day-5-machine-learning-part-II/05 - Homicides prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Crime prediction using statistical learning and urban metrics\n",
    "\n",
    "* Data description: metric distributions are non-gaussians;\n",
    "* Co-linearity: correlations between metrics;\n",
    "* Regression model;\n",
    "* Machine learning approach;\n",
    "- Underfitting and Overfitting;\n",
    "- Prediction vs. data;\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Libraries and settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T13:56:15.680968Z",
     "start_time": "2019-07-04T13:56:14.770300Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>.container { width:90% !important; }</style>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#sets the notebook width\n",
    "from IPython.core.display import display, HTML\n",
    "display(HTML(\"<style>.container { width:90% !important; }</style>\"))\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T13:56:15.705237Z",
     "start_time": "2019-07-04T13:56:15.684562Z"
    }
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T13:56:20.794020Z",
     "start_time": "2019-07-04T13:56:18.220477Z"
    }
   },
   "outputs": [],
   "source": [
    "# Standard Library\n",
    "import os\n",
    "import time\n",
    "import math\n",
    "import subprocess\n",
    "import random\n",
    "\n",
    "# Third Party Libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "import matplotlib.ticker as plticker\n",
    "from matplotlib import ticker\n",
    "import seaborn as sns\n",
    "from scipy import stats\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.stats.multitest as smm            \n",
    "from matplotlib import gridspec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:02:06.804173Z",
     "start_time": "2019-07-04T14:02:06.742587Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lgaalves/miniconda3/lib/python3.6/site-packages/matplotlib/__init__.py:846: MatplotlibDeprecationWarning: \n",
      "The text.latex.unicode rcparam was deprecated in Matplotlib 2.2 and will be removed in 3.1.\n",
      "  \"2.2\", name=key, obj_type=\"rcparam\", addendum=addendum)\n"
     ]
    }
   ],
   "source": [
    "def stdfigsize(scale=1, nx=1, ny=1, ratio=1.3):\n",
    "    \"\"\"\n",
    "    Returns a tuple to be used as figure size.\n",
    "    -------\n",
    "    returns (7*ratio*scale*nx, 7.*scale*ny)\n",
    "    By default: ratio=1.3\n",
    "    If ratio<0 them ratio = golden ratio\n",
    "    \"\"\"\n",
    "    if ratio < 0:\n",
    "        ratio = 1.61803398875\n",
    "    return((7*ratio*scale*nx, 7.*scale*ny))\n",
    "\n",
    "def stdrcparams(usetex=False):\n",
    "    \"\"\"\n",
    "    Set several mpl.rcParams and sns.set_style for my taste.\n",
    "    ----\n",
    "    usetex = True\n",
    "    ----\n",
    "    \"\"\"\n",
    "    sns.set_style(\"white\")\n",
    "    sns.set_style({\"xtick.direction\": \"in\",\n",
    "                    \"ytick.direction\": \"in\"})\n",
    "    rcparams = {'text.usetex': usetex,\n",
    "              'font.family': 'sans-serif',\n",
    "              'font.sans-serif': ['Helvetica'],\n",
    "              'text.latex.unicode': True,\n",
    "              'text.latex.preamble': [r\"\\usepackage[T1]{fontenc}\",\n",
    "                                      r\"\\usepackage{lmodern}\",\n",
    "                                      r\"\\usepackage{amsmath}\",\n",
    "                                      r\"\\usepackage{mathptmx}\"\n",
    "                                      ],\n",
    "              'axes.labelsize': 30,\n",
    "              'axes.titlesize': 30,\n",
    "              'ytick.right': 'on',\n",
    "              'xtick.top': 'on',\n",
    "              'xtick.labelsize': '25',\n",
    "              'ytick.labelsize': '25',\n",
    "              'axes.linewidth': 1.8,\n",
    "              'xtick.major.width': 1.8,\n",
    "              'xtick.minor.width': 1.8,\n",
    "              'xtick.major.size': 14,\n",
    "              'xtick.minor.size': 7,\n",
    "              'xtick.major.pad': 10,\n",
    "              'xtick.minor.pad': 10,\n",
    "              'ytick.major.width': 1.8,\n",
    "              'ytick.minor.width': 1.8,\n",
    "              'ytick.major.size': 14,\n",
    "              'ytick.minor.size': 7,\n",
    "              'ytick.major.pad': 10,\n",
    "              'ytick.minor.pad': 10,\n",
    "              'axes.labelpad': 15,\n",
    "              'axes.titlepad': 15\n",
    "              }\n",
    "    mpl.rcParams.update(rcparams) \n",
    "\n",
    "stdrcparams(usetex=True)\n",
    "figsize=stdfigsize()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T13:56:21.139071Z",
     "start_time": "2019-07-04T13:56:21.087362Z"
    }
   },
   "outputs": [],
   "source": [
    "def winm(x, y, nw):\n",
    "    xw = []\n",
    "    yw = []\n",
    "    step = (max(x) - min(x)) / nw\n",
    "    lw = [min(x) + step * i for i in range(0, nw)]\n",
    "    for i in range(0, len(lw) - 1):\n",
    "        if len(y[x > lw[i]][x[x > lw[i]] < lw[i + 1]]) > 0:\n",
    "            xw.append(np.mean(x[x > lw[i]][x[x > lw[i]] < lw[i + 1]]))\n",
    "            yw.append(np.mean(y[x > lw[i]][x[x > lw[i]] < lw[i + 1]]))\n",
    "    return (np.array(xw), np.array(yw))\n",
    "\n",
    "def f(x,a,b):\n",
    "    return a*x+b\n",
    "\n",
    "def box_plot_style(ax,bp,ymin=0,ymax=1,ylabel='y',xticklabels=None,fontsize=20):\n",
    "    ## change outline color, fill color and linewidth of the boxes\n",
    "    for box in bp['boxes']:\n",
    "        # change outline color\n",
    "        box.set( color='black', linewidth=1)\n",
    "        # change fill color\n",
    "        box.set( facecolor='#bdbdbd',zorder=3)\n",
    "\n",
    "    ## change color and linewidth of the whiskers\n",
    "    for whisker in bp['whiskers']:\n",
    "        whisker.set(linestyle='-',color='black', linewidth=1,zorder=4)\n",
    "\n",
    "    ## change color and linewidth of the caps\n",
    "    for cap in bp['caps']:\n",
    "        cap.set(color='black', linewidth=1,zorder=4)\n",
    "\n",
    "    ## change color and linewidth of the medians\n",
    "    for median in bp['medians']:\n",
    "        median.set(color='k', linewidth=1,zorder=4)\n",
    "\n",
    "    ## change the style of fliers and their fill\n",
    "    for flier in bp['fliers']:\n",
    "        flier.set(marker='o', markerfacecolor='k',\n",
    "                  markeredgecolor='k',markersize=4,zorder=3)\n",
    "\n",
    "    ## Remove top axes and right axes ticks\n",
    "    ax.get_xaxis().tick_bottom()\n",
    "    ax.get_yaxis().tick_left()\n",
    "    \n",
    "    ax.set_ylim(ymin,ymax)\n",
    "    ax.set_xlabel(' ',ha='center')\n",
    "    ax.set_ylabel(ylabel)\n",
    "    if xticklabels is not None:\n",
    "        ax.set_xticklabels(xticklabels,rotation=45, ha='right',fontsize=fontsize)\n",
    "\n",
    "def test2samples(data_to_plot):\n",
    "    data_to_plot=data_to_plot.T\n",
    "    m,n=data_to_plot.shape\n",
    "    indexes=[]\n",
    "    new_p=[]\n",
    "    for j in range(0,m):\n",
    "        for i in range(0,m):\n",
    "            s1=[x for x in data_to_plot[j] if str(x) != 'nan']\n",
    "            s2=[x for x in data_to_plot[i] if str(x) != 'nan']\n",
    "            tstat, pvalue=stats.ttest_ind(s1,s2, equal_var = False)\n",
    "            #ustat, pvalue=stats.mannwhitneyu(s1,s2)\n",
    "            if i <j:\n",
    "                new_p.append(pvalue)\n",
    "                indexes.append((i, j))\n",
    "\n",
    "    # Plot p-values with correction for multiple testing    \n",
    "    a, corrected_p, c, d = smm.multipletests(new_p, method='b')\n",
    "    p_array = np.ones((m, m))\n",
    "    for (i, j), pv in zip(indexes, corrected_p):\n",
    "        p_array[i, j] = p_array[j, i] =  round(pv, ndigits=3)\n",
    "    return p_array\n",
    "\n",
    "def model_scores(prediction,testY):\n",
    "    df=pd.DataFrame(columns=[\"Metric\",\"Score\"])\n",
    "    df.loc[0]=[\"R-square\", metrics.r2_score(testY,prediction)]\n",
    "    df.loc[1]=[\"Explained variance\", metrics.explained_variance_score(testY,prediction)]\n",
    "    df.loc[2]=[\"Mean absolute error\", metrics.mean_absolute_error(testY,prediction)]\n",
    "    df.loc[3]=[\"Mean squared error\", metrics.mean_squared_error(testY,prediction)]\n",
    "    df.loc[4]=[\"Median absolute error\", metrics.median_absolute_error(testY,prediction)]  \n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T13:56:22.545928Z",
     "start_time": "2019-07-04T13:56:22.495641Z"
    }
   },
   "outputs": [],
   "source": [
    "indicators=['Population',\n",
    "       'Homicides', 'Suicides', 'Traffic accidents',\n",
    "       'Elderly population', 'Female population ', 'Male population','GDP',\n",
    "       'Income', 'Unemployment', 'Illiteracy', 'Sanitation', 'Child Labor']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T13:59:43.354402Z",
     "start_time": "2019-07-04T13:59:43.254940Z"
    }
   },
   "outputs": [],
   "source": [
    "datasets=dict()\n",
    "for year in [2000,2010]:\n",
    "    datasets[year]=pd.read_csv(f'data/data-{year}.csv',index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T13:59:59.385008Z",
     "start_time": "2019-07-04T13:59:59.334620Z"
    }
   },
   "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>Population</th>\n",
       "      <th>Homicides</th>\n",
       "      <th>Suicides</th>\n",
       "      <th>Traffic accidents</th>\n",
       "      <th>Elderly population</th>\n",
       "      <th>Female population</th>\n",
       "      <th>Male population</th>\n",
       "      <th>GDP</th>\n",
       "      <th>Income</th>\n",
       "      <th>Unemployment</th>\n",
       "      <th>Illiteracy</th>\n",
       "      <th>Sanitation</th>\n",
       "      <th>Child Labor</th>\n",
       "      <th>CityUF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>26533</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1486</td>\n",
       "      <td>12609</td>\n",
       "      <td>13924</td>\n",
       "      <td>98023.0</td>\n",
       "      <td>371.48</td>\n",
       "      <td>704</td>\n",
       "      <td>2510</td>\n",
       "      <td>25951</td>\n",
       "      <td>1133</td>\n",
       "      <td>ALTA FLORESTA D'OESTE, RO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>74503</td>\n",
       "      <td>46</td>\n",
       "      <td>6</td>\n",
       "      <td>17</td>\n",
       "      <td>3498</td>\n",
       "      <td>36508</td>\n",
       "      <td>37995</td>\n",
       "      <td>381793.4</td>\n",
       "      <td>527.92</td>\n",
       "      <td>2869</td>\n",
       "      <td>5232</td>\n",
       "      <td>71874</td>\n",
       "      <td>1584</td>\n",
       "      <td>ARIQUEMES, RO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7518</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>555</td>\n",
       "      <td>3556</td>\n",
       "      <td>3962</td>\n",
       "      <td>33496.4</td>\n",
       "      <td>344.05</td>\n",
       "      <td>173</td>\n",
       "      <td>841</td>\n",
       "      <td>7488</td>\n",
       "      <td>218</td>\n",
       "      <td>CABIXI, RO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>73568</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>13</td>\n",
       "      <td>4312</td>\n",
       "      <td>36350</td>\n",
       "      <td>37218</td>\n",
       "      <td>385936.5</td>\n",
       "      <td>452.34</td>\n",
       "      <td>2760</td>\n",
       "      <td>5672</td>\n",
       "      <td>73271</td>\n",
       "      <td>1874</td>\n",
       "      <td>CACOAL, RO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>18207</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1294</td>\n",
       "      <td>8890</td>\n",
       "      <td>9317</td>\n",
       "      <td>89764.6</td>\n",
       "      <td>509.20</td>\n",
       "      <td>841</td>\n",
       "      <td>1503</td>\n",
       "      <td>18135</td>\n",
       "      <td>464</td>\n",
       "      <td>CEREJEIRAS, RO</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Population  Homicides  Suicides  Traffic accidents  Elderly population  \\\n",
       "0       26533          3         2                  5                1486   \n",
       "1       74503         46         6                 17                3498   \n",
       "2        7518          0         0                  1                 555   \n",
       "3       73568         13        10                 13                4312   \n",
       "4       18207          2         0                  3                1294   \n",
       "\n",
       "   Female population   Male population       GDP  Income  Unemployment  \\\n",
       "0               12609            13924   98023.0  371.48           704   \n",
       "1               36508            37995  381793.4  527.92          2869   \n",
       "2                3556             3962   33496.4  344.05           173   \n",
       "3               36350            37218  385936.5  452.34          2760   \n",
       "4                8890             9317   89764.6  509.20           841   \n",
       "\n",
       "   Illiteracy  Sanitation  Child Labor                     CityUF  \n",
       "0        2510       25951         1133  ALTA FLORESTA D'OESTE, RO  \n",
       "1        5232       71874         1584              ARIQUEMES, RO  \n",
       "2         841        7488          218                 CABIXI, RO  \n",
       "3        5672       73271         1874                 CACOAL, RO  \n",
       "4        1503       18135          464             CEREJEIRAS, RO  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "datasets[2000].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:26:12.213406Z",
     "start_time": "2019-07-05T11:26:11.769880Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Homicides')"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.loglog(datasets[2000].Population,datasets[2000].Homicides,'o')\n",
    "plt.xlabel('Population')\n",
    "plt.ylabel('Homicides')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-21T23:21:42.796078Z",
     "start_time": "2017-07-21T23:21:42.742142Z"
    }
   },
   "source": [
    "# Correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:01:26.179227Z",
     "start_time": "2019-07-04T14:01:22.451084Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lgaalves/miniconda3/lib/python3.6/site-packages/matplotlib/font_manager.py:1241: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans.\n",
      "  (prop.get_family(), self.defaultFamily[fontext]))\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 786.24x604.8 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 786.24x604.8 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for year in [2000,2010]:\n",
    "    correlation=np.zeros((len(indicators),len(indicators)))\n",
    "    for i in range(0,len(indicators)):\n",
    "        for j in range(0,len(indicators)):\n",
    "            x=np.nan_to_num(datasets[year][indicators[i]])\n",
    "            y=np.nan_to_num(datasets[year][indicators[j]])\n",
    "            r,pvalue=stats.pearsonr(x,y)\n",
    "            correlation[i][j]=r\n",
    "    df_corr=pd.DataFrame(correlation,columns=indicators,index=indicators)     \n",
    "    \n",
    "    fig, ax = plt.subplots(figsize=stdfigsize(scale=1.2))\n",
    "    res=sns.heatmap(df_corr,vmin=0 ,vmax=1.,linewidth=1,cmap=\"Spectral\",square=True)\n",
    "    #ax.set_title(\"Year={}\".format(year))\n",
    "    cbar = res.collections[0].colorbar\n",
    "    cbar.set_ticks([0., 0.25,0.5, 0.75, 1])\n",
    "    cbar.set_ticklabels(['0.00', '0.25',\"0.50\", '0.75', '1.00'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-21T23:50:12.448024Z",
     "start_time": "2017-07-21T23:50:11.857720Z"
    }
   },
   "source": [
    "# Regression models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:06:12.610470Z",
     "start_time": "2019-07-04T14:06:12.572661Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data splitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:14:42.392615Z",
     "start_time": "2019-07-05T11:14:41.739066Z"
    }
   },
   "outputs": [],
   "source": [
    "df_dataset=datasets[2000][np.concatenate([indicators,[\"CityUF\"]])]\n",
    "df_dataset[\"Y\"]=np.array(datasets[2010][\"Homicides\"])\n",
    "df_dataset=df_dataset[df_dataset.Population>0]\n",
    "df_dataset=df_dataset.dropna()\n",
    "\n",
    "df_dataset=datasets[2000][indicators]\n",
    "df_dataset[\"Y\"]=np.array(datasets[2010][\"Homicides\"])\n",
    "df_dataset=df_dataset[df_dataset.Population>0]\n",
    "\n",
    "df_dataset=df_dataset.dropna()\n",
    "X=np.array(df_dataset[indicators])\n",
    "Y=np.array(df_dataset[\"Y\"])\n",
    "trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:06:34.911394Z",
     "start_time": "2019-07-04T14:06:34.839887Z"
    }
   },
   "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>Population</th>\n",
       "      <th>Homicides</th>\n",
       "      <th>Suicides</th>\n",
       "      <th>Traffic accidents</th>\n",
       "      <th>Elderly population</th>\n",
       "      <th>Female population</th>\n",
       "      <th>Male population</th>\n",
       "      <th>GDP</th>\n",
       "      <th>Income</th>\n",
       "      <th>Unemployment</th>\n",
       "      <th>Illiteracy</th>\n",
       "      <th>Sanitation</th>\n",
       "      <th>Child Labor</th>\n",
       "      <th>Y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>26533</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>1486</td>\n",
       "      <td>12609</td>\n",
       "      <td>13924</td>\n",
       "      <td>98023.0</td>\n",
       "      <td>371.48</td>\n",
       "      <td>704</td>\n",
       "      <td>2510</td>\n",
       "      <td>25951</td>\n",
       "      <td>1133</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>74503</td>\n",
       "      <td>46</td>\n",
       "      <td>6</td>\n",
       "      <td>17</td>\n",
       "      <td>3498</td>\n",
       "      <td>36508</td>\n",
       "      <td>37995</td>\n",
       "      <td>381793.4</td>\n",
       "      <td>527.92</td>\n",
       "      <td>2869</td>\n",
       "      <td>5232</td>\n",
       "      <td>71874</td>\n",
       "      <td>1584</td>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7518</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>555</td>\n",
       "      <td>3556</td>\n",
       "      <td>3962</td>\n",
       "      <td>33496.4</td>\n",
       "      <td>344.05</td>\n",
       "      <td>173</td>\n",
       "      <td>841</td>\n",
       "      <td>7488</td>\n",
       "      <td>218</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>73568</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>13</td>\n",
       "      <td>4312</td>\n",
       "      <td>36350</td>\n",
       "      <td>37218</td>\n",
       "      <td>385936.5</td>\n",
       "      <td>452.34</td>\n",
       "      <td>2760</td>\n",
       "      <td>5672</td>\n",
       "      <td>73271</td>\n",
       "      <td>1874</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>18207</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1294</td>\n",
       "      <td>8890</td>\n",
       "      <td>9317</td>\n",
       "      <td>89764.6</td>\n",
       "      <td>509.20</td>\n",
       "      <td>841</td>\n",
       "      <td>1503</td>\n",
       "      <td>18135</td>\n",
       "      <td>464</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Population  Homicides  Suicides  Traffic accidents  Elderly population  \\\n",
       "0       26533          3         2                  5                1486   \n",
       "1       74503         46         6                 17                3498   \n",
       "2        7518          0         0                  1                 555   \n",
       "3       73568         13        10                 13                4312   \n",
       "4       18207          2         0                  3                1294   \n",
       "\n",
       "   Female population   Male population       GDP  Income  Unemployment  \\\n",
       "0               12609            13924   98023.0  371.48           704   \n",
       "1               36508            37995  381793.4  527.92          2869   \n",
       "2                3556             3962   33496.4  344.05           173   \n",
       "3               36350            37218  385936.5  452.34          2760   \n",
       "4                8890             9317   89764.6  509.20           841   \n",
       "\n",
       "   Illiteracy  Sanitation  Child Labor   Y  \n",
       "0        2510       25951         1133   7  \n",
       "1        5232       71874         1584  43  \n",
       "2         841        7488          218   0  \n",
       "3        5672       73271         1874  20  \n",
       "4        1503       18135          464   9  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dataset.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:10:12.393484Z",
     "start_time": "2019-07-04T14:10:12.203835Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.feature_selection import f_regression\n",
    "from sklearn.me"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:10:13.581477Z",
     "start_time": "2019-07-04T14:10:13.536101Z"
    }
   },
   "outputs": [],
   "source": [
    "lm = LinearRegression(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:15:19.378440Z",
     "start_time": "2019-07-04T14:15:19.078692Z"
    }
   },
   "outputs": [],
   "source": [
    "prediction=[]\n",
    "for i in range(0,100):\n",
    "    trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.2)\n",
    "    prediction.append([])\n",
    "    lm.fit(trainX,trainY)\n",
    "    prediction_lm=lm.predict(testX)\n",
    "    prediction[i]=prediction_lm.astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:15:26.761393Z",
     "start_time": "2019-07-04T14:15:26.724447Z"
    }
   },
   "outputs": [],
   "source": [
    "prediction_lm= prediction[10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:15:32.645262Z",
     "start_time": "2019-07-04T14:15:32.030421Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "ax.set_xlabel(\"Data\")\n",
    "ax.set_ylabel(\"Prediction\")\n",
    "ax.plot(testY,prediction_lm,\"o\",markersize=12,color=\"red\",alpha=0.6)\n",
    "ax.plot(np.arange(1,max(testY),0.1),np.arange(1,max(testY),0.1),\"k--\",linewidth=3)\n",
    "#ax.legend(fontsize=14)\n",
    "ax.set_xscale('log')\n",
    "ax.set_yscale('log')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:17:20.339634Z",
     "start_time": "2019-07-04T14:17:20.284853Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KstestResult(statistic=0.36200485897741064, pvalue=7.04762181832909e-130)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x=(np.array(prediction_lm)-np.array(testY))\n",
    "stats.kstest(x,'norm')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:17:28.384048Z",
     "start_time": "2019-07-04T14:17:27.161344Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Normal quantiles')"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=figsize)\n",
    "ax = fig.add_subplot(111)\n",
    "x=(np.array(prediction_lm)-np.array(testY))\n",
    "res = stats.probplot(x, dist=stats.norm, plot=ax)\n",
    "ax.set_title('')\n",
    "ax.set_ylabel(\"Data quantiles\")\n",
    "ax.set_xlabel(\"Normal quantiles\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:20:43.223367Z",
     "start_time": "2019-07-04T14:20:40.217098Z"
    }
   },
   "outputs": [],
   "source": [
    "prediction=[]\n",
    "accuracy=[]\n",
    "for i in range(0,1000):\n",
    "    prediction.append([])\n",
    "    trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.2)\n",
    "    lm.fit(trainX,trainY)\n",
    "    prediction_lm=lm.predict(testX)\n",
    "    accuracy.append(metrics.r2_score(testY,prediction_lm))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:20:46.187474Z",
     "start_time": "2019-07-04T14:20:45.816677Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7549262202184088 0.15764615395949527\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "ax.set_ylabel(\"PDF\")\n",
    "ax.set_xlabel(\"Adjusted$-R^2$\")\n",
    "sns.distplot(accuracy,kde=True)\n",
    "#ax.set_yscale(\"log\")\n",
    "print(np.mean(accuracy),np.std(accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:31:56.956110Z",
     "start_time": "2019-07-04T14:31:56.295914Z"
    }
   },
   "outputs": [],
   "source": [
    "importance_df_lm=pd.DataFrame()\n",
    "importance_df_lm[\"Feature\"]=indicators\n",
    "for i in range(0,100):\n",
    "    df=df_dataset.sample(frac=1,replace=True)\n",
    "    X=np.array(df[indicators])\n",
    "    Y=np.array(df[\"Y\"])\n",
    "    lm.fit(X,Y)\n",
    "    importance_df_lm[i]=lm.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:31:57.045679Z",
     "start_time": "2019-07-04T14:31:56.958108Z"
    }
   },
   "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>Feature</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>...</th>\n",
       "      <th>90</th>\n",
       "      <th>91</th>\n",
       "      <th>92</th>\n",
       "      <th>93</th>\n",
       "      <th>94</th>\n",
       "      <th>95</th>\n",
       "      <th>96</th>\n",
       "      <th>97</th>\n",
       "      <th>98</th>\n",
       "      <th>99</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Population</td>\n",
       "      <td>0.003433</td>\n",
       "      <td>0.001205</td>\n",
       "      <td>0.001659</td>\n",
       "      <td>0.003735</td>\n",
       "      <td>0.002916</td>\n",
       "      <td>0.000495</td>\n",
       "      <td>0.000621</td>\n",
       "      <td>0.004190</td>\n",
       "      <td>0.002260</td>\n",
       "      <td>...</td>\n",
       "      <td>0.001934</td>\n",
       "      <td>0.001173</td>\n",
       "      <td>0.000632</td>\n",
       "      <td>0.002291</td>\n",
       "      <td>0.001555</td>\n",
       "      <td>0.002625</td>\n",
       "      <td>0.002546</td>\n",
       "      <td>0.003368</td>\n",
       "      <td>0.002198</td>\n",
       "      <td>0.001237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Homicides</td>\n",
       "      <td>-0.296301</td>\n",
       "      <td>-0.296986</td>\n",
       "      <td>-0.285744</td>\n",
       "      <td>-0.207770</td>\n",
       "      <td>-0.221917</td>\n",
       "      <td>-0.107107</td>\n",
       "      <td>0.089143</td>\n",
       "      <td>-0.138719</td>\n",
       "      <td>-0.115571</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.303483</td>\n",
       "      <td>-0.201105</td>\n",
       "      <td>-0.146938</td>\n",
       "      <td>-0.145022</td>\n",
       "      <td>-0.071291</td>\n",
       "      <td>-0.421707</td>\n",
       "      <td>-0.129343</td>\n",
       "      <td>-0.008964</td>\n",
       "      <td>-0.214746</td>\n",
       "      <td>-0.151525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Suicides</td>\n",
       "      <td>-0.990634</td>\n",
       "      <td>-1.633985</td>\n",
       "      <td>-1.751442</td>\n",
       "      <td>-0.876494</td>\n",
       "      <td>-1.976891</td>\n",
       "      <td>-1.156921</td>\n",
       "      <td>-1.380040</td>\n",
       "      <td>-0.398277</td>\n",
       "      <td>-2.004530</td>\n",
       "      <td>...</td>\n",
       "      <td>-2.045403</td>\n",
       "      <td>-2.336204</td>\n",
       "      <td>-1.532211</td>\n",
       "      <td>-0.893832</td>\n",
       "      <td>-1.019878</td>\n",
       "      <td>-1.136724</td>\n",
       "      <td>-0.587082</td>\n",
       "      <td>-1.574812</td>\n",
       "      <td>-3.422511</td>\n",
       "      <td>-1.626796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Traffic accidents</td>\n",
       "      <td>1.197237</td>\n",
       "      <td>0.781002</td>\n",
       "      <td>1.090773</td>\n",
       "      <td>1.032370</td>\n",
       "      <td>1.081396</td>\n",
       "      <td>0.988405</td>\n",
       "      <td>0.529056</td>\n",
       "      <td>0.547046</td>\n",
       "      <td>0.653225</td>\n",
       "      <td>...</td>\n",
       "      <td>0.860859</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.823998</td>\n",
       "      <td>1.281072</td>\n",
       "      <td>0.896525</td>\n",
       "      <td>1.359115</td>\n",
       "      <td>0.644850</td>\n",
       "      <td>1.258148</td>\n",
       "      <td>1.026815</td>\n",
       "      <td>1.249700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Elderly population</td>\n",
       "      <td>-0.003297</td>\n",
       "      <td>-0.002610</td>\n",
       "      <td>-0.002665</td>\n",
       "      <td>-0.003567</td>\n",
       "      <td>-0.002863</td>\n",
       "      <td>-0.006827</td>\n",
       "      <td>-0.004068</td>\n",
       "      <td>-0.005340</td>\n",
       "      <td>-0.004767</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.003216</td>\n",
       "      <td>-0.002267</td>\n",
       "      <td>-0.005167</td>\n",
       "      <td>-0.004169</td>\n",
       "      <td>-0.004141</td>\n",
       "      <td>-0.002949</td>\n",
       "      <td>-0.006443</td>\n",
       "      <td>-0.002934</td>\n",
       "      <td>-0.002590</td>\n",
       "      <td>-0.004539</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 101 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              Feature         0         1         2         3         4  \\\n",
       "0          Population  0.003433  0.001205  0.001659  0.003735  0.002916   \n",
       "1           Homicides -0.296301 -0.296986 -0.285744 -0.207770 -0.221917   \n",
       "2            Suicides -0.990634 -1.633985 -1.751442 -0.876494 -1.976891   \n",
       "3   Traffic accidents  1.197237  0.781002  1.090773  1.032370  1.081396   \n",
       "4  Elderly population -0.003297 -0.002610 -0.002665 -0.003567 -0.002863   \n",
       "\n",
       "          5         6         7         8  ...        90        91        92  \\\n",
       "0  0.000495  0.000621  0.004190  0.002260  ...  0.001934  0.001173  0.000632   \n",
       "1 -0.107107  0.089143 -0.138719 -0.115571  ... -0.303483 -0.201105 -0.146938   \n",
       "2 -1.156921 -1.380040 -0.398277 -2.004530  ... -2.045403 -2.336204 -1.532211   \n",
       "3  0.988405  0.529056  0.547046  0.653225  ...  0.860859  0.998039  0.823998   \n",
       "4 -0.006827 -0.004068 -0.005340 -0.004767  ... -0.003216 -0.002267 -0.005167   \n",
       "\n",
       "         93        94        95        96        97        98        99  \n",
       "0  0.002291  0.001555  0.002625  0.002546  0.003368  0.002198  0.001237  \n",
       "1 -0.145022 -0.071291 -0.421707 -0.129343 -0.008964 -0.214746 -0.151525  \n",
       "2 -0.893832 -1.019878 -1.136724 -0.587082 -1.574812 -3.422511 -1.626796  \n",
       "3  1.281072  0.896525  1.359115  0.644850  1.258148  1.026815  1.249700  \n",
       "4 -0.004169 -0.004141 -0.002949 -0.006443 -0.002934 -0.002590 -0.004539  \n",
       "\n",
       "[5 rows x 101 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importance_df_lm.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:31:57.113897Z",
     "start_time": "2019-07-04T14:31:57.051146Z"
    }
   },
   "outputs": [],
   "source": [
    "importance_df_lm=importance_df_lm.set_index(\"Feature\").T\n",
    "features=list(importance_df_lm.columns)\n",
    "medians=[]\n",
    "for feature in features:\n",
    "    m=np.median(importance_df_lm[feature])\n",
    "    medians.append(m)\n",
    "\n",
    "importance_df_lm=importance_df_lm.T\n",
    "importance_df_lm[\"Medians\"]=medians\n",
    "\n",
    "importance_df_lm.sort_values(by=\"Medians\",inplace=True,ascending=False)\n",
    "del importance_df_lm['Medians']\n",
    "\n",
    "importance_df_lm=importance_df_lm.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:28:04.729283Z",
     "start_time": "2019-07-04T14:28:04.173575Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0, 'Traffic accidents'),\n",
       " Text(0, 0, 'Female population '),\n",
       " Text(0, 0, 'Unemployment'),\n",
       " Text(0, 0, 'Population'),\n",
       " Text(0, 0, 'Illiteracy'),\n",
       " Text(0, 0, 'Male population'),\n",
       " Text(0, 0, 'GDP'),\n",
       " Text(0, 0, 'Income'),\n",
       " Text(0, 0, 'Elderly population'),\n",
       " Text(0, 0, 'Sanitation'),\n",
       " Text(0, 0, 'Child Labor'),\n",
       " Text(0, 0, 'Homicides'),\n",
       " Text(0, 0, 'Suicides')]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a figure instance\n",
    "fig = plt.figure(1, figsize=figsize)\n",
    "ax = fig.add_subplot(111)\n",
    "ax.grid()\n",
    "bp = ax.boxplot(np.array(importance_df_lm), patch_artist=True,showfliers=True)\n",
    "## change outline color, fill color and linewidth of the boxes\n",
    "for box in bp['boxes']:\n",
    "    # change outline color\n",
    "    box.set( color='black', linewidth=1)\n",
    "    # change fill color\n",
    "    box.set( facecolor='#bdbdbd',zorder=3)\n",
    "\n",
    "## change color and linewidth of the whiskers\n",
    "for whisker in bp['whiskers']:\n",
    "    whisker.set(linestyle='-',color='black', linewidth=1,zorder=4)\n",
    "\n",
    "## change color and linewidth of the caps\n",
    "for cap in bp['caps']:\n",
    "    cap.set(color='black', linewidth=1,zorder=4)\n",
    "\n",
    "## change color and linewidth of the medians\n",
    "for median in bp['medians']:\n",
    "    median.set(color='k', linewidth=1,zorder=4)\n",
    "\n",
    "## change the style of fliers and their fill\n",
    "for flier in bp['fliers']:\n",
    "    flier.set(marker='o', markerfacecolor='k',\n",
    "              markeredgecolor='k',markersize=4,zorder=3)\n",
    "ax.set_xlabel('Features')\n",
    "ax.set_ylabel('Feature importance')\n",
    "#ax.set_yscale('log')\n",
    "ax.set_ylim(-1,2)\n",
    "ax.set_xticklabels(labels=list(importance_df_lm.columns),rotation=90)\n",
    "#plt.savefig(\"../figures/features_importance.pdf\",bbox_inches='tight')    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:30:32.458816Z",
     "start_time": "2019-07-04T14:30:31.812291Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lgaalves/miniconda3/lib/python3.6/site-packages/matplotlib/font_manager.py:1241: UserWarning: findfont: Font family ['sans-serif'] not found. Falling back to DejaVu Sans.\n",
      "  (prop.get_family(), self.defaultFamily[fontext]))\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 786.24x604.8 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p=test2samples(np.array(importance_df_lm))\n",
    "p=pd.DataFrame(p,index=list(importance_df_lm.columns),columns=list(importance_df_lm.columns))\n",
    "fig, ax = plt.subplots(figsize=stdfigsize(scale=1.2))\n",
    "res=sns.heatmap(p,vmin=0,vmax=1.,linewidths=1,annot=False,cmap=\"Reds\",square=True)\n",
    "cbar = res.collections[0].colorbar\n",
    "cbar.set_ticks([0,0.25, 0.5,0.75, 1])\n",
    "cbar.set_ticklabels([\"0.00\",\"0.25\", \"0.50\",\"0.75\", \"1.00\"])\n",
    "ax.tick_params(labelsize=22)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-22T18:04:53.173586Z",
     "start_time": "2017-07-22T18:04:53.142780Z"
    }
   },
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:39:36.424270Z",
     "start_time": "2019-07-04T14:39:36.339532Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn import preprocessing\n",
    "from sklearn import metrics\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import validation_curve\n",
    "from sklearn.model_selection import learning_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:40:04.713284Z",
     "start_time": "2019-07-04T14:40:04.561902Z"
    }
   },
   "outputs": [],
   "source": [
    "df_dataset=datasets[2000][np.concatenate([indicators,[\"CityUF\"]])]\n",
    "df_dataset[\"Y\"]=np.array(datasets[2010][\"Homicides\"])\n",
    "df_dataset=df_dataset[df_dataset.Population>0]\n",
    "df_dataset=df_dataset.dropna()\n",
    "\n",
    "df_dataset=datasets[2000][indicators]\n",
    "df_dataset[\"Y\"]=np.array(datasets[2010][\"Homicides\"])\n",
    "df_dataset=df_dataset[df_dataset.Population>0]\n",
    "\n",
    "df_dataset=df_dataset.dropna()\n",
    "X=np.array(df_dataset[indicators])\n",
    "Y=np.array(df_dataset[\"Y\"])\n",
    "trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:40:15.342670Z",
     "start_time": "2019-07-04T14:40:15.287989Z"
    }
   },
   "outputs": [],
   "source": [
    "rf=RandomForestRegressor(n_estimators=200,\n",
    "                    criterion='mse', \n",
    "                    max_depth=50, \n",
    "                    min_samples_split=2, \n",
    "                    min_samples_leaf=1, \n",
    "                    min_weight_fraction_leaf=0.0, \n",
    "                    max_features='auto', \n",
    "                    max_leaf_nodes=None, \n",
    "                    min_impurity_split=1e-07, \n",
    "                    bootstrap=True, \n",
    "                    oob_score=False, \n",
    "                    n_jobs=-1, \n",
    "                    random_state=None, \n",
    "                    verbose=0,\n",
    "                    warm_start=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Underfitting and Overfitting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-23T14:31:19.302881Z",
     "start_time": "2017-07-23T14:31:19.268251Z"
    }
   },
   "source": [
    "In statistics and machine learning, the bias–variance tradeoff (or dilemma) is the problem of simultaneously minimizing two sources of error that prevent supervised learning algorithms from generalizing beyond their training set:\n",
    "\n",
    "**Underfitting**: The bias is error from erroneous assumptions in the learning algorithm. High bias can cause an algorithm to miss the relevant relations between features and target outputs (underfitting).\n",
    "\n",
    "**Overfitting**: The variance is error from sensitivity to small fluctuations in the training set. High variance can cause overfitting: modeling the random noise in the training data, rather than the intended outputs.\n",
    "\n",
    "\n",
    "Finding an $\\hat {f}$ that generalizes to points outside of the training set can be done with any of the countless algorithms used for supervised learning. It turns out that whichever function $\\hat {f}$ we select, we can decompose its expected error on an unseen sample $x$ as follows*\n",
    "\n",
    "\\begin{aligned}{\\mathrm  {E}}{\\Big [}{\\big (}y-{\\hat  {f}}(x){\\big )}^{2}{\\Big ]}&={\\mathrm  {Bias}}{\\big [}{\\hat  {f}}(x){\\big ]}^{2}+{\\mathrm  {Var}}{\\big [}{\\hat  {f}}(x){\\big ]}+\\sigma ^{2}\\\\\\end{aligned}\n",
    "Where:\n",
    "\n",
    "\\begin{aligned}\\mathrm {Bias} {\\big [}{\\hat {f}}(x){\\big ]}=\\mathrm {E} {\\big [}{\\hat {f}}(x)-f(x){\\big ]}\\end{aligned}\n",
    "\n",
    "and\n",
    "\n",
    "\\begin{aligned}\\mathrm {Var} {\\big [}{\\hat {f}}(x){\\big ]}=\\mathrm {E} [{\\hat {f}}(x)^{2}]-\\mathrm {E} [{\\hat {f}}(x)]^{2}\\end{aligned}\n",
    "\n",
    "\n",
    "* Gareth James; Daniela Witten; Trevor Hastie; Robert Tibshirani (2013). An Introduction to Statistical Learning. Springer.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Validation curve"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Number of trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:48:59.996792Z",
     "start_time": "2019-07-04T14:42:09.381395Z"
    }
   },
   "outputs": [],
   "source": [
    "X, y = trainX,trainY\n",
    "param_range =np.logspace(0,3,10).astype(int)\n",
    "train_scores, test_scores = validation_curve(RandomForestRegressor().set_params(**{'bootstrap': True,\n",
    "                                                           'max_depth': 50,\n",
    "                                                           'max_features': 'auto',\n",
    "                                                           'min_samples_split': 2}), \n",
    "                                             X, y, \n",
    "                                             param_name=\"n_estimators\", \n",
    "                                             param_range=param_range,\n",
    "                                             cv=10, \n",
    "                                             scoring=None, \n",
    "                                             n_jobs=-1)\n",
    "\n",
    "train_scores_mean = np.mean(train_scores, axis=1)\n",
    "train_scores_std = np.std(train_scores, axis=1)\n",
    "test_scores_mean = np.mean(test_scores, axis=1)\n",
    "test_scores_std = np.std(test_scores, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:41:54.813898Z",
     "start_time": "2019-07-04T14:41:54.773565Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([   1,    2,    4,   10,   21,   46,  100,  215,  464, 1000])"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.logspace(0,3,10).astype(int)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-24T16:23:04.984761Z",
     "start_time": "2017-07-24T16:23:04.944709Z"
    }
   },
   "source": [
    "In this plot you can see the training scores and validation scores of a Random Forest Regressor for different values of the parameter n_estimators (Number of trees). For very low values of n_estimators, you can see that both the training score and the validation score are slightly lower. This is called underfitting. Medium values of n_estimators will result in high values for both scores, i.e. the regressor is performing fairly well. If n_estimators is too high, the regressors will overfit, which means that the training score is good but the validation score is poor. We do not see this here because up to 10000 trees the model is still performing well.\n",
    "\n",
    "See more at: \n",
    "* http://scikit-learn.org/stable/modules/learning_curve.html\n",
    "* http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.validation_curve.html\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:49:00.696123Z",
     "start_time": "2019-07-04T14:48:59.999181Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1c238d58d0>"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "\n",
    "ax.set_xlabel(\"Number of trees\")\n",
    "ax.set_ylabel(\"Score\")\n",
    "ax.set_ylim(0.2,1.05)\n",
    "lw = 2\n",
    "#ax.grid()\n",
    "ax.plot(param_range, train_scores_mean, label=\"Training score\",\n",
    "             color=\"#e41a1c\", lw=lw)\n",
    "ax.fill_between(param_range, train_scores_mean - train_scores_std,\n",
    "                 train_scores_mean + train_scores_std, alpha=0.2,\n",
    "                 color=\"#e41a1c\", lw=lw)\n",
    "ax.plot(param_range, test_scores_mean, label=\"Cross-validation score\",\n",
    "             color=\"#377eb8\", lw=lw)\n",
    "ax.fill_between(param_range, test_scores_mean - test_scores_std,\n",
    "                 test_scores_mean + test_scores_std, alpha=0.2,\n",
    "                 color=\"#377eb8\", lw=lw)\n",
    "ax.set_xscale(\"log\")\n",
    "ax.legend(loc=\"lower left\",fontsize=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2017-07-24T16:12:08.828646Z",
     "start_time": "2017-07-24T16:12:02.998006Z"
    }
   },
   "source": [
    "#### Maximal depth"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:59:24.691745Z",
     "start_time": "2019-07-04T14:57:29.327317Z"
    }
   },
   "outputs": [],
   "source": [
    "X, y = trainX,trainY\n",
    "param_range_max_depth =np.logspace(0.1,2,10).astype(int)\n",
    "train_scores, test_scores = validation_curve(RandomForestRegressor(n_estimators=100), \n",
    "                                             X, y, \n",
    "                                             param_name=\"max_depth\", \n",
    "                                             param_range=param_range_max_depth,\n",
    "                                             cv=10, \n",
    "                                             scoring=None, \n",
    "                                             n_jobs=-1)\n",
    "\n",
    "train_scores_mean_max_depth = np.mean(train_scores, axis=1)\n",
    "train_scores_std_max_depth = np.std(train_scores, axis=1)\n",
    "test_scores_mean_max_depth = np.mean(test_scores, axis=1)\n",
    "test_scores_std_max_depth = np.std(test_scores, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:59:24.743529Z",
     "start_time": "2019-07-04T14:59:24.693970Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  1,   2,   3,   5,   8,  14,  23,  37,  61, 100])"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "param_range_max_depth"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T14:59:25.315463Z",
     "start_time": "2019-07-04T14:59:24.747172Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1c22e6f7f0>"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "\n",
    "ax.set_xlabel(\"Maximum depth\")\n",
    "ax.set_ylabel(\"Score\")\n",
    "ax.set_ylim(0.5,1.05)\n",
    "lw = 2\n",
    "ax.plot(param_range_max_depth, train_scores_mean_max_depth, label=\"Training score\",\n",
    "             color=\"#e41a1c\", lw=lw)\n",
    "ax.fill_between(param_range_max_depth, train_scores_mean_max_depth - train_scores_std_max_depth,\n",
    "                 train_scores_mean_max_depth + train_scores_std_max_depth, alpha=0.2,\n",
    "                 color=\"#e41a1c\", lw=lw)\n",
    "ax.plot(param_range_max_depth, test_scores_mean_max_depth, label=\"Cross-validation score\",\n",
    "             color=\"#377eb8\", lw=lw)\n",
    "ax.fill_between(param_range_max_depth, test_scores_mean_max_depth - test_scores_std_max_depth,\n",
    "                 test_scores_mean_max_depth + test_scores_std_max_depth, alpha=0.2,\n",
    "                 color=\"#377eb8\", lw=lw)\n",
    "ax.set_xscale(\"log\")\n",
    "ax.legend(loc=\"lower right\",fontsize=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Learning curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:12:08.186784Z",
     "start_time": "2019-07-05T11:10:33.770389Z"
    }
   },
   "outputs": [],
   "source": [
    "X, y = trainX,trainY\n",
    "estimator = RandomForestRegressor(n_jobs=-1).set_params(**{'bootstrap': True,\n",
    "                                                           'max_depth': 50,\n",
    "                                                           'max_features': 'auto',\n",
    "                                                           'min_samples_split': 2,\n",
    "                                                           'n_estimators': 100})\n",
    "train_sizes_lc=np.linspace(.1, 1.0, 10)\n",
    "\n",
    "train_sizes_lc, train_scores, test_scores = learning_curve(estimator, \n",
    "                                                        X, \n",
    "                                                        y, \n",
    "                                                        cv=10, \n",
    "                                                        n_jobs=-1, \n",
    "                                                        train_sizes=train_sizes_lc)\n",
    "train_scores_mean_lc = np.mean(train_scores, axis=1)\n",
    "train_scores_std_lc = np.std(train_scores, axis=1)\n",
    "test_scores_mean_lc = np.mean(test_scores, axis=1)\n",
    "test_scores_std_lc = np.std(test_scores, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The shape of the curve for the training score is very high at the beginning and decreases and the cross-validation score is very low at the beginning and increases. However, we can see a maximum around 2500, showning that, actually, given more data is not making it better. \n",
    "\n",
    "See more at:\n",
    "http://scikit-learn.org/stable/auto_examples/model_selection/plot_learning_curve.html#sphx-glr-auto-examples-model-selection-plot-learning-curve-py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:12:08.509114Z",
     "start_time": "2019-07-05T11:12:08.189984Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1c29e0ff98>"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "ax.set_xlabel(\"Fraction of training data\")\n",
    "ax.set_ylabel(\"Score\")\n",
    "ax.set_ylim(0.5,1.05)\n",
    "#ax.grid()\n",
    "ax.fill_between(np.linspace(.1, 1.0, 10), train_scores_mean_lc - train_scores_std_lc,\n",
    "                 train_scores_mean_lc + train_scores_std_lc, alpha=0.2,\n",
    "                 color=\"#e41a1c\")\n",
    "ax.fill_between(np.linspace(.1, 1.0, 10), test_scores_mean_lc - test_scores_std_lc,\n",
    "                 test_scores_mean_lc + test_scores_std_lc, alpha=0.2, color=\"#377eb8\")\n",
    "ax.plot(np.linspace(.1, 1.0, 10), train_scores_mean_lc, '-', color=\"#e41a1c\",\n",
    "         label=\"Training score\")\n",
    "ax.plot(np.linspace(.1, 1.0, 10), test_scores_mean_lc, '-', color=\"#377eb8\",\n",
    "         label=\"Cross-validation score\")\n",
    "ax.legend(loc=\"lower right\",fontsize=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tunning model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The main parameters to adjust when using these methods is n_estimators and max_features. The former is the number of trees in the forest. The larger the better, but also the longer it will take to compute. In addition, note that results will stop getting significantly better beyond a critical number of trees. The latter is the size of the random subsets of features to consider when splitting a node. The lower the greater the reduction of variance, but also the greater the increase in bias. Empirical good default values are max_features=n_features for regression problems, and max_features=sqrt(n_features) for classification tasks (where n_features is the number of features in the data). Good results are often achieved when setting max_depth=None in combination with min_samples_split=1 (i.e., when fully developing the trees). Bear in mind though that these values are usually not optimal, and might result in models that consume a lot of RAM. The best parameter values should always be cross-validated. In addition, note that in random forests, bootstrap samples are used by default (bootstrap=True) while the default strategy for extra-trees is to use the whole dataset (bootstrap=False). When using bootstrap sampling the generalization accuracy can be estimated on the left out or out-of-bag samples. This can be enabled by setting oob_score=True."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:12:08.639555Z",
     "start_time": "2019-07-05T11:12:08.512529Z"
    }
   },
   "outputs": [],
   "source": [
    "RandomForestRegressor?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:13:20.805628Z",
     "start_time": "2019-07-05T11:13:20.764250Z"
    }
   },
   "outputs": [],
   "source": [
    "def Grid_Search_CV_RFR(X_train, y_train):\n",
    "    estimator = RandomForestRegressor()\n",
    "    param_grid = {  \"n_estimators\": [100,200,500],\n",
    "                    \"max_depth\":[10,50,100],\n",
    "                    \"max_features\": [\"auto\"],\n",
    "                    \"min_samples_split\" : [2,4,8],\n",
    "                    \"bootstrap\": [True,False]}\n",
    "    grid = GridSearchCV(estimator, param_grid, n_jobs=-1, cv=10)\n",
    "\n",
    "    grid.fit(X_train, y_train)\n",
    "\n",
    "    return grid.best_score_ , grid.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-06-24T03:13:19.002616Z",
     "start_time": "2019-06-24T03:13:15.837Z"
    }
   },
   "outputs": [],
   "source": [
    "best_score_rf,best_params_rf=Grid_Search_CV_RFR(trainX, trainY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:13:52.891886Z",
     "start_time": "2019-07-05T11:13:52.844317Z"
    }
   },
   "outputs": [],
   "source": [
    "best_score_rf,best_params_rf=(0.89291467777356359,\n",
    " {'bootstrap': True,\n",
    "  'max_depth': 100,\n",
    "  'max_features': 'auto',\n",
    "  'min_samples_split': 4,\n",
    "  'n_estimators': 200})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fitting the data and prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:14:26.568450Z",
     "start_time": "2019-07-05T11:14:26.529657Z"
    }
   },
   "outputs": [],
   "source": [
    "rf = RandomForestRegressor(n_jobs=-1).set_params(**{'bootstrap': True,\n",
    "  'max_depth': 100,\n",
    "  'max_features': 'auto',\n",
    "  'min_samples_split': 4,\n",
    "  'n_estimators': 200})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:14:49.051304Z",
     "start_time": "2019-07-05T11:14:48.997275Z"
    }
   },
   "outputs": [],
   "source": [
    "trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:29:01.859437Z",
     "start_time": "2019-07-05T11:29:01.820280Z"
    }
   },
   "outputs": [],
   "source": [
    "prediction=[]\n",
    "accuracy=[]\n",
    "for i in range(0,10):\n",
    "    trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.1)\n",
    "    prediction.append([])\n",
    "    rf.fit(trainX,trainY)\n",
    "    prediction_rf=rf.predict(testX)\n",
    "    prediction[i]=prediction_rf.astype(int)\n",
    "    accuracy.append(rf.score(testX,testY))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:17:42.644885Z",
     "start_time": "2019-07-05T11:17:42.579406Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3.5, 1. , 1.5, ..., 1.5, 0. , 3. ])"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction_rf=np.median(prediction,axis=0)\n",
    "prediction_rf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:28:53.775771Z",
     "start_time": "2019-07-05T11:27:17.404Z"
    }
   },
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "ax.set_xlabel(\"Data\")\n",
    "ax.set_ylabel(\"Prediction\")\n",
    "ax.plot(testY,prediction_rf,\"o\",markersize=12,color=\"red\",alpha=0.6)\n",
    "ax.plot(np.arange(1,max(testY),0.1),np.arange(1,max(testY),0.1),\"k--\",linewidth=3)\n",
    "#ax.legend(fontsize=14)\n",
    "ax.set_xscale('log')\n",
    "ax.set_yscale('log')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:22:07.550309Z",
     "start_time": "2019-07-05T11:22:07.313105Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1c29e209b0>]"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "ax.set_ylabel(\"Residual\")\n",
    "ax.set_xlabel(\"Data points\")\n",
    "ax.plot((np.array(prediction_rf)-np.array(testY)),'o')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:28:53.779440Z",
     "start_time": "2019-07-05T11:27:26.253Z"
    }
   },
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "ax.set_ylabel(\"PDF\")\n",
    "ax.set_xlabel(\"Residuals\")\n",
    "sns.distplot((np.array(prediction_rf)-np.array(testY)),kde=False)\n",
    "print(np.mean(np.array(prediction_rf)-np.array(testY)),np.std(np.array(prediction_rf)-np.array(testY)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:23:10.723773Z",
     "start_time": "2019-07-05T11:22:38.386270Z"
    }
   },
   "outputs": [],
   "source": [
    "prediction=[]\n",
    "accuracy=[]\n",
    "for i in range(0,10):\n",
    "    prediction.append([])\n",
    "    trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.2)\n",
    "    rf.fit(trainX,trainY)\n",
    "    prediction_rf=rf.predict(testX)\n",
    "    accuracy.append(metrics.r2_score(testY,prediction_rf))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:23:11.042615Z",
     "start_time": "2019-07-05T11:23:10.729145Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8087712790684985 0.10440155751702972\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=figsize)\n",
    "ax.set_ylabel(\"PDF\")\n",
    "ax.set_xlabel(\"Adjusted$-R^2$\")\n",
    "sns.distplot(accuracy,kde=True)\n",
    "#ax.set_yscale(\"log\")\n",
    "print(np.mean(accuracy),np.std(accuracy))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:28:53.782300Z",
     "start_time": "2019-07-05T11:27:55.428Z"
    }
   },
   "outputs": [],
   "source": [
    "max(accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:23:11.188432Z",
     "start_time": "2019-07-05T11:23:11.099904Z"
    }
   },
   "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>Metric</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>R-square</td>\n",
       "      <td>0.727775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Explained variance</td>\n",
       "      <td>0.728118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Mean absolute error</td>\n",
       "      <td>5.009861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Mean squared error</td>\n",
       "      <td>458.510678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Median absolute error</td>\n",
       "      <td>1.100125</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Metric       Score\n",
       "0               R-square    0.727775\n",
       "1     Explained variance    0.728118\n",
       "2    Mean absolute error    5.009861\n",
       "3     Mean squared error  458.510678\n",
       "4  Median absolute error    1.100125"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores_df=model_scores(prediction_rf,testY)\n",
    "scores_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Features importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:29:39.859980Z",
     "start_time": "2019-07-05T11:29:06.962915Z"
    }
   },
   "outputs": [],
   "source": [
    "importance_df=pd.DataFrame()\n",
    "importance_df[\"Feature\"]=indicators\n",
    "for i in range(0,10):\n",
    "    X=np.array(df_dataset[indicators])\n",
    "    Y=np.array(df_dataset[\"Y\"])\n",
    "    trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.2)\n",
    "    rf.fit(trainX,trainY)\n",
    "    importance_df[i]=rf.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:29:39.936426Z",
     "start_time": "2019-07-05T11:29:39.865101Z"
    }
   },
   "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>Feature</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Child Labor</td>\n",
       "      <td>0.041079</td>\n",
       "      <td>0.028574</td>\n",
       "      <td>0.061258</td>\n",
       "      <td>0.020963</td>\n",
       "      <td>0.050365</td>\n",
       "      <td>0.036671</td>\n",
       "      <td>0.041241</td>\n",
       "      <td>0.016183</td>\n",
       "      <td>0.063038</td>\n",
       "      <td>0.025254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Elderly population</td>\n",
       "      <td>0.051066</td>\n",
       "      <td>0.038709</td>\n",
       "      <td>0.019813</td>\n",
       "      <td>0.030695</td>\n",
       "      <td>0.017829</td>\n",
       "      <td>0.010706</td>\n",
       "      <td>0.030543</td>\n",
       "      <td>0.023897</td>\n",
       "      <td>0.022193</td>\n",
       "      <td>0.019704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Female population</td>\n",
       "      <td>0.138356</td>\n",
       "      <td>0.038144</td>\n",
       "      <td>0.066694</td>\n",
       "      <td>0.059373</td>\n",
       "      <td>0.062989</td>\n",
       "      <td>0.058436</td>\n",
       "      <td>0.084579</td>\n",
       "      <td>0.075867</td>\n",
       "      <td>0.081445</td>\n",
       "      <td>0.044951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GDP</td>\n",
       "      <td>0.016414</td>\n",
       "      <td>0.004092</td>\n",
       "      <td>0.005868</td>\n",
       "      <td>0.009894</td>\n",
       "      <td>0.007372</td>\n",
       "      <td>0.005097</td>\n",
       "      <td>0.008424</td>\n",
       "      <td>0.004955</td>\n",
       "      <td>0.006709</td>\n",
       "      <td>0.008924</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Homicides</td>\n",
       "      <td>0.044421</td>\n",
       "      <td>0.065223</td>\n",
       "      <td>0.034927</td>\n",
       "      <td>0.043111</td>\n",
       "      <td>0.052627</td>\n",
       "      <td>0.050583</td>\n",
       "      <td>0.025946</td>\n",
       "      <td>0.023074</td>\n",
       "      <td>0.060419</td>\n",
       "      <td>0.042505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Illiteracy</td>\n",
       "      <td>0.188686</td>\n",
       "      <td>0.336791</td>\n",
       "      <td>0.081536</td>\n",
       "      <td>0.327792</td>\n",
       "      <td>0.249781</td>\n",
       "      <td>0.271592</td>\n",
       "      <td>0.130126</td>\n",
       "      <td>0.110609</td>\n",
       "      <td>0.152723</td>\n",
       "      <td>0.296007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Income</td>\n",
       "      <td>0.006040</td>\n",
       "      <td>0.009417</td>\n",
       "      <td>0.007358</td>\n",
       "      <td>0.016700</td>\n",
       "      <td>0.011846</td>\n",
       "      <td>0.014553</td>\n",
       "      <td>0.009475</td>\n",
       "      <td>0.011090</td>\n",
       "      <td>0.013474</td>\n",
       "      <td>0.009436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Male population</td>\n",
       "      <td>0.084958</td>\n",
       "      <td>0.073214</td>\n",
       "      <td>0.109259</td>\n",
       "      <td>0.116867</td>\n",
       "      <td>0.082133</td>\n",
       "      <td>0.100747</td>\n",
       "      <td>0.154591</td>\n",
       "      <td>0.090640</td>\n",
       "      <td>0.073342</td>\n",
       "      <td>0.097411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Population</td>\n",
       "      <td>0.113927</td>\n",
       "      <td>0.065636</td>\n",
       "      <td>0.117940</td>\n",
       "      <td>0.067291</td>\n",
       "      <td>0.096359</td>\n",
       "      <td>0.087083</td>\n",
       "      <td>0.114373</td>\n",
       "      <td>0.057749</td>\n",
       "      <td>0.050471</td>\n",
       "      <td>0.063423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Sanitation</td>\n",
       "      <td>0.090450</td>\n",
       "      <td>0.086323</td>\n",
       "      <td>0.068805</td>\n",
       "      <td>0.062347</td>\n",
       "      <td>0.054182</td>\n",
       "      <td>0.068763</td>\n",
       "      <td>0.139018</td>\n",
       "      <td>0.057552</td>\n",
       "      <td>0.053477</td>\n",
       "      <td>0.093638</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Suicides</td>\n",
       "      <td>0.012408</td>\n",
       "      <td>0.006024</td>\n",
       "      <td>0.003969</td>\n",
       "      <td>0.020188</td>\n",
       "      <td>0.014370</td>\n",
       "      <td>0.004351</td>\n",
       "      <td>0.013960</td>\n",
       "      <td>0.009801</td>\n",
       "      <td>0.011101</td>\n",
       "      <td>0.032144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Traffic accidents</td>\n",
       "      <td>0.019734</td>\n",
       "      <td>0.024727</td>\n",
       "      <td>0.021360</td>\n",
       "      <td>0.038092</td>\n",
       "      <td>0.022851</td>\n",
       "      <td>0.028370</td>\n",
       "      <td>0.048724</td>\n",
       "      <td>0.025381</td>\n",
       "      <td>0.043333</td>\n",
       "      <td>0.028425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Unemployment</td>\n",
       "      <td>0.192460</td>\n",
       "      <td>0.223125</td>\n",
       "      <td>0.401215</td>\n",
       "      <td>0.186685</td>\n",
       "      <td>0.277297</td>\n",
       "      <td>0.263049</td>\n",
       "      <td>0.198998</td>\n",
       "      <td>0.493202</td>\n",
       "      <td>0.368274</td>\n",
       "      <td>0.238178</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Feature         0         1         2         3         4  \\\n",
       "0          Child Labor  0.041079  0.028574  0.061258  0.020963  0.050365   \n",
       "1   Elderly population  0.051066  0.038709  0.019813  0.030695  0.017829   \n",
       "2   Female population   0.138356  0.038144  0.066694  0.059373  0.062989   \n",
       "3                  GDP  0.016414  0.004092  0.005868  0.009894  0.007372   \n",
       "4            Homicides  0.044421  0.065223  0.034927  0.043111  0.052627   \n",
       "5           Illiteracy  0.188686  0.336791  0.081536  0.327792  0.249781   \n",
       "6               Income  0.006040  0.009417  0.007358  0.016700  0.011846   \n",
       "7      Male population  0.084958  0.073214  0.109259  0.116867  0.082133   \n",
       "8           Population  0.113927  0.065636  0.117940  0.067291  0.096359   \n",
       "9           Sanitation  0.090450  0.086323  0.068805  0.062347  0.054182   \n",
       "10            Suicides  0.012408  0.006024  0.003969  0.020188  0.014370   \n",
       "11   Traffic accidents  0.019734  0.024727  0.021360  0.038092  0.022851   \n",
       "12        Unemployment  0.192460  0.223125  0.401215  0.186685  0.277297   \n",
       "\n",
       "           5         6         7         8         9  \n",
       "0   0.036671  0.041241  0.016183  0.063038  0.025254  \n",
       "1   0.010706  0.030543  0.023897  0.022193  0.019704  \n",
       "2   0.058436  0.084579  0.075867  0.081445  0.044951  \n",
       "3   0.005097  0.008424  0.004955  0.006709  0.008924  \n",
       "4   0.050583  0.025946  0.023074  0.060419  0.042505  \n",
       "5   0.271592  0.130126  0.110609  0.152723  0.296007  \n",
       "6   0.014553  0.009475  0.011090  0.013474  0.009436  \n",
       "7   0.100747  0.154591  0.090640  0.073342  0.097411  \n",
       "8   0.087083  0.114373  0.057749  0.050471  0.063423  \n",
       "9   0.068763  0.139018  0.057552  0.053477  0.093638  \n",
       "10  0.004351  0.013960  0.009801  0.011101  0.032144  \n",
       "11  0.028370  0.048724  0.025381  0.043333  0.028425  \n",
       "12  0.263049  0.198998  0.493202  0.368274  0.238178  "
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "importance_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:29:39.991996Z",
     "start_time": "2019-07-05T11:29:39.941008Z"
    }
   },
   "outputs": [],
   "source": [
    "importance_df=importance_df.set_index(\"Feature\").T\n",
    "features=list(importance_df.columns)\n",
    "medians=[]\n",
    "for feature in features:\n",
    "    m=np.median(importance_df[feature])\n",
    "    medians.append(m)\n",
    "\n",
    "importance_df=importance_df.T\n",
    "importance_df[\"Medians\"]=medians\n",
    "\n",
    "importance_df.sort_values(by=\"Medians\",inplace=True,ascending=False)\n",
    "del importance_df['Medians']\n",
    "\n",
    "importance_df=importance_df.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-05T11:29:40.643463Z",
     "start_time": "2019-07-05T11:29:39.994524Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0, 'Unemployment'),\n",
       " Text(0, 0, 'Illiteracy'),\n",
       " Text(0, 0, 'Male population'),\n",
       " Text(0, 0, 'Population'),\n",
       " Text(0, 0, 'Sanitation'),\n",
       " Text(0, 0, 'Female population '),\n",
       " Text(0, 0, 'Homicides'),\n",
       " Text(0, 0, 'Child Labor'),\n",
       " Text(0, 0, 'Traffic accidents'),\n",
       " Text(0, 0, 'Elderly population'),\n",
       " Text(0, 0, 'Suicides'),\n",
       " Text(0, 0, 'Income'),\n",
       " Text(0, 0, 'GDP')]"
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAl8AAAJ0CAYAAAAs1EboAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3U1wE3e67/FfJ7NEsk9ledS+Vdn4ReKsBjJSsuPFIlVZRDlBrO6AscnqYmewVxfEQLI6FjljVje2wNzVRWbGs6AqyBB2Qc6d7C6SDJtUHbdYpga1vLyTvguudCws2bLcar34+6lyYbpb/X+w2+jx/+X5G47jOAIAAIAn3ul0AAAAAIcJyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD/2m0wEcNv/yL/+if/zjH/qnf/qnTocCAADa5O9//7veffdd/Z//8392nCP58tg//vEP/frrr67d75dffpEkvffee67d0y3E1hpiaw2xtYbYWkNsrTlMse32Xk/y5bFKj9cPP/zgyv0++ugjV+/nJmJrDbG1hthaQ2ytIbbWHKbYKverhzlfAAAAHiL5AgAA8BDJ1wFls9lOhwAAAHpI3875sm1bV69eVblcliSVSiVdunRJ0Wi0pfsdO3ZMtm3vOD47O6tIJHKgWAEAwOHRl8lXPp9XLBbT+Pi4lpeXa47F43HdvHlzX/fLZDJ1Ey9JisfjB44XAAAcHn2ZfJ0/f15+v1+3b9+uHgsGg5qcnFQqlVIkEtlXD9ji4qIWFhZ29HD5/X7XYgYAAIdD3835SqfTsm1bZ86c2XHu3Llzkt4kU82qzOmKRqPy+/01HwAAAPvVl8mXpLrzsEzTlN/vVz6fVz6fb+p+S0tLsm1b8/PzTK4HAAAH1lfJl23b1aQqGAzWvSYUCklqbpWiZVnKZrOyLEupVEoXLlzQ8PCwEolEwzlgAAAAu+mrOV+5XK76+cDAQN1rfD6fpDeJ1V4GBgY0Ozur58+fq1AoVF+TTqeVzWa1urrK8CMAAH1ut2r1jfzyyy8Ntyrqq+Rre29Uo6RocHBQUnPJl9/v19TUVPXv+Xxe3377rdbW1mRZlq5evVozqR8AAGAvfZV8lUqlpq9tJvl6WzAY1O3bt5VOp5VIJLS2tqZsNkudLwAA+lgr+z3u1lvWV8lXo6HGeg4yXBiPx5XJZJTNZpXP5zuafHXj5qQVxNYaYmsNsbWG2FpDbK0htjf6asK9aZrVzxtNiH/9+rUkKRAIHKitynBkKz1oAADg8Oqr5Gv7CsdGQ5CV7YaOHj16oLYqid72hA8AAGAvfZV8Sf+ZgDWq41VZEXnQocJKz5rXyVcoFJJhGLt+VMppAACA7tN3yVdlr8V6dbwsy5Jt2zJNs2EdsGZV7u/1fK9cLifHcWo+JNX8fXvJDQAA0F36Mvny+/169OjRjnOZTEaSNDs7W3PcsizNz8/vmL+VTqerr3nb4uKibt68SZ0vAACwL32XfEnSvXv3JEmJRKJ6LJ/PK5lMKh6P79hUO5FIKJVKaX5+vnrMtm0lEglNT0/r5MmT1Z6ufD6vWCym2dnZai8bAABAs/qq1ERFMBjU06dPdfXqVcViMQUCAZXLZS0sLOxIvKQ3m2bncjl9/PHH1WN+v1+zs7NKp9OyLEsXLlyQaZoaHx/XwsICE+0BAEBLDKcyaQieqBRdc7OeiGEY4tsIAED32O39vi+HHQEAALoVyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB4i+QIAAPAQyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB4i+QIAAPAQyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB4i+QIAAPAQyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB4i+QIAAPAQyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB4i+QIAAPAQyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB4i+QIAAPAQyRcAAICHSL4AAAA8RPIFAADgIZKvFuTzeQ0PDyufz3c6FAAA0GN+0+kA2sW2bV29elXlclmSVCqVdOnSJUWj0QPfe3p6+sD3AAAAh1NfJl/5fF6xWEzj4+NaXl6uORaPx3Xz5s2W7z0/Py/LstwKFQAAHDJ9Oex4/vx5+f1+3b59u3osGAxqcnJS6XRamUympfvm83mtr68rEom4FSoAADhk+i75SqfTsm1bZ86c2XHu3LlzkqTFxcWW7n3t2jUtLCwcKD4AAHC4uT7suLGxoWw2q83NTYVCIX3++eeSpHK5rEwmo4GBAZ0+fdrtZqvS6bQk1e2dMk1Tfr9f+Xxe+XxewWCw6fsuLS3pzJkzMk3TtVgBAMDh41ryVSwWNT09rUKhIMdxZBiGSqVSNfny+XxaXFxUsVjUTz/9pCNHjrjVdJVt29UViI0Sq1AopGw2q2w223TyZVmWHj16pNXVVddiBQAAh5MryVe5XFYsFpNt2zJNU/F4XMlkcsd1V65c0czMjBYXF/WHP/zBjaZr5HK56ucDAwN1r/H5fJK0r0nz09PT+uqrrw4WHAAA6EkfffTRvl/zyy+/6L333qt7zpU5X/Pz87JtW/F4XE+ePNHk5GTd6yplHp49e+ZGszvYtl393O/3171mcHBQUvPJVzqdVjgc3tcQJQAAQCOu9HxlMhkZhqEbN27sea1pmioUCm40u0OpVGr62maSL8uytLS0pO+///4gYQEAgB72ww8/7Ps1u/WWudLzZdt2w56mt+0nQdqvRkON9TQTbyKROFBNMAAAgLe5knyFw2HZtq1Xr17tel2hUJBt2xobG3Oj2R22r0TcPgS53evXryVJgUBg13ul02mZpklNLwAA4CpXhh0jkYjW19c1PT2te/fu1V3JuLW1penpaRmGUa235bbt87JKpVLd3q3KdkNHjx7d9V6ZTEbZbLZauqKeWCxWbZeVkAAAoBmuJF9TU1N69OiRcrmcTpw4oUuXLlXPbWxsKJfLKZlMyrZtjY+PV8tPtEMwGKzW8apXk6uyInKvHq1IJFJdGfm29fV12bZdvYbaXwAAoFmG4ziOGzcql8v6/e9/r0KhIMMwdpx3HEeRSES3b99uS42vinQ6rUQiUXcPR8uydPLkSZmmeaBJ9BcuXFA2m9Xq6uq+V0FWJuC1MnmvEcMw5NK3EQAAuGC393vXthfy+XxaXV3VjRs3NDo6Kp/PJ8dx5PP5FA6HtbCwoLt377Y18ZKkeDwuv9+vR48e7ThX2dNxdna25rhlWWyYDQAAPOH69kLxeFzxeFzSm96wRkN37XTv3j2dP3++ZrViPp9XMplUPB6v1hurSCQSymazsiyrZjNuAAAAt7mWfD148ECBQEDhcLh6bHviVS6XlUwmNTAw0Jbq9tsFg0E9ffpUV69eVSwWUyAQULlc1sLCwo7ES3pT/DWXy+njjz9ua1wAAACuzPlaW1vT9PS0PvzwQ925c6fhdbFYTBsbG1pdXdXo6OhBm+1JzPkCAKD/tX3OVzqdlmEYDbcVqpidnZXjOLuWbwAAAOhnriRflfIN24cc66nU1nr+/LkbzQIAAPQcT7cX8vl88vv9KhaLbjQLAADQc1yZcO/3+xtu5/M227br1gEDAAA4DFzp+aoUGv3xxx93vW59fV2S2ra3IwAAQLdzJfmKx+NyHEfT09MNN9cuFovVvR2/+OILN5oFAADoOa4MO0ajUYXDYa2vr+vkyZOKRqOKRCIKBAIqFovK5XJaWVmR4ziKRqM6ffq0G80CAAD0HNeKrC4vL+vy5ct6/PixMplMdSufikri9ac//cmtJgEAAHqOq9sL3b59W4VCQf/jf/wPbWxsyLIsmaapsbExnTt3bs9SFAAAAP3O9b0dx8bG2B8RAACgAVcm3AMAAKA5rvd8FYtFlcvlPa87rHs7AgCAw8215OvWrVtaWVlpqtiqYRgqFApuNQ0AANAzXEm+UqmUlpaWmr7ecRw3mgUAAOg5riRf3333nQzDUDgc1uzsrEzTdOO2AAAAfceV5KtQKMgwDN2+fVtHjhxx45YAAAB9ybWNtQcHB0m8AAAA9uBKqYloNCrLsty4FQAAQF9zJfmam5uTz+fTl19+6cbtAAAA+pYrw46ZTEbRaFQrKyvKZrMKh8MyTVODg4MNX3Px4kU3mgYAAOgpriRf165dk2EYkqRSqaS1tbU9X0PyBQAADiNXkq+xsTE3bgMAAND3XEm+VldX3bgNAABA32NjbQAAAA+RfAEAAHjItY21JWlra0vPnz9XoVDQ69evd732ypUrbjYNAADQE1xLvtbX1zUzMyPbtpu6nuQLAAAcRq4MO1qWpQsXLqhUKslxHI2Ojmp0dFSO4ygQCGh0dFSBQECO48hxHIXDYTea7XtDQ0MyDGPPD0l7XjM0NNThfw0AAJBc6vlKJpOSpKmpqWqPlm3bOn78uC5duqTPP/9ckrS0tKRvvvlGN2/edKPZvmdZlp4+ferKvU6cOOHKfQAAwMG40vO1vr6ugYGBmqFEv98vSTXDkFNTU3IcR6lUyo1mAQAAeo4ryZdt2woEAjuO+/3+HRPvTdNUNpt1o1kAAICe40ryZZqmisXijuOhUGhHolUqlVQqldxoFgAAoOe4knyNjo7Ktm29evWq5ng4HFahUNCLFy8kvRmetG276RWRAAAA/caV5OvcuXNyHEeLi4s1x6PRqBzH0aeffqqLFy9qYmJChmEoEom40Sy6TCgU2nPVZSgU6nSYAAB0lCvJVyQS0ezsrILBYM1x0zR148YNOY6jZ8+eVUtPLCwsuNEsukwul6uWE6l8SKr5ey6X63CUAAB0lmtFVicnJ+sej8fjGhoaUj6fl2maGh8fd6tJAACAnuPq9kKNhMNhCqsCAACoAxtrb2xseN0kAABA13Al+frggw/0xz/+cc/rEomEYrFYdfUjAADAYeNK8lUqlWRZ1p7Xzc7OynEcffvtt240CwAA0HM8HXb0+/3y+/3K5/NeNgsAANA1PJlwL0lbW1t69OiRbNtWuVz2qlkAAICu0lLylUqldOvWrZpj2WxWo6Oje77WMAyNjY210iwAAEDPayn5ev36dbWAZsXbf2/ENE19/fXXrTQLAADQ81pKvr744gvF4/Hq30+dOqVgMKivvvqq4Wt8Pp8GBwfl8/laaRIAAKAvtJR8+Xy+HUnU4OAgw4kAAAB7cGXC/eTkpI4ePerGrQAAAPqaK8nXuXPn3LgNAABA33Ml+YrFYtra2tLf/vY3HTlyxI1bAgAA9CVXiqwGg0E5jqNcLufG7QAAAPqWK8lXPB6X4zhKJpNu3A4AAKBvuZJ8RaNRhcNh5fN5ffnll9ra2nLjtgAAAH3HlTlfjx8/1pkzZ5TP55XJZJTNZhUOh3ddAXnx4kU3mgYAAOgpriRfly9flmEYkt5Uui+VSlpbW9Pa2lrD17Q7+bJtW1evXq3uI1kqlXTp0iVFo9F93yufzyuZTFbntIVCIc3OzioYDLoaMwAA6H+uJF/dVlw1n88rFotpfHxcy8vLNcfi8bhu3rzZ9L3S6bQSiYTGx8cVDodVKBSUzWYVi8W0vLysSCTSrn8GAADoQ64kX6urq27cxjXnz5+X3+/X7du3q8eCwaAmJyeVSqUUiUSa6gHL5/NKp9P66aef5Pf7q8cvXLigbDarpaUlki8AALAvrky47ybpdFq2bevMmTM7zlWKwS4uLjZ1r++++0737t2rSbwkaXZ2VpJkWdYBowUAAIdN25Kvra0tFYtFz1c+ptNpSarbI2Wapvx+v/L5vPL5/J73mpub25F4Saoe67bhVgAA0P1cT74ePHig06dP69ixYzp16pSOHTumDz74QHfv3nW7qR1s264mVY0mw4dCIUlSNpttuZ1sNivTNDU3N9fyPQAAwOHkypwvSSqXy5qZmVE2m5XjODXnSqWS5ufnq8N47dqCaHuF/YGBgbrX+Hw+Sa0PGVqWpXQ6rdXV1bq9YgAAoL989NFH+37NL7/8ovfee6/uOdd6vqanp/Xs2TP5fD7dvHlTq6urevLkiZaXlzU7Oyufz6d8Pq/p6Wm3mtzBtu3q540So8HBQUn7T75s29bS0pJOnjxZ/Xdsbw8AAKAZrvR8raysKJvNKhgM6i9/+UvNOdM0FQ6HdfbsWcViMWWzWT148ECff/65G03XKJVKTV+7n+TLtm0lk0lZliW/3y/btqvlJr7//vtWQgUAAD3ihx9+2Pdrdustc6Xn6/79+zIMQ19//XXDa/x+vxYWFuQ4jjKZjBvN7tBoqLFRPPu59ubNm1peXtZPP/1Us9qxMsEfAACgGa4kX4VCQX6/X6Ojo7teFwwG5ff7a+Zmuck0zernjYYEX79+LUkKBAIttzM1NaV4PC5JTa2aBAAAqOirOl/bVzg2GoKsbDe0276TzagkX9T6AgAA++FK8jU2NibbtvXq1atdrysUCrJtu1ruoR0qCVijHqlKr9tBK9NXetmocA8AAPbDleQrHo/LcZxdVzJubW1penpahmG0tLn1fmKR6tfxsixLtm3LNM0Db4pd6fEi+QIAAPvhWvIVDoeVy+X0wQcf6MGDB9rY2NDW1pbW19d1584dnThxQsViUZFIpC0rHbfH4vf79ejRox3nKhP9KxPmKyzL0vz8/I4hRNu2G84d+/bbbxWPxw+cxPWqoaEhGYax54ekPa8ZGhrq8L8GAADvuFZk9fbt27p8+bLW19eVSCR2nHccR2NjY1pYWHCryYbu3bun8+fPK5FI6ObNm5LeDEMmk0nF4/EdPW+JRELZbFaWZdVsxh2LxWRZliKRiGZnZxUMBqtlJwYHB6v3Powsy9LTp09dudeJEydcuQ8AAL3AteTL5/NpeXlZmUxGi4uLKhQK1XN+v1+XLl3S5OSkW83tKhgM6unTp7p69apisZgCgYDK5bIWFhbqDnlGo1Hlcjl9/PHHNcdnZ2eVTqeVy+V0/vx5hUIhmaapqampmpWVAAAAzTKct/cCcpFlWRocHKxu6YP/LLrWTME2wzBc7V1y81vdzbEBANBpu73fu9bzVQ+9QwAAALXaknwVi0VZlqVisahAIKCjR4+2bTNtAACAXuJq8nXnzh0tLi7WXSEYDAb19ddfa2RkxM0mAQAAeoprFe4vXryoZDKpUqkkx3HkOI58Pl/181wup08//VR37951q0kAAICe40rydf36dT179qxaTmJ1dVUvXrzQ3/72N7148UJ3795VIBCQ4zian5/Xjz/+6EazAAAAPceV5CudTsswDH344YdaXV3V2NhYzflIJKLvv/9eY2NjchxHS0tLbjQLAADQc1xJvvx+vyTpxo0bu15XKbBa2V8RAADgsHEl+QqFQvL7/QoEArteZ5qm/H6/BgYG3GgWAACg57iSfIXD4YZ7INbz9rAkAADAYeFK8jU1NSWfz7fnSsa1tTWVy+UdG1sDAAAcFq7U+bpz544CgYDm5+f1H//xH9U5YNvZtq2VlRWNjY0pnU43vNeVK1fcCAkAAKAruZJ8zc/PyzAMOY6jlZWVhtc5jqNCoVCz6fbbSL4AAEA/cyX5Yg4XAABAc1xJvlZXV924DQAAQN9zbXshAAAA7I3kCwAAwEMkXwAAAB5yZc6XJK2vryuVSimXy+1ZcNUwjF1XPAIAAPQrV5Kv9fV1TUxMSHpTTgIAAAD1uZJ8LS0tyXEcmaapqakpmabpxm0BAAD6jivJVy6Xk2EYWl5e3nNzbQAAgMPMtQn3fr+fxAsAAGAPriRfv/vd79y4DQAAQN9zZdjxiy++0OPHj3X37t3qxHu448SJE50OAQAAuMiV5CsYDFYTr2fPnikaje45BBkOh91ouu89ffrUlfuQxAEA0B1cq/OVzWarf1Y+b4Q6XwAA4LByJflKpVJKpVIyDKOpOl/UAgMAAIeVK8lXOp2WYRgKh8O6ceMGdb4AAAAacCX5sixLhmHo9u3bOnLkiBu3BAAA6EuuJF9+v1+Dg4MkXgAAAHtwpc5XNBqVZVlu3AoAAKCvuZJ8zc3Nyefz6csvv3TjdgAAAH3LlWHHTCajaDSqlZUVZbNZhcNhmaapwcHBhq+5ePGiG00DAAD0FFeSr2vXrskwDElSqVTS2tranq8h+QIAAIeRK8nX2NiYG7cBAADoe64kX6urq27cBgAAoO+5MuEeAAAAzSH5AgAA8BDJFwAAgIf2Nefr+vXrkt4UVQ2Hw9XjDx482HfDn3/++b5fAwAA0Ov2lXxVNtAeGhqqSb62l5poFskXAAA4jPaVfPl8PhmGIb/fX3OcUhMAAADN2Vfy9dNPP9U9TqkJAACA5jDhHgAAwEMkXwAAAB4i+QIAAPCQK9sL4XA6ceJEp0MAAKDnkHyhZU+fPnXlPiRxAIDDhGFHAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB7q2wn3tm3r6tWrKpfLkqRSqaRLly4pGo3u+17ZbFZLS0vK5XKSpFAopNnZWQWDQVdjBgAA/a8ve77y+byOHTsmSVpeXtby8rK++uorTU9PK5FI7OteS0tLunDhgrLZrGzblm3bymazisViymQy7QgfAAD0MdeTr42NDd25c0fXr1/XgwcPqsfL5bIePHigx48fu93kDufPn5ff79ft27erx4LBoCYnJ5VOp5tOmrLZrBYXF7W8vKyXL1/q5cuXWlhYqG4sPj09Ldu22/JvAAAA/cm15KtYLOqzzz5TLBbT/Py8VlZW9OzZs+p5n8+nxcVFTU9Pa2try61md0in07JtW2fOnNlx7ty5c5KkxcXFpu61tLSke/fuKRKJVI9Fo1Hdu3ev+vdsNnuwgAEAwKHiypyvcrmsWCwm27Zlmqbi8biSyeSO665cuaKZmRktLi7qD3/4gxtN75BOpyWpJmGqME1Tfr9f+Xxe+Xx+1zlblX9LvWuCwaCCwaDy+byeP3/e0jyyZpim6VoBUtM0XbkPAAA4GFd6vubn52XbtuLxuJ48eaLJycm611WSlO09Ym6ybVv5fF6SGiZWoVBI0t49Vn6/Xzdv3mx4PhAISJKGhoZaCbUpm5ubchxnzw9Je16zubnZtjgBAEDzXOn5ymQyMgxDN27c2PNa0zRVKBTcaHaHympESRoYGKh7jc/nkyRZlnWgtiqrKOv1sAEAgP7x0Ucf7fs1v/zyi957772651zp+bJtuzoJfS+lUsmNJhvGUdEonsHBQUkHT76y2awikQjDeQAAYF9c6fkKh8P68ccf9erVK/3zP/9zw+sKhYJs225bfaz9JHYHSb4q88oWFhZavgcAAOgNP/zww75fs1tvmSs9X5FIRI7j7LqScWtrS9PT0zIMo7rq0G2Nhhrraban7m22bSuZTNaUnAAAAGiWK8nX1NSUxsbGlMvldOLECd25c6d6bmNjQw8ePNCJEydULBY1Pj6uzz//3I1md9g+BNio/tbr168l/eeE+f26evVqy5XyAQAAXNte6H/+z/+p3//+9yoUCkomkzIMQ2tra1pbW5P0ZjVeJBLR119/7VaTO2wfziyVSnV7pioT5Y8ePbrv+y8tLck0TU1NTbUeJAAAONRcK7Lq8/m0urqqGzduaHR0VD6fT47jyOfzKRwOa2FhQXfv3tWRI0fcarKuSgJWKTnxtsqKyP2uUsxkMrIsS3NzcwcLEAAAHGqub6wdj8cVj8clvellqpR28Eo8HlcikVA2m90xNGhZ1q7FUxvJZrPKZrMN635VVj4CAADsxZWer2KxqGKxuOO414mX9Cb58vv9evTo0Y5zlT0dZ2dna45blqX5+fm6KyDz+bzu379fN/GybVtLS0sHLlsBAAAOD1d6vmKxmLa2tvS3v/2t7cOKzbh3757Onz+vRCJRTZry+bySyaTi8fiOHrFKT5llWTWbcefzecViMUnS8PBww/Z++umnNvwrAABAP3Kl5ysYDMpxnJoK850UDAb19OlTvX79WrFYTJcvX66Wh6jXgxWNRuX3+/Xxxx9Xj9m2rfPnz+/Z1vj4OCUnAABA0wynsjngAWQyGc3MzCgUCunPf/6zG3H1rUrRtVYKtklv9qZstJigIhgMtj0RNgxDT58+deVeJ06ckAuPIQAAXWO393tXer6i0ajC4bDy+by+/PLLhoVWcXC5XG7PTbS7pQcSAADs5Mqcr8ePH+vMmTPK5/PKZDLKZrMKh8O71tK6ePGiG00DAAD0FFeSr8uXL8swDElviqmWSqWaAqv1kHwBAIDDyJXka2xszI3bAAAA9D1Xkq/V1VU3bgMAAND3XNteCAAAAHsj+QIAAPAQyRcAAICHXJnztbGxse/XjI6OutE0AABAT3El+fr000+rpSaaYRiGCoWCG00DAAD0FFeSL0n72h6GrWQAAMBh5Ury9dNPP+16/vXr18rn80okEopEIvrTn/7kRrMAAAA9x5Xky+fz7XneNE2Zpql//dd/1YsXLzQyMuJG0wAAAD3F09WOwWBQgUBA/+t//S8vmwUAAOganpeaME1T6+vrXjcLAADQFVybcN8sy7JULBa9bhYAAKAruJJ83blzp6nrstmsLMvSwMCAG80CAAD0HFeSr/n5+abqfFVKTESjUTeaBQAA6DmuJF9jY2NNXTc4OKhIJKLJyUk3mgUAAOg5riRfq6urbtwG8MTPP/+sTz75RC9fvtTw8LAePnyo999/v9NhAQAOCTbWxqHzySef6MWLF/rHP/6hFy9e6JNPPul0SACAQ8Tz5KuVTbgBN718+VK//vqrJOnXX3/Vy5cvOxwRAOAwcSX5+uCDD/THP/5xz+sSiYRisZhevHjhRrNAS4aHh/XOO28e/XfeeUfDw8MdjggAcJi4knyVSiVZlrXndbOzs3IcR99++60bzQItefjwoUZGRvTuu+9qZGREDx8+7HRIAIBDxNMiq36/X36/X/l83stmgRrvv/8+zyAAoGM8S762trb06NEj2batcrnsVbMAAABdpaXkK5VK6datWzXHstmsRkdH93ytYRjjwDtmAAAgAElEQVRN1wUDAADoNy0lX69fv65Wq694+++NmKapr7/+upVmAQAAel5LydcXX3yheDxe/fupU6cUDAb11VdfNXyNz+fT4OCgfD5fK00CAAD0hZaSL5/PtyOJGhwcZDgRAABgD66UmpicnNTZs2fduBWALvXzzz8rGAzqN7/5jYLBoH7++edOhwQAPcmV5Gt2dlbj4+Nu3ApAl+rmbZlIDAH0EldLTWxtben58+cqFAp6/fr1rtdeuXLFzaYBtFk3b8tUSQx//fXXamJILTcA3cq15Gt9fV0zMzOybbup60m+gN4yPDxcTXC6bVumbk4MAeBtrgw7WpalCxcuqFQqyXEcjY6OanR0VI7jKBAIaHR0VIFAQI7jyHEchcNhN5oF4KFu3paJ/ToB9BJXer6SyaQkaWpqqtqjZdu2jh8/rkuXLunzzz+XJC0tLembb77RzZs33WgWHWSapk6cOOHavdD9unlbpocPH+qTTz7Ry5cvNTw83FWJIQC8zZWer/X1dQ0MDNQMJfr9fkmqGYacmpqS4zhKpVJuNIsO2tzcrPZk7vYhac9rNjc3O/yvQa+rJIb/9//+X+Xzeb3//vudDgkAGnIl+bJtW4FAYMdxv9+/Y+K9aZrKZrNuNAsAANBzXEm+TNNUsVjccTwUCu1ItEqlkkqlkhvNAgAA9BxXkq/R0VHZtq1Xr17VHA+HwyoUCnrx4oWkN8OTtm03vSISAACg37iSfJ07d06O42hxcbHmeDQaleM4+vTTT3Xx4kVNTEzIMAxFIhE3mgUAAOg5riRfkUhEs7OzCgaDNcdN09SNGzfkOI6ePXtWLT2xsLDgRrMAAAA9x7Uiq5OTk3WPx+NxDQ0NKZ/PyzRNtiECAACHmqvbCzUSDocprAoAAKA2JF8bGxvKZrPa3NxUKBSqFlgtl8vKZDIaGBjQ6dOn3W4WAACgJ7iWfBWLRU1PT6tQKMhxHBmGoVKpVE2+fD6fFhcXVSwW9dNPP+nIkSNuNQ0AANAzXJlwXy6XFYvFlM/nFQgENDs7W61uvt2VK1fqrooEAAA4LFxJvubn52XbtuLxuJ48edJw8n00GpUkPXv2zI1mgb4QCoVkGMaeH6FQqNOhAgBc4ErylclkZBiGbty4see1pmmqUCi40SzQtG5OcHK5XFN7YuZyOc9jAwC4z5U5X7Zta2BgoKlr2VoInVAvcTEMo+7wOAAA7eRKz1c4HK67vdDbCoWCbNvW2NiYG80CAAD0HNcq3DuOo+npaW1tbdW9ZmtrS9PT0zIMQ+fOnXOjWQAAgJ7jSvI1NTWlsbEx5XI5nThxQnfu3Kme29jY0IMHD3TixAkVi0WNj49Xy0/0omw2q2w22+kwAABAjzIclya9lMtl/f73v1ehUJBhGDvOO46jSCSi27dve1Ljy7ZtXb16VeVyWdKbuWaXLl2qrrjcr3w+r2QyqWw2q9nZWU1NTbV0n48++kiS9MMPP7T0+l7TzfOqOhHb0NCQLMty5V6maWpzc9OVewEA3LXb+71rRVZ9Pp9WV1eVTqeVTqdVLBZl27b8fr9CoZDi8bhn+zrm83nFYjGNj49reXm55lg8HtfNmzebvpdt20qn07Isix4vHJhlWXr69Kkr9zpx4oQr9wEAeMv17YXi8bji8bikN71hPp/P7Sb2dP78efn9ft2+fbt6LBgManJyUqlUSpFIpOkeML/fX+3lyuVyyufzbYkZ7tlP71K9Xtrt6F0CALit6eTrs88+09DQkP793/+96Zt3IvFKp9PVgq9vO3funFKplBYXF1safmy2nAY6i94lAEA3a3rCfT6f33PY7fr16zWT7TshnU5LerMC822macrv9yufz9ODBQAAOsKV1Y4V6XRa3333nZu33BfbtqtJVTAYrHtNpYI587cAAEAnuD7nq5O2VzFvNERYGQp1a8UZAADob5WVi/vxyy+/6L333qt7ztWer06zbbv6ud/vr3vN4OCgJJIvAADQGX3V87WffSNJvgAAQDNaqc25W29ZX/V87Wc1YqOeMQAAgHbqq+TLNM3q59uHILd7/fq1JCkQCHgSEwAAwHZ9Ney4fYVjqVSq27tV2W7o6NGjnsUF71Gfyx2hUKipsizBYLBmwQsAoLF9J1+vXr3adT+8crm85zVS+3qegsFgtY7X9p6wisobRL06YOgfFFl1R72Eqpv36wSAXrCv5Mu2bZ08ebLhecMwZFnWrtdUrisUCvtpumnxeFyJRELZbHZHFXvLsmTbtkzTbFgHDAAAoJ32NefLcRzXPtolHo/L7/fr0aNHO85lMhlJ0uzsbM1xy7I0Pz/PCkgcakNDQzIMY88PSU1dNzQ01OF/EQB0p6Z7vm7evNnOOFx17949nT9/XolEohp3Pp9XMplUPB7f0SNW6SmzLKtmM+63kZyhn7m5J6bEkC0ANNJ08nX27Nl2xuGqYDCop0+f6urVq4rFYgoEAiqXy1pYWKi7oXY0GlUul9PHH39c937pdFrPnj2rJl+Li4uyLEuRSKSlDboBAMDh1VerHbfz+/279mJtF4/HFY/HWz4PAADQrL6q8wXg8AiFQk3NPQuFQp0OFQBq9G3PFw4v0zRdm29Ur1wJugNlMAD0Knq+4Jp6PRGSPO+F2NzcbHrF7V7XbG5utj1eAMDhQs8XXEOFc7TL0NBQ06uNK0n/bkzTJLEG0DEkXwC6HmUwAPQTki/AY7zxA8DhRvIFeIx9JwHgcGPCPQAAgIdIvnAoNLMSk5pQAAAvMOyIQ4GVmACAbkHPFwAAgIfo+QJQxSR+AGg/ki8AVdTSAoD2Y9gRAADAQ/R8AR5i0+/W9VJPWigUUj6f3/O6YDDIYhDgECL5AjzU7H6ChmFUN//GG700JFovoeqW7ymJIdB5JF8AcIh0c2IIHBbM+QIAAPAQPV8AJLk7H61yPzfv1a2xAcB+kXwBkNTd89G6OTYA2C+GHQEAADxE8gUAAOAhki8ADYVCIRmGUfMhacexUCjU4UgB7/z8888KBoP6zW9+o2AwqJ9//rnTIaHHkHwBaCiXy8lxnD0/DnM9qKGhoR3JaL0PaWfSWu9jaGiow/8i7OWTTz7Rixcv9I9//EMvXrzQJ5980umQ0GOYcA90WKOil5U37AqKXnYny7J6qgAsDu7ly5f69ddfJUm//vqrXr582eGI0Gvo+QI6jN4l4I16w9z1Pjo9zD08PKx33nnz9vnOO+9oeHi4o/Gg95B8AUCf6rUh0Xq/iEjqul9EHj58qJGREb377rsaGRnRw4cPOxoPeg/DjgDQpxgSbY/333+/qf0xgUbo+QIAAPAQPV8AcED0CAHYD5IvADgghvYA7AfDjgAAAB6i5wsA+hg9ae5oVI/vbdTjQzNIvgD0JIrTNochUXfUe4YMw6iWwwD2g+QLQE/qloTKNE1XkxLTNF27VzcbGhqSZVlNXft2Ql2PaZra3Nw8aFiAJ0i+AOAAmn3Dp5ekFjXI2uPnn3/WJ598opcvX2p4eFgPHz7U+++/3+mw8BYm3AMA0CfY9Ls3kHwBANAn2PS7N5B8AQCwTa/tibkdm373BuZ8AQCwTS/PR3v48OGOOV/oPiRfAAD0CTb97g0kXwDQp7q9DAYrFHFYkXwBQJ/q9jIYvTq0BxwUyRcAuIzq+3ujVw6HGckXALjssCZU+0GvHA4zSk0AANDDQqFQUyUvQqFQp0PF/0fPFwCgK3TLcG23D4m+jU2/D87rbZlIvgAAXaFbhmu7fUgU7qtsy/Trr79Wt2VqZ8kOhh0B4BCpN0Ql7azUzhAV3NArQ6Jeb8tEzxcAHCLd0ruEw6FXhkSHh4erPV9ebMtEzxcAAD2im/ed7ObY9vLw4UONjIzo3Xff1cjISNu3ZaLnCwCAHmFZVtfer5f3xPR6WyaSLwAA9tAtKzGl7q5BRk2z5vRt8mXbtq5evapyuSxJKpVKunTpkqLRaEfvBQDoPcyV25tpmq72pLlZomNoaMj12JpdFVtPXyZf+XxesVhM4+PjWl5erjkWj8d18+bNjtwLAIB+1c0lOrptSNRwum3JgQuOHTsmSfrpp59qjs/PzyuVSmlhYaHpXis37yVJH330kSTphx9+aPo1AABI3deDs5dGw7Vva/dw7dvDw27YK33a7f2+71Y7ptNp2batM2fO7Dh37tw5SdLi4qLn9wIA4KA2NzflOM6eH5Kauq6diZf0Zri2mTjaPazr9i4DB71fXyZfkhSJRHacM01Tfr9f+Xy+qUzczXsBANAOFM7dW72kNRgMNvXaYDDoetLaV8mXbdvVRKjRF7Xy8GWzWc/uBQBAu3RL71Kv6eTXra8m3G//Ag0MDNS9xufzSdq7tomb9wIAAL2rMn9rP3755Re99957dc/1Xc9Xhd/vr3vN4OCgpL0TJjfvBQAAUNFXPV+lUqnpa/dKmNy8FwAA6F2tVCjYrbesr3q+Gg0P1tOoN6sd9wIAAKjoq+Rr+9LP7cOG271+/VqSFAgEPLsXAABARV8lX9tXJTYaNqxsEXT06FHP7gUAAFDRV8mX9J9JU6PaW5VVjPVqd7XzXgAAAFIfJl/xeFxS/dpblmXJtm2ZptlUcTU37wUAACD1afLl9/v16NGjHecymYwkaXZ2tua4ZVman5/fsWqxlXsBAADspu+SL0m6d++eJCmRSFSP5fN5JZNJxePxHRthJxIJpVIpzc/PH/heAAAAu+mrOl8VwWBQT58+1dWrVxWLxRQIBFQul7WwsFA3WYpGo8rlcvr4448PfC8AAIDdGE5l+3N4olJ0rZWCbQAAoDfs9n7fl8OOAAAA3YrkCwAAwEN9Oeerm/3973/Xr7/+2tIO6QAAoDf88ssveued+n1cJF8ee/fddzsdAgAAaLN33nmn4Xs+E+4BAAA8xJwvAAAAD5F8AQAAeIjkCwAAwEMkXwAAAB4i+QIAAPAQyRcAAICHSL4AAAA8RPIFAADgIZIvAAAAD5F8AcAhc/36dT148KDTYQCHFtsLAYCHtra29Pz5c0lSOBzuSAynTp2SYRh6/PhxR9pvRTd83XoRX7edHj9+rOfPn2toaEhnzpzRkSNHPI+B5KvHPHjwQOFwWIFAoO55y7KUSqU0NTXV8Jp22djY0OjoqKdttqJYLO563uuv29u6OT5ia97KyopSqZTu3r1bbXttbU0zMzPVa4aGhvSXv/zF8//8l5aW9M0332h5eVm/+93vPG17L938dduO523/isWi0um0LMvSF198UX2/WFtbUy6X05UrV9ravmVZunjxoizLUiX1eeedd7SwsKBTp061te23kXz1mJGREc3NzenixYsNrzl16pQ+/PBD/fGPf/QuML2J7b/8l/+ieDyus2fPdvQ/xnouXryobDa76zWGYahQKHgUUa1ujo/Y9i8Wi2lra6vau1Qul3Xs2DEZhqErV65oc3NTKysrOnfunOc/q5X4Xr16pa+//trzN57ddPvXjeetNTMzM1pbW5PjODIMQ3fv3q32xK2vr2tiYkK3b99u67N46tQpWZal8fFxffjhh/qP//gP3blzR4Zh6K9//atGRkba1vbbfuNZS/DM2NjYnv85tEM4HNb6+rr+7d/+TfPz84pEIopGo/r88889j+Vt09PTevbsmaQ3X5/BwcEOR1Srm+MjttYUCgVFo9Hq3x89eiRJOnv2rCYnJyVJuVyuIz+rMzMzGhoaUqFQ0OXLl+X3+xUKhepee+fOHU9j6+avG89ba1KplDKZjOLxuCYnJ3ckWJXRnPv377ct+UqlUrIsS1999VXNe9K5c+d06tQpXb16VX/+85/b0nY9JF894O3u7VKp1LDLu1AoaH19XeVy2YvQaiwvL6tcLuu7775TOp3Ws2fPlM1mlUgkFI1GFY/HOzbEsb6+roGBAX3//ffy+XwdiWE33RwfsbXONM3q57lcToZh1LxBhkKhjkx8z2Qy1c8dx1GpVKomFdsZhuFlWFXd+nXjeWvNd999p2AwqBs3bjR8bwqHw1pbW2tbDNlsVsFgcEdngGmaunjxou7evautrS3PRmxIvnrAyZMna/4TXFpa0tLSUsPrHcdRJBLxIrQdfD6f4vG44vG4yuWy7t+/r0ePHunRo0fKZDLy+/3VYUkv50TYtq14PN6V/2FK3R0fsbXG7/fX/JJUSXi2T3ouFovy+/2ex/bkyRPP22xWN3/deN5aUygUFI/HPW93u1wu1zCGjz76SHfu3NHz5889W5RA8tUDzp49W/18ZWVFgUBAY2NjDa//8MMPa17TKT6fT1NTU5qamlK5XNa3336rVCqlxcVFLS0taWhoSJOTk54MS4bDYZVKpba306pujo/YWhONRvXgwQM5jiPLslQul3f8XGaz2ZqeCa9s7yHpNt38deN5a83Y2JjW19d3vSaTyejo0aNti8G2bQ0NDdU9V+kI8HTEyEFPGR4edlKpVKfDaEq5XHZWVlacWCzmjIyMOCMjI87JkyedpaUl59/+7d+cY8eOOcPDw87p06edYrHY1lg2NzedY8eOOSsrK21tp1XdHB+xte7TTz91hoeHneHhYefTTz+tObe0tOQMDw872Wy2Q9G9+RlNJBLVn9FYLOZcv37dKZfLHYvJcWq/brFYrOZcJ79uPG+tuX//vjM8POzcunXLKZfLNXGUy2VnYmLCGRkZcdbW1toWw/DwcMPv2+bmpjM8PNzW9t/Gascek0gkdObMma6u1/L48WOl0+nqxE7HcTQ+Pq5z587tiDuTySiRSOif/umf2jrePzMzo3w+X+12DwQCDSfLej3BWOru+IjtYCq/Tb89VFUul5XL5Tr2s7y+vq6ZmZm6PTmdWn6/XTd+3XjeWnf58mU9fvxYAwMDsm27OjUmm83KcRzF43HduHGjbe2PjIzsmGxfYVmWTp06pdu3b+v06dNti2E7hh17zM2bNzsdQkO3bt3SysqKbNuW4zgyTbM6/6vRHIloNKpsNtv2SaDbJxiXSqWGQwedmmDczfER28GUSqW6tY2y2WzH3gwty9KFCxfk9/u1sLCg8fFxSaoumEkmk7p8+bK+//57/fM//7Onse1Vy/D169fKZDIyTdPzWlo8b627ffu20um0ksmkHMepLvAwTVOzs7PVZ7CdNjc36w5/VhLW58+f7zqfz82vHT1fPazbivxVaqQ06uVqZGJiQpZltXUSsGVZTV/bifkw3RwfsbWuG2ob1TM9Pa3Hjx/ryZMndf+fqPQEnDlzRv/+7//uaWzdXMuQ580d5XJZlmXJNE3PFi+MjIzsmhRXvmaNuF27jZ6vHtStRf5mZ2dbWgm0sLDQ9h/Abp5gLHV3fMTWmm6obdRIoVDQ+Ph4w1/QTNNUJBJRPp/3NK5mdaqWIc+bO3w+366LxtrB6/b2QvLVY7q5yN/k5KTu3LmjSCRSd5uhlZUVZbNZ/elPf6o53o3LtoGD6obaRo2USiUNDAzsek0gEPAs+eqVWobdrJuft26wurra6RBqkHz1mG4u8pdKpXTr1q2GD3kwGFQikdCDBw86VvV+a2tL8/PzyuVyKhQKGhsb09GjRzU7O9sV2yF1c3zEtj/dUNuokWAw2PGl/9v1Ui1DieetFRsbG/ruu+92Hbo1DMPzYe5OIfnqMd1c5C+dTmtsbKzh5trBYFCmaSqTyXQk+aq3uiufzyufz2tlZaXjq7u6OT5i279uqG3USDwe18zMjL788ssdb3ZbW1u6evWqyuWyZ2/mvVTLkOdt/7Zv8L3bNPPDlHxR56vHnD9/3pmenu50GHUNDw/vGdvly5ed48ePexTRf6rUcTl27JiTyWSqx23bdu7fv+/89re/dUZGRtpeb6wX4yO21nRDbaPdXLt2zRkeHnaOHz/uTExMOIlEohpTMz/L7dLNtQx53lpTqT927do1J5vN7vrhtWw266ysrDjJZNJJpVLO2tqaJ3XuSL56TDcX+RseHnYmJiZ2vaZTydfly5edkZERx7Ksuucr/6nOzMx4HNkb3RwfsbXuv/23/1ZNcEZGRpyJiYmaBCeRSHQkropHjx45v/3tb6uFOSuJRTqd7lhMlTfobsTz1ppm3hu8ZFmWk0gkqsW/K1+f7X+fmJhwNjY22hYDw4495tatWxoYGFAikVAymeyqIn+VFUiNNictl8taX1/vSLd3t6/u6ub4iK113VDbaDfRaFTRaFSWZalYLCoQCHR8RV831zLkeWtNJ2qyNfLgwQMlEglJqtajHBsbk2masixLlmWpUCjo2bNnymazmpub08TEhOtxkHz1mG4u8jc7O6uJiQmdP39eCwsLNcUZi8WipqenVS6XNTk56Xls3ba6623dHB+xHcz2jea9rm3ULNM0O5509QKet9acPXu27YW0m1FJvBzH0dmzZzU3N9fwazM/P687d+5ofn5eklxPwEi+ekw7C5EeVCQS0cWLF3Xnzh2dPHmy+h965bcJx3E0OTnZkQrL3ba6623dHB+xuaMTtY32o9Lz1U26rZA0z1trpqamlMvldPr0aU1NTSkUCjW8ttGCrYOyLEvXrl3TwMCA/vznP+/5y8bc3Jy++OILxWIxzc/P6/Tp064+byRfPabbfzudm5vThx9+qGQyqUKhoM3NTUlvhiTn5uY6trVFt63u6qX4iG13eyUIzehE0rO1taVvv/22uiVYxdDQkCYnJztWDkbq3kLSPG+t2draqv4SXhnyq6ed39N0Oi3DMHTv3r2m30cr22999tlnSqVSru6owPZCPapYLNbdv2ttbU25XE5XrlzpcIRvVLq9u0EikdDKyooGBgYUCoUUCARULBarG7tGo9EdBWCJj9j2ste2JXvpRBJhWZb+9V//tWYf1u291IZhKBKJdGRz6Onp6Woh0L0KSd+9e9ersKp43vZvYmKimkzvVZ+tXd/TU6dOaWhoqKVnemJiQsViUY8fP3YtHpKvHtQr+3d1o0wmo2vXrtVUgPb7/Zqdne1Y3aDtujk+YqvvwoULDd8Ms9ms/H5/3Z6GQqFQPed19e3Km2EkEtGNGzdqfkEql8v67//9v+vJkyeamprSH/7wB09jO378uAzD6MpC0hU8b/szMjLS8eLgzewZ2kilgPjGxoZr8ZB89ZhUKqVkMlmzf9fy8nLNcN5BMnw3rK+vK5vN1gxlbGcYhucb4r6tm1Z31dPN8RFbcwqFgs6fP6/V1dW6b4b5fF4XLlxoeL6dKm+G//t//++G18RiMW1tbbn6234zRkZGFI/HdePGDU/bbQXPW3OOHz+uM2fOdPR7OjIyoq+++qql4fSVlRVdv37d1eSLOV89ptv376rM1ajk9IZh1Hxe0enkq9tXd3VzfMTWnMXFRUUikYZvdMFgUOFwWLdu3fK8qnelJMJu4vF4R35Ow+Fww1Xc3YbnrTm/+93veuZ76pV3Oh0A9qdQKOy6UqSTksmknj17ptOnT2thYUF3796V4ziam5vT8vJy9TeOhYWFDkcKtF82m92zLMHAwMCeE8vbIRwO7zlx2zCMtq08283NmzeVzWa7ojRBL+nm521ubo7v6Vvo+eox3b5/14cffrgjuTJNU+FwWOFwWLlcrrrkuB3K5bI+++wzGYZR0/tX2VdsL+3eW6yb4yM292UyGc3NzTXcbHmvn+V2mZqa0unTp/Xq1auaenwV5XJZqVSqI8NE3VBImufNXd3wPZWkzc3Nlr4GlVX7biL56jHxeFzXr1/XN998o0uXLtWc29raqhYy7cSyf8uyFI1Ga46ZplnT3RyJRJTJZNo6iXdzc3PHhNTtxWl348V/mt0cH7G5JxqNamVlpW7R4crParFYbGvV8d0mF/t8PsVisbrDVMViUbZt69atW/rzn//ctvjq6ZZC0jxv7umW72kqlVIqlWprG81iwn0Punz5sh4/fqyBgQHZtl2du1GZa9Wpyar1JvpPT0/LMIzq0uvp6Wk9fvzY1YmLb6v8ZrN9EYJlWU2/vt1zOLo5PmJzVywWU6FQkGEY8vv9CoVCNUWHTdPUX//614Y9FQc1MjJyoNcbhtHWn9V6uuV7yvPmnm74usVisQPfw81VoiRfPaqyf9f2Sfed3r8rkUhobW2tZgVVOp3WH//4R/31r3/V3//+d124cGHPVVZAP1laWtLi4uKOsgSXLl1q+1Zb+3nTa6RbJpSjOZ183rAPbduyG56wbdvJ5/OObdudDsXJ5XLOsWPHnLW1tZrjJ06cqNk5PplMeh7bysqKY1lWw/Obm5tOIpHY9Zp26ub4iM09m5ubnQ6hZ1iW5SSTSWd6etopFArV45lMpiP/hzgOzxvcQ/KFttvc3HSuXbvmXLhwwclkMh2JYXh42EmlUrtec/LkSef69eveBPSWbo6P2OC16enp6i9rIyMjTjabrZ7LZrPOyMiI8/jxY8/j4nk7mHK57CQSCScWizkjIyNOLBZzrl+/7pTL5U6H5jkm3KPtTNPUzZs3Ox3GnsbGxjqyDLtZ3RwfsfWmbiyInEqllMlkagpJbxcOhxUIBHT//v2u3MWD562+9fV1zczM1Ey2z+fzyufzWllZ0cLCQld+P9uF5KsHbWxs6Lvvvtt1PkcnVttcv35doVCooxvybvd2HaNSqdSwtlGhUND6+nrDwrXt0M3xEVvzerUsQbcWRO62QtI8bwdnWZYuXLhQ3ai6Mi+5XC7ru+++UzKZ1OXLl/X999/XLX3illu3bmloaKjhe1QqldLKysqO436/3/VVv0y47zFra2vVH7LdvnWdWKV06tQpGYbh+XYkjWzfgNb5//tg7sZxHEUiEc826+3m+IiteeVyWceOHdvxM9fsSsNO/Kwmk0mlUimNj4/r448/ls/n08TEhObm5jQ2NqZHjx7pwYMHWlhYaFtNvka2by9U+dq+vYVavcU97YyH5+1gKqvcnzx5Ure0iWVZOnXqlM6cOdO2xHB9fV0XLlzYdePzys/F2wzD0PLysn73u9+5Fg89Xz3m22+/leM4Onv2rFDctmsAACAASURBVM6cOdPpcGqcPXtW33zzjX788UdXH9KDxFOxsrKiQCCgsbGxhtd/+OGHnm4Q3c3xEVvzfD6flpeXdxx/8uRJ29o8qG4oiNxItxWS5nk7uEKhoPHx8YZbH1W2u8rn822L4f79+zIMQ19//fWu19VLTkdGRpROp919X/N+mhkOYnh42JmYmOh0GA19+umnzvHjxzsyGXY3zUyU7aRujo/Y+k+9VccnT550VlZWqn+fn593Tp065XVozv37953h4WHn1q1bTrlcdoaHh6sT7svlsjMxMeGMjIzsWFXtBZ631hw7dsxJJBK7XnPt2jXn+PHjbYvh5MmTez7P8/PzzsjISEuv3S96vnqMaZqe70jfrJmZGQ0NDalQKOjy5cvVIn/1tHMLiXrOnj2762+rndbN8RFb6+7cuaNIJFJ3j8SVlRVls9mGQyDtZJqmCoVCzbGxsTE9e/asOh+mUpzTa/F4XM+ePdPi4qLS6bQMw6hWJt9eSNrrHjmJ561VwWCw472Z9XZgeZvzpgLEjuNjY2PuT6dxNZVD2y0uLnbkt9FmDA8PN/VR7zcLoN8sLS05IyMjNTWqtsvlcs7w8HBNb5NX6vUy3L9/3xkZGXE2NjacbDbrDA8Pt7UnYi/37993fvvb39b833Hy5MmOlavpdt38vD169MgZHh52ZmZmdpwrl8vV0iLt7M0cHh52pqenW3rt5cuXXX/fYsJ9D5qentbGxoampqYa9ixJqvvbTzt1wxYSe+nGpfXbdXN8xLY/p06dkt/v11/+8pddr3l7Sy4v5PN5XbhwQV9//XVND9LJkyf16tUrSW96AaampnTlyhVPY3tbuVyWZVkyTVM+n6+jsVTwvO1fIpHQysqKBgYGFAqFFAgEVCwWq72Zu02Ed0MsFtOrV69aWqRx/Phxmaa569d2v0i+eszW1pb+63/9r3uuVjEMY8ewwmHX7NJ6r1cC9UJ8xLZ/IyMje76hTE9P68cff+ya7bYsy9LS0pKKxaLi8XjHtirrZjxvrctkMrp27dqOrY9mZ2fbvmBnaWlJ33zzjebm5jQxMdH061KplG7duqXZ2dldN6rfL+Z89ZjLly9Xk6rKhtrdplgsKp1Oy7IsffHFF9UeuLW1NeVyuY78Jp1MJvXs2bOmltZ3QjfHR2yt87JunBu8LIjcqFbWfng9/5Xn7WCi0aii0agsy1KxWFQgEPBsFOTcuXNaXFzU/Py8TNNsqqDr48ePlUwmNTAwoHg87mo8JF89JpvNamBgQN9//33XdMFvNzMzo7W1tWo9nO0PrN/vVyqV0r/8y794Xsm4m5fWd3t8xNaaSqXzra0tHTlyZMf5crms9fV1z0omdJuTJ0/uWTNrN53o3ed5a932xQCmadYkXV4sBvD5fFpYWNDExIQuX76sSCSiubm5uvXRNjY2lEwmqzsV/OlPf6r7NT2Id1y9G9rO7/crGo12ZeJV2Rbk7NmzevLkyY5VI9u3BfGaZVk7VimZplmz1UUkElEmk/E6NEndHR+xtWZ2dlaO4+j8+fPVeVQVxWJR58+fV7lc1uTkpOexNWNlZaXpqumtqCQr9T4cx5HP59Po6OiOj8q5ZouKuonnrTWpVErJZLLh+WAwqEwmowcPHrQ1jkrx2yNHjujZs2f69NNP9cEHH+izzz7TzMyMPvvsM/2/9u4nKo10zR/4t/r8liOk1215lxLAu2q9oZzVXI0ku+gMldVMRDGrEe5t2SWY2L0aMXfMLgG1dyM41+wUMH13FvbNMoC97RRZp/mzZxb5UVci4F+q3sLv55w+Jy2lPhZF1VNvve/z/OEPf8D09DQODw+NmmonC/xeF4582cydO3daPugiEa0tyEkiL60HxI6PsV2OoiiYm5vDxsYGJiYmjLv9ZjyNRgPz8/M9ObFfh0Kh0NPPartCocDngpyPHj3C7u5u28eKzcUCL1++7FlsnfB4u5xUKgW3291xEZjH44Esy8hkMj1vT6coCv72t79hdXUV6XQalUoFlUrlVIHXQCCAaDTas4EOJl82E41GMTMzg52dHWF6KDaVSqVrfy5+XdolfYqi4NmzZ/jll1/w6dMnZLNZOJ1OxsfYrk00GsX4+Dji8ThKpRI+fPgA4PMFOxqNCpt4Wen169dQFKXjfC6PxwOfz4e1tTXTexTyeLuc89TYcrvdODo6MiWegYEBrKysYGVlBfl8Hrquo1qtwuFwGI+Qe43Jl82sra3B6XQiFoshHo9jcHAQt27darut2cuJRWsLcpKqqshkMsjlcsZcDFVVkUgk8ODBA2M7M9sL2SU+xnY1iqJgd3cXAIySCdSZpmlntk5zOp2WjKDzeLs8URcDNB91m42lJmxG5OapqVQKy8vLWFhYwMLCAr799lvjeXm9Xkc4HIamaZY06+1E9KX1IsfH2PpbLBbDzs6O6eeRsbExSJKEn376qeMk58nJSVSrVZbosInp6WkcHx/j3bt3HRcD/PGPf8TIyIglNciswOTLZkQvZLq4uIhcLgen04lqtWqUwzjZFuT58+emx0XUK7VaDTMzM5AkqWU05ryT1SVJMv3x2XlYlXw1i3F6vV6sr6/jm2++MV47eRM3NTVlSascujhN0xAMBtu+p+VyGeFwGKVSCZubmzfmUTyTL7p2qVQK8Xi8ZZhZlmUsLS3xjtCGdnZ2jJWq7ei6jmQyiVAoJGzf0V6q1WoYHR09Ndos8ij1eViVfAGfR0pKpRIkSTJ6xJ6cOC7LMt68eXPty//twK7J/urqKjY2NiBJUsfFAEtLS6bHZRUmXzZzfHxsetugyxKxLQhdnMvlQjQa7VrdeXJyEuPj45a2ZrJSc67jybt2EUapz5qD2U0ikUA+n7csMUwkEnj9+vWpaugLCwvClucwg52TfU3TjMUATVYvBrAKky+bcblc+N3vfgdVVREIBIS68xNhhOSqdYl6fVcoenztnCf5avYbzeVyPYnBjvtNBC6X60qFTAHr2m2dZPbEcdGPN1GT/YsQaTGAFZh82czs7KzxwZMkCYqiwO/3C1F2QoQRkqsWXuz1XaHo8TWdbP0yMTGBhYWFjiu4SqUSnjx5glqt1rPY7LLfRHPVKvIAcHBwcE3R2AePN+o1Jl82VKvVsLe3h1QqZcyLAD73zVJVFXfu3LEkLhFGSK6juGEv78ZEj6/p5IhJs1VUN41Gw6ge3Qt22W/UH3i89c5ZPT1vyrxRJl82V6vVsL29jf39/ZYJqs3Hkr0+kEUbIaHrEYvFjH+n02nIsnyqrcpJ4+PjltY2EtXx8TH29va6Xsxv6iNRu04cF9VFGpVbkeDMzc0ZvRI7saJfp1WYfPWRWq2GV69eIZlMAvh8IA8NDWF+fr5njyVFGyGh63eeEU06LZvNGolEt9PsTX1EZeeJ4yI67/w+KxKccDhsJNhut7tjYXAAN+bawAr3Nlev17G/v4/t7W3jAyXLMlRVxadPn7Czs4OnT58imUxic3Ozpb7KdTg52pFOpzE4OCjkCInod4UixxcIBLq+p1YSeb+9evUKjUYDgUDgzIrtN1GzafGXRJ5jJvLx5vP5OiZfzZIOTqfTkmkp+XweTqcTb9++5cr3/48jXzaVy+WQSqWMYdxGo4GpqSk8fPjw1JLdTCaDWCyGr7/+uqctOUQeIRH5rhAQPz5RibzfXC4XxsfHb0zF7ptA5OPtLJqmYW5uDgcHB6Ynhi6XiwW2v8CRL5tZW1tDOp1GtVo1ig2qqgpVVTveUfj9fmiahp2dnZ7GJvIIich3hYD48YlK5P0my/KNmTx83TY2NqAoStuahul0GpqmWVLdXuTj7SyKouDu3btIJpOm1+Pz+XyoVCqm/k7RceTLZprzITqNcnUSDAah67rQQ/pWsvKu8DzMiK8fJ0Bb+b4mEgns7Oz0bGVvv0omk1hbW8Pu7m7b5KtYLGJmZgbff/+9ECV2ThL9PJJMJpFOp00/JnVdx8zMDKLRqHDvmVWYfNlMMpnsOsrVSa1W47P2M4TDYXz99dfCVmnvdXz9OgHayve1WVolFArB6/V23M4uXSvMMDk5CYfDgb/+9a9dtxkaGhLyka7I55FwOIxcLmf65zQSiaBYLKJcLsPhcGBwcLDjpHsR39Ne4GNHm2m21ajX61hdXUWhUECpVILb7cbIyAiWlpbaVr2/zsSrH0dIAGBkZATpdNrqMDrqdXx2nAB9Hla9r/V63XgUdbJ0x5dEnB9kJV3X4ff7u27jdrtxdHRkUkQXY9XxdtZc2+ax6HA4TIroHzKZjPHvSqXS8RHkVQsC2wmTLxvK5/OIRCItB3CxWESxWEQ6ncb6+jomJyd7GsOHDx9OfVBOfsC6ETX5ev/+/bUUV+wVM+Jr9xjb7sUirXpfFxcXjaRKURTTf7+dneznaDdWHW+Hh4dnbuNwOLCysmJCNK3sfgPXC0y+bEbXdczOzsLhcGB9fR1TU1MA/lH1Ph6PY3FxEW/fvr32shJNdhwhEfmuEBA/PlGJvN80TbN8ef1FSiN0YvbcJbfbDU3TUK/X247i12o15PN5jIyMmBoXIPbxdtb599atW5Ydh3a/gesFJl82E4/HIUkSdnd3W06KAwMDUFUViqJgcnIS8Xi8p6NLdhshEfmuEBA/vibRWoOIvN8cDgf8fr+lcy2v2tvRikeiS0tLCAaDePToEdbX11tuIsvlMsLhMGq1mjEFw0wiH28in3/pNCZfNlMqlTA1NdXxIifLMhRFQbFYNDkysYl8VwiIH5+orUFE3m937tyxfHl9t9IImqYZk5+/VCqVOr7Wa4qiYG5uDhsbG5iYmIAsy5Bl2RhVajQamJ+fP/dK7+sk8vEmgnw+f+WfYcX7agUmXzZTqVTgdDq7bjM4OMjk6wui3xWKHF84HDbu+M9qDWI2kfdbNBrFzMwMdnZ2LFte3256APA5uXr06NGpEfSmYrGI2dlZvHz5stchthWNRjE+Po54PI5SqYQPHz4A+Hz8RaNRyy7QIh9vX8rn89A0DdVqFV6v15RjcHZ21nYjrVZhqQmbmZ2dxcePH7vWaRkbG8PIyEjPluyed1VjJ6JMuC+Xy0LW4mkSJb6xsTFIkmSb1iCi7DeRl9c3P8PdCpWGw2F89dVXQnxWdV0XNvGx4nibmZmB1+vtWDE+Eokgm80aPUWbfX7X19fPXTrmMq76mBsQe+7wdeLIl82oqopIJII//elPp06K9XodT548Qa1Wg6qqPYvhvKsaO7Eq+arX63j16pXRIaCp183Hz0vU+KrV6qVqy5lF1P0m8vJ6TdPO7DfpdDp72o6smy8r3J9MvKyscA9Yf7yl02mUSiU8fPiw7euxWAyZTAayLCMUCkGWZezv7yOdTuM//uM/8NNPP7VdyHAd3r5925Of24848mVDsVgM6XQaTqcTXq8Xg4ODKJfL0DQNjUYDfr+/pyem61hGbfZdrK7r+Nd//deWtkwDAwMol8uoVquQJAmKolhW4E/k+GZnZ+F0Oi272HUj8n67yOfE7M9DczSz24V4cnIS1WoVP//8s6mxiVzhXoTjLRgMIp/Pty2Uquu6UYB2d3e35b1NJBJYW1vDw4cPhSwAe+M0yJb29/cb3377bWN4eNj4b3R0tJFKpawOTUizs7ON4eHhxsTERKNYLLa8VigUGn/84x8bLpersba2xvi+8OHDh8bo6GgjnU6b/rvPIvJ+E9nTp08bw8PDjZmZmUa5XG55rVarNYLBYMPlcjXC4bDpsU1MTDSmp6fP3CYYDJoU0T+IcLx12z+JRKLhcrk6flaHh4fP3LdkDo582Zyu68acA1HnRIjA5XLB6XR2vNM/ecdoRS8+keMTee6SyPtNdNPT0yiVSpAkCQ6HA16vt2VFoSzLePPmTc8eUXXicrnOHL0Ph8M4OjoyfVROhOPN5XJhfHy87Wdtenoax8fHHXtLNl8XsQ3YTcM5XzbXXIZN3cmyDI/H0/FCIssy7t69a/rJ/OTvFzU+kecuibzfmi7aCswsu7u7SCQSeP36NSqVirGi1eFwYGFhwZI6Wk2iVrgX4Xhrlt34Uq1WQ6lUwtDQUMcFALVajYWaBcHkiy7MjqsdfT5fy+TYdm7dugWPx2NSRK1Ejk/k1Uci7zdAjFZg3YRCIYRCIQDirCgUucK9CMebz+fDzs4Ojo6OcOfOHePrr169giRJRteTdnRdt+yzQK342FFQIic4V12qLEmS6cPetVoNY2NjHdsuNSfS/vjjj20n+d70+EQl8n5rPoJyOBz4/vvv27YCq9frPW0FZkeapiEYDMLr9XascF8qlbC5uWl6vS8RjrdarYbR0VFIkoSVlRV4vV7s7e0hkUh0fSSaz+cxOzuLaDR6Zpsk6j0mX4ISOcGx42rHubk5vH//HpIkdazoDXy+6/6SJEn43//93xsdn6hE3m/hcBi5XK7j/Jtmcnbv3j0hammJZHV1FRsbG5AkqWOF+6WlJdPjEuV4y2QyiEQixqP+5mV8a2urY0J61nwwMheTL0HZMcER2VWSWTNG6kSPD2itmN0pDrOXsIu83yYnJ+HxeLpOHA8GgyiXyz2bnC3yCPpZNE0zKtw3WV3hXqTjTdd1ZDIZFAoFyLKMx48fd6zFVyqVEA6H4ff78d13311bDHR5TL7oRrhqMtvrRFb0+Jq9HRsnKmaf/HeT2Y+TRd5vY2NjuHfvXscq5MDnmn3ZbLZnE7RFHkG/CFHmo4l8vJG9cMJ9HzC7vYUdm6eKftITOb54PI7Dw0NMTU3h/v37GBgYQDAYRDQahdvtxv7+PnZ2drC+vm56bCLvN4/Hc+ZnJZPJ9HTiuMiLJS5ClPdZlDjI/ph82ZDV7S3YPPVmyWazGB8fP5VcybIMn88Hn8+HQqGAQqGAu3fvWhSleERoBWbnZEHEx9xE14XJl810a2/x66+/Gn29elnscnBw0JKaTtdB1JpLIsen6zr8fn/L12RZbimfoCgKMpkM/vznP5sdHgAx95vf70cgEDB6EXZqBcaE9bTzPua2KvkS8Xgje+GcL5sJBoPQNA2yLGN9fb1lVU2xWEQ4HMbHjx8RCoUsuxCKql3NpaavvvrK8ppLosbXrNh9MqEPh8OQJMmYTN5c2WfF/CBR91tTJpPB06dPWwqHOhwOLC0tIRAIWBaXqOLxOJLJ5Lkec1uRuIp+vJE9MPmyGRHaW9iR6DWXRI6v3aTwVCqFZ8+e4c2bN/j06ZPRfNvsSvIi77d2sZrdCqxWq2FmZgaSJCGbzRpfP+8qSCtWO7ZL9l0uF16+fGkkW9PT0/jnf/5n028w7XS8keBM7CNJ12BiYuLMZrf/+Z//2RgbGzMpIntYXFxsuFyuhq7rbV//8OFDY3h4uBGJREyO7DOR4ysUCo3R0dFGNptt+XqzibDL5WoMDw834vG46bGJvN9EUK1WG8PDww2Xy9Xy9eHh4XP99+X3maHdsTQxMdHSLHp1dbUxOTlpdmg83ujacM6XzYjQ3sKOSqUSpqamOq4KlWUZiqKgWCyaHNlnIsfn8Xjw97///dTXt7a2kEgkUC6Xoapq17YmvSLyfhPBwMAAtra2Tn1d5FWQsiyfWpDjdrtxeHhoLCZqFlw1G483ui5MvmwmGo1ibGwMHz9+7NjeIpvN4scffzQ/OIFVKhU4nc6u2wwODlp20hQ9vnZkWcbKyoqlMYiw30QvvdLuZ4u8CtLn87U8IgU+L+h49uwZfvnlF3z69AnZbPbM970XRDjeqD8w+bKZSCSCgYEBTE9Pd21v8eTJk1Ov3eQ2NCLUXOpG9PhEJcJ+Y+mV66WqKjKZDHK5nDHHS1VVJBIJPHjwwNjOisUKIhxv1B+YfNnM4eGh8e92q22a2t152bU8xHUQoeaSneM76WT9Ja/X2/O6ct2IsN+6lV6pVCqoVqtCjzSVy+Wur5vdB1Dkx9wiHG/UH7ja0WbY3uLyYrEY0uk0nE5nx5pL3frw3YT4ZmZm4PV6O7bEiUQiyGazLTWXhoaGsL6+fuVWNpclwn7rZHV1FZubm0K06PlSs5ZWNxyVO03k443sg8kX3Sii11yyMr50Oo3l5WWsrKy0Hc1qXnRkWUYoFIIsy9jf3zcuRJ3Kn5hB1Pc1Ho9jY2NDuOQrHA4b86rcbjdu3brVcdvNzU2zwrINUY83sg8mX3QjWVFz6SKsiC8YDCKfz7dNFE7Wj9vd3W1JshKJBNbW1vDw4UPL272I9r6KmnyNjY1BkiS8ffsWAwMDVodjW6Idb2QfnPNFN5Isy5afLHO5HN6/f4+hoSHcu3evJaGxIj5d11s6JpyUzWYhSRJCodCp0a1QKIS1tTW8f//ejDC7kmUZkiSZPk/JbqrVKlRVZeJ1RSKcR8iemHzZ0PHxMfb29rrO/7KiMrXdNO9YzabrOubm5qDrujF3anl52fK2JLquY3x8vO1re3t7ADqXRHC73ZbNDbK60bwd+Xy+rgt26PysOo+QvTH5splsNmu0Bun2xJjJ12miXKSDwSB0XcfU1BTGx8fx66+/YmNjA4uLi3jz5o1lE9dlWW6b0NdqNZRKJQwNDXW8yNRqNTgcjl6HeIoIjebtaGVlBTMzM9jZ2WFyekGinEfI3ph82cyrV6/QaDQQCARw7949q8OxDVEu0slkErqu4/vvv285ST98+BCTk5N48uSJZbXYfD4fdnZ2cHR0hDt37hhff/XqFSRJ6rq0X9d1S7oqLC8vo1KpdG00r2kaXrx4cSMbzXfr4eh0OhGLxRCPxzE4ONhx0j0T138Q5TxC9scJ9zbjcrkwPj7OD/cFBYNBaJrW9SL98eNHhEKhnl6kg8EgKpUK/vrXv556rVmW4N27d5asGqzVahgdHYUkSVhZWYHX68Xe3h4SiUTX1Yz5fB6zs7OIRqOYm5szNWYRGs13K7q5vb2NXC6Hra2triPVvapwf9VRVEmShFssYCVRziNkf0y+bGZychKKonSsw0TtiXCRBj6vMlNVFd99992p15pJzNbWVk/bzXSTyWQQiUSMoqHN00O3mKanp3F8fIyDgwPT575MTk7C4/F0rau0uLiIn3/+GT///HNPYnC5XMJWuL+O/oecUP4PopxHyP742NFmAoEAdnZ2rA7DdmRZhsfj6TiiJMsy7t6927MLdFO1WsXQ0FDb15qJy8naQWbz+/04ODhAJpNBoVCALMt4/Phxx1VxpVIJtVoN8/Pzlkw6FqHRfLcK91aza+K0sbEBRVFw+/btU6+l02lommZJIVNRziNkf0y+bCYUCqFQKODu3bsIhULwer0dt2134rqpRLhI20WziOp5uN1uHBwc9DiizkRoNP/27due/eybKJlMYm1tDbu7u21f93g8iMViliwW4HmErguTL5up1+vQdR26riMWi3Xcjm1BWolwkabrx0bz3TXPFxe9EUsmk5AkyfQ5fACQSqXgdrs7xuzxeCDLMjKZjOnJF88jdF2YfNnM4uKicUFRFMXiaOxDpIv0hw8f2k7Sbj5ufP/+fdfil1bNBxMRG813t729jbW1NczPz7edZ9hJo9HA2toaxsfHTS99ous6/H5/123cbjeOjo5MiugfRDqPkL1xwr3NNCd8si3IxVzlAnKdK77OmpzdaDS6vs4RzVZsNN9dczHEZVbQjo2NYWRkxPSV1edZ0R0Oh3F0dGT63CpRziNkfxz5shmHwwG/38/E64KsnJd0Uqf2PXQ5/Z48XVW5XIbb7b5U6ZLBwcFrWS15UW63G5qmoV6vt427Vqshn89jZGTE9NhEOY+Q/TH5spk7d+6wLcgliHKR7jSJmKgXqtXqpY99WZYtKZewtLSEYDCIR48eYX19vWVuVblcRjgcNlbYmk2U8wjZH5Mvm4lGo2wLQkR9S1EUzM3NYWNjAxMTE0bz6uZCo0ajgfn5ec59JFvjnC+biUQiKBaLKJfLcDgcbAtCRB1NT0+jXq9fagRrbGwMsiy37cZgBk3TEI/HW+Y4ut1uRKNRJl5kexz5splMJmP8u1KpdHwEeRNWchFRd/fu3cOLFy+wubmJYDB47u9LJpOo1Wq4f/9+D6PrTlEU4zG9rut85Ed9hSNfNnORCbA8WRHdbLVaDf/yL/+Cer2Oly9fYnJy8szv2dnZwdOnT7u20SGiq2HyRUTUxzRNQzAYhCRJUBQF0Wi0bcmE4+NjxONxaJqGRqPR8x6j5XIZAE7Vy6rX6+f+GUwMya6YfNlUuVxGKpWCrut4/PixUQ06m82iUChcqKAiEfU3TdOMVYKSJMHhcMDr9WJwcBDlchmFQgHVahWNRgMOhwPr6+s9n1flcrnw1Vdfnapbd95G5ax5R3bGOV82FIlEkM1mjYKcqqoarzkcDiSTSfz+978/1yMGIup/iqLgb3/7G1ZXV5FOp1GpVFq6AzQFAgFEo1FT6gh2akjOWnh0E3Dky2aSySTi8ThUVcX8/DwmJydPPR6YnJzE0NAQVzvSjcGR4IvJ5/PQdR3VahUOhwOyLHMF4TkcHx9fuE8mUTsc+bKZvb09eDwePH/+3OgF+CWfz4dsNmtyZPbAi3T/4Ujwxfl8PsuTrVwuB1mWTyUzuVwObre7be9Eqz148AC/+93voKoqAoEA55zRpX1ldQB0MaVSCV6v1+owbCkSiWBychKJRALZbBa//fab8VrzIs32IfaSTCaRyWQQCARwcHCALwfyfT4fBgcHsb29bVGE1Ek4HIamaW2/LurNo8/nw6+//or/+q//wujoKObm5rCzs2N1WGRDTL5sxu12I5/Pd90mk8kwQfsCL9L96eRIcKdiwz6fD4VCweTI6CwDAwNt6xSKPBNma2sL7969w/Pnz3H79m0cHh4iFovh9u3b+NOf/oSjoyOrQySb4GNHm1FVFcvLy3jx4gUWFhZaXqvX68aKppOPXoiPa/tVqVTisW5ToVAIa2tr0DQNHo+n5bW9vT28f/++6/dLkoS//OUvvQyxrYGBAaiqClVVUavVsL29jf39eHri4QAAEeJJREFUfezv7yOTycDhcBiPJUV8dEpiYPJlM6qq4vDwEK9fv0YqlYIkSUgmk0gmk0Z9HlVVcffuXatDFQov0v3pvCPBIyMjJkVE5xUKhQAAiUTi1MhksVhEsVjs+v1WJV8nDQwMIBQKIRQKoVar4dWrV0gmk3j9+jUSiQSGhoYwPz/PPrx0CpMvG3r58iVSqRTi8TgajYaxZFyWZSwtLWFqasriCMXDi3R/4kiwvZ1MXC7SvUMU9Xod+/v72N7eNmqOybIMVVXx6dMno1tAMpnE5uYmvvnmG4sjJlGw1ITNNU9asiybUpvHrlKpFJaXl7GwsICFhQV8++23RomO5kVa0zSsr69z1NBmFhcXkcvl4HQ6Ua1WoSgKALSMBD9//tziKOlLGxsbcLvdlq+6vIxcLodUKmUsGGg0GpiamsLDhw9P/T2ZTAaxWAxff/01pzWQgckX3Ri8SPev5kjwyfl8HAkWm8vlQjQaxdzcXMvX//CHP2BpaUnIR3Vra2tIp9NGN4DmKJeqql1vfmOxGHZ2dnB8fGxitCQyPna0sWZvtE442bMVH9f2r5MToDkSbA8Oh6Pt1yuVCqrVqsnRnE8ikQCAjqNcnZTLZZ6PqQVHvmxobm6ubX2ck9j3rDtepImstbi4iF9++QUrKysticnExARCodC55umZndAkk8kzR7naqdVqPM9QCyZfNnOyAKHb7e5Y2wgANjc3zQqLiOhCarUaHjx4gI8fP17q+3mDSXbGx442k8/n4XQ68fbtW95J0Y0RiUSu9P0ilCWgVgMDA3j79i1SqRSKxaJRcDWbzUKWZTbYpr7GkS+bcblcnBjeBS/S/cnlcl3p+yVJ4mRnm+g0Ed9MPI9QrzH5spnZ2Vk4nU7893//t9WhCIkX6f50HTWgZFm+hkio19LpNBRFsXSCOs8j1GtMvmxG13XMzMwgGo0KuRTbarxIE9FV8TxCvcbky2YikQiKxSLK5TIcDgcGBwc7Trrf2NgwOToiouuRz+ehaVrHshOSJOHZs2fmBkV0TZh82cx5h8M57E03CS/U/aVZTqd5eZIkqeXfTWaf45aXl+H1evnUga6MyZfNXGQ4nMPep/Ei3X9EvVDT5cTjcSSTSUxNTeH+/fsYGBhAMBhENBqF2+3G/v4+dnZ2LGkFNjk5CUmSkMvlTP291H9YasJmmFBd3nkv0ky+7CMej+Pw8PBcF2qyh2w2i/Hx8VPvmSzL8Pl88Pl8KBQKKBQKpidfgUAAL168wNHREe7cuWPq76b+8pXVARCZoXmRvnv3LtbX17G5uYlGo4FoNIqtrS3jMQIv0vZy8kI9NTVl9OtsXqhXVlZw+/ZtFAoFiyOl89J1/VSNL1mWjTpgAKAoCjKZjNmhIRQK4fbt2wiHwzg4ODD991P/4MgX3Qgi303T5em6Dr/f3/K1ThfqP//5z2aHR5cgy/KpyvVutxuHh4fGTZKu69eyIvGiIpEIhoaGUCqVsLi4CIfDAa/X23ZbLniibph8CWpmZubc2966dQuKokBRFNy+fbuHUdkXL9L9SeQLNV2Oz+czWqg1KYqCZ8+e4ZdffsGnT5+QzWbhdDpNj+3kaFuj0UClUsHh4eGp7U5OYyBqh8mXoIrF4oW2Pzw8hCRJUFWVc5ba4EW6P4l8oabLUVUVmUwGuVzOGIVWVRWJRAIPHjwwtgsEAqbHxkeNdF2YfAnq3bt35972t99+Q7FYRCqVwvb2NnRd55D3F3iR7k8iX6jpcjweD/7+97+f+vrW1hYSiQTK5TJUVcXU1JTpsXHBE10XlproM6lUCs+ePcPLly8xOTlpdTjCKBaLmJ2dxQ8//NAyp2tiYgIfP34E8PkxQigUwnfffWdVmHRNdF23/EJN/atcLiOVSkHXdTx+/NiY7pHNZlEoFHgOoTMx+epD09PT+Prrrzn6dQ68SBOJRfRCppFIBNlsFo1GA5IkYXNzEz6fD8DnOoLBYJA3v3QmlproQ/fv3+fS+nOSZRkrKyvY3Nxk4kUkAE3TkEgkrA6jrWQyiUwmg0AggIODA3w5duHz+TA4OIjt7W2LIiS7YPJFRETCCAQC0HUdR0dHVodyyt7eHjweD54/f96xp26zbA1RN5xw34f29vY61p7pd/V6/co/45/+6Z+uIRIyy/HxMfb29rquVJUkCX/5y19MjIouKxQKYX9/H+FwGD/88INQj+9KpRJUVbU6DOoDTL76TDwex/Hx8Y2t1P7tt99eqcaOJEmnSlKQuLLZLCKRCACcegR0EpMv+xC5kKnb7UY+n++6TSaTwcjIiEkRkV0x+RLUWR/wk2q1GnRdN1bfTE1N3fgq7QMDAywbcQO8evUKjUYDgUAA9+7dszocugYiFzJVVRXLy8t48eIFFhYWWl6r1+sIh8Oo1WocHaMzcbWjoFwu14VOLs23MRAIYGVlpVdhCW90dBS1Wg2SJBnNlhVF4aPEPuVyuTA+Ps6VvTZUr9fbfi4vUujYirpbi4uLyOVycDqdqFarRj9RTdPQaDSgqiqeP39uelxkL0y+BDUxMXHu5EuWZbjdboyPjxtLnm+yfD6P//mf/0EulzP2oaIo8Pv9uHfvHhOxPjI5OQlFUXixs6Hbt2/btiNHKpVCPB5HrVYzvibLMpaWlrhqms6FyRf1tUwmg1QqZTzGlSQJiqLg4cOHQk3kpctJJBLY2dlBLpezOhS6IJfLZftRouaUD1mWMTAwYHU4ZCNMvuhGqNVq2NvbQyqVQqlUMkbE/H4/VFXFnTt3LI6QLiscDuP4+BihUKjrKl82nRdLPyRfRJfF5ItunFqthu3tbezv7xsrG51OJ+7duwdVVXmRtpF6vY5///d/x/HxcdftuIpVPCInX+Vy+co/Y3Bw8BoioX7F1Y504wwMDCAUCiEUCqFWq0HTNKMpeTqd5kXaRhYXF433qznxmeiqLjLnth0m+3QWJl90Y5XLZWSzWWiaZswJ42R8e9E0DU6nE2/fvuWcG7o2Pp+vY/KlaRocDkfbka1SqdTxNaKT+NiRbpR6vY5UKoW9vT2USiWjRMfU1BQePnzI1aI2MzY2hnv37gn56Iq6u2g5nS9ZMbpUKpXw6NEj7O7utk2wisUiZmdnO75O1MSRL+p79Xod+/v72N7ebkm4FEWBqqpcGm5jd+7cQaVSsToMuqSr3PtbMW7w+vVrKIrSMbHyeDzw+XxYW1tjRwXqiskX9a1cLodUKmUUPwRa633xMZX9RaNRzMzMYGdnB//2b/9mdTh0QX6/H99//73VYZybpmlndlJwOp3IZrMmRUR2xeSL+koul8Pe3p5x8ms0Gky4+tja2hqcTidisRji8TgGBwdx69atttuyCr54nE6n7T6TmUwG0Wi04/zQi7SGo5uLyRf1lcXFRUiSBFmWjfZCzZN7pVI51yMqztWwj5N9ALu9v1b0AaT+4/f7kU6n8ejRI6yvr+Obb74xXmv2diyXy5zKQGfihHvqK3acxEuXJ3ofQOpM5Dpf3UxPTxuFmh0OB7xeL3Rdh67raDQakGUZb9684cpp6oojX9RX3G631SGQiZhQkdl2d3eRSCTw+vVrVCoVHB4eAgAcDgcWFhYwPz9vcYRkBxz5IiLbK5fLSKVS0HUdjx8/NroUZLNZFAoFfPfddxZHSF+y68jXl5q9HYkugiNfRGRrkUgE2WwWjUYDkiRBVVXjNYfDgWQyid///vdspC6Yg4ODjosj7ISJF13GV1YHQER0WclkEplMBoFAAAcHB6dqP/l8PgwODmJ7e9uiCKkTWZZtt9KR6Lpw5IuIbGtvbw8ejwfPnz9HrVZru43P52PdJbqQSCRype+XJIlFVqkrJl9EZFulUqnlMSPRdThZwuQymHzRWZh8EZFtud3uM4taZjIZjIyMmBQR9YODgwOrQ6A+x+SLiGxLVVUsLy/jxYsXWFhYaHmtWfSyVqtxdIwuhJPoqddYaoKIbG1xcRG5XA5OpxPVahWKogCA0dOzH8oZEFF/YfJFRLaXSqUQj8dbJt3LsoylpSW2eqELi0QiuH//Pu7evXvh7w2Hwzg4OGCnDOqKjx2JyPZUVYWqqqjVakbRS5YxoMvKZDJwOp1tk694PI7Nzc2OyVWtVjtV8oToS6zzRUTCy+fzKJfLZ243MDAAt9vNxIt6iskVXRWTLyIS3uzsLJLJZNvXyuUycrmcyREREV0eky8isoVOow3b29sIh8MmR0NEdHlMvoiIiIhMxOSLiIiIyERMvoiIiIhMxFITREREXyiXy21bV+m6DgA4OjpqOw/xt99+63lsZH9MvoiIiL6gaRo0Tev4+uzsbNuvNxoNSJLUq7CoTzD5IiJbuOxIRJPP5+tZbNRfBgcHmUBRT7G9EBEJz+VyXeliKEkS270QkTA48kVEwuNIBBH1E458EREREZmIpSaIiIiITMTki4iIiMhETL6IiIiITMTki4iIiMhETL6IiIiITMTki4iIiMhETL6IiIiITMTki4iIiMhETL6IiIiITMTki4iIiMhETL6IiIiITMTki4iIiMhETL6IiIiITMTki4iIiMhETL6IiIiITMTki4iIiMhETL6IiIiITPT/rA6AiOgyUqkUYrHYubf3eDzY3d3tYUREROfD5IuIbM/hcJy5zeDgoAmREBGdjckXEdne7u4uZFm2OgwionPhnC8ioh6oVqtYXV1FIpGwOhQiEgxHvoiIeqBSqSCZTAIAQqGQxdEQkUg48kVERERkIiZfRERERCZi8kVEN1YikcDs7CyGh4cxOjqKWCyGarXadltN07C4uIiJiQlj+8XFRei63rLd6uoqhoeHMTExYXxteHjY+G90dLRl++npaQwPD7edG1atVo3vKxaLp15vfm8mkwEAZDIZzM7OYnR09FRcF/17iah3OOeLiG6carWKR48eGQmNw+FAtVpFKpXC/v7+qdWTiUQC8Xjc2Nbj8UDXdWSzWWSzWWxtbUFRFADAyMgIFEVBpVIxfn7zNQA9WZVZqVQwOzsLTdOu5e8lot5i8kVEtndylKkdVVWxsrJi/H8zEfF4PFhfXzcSj1gshlQqhXA43FKQ1eFwQFVVhEKhliRldXUVyWQS4XAYP/30ExwOB/x+P/x+P4rFIqanpwEAW1tb1/nnnhKPx1GtVuHxeLC0tASv19tS++yify8R9RaTLyKyPVmWuxZaPZkwpVIpFItFOBwO/Pjjjy3ft7KyAk3TUCwWjWQF+Jy8tRONRpFMJlGtVlEoFFpGuMxUrVYxNTWFly9fnnrtMn8vEfUWky8isr2tra1zPzZLpVIAgIWFhbYJm6qqiMfj0DTtXMmIx+Mxkherki+Hw4Effvih7WvX/fcS0dUx+SKiG6U5ET0ejxvzuLpt183Jyeq//fbb1YO7JJ/P13Hk7zr/XiK6Hky+iOjGqFarRsI0NTXVddsvR4F0XUcmk4GmaSgUCkKtEhwZGWn79av8vUTUO0y+iOjGcDgcxkq/x48fnzvhSKVSiMViAD4nKYFAALdu3YLD4UAikRB21Oiyfy8R9RaTLyK6UbxerzF6dZ5kRNM0I/Ha3d099T3NOVW9UKlUrvwzLvr3ElHvscgqEd0ozZWLzfIMnTRfO1mrq13y0ulnnJyDddlHlO0Kq17URf9eIuo9Jl9EdKP4/X4oioJqtYrp6elTCU4mk8HExAT29/cB/KNMhaZpp4qYdnvk6HQ6jX83f1a1WjVG0b7c7sufXSwW8fTp04v+eadc9O8lot7jY0ciunHW19fx5MkTZLNZTE9Pw+FwwOl0tiRSzaTI7/dDlmXouo7Z2dm227bjcDiM74vFYi0jT6qqGqNoiqIYid3ExITxPbquY2pqCtls1tS/l4h6jyNfRHTjOBwOvHz5EisrK/B4PKhWq9B1HR6PB/Pz83j37h38fr+x/e7u7qkaXqqq4u3bt11re62vrxtJltPpxNTU1Kl5Y6FQCKqqwuFwGEmXoijY3d1tWzTVjL+XiHpLajQaDauDICIiIropOPJFREREZCImX0REREQmYvJFREREZCImX0REREQmYvJFREREZCImX0REREQmYvJFREREZCImX0REREQmYvJFREREZCImX0REREQmYvJFREREZCImX0REREQm+j8D/n+7bTgbvgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 655.2x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a figure instance\n",
    "fig = plt.figure(1, figsize=figsize)\n",
    "ax = fig.add_subplot(111)\n",
    "\n",
    "bp = ax.boxplot(np.array(importance_df), patch_artist=True,showfliers=True)\n",
    "\n",
    "for box in bp['boxes']:\n",
    "    # change outline color\n",
    "    box.set( color='black', linewidth=1)\n",
    "    # change fill color\n",
    "    box.set( facecolor='#bdbdbd',zorder=3)\n",
    "\n",
    "## change color and linewidth of the whiskers\n",
    "for whisker in bp['whiskers']:\n",
    "    whisker.set(linestyle='-',color='black', linewidth=1,zorder=4)\n",
    "\n",
    "## change color and linewidth of the caps\n",
    "for cap in bp['caps']:\n",
    "    cap.set(color='black', linewidth=1,zorder=4)\n",
    "\n",
    "## change color and linewidth of the medians\n",
    "for median in bp['medians']:\n",
    "    median.set(color='k', linewidth=1,zorder=4)\n",
    "\n",
    "## change the style of fliers and their fill\n",
    "for flier in bp['fliers']:\n",
    "    flier.set(marker='o', markerfacecolor='k',\n",
    "              markeredgecolor='k',markersize=4,zorder=3)\n",
    "ax.set_xlabel('Feature')\n",
    "ax.set_ylabel('Feature importance')\n",
    "ax.set_xticklabels(list(importance_df.columns),rotation=90,ha='center')\n",
    "#box_plot_style(ax,bp,ymin=0,ymax=0.46,ylabel='Feature imp,ortance',xticklabels=list(importance_df.columns),fontsize=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-07-04T17:07:57.966295Z",
     "start_time": "2019-07-04T17:07:57.259200Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 786.24x604.8 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p=test2samples(np.array(importance_df))\n",
    "p=pd.DataFrame(p,index=list(importance_df.columns),columns=list(importance_df.columns))\n",
    "\n",
    "fig, ax = plt.subplots(figsize=stdfigsize(scale=1.2))\n",
    "res=sns.heatmap(p,vmin=0,vmax=1.,linewidths=1,annot=False,cmap=\"Reds\",square=True)\n",
    "#sns.clustermap(p,vmin=0,vmax=1.,annot=False,cmap=\"Reds\")\n",
    "\n",
    "cbar = res.collections[0].colorbar\n",
    "cbar.set_ticks([0,0.25, 0.5,0.75, 1])\n",
    "cbar.set_ticklabels([\"0.00\",\"0.25\", \"0.50\",\"0.75\", \"1.00\"])\n",
    "#ax.tick_params(labelsize=22)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "168px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}