LassoRegression_withInteraction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sb\n",
    "sb.set()\n",
    "color = sb.color_palette()\n",
    "import matplotlib as mpl\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "import pickle\n",
    "\n",
    "from sklearn import preprocessing as pp \n",
    "#import the entire linear model\n",
    "from sklearn import linear_model\n",
    "from sklearn.linear_model import Lasso\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import r2_score\n",
    "from scipy.stats import pearsonr \n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.linear_model import LassoCV\n",
    "#from sklearn.preprocessing import PolynomialFeatures\n",
    "\n",
    "from scipy.stats import wilcoxon"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_set = pd.read_pickle(\"./feature_set.pkl\")\n",
    "turnOverRate = pd.read_pickle(\"./turnOverRate.pkl\")\n",
    "LocationCodeComprehensive = pd.read_pickle(\"./LocationCodeComprehensive.pkl\")\n",
    "location_pay = pd.read_pickle(\"./location_pay.pkl\")\n",
    "location_sd = pd.read_pickle(\"./location_sdS_df.pkl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      5015\n",
       "1      5020\n",
       "2      5021\n",
       "3      5022\n",
       "4      5023\n",
       "       ... \n",
       "252    5946\n",
       "253    5947\n",
       "254    5948\n",
       "255    5906\n",
       "256    5907\n",
       "Length: 257, dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LocationCodeComprehensive.rename(columns = {'Unnamed: 2':'location code'}, inplace = True) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>2268</th>\n",
       "      <th>2269</th>\n",
       "      <th>2270</th>\n",
       "      <th>2271</th>\n",
       "      <th>2272</th>\n",
       "      <th>2273</th>\n",
       "      <th>2274</th>\n",
       "      <th>2275</th>\n",
       "      <th>2276</th>\n",
       "      <th>2277</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.496637</td>\n",
       "      <td>0.084248</td>\n",
       "      <td>0.138599</td>\n",
       "      <td>0.220066</td>\n",
       "      <td>1.221343</td>\n",
       "      <td>1.139475</td>\n",
       "      <td>1.125294</td>\n",
       "      <td>0.994916</td>\n",
       "      <td>0.274221</td>\n",
       "      <td>1.229335</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.244796</td>\n",
       "      <td>1.076369</td>\n",
       "      <td>-0.684921</td>\n",
       "      <td>0.604777</td>\n",
       "      <td>-0.961780</td>\n",
       "      <td>0.612005</td>\n",
       "      <td>-0.540393</td>\n",
       "      <td>-0.529198</td>\n",
       "      <td>0.467275</td>\n",
       "      <td>-0.297339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.549661</td>\n",
       "      <td>-0.119342</td>\n",
       "      <td>-1.306784</td>\n",
       "      <td>-0.972549</td>\n",
       "      <td>-1.140865</td>\n",
       "      <td>-1.123760</td>\n",
       "      <td>-1.501034</td>\n",
       "      <td>-1.556083</td>\n",
       "      <td>-0.941953</td>\n",
       "      <td>-0.617944</td>\n",
       "      <td>...</td>\n",
       "      <td>1.569697</td>\n",
       "      <td>0.607361</td>\n",
       "      <td>0.814480</td>\n",
       "      <td>-0.905170</td>\n",
       "      <td>0.872752</td>\n",
       "      <td>1.170375</td>\n",
       "      <td>-1.300692</td>\n",
       "      <td>0.452851</td>\n",
       "      <td>-0.503275</td>\n",
       "      <td>-0.674900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.577287</td>\n",
       "      <td>-0.276289</td>\n",
       "      <td>-0.402631</td>\n",
       "      <td>0.241940</td>\n",
       "      <td>-0.454408</td>\n",
       "      <td>-0.379317</td>\n",
       "      <td>-0.944135</td>\n",
       "      <td>-0.203789</td>\n",
       "      <td>-0.442965</td>\n",
       "      <td>-0.625386</td>\n",
       "      <td>...</td>\n",
       "      <td>0.405405</td>\n",
       "      <td>-0.525482</td>\n",
       "      <td>0.104186</td>\n",
       "      <td>-0.585362</td>\n",
       "      <td>-0.354012</td>\n",
       "      <td>0.070189</td>\n",
       "      <td>-0.394353</td>\n",
       "      <td>-0.090979</td>\n",
       "      <td>0.511157</td>\n",
       "      <td>-0.101346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.482090</td>\n",
       "      <td>-0.893064</td>\n",
       "      <td>0.082008</td>\n",
       "      <td>0.163632</td>\n",
       "      <td>1.103529</td>\n",
       "      <td>0.281656</td>\n",
       "      <td>0.264472</td>\n",
       "      <td>0.419771</td>\n",
       "      <td>0.451285</td>\n",
       "      <td>-0.343488</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.106702</td>\n",
       "      <td>-0.063108</td>\n",
       "      <td>-0.064710</td>\n",
       "      <td>-0.209771</td>\n",
       "      <td>0.571254</td>\n",
       "      <td>0.585751</td>\n",
       "      <td>1.898836</td>\n",
       "      <td>0.346439</td>\n",
       "      <td>1.123058</td>\n",
       "      <td>1.151557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.994165</td>\n",
       "      <td>-1.591461</td>\n",
       "      <td>-0.656436</td>\n",
       "      <td>0.042593</td>\n",
       "      <td>-0.060927</td>\n",
       "      <td>-0.986578</td>\n",
       "      <td>-1.008201</td>\n",
       "      <td>-0.564969</td>\n",
       "      <td>-0.913683</td>\n",
       "      <td>-0.538270</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.069755</td>\n",
       "      <td>0.002773</td>\n",
       "      <td>-0.052163</td>\n",
       "      <td>0.002675</td>\n",
       "      <td>-0.003844</td>\n",
       "      <td>0.072313</td>\n",
       "      <td>-0.003708</td>\n",
       "      <td>-0.002874</td>\n",
       "      <td>0.000147</td>\n",
       "      <td>-0.002773</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 2278 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       0         1         2         3         4         5         6     \\\n",
       "0  1.496637  0.084248  0.138599  0.220066  1.221343  1.139475  1.125294   \n",
       "1 -0.549661 -0.119342 -1.306784 -0.972549 -1.140865 -1.123760 -1.501034   \n",
       "2  0.577287 -0.276289 -0.402631  0.241940 -0.454408 -0.379317 -0.944135   \n",
       "3 -0.482090 -0.893064  0.082008  0.163632  1.103529  0.281656  0.264472   \n",
       "4 -0.994165 -1.591461 -0.656436  0.042593 -0.060927 -0.986578 -1.008201   \n",
       "\n",
       "       7         8         9     ...      2268      2269      2270      2271  \\\n",
       "0  0.994916  0.274221  1.229335  ... -1.244796  1.076369 -0.684921  0.604777   \n",
       "1 -1.556083 -0.941953 -0.617944  ...  1.569697  0.607361  0.814480 -0.905170   \n",
       "2 -0.203789 -0.442965 -0.625386  ...  0.405405 -0.525482  0.104186 -0.585362   \n",
       "3  0.419771  0.451285 -0.343488  ... -0.106702 -0.063108 -0.064710 -0.209771   \n",
       "4 -0.564969 -0.913683 -0.538270  ... -0.069755  0.002773 -0.052163  0.002675   \n",
       "\n",
       "       2272      2273      2274      2275      2276      2277  \n",
       "0 -0.961780  0.612005 -0.540393 -0.529198  0.467275 -0.297339  \n",
       "1  0.872752  1.170375 -1.300692  0.452851 -0.503275 -0.674900  \n",
       "2 -0.354012  0.070189 -0.394353 -0.090979  0.511157 -0.101346  \n",
       "3  0.571254  0.585751  1.898836  0.346439  1.123058  1.151557  \n",
       "4 -0.003844  0.072313 -0.003708 -0.002874  0.000147 -0.002773  \n",
       "\n",
       "[5 rows x 2278 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_features = pd.concat([feature_set,location_pay,location_sd], axis = 1)\n",
    "\n",
    "\n",
    "poly = pp.PolynomialFeatures(interaction_only=True,include_bias = False)\n",
    "all_features_interaction=poly.fit_transform(all_features)\n",
    "all_features_interaction=pd.DataFrame(data=all_features_interaction)\n",
    "all_features_interaction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "XCV_train, XCV_test, YCV_train, YCV_test = train_test_split(xCV, yCV, test_size=0.2)\n",
    "\n",
    "reg = LassoCV(cv=5, max_iter = 5000).fit(XCV_train, YCV_train)\n",
    "\n",
    "YpredLassoCV2 = reg.predict(XCV_test)\n",
    "test_r_squared2 = r2_score(YCV_test, YpredLassoCV2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.005576188808204874"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_r_squared2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01589966003058707, tolerance: 0.015503297942400502\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.021924468450186363, tolerance: 0.017902453002246285\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.019854360871348842, tolerance: 0.016469981678627375\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.022092264129845662, tolerance: 0.017658948057819927\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.022648308097707126, tolerance: 0.0162276693478065\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.02096270906546327, tolerance: 0.0162276693478065\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.019844708111470766, tolerance: 0.014189426426457838\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.015937409343401265, tolerance: 0.014189426426457838\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0186674556181039, tolerance: 0.018123998137960663\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.019838142926766555, tolerance: 0.018123998137960663\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.017996401649527427, tolerance: 0.01697760761867041\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0182254757775151, tolerance: 0.017416971842239524\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.025440669199157995, tolerance: 0.01825281146340277\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0177405971947433, tolerance: 0.016834257829570022\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.02185443394948583, tolerance: 0.01619711792316609\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.025865779013414025, tolerance: 0.015973819191583867\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01755383220729989, tolerance: 0.016126162706942546\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.025078464937892786, tolerance: 0.016126162706942546\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01991292858952315, tolerance: 0.01707507368771651\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.019080310078666507, tolerance: 0.016201551967609765\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.019349049187821432, tolerance: 0.01659465797249663\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01736093433294328, tolerance: 0.01659465797249663\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.018431100002887404, tolerance: 0.017040668330655238\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01649471591292516, tolerance: 0.015816048439408856\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.02407073362823442, tolerance: 0.0176074783331806\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.021569347769297087, tolerance: 0.017234318379969433\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.018346898529823352, tolerance: 0.017234318379969433\n",
      "  positive)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.021628999973794555, tolerance: 0.01568154768714451\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.015997286267929733, tolerance: 0.014174070621682194\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01910190600897199, tolerance: 0.017036830847278014\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.017858073427351157, tolerance: 0.017328943806471896\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.016372535067002758, tolerance: 0.014692537279637716\n",
      "  positive)\n",
      "/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/linear_model/_coordinate_descent.py:476: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01774507402669212, tolerance: 0.016422669401691587\n",
      "  positive)\n"
     ]
    }
   ],
   "source": [
    "# Try running the cross validation Lasso for different train test split and get the\n",
    "#mean r-squared, alpha, and weights.\n",
    "\n",
    "\n",
    "xCV = all_features_interaction\n",
    "yCV = turnOverRate #Turn over\n",
    "\n",
    "train_number = 200\n",
    "#Run this part for 1000 times and get the values for r-squared, alpha, and the weights.\n",
    "#reg.alpha_, reg.coef_, r2score.\n",
    "\n",
    "#initialize\n",
    "test_r_squared = np.zeros((train_number,1))\n",
    "lasso_alpha = np.zeros((train_number,1))\n",
    "lasso_coef = np.zeros((train_number,all_features_interaction.shape[1]))\n",
    "\n",
    "for trialNo in range(train_number):\n",
    "    \n",
    "    #split 80-20\n",
    "    XCV_train, XCV_test, YCV_train, YCV_test = train_test_split(xCV, yCV, test_size=0.2)\n",
    "\n",
    "    reg = LassoCV(cv=5, max_iter = 5000).fit(XCV_train, YCV_train)\n",
    "\n",
    "    YpredLassoCV = reg.predict(XCV_test)\n",
    "    test_r_squared[trialNo] = r2_score(YCV_test, YpredLassoCV)\n",
    "    lasso_alpha[trialNo] = reg.alpha_\n",
    "    lasso_coef[trialNo,:] = reg.coef_ \n",
    "    #print(\"r^2 on test data : %f\" % r2ScoreLassoCV)\n",
    "\n",
    "    #ActualPredictionCV = pd.DataFrame({'Actual': YCV_test, 'Predicted': YpredLassoCV})\n",
    "\n",
    "    #sb.lmplot(x='Predicted', y='Actual', data=ActualPredictionCV)\n",
    "with open('lasso_interaction_output.pickle', 'wb') as f:\n",
    "    pickle.dump([test_r_squared, lasso_alpha, lasso_coef], f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_r_squared_df=pd.DataFrame(data=test_r_squared)\n",
    "test_r_squared_df.rename(columns = {0:'coefficient of determination'}, inplace = True) \n",
    "\n",
    "sb.distplot(test_r_squared_df,axlabel='coefficent of determination')\n",
    "sb.despine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "#with open('lasso_output.pickle', 'wb') as f:\n",
    "#    pickle.dump([test_r_squared, lasso_alpha, lasso_coef], f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Getting back the objects:\n",
    "with open('lasso_output.pickle', 'rb') as f:  \n",
    "    test_r_squared, lasso_alpha, lasso_coef = pickle.load(f)\n",
    "\n",
    "with open('original_values.pickle', 'rb') as f:  \n",
    "    TurnOver2, dataX2 = pickle.load(f)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "test_r_squared_df=pd.DataFrame(data=test_r_squared)\n",
    "test_r_squared_df.rename(columns = {0:'coefficient of determination'}, inplace = True) \n",
    "\n",
    "sb.distplot(test_r_squared_df,axlabel='coefficent of determination')\n",
    "sb.despine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.52570745e-57]\n"
     ]
    }
   ],
   "source": [
    "#One sample t-test to show the significance.\n",
    "from scipy.stats import ttest_1samp\n",
    "\n",
    "tset, pval = ttest_1samp(test_r_squared, 0)\n",
    "print(pval)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.51059490731279e-61\n"
     ]
    }
   ],
   "source": [
    "sum_ranks, pval_w = wilcoxon(test_r_squared[:,0])\n",
    "print(pval_w)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "lasso_alpha_df = pd.DataFrame(data=lasso_alpha)\n",
    "lasso_coef_df = pd.DataFrame(data=lasso_coef)\n",
    "\n",
    "sb.distplot(lasso_alpha_df)\n",
    "sb.despine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_coef=lasso_coef_df.mean()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'average weight for each feature')"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.plot(mean_coef,linestyle='none',marker='o',markersize=8,color='red') \n",
    "plt.xlabel('feature number')\n",
    "#plt.ylabel('average weight for each feature')\n",
    "plt.title('average weight for each feature')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_set.to_pickle(\"./feature_set.pkl\")\n",
    "\n",
    "lasso_r_squared_1000 = pd.DataFrame(data=test_r_squared)\n",
    "lasso_alpha_1000 = pd.DataFrame(data=lasso_alpha)\n",
    "lasso_coef_1000 = pd.DataFrame(data=lasso_coef)\n",
    "\n",
    "lasso_r_squared_1000.to_pickle(\"./lasso_r_squared_1000.pkl\")\n",
    "lasso_alpha_1000.to_pickle(\"./lasso_alpha_1000.pkl\")\n",
    "lasso_coef_1000.to_pickle(\"./lasso_coef_1000.pkl\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_set = pd.read_pickle(\"./feature_set.pkl\")\n",
    "\n",
    "lasso_r_squared_df = pd.read_pickle(\"./lasso_r_squared_1000.pkl\")\n",
    "lasso_alpha_df = pd.read_pickle(\"./lasso_alpha_1000.pkl\")\n",
    "lasso_coef_df = pd.read_pickle(\"./lasso_coef_1000.pkl\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "abs_mean_coef=abs(mean_coef)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "55    0.157820\n",
       "62    0.127956\n",
       "21    0.086186\n",
       "64    0.075024\n",
       "24    0.042282\n",
       "22    0.027030\n",
       "66    0.020481\n",
       "50    0.011860\n",
       "51    0.010044\n",
       "19    0.008916\n",
       "dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_coef=abs_mean_coef.sort_values(ascending=False)\n",
    "top_coef.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_features = pd.concat([feature_set,location_pay,location_sd], axis = 1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Choose the features with top n coefficients \n",
    "\n",
    "features_number = 3\n",
    "#get the top n features and construct a new feature.\n",
    "top_coef=abs_mean_coef.sort_values(ascending=False)\n",
    "\n",
    "selected_features = all_features.iloc[:,top_coef.index[0]]\n",
    "\n",
    "for features in range(features_number-1):\n",
    "    selected_features = pd.concat([selected_features, all_features.iloc[:,top_coef.index[features+1]]], axis = 1)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>55</th>\n",
       "      <th>1</th>\n",
       "      <th>21</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.467229</td>\n",
       "      <td>-1.180298</td>\n",
       "      <td>1.103600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.504403</td>\n",
       "      <td>-1.045167</td>\n",
       "      <td>-0.092121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.888779</td>\n",
       "      <td>0.775737</td>\n",
       "      <td>-0.562840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.026882</td>\n",
       "      <td>-0.108571</td>\n",
       "      <td>-0.092121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.679443</td>\n",
       "      <td>-0.224316</td>\n",
       "      <td>-1.280074</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         55        1         21\n",
       "0  0.467229 -1.180298  1.103600\n",
       "1 -1.504403 -1.045167 -0.092121\n",
       "2  0.888779  0.775737 -0.562840\n",
       "3 -0.026882 -0.108571 -0.092121\n",
       "4  0.679443 -0.224316 -1.280074"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_features.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Look at the distribution of the turnover rate (in the normalized space)\n",
    "#And see if there are natural boundaries for the classes.\n",
    "sb.distplot(turnOverRate)\n",
    "sb.despine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [1.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [2.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [1.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.],\n",
       "       [0.]])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "turn_over_label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "#looks like we can split between -2 and -1, but this is very unbalanced classes\n",
    "#Make labels as 0, 1, and 2.\n",
    "turn_over_label=np.zeros((turnOverRate.shape[0],1))\n",
    "turn_over_label[turnOverRate<-2] = 2\n",
    "idx = (turnOverRate>=-2)&(turnOverRate<=-1)\n",
    "turn_over_label[idx] = 1\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "257"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "turnOverRate.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Try once and see how it looks.\n",
    "X_train, X_test, y_train, y_test = train_test_split(selected_features, turn_over_label, test_size = 0.2,stratify=turn_over_label)\n",
    "\n",
    "#resample because groups 1 and 2 are much smaller than 0. Especially 2 is smaller\n",
    "sm = SMOTE()\n",
    "X_train_res, y_train_res = sm.fit_sample(X_train, y_train.ravel())\n",
    "\n",
    "clf = LogisticRegressionCV(cv=5).fit(X_train_res, y_train_res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_prob = clf.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "ename": "IndexError",
     "evalue": "too many indices for array",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-42-4039a3eb6c45>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmacro_roc_auc_ovo\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mroc_auc_score\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_prob\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmulti_class\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"ovo\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0maverage\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"macro\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/metrics/_ranking.py\u001b[0m in \u001b[0;36mroc_auc_score\u001b[0;34m(y_true, y_score, average, sample_weight, max_fpr, multi_class, labels)\u001b[0m\n\u001b[1;32m    381\u001b[0m             \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"multi_class must be in ('ovo', 'ovr')\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    382\u001b[0m         return _multiclass_roc_auc_score(y_true, y_score, labels,\n\u001b[0;32m--> 383\u001b[0;31m                                          multi_class, average, sample_weight)\n\u001b[0m\u001b[1;32m    384\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0my_type\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m\"binary\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    385\u001b[0m         \u001b[0mlabels\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munique\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/metrics/_ranking.py\u001b[0m in \u001b[0;36m_multiclass_roc_auc_score\u001b[0;34m(y_true, y_score, labels, multi_class, average, sample_weight)\u001b[0m\n\u001b[1;32m    487\u001b[0m         return _average_multiclass_ovo_score(_binary_roc_auc_score,\n\u001b[1;32m    488\u001b[0m                                              \u001b[0my_true_encoded\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 489\u001b[0;31m                                              y_score, average=average)\n\u001b[0m\u001b[1;32m    490\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    491\u001b[0m         \u001b[0;31m# ovr is same as multi-label\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/metrics/_base.py\u001b[0m in \u001b[0;36m_average_multiclass_ovo_score\u001b[0;34m(binary_metric, y_true, y_score, average)\u001b[0m\n\u001b[1;32m    196\u001b[0m         \u001b[0mb_true\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mb_mask\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mab_mask\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    197\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 198\u001b[0;31m         \u001b[0ma_true_score\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbinary_metric\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_score\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mab_mask\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ma\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    199\u001b[0m         \u001b[0mb_true_score\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbinary_metric\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_score\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mab_mask\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    200\u001b[0m         \u001b[0mpair_scores\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mix\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0ma_true_score\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mb_true_score\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m/\u001b[0m \u001b[0;36m2\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIndexError\u001b[0m: too many indices for array"
     ]
    }
   ],
   "source": [
    "macro_roc_auc_ovo = roc_auc_score(y_test, y_prob, multi_class=\"ovo\",average=\"macro\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.45410518225735624"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "macro_roc_auc_ovr = roc_auc_score(y_test, y_prob, multi_class=\"ovr\",\n",
    "                                  average=\"macro\")\n",
    "macro_roc_auc_ovr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Expected n_neighbors <= n_samples,  but n_samples = 5, n_neighbors = 6",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-22-e6ee4d5855bb>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     17\u001b[0m     \u001b[0;31m#resample because groups 1 and 2 are much smaller than 0. Especially 2 is smaller\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     18\u001b[0m     \u001b[0msm\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mSMOTE\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 19\u001b[0;31m     \u001b[0mX_train_res\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train_res\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msm\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_sample\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mravel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     20\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     21\u001b[0m     \u001b[0mclf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mLogisticRegressionCV\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcv\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_train_res\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train_res\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/imblearn/base.py\u001b[0m in \u001b[0;36mfit_resample\u001b[0;34m(self, X, y)\u001b[0m\n\u001b[1;32m     79\u001b[0m         )\n\u001b[1;32m     80\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 81\u001b[0;31m         \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_fit_resample\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     82\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     83\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_X_columns\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_y_name\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/imblearn/over_sampling/_smote.py\u001b[0m in \u001b[0;36m_fit_resample\u001b[0;34m(self, X, y)\u001b[0m\n\u001b[1;32m    725\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    726\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnn_k_\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_class\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 727\u001b[0;31m             \u001b[0mnns\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnn_k_\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkneighbors\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_class\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreturn_distance\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    728\u001b[0m             X_new, y_new = self._make_samples(\n\u001b[1;32m    729\u001b[0m                 \u001b[0mX_class\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mclass_sample\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_class\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnns\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mn_samples\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1.0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/home/akira/anaconda3/envs/unsupervisedLearning/lib/python3.6/site-packages/sklearn/neighbors/_base.py\u001b[0m in \u001b[0;36mkneighbors\u001b[0;34m(self, X, n_neighbors, return_distance)\u001b[0m\n\u001b[1;32m    615\u001b[0m                 \u001b[0;34m\"Expected n_neighbors <= n_samples, \"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    616\u001b[0m                 \u001b[0;34m\" but n_samples = %d, n_neighbors = %d\"\u001b[0m \u001b[0;34m%\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 617\u001b[0;31m                 \u001b[0;34m(\u001b[0m\u001b[0mn_samples_fit\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mn_neighbors\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    618\u001b[0m             )\n\u001b[1;32m    619\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Expected n_neighbors <= n_samples,  but n_samples = 5, n_neighbors = 6"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegressionCV\n",
    "from imblearn.over_sampling import SMOTE\n",
    "\n",
    "train_number = 1000\n",
    "#Run this part for 1000 times and get the values for r-squared, alpha, and the weights.\n",
    "#reg.alpha_, reg.coef_, r2score.\n",
    "\n",
    "#initialize\n",
    "logit_r_squared = np.zeros((train_number,1))\n",
    "logit_coef = np.zeros((3,7,train_number))\n",
    "\n",
    "for trialNo in range(train_number):\n",
    "    \n",
    "\n",
    "    X_train, X_test, y_train, y_test = train_test_sstratify=y,plit(selected_features, turn_over_label, test_size = 0.2,stratify=turn_over_label)\n",
    "\n",
    "    #resample because groups 1 and 2 are much smaller than 0. Especially 2 is smaller\n",
    "    sm = SMOTE()\n",
    "    X_train_res, y_train_res = sm.fit_sample(X_train, y_train.ravel())\n",
    "\n",
    "    clf = LogisticRegressionCV(cv=5).fit(X_train_res, y_train_res)\n",
    "    predicted_label = clf.predict(X_test)\n",
    "    \n",
    "    logit_r_squared[trialNo] = clf.score(X_test, y_test)\n",
    "    logit_coef[:,:,trialNo] = clf.coef_ \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.coef_.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.467229</td>\n",
       "      <td>-1.180298</td>\n",
       "      <td>1.103600</td>\n",
       "      <td>-0.551469</td>\n",
       "      <td>0.515634</td>\n",
       "      <td>-1.302577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.504403</td>\n",
       "      <td>-1.045167</td>\n",
       "      <td>-0.092121</td>\n",
       "      <td>1.572352</td>\n",
       "      <td>0.138588</td>\n",
       "      <td>0.096282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.888779</td>\n",
       "      <td>0.775737</td>\n",
       "      <td>-0.562840</td>\n",
       "      <td>0.689458</td>\n",
       "      <td>-0.500240</td>\n",
       "      <td>-0.436616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.026882</td>\n",
       "      <td>-0.108571</td>\n",
       "      <td>-0.092121</td>\n",
       "      <td>0.002919</td>\n",
       "      <td>0.002476</td>\n",
       "      <td>0.010002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.679443</td>\n",
       "      <td>-0.224316</td>\n",
       "      <td>-1.280074</td>\n",
       "      <td>-0.152410</td>\n",
       "      <td>-0.869737</td>\n",
       "      <td>0.287141</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3         4         5\n",
       "0  0.467229 -1.180298  1.103600 -0.551469  0.515634 -1.302577\n",
       "1 -1.504403 -1.045167 -0.092121  1.572352  0.138588  0.096282\n",
       "2  0.888779  0.775737 -0.562840  0.689458 -0.500240 -0.436616\n",
       "3 -0.026882 -0.108571 -0.092121  0.002919  0.002476  0.010002\n",
       "4  0.679443 -0.224316 -1.280074 -0.152410 -0.869737  0.287141"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "poly = pp.PolynomialFeatures(interaction_only=True,include_bias = False)\n",
    "selected_features_interaction=poly.fit_transform(selected_features)\n",
    "selected_features_interaction=pd.DataFrame(data=selected_features_interaction)\n",
    "selected_features_interaction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['x0', 'x1', 'x2', 'x0 x1', 'x0 x2', 'x1 x2']"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "poly.get_feature_names()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#make a model with interactions.\n",
    "#Run N times and take the average.\n",
    "#Try to use the shuffle split function => later\n",
    "x_linear = selected_features_interaction\n",
    "y_linear = turnOverRate #Turn over\n",
    "\n",
    "train_number = 450\n",
    "#Run this part for 1000 times and get the values for r-squared, alpha, and the weights.\n",
    "#reg.alpha_, reg.coef_, r2score.\n",
    "\n",
    "#initialize\n",
    "linear_r_squared = np.zeros((train_number,1))\n",
    "linear_coef = np.zeros((train_number,x_linear.shape[1]))\n",
    "\n",
    "for trialNo in range(train_number):\n",
    "    \n",
    "    #split 80-20\n",
    "    Xl_train, Xl_test, Yl_train, Yl_test = train_test_split(x_linear, y_linear, test_size=0.2)\n",
    "    \n",
    "\n",
    "    linear_reg = linear_model.LinearRegression()\n",
    "    linear_reg.fit(Xl_train,Yl_train)\n",
    "    \n",
    "    \n",
    "    Ypredl = linear_reg.predict(Xl_test)\n",
    "    \n",
    "    linear_r_squared[trialNo] = r2_score(Yl_test, Ypredl)\n",
    "    linear_coef[trialNo,:] = linear_reg.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "linear_r_squared_df=pd.DataFrame(data=linear_r_squared)\n",
    "#linear_r_squared_df.hist()\n",
    "sb.distplot(linear_r_squared_df)\n",
    "sb.despine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.72314810927932e-42\n"
     ]
    }
   ],
   "source": [
    "sum_ranks, pval_w = wilcoxon(linear_r_squared[:,0])\n",
    "print(pval_w)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'mean_linear_coef' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-38-342911561133>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmean_linear_coef\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'mean_linear_coef' is not defined"
     ]
    }
   ],
   "source": [
    "mean_linear_coef[5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fdd6331c2e8>]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "linear_coef_df = pd.DataFrame(data=linear_coef)\n",
    "\n",
    "mean_linear_coef=linear_coef_df.mean()\n",
    "\n",
    "plt.plot(mean_linear_coef,alpha=0.7,linestyle='none',marker='*',markersize=5,color='red',label=r'Lasso; $\\alpha = 0.05$',zorder=7) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "index = LocationCodeComprehensive==5015"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7fdd6328b278>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 449.55x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "PredictionMatrix=x_linear.multiply(mean_linear_coef)\n",
    "PredictionValue=PredictionMatrix.sum(axis=1)\n",
    "\n",
    "#Put together into a single data frame so we can plot easier\n",
    "frameMVP = { 'turnover rate': turnOverRate, 'Predicted turnover': PredictionValue, 'current location': index} \n",
    "dataForPlotMVP = pd.DataFrame(frameMVP)\n",
    "sb.lmplot(x='Predicted turnover', y='turnover rate', data=dataForPlotMVP, hue = 'current location')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fdd6328b2b0>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots()\n",
    "\n",
    "sb.regplot(x='Predicted turnover', y='turnover rate', data=dataForPlotMVP, ax = ax)\n",
    "\n",
    "frameMVP2 = { 'turnover rate': turnOverRate[index], 'Predicted turnover': PredictionValue[index]} \n",
    "dataForPlotMVP2 = pd.DataFrame(frameMVP2)\n",
    "sb.scatterplot(x='Predicted turnover', y='turnover rate', data=dataForPlotMVP2, color=sb.xkcd_rgb[\"pale red\"], ax = ax)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Getting back the objects:\n",
    "with open('scaler.pickle', 'rb') as f:  \n",
    "    scalerEmployeeResponse, scalerOccupancyTurnover = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "location = 5015"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0492875665990797\n",
      "0    0.965517\n",
      "Name: 55, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "TurnOver2=TurnOver2.reset_index(drop=True)\n",
    "dataX2=dataX2.reset_index(drop=True)\n",
    "base_turnover=TurnOver2[index]\n",
    "percent_turnover = 25\n",
    "new_turnover = base_turnover*(1-(50/100))\n",
    "log_new_turnover = np.log(new_turnover+0.1)\n",
    "scaled_turnover = (log_new_turnover-scalerOccupancyTurnover.mean_[1])/scalerOccupancyTurnover.scale_[1]\n",
    "current_turnover = turnOverRate[index]\n",
    "turnover_change = scaled_turnover-current_turnover\n",
    "#currently assumes we use the top component.\n",
    "#Use the dataframe selected_features_interaction because it's column is reset.\n",
    "feature_change = turnover_change/(mean_linear_coef[0]+mean_linear_coef[3]*selected_features_interaction.loc[LocationCodeComprehensive==location,1]+mean_linear_coef[4]*selected_features_interaction.loc[LocationCodeComprehensive==location,2])\n",
    "#bring it back to what it means in the raw data scale.\n",
    "new_feature = selected_features_interaction.loc[LocationCodeComprehensive==location,0]+feature_change\n",
    "new_feature_before_scale = (new_feature*scalerEmployeeResponse.scale_[55])+scalerEmployeeResponse.mean_[55]\n",
    "new_feature_before_transform = 1.1-np.exp(0.1-new_feature_before_scale)\n",
    "\n",
    "current_feature = dataX2.loc[LocationCodeComprehensive==location,55]\n",
    "feature_change = new_feature_before_transform - current_feature\n",
    "print(new_feature_before_transform[0])\n",
    "print(current_feature)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.467229</td>\n",
       "      <td>-1.180298</td>\n",
       "      <td>1.103600</td>\n",
       "      <td>-0.551469</td>\n",
       "      <td>0.515634</td>\n",
       "      <td>-1.302577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.504403</td>\n",
       "      <td>-1.045167</td>\n",
       "      <td>-0.092121</td>\n",
       "      <td>1.572352</td>\n",
       "      <td>0.138588</td>\n",
       "      <td>0.096282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.888779</td>\n",
       "      <td>0.775737</td>\n",
       "      <td>-0.562840</td>\n",
       "      <td>0.689458</td>\n",
       "      <td>-0.500240</td>\n",
       "      <td>-0.436616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.026882</td>\n",
       "      <td>-0.108571</td>\n",
       "      <td>-0.092121</td>\n",
       "      <td>0.002919</td>\n",
       "      <td>0.002476</td>\n",
       "      <td>0.010002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.679443</td>\n",
       "      <td>-0.224316</td>\n",
       "      <td>-1.280074</td>\n",
       "      <td>-0.152410</td>\n",
       "      <td>-0.869737</td>\n",
       "      <td>0.287141</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3         4         5\n",
       "0  0.467229 -1.180298  1.103600 -0.551469  0.515634 -1.302577\n",
       "1 -1.504403 -1.045167 -0.092121  1.572352  0.138588  0.096282\n",
       "2  0.888779  0.775737 -0.562840  0.689458 -0.500240 -0.436616\n",
       "3 -0.026882 -0.108571 -0.092121  0.002919  0.002476  0.010002\n",
       "4  0.679443 -0.224316 -1.280074 -0.152410 -0.869737  0.287141"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_features_interaction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.38014280864088534, 2.9262203153695e-10)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pearsonr(PredictionValue, turnOverRate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "#From Fabian Pedregosa-Izquierdo, f@bianp.net\n",
    "import numpy as np\n",
    "from scipy import stats, linalg\n",
    "\n",
    "def partial_corr(C):\n",
    "    \"\"\"\n",
    "    Returns the sample linear partial correlation coefficients between pairs of variables in C, controlling \n",
    "    for the remaining variables in C.\n",
    "    Parameters\n",
    "    ----------\n",
    "    C : array-like, shape (n, p)\n",
    "        Array with the different variables. Each column of C is taken as a variable\n",
    "    Returns\n",
    "    -------\n",
    "    P : array-like, shape (p, p)\n",
    "        P[i, j] contains the partial correlation of C[:, i] and C[:, j] controlling\n",
    "        for the remaining variables in C.\n",
    "    \"\"\"\n",
    "    \n",
    "    C = np.asarray(C)\n",
    "    p = C.shape[1]\n",
    "    P_corr = np.zeros((p, p), dtype=np.float)\n",
    "    for i in range(p):\n",
    "        P_corr[i, i] = 1\n",
    "        for j in range(i+1, p):\n",
    "            idx = np.ones(p, dtype=np.bool)\n",
    "            idx[i] = False\n",
    "            idx[j] = False\n",
    "            beta_i = linalg.lstsq(C[:, idx], C[:, j])[0]\n",
    "            beta_j = linalg.lstsq(C[:, idx], C[:, i])[0]\n",
    "\n",
    "            res_j = C[:, j] - C[:, idx].dot( beta_i)\n",
    "            res_i = C[:, i] - C[:, idx].dot(beta_j)\n",
    "            \n",
    "            corr = stats.pearsonr(res_i, res_j)[0]\n",
    "            P_corr[i, j] = corr\n",
    "            P_corr[j, i] = corr\n",
    "        \n",
    "    return P_corr\n",
    "\n",
    "turnOver_features = pd.concat([turnOverRate, selected_features], axis = 1)\n",
    "partial_corr_array = turnOver_features.to_numpy()\n",
    "\n",
    "# Calculate the partial correlation coefficients\n",
    "partial_corr_results = partial_corr(partial_corr_array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.285774</td>\n",
       "      <td>-0.180665</td>\n",
       "      <td>0.209384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.285774</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.116342</td>\n",
       "      <td>0.644236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.180665</td>\n",
       "      <td>0.116342</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.175639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.209384</td>\n",
       "      <td>0.644236</td>\n",
       "      <td>-0.175639</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3\n",
       "0  1.000000 -0.285774 -0.180665  0.209384\n",
       "1 -0.285774  1.000000  0.116342  0.644236\n",
       "2 -0.180665  0.116342  1.000000 -0.175639\n",
       "3  0.209384  0.644236 -0.175639  1.000000"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "partial_corr_results_df = pd.DataFrame(data=partial_corr_results)\n",
    "partial_corr_results_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb125a8b080>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAAFsCAYAAAApGMkVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nO3deXxTVf7G8U9a9q1A2UpRVj06songPiwqyCKKiDAoiyPiuOEKOC6AIiiDiozKUEcHZVEGQX6iwAiogLgNIqvbUZSthbIUapEBgSa/P24oLQRoSZOUw/P2lRfJvecm30T69PA9N4kvEAggIiJuiot1ASIiEjkKeRERhynkRUQcppAXEXGYQl5ExGEKeRERhxWLdQEiIqc6Y8xzwA1AHaCRtfabEGPigReB9kAAGGWtfe1E+8KlmbyISPjeBVoCG44z5magAXAWcAnwhDGmTj72hUUhLyISJmvtp9baTScY1gN41Vrrt9Zux/vFcGM+9oVF7RoRkWMwxlQEKobYlWmtzSzg3Z1J3pn+RuCMfOwLS7RDXp+hICL55Qvn4AM7fimMvHkSGHaM7U8Uwv1HXFRD/sCOX6L5cEVW8Sr1KFYiOdZlFAkH96dRO7FxrMsoEjZkrObVWr1iXUaR0D91SqxLOGQs8EaI7QWdxYM3O68NfBW8nXv2frx9YVG7RkTc5M8O+y6CLZmTCfRQpgP9jTEzgUSgC95i7Yn2hUULryLipoA//Es+GWNeNMakArWAD40x3wa3zzXGNA8Omwz8AvwEfAkMt9b+ko99YfFF86OGC6lHdspTu+YwtWsOU7vmsGC7Jrye/Jbvw86b4knnhlVDUaCZvIiIw9STFxEnBQrQbnGZQl5E3ORXyINCXkRcpZk8oJ68iIjTNJMXETcVwnnyLlDIi4ib1K4BFPIi4iotvAIKeRFxlE6h9GjhVUTEYZrJi4ib1K4BFPIi4iq1awCFvIi4SqdQAgp5EXGVZvKAFl5FRJymmbyIuEkLr4BCXkRcpXYNoJAXEVdpJg+oJy8i4jTN5EXESYGATqEEhbyIuEo9eUAhLyKuUk8eUMiLiKs0kwe08Coi4jTN5EXETfrsGkAhLyKuUrsGUMiLiKu08Aoo5EXEVZrJA1p4FRFxmmbyIuImtWsAhbyIuEohD5xG7ZrHnx5Dy05/okuvO2JdStS8MGY4P3z3Kcu/XsD5TRsetb906VK89+4kvlmzmFUrP+bpkY/k7DvzzGTmfzCN5V8v4KMF00lOTopm6YXuiWceZvFXs/ngkxk0bHxuyDGDHhvAF6vn892GL/Nsr5lcg3+/+xpzF07jg09m0Oaqy6NRckTUat2YGxc/S/dPn6fJ3Z2P2t+ofwe6ffw3ui54mo7/foRyyYk5+9pPGUyfb1/h6jceimbJJy0QyA774oLTJuS7dGxLypgRsS4jajq0v4KzGtTlnD9czp13Psy4l58JOW7MCyk0bNSK5i2u5tJLWtD+6jYAjP7bUCa/OYNmF7RlxMixjBzxSMjjTwVtrrqcuvVq06rFNTzy4HBGPPd4yHEfzlvMdW1vOmr7gIduZ/as+XRs04MB/Qfz1LOPRbrkiPDF+bhsRF8+6D2aGW0GU/+6i6l4Vs08Y3Z8u57/6ziEmW0fZd2cpVz4WM+cfavHz2HRfSnRLlvClK+QN8YkGmOaBi+JJz6i6GnetBEJFcrHuoyo6dz5aia/OQOA/y5dTkLFBGrUqJZnzN69+1i0+HMADhw4wPIVa3Jm7OeeexYff/wpAAsXfca1ndtFsfrC1bZDG96Z9j4AK5atpkJCeapVr3LUuBXLVrNt646jtgcCAcqVLwtA+fLl2Ja+PbIFR0jVpvXJWr+V3Ru34z+Qzc+zvqR2uwvyjNny+fdk79sPwLblaymbVDln3+bPvuXAnn1RrTksfn/4FwccN+SNMfWNMR8Ba4E3g5e1xpiPjDFnRaNAOTnJNWuQumlzzu201C0k16xxzPEJCRW4plNbPl7oBfvq1d/R9fqOAHTp0oEKFcpTuXKlyBYdITWSqrE5LT3ndvrmrVRPqnacI/IaO3o81994DV+uWcAb0/7B0L+G/ldRUVc2qRK/bdmZc3tP+k7KJh37/6np2YrUhauiUVpkBPzhXxxwopn8JGACkGitPc9aex6QCLwe3CdFlM/nO2pbIBAIOTY+Pp43J4/j5XETWLduIwCDH36Kli0v5qul82j5x4tJTd3CwYMHI1pzpBTktQjl2q4dmDF1Fhc3asstPe5i7PinQ95n0Rei5mO8DA26XkaVxvVYlTInsiVFkmbywInPrkm01r6Ze4O11g9MMcaEbmxKzNx5R1/69bsZgGXLVlLrjMP91uRaSWzesjXkcSnjR/PT2nW8+NJrOdu2bNnKjd37A1C2bBm6Xt+JrKzdEay+cPXp14M/9b4BgNUrvqVm8uF/xdSoWb1ALZceva6nz413ArB82WpKlixJ5cRKZOzYeYIji5Y9W3ZSLlf7pWyNyuxJ33XUuJqXn0fTAdcyu9tI/PtPzV/sgDMz8XCdaCa/0xjT0xiTMwUwxviMMTcDmZEtTQpqfMpEmrdoR/MW7XjvvXn0vrkbABdd2IysX7NIT9921DHDnxxMQkJ5HnxoWJ7tiYmVcmarf314AG9M/Hfkn0AhmvSvaXRs3Z2Orbszf+7H3NDDO5Pk/OaN2Z21O2Tv/Vg2p6ZzWauLAGhwdl1KlipxygU8wPZVv1Chbg3Kn1GVuOLx1L/uYjYuWJ5nTOJ5tfnjqFuZf+sY9mVkxahSKUwnmsn3BVKAccaYtOC2ZGBlcN8pY9CwUXy1YjWZmVlc2aUXd/XrzQ2dr451WREz9z8f0b79FdjvP+N/e/dy220P5uxb9tV8mrdoR3JyEo8+ch/f//ATXy2dB8A//vE6E16fSqtWlzLyqUcIEGDJki8ZcO+peUYJwMcLltCm7R/5ZNkc9u7dx8ABQ3L2zV30Nh1bdwfgkWEPcF23jpQuU4ov1yzg35NnMnb0eEYMfY5RLwyj3x29CQQCPHT3kGM9VJEWyPbz+ZCJdHhzML64OOy0xez6MY0LBt7A9lXr2LhgORc93pNiZUtxVcq9APyWlsH8W8cA0PmdISQ0SKJ42VL0/OpFlgx8ldTFa2L5lI7PkXZLuHz56U0aY6oCZwRvbrLWntTpBQd2/JL/RqjDilepR7ESybEuo0g4uD+N2omNY11GkbAhYzWv1uoV6zKKhP6pUyDkIkL+7Z33cth5U/rqe07FxZc88vWO12Con5rnjYnI6UkzeeA0ejOUiMjpSJ9dIyJu0kweUMiLiKt0CiWgkBcRV2kmDyjkRcRVmskDWngVEXGaZvIi4ia1awCFvIi4Su0aQCEvIq7STB5QyIuIqxTygBZeRUScppm8iLipAF8M4zKFvIi4Se0aQCEvIq5SyAMKeRGRQmGMORuYiPc92BlAH2vtT0eMmQTk/gKFxkAXa+17xpgngLuAzcF9n1lr7w63LoW8iLgp+ufJpwDjrLVTjDG9gFeAK3IPsNb2OXTdGNME+BiYl2vIJGvtwMIsSmfXiIib/P7wL/lkjKkGNAOmBjdNBZoFv1XvWPoBb1prfz/p55gPmsmLiJsK4ewaY0xFoGKIXZnW2sxct88A0qy12QDW2mxjzObg9qO+Vc8YUwK4CbjqiF1/Msa0A9KBYdbaL8J9DprJi4ibCmcmfz+wLsTl/jCr6wJstNauzLUtBahrrW0MPAvMMsYkhvk4CnkRkeMYC9QNcRl7xLhNQLIxJh4g+GfN4PZQbgUm5N5grU231h4IXl8QPLZhuE9A7RoRcVMhnEIZbMlk5mPcNmPMSqAnMCX45wprbahWTS3gj3jtmtzbk621acHrTYE6gA33OSjkRcRN0T+75g5gojFmKLAL6ANgjJkLDLXWLguO6wu8b63decTxTxtjLgCygf1Ab2tterhFKeRFxEkBf3Q/1sBa+wNwUYjtHY+4PfIYx/eNRF0KeRFxk97xCmjhVUTEaZrJi4ib9M1QgEJeRFwV5Z58UaWQFxE3qScPqCcvIuI0zeRFxE2ayQMKeRFxlb7+D1DIi4irNJMHFPIi4iqdXQNo4VVExGmayYuIm/RmKCDKIV+8Sr1oPlyRdnB/WqxLKDI2ZKyOdQlFRv/UKbEuwR1q1wBRDvliJZKj+XBF1sH9aRzY8UusyygSilepx+N1bjrxwNPAiPVvsWdI91iXUSSUfertsO8joIVXQO0aEXGVZvKAFl5FRJymmbyIuEkLr4BCXkRcpXYNoJAXEVdp4RVQT15ExGmayYuIm9SuARTyIuIqLbwCCnkRcZVm8oBCXkQcpXe8erTwKiLiMM3kRcRNatcACnkRcZVCHlDIi4irdHYNoJAXEVdpJg9o4VVExGmayYuIkwKayQMKeRFxlUIeUMiLiKv0ZihAPXkREadpJi8iblK7BlDIi4irFPKAQl5EHBUIKORBIS8irtJMHtDCq4iI0zSTFxE3aSYPKORFxFF6x6tHIS8iblLIAwp5EXGV3vAKaOFVRMRpmsmLiJPUk/co5EXETQp5QCEvIq5STx5QT15ExGmayYuIk9ST9zgX8i+MGU6H9lfwv7176dfvAVas/CbP/tKlSzFt6j+pV7822dnZzJmzgEcfewaAM89M5rV/jqFK1crs2plJn1vuJS1tSyyeRkQ9/vQYPvlsKZUrVeTdKSmxLifizmrVmI5D+xAXH8fX0xbyyfj38+yvc+E5dBzam+rnnMnbA17i2/8sBaDGH2pz7YhbKVmuNIFsP4vGvcs3s7+MxVMoNPENmlCi05/BF8fBrz/iwJJZR49peAkl2twIBPCnb+D36S8e3lmyNKXvfYHs75ayf86E6BV+MtSuARxr13RofwVnNajLOX+4nDvvfJhxLz8TctyYF1Jo2KgVzVtczaWXtKD91W0AGP23oUx+cwbNLmjLiJFjGTnikWiWHzVdOrYlZcyIWJcRFb44H52H/5lJt4zmxbaDaHTtpVRtkJxnTObmHbwzMIXVsz7Ps/3A3t9558HxvNRuMBP7jqLT0N6UqlAmmuUXLp+PEp37sW/S0+x96QHiG1+Gr2re18JXuQbFW3Zh76tD2PvSQ/w+9408+0tc2QP/+u+iWPTJC/gDYV9c4FTId+58NZPfnAHAf5cuJ6FiAjVqVMszZu/efSxa7P0wHzhwgOUr1pCcnATAueeexccffwrAwkWfcW3ndlGsPnqaN21EQoXysS4jKmo1bUDGhq3s2rSN7APZrHn/C85td0GeMZmpO9j6wyYCgbxTv4x16WSsTwdg97ZMfsvIomzlClGrvbDF1WqAPyOdwK5tkJ1N9prPKXZuizxjijW/koP/nQf79ngb9mQdPr5mXXzlEsheuyqaZZ88fyFcHOBUyCfXrEHqps05t9NSt5Bcs8YxxyckVOCaTm35eKEX7KtXf0fX6zsC0KVLBypUKE/lypUiW7REVIXqlfh1c0bO7awtO6lQvXKB7ye5SX3iixdj54athVleVPkqVCbw6+HXIvBrBr7yeV+LuCo18SUmUeq24ZS6fQTxDZoED/ZRon0f9n8wJZolSyE46ZA3xqwpzEIKg8/nO2rbsb44ID4+njcnj+PlcRNYt24jAIMffoqWLS/mq6XzaPnHi0lN3cLBgwcjWrNEWAH+ThxLuaoV6TbmTmYOeuUU/yKKo18LOOL5xMURl5jEvglP8vvbf6dElzugVBmKXdiO7B9XEMjKCHEfRVPAH/7FBcddeDXG/OE4uxMLuZaTcucdfenX72YAli1bSa0zaubsS66VxOYtoWdeKeNH89Padbz40ms527Zs2cqN3fsDULZsGbpe34msrN0RrF4iLSt9Jwk1D/9VrZBUmd3bduX7+JLlStPn9UF8+Px0UlesjUSJURPIysCXcPi18CUkEtid97UI/LoTf+qP4M8mkLmdwI7NxCUmEX/G2cTVPpdiF7bDV6IUxBcjsH8fBxa8Fe2nkX9RDmljzNnARLxszAD6WGt/OmLME8BdwKGWw2fW2ruD+8oArwMXAAeBgdba2eHWdaKza74B1hN6ClAl3AcvDONTJjI+ZSIAHTtcyV133sK0abO46MJmZP2aRXr6tqOOGf7kYBISynP7Xwbm2Z6YWImdOzMJBAL89eEBvDHx31F5DhI5aat+JrFODSrVqkrW1p006nwJ0+99OV/HxheP56ZXHmDFzCV8O/e/Ea408vxpPxOXmISvYlUCu3cS3+jSvGfOANnfLyW+8eWwYjGUKY+vShL+nVv5fcZLOWOKnd+KuJr1i3bAE5OZeAowzlo7xRjTC3gFuCLEuEnW2oEhtg8EdltrGxhjzgKWGGMaWGt/C6eoE4X8euCP1tq0I3cYYzaF88CRMPc/H9G+/RXY7z/jf3v3ctttD+bsW/bVfJq3aEdychKPPnIf3//wE18tnQfAP/7xOhNen0qrVpcy8qlHCBBgyZIvGXDvY7F6KhE1aNgovlqxmszMLK7s0ou7+vXmhs5Xx7qsiPBn+5k99A36Tvqrdwrl24vY9lMaVz7QjbQ1v/DDh8tJblyPm155gNIJZTnnymZc8UA3Xmo3mIadLqbOhedQplI5mnVrCcA7A18h/bsNMX5WJ8nvZ//sCZTq+xjExXFw+UIC21IpfkV3/Jt/JvuHr8leu4r4Bk0oPWAMBPzsnzcF9oaVMbETxZA3xlQDmgFtg5umAi8bY6paa7fn8256AH0BrLU/GWOWAR2A6eHU5jtej9EY8yzwf9baz0Ps+7u19r6CPFixEsmnckOz0Bzcn8aBHb/EuowioXiVejxe56ZYl1EkjFj/FnuGdI91GUVC2afehtAdhHzbcXWrsPPmsvXplYCKIXZlWmszD90wxlyAN0M/L9e274Be1trlubY9AdwG7ATSgWHW2i+C+3YD9Q79UjDG/ANYa60dE85zOO7Cq7V2UKiAD+4rUMCLiERTIS283g+sC3G5/yTLSgHqWmsbA88Cs4wxEV3fdOoUShGRQwop5McCdUNcxh7xcJuAZGNMPEDwz5rB7TmstenW2gPB6wuC+xsGd28EaucafuaRx58M5z7WQEQECmfhNdiSyczHuG3GmJVAT2BK8M8VR/bjjTHJh9Y4jTFNgTqADe6eDvwFWBZceG0RvJ+wKORFRArHHcBEY8xQYBfQB8AYMxcYaq1dBjwd7N9nA/uB3tba9ODxzwJvGGPWBvffbq0N+xxuhbyIuCkQ1rptgVlrfwAuCrG9Y67rfY9z/B7gxsKuSyEvIk5y5R2r4VLIi4iTAv7ozuSLKoW8iDhJM3mPTqEUEXGYZvIi4qRAlBdeiyqFvIg4Se0aj0JeRJykhVePevIiIg7TTF5EnHRKf4lXIVLIi4iT1K7xKORFxEkKeY9CXkScpHaNRwuvIiIO00xeRJykdo1HIS8iTtI7Xj0KeRFxkt7x6lHIi4iT/JrJA1p4FRFxmmbyIuIk9eQ9CnkRcZLOrvEo5EXESXozlEc9eRERh2kmLyJOUrvGo5AXESfpFEqPQl5EnKSzazwKeRFxkhZePVp4FRFxmGbyIuIk9eQ9CnkRcZJ68h6FvIg4ST15jy8Q3VdCL7uI5FdYU/FltbqEnTfNU9895f85oIVXERGHRbVdUzuxcTQfrsjakLGax+vcFOsyioQR69/iwI5fYl1GkVC8Sj3qJjaJdRlFwrqMVWHfh3ryHvXkRcRJOrvGo5AXESdpAdCjnryIiMM0kxcRJ6ld41HIi4iTtPDqUciLiJP8sS6giFDIi4iTAuG9l8oZWngVEXGYZvIi4iS/zqEEFPIi4ii/2jWAQl5EHKWevEchLyJO0tk1Hi28iog4TDN5EXGS2jUehbyIOEntGo9CXkScpJD3qCcvIuIwzeRFxEnqyXsU8iLiJL8yHlDIi4ij9I5Xj0JeRJykj67xKORFRAqBMeZsYCKQCGQAfay1Px0xZgjwJ+Bg8PKotXZecN8bwFXAjuDw6dbakeHWpbNrRMRJ/kK4FFAKMM5aezYwDnglxJilQAtrbRPgVmCaMaZ0rv2jrLVNg5ewAx4U8iLiKL/PF/Ylv4wx1YBmwNTgpqlAM2NM1dzjrLXzrLX/C95cDfjwZv4Ro3aNiDipMHryxpiKQMUQuzKttZm5bp8BpFlrswGstdnGmM3B7duPcfd9gJ+ttam5tj1ojPkL8DPwiLX2+3Cfg2byIuKkQmrX3A+sC3G5P5zajDGtgKeAnrk2PwY0sNY2AmYCHxhj4sN5HFDIi4gcz1igbojL2CPGbQKSD4Vy8M+awe15GGMuAaYAXay19tB2a22atdYfvD4JKAfUCvcJqF0jIk4qjDdDBVsymfkYt80YsxJvZj4l+OcKa22eVo0xpgUwDehmrV1+xL5ka21a8PrVQDaQFu5zUMiLiJNi8GaoO4CJxpihwC68njvGmLnAUGvtMuAfQGngFWPMoeN6W2vXBI+tjtcpygKutdYeDLcohbyIOCnab4ay1v4AXBRie8dc11sc5/irIlGXevIiIg7TTF5EnKQPKPMo5EXESfrSEI9CXkScpA8o8zjXk3/imYdZ/NVsPvhkBg0bnxtyzKDHBvDF6vl8t+HLPNtrJtfg3+++xtyF0/jgkxm0ueryaJQcEWe1asx9Hz3HA4vG0PLOzkftr3PhOdw1eyRPrp3MeR0uzNle4w+1uX3mkwyYP5p7/jOKhtdcHM2yo+7xp8fQstOf6NLrjliXEjXDnnmYhV+9z38+mc55jc8JOWbgY/fw2ep5fLPhizzbaybX4K13X2P2wmn855PptC7CPyN+X/gXFzgV8m2uupy69WrTqsU1PPLgcEY893jIcR/OW8x1bW86avuAh25n9qz5dGzTgwH9B/PUs49FuuSI8MX56Dz8z0y6ZTQvth1Eo2svpWqD5DxjMjfv4J2BKaye9Xme7Qf2/s47D47npXaDmdh3FJ2G9qZUhTLRLD+qunRsS8qYEbEuI2paX3U5deqdSZsWnU/4M9Kl7c1Hbb/nof7MmTWPa9r0YED/h3nq2UcjXbKEyal2TdsObXhn2vsArFi2mgoJ5alWvQrbtu7IM27FstUhjw8EApQrXxaA8uXLsS39WB85UbTVatqAjA1b2bVpGwBr3v+Cc9tdwPa1h99XkZnqvSaBQN7OZca69Jzru7dl8ltGFmUrV2Bf1v9wUfOmjUjbsjXWZURN2w5tmBn8GVm5bA0VEspTtXoVth/xM7Jy2ZqQxwcCUK58OcD7GdlahH9G1JP3HDfkjTGJwN+AM4FZ1tpxufa9Y629IcL1FUiNpGpsTjscUumbt1I9qdpRIX8sY0ePZ/KMV7il/02UKVOam7r2j1SpEVWheiV+3ZyRcztry05qNW1Q4PtJblKf+OLF2Lnh9AlB11VPqsaWtMP/P7ds3kqNpGpHhfyxjB09nkkzUujbvydlypSmV9fbI1Vq2BTynhO1a14BduJ9TnIXY8xMY8yhXwz1IlrZSfCF+GjQQCD/yy/Xdu3AjKmzuLhRW27pcRdjxz8d8j6LvDBfB4ByVSvSbcydzBz0SoGPlaIr1F/ngv6MvDP1PS5t1I4/97ibMeNHFtmfkYAv/IsLThTyDay1g621M4F2wBZgtjGmVORLy58+/Xowd9HbzF30NlvTt1MzuUbOvho1qxeo5dKj1/XMfnceAMuXraZkyZJUTqxU6DVHWlb6ThJqHv6I6gpJldm9bVe+jy9ZrjR9Xh/Eh89PJ3XF2kiUKFHUu18P5iyaxpxF09iWvp2k5Oo5+5JqVi9Qy6V7r+uZE/wZWVHEf0Zi8KUhRdKJQr7koSvW2oC19m5gDTAHKBJBP+lf0+jYujsdW3dn/tyPuaGHdybJ+c0bsztrd75bNQCbU9O5rJX3ruQGZ9elZKkSZOzYGZG6Iylt1c8k1qlBpVpViS8eT6POl/DDgq/zdWx88XhueuUBVsxcwrdz/xvhSiUaJv9rGp1a96BT6x7Mn7uQrsGfkabNG7E767d8t2oANqdu4dLgz0j9U/hn5HRyopD/xRjTMvcGa+0g4Evg7IhVdZI+XrCEjRtS+WTZHEa9MIzHBx3+9qy5i97Ouf7IsAf4cs0CSpcpxZdrFnD/4DsBGDH0OXr2voH/LJ7Oi//8Gw/dPSTqz6Ew+LP9zB76Bn0n/ZX7PnyOb2Z/ybaf0rjygW6cc1UzAJIb12PQFy/RsONFXPd0PwbMHw1Aw04XU+fCc2jWrSV3z32au+c+TY0/1I7l04moQcNGcfNfHmD9xlSu7NKLd96fF+uSImrhgiVs2pDKomWzeeaFYQzJ9TMyZ9G0nOt/HXY/n6+ZT+kypfh8zXzuG+ydYjpy6PP8qXdX5i5+mxf/OYpBdw+N+nPIL83kPb7j9eOMMZWBgLX2qH/rG2P+YK39riAPVjuxsZq7wIaM1Txe5+hTOE9HI9a/xYEdv8S6jCKheJV61E1sEusyioR1GauA8D5G8qUzeoWdNwM2TTnlO/PHPbvGWnvMf4cVNOBFRKLJlTczhcupN0OJiEheTr0ZSkTkEFd66uFSyIuIkxTyHoW8iDhJZ3l4FPIi4iQtvHq08Coi4jDN5EXESerJexTyIuIk9eQ9CnkRcZJfMQ+oJy8i4jTN5EXESerJexTyIuIkNWs8CnkRcZJm8h6FvIg4SW+G8mjhVUTEYZrJi4iTdAqlRyEvIk5SxHsU8iLiJC28ehTyIuIktWs8WngVEXGYZvIi4iTN4z0KeRFxknryHoW8iDhJPXmPevIiIg7TTF5EnKR5vEchLyJOUk/eo5AXEScFNJcHFPIi4ijN5D1aeBURcZhm8iLiJJ1C6VHIi4iTFPEehbyIOEkzeY9CXkScpIVXjxZeRUQcppm8iDhJ58l7fIFAVF8Iveoikl++cA6+tU63sPNmwvoZYdVQFGgmLyJO0kzeE9WQf7VWr2g+XJHVP3UKe4Z0j3UZRULZp96mbmKTWJdRJKzLWMWBHb/EuowioXiVerEuwRmayYuIk3R2jUchLyJO8kd3vbHIUsiLiJMU8R6FvIg4KdrveDXGnA1MBBKBDKCPtfanI8bEAy8C7fF+D42y1r52on3h0JuhREQKRwowzlp7NjAOeCXEmJuBBsBZwCXAE8aYOvnYd9IU8iLipIaVbp8AAAyLSURBVEAh/JdfxphqQDNganDTVKCZMabqEUN7AK9aa/3W2u3Au8CN+dh30tSuEREnFcbZNcaYikDFELsyrbWZuW6fAaRZa7MBrLXZxpjNwe3bc407E9iQ6/bG4JgT7TtpmsmLiJP8BMK+APcD60Jc7o/dMysYhbyIOKmQ2jVjgbohLmOPeLhNQHJw8fTQImrN4PbcNgK1c90+M9eY4+07aWrXiIgcQ7Alk5mPcduMMSuBnsCU4J8rgr313KYD/Y0xM/HOwukCtMzHvpOmkBcRJ8XgHa93ABONMUOBXUAfAGPMXGCotXYZMBm4CDh0auVwa+2hz7I43r6TppAXESdF+RN2sdb+gBfSR27vmOt6NnDnMY4/5r5wKORFxEn6+j+PFl5FRBymmbyIOEmfQulRyIuIk/SlIR6FvIg4ST15j0JeRJwU7bNriiotvIqIOEwzeRFxkhZePQp5EXGSFl49CnkRcZIWXj0KeRFxkhZePVp4FRFxmGbyIuIktWs8CnkRcZIWXj0KeRFxkl89eUA9eRERp2kmLyJO0jzeo5AXESdp4dWjkBcRJynkPQp5EXGS3gzl0cKriIjDNJMXESepXeNxKuRrtW7MJU/2xhcfh526iFXj3s+zv1H/DpierfFnZ7MvYzefPPRPfkvLAKD9lMFUO78+W7/6kXm3PB+L8gtVfIMmlOj0Z/DFcfDrjziwZNbRYxpeQok2NwIB/Okb+H36i4d3lixN6XtfIPu7peyfMyF6hUfIsGcepvVVl7Nv7z4G3jOEb1f/cNSYgY/dw/U9OpOQUIGGtS/J2V4zuQbPjRtBhYTyxMfH8bfhf2fRh59Gs/yoePzpMXzy2VIqV6rIu1NSYl1O2PRmKI8z7RpfnI/LRvTlg96jmdFmMPWvu5iKZ9XMM2bHt+v5v45DmNn2UdbNWcqFj/XM2bd6/BwW3Xfq/8UGwOejROd+7Jv0NHtfeoD4xpfhq5qcd0jlGhRv2YW9rw5h70sP8fvcN/LsL3FlD/zrv4ti0ZHT+qrLqVPvTNq06MwjDw5nxHOPhxz34bzFdGl781Hb73moP3NmzeOaNj0Y0P9hnnr20UiXHBNdOrYlZcyIWJdRaAKBQNgXFxQ45I0xlSJRSLiqNq1P1vqt7N64Hf+BbH6e9SW1212QZ8yWz78ne99+ALYtX0vZpMo5+zZ/9i0H9uyLas2RElerAf6MdAK7tkF2NtlrPqfYuS3yjCnW/EoO/nce7NvjbdiTdfj4mnXxlUsge+2qaJYdMW07tGHmNO9fdSuXraFCQnmqVq9y1LiVy9awfeuOo7YHAlCufDkAypcvx9b07ZEtOEaaN21EQoXysS6j0PgJhH1xwXHbNcaYJsAEIBvoCzwHtDHGZACdrbUrI19i/pRNqsRvW3bm3N6TvpNq59c/5njTsxWpC90IsSP5KlQm8GtGzu3ArxnE1Torz5i4KjXxA6VuGw5xcRz4eLoX6j4fJdr34fcZLxNfv2GUK4+M6knV2JK2Nef2ls1bqZFULWSghzJ29HgmzUihb/+elClTml5db49UqSKF7kQz+ReBJ4GXgQ+At6y1ZYC78AK/CPEdvekYv4gbdL2MKo3rsSplTmRLipkQr8WRL0ZcHHGJSeyb8CS/v/13SnS5A0qVodiF7cj+cQWBrIwQ93Fq8oX6q1GAf4pf27UD70x9j0sbtePPPe5mzPiR+ELdqRQpatd4TrTwWt5a+x6AMeYpa+2bANba940xwyNeXQHs2bKTcrnaL2VrVGZP+q6jxtW8/DyaDriW2d1G4t9/MJolRk0gKwNfQmLObV9CIoHdeV+LwK878af+CP5sApnbCezYTFxiEvFnnE1c7XMpdmE7fCVKQXwxAvv3cWDBW9F+GmHp3a8Hf+rdFYDVK74lKbl6zr6kmtUL1HLp3ut6brnxTgBWLFtNyZIlqZxYiYwdO09wpMSSK+2WcJ1oJp97ujK/gMdG1fZVv1Chbg3Kn1GVuOLx1L/uYjYuWJ5nTOJ5tfnjqFuZf+sY9mVkHeOeTn3+tJ+JS0zCV7EqxMcT3+hSDv6wLM+Y7O+XElc32I4pUx5flST8O7fy+4yX2Pv8Xewdcw/7503m4MpPTrmAB5j8r2l0at2DTq17MH/uQrr26AxA0+aN2J31W75bNQCbU7dwaauLAKh/dl1KliqhgD8FBArhPxecaCa/3hhT3lq721rb/9BGY0wt4H+RLa1gAtl+Ph8ykQ5vDsYXF4edtphdP6ZxwcAb2L5qHRsXLOeix3tSrGwprkq5F4Df0jKYf+sYADq/M4SEBkkUL1uKnl+9yJKBr5K6eE0sn9LJ8/vZP3sCpfo+BnFxHFy+kMC2VIpf0R3/5p/J/uFrsteuIr5BE0oPGAMBP/vnTYG9v8W68ohYuGAJbdpezqJls9m7dx+DBwzN2TdnkffLAOCvw+7n2m4dKV2mFJ+vmc+0yTP5++gURg59nmdeGEq/O3oRCAQYdPfQYz3UKW3QsFF8tWI1mZlZXNmlF3f1680Nna+OdVkSJt/J9J2MMWWBstbabQU57tVavdz41Rim/qlT2DOke6zLKBLKPvU2dRObxLqMImFdxioO7Pgl1mUUCcWr1IPQi0v51rD6xWHnzTdbvzzlF19O6s1Q1to9wJ5CrkVEpNC40m4Jl1PveBUROUTfDOVRyIuIkzST9xSpM2RERKRwaSYvIk5Su8ajkBcRJ6ld41HIi4iTNJP3KORFxEmayXu08Coi4jDN5EXESYGAP9YlFAkKeRFxkj6F0qOQFxEnufJ58OFST15ExGGayYuIk9Su8SjkRcRJatd4FPIi4iS9GcqjkBcRJ+nNUB4tvIqIOEwzeRFxknryHoW8iDhJZ9d4FPIi4iTN5D3qyYuIOEwzeRFxkk6h9CjkRcRJRaldY4wpA7wOXAAcBAZaa2eHGHcdMBQoCfiACdba54P7bgHGAuuDw9dZa68/0WMr5EXESUVs4XUgsNta28AYcxawxBjTwFr72xHj0oHO1trNxpgE4GtjzFJr7ZLg/g+ttd0K8sDqyYuIkwKBQNiXQtQDSAGw1v4ELAM6HDnIWvtfa+3m4PVfge+B2uE8sGbyIiLHYIypCFQMsSvTWptZgLs6E9iQ6/ZG4IwTPPY5wMXAX3JtbmWMWQlkAX+z1s450QMr5EXESYW08Ho/MCzE9ieBJw7dMMYsxwvyUKoX9EGNMUnALODuQzN7YDYwzVq71xhzPvCBMaa1tfb7492XQl5EnFRIn10zFngjxPY8s3hrbbPj3YkxZiNe22V7cNOZwMJjjK0GfAg8a619O9dj7Mh1fYUx5lPgQryWzjEp5EXESYUxkw+2ZArSljmW6Xhtl2XBhdcWQM8jBxljEoEFwMvW2teO2JdsrU0LXq+N18oZcaIHVsiLiJOK0imUwLPAG8aYtUA2cLu1djeAMWY4sNlamwL8FTgb+Isx5lAv/u/W2teBu4OnWB4Mbn/UWrviRA+skBcRiTBr7R7gxmPsG5rr+iBg0DHGPQo8WtDHVsiLiJP0efIehbyIOKmItWtiRiEvIk5SyHv0jlcREYf59NtORMRdmsmLiDhMIS8i4jCFvIiIwxTyIiIOU8iLiDhMIS8i4jCFvIiIwxTyIiIOU8iLiDjstPnsGmPM2cBEIBHIAPoEv1D3tGOMeQ64AagDNLLWfhPbimIj+AUNk4H6wO/AWuAv1trtxz3QUcaYd4G6gB/4DRhgrV0Z26okXKfTTD4FGGetPRsYB7wS43pi6V2gJXm/WPh0FABGW2uNtbYx8DMwKsY1xVJfa20Ta+35wHPAhFgXJOE7LUI++J2JzYCpwU1TgWbGmKqxqyp2rLWfWms3xbqOWLPW7rTWLsq16Uu87+E8LVlrf811MwFvRi+nuNOlXXMGkGatzQaw1mYbYzYHt5+W/zSXvIwxccCdwHuxriWWjDGvAe0AH9A+xuVIITgtZvIi+fASXh/65VgXEkvW2tustWfifc3cs7GuR8J3uoT8JiDZGBMPEPyzZnC7nOaCC9FnAT2stWpRANbayUCb4OK0nMJOi5C31m4DVgI9g5t6AitO17Mo5DBjzEjgAqCLtfb3WNcTK8aYcsaYM3Ld7gzsDF7kFHbafGmIMeYcvFMoKwG78E6htLGtKjaMMS8CXYEawA4gw1p7Xmyrij5jzHnAN8CPwN7g5nXW2utjV1VsGGOqA7OAskA2XrgPtNYuj2lhErbTJuRFRE5Hp0W7RkTkdKWQFxFxmEJeRMRhCnkREYcp5EVEHKaQFxFxmEJeRMRhCnkREYf9P6OLm3fyJRM0AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, ax = plt.subplots(figsize=(6, 6))\n",
    "sb.heatmap(partial_corr_results, annot=True, linewidths=.5, ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(257, 60)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "questionare=all_features.iloc[:,0:60]\n",
    "questionare.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7fb126523780>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "#Put together into a single data frame so we can plot easier\n",
    "frameMVP = { 'turnover rate': turnOverRate, 'Provide excellent service': all_features.iloc[:,55]} \n",
    "dataForPlotMVP = pd.DataFrame(frameMVP)\n",
    "# Recommended way\n",
    "sb.lmplot(x='Provide excellent service', y='turnover rate', data=dataForPlotMVP)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f3be7d22860>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAFgCAYAAACFYaNMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nOy9e5Ql1X3f+9lVdZ797ulmhmEYHoOnQJJHMhIBJCUGETCWLBwrZFkRku0Qr+vr68TOY11W7k2WnXjl4WBfO/aK7NhX5i7bErKTCXaQkDEKFrYAaSKMpBECCtMCDcM8mJ7u6ed5VdW+f1TV6Tqnz+nzfnX/PmvBdFdXndpVp+q7f/u3f/v3U1prBEEQhP5jDLoBgiAIexURYEEQhAEhAiwIgjAgRIAFQRAGxG4RYAu4OvxXEARhJNgtgnUIeO3ixXV8v/dRHTMzWZaXN3t+nn4i1zQayDUNP7WuZ35+QtXad7dYwH3FssxBN6HryDWNBnJNw08r1yMCLAiCMCBEgAVBEAaECLAgCMKAEAEWBEEYECLAgiAIA0IEWBAEYUCIAAuCIAwIEWBBEIQBIQIsCIIwIHbLUuQ9wcmFRR4/cYrFlTxzU2nuvvkwx47MDbpZgiC0iQjwiHByYZHPfPEVTNMgm7a4tFHkM198BUBEWBBGFHFBjAiPnziFaRqkEiZKKVIJE9M0ePzEqUE3TRCENhEBHhEWV/IkrcqvK2kZLK7kB9QiQRA6RQR4RJibSlN0/YptRddnbio9oBYJgtApIsAjwt03H8bzfAolD601hZKH5/ncffPhQTdNEIQ2kUm4ESGaaJMoCEHYPYgAjxDHjsyJ4ArCLkJcEIIgCANCBFgQBGFAiAALgiAMCBFgQRCEASECLAiCMCBEgAVBEAaECLAgCMKAEAEWBEEYECLAgiAIA2LPr4STJOeCIAyKPS3AkuRcEIRBsqddEJLkXBCEQbKnBViSnAuCMEiGxgVh2/Y+4A+AI0ABeBX4KcdxLvTqnHNTaS5tFEklzPI2SXIuCEK/GCYLWAMPOo5jO45zDFgAfqmXJ5Qk54IgDJKhsYAdx1kCnopt+irw0708pyQ5FwRhkCit9aDbsA3btg3gCeBRx3F+o4lDrgZe62mjBGGP8NxL53nkqVc5v7TJ/tksH7ntOt5zw/5BN6vvdPk+qJobh1SAPwlcAXzEcRy/0f6EAnzx4jq+3/vrmZ+f4MKFtZ6fp5/INY0Gvb6meGhm0jIouj6e53PfnUd7NjIcxu+pk/tQ63rm5ydqCvAw+YABsG37V4DvAX60SfEVBKFLSGhmQL/uw1AJsG3b/w54N/B3HMcpDLo9grDXkNDMgH7dh6ERYNu23w7838BB4Fnbtr9h2/YfD7hZgrCnmJtKU3QrB557MTSzX/dhmKIgvk0dR7UgCP3h7psP85kvvkIBKnyfey00s1/3YWgEWBBGjd2YyElCMwP6dR9EgAWhDXZzIqdjR+ZG/hq6QT/ugwiwIIS0YtHGZ8kBUgmTQrhdxEtoFhFgQaB1i3ZxJU82Xfn67MVoAaEzhiYKQhAGSatxnxItIHQDEWBBoPW4T0nkJHQDcUEIAq2nJpVoAaEbiAALAu3FfUq0gNApIsCCgFi0wmAQARaEELFohX4jk3CCIAgDQgRYEARhQIgAC4IgDAgRYEEQhAEhk3BC19iN2cEEoZeIAAtdYTdnBxOEXiECLHSFWtnBVosev/Poi2TTlljEglAD8QELXaE6l0Ku4LK6USBf8ios4pMLiwNspSAMF2IBjxDD7GOtzqWwulEEFAnLKGcXq5cvd5ivSxB6iQjwiNBNH2stwbtjfqKj9lXnUii5HijFZDZR3qdWdrF++I5F4IVhRQR4ROhWBYZ6gjc1leWquWzb7avOpZBKWCQTBtn0lgDXyi7W68oSozY5KJ3F3kIEeEToVgWGeoL3yFOv8k/vPdZRG+O5FCLhK5S8HbOL9bqyRPX1ep7PynqRTz7yLY5cMcX1h6d5+dSloRC8UessesVe6oREgEeEVvPV1qOe4L21tNmVdkY0m12sW9dVj/j1Lq/mWd0slf92+sI6r5y+xGQ2yeRYcuCCJ3Xm9l4nJAI8IrSTr7YW9QTvstn23Q/1aCa7WLeuq5qTC4s8efwkK+sFVjeKpJNmhfgqYCPnohTkih5T4ztPFPYDqTO39zohCUMbEY4dmeO+O48yPZZkM+8yPZbkvjuPtvxQ1iul85HbrutRy3emW9cVJ7KilldzTE+k8Hy/QnwBTFOhAV+DG6vtNkjBkzpzrZeGGnXEAh5iavnCHvjYjR19Zj3XwHtu2M+FC2vdaHZbbeqmdRNZUemkhWkEYXBvLecq9vF8Xf7Zir3wrQhet32VvRoNjBK9dkkNGyLAQ0ovfWG7LfF4tRC+ubjBzESq/PdMysIyFG5MdPXWj2SSJlrrlgSvF9+PVOXYe52QCPCQstd8Ye0SCaHraTbzJZbX8vg++L7m8rmxrR1V8I9pKHytywI8mU1wYDbbsuD16vvZbZ1jq+y1TkgEeEjp54RMNGF19sL6yD3wj584hetp1jaDlXemYaC1z0beZXmtwFjaouj6+L5maixBoeTjej6WaTCRTaA1bbl1ZMKsd+ylTkgEeEjply8ssiBTSRPQfOfMKr9x/CQH58a497YjwHBbI4sreTbzJUBhxKxc7WvyRQ9FcC8jF8T0xNb9LJQ8pseSbZ13r/kqhd4gURBDSr1ohW77wqKhtOdplteL+FqXJ60e+sLLPPTYS1zaKA5tQp25qTSu55fFF0ADSctkPJPgwZ9+Lw987Ebuve1IV+9nv74fYXcjFvCQ0i9fWDSUvnAphwIMpUCBrzX5ggvAzGRg1Q2jH/rumw+z8OYqnq8xVCC+GsikzIrY5m7fz73mqxR6gwjwENMPX1g0lC65fjRPha/BMsOEOqiK/YfNz3nsyBwfuvUwj331FJ6vSZgGmZRJwjK2xTZ3+37uJV+l0BvEBbHHiYbSprG1MAE0k2NJTMPANCoFeBj9nPe8/1p+5kfewdFDU0yEUQ333XmU99ywf9BNE4QdEQt4jxNZcH/89Gu8cW4Ny1RMjacwDEU6ZUHo3xz2mEyxRoVRRARY4NiROe645Rqe/OprZZ/m9FiSj34gGMIPys95cmGR408tcH5pE1Dsn0lz7+3XdbxMWfy2wrAgAiyUqWdFDkKgTi4s8tAXXmYjF8T3KqU5u5Tjocde4v4P3dBWm/Zapi1h+BEBFoaSx0+cIl9wUcooh5j5BLG9x59aaMuKldWFwrAhk3DCULK4ksfzK+N7FeB6PmcWN9qKTd5rmbaE4UcsYGEomZtKs7pRxNeURTiK8U20acXK6rXuIH707iEWsDCU3H3zYdIpC619PF+XE+hoDZNjiYp9m7ViZfVa50R+9GFeHTlKiAUsDGUynmNH5rj/g9dXREEcmA0sVVdX7tusFdvJ6jWx+gLEj95dRID3OPFkPMMWGVArKqNc7JPm88V2Kp4SPbGFZIHrLuKC2OPEq0coFdRFM02Dx0+cGnTTatJqCaNuDJnjVt8o3KNeImWTuotYwHucfls03RjKt7LqrRtDZrH6tuhWxYr4c3D5/Dh3fN/BPTeagCETYNu2fwX4u8DVwPc6jvPCYFs0eHrte4wiA5J9iAxoNJTvxbV2QzwlemKLbmSBq34Olldze9alM1QCDPwJ8OvAlwfdkGGgH77HyKLJF10MpXqa72EnaxToybV2Qzz3Wp2yRnSad6P6OUgnTFxP78mJvKHyATuO87TjOG8Muh3DQj98j5FPdWYy07Wy8PWoXgiRK7gsr+Z55Y1L/M6jL+J6uuvX2o3Qs1b9zsLOyIKYLYbNAu6IffvG+3au+fmJnp9jab3IRCaYHIuwTMXyerHj8z/30nkeeepVzi9tsn82y0duu67n6Rsvnx9neTVHOmGykXdZWi2g0SQTJoWSR6HkkU5ZjIUug2avdadrueOWa5iayvLIU6/y1tIml7V5rXfMT3DHLde0d+Fdph/PXi+JPwcRvtZcPj8+8tcW0ex17CoBvnhxHd/XjXfskPn5CS5cWOvpOU4uLLK+UeTipU0SlsnkWJJMyqJQ8pgZT3Z0/rhrI50wuLC8yW8/cpKVOzrLNFZ9jmo/4R3fd7BcwXh5NY/WGhRMZBOsbpZwXZ/l1TxJK6hk0cy11rqW3zz+De678yh33HINFy6scdVcln9677GK43r9/fWKfjx7vSb+HCQtA19rCkWPO77v4MhfG9T+juoJ8lC5IISASFSSCQOUouT6XFzJsbpR7IrvsZZrw7JU11wb9UK/gPJQvuQGlYlnJ1Jk0wkmswlQUHJbcxVIiNjoUe3SmZnM7FmXzq6ygHcLkahk0wkSlsHqZomS61Moevz4PW/rOB/uwpur+FpjWQaT2QTZdIJUwuyaD26nybYHPnYjx47M8eDDz1dMjmXTCUquT7Hks5l3m55dlxCx0SQ+kbcbrPp2GSoBtm37N4CPAAeA/2nb9kXHcd4+4Gb1nbioZNOBQGqt2cy7HYvvZ774CkoBGjzPZ2mtAEAqaXYtrKoZUawVWZCwDH7iB69v6Rp7FSJW7UK5/vA0L5+6tOeXIgvdZagE2HGcnwV+dtDtGDS9EpXIMp2eSLG0mgcUCs3KepHZyXTXwqqaaX+3qgrvFCL23Evn+aMnXm4r70M8JO7c0iavnL7E1FiKiWxiTy9FFrrLUAmwENCruNPIMlVKMTsZpHssuT4a+KmPHOOquWzDz9iJR5/+Dk987TSbBRcFjGUsZifTddvfjTpux47M8frZVZ742mnyRZd00uKumw4B8NuPnARFy3HF1S6UXNFDocgVXCbHkrsuAY0kGhocIsBDSLesw2rilmkmZZWjKqbHkrznhv0d+eEeffo7PPrs6ygUCVPhepr1nIvv5zi8f6JnL/XJhUWeeeEck+NJ5qxA7J954RzPORewLIVpBPPMrYhmtQvFdYPE8K63lQNht/iZJdHQYBEBHlLqZQLrRJR7uaLria+dRqHKZewTlsLzNaB44GM3dvz59ag34Xd+aZPDByZwva2wxGZFs9qFYlkGbuijjujlcu3q7xjoWd4ESS85WESAR4RuWCq9sqwB8kW3LL4RWms2Cy4P/NazPRva1pvwA0Wh5JUtYIC1zRKFotewPdUdVSZpsup6ZFIWWuueLUWu9R0/9IWXQWuymURP8ib0M4pk1Fwd/WivCPAAaOeL7ZalEreso3Z8+olXOras0snAnWGGGuz5Pn6wzqKnQ9t6E377Z9K4rsZVXjnaw/M0SkHR9Vhey7Pw5iofuvUw97z/2orPrO6oDsxmue1dB3seBVHrO14KhXBmMrC2u503oV+JhnYyIO4YwtVv/XLNiAD3mXa/2G5bKt3OSHXXTYd49NnX8fyghlvkLp0aT5YXSPRiaFvPrfLRO48yNZXld//Ht3hrKU/gDAlKGrmexjIVvtY89tVTXH355LY21XIB3dO1Vtem1nfs+X7Y8i26aaH2K9FQrc5ltejxO4++yMNPvsrseHKoLOJ+uWZEgPtMu19sty2VbmekiqzIKBpBEdRumxpPlffpxdB2J7fK/PwEf/TEy1w2m+XCcg5Pb/mDPV9jGYGfelj8nbW+47gLJaLo+qQTBg8+/HzHFnkzbqluDMWrO5dcwWV1o4BGcXlm+Cb/+uWaEQHuM+1+sfUslesPT7f1IvbiAbvn/deWhTha6RanVxNXO4WzRddpWQZe0SvbklpvVVgelmiGWt9xOmVBuDQ7aRnkiy6buRIohau7497Z6f51ayhe3bmsbhQBRcIyejpCapd+uWYkF0SfabekS62UiO97xwGeeeFcW+V2elFa5uTCIg8+/DwP/NazrG8W2cy7A69AHF3nZDaopByVto9+zqS6twKwU2p9x/d/8Hru/9ANFXkTJseSZNNWX/JfdCvXRnVa0JLrgaL8vcBwhfb1q4K2WMB9phOfW7Wl8uDDz7ftp6puR77odvSAVVtKRdcHHQzzW8nt0G2i6zRNg4msxdqmC0DCMhhLWyQsY6gSq9ezRuN5E/7BL/5Z3yIXujVSqnZ1pBIWyUSQ7yRimKqM9DJiKI4IcJ/p5hfb6cuRSpjlku8HLxvjvjuPtP2APX7iFK6nWdss4HpBprNMymI8k+AX/+HNLX9et0KA4vfbdX2m5pKgFPmiV87xEEWCjEJoFPS3RFI3z1UdgfOZL75CoeRhmWpgI6Sd6MZKzUaIAA+Abn2xrb4ckai9ubhBvuAxnk1w+dwYRdenUPA6asuZxQ028iWUMjBUsBJudaOA5/mND67Rzoe+8DL5govn+6yG8bD3f7C1RD0R9e73qK4C62eJpF6dK94xLq8XmRmyKIh+IQI8wrQyMQdbNdeKJR9fa9Y2iyQsA601qxtFPvnItzhyxVRbL0Kw4kwRrcUwFHhaVaxEa5bjTy2wkSuilIFpGPgaNnJFjj+10NUXdFRXgdUaRfXKku/lUDzqGCUdpTCS1HsRn3nh3DarLmUZZbFxvSC3gUaxvJrHJ5iN9bVu2wo0TQVu8BlhtktQ4fYWidwi1WIebO8eo5xLuNZwvleWfD+G4nsVEeARp9mJufPLeS4Ps51ZloEXinDJ1YFIKkUiPK4dK/CKuTHOLW2SK3q4ro8VLuE9MNtOhjWFUtWWs8b16Oqy5t1Sbn5ULXlBBHikqTVRVT83QpDDIJUwmcwmgqW5vg7CskKtmxxLlvdvxgqMnz+dNPF9zcxEqmNf4f6ZNGeXcvgE1rSvNZ4PptHdZc27pdz8KFvyex0R4BGl3rAznTDKQhtRdH32z2aDysNAJmUx4fps5IOQLNNQ7JvOlEuFN2MF1gw7UwpL0XHY2b23X8dDj71Evujh+bqcU2LfVKarQfv9CjWqRTtRHicXFnny+EnOXlivOKZfVUH24iRZrxEBHlHqDTtRCs/1tudG+MDR8nFRgpmKyTmDljJ91To/wHg2yS92mH7y2JE57v/QDeW2rqwXmJ5IkUltPa7dsvAG4d9sx2cbHZNKmtuOadeS30lgRzVCZNRoWoBt274BuBc44DjOz9i2fT2QdBznZM9aJ9Sl3rBzM+/y8buO1n2x6r08T379zDbLqp3zd2vYGxfGbi1rbpRrt19WXjs+2+iYdNKiFI5w4oVOW72ORgIrfuX+0JQA27b994BPAo8AHwN+BhgHfgn42z1rXR8Y1WHWTsPOVq26Y0fmuOOWa1oKBernBFY3fLU1c+0+9hIoRTZt9dXKa6fzanRMq995I4EVv3J/aNYC/kXgLsdxvmHb9o+G274JvLM3zeoPozzM6sWwsx/nb4du+Gp3zLU7kSpvq7byetFBt9N5Rccku9ThNRLY3RIhMuw0K8CXEQguVOYyaT3KfogY5WFWO6LUzaTY/Z7A6tRXWzvX7vZHOC5CjTrodsW5nc4rOiZfdDGU6rjDaySwuyVCZNhpVoD/CvgE8PuxbR8F/lfXW9RHRn2Y1c1h5x23XNPz8w+S2rl2FdXJzuMitNP9ApoaPe0k0q2Id/S3Vn319WgksIOMENlLNCvAPws8Ydv2PwTGbNv+M+AocFfPWtYGuYKLBixDYRhBgUi9g42+14ZZve5whtmfXjPXbtIEpcq5dqtFaKf71czoqZEF3eq9acdXv9Nnwc4CO0od7KjSlAA7jvNyGPXwQ8DngTeAzzuOs97LxrXKeq5EyfVRCpRSGAYkDBMrYZAwFYZSWKaBUsHigx+8+TCf3kPDrF52OMPuT68lOB/9wHXbtsVFaKf71UxnNuwurp2SFA1rR7rbaDYK4jccx/lZ4L9Wbf9PjuP8k560rAO0DmJafR9cXAgjmCJhNg1ImCZHDk3ziR+wefqbZ3nrUo4DMxk+cOMh3nHtvsFeQI/odOLuzcUNvLCe2sG5sW3D6WEWG2ica7eane7X4ydONezMRtHFNewd6W6jWRfETxC4Iar5BDB0AlyPuDCXXBcKMDeV4Ue+/9pAmM0gH8Jm0cVSBqYJhmEEiWvamG4cNkuik4m7kuuzWXBBQ6GkOb9cWcRzcSWPUnB+bSsf8EQ2MZRic3JhkeNfepXzy3lAs382y723bc+F3Oh+NerMRtHFNQod6W5iRwG2bfv+aL/YzxHXAo1r34wAW8KsKZW28tdGFrNlGlimwgz/DfLUagxD1RXnYbUk2p24W90soQDDUPhakSu4TE+kyi9mOmny5uIGhLXWPM+jUPK4Ym6sZ9fSDicXFnnosZfYyLtlV9TZixt18w03spp36sxGMZJgFK32UaaRBfyJ8N9k7GcI3rHzwI/3olHDQiTMRd+jWNrariyLSyv5IN2iobAMA9MKJv1MZWAa8PTJs0yMJbEMA8/3UUqNpCURvZCu61ekh3Q9v+LFzBXcio5Ih//LFdy+t3knHj9xinzRQykwlAIFvg461Va/m2Y6s3jVkf0zaT5659Gh/v5H0WofZXYUYMdxbgewbfvfOo7zr/rTpNHAD8vq+r6mhF/2M0NgOZ9Z3GAsm8CyFGkzgWkqPE+zkSuhNRhGe26NfnJyYZHNvMvSWgGtNVqDZSp8DZZpVLyYqxvFIKArtCpVmBR4tWoJ8aBZXMkHo5dY9JmhwPP9CiuvOtMbWpMv+S0lzolGQOWqI27r1UH6zSha7aNMs1EQZfG1bbsieNJxnOF/qvqM1pBOmiyvFUhaW5aE5/vMjKe4sJLDMMAyDCzLCCxnI3RzNAid6xfx5C+FUmDF+hpKro9hQCaVqnoxVeAzV1vK5uugSsYwMTeVZnWjWE4cD8F1mYZR7kzi4qlU4KIAxcxEsmlX0qj6UiX+t780GwVxkCAXxN8Cpqv+bG4/Qvib7zzIo8+8RhFImAYlL7Akbn7bfnw/itDwoBjUYlOEoXMmJM0gdM4yInFWZX9lv4gEJJs2SVgGqxtFCqWobpxifbPI/liy9eocvlGe4QOzjdNa9vplr7BmE0YwGilpfHR4TzXpdFCT7OTCIr/z6IsUSi4Jy8TzNUoF+ZTXci4HZpNNCeko+1Il/rd/NBsF8dvAJnAH8BcEQvyvgS/0plmjj314hnuAL3/zDMtrBWYmUvzNdx7EPjxTc/9AsDS+C24YoRGJslJgmgYJ08CyAmvZNBSWWdta7oaoxQUkk7LIpCw28yUuruSZn8mUh6eRNVidw9c0FOmUyb23X1f3HDtNVEJ3rLBaeYsTpkE2ZbG6UUIpzf7ZMe697QgQRDbkSx5mWFg0iOhQKMANXQjNCOlu9qUOW3TPKNOsAL8XOOw4zoZt29pxnG+Gq+KeBf7f3jVvtLEPz9QV3GaIRBkNnu9RLFVZy2E8c3mhiaF45Y1LPPKX3wFgLNN+9MXcVJrzyzlyBbccVlYK/601rH7gYzdW5PBt5sWsN0w//tQChZLXlQiSenmLp8eS/PL/8b6KfaNyTolYySYFeF5QtslqIWH9bvWlDmt0z6jSrAB7QDSdfcm27XlgFbiiJ60SdqRsLXvgepULTf70K9/FNA3G0lZY8Rg28iX+4htneOd1c027Ma4/PM0rpy+hwuKYJdfH8zXZbKXHqdWUiHHraWW9yNR4osJKTFoGZxc32Ded6Yr/tBVXQLRvVLLJ1xrDAM8P3CkTGYtCyWtKSPvtS+2XVTqqvu126Mc9bVaATwAfBP4Y+DPgj4Ac8FxXWyN0hNZB9EU6ZVEIfcumoUhYBhdX8iyvFTDNwAdqKQPDCCafDGP7Z7186hKT2WRFkU1cj0Kxcs61lWF1tfW0ulFkea2IUopsOlH+PFDl8kgR7fpPW3EFRPtGbVndLIXJzxXzU2nyJZ/psWTTL2K/fKn9tEq76duOBG5pvcjsePP3tR/06542K8CfIKhcDsHKt38OTAD/qWstEbrCzESK1VypHH3h+ZrcZpHJTIJ80SMYzATEc2Zo02QtVyqLs+drLp/LBpOFno/r6bIPuF7ymkYWQ7X1ND2R4uJKjpX1IpmUVf68/TPpmnXt4lEKzVomrbgC4vtmUhamGbgi7hvy2N1+WqXd8m3HBW6iA1dZr+jXPW0owLZtm8CvA/8bgOM4OeDfdq0FQlepF33xN995cNu+8aXZRddnI7e12sRQsJF3yaYSjGUSmIZBJmWybzJFJpXgwqUc81Np/vZ7DvG91841ZTFUW0+ZlMXsZJpLa4WKQp5Qf5nvyYVFHvrCy+FqRJ/VjWLdVWzxczcj2KMagtXPiItu+bbjAtfNQqvdol/3tKEAO47j2bZ9FzD08b6f+vyLbObdcNbeJJOyyIYz+NU/R3+3zBrj7wHinFpuOnKiFq1GX9Tjlrcf4NFnXmMz75IwDVzfx1CKD7/3ao5eOY1hGFhGkD8jX/L4yrfPMzOZxjSCBSfJhCZXcPmz2AtVMyevaXDkiqlyXbM4tYTw53/3BBu5IkoZmIaBr2EjV+T4UwvbjvnRu67nqrlsS66AUQzB6mfERbc6qWEP0+vXPW3WBfFrwL+xbfsXHMcpNdx7QCyvFXhrOdfSMUnLiIlyIMyVQh0T77RFJmmSyabQWqNUdxcZOKeWefSZ14LiiymL1VyJR595jXugZRHeaf9aIn/rbGXOhnpCfu3BKVxPg+fFF/+x8OYlxrNJkpaJaSiSpkE2ZVJyNcWSj2nCD733ao4/9SqGqTCVIld061pP9YQwWtYbXxbtacXZi5vbLPDffuQkH73jupET1Fbpd8RFNzqpYQ/Ti9/TlGVQ8jRozQdvuaq8T73X3/c1Oowxj1aQbuRKU2OZxEr1vs0K8D8GDgD/zLbtC8TquDiOMzRxNXfedCVvLeXIFV1yeZfNgkuu6JIveMHPBZeSu30SqegWWWlxyayhIJ3cEuVs2iKd3BLqass7HRP2elb3l795BtM0yv7bpGVSDLd3Es4Wp57IT0ykOTiTqdi3lTC6yWwy9D1v3d+i6zGZSbC0FmRKW14vYCjFymoBw1Tsn8lw1+2HOXpohlIY9hUkOAqe7NoRGwqlKv+gVBQqVumz83x/aIa0vWQUXSdxgbNM1XR0SbtUi2UY3QlloQxDPglWRl5/eJZP/IDNV184x9J6kQNjCW5+236uOjDJynoRHWad8oNPAK3Li4+KKJYv5cPEVEHSLixrrBMB/ngH19433nXd3DaBrRJyhLAAACAASURBVMb1grSKubxbFupc0WOz4vdArHOhcG/mXfLFymQzvib4WxvJZhKWERNosyzUby5ukLRMSiUfw1DBBJkKhmv5oksyYVYs9W2HeiL/xIlT/MTddtsukGrf80a+xPpmiYuX8vz8p04wnrEoeZpM2iKZNCl5Pm8ubnBxJc/SVL78gkTXbBlGUNUkjG9WSmGguP7wFOcvBWkkta/xwnwcWmumxpIopQK/tgbTMFleLezY7t2yqGDUXCfxTmNlo8S+yRQ/cNPhilzc8Uddh2Lm+5VCiQatdPnd1KEQooOl8L6v0eExPqC1vyW45Q+P/in/gCZIVftD76ss1bWRb+wA0OEzGf+9Hko3Gxg63FwNvPbywoWGAtwuvtYUS4FQJ9MJzr21Xraq4/9FopwveOXfu9UmFVrd2Srh3nKPBOXVM0kz+D3298jq/uWHnyedsircJ1priq7PD916Vdk6jk/g3fO+a7aJcC2hhkDgowUcvg5GCkoFsbQAMxNJMqko5CywkH/yw29v+LlPf/NMeci6ulkArYLyU6YilbSYyCTw0SQtK8h0ZgTXNZ5J8mN3Xx+2IxBxZQQLLJw3lvmvT75KrujhhT7uVMLgvrts3nb1bM3voFuvS7vCPz8/0ZWSRDvRTB8fCViw7DwugFC2LeP3KlyfrmPbA0HVTE9lubi8ge/pMH9IdBLwwo62LJh6u1AOG7OzYywtbZR/NwzF7MzYFZfPjZ2p3rdZC3jPYyhFOhm4GWZnxxhPNp8Cw/X8skUdiXKFaIdW9+KlPOeWNsMHLbDi4ugwvWM7KR4Toa87X3BZy5XC0kyqnJVtPJPkT7/63XAZcRC+ZhkGWuttLpC6vur3XcNPfvjtfOpz3+aNt9ZRoQBD8CIBrOfdsgAnTIPltcKOn3v8qQVQikzKJJkwKXo+GoPxjEmx5FeIdHBsqdx5oOFvXL+ftSr3UiQw/+3PX2VprYARCrKnFCXX57//xQJzU5lgPwWKyAKP/x7+HC4VV+EHL5xe4asvnmNlvcjMeJJb3nGAo1dOh0oR7PjXbyzzha9+F8NQHNiXpeT5fP7Z1zHN4EP/4utvsrSaZ2Yyxe3fd4gbrt66957nl0Uq8i+WRa1s5cX8j8SEMBKvGsIYER3nVUtbqH5b56ohftWWZD2qjjWTCdY3h3ZqqaeIAPeBoDpEkols433jFuD0eJJb3n6AQ5eNsxmKd1mwYxZ3ruCytJpnabUQCA/bLbWS61Nyt4TI9byKv2/m46Je+TIsrRT4lT/8etltcmZxIxBoU1M0/HAiTPPE104xPZ7i4koeNyxdVI0XGw2UvEBAI2q5R1ZKgUBPjSXL2wDG0gl+7u9tWc7AtknDD77/2m1+7fi9ObO4GfjoqjK4bZ5zw5L15SO2fUYt4h1IwjS4sJLjpVPL20YQjz79Gmt5NwzB8lAoSp7Hf/vSq5S8II9G0jJ561Keh//nK/zgLVfxPYeCHFiuMlhe2fIvVl+TMFoMlQDbtn0U+D1gH3AR+DHHcf56sK3qL/UmvmYm6h8TvfiT48my9ee6HnfddJiDc+PbxPr0W+u8dnaVfNHDNBXj6QSGaXDhUq7CdxWhgaXVAktU+1MrRXxto8SvHz9Z/r3k1f6slfVC2cp/21Uz/PXpS2RSFosreTIpqyLCxKvRnmrLOaL63lUPBbcTDp2rfI1K6bZ84c1Ooi6u5EmnrLJrKlcosZ4LXFUJUzGeTeCH/VLR9fnSX53myMGpcvtqfUfCaNLsQowngR9wHGfnGY3O+S/AJx3H+bRt2x8nyML2gR6fc+Sp9eIDPPfyW/zkh+eb+ozZ2TG+8o3T/I+nv4NhGpjhcNz1fN5tX8bMRKos5F9/ZZGCG4hvNCkShN40JshNsWVt/+U3z/KX3zxb/n01HIpGFSsiy25pNV+OkPC1Zixl8eLrS1y4lOPb37nI2maJmckU3/+uK5qO3Ng3lQk6ndBVEunaeCbRVjjg8lqBdKrylarVWcRXK+YKpSACJzy3r3U5IieTStTtbITdQbMLMa5haylyT7Bt+zLgRuDOcNNngf9s2/a84zgXennuUafZF78R9uEZfvj915Ytv9k6lt81l0/WnKz70K1Xc+Vl4+SKHs53l/nKC+dY3YzExOTKyyZIJ60KH3gwYenW9Hd7sXF1vrjdZfLpJ16p2LayWeL1xx0yKYuJbILJsRQJU9UMBcykLN5tz/Olr5+hVHJxvWBCL5lKkEpaeFo3tGSrreRkIrgX8ST81W4WqIwYWc8FhU6VAitMxq/QZV95reOF3UOzLoh/A/yWbdu/AJymMg64W2EHVwJvOo7jhZ/r2bZ9JtzelABPT2drDll7wezs8BSbnJ8dY2U9TyqWwKZQ8pmfHWupnbOzY9w6O8at7zq04363zo4xMZHmiROnuLiSY99UhrtuPsw7YrP41187xw/f/j3bjn1hYZEnTpxio+AyN5Xhozcf5u3X7qNQ9HjeeYv/8ZevsrIeWMGGgmTCxPc1lmXgun4505uvYXktX9P3GYl7q4tyPD9IXrS4mg9W+YWJiqIwuPOXcpxa3GQsbXFmcZ3Hv/JdElbgMtgoBOGMWhP6cIMFEWj44PuvrfgeKu9fnoRlMDUeiOxSuBLMLflcuLSJ62lSCZMzy7ny/R2mZ69b7LZril/PTlElTYWh2bYdiey2+VPHcbpSEcO27XcDv+84zttj214EPu44zvMNDr+aHoehxWnsW2yfdnyP1ZM/O4WP1Tv+Ky++xYWljbaXLjfDTu2EIIphfbMUVqEIBNEyFZ6vSVkmP3rHdRXt+uWHnyeVNNGociyw52sKRZcPvPsQWhksXQpC4nLF2ORl0SOXdytDnjrECKMhDEORMINk7pmUyTWXT3Ll/okaS+JN0kmLhx57sSJ5Ur7ocmmtgK8p+4OjpED3vO8abn3XoZ49e4Oil+/TIOhFGNo1jXfpmDeAK2zbNkPr1wQOhtv3BO0uRe4k/0N0zmTC7Gj5czPsNEkFQV4IXwdRANFIxvU0lhHEDFe3a8uXGlQIgWC/A7NZ3vuOy3d8sbUOlkiXhTlahFNwOXV+jRdeXyIIOAsmAn1fk06auJ6OlWbaIirS6vmasIQeRdfnG69e5BuvXqx7T8rJ341gwQkEVk4yYZBOWIErxgtCEp987g2uv3aOouuRCMMIhdGm2aKc3wWwbdsA9juOc7bBIS3jOM5btm1/A/j7wKfDf7++l/y/nSxFbrf6RnTOVCIQl14sf45o5KtOl1NAVsZAa1RQisk0KtrVSua3apRSpJImqaTJ9Hilj/WmG/bzjh1GIp7v86lHv81qrhQmBApWaJU8j6RpcP3Vs+SrIk82C16Ywa3S6o5GbL6nIRY1Uiz5FEuV8csr60X+r998BgjcHIFf26pYVRlflLO1JH4rMVU6aZWFXhg8zRblnAZ+E7iXIEh0zLbte4C/0eVy9f878Hu2bf88sAz8WBc/e+jp1mRav87ZqrukOlcxVE5SreZKjGcSrKxXnztY0VbdrlYt/2bb22g/0zC47cZDPPrMa6AgZQVLq5UydnT7xK3u6njuXMErW+Mvvr5U9h/7OlgdVu0t8XzNWq7EWq71BQzpZOUqyurJya2/Ve6XsPaW1d1pZsJmaNYF8V8IBPEq4MVw21eA/wfomgA7jvMycHO3Pm/UaCRQcbr1cETnTMQm8KJz7nSOdtwljSzW6PMmx5IsrwXWX7SEeCVM5DM7WZktq9ryd04t86nPfZvltQLzs2Pc+rbLsA/PNN3eZvdrx+2zk9Ud521Xz2zzlbuux7Fr93HqwiYXL20ylk7wPVdOMzWeqlhVWbnSMli8U21154se+aLXcsduGmrHrIEVGQNjv6eTVtlFNCp0KzNhI5oV4DuAg47jlGzb1gCO41wIQ8eELtHskLrRw1FObBMufQ2KeG6tpY/zgXcf4k9PfDeohBGeM4HB9YeneeK5N7BMg5nJNK7v8+Tzp1lcyfHdc2ucfmsdZSjG0hbZVALDVHi+z8unlrnpbfu3Tdcq4Ka37Wd6IsUzJ8+yvF5kKswwdeSKaQxgLJPgKy+cY3k9z77JFKubLiXXA4Il0+jAN3r6wjrXXREsTPD1VoKVV964xP/8qzdQSjE9kaJQLPHF8BpOvHi+nInO18HESIHtrpZW3ECdFl2tRy1xv/bgPv7qlQskEybj2SQlz+eF1y6GFveBup+ltabk+jWFeXs4oFexrTr0z/M167kS621Y3alEkDGwIgFVmLtkdjoLvj9UVveXv3kG19Os54t4ro9pGaQTZtddc80K8AowB5R9v7ZtH47/LrRH/Nm6/qoZTFPx7LfOsbJeYH4qzXu/93KOHppGGWAoA0PBt76zxGUzWVJJs5xcpuRq/vqNS9z69gMQZhQzDBUU1TQqF/zHh7Nz0xnWNgr8xclz5Asu6aTJ97/zIAtnVhnPJMpCpLVmdbPI0yfPYRoKHwWe5uJKHn9Ck0pa+L7Pdy6tkk3Wf6zeeWSOd8bC1eLX/56Jy7jp+q0+/T/8wXO8dSlIlGMSlK4xjMAP/J6w748f/6nPv0ih4KEVlAyDqYkUyYTmm68uUvJ8ZqfSmMool2KCIGRtIpsIhNwPAnLnptL4/tbwP5kIsrsp1b8lv9Xi/qnPfbstX71SimQiyKMxtYPVXQvP1+SLMZEOMwdGP+erRD2K6d6sYXUXSh6FUntWdzomzHVzdcf93OH2Tqzu88tBWltFuEjH89nw/K6vQmxWgD8F/Hfbtv8lYNi2fSvw7wlcE3uW+MsfWZvBz2G9NUNhRolcDBWKaJRecSupi4p91tzU5dz69stj6RljyV9Czl3cKC9lLZcV0ppzFzdIJrZHBe4UavithUX+/OtnSCVNJseSFF2fP//6m+SLHjMTKYqlrbC+cxc3yzke3NiEUb7occX8OIWSx77J1hJqVzct/vvSWpHJMMVk/FrOXtzctv/JhUVePb2CUiroIDRcXC0wM5FkebXA3FSaxUv5igTghZLHzHiSsbAIJ8BY2mItVwqGzSpYeVfyfOanMsxMpEFrvDDczXV9XN/HD5LCVmQF6xaRG+i1s2tYVvAsJcLMdr2eHzANxVg6UXF/mkFrTcnzYz7uLXHenozKo+j5rG0U2cy72yJMPF+zkStVlMtqllTCLLtK0qla1XHMSiEPLfSkFUwEE46UgHI0jFdjeX0nNCvA/xHIA58EEsBDBMuEf72rrekStUYscYHc2ra1vwpVULE965UKBdMILcvp8SS4XjnTV3C+KlGtyi/QKfHPmBxLdlxNIEqHuPDmKkrBvukMprVVm8vz3G2FMV1PhxaBwjIpi3AUmtXthNqtVE14/MSp0L0QtMlQUd6JItcenKxbNeIH/kZle+949yE+88VXyBcrC49+8ObD2yo1Q/A9+7FO0PfA00ERUy8SaL3VCbbyLMRdTVYYrra0kmdqPEk6aQ3tKjmlgmRCyfGdfd0R8XDBKI57s172wHwU0+2V47sji7ye1X1pvdViC6r8HPnhMw/B8+RpzbMvnK1jjZuYtUqM70CzYWiaoALyUFdBnhxPoT1dzvcaEaUMDH4GdKVvtBY7vShjmSSb22bqmz++UzotQRMvoBnFr15YzjE7mSpbAJap8DyfAuC6HqsbgQWiCcKwTMOoEOFWyrXH23H8qYVymaH9M2nuvX2rhFAr17m4kmdyLMHyehFfb700rq8r2tUoB2/1fumEgZUw+fQTrzA3darimPo5fYOXMHq2vHCRiNaBWPta43k+nhdYivG/x4n7o8czFisbQSWGtc0ihqGaDrkbJUxDkU0nyLZpdZcr4RQ88lUx3nFRj/u+C8XK5JvVC3Tiv5Vcn88/+9267UglTMYyCVKJrVJnk9kE/+Ce7625f7NhaN8kiM39rOM4p5s5ZhCkLAPfaEL5umyd9ptOS9DEK9JGlpVSsLqxVR7+4NwYd998mONfepW3VgtYplFOWBNkvPTDwpiay/eN1SyquRNRdeONXJGozNDZpRwPPfYS93/ohooKD81cZ2Qtz06kWN0s4bo+Vlj2KNq/2aoR0X47VXoGGlaBjp4to3pIBBBa9UoR+pt9PD/o3MruDU8zmU2Gq+oCQdrIu5RKPpOZRM9WLI4icau7VV+372vyxUpre+HNFZ5z3gqMN4Kc3r7WgRuKreXurlfb6q6mIwEG/jXBwohfsG37r4CHgf/mOM5Sk8cLXaaTEjTxirST2QRLawVUOFte7UpYWiuWl9hOppOsbpbwfY3nQ9KCdDrJvbcdaen8JxcW+Z1HX2SzEExymGY47CN4EeJ13Jq9zshaNk2D/TOZchztvbdf19rNiRHvqICK0ulA3b+18r3oMBGPqQwC1+7WEHbfVJr1fImJbBLDhJKbouT6pCyD++60y2V6RtGIGCYMQ5FNBz7gqCDS0SunufbgZDkSZf9MpmaHtz3CxMWwTC5c3Cgve98pP02zLog/Bv7Ytu0J4CMEYvyrtm0/6TjOPW1dtTAw4r7VaKi3ulEEf8uVAIGFly95mCqYdFvPlZjMJsrFTSPfaqtuh+hzIRjeBRN7WxMdjUqT71TOp1ZZ+napVTrd83wW3lzB8zUJy2RybGu43O2y6t//roN85ouvsLZZCkZ3BB3Vh2+9mvnpTODK8HW4/Nkvpw+t5c4QWqeZMMOEZZCwkkyGBQOgdi6IerSUkN1xnDXbth8GLhFMxn2wleOF4aDat2qaBjMT6YoS7g8+/HywECDKVaDA1yqIjphMMz2WbNntAFtWZcIyKMTiTKOhXFD3rn5+p53cAtXWcqf106onAaPKI6ZhkLAUruezFEYhZNOJrpdVb9SpGEphmIqECemEWQ6Ti4bLpfJEYOBzjpcpEoaDZn3AiiAx+seAHwG+S+CG+ImetUzoGc1Yi5H1F7kooomtUosTftXEP/ei6wUhXHF0YI2fXFisaVnv5BbodlXg6o7q0loBUEyNB6FxS6tBOszVjVI5Y1m3y6q30qlEVm9UgDVpASnKwhxZy37obvJCi9kXV8bAaNYCPgOsA38IvM9xnJd61yShHzR6sSPrr+yi2CxRcn3SCYv77jzatthFnwtBTgU/psCWqZiZSGGaRl1BreUWaGXo30o14uqOSuugqnN0T2Yn06ysFyi5XltRIP0iElbTUOXFCXGL2dcaN6xI7Ho+bikmzGIx95RmBfjvOI5zoqctEYaKuPWXKWcp8zsS3+hzH3rspbAsUeWrPZa2yKYTaF3fD9xsbPDJhUWePH6SsxfWy0ILjSMXqol3VA8+/Hy584DgvhiGatsdM2jiURpJK/RTJkxUpr7FXBJXRldpdhLuhG3b30Mw+XYF8Cbwh47jvLLzkcKgaMXSq0WnoW47fe7kWJKNgktVYWZWN0qkwnSJ9XypzcQGR37iVNKsENpUwuzIfdFp/PWo0Mhijguz6/l4rqbke+X6gOLKaJ5mfcAfBj4DfJ7A/2sDX7Nt+xOO4zzaw/YJMZoV1UYTVc3SSajbTuSLHlaYpwJ0eQJOA5fWCkyNJ+uKWjMdQ+QnTieD5dqR0J5f2uTyucrSN624L3rVKY0KtYS5Ip45FpXhuj6lUmAx10qnGeeFhUW+8PR3epr2sR2GKR3lvwd+2HGcL0UbbNu+DfjPgAhwH2hFVPs5UdUOc1Npltfy4bLNymXNWtPQzdGoY6jnJwa1bXl1q5ELveqURp2gmKjCMhSWEeRIVplw5Z/n44ZuDNet9C87p5aDlWWKnldkaYVhS0d5CPhy1banw+1CH2hFVDudqOo1d998mIU3V/F8Xc6nYZqKiUyCA7PZjgUu8hMnq4R2/0yaguuzWvTYzJdwPR9DGdxyg2RV7QWBKAcLVkwCUYaYtexpXj61zMHLxvB9jesGLo1cweXZb50dqAB3Up2mFZrNHPEN4J9Xbftn4XahDyyu5Lclg6knqnNT6WAlWIxux6h2wrEjc3zo1sNBPgMdRENMZBIkLKMr/tS7bz6M5/nkiy5abyUKuvf263jfOw6wnitR8jSWaTKWsXjmhXOcXFjswpUJO3FyYZEHH36e//M3n+WXH/46L76+xKunV9nMlcgXPLTWJC2DmYkUyjSYnkgxnk2QSVpY1lYWwX6wvFYoZ52L6EX2uWYt4J8GPmfb9s8RFMm8EtgAduUquE4nsHpBK5nBWp0sqhUx0Ovrvef913L15ZM9uc/RZzz59TPbrunxE6fYN5XelpZyWNwzu5V6LrR0wiBf9DANgyjhZKEUhPWlE+Z2H7MXFEh1w9SQbli0wKe7E4CtVKfphGajIF62bfsG4FbgcoK44BOO47SepHPI6dYEVrdpRVRbmSyqFzEQ/5xe0Q1/ar3O8tiROe645ZptixaG3T2zW6nnQkOpwPWgvB2f67KP2VRgQrTwNx7L7PthrmYvnAT0g4xz7QhzJwVfW6HppciO47jAl8PKyEBQJdlxnOq1TCPNsE5gNSOq7Vju9SIGBn29zRDvLJWC75xZ5TeOn+Tgviz33n4dd8xPbDumeiSxvJpnLVdCa/hHv/aX3HXTIe55/7X9vpRdT72ObzPv8jN/71380RMvtzUSiscyG6YKBC22AtAPM8z55QUn2ycCa4lzOzX/2qHZMLQbCZKxHwOiMW9U56b+wv0RZJgtpJ0sxnYt90Feb6eunqjz8H0d+uYCP+Fbl/J85ouvMDWV3ZaMJz6S2MgVWdt0ATCNYOj76LOvAzQlwo3aP4yurEGxkwvtPTfs7yhpUj2iTHNW3Jcbc2l4ftyd4eO6mpLnlYW5VzX/4jRrAf8e8DngfmCzwb4jTSu+1mGiXcu9XsRAs9fbrgh1w9UTdR5vLeegIqZYs7Je5Pcee5Gf//H3VBwTH0mcD0sbWWZQegiCXMdPfO10QwFu1P52r69d0R52sR+2RSxax6xmkx2FudSExdwuzQrwVcC/DCtj7GqG7UFplnYt2eh680UXQ6mWrrcTETr+1AIrYfUKKyxFv1MOiFpEnYcbZIivyKjm+ZpT59dqJvWJRhI/+R//PFgQEptaNxTki27Dczfq8NrpEDsR7WGct4gzKotYGglzeaFJk66MRjQrwH8M3AX8WeunGC1G5UGppl3LfaeIgUa0K0LHn1rgzOJGuYCm62mWVvPMTKRacn1EnUfUcURoDShNooGgp5MWhZKHGQtt8nWwvRGNOrx2OsR2RzHDOm9RzSgvYokLc6KGMEfJjEquj2UErrBmRLlZAU4TJGR/GjgX/4PjOD/W8tUMOaP4oHRiudeLGGhEtchs5kusbpR4a2mTBx9+njOLG0xXhe0kLYOzixvbCmj6WpULaNZipyH2p594ZZuweT5MZK0dBe+umw7x6LOv4/mUyy1pNHfd1Hh9UaMOr50Osd1RzDDPW+x2ImGOkhmlEybz+8bA8yqWZScTtWulNSvAL4b/CUPKICz3dNLk3MXNcq5gT2sUioRlcGmjSK7oYW6WKqoFBJaqql1A09PbOoyTC4sc/9KrnLm4iWUaTI0ntw2x56ZOsbyWD5LBEFa5BnIFj2su3z65ExfzmfFUsDDD9UknraajIBp1eO10iO2OYkZ13qIZht23XY9qa3nfVOZCrf0aCrBt2ybB4ovPOI7T3WUgQlfpp+V+cmGR1Y0inq9RCkqh/9UwYGo8FQTH+z7La4UgkbkK4inTKYv9M2lcTUUBTdNQHNiXqWh/5NtcWQ+qAPs6WKE0lrbIFTw++ci3OHLFFG8ubjAzkWJ5vRhUuiYqdbRd8Kr9pUXXZ9JQLafZbNThtdMhtjuKGdV5i0aMgm+7BWpOLDQUYMdxPNu2f9VxnIe63yZhVHn8xCmymQSppBmIaJhb0gx9X0trhbL/S4f/84NEsrzn+st45oVzFQU0o6XC1ecwTaOcM0KpwEpe2ShhmcF5Lm0UyRc8TENtE/TpyRSPnzgVlpQPBLCb/tJGHV6rHWK7o5hRnbdoxKB92/2wvpt1QXzOtu0PO47zua6eXRg64g9dOmGACurAVT+Akd9RhYU9zy1thnXHNKubJRSBTxUCq8zXQbhXNpPg5VOXuO/Oow0f7ugcVliXLrJsAwJXRyphMp5NsJ4rsW8qXRb0zbxLseSFVT22rKdC0SOZCNb0u66PZRlMZHb2FfeTSLSj7yHoPE41fPlHcd6iEYOOUe+H9d3KJNxx27a/QuCOKL8Hu3ESbq8Sf+hAc3YpB2hmJ9M1/K6VfsfJbIKl1UIQ1RBGJGggioE3VOASiF6gZgQjOke8Lt3WrLJmciyY4JvIJvA8n+mxZFnQLRWcNEh5uWU95YtFNvIuSgVt8jyf5bUil89munYfO2Wnl7/W6r7dyiB92/2yvpvNhvYCQU7gLwGvAgux/4RdQvyhW8sFIqWUEZRFD6tJPH7iFLCVcaxQCrJYmabBWNrispkMRhhelrCMcoxtYAEbLb1A0TlM02BmPFleLGGaitnJNJlUYD8UXZ+Dc2M88LEbefCn38sDH7uRfCmI0zy3tMnpt9Y5t7SJ6wYz0+FUXdi20K7uV5qtJoh/D0qpbfd+r1D9jEVZ7frh224l+2AnNJuM59909azCUBIf8rmuH/O7BhZt/AGs5Xf86Aeuq1iAUXJ91nKlsuhlUqmWXqDqc1x7cJLrD0/zzAvnynGW9Sac0gmDs0u5bZYuBMU01zaDfMCWaTCRTZMvetvOPygkrCxgkL7tflnfzeaC+EC9vzmO8+fda44wSOIPXeR31XprLX31A1jPjRB/cbzFDTxPY5mK/TOZtmrTVe/fVBrLskWryhmzIBBc0zTYP7sVnhalPxwW+j30HuZQr0H5tvsVWdKsD/h3q36fJ8gIdxqQ1FG7hPhDN5GxQovRZyKbbnn418sXp5nPzhc95mcyXForbLN0Pc8f6pCtfoaV7bJQr67RL+u7WRfENfHfw9jgfwW0tnRKGGqqH7rLZzPlKIjpseRQWUaNmJtKs54vbbN0r5jbCkcbRosP+jv0sbtkKAAAFctJREFUHnSo1zDTD+u76XzAccLY4H9HYAH/anebJAyS3RLOdPfNh/nDJ1+tmei7W9fYy6F7v74H8TcPlrYEOOROYFclYxcGQy+E7NiROaamsm0n+m6mzbth6L6blzGPAs1OwlXE/gJZgtjgn+lFo4S9Qy+FrFeJvmH3DN136zLmUaFZC/jjVb9vAK84jrPa5fYIe4xRFbLdMnTfrcuYR4VmBfgmx3F+pXqjbdv/zHEc8QELbTOqQrabhu67xe8/ijS7Eu7n62z/V91qiLA3mZtKVyRTh8EJ2cmFRR58+Hke+K1nefDh5zm5sFh330Gu0hJ2DztawLEFGKZt27cD8fWa1yJhaEKHDIsPslVftAzdhW7QyAURLcBIA/F0lJqgMsY/7kWjhL3DsAhZO75oGboLnbKjAEcLMGzb/n3Jeib0imEQslH1RQujTVM+YBFfYbczTL5oYe/Q7CRcT7Ft++O2bZ+0bdu1bfsfDbo9wt5DJtWEQdDJSrhu8g3go8C/GHRDhL3JsPiihb3FUAiw4zgvANi2LUubhYExDL5oYW8xFC4IQRCEvUhfLGDbtp8H6jnT9juO05VyBPv2jXfjY5pifhfW5pJrGg3kmoafZq+nLwLsOM6N/TjPxYvr+L5uvGOHzM9PcOHC7lqDItc0Gsg1DT+1rqeeIA+FD1gQBGHY6EeppqEQYNu2/z7wy8AM8MO2bf8L4C7HcV4cbMsEQdiL9Cvf81AIsOM4nwU+O+h2DDvDXDxxGHnupfM9S8gu7G76lSZ1KARYaMxuqcDQL04uLPKHT74KCrlfQsv0a2m6CPCIMKqJywfF4ydOYVkK0wgiLQd1v2TUMpr0K9+zxAGPCIsreZJW5dclyWLqs7iSr3h5oP/3Kxq1XNooVljhO+UZFoaDfi1NFwt4RGi3R27GAju5sMiTx09y9sJ6R1ZaK9Zery3DqCx9ZAFD9y2YRtcgo5bRpV9L00WAR4R2Epc34zeO9kklzY58pa34qNvxZ7cq2DuVpe8GzVyDpLgcbfqxNF1cECPCsSNz3HfnUabHkmzmXabHktx359EdH5C4BaaUIpUwMU2Dx0+c2rZPOmnV3acZmjlXO/tCe0P5Y0fm+KmPHGvpfnX7eiXFpdAIsYBHiFZ75GYssG5Zaa18TqvnHMahfDPXcP3haR77yil87WOZBtl0AstUkuJSKCMCvItpxm8c7ZPscLa3FR91q/7sdjqJXoehNbqGkwuLPPPCOcYyFrmCR8nzWc+V+NAtvY+CkMiL0UFcELuYZmZyo33yRbej2d5WZo1bnWFuZygfhaE16+ZolUbXEFntU+MpDuzLcuVl4+ybSvPyqUtdOX89JPJitBALeBfTzExu9POTXz/TUhTEo09/hye+dpp80SWdtLjrpkPcd+fRpiyvVmeY25mAXFzJMzWexPW2kjN1cwKs0TV0ewIusmqX1ovMjifr3q9O3DViOfcfEeBdTjN+42NH5rjjlmuazkj16NPf4dFnX0ehMA1FoeTx6LOvc897r+aBjzWX+K4Vf3Y7IUHNhqH1SnS6Gcgfj7iYyOzsTmlX+GWl5WAQARZa5omvnS6LL4CpwPOD7fe8/9qenLPVCchmwtA6EZ1Gx7ZjtdejVsRFPau2XeEf9ETnyYVFjn/pVc4v5wHN/tks9952ZNeLvwiw0DL5olsW3whDBduHhWNH5piayu6YjKcT0Wl0bCOrvRXLuxmrNvq8M4sb5IoeY2mLybFk08I/yJjl5146z0OPvcRG3kUp0BrOXtzgoS+8zP0fvH5Xi7AIsNAy6aRFoeRhxjTY18H2YeI9N+znqrls3b93IjrNHFvPam/V8m4m4iL6vOmJFOZmifXNEp6vuWJurCm3Sr9yH9TikadeJV/0UAoMpUCBrxX5grvrVw1KFITQMnfddAiNxvM1Wof/ornrpkODblpLdLJQopNjW12I0mzERfR5k2NJ9k2nuWJujAc+dmNTAtav3Ae1OL+0iedr4mMqQ4Hn+7t+1eBwmSzC0LDTEDny81ZHQfTK/9srOvHTdnJsq5Z33J2xvF5kpioKohvug37lPqjF/tksl1bz+HpLhH0NpmHs+lWDIsDCNpoZIt/z/mtHTnBrdSrNhs5V04lgtTPcj9wZteqNdct90I/cB7X4yG3X8WsP/xUbeRcfjdYAmnQ6uetXDYoAC9sY9Ix4L6jXqdx359GmQ+eqaVewuhkh0YvP6zfvuWE/93/ohnIUhFKa/bNjEgUh7E12YxavYepUuj3cP3ZkjtfPrm5zCY2SeA3K+h40IsDCNgY5I94rhq1T6abgRHknJseTzFnB5OAzL5zj6ssn96SojRISBSFsY5Az4r1iN6eGbDWqQhgeRICFbbSTe3jY2Y2dSoSUqxpdxAUh1GS3+eQGGWbVa3ajy2ivIAIs7Bl2W6cSMepREHsZEWBBGHF2s3W/2xEBFkYOyVu7nd1q3e92RICFkULy1gq7CYmCEEYKCbkSdhMiwMJIISFXwm5CBFgYKXbzggph7yECLIwUu3lBhbD3kEk4YaSQkCthNyECLIwcEnIl7BbEBSEIgjAgRIAFQRAGhAiwIAjCgBABFgRBGBAiwIIgCANCBFgQBGFAiAALgiAMCBFgQRCEASECLAiCMCBEgAVBEAaECLAgCMKAEAEWBEEYECLAgiAIA2IosqHZtv1J4A6gAKwDP+c4znODbZUgCEJvGRYL+E+B73Uc553AfwD+aMDtEQRB6DlDYQE7jvP52K9fAQ7Ztm04juPXO0YQBGHUUVrrQbehAtu2fwF4p+M4H2nhsKuB13rTIkEQhI5RtTb2xQK2bft5oF7Rrv2O43jhfh8FPgb8rXbOc/HiOr7f+w5lfn6CCxfWen6efiLXNBrINQ0/ta5nfn6i5r59EWDHcW5stI9t2z8C/DvgDsdxzve+VYIgCINlKHzAtm3/EPCrwJ2O47w+4OYIgiD0haEQYOD/A4rAcdu2o213OI5zcXBNEgRB6C1DIcCO48wPug2CIAj9ZljigAVBEPYcIsCCIAgDQgRYEARhQIgAC4IgDAgRYEEQhAEhAiwIgjAgRIAFQRAGhAiwIAjCgBABFgRBGBAiwIIgCANCBFgQBGFAiAALgiAMCBFgQRCEASECLAiCMCBEgAVBEAaECLAgCMKAEAEWBEEYECLAgiAIA0IEWBAEYUCIAAuCIAwIEWBBEIQBIQIsCIIwIESABUEQBoQIsCAIwoAQARYEQRgQIsCCIAgDQgRYEARhQIgAC4IgDAhr0A0QBs/JhUWePH6SsxfWmZtKc/fNhzl2ZK6p444/tcD5pU1AsX8mzb23X9fUsdHxj584xeJKftt5W/nb9YenefnUpW37fvaJl/mTpxbIF13SSYu7bjrEPe+/tuV7U68d7dLudQu7D6W1HnQbusHVwGsXL67j+72/nvn5CS5cWOv5efrByYVFPvPFV0glTQylKLo+nudz351Hd3zxTy4s8tAXXmYjVwQUSoHWMJa2uP9DNzQUjei8pmmQtIyK8wJN/211o8jqZpGpsRQT2UR53+sOTvK/Xr4AgKHA16DR3PPeq5sW4Z3a2K4otnvd0fl207MXsduuqdb1zM9PqFr7igtij/P4iVOYpkE6aaGUIpUwMU2Dx0+canhcvuCilIFpKAwViHC+6DU8Nn7eVMLcdt5W/pYreigUuYJbse+Jly6AAtNQKKWCf1E88bXTLd+bWu1ol3avW9idiAtij7O4kiebrnwMkpbB4kq+4XGe72MaW324AjxfNzy2mfM2+zfX9TEUuJ5fsa+vNUlVaXQYCvJFt2Hbmm1jO3Ry3cLuQyzgPc7cVJqi61dsK7o+c1PphseZhkHc46MJLM5GxzY6byt/s6ygDZZpVOxrKIVX5Y3yNaSTzdsc7d6bdj+zF+cThhsR4D3O3TcfxvN88kUXrTWFkofn+dx98+GGx6VTFlr7eL7G1xqtIZ00Gx4bP2+h5G07byt/yyRNNJpMyqrY9+Yb5kEHFrnWOvgXzV03HWr53tRqR7u0e93C7kRcEHucaHLnya+faSkK4tiROe7/4PUVURAHZpuPgoj22WnGv5m/HZjNctu7DtaMgrjmyjc7ioJopo2t0ul1C7sLiYJog902awtyTaOCXNPwI1EQgiAII4AIsCAIwoAQARYEQRgQIsCCIAgDQgRYEARhQIgAC4IgDAgRYEEQhAGxWxZimACGUTPUrif081z9Qq5pNJBrGn5qXM/VwGmgIhnJblmI8X7gy4NuhCAIwg5cA7we37BbBDgF3AScBbwBt0UQBKEWu9YCFgRBGDlkEk4QBGFAiAALgiAMCBFgQRCEASECLAiCMCBEgAVBEAaECLAgCMKAEAEWBEEYECLAgiAIA2K35ILoO7ZtfxK4AygA68DPOY7z3GBb1Rm2bX8ceAB4G/BPHMf5zwNuUlvYtn0U+D1gH3AR+DHHcf56sK1qH9u2fwX4uwT5BL7XcZwXBtuizrFtex/wB8ARgnfoVeCnHMe5MNCGdYht239CsOTYJ9CFf+w4zjfq7S8WcPv8KcHL8E7gPwB/NOD2dINvAB8FHh50QzrkvwCfdBznKPBJ4LcH3J5O+RPgbwHfHXRDuogGHnQcx3Yc5xiwAPzSgNvUDX7ccZx3Oo7zfcCvAA/ttLMIcJs4jvN5x3FK4a9fAQ7Ztj3S99NxnBccx3mRoPceSWzbvgy4EfhsuOmzwI22bc8PrlWd4TjO047jvDHodnQTx3GWHMd5Krbpq8BVA2pO13AcZyX26xQN3iVxQXSHfwQ85jjOyArXLuJK4E3HcTwAx3E827bPhNtHeni7WwkNl58GHh10W7qBbdufAu4CFHD3TvuKANfBtu3n4f9v7+xjtSzrOP5xQuqgOKgESULrxW9FK8qZlSNQnM4EB65QIhZuynQ1e8HNuUSqYSWpm9iy1YoaTh3WXNlopS0PUM2puFWLvpKCQsd4E0pMcCT98bueuHk4zznP4Rx4eA6/z/Zs5/7d13293ef+3tfLff0uxjU4Pbr2gEu6Evg00UU8pmm2TElylLmbGC9tyzmHemxfDSBpLvBt4BONwqYAN8D2h3oLI2kmcCsw1faWI5+r/tFMmQYBm4Cxkk4srd8TgTOKPTnGKBOM7wKmD7YepO3lkr4v6TTbO7oL09Zjlq1E0jTgTuBi2xtbnJ2kYHsrMZk4u5hmA0+3++z6YETSrcDZwAzbe1udn/4iabikMyvH04GXyq9b0h/wYSJpG/AaB48rTm30pmsHJM0mukwjibK9AlxUJubaBknvJj5DGwnsJD5Dc2tzdfhIWgpcDowBtgM7bE9oba76h6QJwF+AZ4BXi3mD7Zmty1X/kDQa+DkwjNgY4iXgBttrG12TApwkSdIicggiSZKkRaQAJ0mStIgU4CRJkhaRApwkSdIiUoCTJElaRApwMuiRNEXS5j6EXyxpu6R/Hsl8NUj7x5IWH+10u8nHOEm7y0KW5AiRAtzGSLpJ0so62/oGtisHIL2Nki7s4XyfhO5YpHxIvwB4r+0xrc7P0aL+3tp+wfbwXJ5+ZEkBbm9WAefVWimSxgBDCe9fVds7S9jjDkl9XW4/nljosPUopJUc5+Q/THvzBCG4E4GnCIdAvwPeXmd71nYXgKS7iFVVI4D1hOP11eXcVwln7HuAmcALhH/TJyUtJxz5PCzpv8DXbS+pZUTSMMJH8kmSdhfzWYRD9NuAWcW2Arixu6Wnkp4HLrf9VHEOvxyYYPuvkq4GptmeIemkRnFKmgLcSzh4+RLwCPDDunSuB64lVvltrtgvBB6ulOGntudJuozw+TyWWOZ8ne115ZqNwD3AnDjUMNv76tJrWOe9Ieka4EbgVGANcG3lXu4nvIgtAE4n/Dh/3vb+8gJeAnwWeBm4o9TJ0G7yd8i9LXW6oRZe0mMl/QuA9xP/Z/OApcB0wMCnasvyy2rEu4mlxtuAhbZXNFPm44lsAbcxtl8DHueAJ7aPA6uJB6Vqq7Z+nyDE+VTigX1Q0smV85cBDwAdhHvA75S05hKCPL10TZdUrsH2K8AlQFc5P7wIxVeAj5Q0PwB8GLi5QZE6gSmVfD8HTK4cd5a/e4tzTCnfeGB+NQFJCwnhmFwV31KGR+vKMK/srnE/8EVgFLCSEKo3VC6dDVwKdNSLW6G3Ou8WSRcQwj8LeAvhkP2BumDTgHNKPcwCLi72a0pZJhL+kWc0Sqe3e1vhSmAu8SJ6B+EHe1kp1zpgUcn3MOLFdx/wZqJ+vluWHycVUoDbn04OiO0kQoBX19lqwoXte23vsL3P9h3ASYAq8a2xvbKM/S0nHuz+MIdoLW8tDnG+RjzEjcpSE9xJhPjUjidXytFbnK8Di2zvtV3zM3CCpDsJgTq/D855riB8PT9SHPDfDpwCfKwSZqntTZW0DqKJOm/EHOBHtteWHsNNwEclva0S5lu2d9l+gWiVTiz2WcBdtjfb3snA7DaxzPazxen4r4ie1aPlpfMg8MESbhqw0fayUua1wM+ATw5AHgYVOQTR/qwCPidpJDDK9npJW4CfFNv7qLSAJS0AriZcNO4H3kR0X2tUZ/7/A5wsaUiDll0znMHBW+k8X2zd0QncXsatTyS2eVpUBGcE0f1vJs5ttvfUxd1BtIavqNu1oE/5t/26pE1EK7BGj64um6jzntL+vyMX27sl7Shpbyzm+vs1vHJtNV8D4Y6z6nL11W6Oa2mPB86VtKtyfgjxQk8qZAu4/fkjIU7zgd8D2P430FVsXbY3AEiaRIwnzgJG2u4A/kV47m+G3jw3dXe+i4O3mhlXbIdg+++EiFwPrLL9MiEw84mWec1fbG9xdpePnUTLbJmk83opR8P8SzqBsutGL+nVwvenzuvTHkZsNPqPhlcc4EXgrZXjMxsFLAykV65NQKftjspvuO3rBjCNQUEKcJtTur1PAl8mhh5qrCm26vjvG4F9xKTIEEm3EK2xZtlCTPD1dP40SSMqtvuBmyWNknQ6cAsxSdaITmKLp9pww2N1x4cTJwBlD7I5wEOSzu0tfGEFcKmkqZKGEhNee4E/NHl9f+r8PuAqSRPLxOM3gMeb9D+9AviCpLGSOoiXQE/0dm/7wi+BsyTNlTS0/M6R9J4Bin/QkAI8OOgkJjvWVGyri60qwL8mxu6eIbrVe+hb1/SbhPDtknRD/UnbfyPE8bkS5gxgMfGC+BPwZ6JL3dNCg05CtFY1OOYw4qzm8RHgKuAXks5uIryBzxAz+tuJGf/pZQK0GQ67zm3/FlhIjJ++SEx8Nfs99w+A3xB19DQxebiP8FPbHT3e275Qei4Xlbx2Eb2Y24ix76RC+gNOkuMASZcA37Pd9jsPDyZyEi5JBiGSTgHOJ1rBo4lPxB5qaaaSQ8ghiCQZnJxAfJ63kxiCWEeMlSfHEDkEkSRJ0iKyBZwkSdIiUoCTJElaRApwkiRJi0gBTpIkaREpwEmSJC3if6/+mvKLeRxOAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Put together into a single data frame so we can plot easier\n",
    "frameMVP = { 'turnover rate': turnOverRate, 'Want to work for a long time': all_features.iloc[:,47]} \n",
    "dataForPlotMVP = pd.DataFrame(frameMVP)\n",
    "# Recommended way\n",
    "sb.lmplot(x='Want to work for a long time', y='turnover rate', data=dataForPlotMVP)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Now calulate for all the features in the questionare. \n",
    "\n",
    "turnOver_questionare = pd.concat([turnOverRate, questionare], axis = 1)\n",
    "partial_corr_array2 = turnOver_questionare.to_numpy()\n",
    "\n",
    "# Calculate the partial correlation coefficients\n",
    "partial_corr_results2 = partial_corr(partial_corr_array2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.06233445098667855"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "partial_corr_results2[0,48]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.067132823872621, 0.28363705247997695)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pearsonr(all_features.iloc[:,47], turnOverRate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}