV+F+5H1wZ3jdgR2q2hv3Jeqmrtj7\nVIBZXIf7wZ4GFAXu8+ZbD3dw6YU7MyyDy/qBzANGi8g1IlI9m+kC6YFLFGVxCXYu7ozFN87JqnrY\n73MTgZ4+Zyen4rblJG/8WqCNF/+jwNsnUU8baFvVAD7HnUFVBJJwP4bcGg0cxG3vG7xXdq7BrdOp\nwBrcDxevCmgy8ADuQPYrcF4wAXilhj64g8gub/DFwBequi/IeRQHeuK+E4FsxX3HS+O26X9F5Fxv\n3L1ACm5bVgIexB1AsvIjbnuXA94B3heR+GyW2wt3MK+A20cT/MZfBrQA6olICVyCeAe3z68BXvR+\nG5C7/fU00AS3H8rhTtqO4A7cAGW93+lc3AG7H3AhUAsoCYzym187oC7QKYtlNQTWqWqa3/C+wKfA\ne977btnEm+FD3O8HyPyuX4DbbhM4tkSyCkgXkTdEpIv3W8zk/Xb747b/27gTkx+yWN4e3Pofx1un\nNUCjbKPOLoNE6oU7k9qHO/vZgDvAFvPGfQ085jNtJVxxrZjPsGuB2T6ZeoDPuI4EKEkA04E7s4nJ\n9+wuIYv5DPMZfzvuYADubH6iz7jiwCEClyROBUYAy4B03A+wmTfuAnIuSczxG38T8JX3v+DOFtv6\nn/V44zb6jLs543MB4kwGemR39nQC2+oB4KMA8xlPECUJXOnxMHCOz7j/I/uSxFifcV2Bld7/fXBV\nBxnjMrZfdiWJQ7jvbjouQVzgM/5LYEQuvv+Hgd+Bhrn4/XyM9z0GHsOVBs86gd/hLqBRNvtiks/7\nkt76nuGzfS/yGd8T+NZvHi8Dj+Rmf+FKBn9lFZf/98wbNgu43ef92d6yivhMXyubbdALmOc3rDju\n4HuZz3pM8Rl/3HfTG94ZOOzzfhiQ7P1f1dt+jX3G1/W2cwruJPgToJLfPEd644sH+C10xR1Di+JX\nkvCm+x7ok933ID+XJC5T1bKqWkNVb1fVv3zGbfL5vwau6LXFu1CYittpp3njq/hNvyGbZZ6BO1s+\nUX/4/H8A98M5LgZVPYA7eGRJVXep6v2qWh+XBJOBjzPO8IOwye/9B0Ar76y/Le6s69sslqu4UsO1\n3qDr8Dk7FJE+IpLss50b4M4iT0SgbXWy+wDcWXMRgt/v2cXjv+8U96PMznuqWha375biznoz7MCd\nLefkMm8e8cBA4BvfOmtf3pnmPO/iZiruwJCxX0bizhZniMg6Ebk/0AJF5D7vjpfd3nzKkP3+9d0u\n+3DVSFWyGo/7nbbI+O548+8FnE7u9lcF3DYJ9jtSxW9eG7xlVQoQp79duBoLX5fjDtrTvPcTgC4i\nUjGHWKritlGGPt5nUdXNuCq7vhkjVXWFqvZT1Wq431oV4Dm/eS4D1nvHlOOo6jTc9zXQtYdSuJOR\ngPJzksiOb3F5E64kUcFLKmVVtbR3gAV3ncL3zpDsqm82AWcGsczc2gJUy3gjIsVwVRc5UtXtuOJ1\nFVzRej/uTCZjXrG4H1nAWFV1F67evyfuwD/JO9hlZSJwpVcUboFLMBlF41dxB6zy3gFsKe7MOi9l\ntw98+W8H3wPoNtyPONj9nh3/fSe+77Pj7btbgOE+1XIzgU5e9Usw80hX1Q9xZ5nn+48XdwvjB7jv\nSCVvv0zD2y+quldV71XVWrjq03tEpH0W82mDq7a5GjjVm89ust+/mdtX3N2H5XClnszwff7fBHzj\n8xvNqBK6jdztr+24aqmsviNZfad/xyUo3/mmceyNINn9thcDNf0uavfFnURsFJE/cDcmxOF+W9m5\nHO/kTETOA2oDD4jIH958WgDXZXUBXVVX4koV/he+g/EQrpqxuO9AbzlnAb9k9+FoTRKZ1N2WNgN4\nRtythTEicqaItPMmeQ8YLO620lNxF18DGQvcJyJNvDunzvIOjuC+VCd6n/ZkoJuInOfVUw8nmx+f\niDwpIg1EpIiIlAJuA9ao6g5cXWW8iFwiInG4ImuO9zrj6oL74K7JvBNoIlX9GfdDHAtMV9WMs4wS\nuB/TNi/G/pzYFzYnE4CLReRqb/3Li0hSFtP9AtQXkSSv3ny4zzqk4+pjh4tIca/eu28W8wjGVKCh\niFzm/ajuwJ39BkVVf8VVY/7TG5Rx59oHInKO930tLyIPikhX/89738MeuCrIFVksoihu/28D0kSk\nC65KNePzl3rfY8Ed9NNxJUl/pXAHz21AERF5GHeNIztdxd2mWRRXNz5PVQOdlX8G1BGR3iIS572a\niUjd3OwvVT0CjAOeFZEqIhIrIq28ZLnNWzff3+lE4G4Rqeklsv/D3X3mf40hS6qagiuJNQcQkapA\ne9w1oCTv1Qh4kizucvLiqykiL+Cqih/1RvXFVT3W85lPA9wF+C7ed+NeEanmzecMXAk/u2tTgdbh\na9wJnf82bY4rhWRbyo76JOHpg/uxLMcVDydztEj/Ku5H+gvuDoIPA81EVd/HXbB8B3fHxse4syOA\n/wDDvKLyfbkJTt0ta4NwVTlbcPXNW3EloKwUBz7CFQPX4c6Eunvz2o2rwx+Lu5NrPzlXf4Crz6wN\n/KGq2Z454Nb/YnySiaoux91RMheXMBvi6jPzlKpuxFWX3IsrmieTxYU1VV2Fq2+fibuDw/9BvoG4\ns70/cGdgr59gPNuBq4CncFVF9YCFBN53WRkJ3CIip6nq37htuxJ3kMi4m6cC7q6qDJ+KyD5v/BNA\nX+975B/fXmAw7mRoF+5s9hOfSWrjttE+3L57UVVnZxHjdOAL3EnIBtzZenbVMOC+H4/g9lMT4PpA\nE3pxdsRdsP4dt1+e5OgJTm72133AEtyF9p3efGK8KpcngO+932lLXEJ5C5gD/Oat16Ac1svfy7ib\nZfD+JqvqDFX9I+MFPA8kytFbXFv57L+vcQm3maou8U5qrgZe8J2Hqv7mxdoXd/xpAcwXkf245LAU\n97s4EcM4eizL0At3h1q2JHCtgwkV74wmFajtfTFMlBD3XE4K0CvAwbZQEPdQVoqqHvf8UkHjlVJ+\nBtqr3wN10UpETsNdA2msqgezm7aglCTyPRHp5hWlS+Dqj5fg7mIx+ZyIdBKRst7B4kFcVWGui/0m\nOqnq36par6AkCABV3aqqdXNKEGBJIpx64IrZv+OqAK7J5uKxyV9a4e6m2Y67H/4yv7vtjCmwrLrJ\nGGNMQFaSMMYYE1BUNLzlq0KFCpqQkBDpMIwxJqosWrRou6rm9MDfcaIuSSQkJLBw4cJIh2GMMVFF\nRHJqdSBLVt1kjDEmIEsSxhhjArIkYYwxJqCouyaRlcOHD5OSksLBgzk+F2JMnouPj6datWrExcVF\nOhRj8lyBSBIpKSmUKlWKhIQEJOjWtI05earKjh07SElJoWbNmpEOx5g8F7LqJhEZJyJbRWRpgPEi\nIs+LyBoRWSxHe9LKtYMHD1K+fHlLECbsRITy5ctbKdYUWKG8JjEe1xNTIF1wzVPUxrW5/9LJLMwS\nhIkU++6Zgixk1U2qOkdEErKZpAfwptd+0TyvAbXKBakRLWNMGKSnwY418OdS9/dIeqQjColvVm9j\nY8lEevfKqbv2vBXJaxJVOba9+hRv2HFJQkRuwZU2qF79RDsYC62SJUuyb19QfdsH9PvvvzN48GAm\nT56c5fjU1FTeeecdbr/99qCm99evXz+++eYbypQpg6ry7LPP0r79cZ2URcyYMWMoXrw4ffoc13eL\nMc5fu+CPpS4h/LEU/lwCW1dCum/3HgWzZNcG+KTEVUDhSRJBU9VXgFcAmjZtWmBbJKxSpUq2B/zU\n1FRefPHFzCSR0/RZGTlyJFdeeSWzZ8/mlltuYfXq1ScVM0BaWhpFipz8V2nAgAEnPQ9TQBxJh53r\n4I8l8Oeyo0lhj0//WiUqQqUG0OIWqNQQTm8AFepAbMG8y+zal+cCcFmYlxvJ5yQ2c2yfttW8YQXG\n+vXrueiii0hMTKR9+/Zs3LgRgLVr19KyZUsaNmzIsGHDKFmyZOb0DRq4jq2WLVtG8+bNSUpKIjEx\nkdWrV3P//fezdu1akpKSGDJkyDHTp6enc99999GgQQMSExN54YUXso2tVatWbN58dHMvWrSIdu3a\n0aRJEzp16sSWLa5A9+OPP5KYmJi5zIzljR8/nu7du3PRRRdllkZGjhxJs2bNSExM5JFHHgFg//79\nXHLJJTRq1IgGDRrw7rvvAnD//fdTr149EhMTue8+19Hf8OHDefrppwFITk6mZcuWJCYmcvnll7Nr\n1y4ALrjgAoYOHUrz5s2pU6cO33777cnsIpMfHNwDG+bCglfhk8Hw6kXwn2owqilM7g/f/RdSN0KN\nVnDxo3D9B3DvKhiyBvp8DB0fh0Y9oVL9ApsgIimSJYlPgIEiMgnXTd/uvLge8einy1j++56TDs5X\nvSqleaRb/Vx/btCgQfTt25e+ffsybtw4Bg8ezMcff8ydd97JnXfeybXXXsuYMVn3HjhmzBjuvPNO\nevXqxaFDh0hPT2fEiBEsXbqU5ORkwCWVDK+88grr168nOTmZIkWKsHPnzmxj++KLL7jsMndOcvjw\nYQYNGsSUKVOoWLEi7777Lg899BDjxo2jf//+vPrqq7Rq1Yr77z+2e/CffvqJxYsXU65cOWbMmMHq\n1atZsGABqkr37t2ZM2cO27Zto0qVKkydOhWA3bt3s2PHDj766CNWrlyJiJCamnpcfH369OGFF16g\nXbt2PPzwwzz66KM899xzgCu5LFiwgGnTpvHoo48yc+bM4HaIiawjRyB1/fHVRakbj05T7FRXOmjS\nz/09vQFUPAeKBNONuwmFkCUJEZmI6/i7goik4PrCjQNQ1THANFxfxmuAA0D/UMUSKXPnzuXDD12X\n2r179+af//xn5vCPP/4YgOuuuy7zTNpXq1ateOKJJ0hJSeGKK66gdu3a2S5r5syZDBgwILPap1w5\n/+5snSFDhvDggw+SkpLC3Lmu+Prrr7+ydOlSOnToALhSSeXKlUlNTWXv3r20atUqM9bPPvssc14d\nOnTIXM6MGTOYMWMGjRs3BmDfvn2sXr2aNm3acO+99zJ06FAuvfRS2rRpQ1paGvHx8dx4441ceuml\nXHrppcfEuHv3blJTU2nXrh0Affv25aqrrsocf8UVVwDQpEmTYxKlyYdWToM1M11S+HMZHPKu20kM\nlD8LqjaBc/vC6Q1dUihdBaLkbrF35m9kSnL4Kj+Wb9lDvcqlw7a8DKG8u+naHMYrcEdeL/dEzvjz\no+uuu44WLVowdepUunbtyssvv0ytWrVOer4Z1yReeOEFbrjhBhYtWoSqUr9+/cykkSGrM3xfJUqU\nyPxfVXnggQe49dZbj5vup59+Ytq0aQwbNoz27dvz8MMPs2DBAmbNmsXkyZMZNWoUX331VdDrcMop\n7qwyNjaWtLS0oD9nImDyDRATC6cnQtJ1PqWDulC0eKSjOylTkjeH9cBdr3JpeiRVDcuyfEXFheto\ndd555zFp0iR69+7NhAkTaNOmDQAtW7bkgw8+oGfPnkyaNCnLz65bt45atWoxePBgNm7cyOLFi2nU\nqBF79+7NcvoOHTrw8ssvc+GFF2ZWNwUqTQAMHDiQcePGMX36dC688EK2bdvG3LlzadWqFYcPH2bV\nqlXUr1+fUqVKMX/+fFq0aBEwVoBOnTrxr3/9i169elGyZEk2b95MXFwcaWlplCtXjuuvv56yZcsy\nduxY9u3bx4EDB+jatSutW7c+LvmVKVOGU089lW+//ZY2bdrw1ltvZZYqTJQ5kgYtb4OLH4l0JCFR\nr3Jp3r21VaTDCClLEnnkwIEDVKtWLfP9PffcwwsvvED//v0ZOXIkFStW5PXXXwfgueee4/rrr+eJ\nJ56gc+fOlClT5rj5vffee7z11lvExcVx+umn8+CDD1KuXDlat25NgwYN6NKlC3fccbQgdtNNN7Fq\n1SoSExOJi4vj5ptvZuDAgQHjFRGGDRvGU089RadOnZg8eTKDBw9m9+7dpKWlcdddd1G/fn1ee+01\nbr75ZmJiYmjXrl2WsQJ07NiRFStWZFZNlSxZkrfffps1a9YwZMgQYmJiiIuL46WXXmLv3r306NGD\ngwcPZt6K6++NN95gwIABHDhwgFq1amVuO2NMeEVdH9dNmzZV/06HVqxYQd26dSMUUe4dOHCAYsWK\nISJMmjSJiRMnMmXKlEiHlaV9+/Zl3n01YsQItmzZwv/+978IR5X/RNt3MGweqwDnDSqQJYme3i2p\n0VKSEJFFqto0t5+zkkQELFq0iIEDB6KqlC1blnHjxkU6pICmTp3Kf/7zH9LS0qhRowbjx4+PdEgm\niqSr8snPm5m0dm7OE0eZSF1IDjdLEhHQpk0bfvnll0iHEZSePXvSs2fPSIdhopQqbN//NxTAY2mk\nLiSHmyUJY0xIVShxStRUyZjjWc90xhhjArIkYYwxJiCrbjKmEAn3U8JvE113T5rjWUkij8TGxpKU\nlESDBg3o1q1bjk8rB8u3Eb+8NHz4cKpWrUpSUhJJSUnHtcuUl5KTk5k2bVrA8QsWLKBt27acffbZ\nNG7cmJtuuokDBw4wfvz4bJ/1yK2uXbtm7pfnn3+eunXr0qtXLz755BNGjBiRZ8vJzzKeEg4XQahT\nqVTYlmfynpUk8kixYsUyG97r27cvo0eP5qGHHopwVNm7++67s2w3Kifp6enExsYGPX1ycjILFy6k\na9eux437888/ueqqq5g0aVLmg3iTJ08O+GT5yfBNVC+++CIzZ87MfACye/fuQc8nr5pGj5SwPiX8\nmFC/SgG8takQsZJECPg2w71v3z7at2/PueeeS8OGDTMfmlu/fj1169bl5ptvpn79+nTs2JG//voL\ncM9RNGrUiEaNGjF69OjM+R48eJD+/fvTsGFDGjduzOzZswHXbPdll11Ghw4dSEhIYNSoUTz77LM0\nbtyYli1b5tgirK9Zs2bRuHFjGjZsyA033MDff7vOXBISEhg6dCjnnnsu77//PmvXrqVz5840adKE\nNm3asHLlSgDef/99GjRoQKNGjWjbti2HDh3i4Ycf5t133yUpKSmzqfAMo0ePpm/fvpkJAuDKK6+k\nUqVKx0z36aef0qJFCxo3bszFF1/Mn3/+CcA333yTWRpq3Lgxe/fuZcuWLbRt2zazZJfRnHhCQgLb\nt29nwIABrFu3ji5duvDf//73mBLLtm3b+Mc//kGzZs1o1qwZ33//PeBKXr1796Z169b07t076O1p\nTLSL3tOhQD6/33VUkpdObwhdgquOSE9PZ9asWdx4440AxMfH89FHH1G6dGm2b99Oy5YtM89aV69e\nzcSJE3n11Ve5+uqr+eCDD7j++uvp378/o0aNom3btgwZMiRz3qNHj0ZEWLJkCStXrqRjx46sWrUK\ngKVLl/Lzzz9z8OBBzjrrLJ588kl+/vln7r77bt58803uuuuu42L973//y9tvvw3Ak08+Sbt27ejX\nrx+zZs2iTp069OnTh5deeinzs+XLl+enn34CoH379owZM4batWszf/58br/9dr766isee+wxpk+f\nTtWqVUlNTaVo0aI89thjLFy4kFGjRh0Xw9KlS+nbt2+O2/X8889n3rx5iAhjx47lqaee4plnnuHp\np59m9OjRtG7dmn379hEfH88rr7xCp06deOihh0hPT+fAgQPHzGvMmDF88cUXzJ49mwoVKhzzgOCd\nd97J3Xffzfnnn8/GjRvp1KkTK1asAGD58uV89913FCtWLMd4jSkoCl6SiJC//vqLpKQkNm/eTN26\ndTOb3VZVHnzwQebMmUNMTAybN2/OPAuuWbMmSUlJwNFmr1NTU0lNTaVt27aAa2L8888/B+C7775j\n0KBBAJxzzjnUqFEjM0lceOGFlCpVilKlSlGmTBm6desGQMOGDVm8eHGWMftXN/3yyy/UrFmTOnXq\nAEerzTKSRMZDdfv27eOHH344pvnujBJH69at6devH1dffXVmk955ISUlhZ49e7JlyxYOHTpEzZo1\nM5d3zz330KtXL6644gqqVatGs2bNuOGGGzh8+DCXXXZZ5jYOxsyZM1m+fHnm+z179mR2S9u9e3dL\nEKbQKXhJIsgz/ryWcU3iwIEDdOrUidGjRzN48GAmTJjAtm3bWLRoEXFxcSQkJHDw4EHgaJPX4C58\nZ1Q3nQjfecXExGS+j4mJybPmtDOaBj9y5Ahly5bNvAbja8yYMcyfP5+pU6fSpEkTFi1alO0869ev\nz6JFi+jRo0e20w0aNIh77rmH7t278/XXXzN8+HDA9XB3ySWXMG3aNFq3bs306dNp27Ytc+bMYerU\nqfTr14977rkn6H6zjxw5wrx584iPjw+4/nmpsPRJYKKXXZPIY8WLF+f555/nmWeeIS0tjd27d3Pa\naacRFxfH7Nmz2bBhQ7afL1u2LGXLluW7774DYMKECZnj2rRpk/l+1apVbNy4kbPPPjvPYj/77LNZ\nv349a9asAQjYRHfp0qWpWbMm77//PuBKSxnNjKxdu5YWLVrw2GOPUbFiRTZt2kSpUqUCXogeOHAg\nb7zxBvPnz88c9uGHH2aWtjLs3r2bqlVdEwhvvPFG5vC1a9fSsGFDhg4dSrNmzVi5ciUbNmygUqVK\n3Hzzzdx0002ZVWTB6Nix4zFdv2aVCPNSuO82KixNSZi8U/BKEvlA48aNSUxMZOLEifTq1Ytu3brR\nsGFDmjZtyjnnnJPj519//XVuuOEGRISOHTtmDr/99tu57bbbaNiwIUWKFGH8+PHHlCBOVnx8PK+/\n/jpXXXUVaWlpNGvWjAEDBmQ57YQJE7jtttt4/PHHOXz4MNdccw2NGjViyJAhrF69GlWlffv2NGrU\niOrVqzNixAiSkpJ44IEHjmkLqlKlSkyaNIn77ruPrVu3EhMTQ9u2bencufMxyxs+fDhXXXUVp556\nKhdddBG//fYb4Jpdnz17NjExMdSvX58uXbowadIkRo4cSVxcHCVLluTNN98Mehs8//zz3HHHHSQm\nJpKWlkbbtm0DdjGbVwpDnwQmellT4cbkgRP9DkZbc9O5VoCbCo82J9pUuFU3GWOMCciShDHGmIAK\nTJKItmozU3DYd88UZAUiScTHx7Njxw77sZqwU1V27NiR5S2zxhQEBeLupmrVqpGSksK2bdsiHYop\nhOLj4zPbgDKmoCkQSSIuLi7zCVxjjDF5p0BUNxljjAkNSxLGGGMCsiRhjDEmIEsSxhhjArIkYYwx\nJqACcXeTMSafOPwXpPwI67+HDd/DkcMQE3xXtyb/sSRhjDlxf++FTfNhww8uMWxe5BID4np0bHEb\nnJtzz4Mm/7IkYYwJ3l+7YOM8V0pY/z1s+QU0HSQWqiRBy9sg4Xw4owUUKxvpaE0eCGmSEJHOwP+A\nWGCsqo7wG18GeBuo7sXytKq+HsqYjDG5sH+7SwgZJYU/lwIKsUWhahM4/26ocZ5LCqeUjHS0JgRC\nliREJBYYDXQAUoAfReQTVV3uM9kdwHJV7SYiFYFfRWSCqh4KVVzGmGzs2eIlBa+ksP1XN7xIMTij\nGVzwgEsspCplAAAe70lEQVQK1ZpCnPX3XRiEsiTRHFijqusARGQS0APwTRIKlBIRAUoCO4G86ZDZ\nGJOzXRtcKWHDd+7vznVueNFSUL0FNLoGarSGKo2hSNHIxmoiIpRJoiqwyed9CtDCb5pRwCfA70Ap\noKeqHvGfkYjcAtwCUL169ZAEa0yBpwo71h4tKWz4AXZ7P9H4sq6E0PRG9/f0RIi1S5Ym8heuOwHJ\nwEXAmcCXIvKtqh7TM7yqvgK8Aq770rBHaUw0OnIEtq08Nins+9ONK1HRJYPzBru/p9WDGHtsyhwv\nlEliM3CGz/tq3jBf/YER6jqCWCMivwHnAAtCGJcxBdORdPhjiVd95CWFv3a6caWqQM22LiHUOB8q\n1AaRyMZrokIok8SPQG0RqYlLDtcA1/lNsxFoD3wrIpWAs4F1IYzJmIIj/bC7BXX9dy4pbJwHf3uF\n8LI14OwuXlJoDacmWFIwJyRkSUJV00RkIDAddwvsOFVdJiIDvPFjgH8D40VkCSDAUFXdHqqYjIlq\nhw/C7z8dfZp50wI4vN+NK18bGlzhEkKN86CMdYJk8kZIr0mo6jRgmt+wMT7//w50DGUMxkStQ/t9\nmrj4wf2f/rcbd1p9aNzraEmh5GmRjdUUWJG+cG2MyXBwj2viYr13O+rvP8GRNJAYd7dR85tdUqje\nCoqXi3S0ppCwJGFMpBzYSdODc6l7aAm8fD/8sRj0CMQUgSrnwnmDXCnhjBYQXzrS0ZpCypKEMeGy\nb+uxTVxsXcYQ4BBxUKEFtB3iPc3cDIqWiHS0xgCWJIwJnd2bj23iYsdqNzyuuCsd1L+cRxaXZW1c\nHd7u3y6ysRoTgCUJY/KCKuxa71NS+A5SN7hxp5R21xEaX+9aSK3cCGLjAFi5cm7kYjYmCJYkjDkR\nqrB99bFPM+/xnhUtVs5VG7UYAAmtoVID63jHRC1LEsYE48gR2Lr82Mbw9m9z40pWOvp8QsL5UOFs\na+LCFBiWJIzJSnqau9vIt4mLg6luXJkz4Mz2R5NCuVr2NLMpsCxJGOPr0AF+G38TlbbMprgeAGBL\nbBVWFG3OijKJrCjagG1FTodtuNfCrcDWE17c8i17qFfZbm81+VdQSUJEigLVVXVNiOMxJnLS02By\nf2r8Pp0PuYjNZZuxomgDdsVWCNki61UuTY+kqiGbvzEnK8ckISKXAM8CRYGaIpIEPKKql4c6OGPC\nRhU+vRNWfcG40gP5ssSlvHtrq0hHZUzEBXN17TFcZ0GpAKqaDJwVyqCMCbuZwyH5bWh3P1+WuDTS\n0RiTbwSTJA6raqrfMOv4xxQcc0fD989B0xvggvsjHY0x+Uow1yRWiMjVQIzXN8RgYF5owzImTBa/\nB9MfhHo9oOvTdpeSMX6CKUkMBJoAR4APgb+BO0MZlDFhsXomfHwbJLSBK161B96MyUIwJYlOqjoU\nGJoxQESuwCUMY6JTykJ4rzecVheueQeKnBLpiIzJl4IpSQzLYthDeR2IMWGzbRVMuMp11NPrA2uG\n25hsBCxJiEgnoDNQVUSe9RlVGlf1ZEz02b0Z3rrc9dnQ+yMoVSnSERmTr2VX3bQVWAocBJb5DN8L\n2C0gJvoc2AlvXwEHd0P/qa45DWNMtgImCVX9GfhZRCao6sEwxmRM3jt0ACZeAzvXwfUfuOa6jTE5\nCubCdVUReQKoB8RnDFTVOiGLypi8lH4Y3u8HmxbA1W9AzbaRjsiYqBHMhevxwOuAAF2A94B3QxiT\nMXlHFT4ZDKunw6XPuuchjDFBCyZJFFfV6QCqulZVh+GShTH535cPwy/vwAUPuieqjTG5Ekx1098i\nEgOsFZEBwGagVGjDMuaod+ZvZEry5lx/7tJ9H9B776t8Ubwbr69sA78G11WoNd9tzFHBlCTuBkrg\nmuNoDdwE2CmZCZspyZtZvmVPrj7T5sBMeu99lbnxbRhfekCumtuw5ruNOSrbkoSIxAKXq+p83K2v\nvcMSlTF+6lUuHXzT3au/hInPQc22tOo1mVb2NLUxJyzbkoSqpgMXhikWY07eph/hvT5QqT70nGDN\nbRhzkoK5JrFIRD4E3gf2ZwxU1U9CFpUxJ2Lbr/DOVVCyEvSabM1tGJMHgkkSpXDJoavPMAUsSeQj\nJ3pxNxoEdSF5d4prbiO2qGtuo+Rp4QnOmAIuxyShqnYdIgpkXNwtiHfl5Hgh+cBOeOsK+Hsv9J8G\n5WqGLzhjCrhgShImSuTq4m5BcWg/vHM17FoPvT+E0xtGOiJjCpRgboE9YSLSWUR+FZE1IpJlo4Ai\ncoGIJIvIMhH5JpTxmAIm/TC81xc2L4IrX4OE8yMdkTEFTshKEt7ts6OBDkAK8KOIfKKqy32mKQu8\nCHRW1Y0iYhXJJjhHjsCUgbDmS+j2P6jbLdIRGVMg5ViSEJGKIvKyiHzmva8nIv2CmHdzYI2qrlPV\nQ8AkwL/hnOuAD1V1I4Cqbs1V9Kbw+vJfsHgSXDgMmvSLdDTGFFjBNvD3DXCG9341cG8Qn6sKbPJ5\nn+IN81UHOFVEvhaRRSLSJ6sZicgtIrJQRBZu27YtiEWbAu37/8HcUdD8Fmh7X6SjMaZACyZJnKaq\n7+D1Rqeqh8m7numKAE2AS4BOwL9E5LgmyFX1FVVtqqpNK1asmEeLNlEp+R3XaF/9K6Dzk7lqbsMY\nk3vBXJPYLyLlcM9GICLNgGAa0tnM0dIHQDVvmK8UYIeq7veWMwdoBKwKYv6msFk13V2HqHUBXD4G\nYkJ634UxhuBKEkOAT4Fa3t1HE4FBQXzuR6C2iNQUkaLANRz/AN4U4HwRKSIixYEWwIqgozeFx8b5\n7k6m0xtCz7etuQ1jwiSYh+l+FJELgbq4joeWexeic/pcmogMBKYDscA4VV3mNTeOqo5R1RUi8gWw\nGFeFNVZVl57E+piCaOtK9yxE6SquuY1TrKV6Y8IlxyQhIj/hSg/vq+r63MxcVacB0/yGjfF7PxIY\nmZv5mkIk/TBM7u9KDr0/hJJ2TcqYcAqmuukqIA6YIiJzReQuEakS4riMceaPga3L4dL/wqkJkY7G\nmEInxyThdVn6f6raCNfZ0LnAxpBHZsye3+HrEVC7E5zdNefpjTF5LqgnrkWkGnA10NP7zEOhDMoY\nAKY/CEfSoIvd6mpMpARzTeIHoCSuP4nrVXV1yKMyZs0sWPYRXPCgtepqTAQFU5K4WVWXhTwSYzKk\n/Q3ThkC5WtD6zkhHY0yhFjBJiMi1qjoRaC8i7f3Hq+rzIY3MFF4/PA8718L1H0BcfKSjMaZQy64k\ncar3N6t7DjUEsRjj+oWY8zTU7Q5nXRzpaIwp9AImCVV90ft3qqrO8x0nIi1DGpUpvD6/HyQWOv8n\n0pEYYwjuOYkXsxg2Oq8DMYZfP4dVn8MFQ6FMtUhHY4wh+2sSzYFWQEURGewzqjTu4Tpj8s6hA/D5\nP6HiOdDy9khHY4zxZHdNogRQwZvG97rEXtxT2Mbkne+ehdSN0PcziLVzEGPyi+yuScwGZovI66q6\nLowxmcJm+xrXkVBiT6jZJtLRGGN8ZFfd9Iyq3gs8IyLH3c2kqleENDJTOKjCtPugSDx0+HekozHG\n+Mmuuuld7++ocARiCqnlH8O62dDlKShVKdLRGGP8ZFfdtMD7OytjmIiUAaqq6vIwxGYKur/3whcP\nwumJ0PTGSEdjjMlCjrfAisgsESktIqcCycBbImL9P5iT9/UI2Ps7XPIsxAbV1qQxJsyCeU6inKru\nAa4A3lbVJkCn0IZlCrw/l8O8l+DcPnBGs0hHY4wJIJgkUUREKuJue/00xPGYwiDjYnV8aWg/PNLR\nGGOyEUySeAL4BtikqgtEpBbwW2jDMgXa4ndhw/dw8aNQonykozHGZCPHimBVnQRM8nm/DugRyqAK\ngnfmb2RK8uawLW/5lj3Uq1w6bMs7YX+lwoxhUK0ZNO4d6WiMMTkI5sJ1FRF5T0S2eK93rY/rnE1J\n3szyLXvCtrx6lUvTI6lq2JZ3wr56HA7sgEuegZhgCrLGmEgK5paS14HJQMZpX29vmF28zkG9yqV5\n99ZWkQ4j//j9Z1j4GjS7GSo3inQ0xpggBHMqV0lVX1XVv73XWMCeejK5c+QITL0XileAi6yLdGOi\nRTBJYqeIXCNH9QR2hjowU8D89AZsXgQdH4f4MpGOxhgTpGCSxA1AH2C79+rtDTMmOPt3wKxHocb5\nkHh1pKMxxuRCMHc3rQe6hj4UU2DNfMQ1wXHJ0yAS6WiMMbkQzN1NCSLykYj84b0+EJGE0IdmCoSN\n8+Hnt1xHQqfVjXQ0xphcCqa6aSLwCVDde33qDTMme+lp7mJ16arQbmikozHGnIBgkkQJVX1dVQ95\nr/FA8RDHZQqCH8fCn0ug83/glJKRjsYYcwKCeU5imojch3vqWoGewFQRKQ3gNf5nzLH2/gGzn4Az\n20Pd7pGOxhhzgoJJEr28v3f6De+NSxrV8zQiUzDMGAZpB6HrSLtYbUwUC+bupjPCEYgpQH6bA0ve\nh7b/hPJnRjoaY8xJyFXjOSLyYi6n7ywiv4rIGhG5P5vpmolImohcmZv5m3wo7RBMvQ/K1oA290Q6\nGmPMScptC2stg51QRGKB0UAXoB5wrYjUCzDdk8CMXMZi8qN5L8L2X101U1yxSEdjjDlJuU0SO3Ix\nbXNgjaquU9VDuAvfWTUxPgj4ANiay1hMfpO6Cb55Es6+BOpY+4/GFARBJwkROUVVO+Ri3lWBTT7v\nU7xhvvOsClwOvJTDsm8RkYUisnDbtm25CMGE1fQHXK9zXUZEOhJjTB4J5onr5iKyBFjtvW8kIi/k\n0fKfA4aq6pHsJlLVV1S1qao2rVixYh4t2uSp1TNhxafQbgiUtRvejCkogrkF9nngUuBjAFX9RUQu\nDOJzmwHfO6OqecN8NQUmibtFsgLQVUTSVPXjIOZv8ovDB12f1eVrQ6tBkY7GGJOHgkkSMaq6QY69\n1z09iM/9CNQWkZq45HANcJ3vBKpaM+N/ERkPfGYJIgp9/xzs+g36TIEiRSMdjTEmDwWTJDaJSHNA\nvTuRBgGrcvqQqqaJyEBgOhALjFPVZSIywBs/5iTiNvnFznXw7bPQ4B9Q64JIR2OMyWPBJInbcFVO\n1YE/gZnesByp6jRgmt+wLJODqvYLZp4mH1GFz4dCbFHo+ESkozHGhEAwT1xvxVUVGXOslZ/B6hnQ\n6f+gdOVIR2OMCYEck4SIvIpro+kYqnpLSCIy0eHQfvj8fqjUAJrfGulojDEhEkx100yf/+NxzzVs\nCjCtKSzmjIQ9KXDlaxAbzNfIGBONgqluetf3vYi8BXwXsohC5J35G5mS7H8Hbugs37KHepVLh215\nYaMKyz+GH0ZBUi+oHnRLLcaYKJTbZjkAagKV8jqQUJuSvJnlW8LX9UW9yqXpkVQ15wmjycb58FpH\neL8fVKgDHR6LdETGmBAL5prELo5ek4gBdgIBW3TNz+pVLs27t7aKdBjRZ8damPmIe6K65OnQ7XlX\nirBqJmMKvGx/5eKeoGvE0Selj6jqcRexTQG1f7trsG/hOCgSDxc+BK3ugKIlIh2ZMSZMsk0Sqqoi\nMk1VG4QrIJMPHDrgmvz+7jk4fACa9IML7oeSp0U6MmNMmAVTX5AsIo1V9eeQR2Mi60g6/DIRvnoC\n9v7umvy+eDhUrBPpyIwxERIwSYhIEVVNAxoDP4rIWmA/ILhCxrlhitGEmiqsmQVfPgxbl0HVJvCP\nsZDQOtKRGWMiLLuSxALgXKB7mGIxkbBlMXz5L1j3NZyaAFe+DvUvh2MbdDTGFFLZJQkBUNW1YYrF\nhFPqJvjqcVj8LhQrC51HQNMboMgpkY7MGJOPZJckKopIwJ7sVfXZEMRjQu3gbtdq6zyvM8DWg+H8\ne1yiMMYYP9kliVigJF6JwkS5tEOw8DX45in4axck9oSLhkHZM3L+rDGm0MouSWxRVXukNtplNKMx\n81HXMVDNdtDx31C5UaQjM8ZEgRyvSZgotmEuzBgGmxfCafWg1wdwVnu7KG2MCVp2SaJ92KIweWv7\navjyEfh1KpSqDN1HQdJ1EBMb6ciMMVEmYJJQ1Z3hDMTkgX1b4esRsGg8xBV31xxa3gFFi0c6MmNM\nlLIW2gqCQ/th7mj4/n+QdtDdytpuKJSsGOnIjDFRzpJENDuSDskTYPb/wd4tULcbtB8OFc6KdGTG\nmALCkkQ0UoXVX7pmNLatgGrN4arx1gGQMSbPWZKINr8nu2Y0fpsD5WrB1W9C3e52x5IxJiQsSUSL\n1I0w69+w5D0oXh66PAVN+kORopGOzBhTgFmSyO/+2gXfPgPzXwaJcU1onH8XxJeJdGTGmELAkkR+\nlfY3/DjWNaNxcDc0uhYuegjKVIt0ZMaYQsSSRH6jCks/gFmPQeoGOPMi6PAYnN4w0pEZYwohSxL5\nyfrvXTMav/8ElRrA9R+6ZjSMMSZCLEnkB9t+dc1orPocSleFy15yrbRaMxrGmAizJBFJe/+Er/8D\nP70JRUtA+0eg5W0QVyzSkRljDGBJIjL+3gdzR8H3z0P639D8Zmg7BEpUiHRkxhhzDEsS4ZSeBj+/\n5UoP+/6Eej1c6aH8mZGOzBhjshQTypmLSGcR+VVE1ojI/VmM7yUii0VkiYj8ICIFsyccVfj1CxjT\nGj67C05NgBu/dE9LW4IwxuRjIStJiEgsMBroAKQAP4rIJ6q63Gey34B2qrpLRLoArwAtQhVTRGz+\nybWxtP5bKHcm9HwbzrnUmtEwxkSFUFY3NQfWqOo6ABGZBPQAMpOEqv7gM/08oOA8KbZrvWtGY+lk\nKF4Buj4NTfpBbFykIzPGmKCFMklUBTb5vE8h+1LCjcDnWY0QkVuAWwCqV6+eV/GFxoGdrhmNBa+A\nxEKb+6D1nRBfOtKRGWNMruWLC9ciciEuSZyf1XhVfQVXFUXTpk01jKEF7/BBlxi+fRoO7oHGveDC\nh6B0lUhHZowxJyyUSWIzcIbP+2resGOISCIwFuiiqjtCGE9oHDlytBmN3RvhrA7Q4VGoVD/SkRlj\nzEkLZZL4EagtIjVxyeEa4DrfCUSkOvAh0FtVV4UwltD4bQ7M+BdsSYbTE6HHC1DrgkhHZYwxeSZk\nSUJV00RkIDAdiAXGqeoyERngjR8DPAyUB14Ud7dPmqo2DVVMeWbrCteMxurpUOYMuPwVaHgVxIT0\njmJjjAm7kF6TUNVpwDS/YWN8/r8JuCmUMeSpvX/A7Cfg57ehaCm4+FFoMQDi4iMdmTHGhES+uHCd\n7/29F354wb3SD7vE0HYIFC8X6ciMMSakLElkJz0NfnoDvh4B+7dC/Sug/b9c39LGGFMIWJLIiir8\nOs1dd9ixGqqfB9dOhGr5/3KJMcbkJUsS/lIWuY5/Nv4AFerANRPh7C7WjIYxplCyJJFh52/uWYdl\nH0KJinDJs3BuX4i1TWSMKbzsCHhgJ8wZCQtede0qtRsK5w2CU0pFOjJjjIm4wpskDh+E+WPg22fh\n0F5ofD1c8CCUrhzpyIwxJt8ofEniyBFY8h589Tjs3gS1O7lmNE6rG+nIjDEm3ylUSaLB3z/DK/+E\nPxZD5UZw2YtQs22kwzLGmHyr0CSJCw7M4Lbdz0KZ6nDFWGjwD2tGwxhjclBoksT8+NYU1/30HfiY\nNaNhjDFBKjSn0n/FlGBaicstQRhjTC4UmiRhjDEm9yxJGGOMCciShDHGmIAsSRhjjAnIkoQxxpiA\nLEkYY4wJyJKEMcaYgCxJGGOMCciShDHGmIAsSRhjjAnIkoQxxpiALEkYY4wJyJKEMcaYgCxJGGOM\nCciShDHGmIAsSRhjjAnIkoQxxpiALEkYY4wJqND0cV2vSulIh2CMMVGn0CSJR7rVj3QIxhgTdUJa\n3SQinUXkVxFZIyL3ZzFeROR5b/xiETk3lPEYY4zJnZAlCRGJBUYDXYB6wLUiUs9vsi5Abe91C/BS\nqOIxxhiTe6EsSTQH1qjqOlU9BEwCevhN0wN4U515QFkRqRzCmIwxxuRCKJNEVWCTz/sUb1hup0FE\nbhGRhSKycNu2bXkeqDHGmKxFxS2wqvqKqjZV1aYVK1aMdDjGGFNohDJJbAbO8HlfzRuW22mMMcZE\nSCiTxI9AbRGpKSJFgWuAT/ym+QTo493l1BLYrapbQhiTMcaYXAjZcxKqmiYiA4HpQCwwTlWXicgA\nb/wYYBrQFVgDHAD6hyoeY4wxuSeqGukYckVEtgEbTvDjFYDteRhONLB1LhxsnQuHk1nnGqqa64u6\nUZckToaILFTVppGOI5xsnQsHW+fCIRLrHBV3NxljjIkMSxLGGGMCKmxJ4pVIBxABts6Fg61z4RD2\ndS5U1ySMMcbkTmErSRhjjMkFSxLGGGMCKpBJojD2YxHEOvfy1nWJiPwgIo0iEWdeymmdfaZrJiJp\nInJlOOMLhWDWWUQuEJFkEVkmIt+EO8a8FsR3u4yIfCoiv3jrHNUP5YrIOBHZKiJLA4wP7/FLVQvU\nC/d091qgFlAU+AWo5zdNV+BzQICWwPxIxx2GdT4PONX7v0thWGef6b7CPd1/ZaTjDsN+LgssB6p7\n70+LdNxhWOcHgSe9/ysCO4GikY79JNa5LXAusDTA+LAevwpiSaIw9mOR4zqr6g+qust7Ow/XmGI0\nC2Y/AwwCPgC2hjO4EAlmna8DPlTVjQCqGu3rHcw6K1BKRAQoiUsSaeENM++o6hzcOgQS1uNXQUwS\nedaPRRTJ7frciDsTiWY5rrOIVAUup+D0eBjMfq4DnCoiX4vIIhHpE7boQiOYdR4F1AV+B5YAd6rq\nkfCEFxFhPX6FrIE/kz+JyIW4JHF+pGMJg+eAoap6xJ1kFgpFgCZAe6AYMFdE5qnqqsiGFVKdgGTg\nIuBM4EsR+VZV90Q2rIKhICaJwtiPRVDrIyKJwFigi6ruCFNsoRLMOjcFJnkJogLQVUTSVPXj8ISY\n54JZ5xRgh6ruB/aLyBygERCtSSKYde4PjFBXYb9GRH4DzgEWhCfEsAvr8asgVjcVxn4sclxnEakO\nfAj0LiBnlTmus6rWVNUEVU0AJgO3R3GCgOC+21OA80WkiIgUB1oAK8IcZ14KZp034kpOiEgl4Gxg\nXVijDK+wHr8KXElCC2E/FkGu88NAeeBF78w6TaO4Bc0g17lACWadVXWFiHwBLAaOAGNVNctbKaNB\nkPv538B4EVmCu+NnqKpGbRPiIjIRuACoICIpwCNAHETm+GXNchhjjAmoIFY3GWOMySOWJIwxxgRk\nScIYY0xAliSMMcYEZEnCGGNMQJYkTL4kIuleS6YZr4Rspk0I1GJmJIjIWBGp5/3/oN+4H8Icy13e\n8xLGnBC7BdbkSyKyT1VLBjltAvCZqjYIaVAnIDfrcYLzF9zvOMu2ikRkPdA0mp8bMJFlJQkTNbwS\nw7ci8pP3Oi+LaeqLyAKv9LFYRGp7w6/3Gf6yiMRm8dl+IjLFaxxvtYg84jPuHhFZ6r3u8oaVEJGp\nXj8GS0Wkpzf8axFpKiIjgGLeMid44/Z5fyeJyCU+8x8vIleKSKyIjBSRH734bw2wHX4VkTeBpcAZ\nIvKSiCwU15/Co950g4EqwGwRme0N6ygic73t976IhCyBmQIi0m2n28teWb2AdFyjbcnAR96w4kC8\n939tYKH3fwJe2/vAC0Av7/+iuEbu6gKfAnHe8BeBPlkssx+wBfdkejHcAbgprsG8JUAJXFPUy4DG\nwD+AV30+X8b7+zXu7B1gn98y9nl/Lwfe8Ilzk7fMW4Bh3vBTgIVATb95JOCepm7pM6yc9zfWW36i\n9349UMH7vwIwByjhvR8KPBzpfW2v/P0qcM1ymALjL1VN8hsWB4wSkSRcEqmTxefmAg+JSDVcvwqr\nRaQ97kD/o9ckSTEC9y/xpXqNH4rIh7jWchWXqPb7DG8DfAE8IyJP4qq7vs3F+n0O/E9ETgE6A3NU\n9S8R6QgkytFe9MrgEuJvfp/foK4vgQxXi8gtuKZ2KgP1cE1z+GrpDf/e2w5FcdvLmIAsSZhocjfw\nJ65V0xjgoP8EqvqOiMwHLgGmedU1gjtrf8B3WhG5HNcuDsBNGbPwn2WgYFR1lbiuI7sCj4vILFV9\nLJgVUdWDIvI1rpnrnrjOdPBiHaSq03OYxX6f9agJ3Ac0U9VdIjIeiM/iM4JLgtcGE6MxYNckTHQp\nA2xRd5G2N65q5RgiUgtYp6rP41pETQRmAVeKyGneNOVEpIaqfqSqSd5roTeLDt74YsBlwPfAt8Bl\nIlJcRErgqoq+FZEqwAFVfRsYiety0t9hEYkLsD7v4hpnyyiVgGvI7raMz4hIHW+Z2SmNSxq7xbWC\n2sVn3F6glPf/PKC1iJzlzbuEiGRVGjMmk5UkTDR5EfhAXG9rX+BzNu3jaqC3iBwG/gD+T1V3isgw\nYIaIxACHgTuADVl8fgGuu9NqwNsZycM7O8/on2Csqv4sIp2AkSJyxJvnbVnM7xVgsYj8pKq9/MbN\nAN4CpqjrmhNcfx8JwE/enUvbcMkqIFX9RUR+Blbirm1877f8L0Tkd1W9UET6ARO9ai6AYURvXxMm\nDOwWWGM83gG0qaoOjHQsxuQXVt1kjDEmICtJGGOMCchKEsYYYwKyJGGMMSYgSxLGGGMCsiRhjDEm\nIEsSxhhjAvp/4g4Q3EeP2x4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113d00898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "title ='Predicting Survival including RCB as a predictor (ADASYN)'\n",
    "predictive_statistics.plot_compare_roc(fpr3, tpr3,fpr4, tpr4, auc3, auc4, title = title)\n",
    "plt.legend(['Logistic Regression','Random Forest Classifier']);\n",
    "\n",
    "#plt.savefig('LG_vs_RFC_Alive_RCB.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1><center> Prediction of continous outcomes</center></h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# metrics\n",
    "mae = metrics.median_absolute_error\n",
    "\n",
    "def mae_report(Ytest, Yhat, outcome_):\n",
    "    error = mae(Ytest, Yhat)\n",
    "    error = np.round( error, decimals=3)\n",
    "    # report\n",
    "    print('\\n' )\n",
    "    print('==' *40)\n",
    "    print('The median absolute error for testing data set of ' + outcome_ + ' is: ' + str(error))\n",
    "    print('==' *40)\n",
    "\n",
    "def train_test_report(predictor, Xtrain, Ytrain, Xtest, Ytest, outcome):\n",
    "    # train\n",
    "    predictor.fit(Xtrain, Ytrain)\n",
    "    # test\n",
    "    Yhat = predictor.predict(Xtest)\n",
    "    # report\n",
    "    mae_report(Ytest, Yhat, outcome)\n",
    "    \n",
    "# lsq \n",
    "import statsmodels.api as sm\n",
    "def lsq(Xtrain,Ytrain, Xtest, Ytest, outcome =''):\n",
    "    # train\n",
    "    OLS = sm.OLS(Ytrain,Xtrain).fit();\n",
    "    print(OLS.summary())\n",
    "    #test\n",
    "    Yhat = OLS.predict(Xtest)\n",
    "    # report\n",
    "    mae_report(Ytest, Yhat, outcome)\n",
    "\n",
    "# SVR    \n",
    "from sklearn.svm import SVR\n",
    "from sklearn.model_selection import GridSearchCV   \n",
    "\n",
    "# GridSearchCV utility\n",
    "def gridsearch(regressor, grid):\n",
    "    optimized_regressor=  GridSearchCV(  regressor, \n",
    "                               param_grid = grid, \n",
    "                               cv= 3, verbose = 0, n_jobs = -1,\n",
    "                               scoring = metrics.make_scorer(metrics.median_absolute_error))\n",
    "    \n",
    "    return optimized_regressor\n",
    "    \n",
    "\n",
    "\n",
    "def svr(Xtrain,Ytrain, Xtest, Ytest, outcome = ''):\n",
    "    # define regressor\n",
    "    regressor =  SVR()\n",
    "    # define parameter grid search\n",
    "    grid = dict(       kernel = ['rbf','linear','sigmoid'], \n",
    "                       C = np.arange(1,11,1),\n",
    "                       epsilon = np.arange(1,11,1),\n",
    "                       gamma = np.linspace(1/10,10,3))\n",
    "    # perform grid search\n",
    "    grid_search=  gridsearch(regressor, grid)\n",
    "    \n",
    "    # train, test, and report\n",
    "    train_test_report(grid_search, Xtrain, Ytrain, Xtest, Ytest, outcome)\n",
    "\n",
    "    \n",
    "# ElasticNet\n",
    "from sklearn.linear_model import ElasticNet as ENet\n",
    "\n",
    "def ElasticNet(Xtrain,Ytrain, Xtest, Ytest, outcome = ''):\n",
    "    # define regressor\n",
    "    regressor =  ENet(max_iter=5000)\n",
    "    # define parameter grid search\n",
    "    grid = dict(   alpha = np.arange(1,20,.5), l1_ratio = np.arange(.1,1,.05))\n",
    "    # perform grid search\n",
    "    grid_search=  gridsearch(regressor, grid)\n",
    "    # train, test, and report\n",
    "    train_test_report(grid_search, Xtrain, Ytrain, Xtest, Ytest, outcome)\n",
    "    \n",
    "\n",
    "# Random Forest Regressor\n",
    "from sklearn.ensemble import RandomForestRegressor as RFR\n",
    "\n",
    "def RandomForestRegressor(Xtrain,Ytrain, Xtest, Ytest, outcome = ''):\n",
    "    # define regressor\n",
    "    regressor =  RFR( criterion='mse', random_state = RANDOM_STATE)\n",
    "    \n",
    "    #\n",
    "    num_features = Xtrain.shape[1]\n",
    "    \n",
    "    # define parameter grid search\n",
    "    grid = dict(    n_estimators = np.arange(5,100,5), \n",
    "                    max_features = np.arange(1,num_features, 1),\n",
    "                    max_depth = [None, 1, 2, 3, 4, 5])\n",
    "    \n",
    "    # perform grid search\n",
    "    grid_search=  gridsearch(regressor, grid)\n",
    "    \n",
    "    # train, test, and report\n",
    "    train_test_report(grid_search, Xtrain, Ytrain, Xtest, Ytest, outcome)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Organize predictors in the right format and split data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "\n",
    "# allocate continous predictors\n",
    "cont_predictors = ['age','MRI_LD_Baseline', 'MRI_LD_1_3dAC', 'MRI_LD_Int_Reg', 'MRI_LD_PreSurg']\n",
    "contX = df[cont_predictors].values\n",
    "contX = StandardScaler().fit_transform(contX)\n",
    "\n",
    "\n",
    "# allocate categorical predictors\n",
    "cat_pred = ['PCR','White', 'ER+', 'PR+', 'HR+'];\n",
    "catX = pd.pandas.get_dummies(df[cat_pred], drop_first=True).values\n",
    "\n",
    "# concatenate predictors\n",
    "X = np.concatenate( (catX, contX), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## - Recurrence-Free Survival (`RFS`, Continous in months)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "#outcome\n",
    "y = df.RFS.values / 30; # conver to months\n",
    "\n",
    "#split\n",
    "X_train, X_test, y_train, y_test = predictive_statistics.split_data(X, y, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.773\n",
      "Model:                            OLS   Adj. R-squared:                  0.751\n",
      "Method:                 Least Squares   F-statistic:                     36.38\n",
      "Date:                Tue, 20 Jun 2017   Prob (F-statistic):           5.60e-30\n",
      "Time:                        09:37:26   Log-Likelihood:                -535.00\n",
      "No. Observations:                 117   AIC:                             1090.\n",
      "Df Residuals:                     107   BIC:                             1118.\n",
      "Df Model:                          10                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "x1            27.0400      5.212      5.188      0.000      16.708      37.372\n",
      "x2            17.8831      4.243      4.215      0.000       9.473      26.294\n",
      "x3            15.3054     15.097      1.014      0.313     -14.623      45.234\n",
      "x4             7.5932      7.294      1.041      0.300      -6.865      22.052\n",
      "x5             9.6444     16.267      0.593      0.555     -22.603      41.891\n",
      "x6             2.4913      2.422      1.029      0.306      -2.309       7.292\n",
      "x7             9.0136      5.044      1.787      0.077      -0.985      19.012\n",
      "x8            -4.0383      5.040     -0.801      0.425     -14.029       5.953\n",
      "x9            -6.3210      3.774     -1.675      0.097     -13.802       1.160\n",
      "x10            2.0339      3.237      0.628      0.531      -4.384       8.452\n",
      "==============================================================================\n",
      "Omnibus:                       11.597   Durbin-Watson:                   1.873\n",
      "Prob(Omnibus):                  0.003   Jarque-Bera (JB):               12.486\n",
      "Skew:                           0.655   Prob(JB):                      0.00194\n",
      "Kurtosis:                       3.918   Cond. No.                         15.7\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "================================================================================\n",
      "The median absolute error for testing data set of  RFS (months) is: 14.364\n",
      "================================================================================\n"
     ]
    }
   ],
   "source": [
    "# LSQ\n",
    "predictive_statistics.lsq(X_train, y_train, X_test, y_test, outcome =' RFS (months)')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ElasticNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "importlib.reload(predictive_statistics);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "================================================================================\n",
      "The median absolute error for testing data set of  RFS (months) is: 12.536\n",
      "================================================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEKCAYAAAAB0GKPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x116369390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "EN = predictive_statistics.ElasticNet(X_train, y_train, X_test, y_test, outcome =' RFS (months)');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SVM Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "================================================================================\n",
      "The median absolute error for testing data set of  RFS(months) is: 82.589\n",
      "================================================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEKCAYAAADTgGjXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x116013048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "SVR = predictive_statistics.svr(X_train, y_train, X_test, y_test, outcome =' RFS(months)');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "================================================================================\n",
      "The median absolute error for testing data set of  RFS(months) is: 11.76\n",
      "================================================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEKCAYAAAAfGVI8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsvXl8nPV17/8+s2lGu2Rbli3JGLGZHduCkIaAA0mapAlZ\nwA78bu8r7U0D/XWBNLe34fa26W16ext6exdyf/21pktCf70NsQkUyE4gxiEJAclmMxiMhW1Jlixb\nkrXNPnN+fzwzo9E6I3lGGknn/XoZaR49zzPHI/w9z/csnyOqimEYhrF6cS21AYZhGMbSYo7AMAxj\nlWOOwDAMY5VjjsAwDGOVY47AMAxjlWOOwDAMY5VjjsAwDGOVY47AMAxjlWOOwDAMY5XjWWoD8mHt\n2rW6efPmpTbDMAxjWdHR0XFGVdflOm9ZOILNmzfT3t6+1GYYhmEsK0TkeD7nWWjIMAxjlWOOwDAM\nY5VjjsAwDGOVU1RHICK/JyKHROQ1EfmGiPhFpF5EnhKRI6mvdcW0wTAMw5ibojkCEWkC7gHaVPUK\nwA3cAdwHPK2qFwFPp14bhmEYS0Sxq4Y8QEBEYkA5cBL4j8CO1M8fAvYBXyyyHYZhGCXJvsP97N7f\nSddQkJa6cu6+sZUdWxoW1Yai7QhUtQf4K+AE0AsMq+oPgfWq2ps6rQ9YXywbDMMwSpl9h/v50hOH\n6B8NUxvw0j8a5ktPHGLf4f5FtaOYoaE64OPA+cBGoEJEfjX7HHXmZM44K1NE7hKRdhFpP336dLHM\nNAzDWBL2He7nnocP0nM2SN9wmLFInHKfB69b2L2/c1FtKWay+P3AO6p6WlVjwKPALwGnRGQDQOrr\njK5PVR9U1TZVbVu3LmdjnGEYxrIhvRMIRhN4XEI8qZw8G2Y0HCPgddM9FFxUe4rpCE4A14tIuYgI\ncAvwBvAE8JnUOZ8BHi+iDYZhGCXH7v2deN1CmccFKrhEEIHToxFCsQTNdeWLak/RksWq+gsReQQ4\nAMSBg8CDQCWwR0Q+CxwHdhXLBsMwjFKkayhIbcDL2soyTg6HIAmIEo4niSWUu29sXVR7ilo1pKp/\nAvzJlMMRnN2BYRjGqqSlrpz+0TDVAS8AZ8YiROJKhc/Dl2+9fOVUDRmGYRgzc/eNrcQSSjAap8rv\nobHGz8baAF+9Y+uiOwEwR2AYhrHo7NjSwJdvvZyGKj/DoRgNVf4l2QmkWRYy1IZhGCuNHVsalmzh\nn4rtCAzDMFY55ggMwzBWOeYIDMMwVjnmCAzDMFY55ggMwzBWOeYIDMMwVjnmCAzDMFY55ggMwzBW\nOeYIDMMwVjnmCAzDMFY55ggMwzBWOeYIDMMwVjnmCAzDMFY5pj5qGIaxwlBVwrFk3uebIzAMw1gh\nqCqjkThnx2N43JL3deYIDMMwSox9h/vZvb+TrqEgLXXl3H1j65yzC1SVkXCc4WCMeNLZCXjc7rzf\nz3IEhmEYJcS+w/186YlD9I+GqQ146R8N86UnDrHvcP+0c1WV4VCMrsEQA2ORjBOYL+YIDMMwSojd\n+zvxuoVynwcR56vXLeze35k5p1AOIE3RHIGIXCIiL2X9GRGRz4tIvYg8JSJHUl/rimWDYRjGcqNr\nKEjAOzmsE/C66R4KpkJAhXMAaYrmCFT1TVW9RlWvAbYDQeAx4D7gaVW9CHg69dowDMMAWurKCcUS\nk44Fo3Eaq/10DYY4M1o4B5BmsUJDtwBHVfU48HHgodTxh4BPLJINhmEYJc/dN7YSSyjBaJxkMslo\nOEY4luS2bc0FdwBpFssR3AF8I/X9elXtTX3fB6xfJBsMwzBKnh1bGvjTj11GfYWPwfEotQEf99x8\nEde11ud9j2A0zsMvnMj7/KKXj4qID7gV+I9Tf6aqKiI6y3V3AXcBbNq0qag2GoZhFJt8SkLTfQAX\nrq/i/tuumvd7nB6N8NjBHp585STjkUTuC1IsRh/Bh4EDqnoq9fqUiGxQ1V4R2QBMr4kCVPVB4EGA\ntra2GZ2FYRjGciBdEup1y6SS0C/j7ADSDmA4GCOWmH/452j/GHs6unnmcD+JpLNclnnyD/gshiO4\nk4mwEMATwGeAr6S+Pr4INhiGYSwZ2SWhAOU+D8FonN37O9m+uY6zC3AAqkr78SH2vNhFx4mzmeN1\n5V4+ubWJnW0tXPbn+d2rqI5ARCqADwB3Zx3+CrBHRD4LHAd2FdMGwzCMpaZrKIhboPP0GNFEEq9L\nWFPp49iZMU6PRuZ1r2g8yTOH+9nb0c07Z8Yzx8+rL+f27c184LL1+Dwu/N78O4uL6ghUdRxYM+XY\nAE4VkWEYxqqgqszDkf4xXAIugXhSOXk2zHlrKvK+x2g4xpMv9/LYwR4GxqOZ49e01LKrrZnrzq/H\nJfnrC2VjWkOGYRhFJplMoqqAoICms56aO/3ZOxziWx09fPe13oyiqEvgfZc0sLOtmYvXV52zfeYI\nDMMwikQmCRyOs766jKFULsDrdlFX7iMYm72y543eEfa2d7P/yGlS+V/KfW5+5coNfGpbE+ur/QWz\n0xyBYRhGgZlaBdRYHWBgPEJLXXnmnFAsQUNF2aTrkqr8/OgAe9q7ebVnOHN8XWUZn9rWxK9ctYHK\nssIv2+YIDMMwCoSqMhaJT6sCuuPaFh545gihWAK/10U4liSeVO64tgWASCzBD18/xd6ObrqHQpnr\nLlxXya5rm9lx8To87vn1/84nXWCOwDAM4xzJ1QdwXWs993IRD7/YRd9IiMbqAHdc28LFjZU89LNj\n/OtLJxkOxSbO31zHzrYWtm2qReaZAPZ5XFQHvFT68l/ezREYhlHSzHdIy2Iyn0aw61rrMzIRJwaD\nfKujmy89eYhoPDVIxiW8/9L17Gxr5vy1+VcTAbhEqCjzUOX3zKtsNI05AsM4R0p5oSoVFvoZ5erI\nXSpmmgiWzzWv9gyzp72bnx8dIF0vVOX3cOvVG/nENRtZU1k25z2mkv3073ItrHQUzBEYxjlRqgtV\nKXEun9FcHblL8fkuxAEkkspPjpxmT3s3h/tGM8c31Pi5bVszH76ycdr8gbkQESrK3FT7vQt6+p+J\nnI5ARN4N/CrwXmADEAJeA74D/LOqDs9xuWGsaEptoSpFzuUz6hoKUhvwTjqWHtKymCzEAYSiCb73\nWi+PdPTQNxLOHN/SWMWnr23hhgvX4p7HU7zX7aLa76XS75nXdfkwpyMQke8BJ3H0gP4cRyDOD1wM\nvA94XET+h6o+UVCrDGOZUCoLVSlzLp9RS105/aPhjBMBp+yyOasMs5gkk8poOM7ZUDQj5paLM2Mp\nBdCXexmLxAEQ4JcuXMOu7S1c0VQ9rwRwuc9DdcAz6TMoNLnu/G9V9cyUY2PAgdSf/y4ia4timWEs\nA5Z6oVoOzPQZnRmLEIwmuOH+Z+bMGdx9YytfeuIQwWicgNdNKJYgllDuvrG1qDYnk85IyOFQLG8H\n8M6Zcfa0d/H0G/3EU9f4PC5++fL13L6tmZb6/P+fcLuEyjIP1QEv3nmWjS6EOR1B2gmkxONCqpoU\nkYuBLcD3VDU2g6MwjFXDUi1Uy4mpn9GZsQinx6I0VPly5gx2bGngyzjhpe6hIM1FTsYnkspIyHEA\nyTzkH1SVAyfOsqe9ixePDWWO1wa8fGLrRm69eiO15b6837/M66ba76GyzDPvstFzQTSPv6yIdODk\nCOqAnwIvAlFV/TfFNc+hra1N29vbF+OtDGPepCtiFmOhWq5kf0bDoRgVZW7WVk5IJASjcRqq/Hzj\nruunXVPsaqx9h/v522ePcnwwyPpqP3e0teScBhZPJPnxm6fZ097F0dMTCqAtdQF2trXwgUsbKMsz\nkSsiVPjcVAcKl/zNuneHqrblPC9PR3BAVbeJyO8CAVX9SxF5KTWYvuiYIzCMlUPbf3mKYCROLKn4\n3C7WVZVRWeZhOBTjJ1+8GZhcaZS90/ryrZcX1Bk88/op/viJQ7iESR2/984yGnIsEufbr/Ty6IFu\nzoxNKIBe1VzDrrZmrm9dk7cCqMflosrv1P7Pt2s4X/J1BPlmHyRVPfRvgM+mjhXWdRmGseLZd7if\n0XCcpCpul2TkmNdUetm8pjJz3nwqjRayc0gkleFQjK/++G1cQqZ8M+10Hn6xa5Ij+MFrffzjT49x\nZiySqf93Cdx08Tp2tjWzpbE6788g4HNT5fdS4XMvavhnLvJ1BPfizBx+TFUPiUgr8OPimWUYxkpk\n9/5O6sq9DIxH0aSjh5NEGRyP8RefnMir5FtpNN8ehWTKAaRzAL3DIar9k5dBv9dF34ij9/Nm3yh/\n++xRXu6eqJIXnMX8d3ZcyIeubMzr7+0SST39e/HNY4TkYpGXI1DV/cD+rNedwD3FMsowjJVJ11CQ\ntZVllHmcpHE0kcTndhHwuiYt3PlWY+W7c1BVRkLTy0A3pFRBsxu6QtEEAa+HL+x5iZe6JhyA2yXU\nBbzUBLxEE0l++PqpnI6gUJ2/xSYvR5CqFPp9YHP2Nap6c3HMMgxjJZJe4KsDXqpTT/zpRHE2+VZj\n5do5pB3AcGjmRrBsVVCfRxgcjzESjjvOYsA5x+MS1lR4qfZ7M6Ecv2ti1zCVYiZ/i0W+oaG9wN8C\nfw/MPknBMAxjDvJd4PMtG51t59BUG2A4GMvZCHZdaz2/ETmfv3vuHXrOhsg+dft5dexqa+Zfnj/B\nYDA6KZ4fjjkzBrJxu4Qqv5fqIiZ/i0W+jiCuqn9TVEsMw1jxzKcvYMeWhpxJ36mOJRiNE4krn9ra\nxMD43EPhu4eCPNLRww8O9RHJUgC9eUsDO7c3c0GDk7zWJHPOEvB5XNQEvIte+19I5iwfFZF02vwe\nHHmJx4DMp6uqg0W1LoWVjxrGyudcFEr/9tmjnBgM0pBHH8BrKQXQn759JlMBVFHm5mNXbeSTW5tY\nVzVdAfSFzsFpswR2bGmgJuAl4Cvd8E9B+ghE5B3Ambg8HVXVOdsnRaQWJ5x0Reo+/w54E/gmTr7h\nGLBLVYdmuQVgjsAwVjoL7RtIl4GO5OgETiSVn759hj3tXbzeO6EA2lBVxu3bm/nIlY15afmkq38W\nS/rhXClIH4Gqnp+6mV9Vw9k/E5F8Jic/AHxfVW8XER9QDvwh8LSqfkVE7gPuA76Yx70Mw1ihzFeh\nNJlUzubhAEKxBN9/rY9HOrrpHZ5Ywi5ZX8WutmZuvHhdXkqePo+LKr+XqrLSrv5ZKPnmCH4GbMvj\nWAYRqQFuBH4NQFWjQFREPg7sSJ32ELAPcwSGsarJt29gah/AbAyOR3nsYA9PvHyS0XA8c/zdrWvY\ndW0zVzXV5IznizjCbwud+rWcyCVD3Qg0AQER2cpEiKga5+l+Ls4HTgNfE5GrgQ6cxrT1qtqbOqcP\nWL9A2w3DKCHORRsoV99AvmJw75wZ55GObn70xiliCec8r1v44GWN7NzezKY1uRVAi6n7X6rkyhF8\nBueJvg3IDtKPAl9X1UfnuLYNeB54j6r+QkQeAEaA31XV2qzzhlS1bobr7wLuAti0adP248ePz+fv\nZRgrilIfh3mu2kBTrx8YjzA4HqPC56J1XRU7tzdz7fkzJ4BVlZe6zrKnvZtfvDNRv1Lt9/CJa5r4\n+NaN1OWhALoYuv+LTaFF525T1W/N04BG4HlV3Zx6/V6cfMCFwA5V7RWRDcA+Vb1krntZsthYzRRT\ngK1QDubOB5+f9kQ/k6JoPrYc6R9lJBSjNuDF4xbOjEeJJ5TNayq4672tmYqgeCLJs2+d5pvt3bzd\nP5a5T3NdgNu3N/PBy9bnDOm4RKj0e6heROmHxXTqhRad+7aI/F9M7yz+8mwXqGqfiHSJyCWq+iZw\nC/B66s9ngK+kvj6epw2GsSop1jjMQs5bLsSkth1bGnjvxev49O6f0z8aJp5Q+kbCmSavzjPj3P+D\nw9x780X0jYZ59EAP/aMTvQJXNlWzc3sLv3RhbgVQr9uRfljs5G+pzrjO1xE8DgzjxPnn7tKYzO8C\n/ydVMdQJ/DrgAvaIyGeB48CuedzPMFYdxRqHWUgHU+lz8/bpMRIpaem1lWV43JL3pLbsMtCuoSDV\nfg+9w0GmNgWfDcb40++8TjqQ4RK44aK17NrewmUbcyuALnX4p1RnXOf7aTSr6ofme3NVfQknvzCV\nW+Z7L8NYSpYyRl+scZiFcjD7DvczkArfuARiiSQ9Z0PUlnv541+5bM5rY4kkZ4MxxiJx0mHqtBBc\nOtmbjab+4/e4+NAVjdy+vZmNtYFp52WTln6o8nuWvPa/VGdc5/up/ExEriyqJYZRRPYd7ufOB5/n\nhvuf4c4Hn2ff4f55XfulJw7RPxqetJ2fzz3OhbtvbCWWUIJRZ7EMRuMFGYfZUldOKDZZOmwhDmb3\n/k6qA16a6wJ43S4UR6phXWXZrM4yEk/QPxKmazDIaDhGdq7yjmtbiCWmC8SlcQk8fNf13HPLRXM6\nAb/XTUO1n0315dRX+JbcCUDhPvNCk++O4Abg11KdxhGcMlJV1auKZplhFIhzjcsu9XZ+Nn0ecJK0\nC5Fk2L2/k7dOjTAWSVBf4WVNRdmC5y2nn3JFnCdvcCp5hkOxaeeGYwnOBmMEo/FpPwOIxpMMBKPM\nVMPiFlCFzfXlGeXSqaRr/6sDHso8pVf7X6ozrvN1BB8uqhWGUUTOdSEvhe38VAG2hTq39HXReIJw\nLEkskaRvJMJYOM4VTbULCnnlE7oKRROcDUUJRWcWLx4JxXjylZM8dvAkg+MTIyA9LgEUVXC7hQqf\nh7tuvGDa9cul9n8+onuLSb6DaY6nmsLemzr0E1V9uXhmGUbhONeFvFgx+nNhvs4tvQs4cGKIZDJJ\nEsEjgs/tIqFKNPVUupAFaa6n3GA0zlAwRiQ2swPoORviWx3dfP+1PsIpBVB3lgLo4Fh0mthbtqDc\nUid/F0I+qqqLTb6Dae4FPgekG8j+WUQeVNX/XTTLDKNAnOtCXorb+fk4t+zdQyKZJJEERXG7BZcI\n7lSCd6Ghrpmecj/z7vO4qLGKvuHwjNe8fnKEPe1dPPf2mUxlUIXPza9ctYHbtjVPKIA2ME1JtJSS\nvyuFfN3oZ4F3qeo4gIjcD/wcMEdglDznupCX4nZ+Ps4te/dQ5nEzngrPxJNJ3C43qlDmdp1TqCv9\nlDseiTMUjBKNJ6ftAhJJ5WdHB9jb3sVrJ0cyxxuqyrhtWxMfuXIDFWWzL0krQfe/VMnXEQiTJ5Ml\nmFma2jBKjvks5LOViZbadn4+zi1797CuqozgQBDFSbwm1Ym/11R4zynUFYzGGRx3HMBUwrEEPzh0\nikc6uuk5OzHe8aKGSna1tXDTxWtnneglIpT73FT7S1v3f7mTryP4GvALEXks9foTwD8UxyTDKDz5\nLOSl2vU5E/Nxbtm7hyq/l3WVPvrHoihOJU5NhRev272gUNdcOYChYJTHD57kX1/qYSRLAfT61np2\nbm/mmpbaWZ/svW4XVX4PlWXLb+zjciQvrSEAEdmGU0YKTrL4YNGsmoJpDRmLQSH0ckqRfYf7+Q+P\nvMxoOE48mcTjcuFzC0115YxF4gsKdY1H4pwNzewATgwE2dvRzQ9f75ukAPr+S9ezs62ZzWsqZrzn\nC52D7Onoom8kzHn1FUseflsJFFprCOAdIJ66RkRkm6oeWKiBhlFqlEKZaLFQAHFCLQj4vG6++KEt\n815oxyJxzganh4BUlVe6nRGQP+8cyByv9nv42NXOCMj6ipkVQEWEV7rO8tf73sbncVFf7ivp3dhK\nJN+qoT/DkaM+CpkxnwrcXByzDGPxKcUy0UKwe38nNQEvG2omunDn00ehqikHEJvW8ZtIKs++dZq9\n7d28eWpiBOTGWj87tzfzwcsbCcyiAJpd+/+Hj76Kz+MqOQ2e1UK+O4JdwAWpKWOGsSIpxTLRfMil\ng7TQnY6qMhKOMxyMEU9OdgDBaJzvvtrHtw50c2pkQofysg1V7Lq2hfdcsHbWxq5AKvn74juDGbtP\nj0ZorJ48NH6l7MaWA/k6gteAWmBxxFUMYwlYaJnoUgrS5ZPgrirz8Hb/GAl1lEHXVZXhds2uDJpM\nKqPhOMOh6Q7g9GiExw728OQrJxmPOPkBwVEA3bm9mSuaaiad/0LnYKYhrKm2nN+8qZUPXN44ze4z\noxF6zoYBychHrITd2HIhX0fwF8BBEXmNLBlqVb21KFYZxhIx3zLRpa40ytVhvO9wP6fHIsSTE8qg\n3UMh6mZQBk2PgxwJx0hM0X8+enqMve3dPHO4n3jqZ2UeFx+6vJHbtjfNuGC/0DnIV585gs/jYk2F\nj7OhKH/2nTfwul3T7G6s8dM9FOLUaBhQTo066qM+t4t9h/stPFRk8nUEDwH3A68Cs8sCGsYqY6kF\n6XKFfdL5gQqfhzNjEaKJJB63sKbCl7EvnkgyHIoxGo5PmgesqrQfH2JPezcdx4cyx+vKvXzimiZu\nvWYjNbOIv5X7PDx6sIeAz5312bgyn03a7v6RMKfHIpnu4qQq3WdDeF0ummv9RBNJSxovAvk6gqCq\nfrWolhjGMmSpK41yJbgzyqC+iZBLWhk0Gk9yNhRlPJKYJAMdSyR55nA/ezu66Tw9njl+Xn05O9ua\nef+l62cc6zh17OPJ4dCsn01LXTlv9o0wGIwhpOSMgaRCfcBLU9YOw5LGxSdfR/ATEfkL4Akmh4as\nfNRY1Sx1pVGuBPdM9qV7I6Y6q9FwjG+/0sujB3sYGJuoC7mmpYZdbS1cd379jCMgZxv7ONdnc/eN\nrfz6118EJsoQ0wwFYzTVTby2pHHxydcRbE19ze6qsfJRY9lQrIRuISuNFmJjrgR3tn1lHhfjkTjR\nhHLX9ubMPfqGwzxyoJvvvtpLOOZEfl0COy5pYFdbMxevr5rxvQM+NzUB76zKn3N9Nmn7JLUVEAG3\nCPGU5EU2ljQuPnl3Fi8l1llsnAvZCd3sBenLt15eEGeQXsDPRZAuHxsX6sx++Foff7v/KCfPTpZy\nPtw3wp4Xu9l/5HQmRh/wuvmVqxr51LZmGqv90+4138Evc302V/3nHxCKJfC4JsJMkXgCVWhdV1GU\n39VqI9/O4jkdgYj8KvAvqjpjglhELgA2qOpzC7Y0D8wRGOdCqUpHZC/sI6EY5T4366omFt9sGxfi\nzGaSgUiq8vOjA+zt6OaV7uHM8bWVPj61tYmPXrWRSv/0J/x081eVf3L451z46o/e4oFn3sYlzg4k\nqc6f686r5VDvKOPRBBU+N79xw/nc8/6LC/Keq41CSUyswSkb7QA6gNOAH7gQuAk4A9x3jrYaRlFZ\n6oTuTEwtO+0dDhGKJSjzuDNJ3anVP/lUJ83WBRyJJXjqjVPsae+me2hCAbR1XQWfbmthxyXrZtT2\nL+bgl/Ti/vfPvZNZ9G/Zso6OE8OsqypjU8rhPXKgh6uaa21HUETm/O2q6gMi8v/g5ALeA1wFhIA3\ngH+rqifmul5EjgGjOLLVcVVtE5F64JvAZuAYsEtVh2a7h2GcK0ud0J2JqQu73+MmmkhyZiwyY0NV\nLmc2WxPY2WCUx186yeMvneRs1gzhazfXsauthW2bpiuALubgl3vef/Gkp/07H3x+SctxVys53byq\nJoCnUn8WwvtU9UzW6/uAp1X1KyJyX+r1Fxd4b8PISSlKR0xd2NdVldEzFCIST6Kq02ys9Ll569Qo\n0cREKNfnFlrXVjA0Hp3WBNY1GOSRA9384NCpjECcxyXccqkzArJ1XeU0m8q8bqpT0s9LNfilFHdv\nq4E5HYGIfA7Yp6pHxPk/4x+A24DjwK8tsHz048CO1PcPAfswR2AUkeUwYazK72VtVYLxSILhUGyS\njfsO99M7Ep7kBACiCaVrKMgPXuvjutZ6VJVXe4bZ297Nz44OZMoyK8s8fOzqDXxyaxNrKx09n7T0\nQ+9IiOZa570+eEVjwf5+C01sl+LubTWQa0dwL/D11Pd3AlcDrTjlpA8wMcx+NhT4kYgkgN2q+iCw\nXlV7Uz/vA9YvwG7DmBfLYcKY1+3mq3dcNc3O3fs7icSSmaarbGJJ5RsvnCAYS7CnvYvDfRMKoI3V\nfm7f3syHr2icNN0rW/phbUr64b989w18Htek917oYn4ushuluHtbDeRyBHFVTQcWPwr8k6oO4Czu\nf5nH/W9Q1R4RaQCeEpHD2T9UVRWRGcuWROQu4C6ATZs25fFWhlHaTF1Yb9/WxM87BzO7lHe31rN7\nfyd/9PhrkxberqEgsURymhMAiCeU104O83LPRAXQlsYqdrW18N6LpiuA+r3uOaUfsktVF7qYn4vs\nRinu3lYDuRxBUkQ2AEPALcCfZ/0sMPMlE6hqT+prf2rM5XXAKRHZoKq9qXvPqGia2j08CE75aM6/\niWGUMDMtrI8c6MmUf8628P6nWIKGqjL6R8IkEtP/GaTTAgL80oVr2LW9hSuaqqfF+J3afy9+r3tO\n6Yc057KYn2ucv9R2b6uBXCUBXwLacap7nlDVQwAichPQOdeFIlIhIlXp74EP4shZPwF8JnXaZ4DH\nF2q8YSwXshdWZyC7B69b2L2/c9rPgVSzlvI3z3ZydVPNJDG4qbzr/Hq+/uvX8mcfv4Irm2syTsAl\njr5QS305DdV+/KkBMS115YSmjJicGofvGgpOGyiT72Kez/2N0iJX+ei3ReQ8oGpKiWc78Okc914P\nPJb6n9KD05j2fRF5EdgjIp/FSTrvWrD1hlEkCi1JkespuWsoSI3fQyKpJJKKqlLmcXHszCgnhsZT\ni/tkZ1AX8PLb77uQmy+dbJfH5aIm4Ez+mmk4TD5x+HNJ2lqcf/mRq2rov6rqHwJDIvIBVX0KQFXH\n57oudU4nTnJ56vEBnDCTYZQkxZgxMNfCmkgqjdV++kfD+FOyDarKwHiUkUiCpE48XXvdjsRDS105\n/+uOaya9h9/rNKNVlk3+Zz2TU/vyrZfPGYdPL+ZnxsIMB2NEEs7Q+49fvTHnZ7d7fyfBaJxoPInP\nLVy0vtri/CVOrhzBh4A/TH1/PwvvJTCMZUMxZgzM9JQcjSe549oW9rZ3MTgWoXsohFscMbfxSJJE\nVjgo4HVRV+6jwucGgf7RcOZnFWUealLx/6nM6tRuvXxOeY0dWxq4vfssf73vKImk4ve4qfJ75uzy\nzX6vxmrHFudGAAAgAElEQVT/jCJzxWApJ8StFIrbNmgYy4x9h/s5cGKIE4NBOk+PMZLqxj3XpqYd\nWxr48q2X01DlZygYpSbg5bd3XEgomuCBp48QiicIeN3EkjASTpBQRYDagJeGqjJa6sozjV7hWJLG\nmgA1qfj/+qz4/1Ry5Sbm4uedgzTXBbh0QzWt6ypZV+Wf89pzea+FknY+/aPhSY5u32Gbqjsfcu0I\nGkTkCzhFCenvM6jq/yiaZYaxyKQXlfSglHhCOTns6PJ43LPP+M0HVWXb5jr+suGqSRpAv/n/dTAc\njBHMSq4KsKbSxwN3XEPXQIgHnjlCKJbA73WlOo/hnpsvZE1l2QzvNJl8Knhme6Keb/XPUnQFL/WE\nuJVCLkfwd0DVDN8bxoojvag01vg5eTYMAqJwajRMQ5V/QcnOWCLJSCjGWCSekYBIqvLCO4Psae/m\nrf6xzLlul1AX8FId8BCMJthQE2BDTYB7uYg9HV2cGgmzqb5iXqGPXEnfufIh800YL0VXsElSFIZc\nVUN/OtvPUiWhhrFiyIx1FGFjLZwejRCJJxGVeevhh2OJjANIyzmcHA5S5nETiSfpH80M+sPrEuor\nfI7EswihWILG6gAiQkWZm09ub+KOdy2sqTJXBc9cT9Tzrf5Zimohk6QoDDlzBCLSJCJtIuJLvW4Q\nkf8KHCm6dYaxiGTXv1f5vbSuq+S8NRVs21SXtxMYi8TpORvi5NlQxgn8zx+9ReeZMQbGonQNhTJO\nYPt5dfy7X9rMmkofPo8LEWcRiyeVz96wmZa6AA1V/rwGwMxGdm5iOBSjocrP7dua2L2/kxvuf4YD\nJ4aIJyaPG0k/Uc907VwOcb7nF4K7b2wlllCC0TiqzlcrVZ0/ucpHPw/8J+BtoExE/l+c6qF/ArYX\n3zzDWDwW+kSbloAeCU+eAdAzFOKvfvgmA+PRSR0A5T43LbXl/LfbrwLg4vVVPPxiF6dGQjTXl/Nb\nN13A+y4tnARXdqfu1FDQmdEIPWfDgMwofz3fLt/F7go2SYrCkCtHcBdwiaoOisgm4C3gParaUXzT\nDGNxme+iEk8kGQ7FGA3HJ3X+Hjo5zJ72bp47cmaSA/C4YE1FGdUBD8PhieHwN21Zx61bNxZl+MtU\npoaCGmv8dA+FODUapsrvWZbNXyZJce7k+j8vrKqDAKp6QkTeNCdgrGTyWVQi8QTfe6WXr//8OL3D\nITZUB9jV1kw0meRrzx3j+ODkRKVLwC0AwsB4hEQySXN9BZV+Dy+fOMs//vTYotXAT02uVvm9NNUq\nfSORafLXxuohlyNoFpGvZr3ekP1aVe8pjlmGUXqEognOhqI8e/g0DzxzBI9LqCxzc2xgnD9+4tCk\nwTBlHhfxZJJk0ikHRQQR0CSMhOP87vsu5PWeEf7sO28UtIM5FzMlVz1uF9s21S3p/GZjacnlCP7D\nlNe2GzBWFcmkMhqJMxKaiP8//GIXgjMcvudsjKz1H7/X0fgPeN28MzCOxwUigtslxBJJvC6hwu/l\nlsvWL8lYRtMBMmYiV/noQ4tliGGUErPF/48NjHP41AiR2MR8AAGq/B5Gw3Fa6gKIswfA63YRTyRJ\nqnJJQzUAwWichio/sDQ18JZcNWai+Nkpw1hGZNf/p1FVXu4eZk97F893DmaOu8SRgKgt9xJLOKqh\n4ViSRFIZCkYJp5yFzy0zziFeqhp4S64aUzFHYBg4T+pngzHCWVIP8USSZ986zZ72bo5kdQCvqfAR\nSySp8nso97kJx5LEk8qutmYef/kkw8EYbpfgdQvxhKJA30iYixqqJj19W5jGKBXMERirFlVlLBJn\nOBQjGp+o/x+PxPnuq71860DPpA7gKzZWs7OthV+6YA0dx4Z4+MUu+kZCbKgJ8Bs3nM+Hr9zA852D\nBCOOaJzP7WJjbRlul9BQ5Z+WjLUwjVEqmCMwVh2JpDISijESjk2q9OkfCfPowR6+80ov41FnZ+AS\nuOHCtexqa+GyjdWZc69rrefGS9ZRE/BSkaX/PxqJc2FD5aRRkao6a9zfwjRGKWCOwFg1ROIJhkMx\nxiMJNCsB/Hb/GHvau/jxm6czjsHvcfGhKxq5bXszTbUBXugc5AvffJnekRDNtc6T+wevaJz2HqZ9\nYyxHzBEYk1iJQz7GU+Gf7Pi/qvLCsUH2tndz4MTZzPG6ci+f3NrEx67eSE2qoueFzkEeeOYIPo9T\nGno2FOW/fPcNfB7XtM+mkHH/lfi7MEqTXFpDX5jr5zaPYGVRjBGNS0UiqYyGnfLPbP2faDzJ04f7\n2dvexbGBiXDNeWvK2dXWwi1bGvB5JrQYPS4XjxzoJuB1UVHmOIZyn2vWev9Cxf2X4+/CHNfyJdeO\nID1/4BLgWuCJ1OuPAS8UyyhjaVgJQz4i8QQjoThjkfik8M9IKMa3X+nlsYM9DIxP6Pxs21TLzrZm\nrttcPymu7/OkBsCXeegbCc+r3r8Qcf/l9rtYjo7LmCCveQQish/Ypqqjqdf/GfhO0a0zFpXlPORj\nPOKof4aiiYz+f+9IiPpyH3XlPg6cGCKcqgxyCbzvkgZ2tTVz0frJs5YCPjc1Ae+kGL8NXMnNcnNc\nxmTyzRGsB6JZr6OpY8YKYrklOmeSf0jH8xNJR5v+1MhE+We5z81Hr9rAp7Y20VDtn3SvyjIP1bMM\ngLeBK7lZbo7LmEy+w+v/CXhBRP5zajfwCyAv+QkRcYvIQRH5dup1vYg8JSJHUl/rFmS5UXCWy5CP\nWCLJwFiEE4NBBsYiGSeQSCp/u/8oA2NR+kcjjEWc5LBbhI01fr551/X85k0XZJyAiKPB31JfTsMc\nA+Bt4Epusof6pCllx2VMRrLjqHOeKLINeG/q5X5VPZjndV8A2oBqVf2oiPwlMKiqXxGR+4A6Vf3i\nXPdoa2vT9vb2vOw0zo10wm8xG5zyTTIGo3FGw3HGs+QfwJGF+MGhU3zrQDfdQ6HM8TKPi7pyL5Vl\nbsYiCf7lc05Dl9slVPu9HDw+xN89907JJjeX4nexULJzBNm7pmI7TGNuRKRDVdtynjcPR3ADcJGq\nfk1E1gGVqvpOjmuacXYOfw58IeUI3gR2qGqviGwA9qnqJXPdxxzByiXXApJIKmMzTP8CGApGefzg\nSf71pR5GwhPOwe9xsbbSUQCV1AzgNRVlfPXOrdQEvFT5Pex/67QtXAVmOTmu1UK+jiCvHIGI/AnO\nU/0lwNcAL/DPwHtyXPq/gD9govoIYL2q9qa+72OWXIOI3IUzIY1NmxY2uNsofWZLMv7Ns0e5vKlm\nWvUPwInBII90dPODQ33EEs7PvG7h/ZeuZ8v6Kh5u73IqgFIzgBNJ5a4bz6elPpCpDCrF5OZyL7+0\nLunlS77J4k8CW4EDAKp6UkSq5rpARD4K9Ktqh4jsmOkcVVURmXFLoqoPAg+CsyPI006jxMi1uGUn\nGVWVpDox/eMD44yGY5nzfnF0gH/46TFODAUn6QJV+T3cevVGPnHNRtZUlgGwvtqfmQHcUl/Bb+24\nYNoCVWrJTSu/NJaSfB1BNHvRFpGKPK55D3CriHwE8APVIvLPwCkR2ZAVGupfkOVGyZPP4tZSV86p\nkRBlHjcJVVDnKb6xOgA4CeB/fO4dHjnQnXn6ByfO/9ErN3DXTa0EpiR5b7wk9wzgfKpyFvMJvRR3\nKMbqId+qoT0ishuoFZHPAT8C/n6uC1T1P6pqs6puBu4AnlHVX8VpSvtM6rTPAI8vyHKjaOw73M+d\nDz7PDfc/w50PPs++wwvz1dmLm4jz1esWdu/vBJza/53bmwnFkpkQUCiWIJ5UPnnNRh7p6OZX/+EX\nfOPFrowT8HtcbKj2s6G6jOMDwUlOIOBzs7E2wMbaQM5B8LmqctJOrH80PMmJLfSzyEXXUHCaQ7Py\nS2OxyGtHoKp/JSIfAEZw8gRfUtWnFvieX8FxLJ8FjgO7FnifZU2pxoMLGaKYKfzi97g4PjBO12CQ\nWCLJNZtquffmizKSzvUVZayrLOMvn3qT8chEOWJFmZv6cl9msVSUvpFQ6mceambpAZiNXFIQi/2E\nvtz6BoyVRb7J4vtTJZ5PzXAsJ6q6D9iX+n4AuGXelq4gSjkeXMgFMHtxS6ozwWs84oxqzK4Auq61\nnjVVPva2d/PM4X5e7x0BnPLPX768kSOnRhmLxCc9MYdjSZpqy2mqC1DmmTg+Hwc7V3JzsXMINqTG\nWEryDQ19YIZjHy6kIauJXCGTpaSQIYq73ns+kXiS4VCUaDzBeCROPKnccW0L4CSHXzw2yB888gqf\n+6cOfvj6KeJJpa7cy6+/ZzMPf+56Pv/+i/jMuzcTTzpho7FIjOOD4/ScDRGMxvn52wOZ9ytkOGex\nG6SWomktF4UKERqlTy710f8b+C3gAhF5JetHVcDPimnYSqbUKlayKUSIIpZIMhKK0dpQye/suDAT\n9mmsDnDHtS1sPa+WHx7qY097N51nxjPXbaovZ+f2Zj5w2fpJCqDXtdbzebmIv/tJJ51nQvjcLlrq\n/EQTSX7/kZdZV1mWkZoo97mpCTidw+eym1mKJ/RSKr8s5V2rUXhyhYb+Bfge8BfAfVnHR1V1cOZL\njFyUcjz4XBbAYDTOSChOMDrR3HVdaz3XtdYDMBaO8+QrJ/lvT73JwNiEdNXVzTXsamvhXa31uLIU\nQAFcKRmI29ta+NeXTqKQ+dxGQjHOBmOMheOsry5jJBxnJBxnaDzKhtoAVX7vgh3sTDmEd7fWs3t/\nJ3/0+GsFzeuUYr7IqphWF7nUR4eBYRF5AEcWIq0+Wi0i71LVXyyGkSuNUo4Hz1dPfzbd/2z6hsN8\n60A33321LxNucQncdPE6drW1cEnj9JaUtAxEdcCL2+U4h6k7qTNjEVwC8aTSOxxBBFQhmlROng2z\nsda5T3Nd+YIW2+wn9GI9IZfqk3cp71qNwpNvH8HfANuyXo/NcMzIk1IfWp5PiCIcSzASijEeTUzr\n/E3zZt8oe9q7ePat06RHAwe8bj5ypTMCsnGKAig4g2DSMhAu1+TdwdSdVDSRRICkKm6X4HW5iCaS\nOOYofcNhGqr9vLu1ni89cYhYIsFwMEbvcIgDJ4b47R0XcM/7L87rMynWE3KpPnmX8q7VKDz5OgLR\nrH/tqpoUERtzeQ6UUjw4X9Kyz6Ph2KTu3knnqPJ85wB72rt5pXs4c3xtpY9PbW3io1dtpNI//X+d\n7EEwMiU8lGbqTsrtEuIJRUQQccJIHhWSqiigwJdvvZzd+zuJJRIMjMUQAa/bRSKp/PW+o1zVXJvX\n76FYT8il+uRdyrtWo/Dku5h3isg9OLsAcBLIS1/iYiwK0XiSkbATi0/O8vQfiSV46o1T7G3vpitL\nAbR1XQW7tjfzvi0NeN3Ti9T8Xje15d5pDWCzhXKyd1Kb68sZGI8yGo6TTCoCiEBLbTket9BQ5WfH\nlgb+6PHXGA7GMs4CwJ0KKeX75F2sJ+RSffIu9V2rUVjydQS/CXwV+COcB62nSQnCGSsTVWU86oR/\nwlPKKLMZDsZ4/OUe/vXgSc6GJrSB2s6r49PXtrBtU+2MT/jlPg+15TM3geWKm2cvRvsO93P/9w/z\nVv8YXjdsrPLjccukp9eWunJ6h0OTHJGq06eQ75N3sZ6QF+PJezanmitvshx3rcbCyFuGeikpVRnq\nUqz2OFfSyd+RUJx4cubwD0D3UJBHOnr4waE+Iqkwkccl3HJpA7dvb+aCdZUzXpdPF/CdDz4/7Sk5\nGHUa0b5x1/UzXjOXBPK+w/3c/c8dTi5BBFVIoqyp8HH+2spZ7zmf9zgXiinfPJvM9+3bmnjkQI/J\ncK9wCjKPQET+QFX/UkT+N85OYBKqes+5mZkfC3EExV6kV9ogjkg8wXAoxnhk9uSvqnLo5AjfbO/i\nZ28PZP6HqCzz8LGrN/DJrU2sTSmATqWyzENNuXdSF/Bs3HD/M9QGvJN2EqrKcCjGT75487z/bgBf\n/dFb/PW+oySSSpnHRZXfg8/jXra/r3yZzameHo2wrqpsXs7WWH4Uah7BG6mvpfc4PgeLUZJXqtUe\n8yHf8E8iqTz39hn+8bl3JsX/68q9/Jt3beLDV2wg4Ju+wIuI4wAC3kkNYrkoRtz8nvdfzFXNtasu\n5j1bMno8mmCTidwZKXL1ETyZ+prXfOJSoRiL9NQdxpH+0Wnlj8vlH1I8kWQ07Ix9nCv8E4ol+P5r\nfTzS0U3vcDhzvMzjorLMjdftorm2fJoTEBGq/B5qA148MySIc1GsuPlqjHnP5lQrfM7nWmpJamNp\nyCUx8SQzhITSqOqtBbeoABS6JG+mHcZoOI7XHWFt5YQzmO8/pMXOMYSiCUbDc9f+AwyMRXjsYA9P\nvtLLaPYISK+LtRVlBLyuzAjIh1/synQOu1IOoGaBDiCNVawUjtmc6m/ccD6PHOix8lADyB0a+qvU\n108BjTjjKQHuBE4Vy6hzpdChhZl2GHXlXgbHY5T7PAv6h7RYHaXp2v+R0PSZv1N558w4e9u7efrw\nqYz+v8/j4pcvW8/Pjp6hvsKHMBG393td9I2EMjIQNVldwAthJSbfl5q5nOpqDJUZM5MrNPQsgIj8\n9ykJhydFpGTzBoUOLcy0w1hbWUY8kaShyr+gf0jFzjGkhd9G56j9BydPcPDEWfZ0dPPCOxPyUTUB\nLx+/ZiMfv2YjdeU+ugZDDIxHJimTRuJJmuvKaakvPycHAKUrtbASmC0kthpDZcbM5NtHUCEirara\nCSAi5wP5jKtcEgodWphth3HR+uoFV1gUq6N0POLE/rOF32Yinkiy763T7Gnv5u3+sczx5roAO7c3\n88HL1lOWtejfcW0LDzxzhFAsgd/rJpaScvid9114zk4AVkby3TCWK/k6gt8D9olIJyDAecDdRbOq\nABTyaacYyctChq/yTf4CjEXifOeVXh490MPpsUjm+JVNNexqa+bdF6yZpgAKjoro77kuYm97N73D\nzlD4QoYSSlVqwTBWA/mOqvy+iFwEbEkdOqyqkbmuWWoKGW8uRvKyEM4l3+QvwKmRMI8e6OGJl09m\nGsDAcQC/eVMrl26onvVaj8tFbYWXnW0t7Lp2U972zYdSlVowjNVAvqMqy4EvAOep6udE5CIRuURV\nv11c8xZGMeLNhY6nLtS5zCf5C/DWqVH2tnfz4zf7MwqgAtQEnCT3mbEIo6GZw0hpB1A1hxBcoTCR\nM8NYOvINDX0N6ADenXrdA+wFStIRLJd483ycSziWYDQcZywSz/n0r6q8cGyQPe3dHDxxNnPc4xIq\nytysrSjLxPWnloA65y2eA0iTj2O0qiLDKA75OoILVPXTInIngKoGZbFWiAWwUuLNqspYxJm6FZmj\n8zdNNJ7k6TdOsaejm+MDE3/X89dWsHN7M1//2TvUBLwzloDC0jiAbOZyjFZVZBjFI19HEBWRAKnm\nMhG5AJgzRyAifmA/UJZ6n0dU9U9EpB74JrAZOAbsUtWhBVk/C8s93hyNJxkNxxiLxEkkc4sCDodi\nPPnySR472MNQcEIBdPumWnZd20LbeXWICD88dGpaCWg4lmRDTYA1lWVU+5fGAeTDctnlGcZyJF9H\n8CfA94EWEfk/wHuAX8txTQS4WVXHRMQLPCci38NpTntaVb8iIvfhzEL+4oKsn4XlGG9WVYLRBCPh\nGKFo7qd/gJ6zIR7p6OYHr/URTiWA3S7h5i0N7NrezAUNkxVAJ5eAuojEnRLQe26+kJopO6hSY6Xs\n8gyjFMnpCFIhoMM4C/j1OLnGe1X1zFzXpSaapQvUvak/Cnwc2JE6/hCwjwI7guUkURBPJBkJxxnL\no/QzzaGTw+xp7+a5I2cy+h8VPjcfvWoDn9rWzLqqmRVAr2ut514u4pvtXfSPhtlU4BLQYrLcd3mG\nUcrkNY9ARF5V1SvnfXMRN06S+ULgr1X1iyJyVlVrUz8XYCj9esq1d5EafrNp06btx48fn+/blzTB\naJyRUO7GrzSJpPLTo2fY297NoZMjmeMNVWXctr2Zj1zRSEXZ3H7d53FRW+6jMsd5pchKk/02jMWg\nUDLUaQ6IyLWq+uJ8jFDVBHCNiNQCj4nIFVN+riIyoydS1QeBB8GZRzCf9y1V0kNfRsPxvEo/wakW\n+sGhPvZ2dHPy7IQC6MXrK9nV1sJNF6/L2dlb5nVTG/DmdBSlzHLa5RnGciPfleFdwK+KyDFgHCc8\npKp6VT4Xq+pZEfkx8CHglIhsUNVeEdkA9C/A7nmx1GWH4Zij+Z9P41eawfEoj7/Uw+MvnWQkSwH0\n+tZ6drW1cHVzTc7EbrnPUQKdaVbAcsS0cQyjOOTrCH55vjcWkXVALOUEAsAHgPuBJ4DPAF9JfX18\nvveeD0tVdphMKmNRp/ErGs/v6R/g+MA4ezu6eer1CQVQr1v44GWN3L69ifPW5JZ4qihz5gHnMw3M\nMAwj1zwCP87g+guBV4F/UNX8gtqwAXgolSdwAXtU9dsi8nNgj4h8FjgO7Fqw9Xmw2GWH0XiSkXCM\nsRyqn9moKi93D7OnvYvnOycUQKv9npQCaBP1Fb457yHiNIvVBnzzmgZmGIaRa0fwEBADfgJ8GLgM\nuDefG6vqK8DWGY4PALfMz8yFs9Cyw/mGk8Yj8XmVfoJTMfTsW2fY097FkSwFUJ/HxYcvb+Tum1rn\nHPIOhRsGsxpZ6pChYZQKuRzBZelqIRH5B+CF4ptUWBZSdphvOGk+qp/ZBKNxvvNqH9/q6KZ/dKIv\nz+d2sabSi1uEF44N8u7WNZOkH7Jxu4SagJdqvxdXAWSgVxvWqWwYE+R6hMy0qc4jJFRS3H1jK7GE\nEozGU01b8ZzNZdnhJBHnq9ct7N7fCTjJ3/6RMF1DIYaC0bydwOnRCLufPcqnH3yev9l3lP7RCC5x\nhsA0VPnYvKacqjIv5T4PHpfw8Itd0+7hcblYU1FGS105teU+cwILJNfv2DBWE7l2BFeLSLpoXYBA\n6nW6amh27eISYSFlhzOFk/weFycGxukeCs4r+QtwtH+MPR3dPHO4PyMZ4fe4+OUrGrl9ezO/v/dl\nqv2TfxXZGkAAXreLmvKl0wFaaVinsmFMkGtU5YooO5lv2WF2OCmpSjKpjEfjrKvy5+0EVJUXjw2x\nt72LjiwF0LpyL5/c2sTHrt6YkXXYUB2YUQOosTqA1+2ittxLlb+0JSCWG9apbBgTLN8OoyJy13vP\n54+fOEQ8EcXncRGOJYknlTuubcl5bTSe5JnD/ezt6OadM+OZ4+fVl7OzrZn3X7p+WlXPVA2gcCxJ\nIqncfVMrLfW2MBWD5ahHZRjFwhxBFunk7wUNVfzOjgt5+MUu+kZCNFYHuOPallkTtwCj4RhPvtzL\nYwd7GBiPZo5f01LLrrZmrju/fsYRkDChAfTwi12cGnHGQP7WjgssaVlErFPZMCbIS2toqWlra9P2\n9vai3T+UUv0cj8w/H947HOJbB3r47qu9hGNO2Mgl8L5LGtjZ1szF66vyus9KkIEwDKO0KLTW0Ioj\nkVTGwk7tf766P9m80TvCnvZufnLkdGYEZLnPza9cuYFPbWtifbU/r/v4vW7qyn0rRgbCMIzlx6pz\nBJF4gpFQfiMfp5JU5edHB9jT3s2rPcOZ4+sqy7htexMfuXJD3sqeAZ/jAHI1jBmGYRSbVeEI5jvy\ncSqRWIIfvn6KvR3ddA9NlHReuK6SXdc2s+PidXl39ZoDMAyj1FjRjmC+Ix+nMhSM8vhLJ3n8pZMM\nhyZGQF53fj27tjezdVNt3jX95gAMwyhVVpwjSD/9j4bjhBfw9A9wYjDojIA81DdJAfSWLevZ2dbM\n+WtzK4CmsRyAYRilzopxBOf69K+q7Hmxm2+8eGKS/n+V38OtV2/kE9dsZE3lzCMgZ8IcgGEYy4Vl\n7QgK8fSfSCo/OXKar/30GF1Z8X+3S6jwufn9D1zCDRevzft+AZ8jBW0OwDCM5cKydATn+vQPTu/A\n917r5ZGOHvpGJkZA+j0u6sp9VJa5CceTfP1n7/DowR56R0JsmKOxrNznDIOxHIBhGMuNZeMIHOXQ\nxLw1/6dyZizCYwd7ePLlXsZSDWSCMwNgXZWPcu/ER5JIJuk5G6GpVqn2exgYj/DAM0e4l4syzqCy\nzEONTQMzDGMZsywcQSKpdA2G5qX5P5XO02Ps7ejm6Tf6iad2ET6Piw9d3sht25r4n08dYWA8Muma\nM+NRPC7JiMGlNWkebu/ilsvWU1vu5adHzthwE8MwljXLwhHEk7ogJ6CqdBwfYk97N+3HhzLHawOO\nAuitV2+kptxR9ZxJ+C2eUBqrsxLEAhU+N2dGw6yrKrPhJoZhrAiWhSOYL7FEkh+/eZo97V10np5Q\nAN1UX87O7c184LLpCqDZwm9poTmv2+XIT4gzEtLjEkKxBC31Tvlooech2+hEwzCWghXlCMbCcb79\nykkePdjDmbEJBdCrm2vY2dbM9a1rZlUABccZZCeCX3hnkP/9zNvEE0nKfZ5pUsWFHG5iuwvDMJaK\nFeEI+kbCPHqgm++80kcoVUbqErjp4nXsamvhksb8FECzqfR7uG17M43V/lmligs53KSQuwvbWRiG\nMR+K5ghEpAX4J2A9oMCDqvqAiNQD3wQ2A8eAXao6NNt95uLNvlH2tHfx7FsTCqABr5uPXNnIbdua\naazJTwE0m0q/h9qALxM6mmu6WfZwk3giyamRCLFkEq9L2He4f16Lb6F2F7azMAxjvhRzRxAH/r2q\nHhCRKqBDRJ4Cfg14WlW/IiL3AfcBX8z3pklVnu8cYG97Ny93TyiArqn0cdvWJj561UYq/fP/a1WW\neagt903LHcxFerjJ/d8/zLGBIF630FwbIJbUeS++hdpdFDpvYRjGyqdojkBVe4He1PejIvIG0AR8\nHNiROu0hYB95OIJoPOkogLZ3TeoAbl1Xwa62Ft53yTq8eSqAZrMQB5DNji0N7N7fyeY15ZMW8fku\nvoUanWhD2Q3DmC+LkiMQkc3AVuAXwPqUkwDowwkdzcnAWIQ7/+55hoITCqBt59Wxq62Z7efV5a0A\nmj8kIgcAAAxOSURBVM3UENC5UIjFt1CjE20ou2EY86XojkBEKoFvAZ9X1ZHsRVtVVURm1IgQkbuA\nuwB8jRfiC8bwuIRbLm3g9u3NXLCuciG2pHYA3gXtHmajUIvvXPmIfLGh7IZhzJeiOgIR8eI4gf+j\nqo+mDp8SkQ2q2isiG4D+ma5V1QeBBwECGy/WO65t4ZNbm1hXlb8CaJYdVPk91Aa8eQ+QySZXFU4p\nLb42lN0wjPlStOH14jz6PwQMqurns47/N2AgK1lcr6p/MNe9Lr96qz75o58sxAaq/R5qFugAYHIV\nTvYi/+VbL5+0uKadhS2+hmGUCqUwvP49wL8FXhWRl1LH/hD4CrBHRD4LHAd25brRXE1gs51f5XeS\nwG7X/PMH2eRbhVOIsI5hGMZSUMyqoedwhD1n4pZivKdLhOqAl5qA95wdQBqrwjEMY6WzIjqL3S6h\n2u84AFeBHECalVyFYx3IhmEAFK50Zglwu4T6Ch8tdeXUVfgK7gTASQTHEkowGk/NRIiviCqcdO6j\nfzQ8qQN53+EZc/eGYaxglqUj8LhcrKkoo6WunNry4jiANDu2NPDlWy+nocrPcChGQ5V/WqJ4OZKd\n+xBxvnrdwu79nUttmmEYi8yyCg15XC5qyr1U+z0LaiJbKCsxEWy5D8Mw0iwLRyDA2qoyqsoW1wGs\nZFZy7sMwjPmxLEJDPo+Lar/XnEABWam5D8Mw5s+y2BEUCquSmcA6kA3DSLNqHIHp9E9nJeY+DMOY\nP8siNFQIrErGMAxjZlaNI+gaChLwuicdsyoZwzCMFRwampoPqCpzhs9blYxhGMZkVqQjmCkfMByK\nZYSPlloq2jAMo5RYkY5gJsVQAK9LqKsosyoZwzCMLFakI5ita3Y4FOP7v3f9ElllGIZRmqzIZHFL\nXTmhWGLSMcsHGIZhzMyKdATWNWsYhpE/K9IRrFTFUMMwjGKwInMEYF2zhmEY+bIidwSGYRhG/pgj\nMAzDWOWYIzAMw1jlmCMwDMNY5ZgjMAzDWOWIqi61DTkRkdPA8SU0YS1wZgnfPxelbh+YjYWg1O2D\n0rex1O2Dwtp4nqquy3XSsnAES42ItKtq21LbMRulbh+YjYWg1O2D0rex1O2DpbHRQkOGYRirHHME\nhmEYqxxzBPnx4FIbkINStw/MxkJQ6vZB6dtY6vbBEthoOQLDMIxVju0IDMMwVjnmCLIQkX8UkX4R\neS3rWL2IPCUiR1Jf65bYxhYR+bGIvC4ih0Tk3lKyU0T8IvKCiLycsu9PS8m+Kba6ReSgiHy7FG0U\nkWMi8qqIvCQi7aVmo4jUisgjInJYRN4QkXeXmH2XpD679J8REfl8idn4e6l/J6+JyDdS/34W3T5z\nBJP5OvChKcfuA55W1YuAp1Ovl5I48O9V9TLgeuC3ReQySsfOCHCzql4NXAN8SESuLyH7srkXeCPr\ndSna+D5VvSarnLCUbHwA+L6qbgGuxvksS8Y+VX0z9dldA2wHgsBjpWKjiDQB9wBtqnoF4AbuWBL7\nVNX+ZP0BNgOvZb1+E9iQ+n4D8OZS2zjF3seBD5SinUA5cAB4V6nZBzSn/pHdDHy7FH/XwDFg7ZRj\nJWEjUAO8QyrPWGr2zWDvB4GflpKNQBPQBdTjjAT4dsrORbfPdgS5Wa+qvanv+4D1S2lMNiKyGdgK\n/IISsjMVcnkJ6AeeUtWSsi/F/wL+AEhmHSs1GxX4kYh0iMhdqWOlYuP5wGnga6nw2t+LSEUJ2TeV\nO4BvpL4vCRtVtQf4K+AE0AsMq+oPl8I+cwTzQB0XXRJlViJSCXwL+LyqjmT/bKntVNWEOtvxZuA6\nEbliys+X1D4R+SjQr6ods52z1DamuOH/b+9sY6yqrjD8vAIqoNVCSSp+TcFatcVa2lTFtgG0TewP\nraJBBYFC0jQlTTs/rVVjgon40dra2B/aaiImNcgIVpsqBcdQIRErOAy0FT+g2ig2ptiOorXD8sda\nA5ub8c5IuNzD3PUkJ+yzz9pnv+dwOWvvdThrx328EA8BfqM82GSNw4HJwK/N7EvAO9SEMCpyD5F0\nOHARsLT2WDM1Ruz/YtypjgdGS5pd2hwsfekIBmaHpOMA4s83m6wHSSNwJ/CAmXVEdeV0mtlO4En8\nvUuV9J0HXCRpG/A7YLqkJVRLY9+IETN7E49tf5XqaHwNeC1mewAP4Y6hKvpKLgSeM7MdsV8VjRcA\nr5jZv8zsA6ADmNIMfekIBuYRYG6U5+Ix+aYhScBvgL+a2c+KQ5XQKWmcpGOjPBJ/f/G3qugDMLNr\nzOwEM2vDQwarzWw2FdIoabSko/vKeOy4m4poNLM3gFclfS6qzge2UBF9NVzJ3rAQVEfjP4BzJI2K\nf9fn4y/cD76+ZrwkqeqG/1heBz7ARzwLgLH4S8WtwJ+AMU3W+DV8qtgFbIzt21XRCZwJbAh93cD1\nUV8Jff3oncrel8WV0QhMAJ6PbTNwbQU1ngU8G3/Xy4FPVklfaBwNvAUcU9RVRiNwIz5Q6gbuB45o\nhr78sjhJkqTFydBQkiRJi5OOIEmSpMVJR5AkSdLipCNIkiRpcdIRJEmStDjpCJIkSVqcdARJQ5B0\ngqQVkUr3JUm/iE/9kTRP0q+arbEWST2DsOlLDd0l6SlJJxfHemvSHrfFx0IPRJtuSX+O9CBIGhnn\nGNbAa2qTdFWx/7HuvaTbJE1vjLqkKqQjSA448ZVkB7DcPJXuqcBRwE0N7HN4o87dD9PM7EygE/hp\nUb/LIu1xbNvwVNc7zGySearhBfgHiwDzgQ4z622g1jbgqoGM6nAn1UjHnTSQdARJI5gOvGdm94In\noQPagfmSRoXNiZI6Y8ZwA+xJq/CYfFGbbkkzo/7LMXL+i6THizwsnZLukC/acq2k7ZIOK871qqQR\nkiZK+mO0XyPptLD5jKR1MVpftB/XuQ5PJVyP44B/9u2Y58h/P3ZnEekDJE2Na1wh6WVJN0uaJV/k\nZ5OkiWHXJml1zEhWSTop6u+T9EtJa6P9ZdHHzcDXY4bSHnXj435slXRLtB8W5+iO/tpD73ZgrKRP\n78f9SQ4Vmvn5d25Dc8MX2/h5P/Ub8BQU8/BUHmOBkfjn9V8BZgB3F/bHACOAtcC4qJsJ/DbKncBd\nhf0KfLTeZ3dPlFcBn43y2XhuIfCcLnOivBDoGcS1bSPWCMBTWX+vONbL3rQfD0fdWXjSsHXAokLH\n4cAbRdupwE7ccRyBO48b49iPgDui/HtgbpTn47Mu8EWVluKDuzOAF4vzPlr0Mw94Oe7tkcB24ER8\n4ZaVhd2xRfluYEazf1e5NW47mNPpJClZaWZvAUjqwHMo/QG4XdJi/OG1Rp7C+gvASo84MQx3In08\nWFOeiWc8vQK4K+LxU4Cl0R78QQuehXRGlO8HFg9S+5OSxgA9wHVF/S7ztNF7MLONkibgSeMuANZL\nOhd4G3/wl6y3yEMv6SXgiajfBEyL8rnApYXmW4r2y81sN7BFUr0c9qvM7O3oZwtwMp7PaIKkO4HH\nir7BHdn4OudLDnEyNJQ0gi34CHMPkj4BnAS8GFW1Sa7MzF7AUxlvAhZJuh4QsNn2xt0nmdm3inbv\nFOVH8KUxx0T/q/Hf+E7bN3Z/etnvflzfNPzhuRFPGlYXM+sxsw4z+wGwBE8SuAsfkZe8X5R3F/u7\nYVCDtrK9PtJqX7teYLiZ/RtfbrIT+D5wT2FzZOhNhijpCJJGsAoYJWkOePwZuB24z8zeDZtvyhfp\nHgl8B3ha0njgXTNbAtyKO4W/A+NiFE3E/D/fX6dm1gOsx9fSfdR8gZz/AK9IujzaS9IXo8nT+MwB\nPF4/aMzs/8CPgTnhePpF0nmKxcfl/2vqDGB7PHiHSap1BgOxtkbzmgHs/wscPdBJJX0KOMzMluEv\nwCcXh0/Fw3fJECUdQXLAMTMDLgEul7QVeAF4D/hJYfYMvrhOF7DMzJ4FJgHPyJe5vAFYZGb/Ay4D\nFkt6Hh+FT6nT/YPAbPYNGc0CFkT7zfiqUOCx94WSNjHwS9/+rvN1PHX5wjpmE4Gnoo8NeNrmZXHs\nCTwk9nH4IfBdSV3A1fg11KML6I0X8O117I4HOuPeLwGugT2LIJ0SupMhSqahTpImIWky0G5mVzdb\ny0ch6RJgspldN6BxcsiSM4IkaRJm9hz+4rlhH5QdAIbjYb1kCJMzgiRJkhYnZwRJkiQtTjqCJEmS\nFicdQZIkSYuTjiBJkqTFSUeQJEnS4nwIFhPZjkJroIYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x115ad1400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "RFR = predictive_statistics.RandomForestRegressor(X_train, y_train, X_test, y_test, outcome =' RFS(months)')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.0  Survival Length (`Survival_length`, Continous, months)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = df.Survival_length.values / 30\n",
    "\n",
    "#split\n",
    "X_train, X_test, y_train, y_test = predictive_statistics.split_data(X, y, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.795\n",
      "Model:                            OLS   Adj. R-squared:                  0.775\n",
      "Method:                 Least Squares   F-statistic:                     41.39\n",
      "Date:                Tue, 20 Jun 2017   Prob (F-statistic):           2.78e-32\n",
      "Time:                        10:03:09   Log-Likelihood:                -532.02\n",
      "No. Observations:                 117   AIC:                             1084.\n",
      "Df Residuals:                     107   BIC:                             1112.\n",
      "Df Model:                          10                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "x1            25.8238      5.081      5.083      0.000      15.752      35.896\n",
      "x2            21.5604      4.136      5.213      0.000      13.361      29.759\n",
      "x3            11.1381     14.718      0.757      0.451     -18.039      40.315\n",
      "x4             6.6792      7.110      0.939      0.350      -7.416      20.775\n",
      "x5            13.6270     15.858      0.859      0.392     -17.810      45.064\n",
      "x6             1.0752      2.361      0.455      0.650      -3.605       5.755\n",
      "x7             8.8649      4.917      1.803      0.074      -0.882      18.612\n",
      "x8            -3.8696      4.913     -0.788      0.433     -13.610       5.871\n",
      "x9            -4.7785      3.679     -1.299      0.197     -12.072       2.515\n",
      "x10            1.9332      3.156      0.613      0.541      -4.324       8.190\n",
      "==============================================================================\n",
      "Omnibus:                       13.355   Durbin-Watson:                   1.709\n",
      "Prob(Omnibus):                  0.001   Jarque-Bera (JB):               14.827\n",
      "Skew:                           0.724   Prob(JB):                     0.000603\n",
      "Kurtosis:                       3.973   Cond. No.                         15.7\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "================================================================================\n",
      "The median absolute error for testing data set of Survival_length (months) is: 15.819\n",
      "================================================================================\n"
     ]
    }
   ],
   "source": [
    "# # LSQ\n",
    "predictive_statistics.lsq(X_train, y_train, X_test, y_test, outcome ='Survival_length (months)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "================================================================================\n",
      "The median absolute error for testing data set of Survival length (months) is: 102.682\n",
      "================================================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEKCAYAAADTgGjXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRAzDMIySMSViGIZhlMz/B06/AC+FkpmVAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x115e798d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictive_statistics.svr(X_train, y_train, X_test, y_test, outcome ='Survival length (months)');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "================================================================================\n",
      "The median absolute error for testing data set of Survival length (months) is: 9.136\n",
      "================================================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEKCAYAAAAB0GKPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11608e588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictive_statistics.ElasticNet(X_train, y_train, X_test, y_test, outcome = 'Survival length (months)');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "================================================================================\n",
      "The median absolute error for testing data set of Survival length (months) is: 9.254\n",
      "================================================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEKCAYAAAAfGVI8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x115ff1cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictive_statistics.RandomForestRegressor(X_train, y_train, X_test, y_test, outcome = 'Survival length (months)');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}