notebook/boston_randomforest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RANDOM FOREST USING BOSTON DATASET\n",
    "An excellent explanation is provided in this youtube video https://www.youtube.com/watch?v=loNcrMjYh64\n",
    "  * The key method is that it using 'bagging', i.e., the sample selection of samples to construct many decision trees.\n",
    "  * Then prediction is based on the highest 'votes' that from all the decision trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as py\n",
    "from sklearn.datasets import load_boston\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.cross_validation import train_test_split\n",
    "import sklearn.metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Boston House Prices dataset\n",
      "\n",
      "Notes\n",
      "------\n",
      "Data Set Characteristics:  \n",
      "\n",
      "    :Number of Instances: 506 \n",
      "\n",
      "    :Number of Attributes: 13 numeric/categorical predictive\n",
      "    \n",
      "    :Median Value (attribute 14) is usually the target\n",
      "\n",
      "    :Attribute Information (in order):\n",
      "        - CRIM     per capita crime rate by town\n",
      "        - ZN       proportion of residential land zoned for lots over 25,000 sq.ft.\n",
      "        - INDUS    proportion of non-retail business acres per town\n",
      "        - CHAS     Charles River dummy variable (= 1 if tract bounds river; 0 otherwise)\n",
      "        - NOX      nitric oxides concentration (parts per 10 million)\n",
      "        - RM       average number of rooms per dwelling\n",
      "        - AGE      proportion of owner-occupied units built prior to 1940\n",
      "        - DIS      weighted distances to five Boston employment centres\n",
      "        - RAD      index of accessibility to radial highways\n",
      "        - TAX      full-value property-tax rate per $10,000\n",
      "        - PTRATIO  pupil-teacher ratio by town\n",
      "        - B        1000(Bk - 0.63)^2 where Bk is the proportion of blacks by town\n",
      "        - LSTAT    % lower status of the population\n",
      "        - MEDV     Median value of owner-occupied homes in $1000's\n",
      "\n",
      "    :Missing Attribute Values: None\n",
      "\n",
      "    :Creator: Harrison, D. and Rubinfeld, D.L.\n",
      "\n",
      "This is a copy of UCI ML housing dataset.\n",
      "http://archive.ics.uci.edu/ml/datasets/Housing\n",
      "\n",
      "\n",
      "This dataset was taken from the StatLib library which is maintained at Carnegie Mellon University.\n",
      "\n",
      "The Boston house-price data of Harrison, D. and Rubinfeld, D.L. 'Hedonic\n",
      "prices and the demand for clean air', J. Environ. Economics & Management,\n",
      "vol.5, 81-102, 1978.   Used in Belsley, Kuh & Welsch, 'Regression diagnostics\n",
      "...', Wiley, 1980.   N.B. Various transformations are used in the table on\n",
      "pages 244-261 of the latter.\n",
      "\n",
      "The Boston house-price data has been used in many machine learning papers that address regression\n",
      "problems.   \n",
      "     \n",
      "**References**\n",
      "\n",
      "   - Belsley, Kuh & Welsch, 'Regression diagnostics: Identifying Influential Data and Sources of Collinearity', Wiley, 1980. 244-261.\n",
      "   - Quinlan,R. (1993). Combining Instance-Based and Model-Based Learning. In Proceedings on the Tenth International Conference of Machine Learning, 236-243, University of Massachusetts, Amherst. Morgan Kaufmann.\n",
      "   - many more! (see http://archive.ics.uci.edu/ml/datasets/Housing)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# lets learn about more about the dataset\n",
    "print boston['DESCR']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00632</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.538</td>\n",
       "      <td>6.575</td>\n",
       "      <td>65.2</td>\n",
       "      <td>4.0900</td>\n",
       "      <td>1.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>15.3</td>\n",
       "      <td>396.90</td>\n",
       "      <td>4.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.02731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>6.421</td>\n",
       "      <td>78.9</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>396.90</td>\n",
       "      <td>9.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.02729</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>7.185</td>\n",
       "      <td>61.1</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>392.83</td>\n",
       "      <td>4.03</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      CRIM    ZN  INDUS  CHAS    NOX     RM   AGE     DIS  RAD    TAX  \\\n",
       "0  0.00632  18.0   2.31   0.0  0.538  6.575  65.2  4.0900  1.0  296.0   \n",
       "1  0.02731   0.0   7.07   0.0  0.469  6.421  78.9  4.9671  2.0  242.0   \n",
       "2  0.02729   0.0   7.07   0.0  0.469  7.185  61.1  4.9671  2.0  242.0   \n",
       "\n",
       "   PTRATIO       B  LSTAT  \n",
       "0     15.3  396.90   4.98  \n",
       "1     17.8  396.90   9.14  \n",
       "2     17.8  392.83   4.03  "
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "boston = load_boston()\n",
    "# the boston dataset consists of two arrays: features (.data) & target (.target)\n",
    "# lets add the feature into a dataframe\n",
    "df=pd.DataFrame(boston.data, columns=boston.feature_names)\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    506.000000\n",
       "mean      22.532806\n",
       "std        9.197104\n",
       "min        5.000000\n",
       "25%       17.025000\n",
       "50%       21.200000\n",
       "75%       25.000000\n",
       "max       50.000000\n",
       "Name: PRICE, dtype: float64"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# now add the target to the feature dataframe\n",
    "df['PRICE']=pd.DataFrame(boston.target)\n",
    "df['PRICE'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "      <th>PRICE</th>\n",
       "      <th>PRICE_CAT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00632</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.538</td>\n",
       "      <td>6.575</td>\n",
       "      <td>65.2</td>\n",
       "      <td>4.0900</td>\n",
       "      <td>1.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>15.3</td>\n",
       "      <td>396.90</td>\n",
       "      <td>4.98</td>\n",
       "      <td>24.0</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.02731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>6.421</td>\n",
       "      <td>78.9</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>396.90</td>\n",
       "      <td>9.14</td>\n",
       "      <td>21.6</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.02729</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>7.185</td>\n",
       "      <td>61.1</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>392.83</td>\n",
       "      <td>4.03</td>\n",
       "      <td>34.7</td>\n",
       "      <td>expensive</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      CRIM    ZN  INDUS  CHAS    NOX     RM   AGE     DIS  RAD    TAX  \\\n",
       "0  0.00632  18.0   2.31   0.0  0.538  6.575  65.2  4.0900  1.0  296.0   \n",
       "1  0.02731   0.0   7.07   0.0  0.469  6.421  78.9  4.9671  2.0  242.0   \n",
       "2  0.02729   0.0   7.07   0.0  0.469  7.185  61.1  4.9671  2.0  242.0   \n",
       "\n",
       "   PTRATIO       B  LSTAT  PRICE  PRICE_CAT  \n",
       "0     15.3  396.90   4.98   24.0     medium  \n",
       "1     17.8  396.90   9.14   21.6     medium  \n",
       "2     17.8  392.83   4.03   34.7  expensive  "
      ]
     },
     "execution_count": 164,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# rename housing price into 3 categories: cheap, medium, expensive\n",
    "# this is based the 25, 50, 75 percentiles\n",
    "def price(x):\n",
    "    if x < 17:\n",
    "        return 'cheap'\n",
    "    elif x < 25:\n",
    "        return 'medium'\n",
    "    else:\n",
    "        return 'expensive'\n",
    "\n",
    "df['PRICE_CAT'] = df['PRICE'].apply(price)\n",
    "df.head(3)\n",
    "\n",
    "# Or alternatively, it is much easier to use a quartile cut\n",
    "# df['priceCAT2'] = pd.qcut(df['PRICE'], 3, labels=['cheap','medium','expensive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CRIM</th>\n",
       "      <th>ZN</th>\n",
       "      <th>INDUS</th>\n",
       "      <th>CHAS</th>\n",
       "      <th>NOX</th>\n",
       "      <th>RM</th>\n",
       "      <th>AGE</th>\n",
       "      <th>DIS</th>\n",
       "      <th>RAD</th>\n",
       "      <th>TAX</th>\n",
       "      <th>PTRATIO</th>\n",
       "      <th>B</th>\n",
       "      <th>LSTAT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00632</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.538</td>\n",
       "      <td>6.575</td>\n",
       "      <td>65.2</td>\n",
       "      <td>4.0900</td>\n",
       "      <td>1.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>15.3</td>\n",
       "      <td>396.9</td>\n",
       "      <td>4.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.02731</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.469</td>\n",
       "      <td>6.421</td>\n",
       "      <td>78.9</td>\n",
       "      <td>4.9671</td>\n",
       "      <td>2.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>17.8</td>\n",
       "      <td>396.9</td>\n",
       "      <td>9.14</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      CRIM    ZN  INDUS  CHAS    NOX     RM   AGE     DIS  RAD    TAX  \\\n",
       "0  0.00632  18.0   2.31   0.0  0.538  6.575  65.2  4.0900  1.0  296.0   \n",
       "1  0.02731   0.0   7.07   0.0  0.469  6.421  78.9  4.9671  2.0  242.0   \n",
       "\n",
       "   PTRATIO      B  LSTAT  \n",
       "0     15.3  396.9   4.98  \n",
       "1     17.8  396.9   9.14  "
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# time to split the dataframes into training & test\n",
    "feature = df[df.columns[:-2]]\n",
    "target = df['PRICE_CAT']\n",
    "feature.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(404, 13)\n",
      "(102, 13)\n"
     ]
    }
   ],
   "source": [
    "# now to randomly assign training and testing dataframes\n",
    "# 80% will be for training, 20% will be for testing\n",
    "train_feature, test_feature, train_target, test_target = train_test_split(feature, target, test_size=.2)\n",
    "\n",
    "print train_feature.shape\n",
    "print test_feature.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### PREPARE THE CLASSIFIER"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# use 100 decision trees\n",
    "clf = RandomForestClassifier(n_estimators=100)\n",
    "clf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### FIT THE MODEL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = clf.fit(train_feature, train_target)\n",
    "model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### TEST THE MODEL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "predictions = model.predict(test_feature)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SCORE THE MODEL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy \n",
      "82.3529411765 % \n",
      "\n",
      "confusion matrix \n",
      "[[21  0  3]\n",
      " [ 0 21  4]\n",
      " [ 8  3 42]]\n"
     ]
    }
   ],
   "source": [
    "# check how many % the model is accurate\n",
    "print 'accuracy', '\\n', sklearn.metrics.accuracy_score(test_target, predictions)*100, '%', '\\n'\n",
    "print 'confusion matrix', '\\n', sklearn.metrics.confusion_matrix(test_target,predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>RM</th>\n",
       "      <td>0.225612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LSTAT</th>\n",
       "      <td>0.192478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CRIM</th>\n",
       "      <td>0.108510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DIS</th>\n",
       "      <td>0.088056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AGE</th>\n",
       "      <td>0.074202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NOX</th>\n",
       "      <td>0.067718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <td>0.057706</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PTRATIO</th>\n",
       "      <td>0.051702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TAX</th>\n",
       "      <td>0.047568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>INDUS</th>\n",
       "      <td>0.037871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RAD</th>\n",
       "      <td>0.026538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZN</th>\n",
       "      <td>0.012635</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CHAS</th>\n",
       "      <td>0.009405</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                0\n",
       "RM       0.225612\n",
       "LSTAT    0.192478\n",
       "CRIM     0.108510\n",
       "DIS      0.088056\n",
       "AGE      0.074202\n",
       "NOX      0.067718\n",
       "B        0.057706\n",
       "PTRATIO  0.051702\n",
       "TAX      0.047568\n",
       "INDUS    0.037871\n",
       "RAD      0.026538\n",
       "ZN       0.012635\n",
       "CHAS     0.009405"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# rank the importance of features\n",
    "df2= pd.DataFrame(model.feature_importances_, index=df.columns[:-2])\n",
    "df2.sort_values(by=0,ascending=False)\n",
    "\n",
    "# RM: average number of rooms, status of people living in the area, and crime rate are the 3 top factors \n",
    "# determining the housing price"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x117f3f0d0>]"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAECCAYAAAD3vwBsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJztvXmUZFd54PmLPXKJiMzKzNqzqqSS6kqAkJBkFlm2DDZ2\nY5bGY2amaZi21RZj8NJu7LGN2ofuc9rbnMbGDHAAAQbjGduA1Y3HxoPwgmyQAGEkJFQC3axFqr0q\nl6qMyC32N3+8d1+8eLG9yIzMDGV8v3PqVMZb4t1347373W+9IcuyEARBEAaP8HY3QBAEQdgeRAAI\ngiAMKCIABEEQBhQRAIIgCAOKCABBEIQBRQSAIAjCgBLtdIBSKgR8BLgVyAP3aa1Pe/a/DfhVoAx8\nWmv9MaVUFPgMcMTZ/g6t9Uzvmy8IgiCslyAawJuBhNb6LuB+4P2+/e8DXgPcDfyaUioD/CQQ0Vr/\nIPDbwO/1rsmCIAhCLwgiAO4GHgLQWj8G3Onb/xQwDgw5ny1gBog62kMGKPaktYIgCELPCCIA0kDW\n87mslPKe9wzwOPA08EWtdQ5YBq4DngUeAD7Ym+YKgiAIvSKIAMgBKe85WusqgFLqFuD1wGFse/8e\npdRbgHcDD2mtFbbv4E+VUvFeNlwQBEHYGB2dwMCjwBuAB5VSr8Se6RuywCpQ0FpbSqlZYAy4CpSc\nYxad60TaXcSyLCsUCnXZfEEQhIFn3QNnqFMxOE8U0EudTfcCdwAjWutPKqV+Hvj3QAE4BbwDSACf\nAvYBMeADWuvPdWiLNTe3tN772FFMTaWQvrCRvqghfVFD+qLG1FRq8wTAFiICwEEe7hrSFzWkL2pI\nX9TYiACQRDBBEIQBRQSAIAjCgCICQBAEYUARASAIgjCgiAAQBEEYUEQACIIgDCgiAARBEAYUEQCC\nIAgDiggAQRCEAUUEgCAIwoAiAkAQBGFAEQEgCIIwoIgAEARBGFBEAAiCIAwoIgAEQRAGFBEAgiAI\nA4oIAEEQhAFFBIAgCMKAIgJAEARhQBEB0EPyxTKr+dJ2N0MQBCEQ0U4HKKVCwEeAW4E8cJ/W+rRn\n/9uAXwXKwKe11h9ztr8HeBMQAz6itf5075vfX3zwwe+SWy3xO/e9YrubIgiC0JEgGsCbgYTW+i7g\nfuD9vv3vA14D3A38mlIqo5S6B3iVc86PANO9a3L/MreYZ35xbbubIQiCEIggAuBu4CEArfVjwJ2+\n/U8B48CQ89kCfgI4rpT6K+CvgS/2pLV9TqFUoViuUq1a290UQRCEjgQRAGkg6/lcVkp5z3sGeBx4\nGvii1joHTAJ3AG8B3gX8eW+a29/ki2Xn/8o2t0QQBKEzQQRADkh5z9FaVwGUUrcArwcOA0eAPUqp\ntwALwJe11mWt9QyQV0pN9rTlfUa5UqVcsWf+hZIIAEEQ+p+OTmDgUeANwINKqVdiz/QNWWAVKGit\nLaXULDAGPAL8B+CPlFL7gWFsodCWqalUp0P6lqXVovv30Ehiw/fyQu6LXiN9UUP6oob0xcYJIgC+\nALxWKfWo8/lepdRbgRGt9SeVUh8HHlFKFYBTwJ9orctKqR9SSn0LCAG/oLXuaBifm1ta521sPwvZ\nvPv3pSs5khsIsJ2aSr2g+6KXSF/UkL6oIX1RYyOCsKMAcAbud/k2z3j2PwA80OS896y7VS9AjP3f\n/ltMQIIg9D+SCNYj8h67f0EEgCAILwBEAPQI76zfqw0IgiD0KyIAekShTgCIBiAIQv8jAqBHiA9A\nEIQXGiIAeoSYgARBeKEhAqBHeE1A3SSCrRXKPPydC1Sq1c1oliAIQktEAPSItXX6AL5+/DL/95c1\nT56Y34xmCYIgtEQEQI+o0wC6EAAmg3hWqogKgrDFiADoEet1Ahtz0bwnk1gQBGErEAHQI7x2/26c\nwEZYLIgAEARhixEB0CPMQB6iSw3ACICcCABBELYWEQA9wgz6o8OxrqKAzHnz2TyWJQvJCIKwdYgA\n6BH5YplEPMJQPNqVBmDMRYVihZW85A8IgrB1iADoEYVihWQsQiIe6VIA1I4VP4AgCFuJCIAekS9W\nSMYjJOMR8sVyYHOO11wkkUCCIGwlIgB6RL5UIRG3NQDLgmI5WGZvnQYgjmBBELYQEQA9oGpZtgko\nHiUZt9fYCZoMli+WCYXsv8UEJAjCViICoAcUHTNOMh4hGYsAwXIBLMsiX6ywe3wYgPmsZAMLgrB1\niADoAcaMY3wA3m3tKJarWBZMZZLEY2ExAQmCsKWIAOiSmXOL/NHnn2KtUJvhG3NPwokCgmACoOAR\nHBPp5I41Aemz1/jDzz3J6haHuV5aWOG//fkTzHeoszS/uMZ/+/MnuLSwsu5rPXlyng//j6ddbdDL\n579ykr/9xvMN21fzZX7v/3mc+x/4hvvv81852fFaX/rmGf7y4c7HDTLFUoX/6y+f4qmTjUUWn7uU\n431/8R1yTh2uXmJZFn/8xe/xz09e6Pl3bwYdBYBSKqSU+qhS6utKqa8opa737X+bUupxpdRjSql3\n+vbtVkqdVUod63XDt4tvPnOZp08vcPpSzt1W0wCirgYQJBnMmIkS8QgTmSQr+XKdYNkp/OPj53nm\nuaucOL+4pdd95LuXePbsIk82GQS8fP34ZZ49u8gTM3Prvta3vn+FJ2bmePZs/T2u5Et8+Vtn+afv\nNA4IZy7nOHk+y+JKkXypwnw2z8NPXugYQfYPj5/n7799XhIH2/D85SWeOrXAP3z7XMO+h5+4wPfP\nXOP7z1/r+XUXl4s8evwyX/rm2Z5/92YQRAN4M5DQWt8F3A+837f/fcBrgLuBX1NKZQCUUlHgY8Bq\n75q7/cw7Zpolz+zBO5AbJ3AQDcArOCYzQ8DOiwSyLIuZc/aguNX3FvS62j2usO5rGe3GXNNw4lwW\nC1haLTWcs7Rmb3vLPUf5o1+6m1uun+iYEFiuVFlcKlCuVGXluTaY3/zkhRzlSn1E3mY+j+Y7ZxfX\nuLa0/udpqwgiAO4GHgLQWj8G3Onb/xQwDgw5n8205A+AjwIXN97M/sGYabwvtHkRh+p8AJ1n8vk6\nE1Ci7vt3CpevrpJz+mor8xwKxQrPX17qeN1ypcqpC1nnuPU74Y3mps/VzyrN52K52hAZZp6h1HAM\ngIl0Emj/DFzN5d0XzAgQoRHzmxdKFc5cWXK3X1squKXXN+N59D5D/mehHwkiANJA1vO5rJTynvcM\n8DjwNPBFrXVOKfWzwKzW+u+x66PtCCzLaisAkvEIiVgXPgBP9NBExn75d1oymPaYRLZSuJ28kKVS\ntTpe9/lLS27Oxkbat+oIgOcvLdUN9N77X/LZnM3n1JAjAAI8A942+r9PqOHtpxnPb6DPXmt6zGZf\nt1+JBjgmB6Q8n8Na6yqAUuoW4PXAYWAF+DOl1FuAe4GqUuq1wG3Anyql3qS1nm13oampVLvd2052\nueAOFmWr1t5Y4ioAkxOjjKXsmXw0Fu14P/GLth9hYnyEG6fHAFgr2d/f730RlDOzNcdqbrW0rvta\nzznnv33e/fvacqHld/zTdy+5f19dKjA5OUoo1P2cxQz6larF/EqR2w7sZjVf4qxn9hlJxOraUXKm\n8ocOjjM1leL66XH7u6pWy/YWPWb/cIBnbCfT7t692tFzV5bdY8/On3a3L64Ue95/q8WauenUpVzf\n/z5BBMCjwBuAB5VSr8Se6Ruy2Db+gtbaUkrNAmNa63vMAUqph4Gf7zT4A8zNLXU6ZFt5zuP4nVtY\ncds7v7AMQKlQohCxB4+Fa6sd72d23j6vXCwTdtYEPnfZvka/90UQLMviuyfnSA/HSMQjXPL0WVCm\nplLr6ovvPHuFUAiO7E3z3KUc5y8suhFaXp549goAR/enOXUxx+mzV0kPx7u+3rJnwPnW05c4MD7E\n06cXqFoQj4Uplqqcu7jI+FDtlZtzoo7K+SJzc0vEQvbo/vyFxab3PDWV4jmPI/38pSzXTY103dad\nQKfn4uL8CqNDMYaTUZ45Pc+VKznC4RBPzcySiEeYzCS5cnWF2dncugR+K85dsd/fowfSnLqQ49Tz\nC6RHun+eumEjQiaICegLQEEp9Sjwh8C7lVJvVUrdp7U+C3wceEQp9VUgA/yJ7/wdE6rQSv3OO6ac\nhMcHECQT2Gs6yozGiUZCO8oENOc4wo5NjzGZGSK3UqRU3nzHZbFU4fSlHId2p5jebQ+QzRx+lWqV\nE+ez7N01zHX70/Zx6+j/StV2yB7aPUqImlPZOBtfenQSaHQEm88jPhNQuzZ49y03cSwL9sTjai7P\nRCaJmh5jrVDh3OwyuZUilxZWufFAht1jQxRL1TrB3QsWsnlGklFudX5zf1BAv9FRA9BaW8C7fJtn\nPPsfAB5oc/5r1t26PsM7OHtVzDofQBdOYG8eQDgUYlc6uaOigMxAeGx6jLNXbG1nIVdg767hTb3u\nc5dylCsW6tAYI8moc908+yfrZ8tnryxTKFZsAeVxwF63L93V9dYK9u84kUlCCE5fzFEqV9BnFwmH\nQrzshkm+/exsowBYKzGSjBKN2POw1FCMeLR9QqB3X7PIIsE2NZbKVSbTSY5Nj/G1715Cn73GLuc3\nPjY9Rm7FnsAt5PKk1qHxNcOyLBZyefaOD3PMMenqc4vcedPunnz/ZiCJYF1gZl+hUP3L5w7kMU8U\nUKA8gJrmAHYUSG6l2NWCMv2McYKpQ+NMug7OzS93YRyvRvOwr9s4qGq3fWNMtDmuEyYCaDgR5dj0\nGOVKle+fWeS5SzkO7x1lasz+7mZO4FHP4BMKhZjItE8InM/m3dpR4gRujnnGJjJJ1KHaQOz9vd3n\ncbF3E66ltRLFUpWJTJLr9qWJRcN1QQD9iAiALjCzr/2TI6yslag6USZmtu9NBOsuE9iepRoTwNy1\nnZE6oc8tMpKMcmBqJJB5o5fXBVsAtLuuUc/VdG1AWI8GZnIAhpJRlOPIfeixM1SqFsemx9wwT6/W\nWLUsltdK7j5Du4TAStXi2lLB1WQkDLQ55reeyCSZzAwxkU4yc26RZ89dIx4Nc92+dO256KHG7b1u\nLBrm6P40F+aWe25m6iUiALpgPpsnEY+wd3wYC1jO2z+sdyYfjYSJhEMB8wBqCWSAOwjNXnvhF4Vb\nyOaZz+Y5Nj1GOBSqxbhvsonLxPUfnBphdCjW8rrVqp2gNplJsiud3JCAWq3TADIAbkawmh53B3mv\nzX41X8ayaiGghsk2/XQ1m6dStTgwOUI0EhITUAtM35m+VIfGWMmXuTC3wtEDGaKR8KZMSMx31a47\njgVbngHfDSIAumAht8ZkOlmb0a3WC4BkPEIoFCIRi3TtBIZaItBO0ABmPLNwwGMC2lwBYOL6zXXH\nUnEi4VCD6en83DKrhbJrIhhJRknEIxs2AaWG4xxwZugh4MbpDEOJKJFwqM5k4+YA+OzP7XIBZp3n\nYiKTJDUcFxNQC+Y9M3GoPYPev9uZBjd+3aG6a/WzGUgEQAuW10pUPbVWVvMl1goVJjJJRt0Znf0C\nFkoVopGQ68xLJoItC2n8BKaEtBkkr1zdmABYzTefGVaq1Z7WGqpWrZbXMlmQZoAdSyUIhRpnXIVS\nhfOzy+6/jc7Iate1TTGRcJjxVKLhe/W52gwdbPv7ZAAn/Gq+1FCDx2sCAjjm3PPB3aOMJGOEQiFG\nh2N1M3Z/FrCh3czUCIDJzBCjQ7EtMQEtr5Xqfp/NFDqr+fp3rl2bjPm1Ge5MPFPTAAxq2iPwY5He\nmoBy9dc9uj9NNBJyn7VuWF5rfM42AxEATbi0sMK7P/QI/+BJJvLOKlJD9qzNqwEYOz7YNv1gtYDK\nhEMhYtGw+90AcxswAZ04v8gvfeBrHD+90LDvv//zaX7jo1/vmU3yS4+d4d0ffrRpzZMT57Mk4xEO\n7bZjlKMRZyD2vXB/9Lkn+c+f+pb779c/+vW6fItu8TqADRPpJNnlIiXPKm3GQX3MMzhMZJKsFcot\nhdqZy0v88ge+xuO6vmic1wQEcJMjfJSnDamhOEtrnTWAyXTrmlCuBuBooYViZVPDagulCvc/8I26\n3+c9D3xjUwoWXlpY4Vc++AjfOH657XGXr67yqx9+hP/3q6daHrOQyzOUiDCctIXr7rEhxlMJopEQ\n1zvhvkEc7t2y4NM84rEI1+1Lc/bKUld9dupill/54Nf41vc7pk5tGBEATfjuqQUqVYvveKpDeu17\nNROQowEUy64ZB+yy0EGdwAnHbAQwnkoQDoXcF309mHBLb/0Tw7krS6zkyz1TSZ+YmaNUrnJudrlu\ne9WymFtcY9/EMOFwLclmIp3kmlPIDOwZ34nzWSYzSX709oPccWwKgKdPNQqvIFSqVU5csOP6M57k\nm4lMEgu4umT/hpZloc8tMp5KMOW8rKZ90NoscPqiXdjt7Gx93675BMDLbpzkf371UX7yVYfdY1LD\nMdYKFVcIuRrAUHMNoKkJ6GotusUIjs30A5y6kGUlX+bo/jQ/evtBbjiYYa1Q4eSFbOeTu+T5S0tU\nqlbT59bL06cWKFcsvnn8UtP9lmUxn827vyXYg/19r7+Zd/7rlxCP1d7TiXSS1UK5Z2XK57N5ErGI\nG3oM9kTEsuwJUVCePDGPZdnhxJuNCIAmmAHylBPPDbUqoM1evrwzkBuS8QjlSrWhCqEfs5C8wTZX\nxJndgAnICKWm1Sedbb0oUrVWKHPmshPb77OvZ5eLlCuWaws1TGaSWBauxjBz3h5QX/Xivbztx4/x\nM6+7yWnf+gSUiev3qvzmunY77d/w4sIqy2sl1KGxuizQTpFAZlD2960ZQMyMMxoJ87pXHGZsNOEe\n4zqCHe2rpgHUC4DMqO2zaGsCSiddwbGZAsC8B6+/6whv+/FjvOmuI3Xbe4nx0XRKbnOT7M4uNl17\nYSVfplCsuDZ+w81HdnG7M8EwbCTyqxkLuTyTmWTdM1ULQw3+zpl73IqQaREAPqqW5Xrty5Uqz12y\nZyRe9a6ZEzjpEwDQeU0A/3n29w9xNZfvKDxaYdrUrvxwL7ITT17Iuvbaed8L5I/CMPjt224YpvOS\njA7FODg1wqkL2XXdvxvnPV0vAPxVNmecgmDH/Md1iAwx99UgAAr256FEY6kJQ81sWC+g/SYgEzHV\nbFCau7bK6JBdVsOvhW4G+twiIeDYQTuy6eiBDOFQaFOyW2t92/p+vKXFy5Vq0xmy+576nr1m9DIS\nyPYRlt3vNNzQZZ8VSxWec+5rK5JCRQD4uDC3wkq+TNp5wUz1wHoTkPMyr9l25UrVch25QOByEE0F\nQDpJ1TNL7hZ3gFmrf5Esy3L3nbuy3NLOHRTvA+1/gfy2UIPfxKLPLhIJhzi6P+Mec2x6jGK5yvOX\nuq//4488cq/rM6vocx0ERYsXr1YJtr5vTSaw0QCa4Z80GGHs1wBMe3MrxboZbtWymL225t7LZpuA\nSuUKpy/mmN496t7XUCLK4b2jPHcp1/NkxWZVdv1cnF9hea1UezebDKr+CKB21J7Hjc+051sInmQ8\nyuG9qYYqsa04dTEXqIptrxAB4MMM+D/x8kP2Z6OO5fLEomHSI3FGPeq3NwnMkHD+XmvzgxsTkfc8\nCFYSuB2tNIB8sUK5Yj9YFrb5ZSPos4uEQjQ1V7iZmL6XwbvojW1CWuLIvlSd+cxE73RrpjJx/VNj\nSTflv3bd2sBu7P/pkXhDSYpOoaqtTUABNADfjL2VCQhomqS0tGJPNiZdAbC5GsDpi/ZCKsd85jQ1\nPU6larlrKPQKt2/bBCiYd9F9N882PiP+SJx29NIE1O666tAYlarFyYud+8zcUywa3pIVAkUA+DAP\n2R037Wb/5AgnHXPEQjbPrrRt34tFwwwlIiytlmrrATczAbURAGYGlYjVDxp+e3W3mBfIb0s1A8Uu\nZ+GZjajxhVLFLnOwJ8WudKKJCcjWXvwvg1flPnXRNiGZMEyDt4ZKN7hx/b7vA+zfzbnu7OIa2eUi\nx6bHGqpApkbiRCPhpn1fKlfIrtQP3oa1gu0DioRbv041rbH2+yTiEWLRRqHRLBnM9UGlfRrAJoWC\n+sNkDUYg9NIMVLUs95lZWi22DH8017z92BRH9tnVW/2mwlbaZzN6aQJqd12jaQZZH2DGMbvdeoNd\nTG6zzUAiADwYG6OJDlHTYxRLVWbOLbK8VmIyXXPq2XHYxYZkLqjF9bfLBnbLQCT8PoCNzUpamYDM\nrPXWGyaJhEMbcuSddhZbUYfGmoZYtjYBOaue5fJNwzUBMs7M/MT5LJVqcD+A9vkTvEQjYTKj8brr\n+s0/YOzvjaGqAFc9y0Wu5Mt1bVstlNwIoFY0MwH5I4AMzbRAf5+ObrITuGZOy9RtP3YwY1c87aEj\neGml6A7k5YrVNILOsiz02UUyo3F2jw/xkusnKJWrDSHDrbTPZqRHeleBt5UJCOBG02cdhGapXOXU\nxRwHpkY5vGcU2HwzkAgAD5cWVllaLaGc2aEZTExssjeqJTUcZ3m11FDQDYJpAGvF+iQww+QG7JKm\nvgxAsVSts9OameJkOsmRfSnOXO4uNtmLd3ZoBqSr3tlqdo3hRJQh36AYi0bIjMSZz67ZDsaQ/XL4\nUYfGKBQrbkhrEGZaCBTDZGaIq7kCz56pT1BrPC5Zp9m59+QTCitrtb5bzZc7CgBT9G3ZmeEurRZb\nVqFspgX6k5s20wRUrlQ5eSHLgcmRhjYOJ2NM7x6ti5DbKP4BuNk9Xbm2Rnal6L6bL3HKLfsF0UIu\nTzwabmpa89PO4d4t7UxAw8kY03tG3SqxrXjuUo5SuepOrGDzM+dFAHhwi4g5g4MZTL7tJP54Z7Sp\noZhdnGvZnhnW+wA6F4TzF4Iz7EqvXy019WUMzUoPjA7HODY9RtWyOBXAJtkMfXbRLXPgptTnajH2\nJhyuGROZJFdzBZ67aJuQ/EICarPzoLNMY9fflU60vW7VsvjOiXlGktGG0tDe46BRAzO/hzHZmf60\nLIu1QsXNAm6FVwMw/phWg1QzZ7TfBDQyFLOr0m6CCej5y0sUS432f8OxQ2N1EXIbxdyn27dN7mnG\n57h/0fW76ra735W11wEIusjLhBH4G3RqL2TzRCNhUi0Wf1HT4y0jlwz1xQlbJwT2EhEAHowDxjxk\nY6MJ9owPuQ+HN6zRzIzmnAWm66OA7MGgnQnIXwjOEIuG2dXCDNGJxnLDtRdp2RN2aOy661HjS+UK\npy7m3DIH/hDLZU9J3GZMpJNUqpZbKbMZZntQO7OJDlFN7Pre64LtvzAF6tod5+9/MxMzqrnp20Kp\nQtWyOmsAyRgh7N+onQMYamUzmpmAjIALh0K2GXITTED+98BPTUD3ZtHzVn3brE3HnCCB8VSSfRPD\nnLhQMxWuFcqs5MuBzD8G//O7Xuzks0TL5yqIbytoFdteIgLAwdj//dEhylcqwGBe3nkjALyZwAHW\nBCg08R0YpsaHuZortK130gzz4kSc7NtWtWduPJghFFpfwtVzl5YoV6ruIOC3V7ezhUK9itzKDLMr\nnWRqzC7hWwnQB63CP1tft9FRXDuueZEw8yIe3muXEjCzVDcJrIMACIdDjDj1e2pZwM1ni9FImF2+\n+kVmpSlvqKlthuy9CWjmnK0ZthIA63XUt6Khb1cbQ5j1uUVGh2Lsn/C8m9O2qdBNSOwiAsjQi0ig\nQrFi+wjbXNf4UlpNusqVKifPZ9k3MUx6JO4mBIoJaIuYXVxjsUl0iHdQmcy00QA8A8CQEQCF1gKg\nme/AsGd8mErVYnG5u1wA8+IYAdaq+uRQIsqhPSmeu5hrmk3ZDu1LovLbq/0zVT9GYIRoP2Cr6XFW\nC2XOXu6cDl9zALce2L3Cu9XA5j2uMbdhjVAIDrmzVLs/jR+lkwkIbOG7tFqqE8Yt25FOsrhsl82w\nLIv5XJ6p8fqw1dRQjJV8ed1Jg82wl8lcZM+uYTKeTOb6+7Arnp5cZ8KeHzP4Htln143yR7AtZPNc\nzRUaNDy/pthNBJChFzNt0/521zV91irJ8eyVZQqlivts9tI/0Y4gi8LvWJ55/ionnIfn4oKdZu8f\nHIy5JBIONU3tn/PZhqE2qHvtit9+drZuIYpa/kAzDaBm//PHtLfDDCz7Jke44JhF3H0m8ciJHlHT\nY5y5vMRf/OOJuro5nfiXZ+0CVcY+PG6qfDoPaqeXwWgGB6ZsE1Irjk2P8cjTl/jzLz/Lbue7juxN\nc9uNk3XHeeP694wPNfuquusOJaJM7x7teFyDDyCXZ2w0wVjKfgZMX/sLwbUjNRTj8sIq2RVbsI+2\nEwCZJDPnszz4T6eIRcMUihV2+wWAc/7KWskdrL/23YstB7NkPMqP3nHQLT5oOHtliSeculcr+TL5\nYoWXTzc6570cmx7jwvwKn/3HE25EUlBuOTpRl/w3n80znIiy2105rV4A6BYanhH4Xz9+iXyx7AYN\nrMcE9Nj3rnQ94fK2P8h1jx2y++z5y0vccKC+f03ei7844ffPXKNYqtTVMOolAy0APvk333Nju8Ge\nlb7oSP0sciKT5ODUCJFIuK6wmXn5zMtWXwqi3gdw9soSH/mr49x9yz7+/etvtve5paAbf4Ldzgx+\nPpvnxoPB78fMSo2aXG8CKhKNhNx23nL9BH/3L+f45ycvBr+Aw+E9KdKOBhSNhBkbTbj1gDplYh6Y\nGiESDnHrDRNtr/GiI+NEwiG+6akOGQmH+PB//OE6rWn2mh3X/wM37W7r+JvMJBlJRnnxdbvqfkc/\ndtXIMJfmV9xtlWqVa0tFrj+Q9tTgsfs6qAkI7FmghV3R0nxuxcGpUeAKf/cv59xth50Zsvf77LbY\nAuDc7DKf/v+e7dCGGD94y766bZ95SDeEU77kuva/z0uu38XD37nAV5640Pa4Zjz2vSv8/s+/CnCC\nBrJ5do8PtYxsMm27wRcxNp5KcGBqhPNzK5yfq/1eB6ZaC3g/+ybs51GfW9ywSetgh+uq6TEefuIC\n+uy1RgHgWUDI4J2M7JtoHrSwUTo+tUqpEPAR4FYgD9yntT7t2f824FeBMvBprfXHlFJR4FPAESAO\n/K7W+m963/yNsVYos3fXMP/uJxRgJwI16+jf+Le3N2wbdey3xkbtrwYKNTOPWR3KW+XTmIeaaQBm\nptetWuop+PhUAAAgAElEQVRqAM491JuASqSG4+4g+aIj4/yXn/2BdYWCHpiq76PJTJJTF3JUqlWP\nCaj5bHwyM8Tv//wr67SpZuxKJ/mvP/dyQtEoi4urPPL0Jb5+/DInL2Z58ZFd7nHt4v+9xGMRfvu+\nVzTtby/hsF0y+MS5RVbzJYaTMa7lClQti8kmhQC7NQGBHW7s/dyM1/7ANDceHHPNBeFwiJe/dD+L\nnmdo1CeMnnXMc2+86wg3H66fyCzk8vzx334ffXaxTgCsFco8fznHoT2j/JvX3AjYGuyRvfXCxs9t\nN0zy3p+5M1B5Ay9/9chzzJxb5NpSgfFUwi7eVqrU960vCshMKpppeL/+b17GRY+wHh2KcbCNhucn\nPRLnd+57xbpLrxiC9JmbEHYuy+tfVdterVqcOJ91y1YbvP6JbRMAwJuBhNb6LqXUK4D3O9sM7wNu\nBlaB7yml/gL4KWBea/3vlFLjwJNAXwkAy7Iolqukh2PcdLi17RhoquL6X95mxeCMADB2c28suZsJ\n3FQArC8EzLw47pqxPiew9wUKhUIc7vDABmUik+TE+SzXlgos5BpL4vppJRz87JsYYWoqxdzcEoVS\nha8fv4w+u1gvADpErHjpJHQManqMmXOLzJzPctsNkzWzVpNS4N2YgEwugBmw2mkA0Ui4Ycbrzxr2\nrzVsciHufuk+dyF6Q7Vq8Rf/cKKhxMbJC1ksy9YIO70HXkKhENftSwc+3nDrpZzdt+cWecWL9tQl\nbiXjEWepy3oNYCGbZygRbVprKT0SJ92FCbMZe3YNs8dXFmQzyIwm2LNrmBPnF6lUq27m+LnZZdYK\nZe5Q9dVKtyISKIgT+G7gIQCt9WPAnb79TwHjgHniLODzwHs91+i7xUvNzCq2TtuaXwAkmuQBFIp2\niKBxUnlr4bfzARgNoNsIABMRsmd8iHAo5GYDF0sVCqVKoOSY9eANpZvvMg47KDceHCMUqlXyNMw4\n0SH7WsT1rwejTZgB1WvWikbCDCWirn+lOxNQrO77WmUCB8WrjXjXOGjmgA+HQxybHmNuMV+XtNcu\nM3ozOOaWR2503IZCIWepy9pwYRzg3dj1+xk1PUbel+Toz3EwbEUyWBABkAa8GUNlpZT3vGeAx4Gn\ngS9qrXNa61Wt9YpSKgX8JfBbPWtxjyg6pQvi0fUFQiVikTpnmncgDzvrAueLFS7O29VFASwLFh1V\nM98iEQzsiKLRodi6TECJeIR4LFK3BOHyWi0HYDMwA46ZyXQThheU4WSUQ7tTnL7kWaNhcY2FXKFt\nXP96OLo/49qFoXkW7kZMQGDP8DuZo4J+39JqseUaB16a5VfMnFskHApx9EB7p2+vOLwnRSIWaYjc\ncfvWl9tQq++/cwQA1P8GrarT9nq9gmYEMQHlAK+tIKy1rgIopW4BXg8cBlaAP1NK/bTW+r8rpaaB\n/wF8WGv9uSCNmZrqjUkiCGFH9UyNJNZ93cxogvnFNcIhOLAvU/fiDSejlKtVLnpWcVrI5imHwkxN\npahiH3twf6apart3Ypizl5eYnBwNPJteKZQZG7XvZzxlt21qKkXW8Tfsdkwqveb6Q7ZJ5sysbdo4\nsCfV0+uY77pVTXHmyhILq2VuOTrG005Zhztu3tPz+7rRMQONpJKsOML6xiMTTE2l2JVOcuLcIpOT\no1Sd3+bA3kzHNkzvq5X3GBuNs3t39yYU7zUOlZz6ORZcdJYRvfNFe1u24xUv3c/nHz7JmbkV3jiV\nIl+07f9HD2Y4dDC4+Wej3HzdLp6cmSM+FGfFuYcbDjt9OzbE2dll0mPDJGIRss7aHAf3Nj5TWzle\n9IpX3RblE1/8Hs9fWbbHAcf+PzU+xM037q47dnzXCOEQ5FZLm3avQQTAo8AbgAeVUq/Enukbsti2\n/4LW2lJKzQLjSqndwJeBX9RaPxy0MXNzvUktD4JxyFYr1XVfdyQRZR7b5DM/X1+3JhYNs7xW4tvf\nvwLAy2/azZceO8ups1fZm0mw5IScLWXXWFmql/BTUykyw3GK5SqnzlwNFKZpWRbZ5QLTu22b+VA8\nwkq+zKXLWc5dsF+iSGhz+jhq2S/xd0/aoYQj8UjPrmN8AACHHDPPt757kb3pBP/yjB0htH98qOf3\ndf2+NM+eucZjT13gvLNMYahcYW5uiWQsYi9feP4aV508kMJqgbm5dt8IFU9m+HAy2nWbvX0BUHYW\nopm9usq88zzvG0u2/N50wtY6npqZY25uie89f5VyxeLovvSWvnvX7U3x5MwcX//Oec47eR7hqtO3\njlb9/Nmr7EonOfn8VQCGY/XPlL8vXkhMZpIcPzXPldkcl+ZXWFot8pLr9ja9n7FUgkvzK23vdSPC\nIYj94wtAQSn1KPCHwLuVUm9VSt2ntT4LfBx4RCn1VSAD/Anwn4Ax4L1KqYeVUl9RSgXzwG0RxgQU\ni60/F86o4P6SzmCbhPLFCjPn7AqGNzvhpUblzRcrxGPhliGJ3TqA/PVljLlnuS7zdHN9AOY6m6Wu\n+zNQZ84udozr78W1FrJ50iNxNxbbW9fHXQ84kAko3vTv9WKCE5ZXiy3XOPASCduO5ctXV8kuF2oV\nWTtEUPUarxnErKNr7mXU07ewvuzefkdNj7GSL3N+drljFNukJyFwM+j41GqtLeBdvs0znv0PAA/4\n9v9H51/fUtqgDwBqA0FTO34sQqFYoVCs8PKbdzeUGMiXKk3PM3iLkl2/v7OpwM30HTICoPYidao9\ns1HisQjpkTg5J6disxx2o0MxDjhLRs4vrjG7uMatRyfaxvWvF1Mu49mz17i6lGd6d22WVXO+Flkt\nlIlGQk3r+vvx9n8vhHE0EmY4EeXMlSXWCpWOuRBgDz7HT19l5nzWrT1/rElF1s3kun1pYtGwK1y9\nQQPevgVPeecdJACOHRrj0eOXmTm3yMkL7ctumITAa0uFhsiuXjCwpSBMCYQgL24rzMPazJnnLQ2h\nDo3X1cIHOwrIXwraS7dlof1rzHpfpKVNdgJD/aC/mS+rcpaM/LKTINWu/MNGGEpEObwnxakLOWeB\n+9o9eWvxBykFbbAjiOzfvF0WcDekhmPukpSdciGglmj0zHMLnPIt+bhVxKJhju5Pc27WXsTH++z4\nF7tfT3mHfqe26t1i3RoHzdjoCoGdGFgB0AsNwAwEzQSA1yx0bHqMWNSeJZsHutBkPWAv3ZqA/PVl\ntlIDgJqKHo2ENxyX3Q7z8pgM5nb1hDZK6zpQteibtUKZoS4GUFMArlfC2CtIgvTFkX0p4tEwXz9+\npemSj1tFkL4Fp75/LLxp5svtYCqTZDyV4KmTC3VrHDTDLQstAqC3bDQMFDqYgOI1e7EpzTCZsYs7\nVasWhWKlaRKYodulIb31/qE+S9SvHWwGRmC1K4nbC4y5olypkohHOLy39/Z/Q10l2CalwJdWS6wW\nygy3WQvYj19AbxQjUEaHYi3XOPASjYQ5eiDj2pS3Kv7fj/e69VV267OBF7J2DkCv80q2k1AohJoe\nC/QbtKpN1SsGWAA4JqANFFkyD2uzgdxs81YXNbXwZxfXsGguOAzDyRhDiUjDSlSt8Jt5vIPU0lqJ\ncCgUyFG5XsyDutmqemY04To6bzyQabsO70Y5Nj2GGXaalQK/tlygVK4GNgHZ5zq/T4tS0N1i2tJN\nLoR3wLlxmwTA9QcybtnyeuFar12t5Ms7yvxj8GpexwJUsV3PCoFB2PHF4L57aoFvfu8yP/f6m+sG\ni1KplxpAsyggu2uPNZnpXJizQ0bbaQBgvxgX51f5z3/8WJO9IX7szoP88K37Ae+CL/UzTBMFNDoU\n3dSZeU0D2PyX9dj0GJevrm6q+QdgJBnjwNQo5+eWfYsB2X0768Ted2MCGu21BuAIlG76why7f3LE\nLeq31SRiEa7bl+bkhWxzDWC15IkA6r3zc7sxQti/xoEf4zt8XM9x5rI9DuxKJ/ml/+kWopGNT352\nvAbw1acu8s1nrrgvq8ENA92AAJjePcqNBzO89PrGyokvPjLO4b0p7lS15A4zOF5wKhd2ygT9gZv3\nMJSIcG2p0PDvwtwyX/7WWfdYb71/7/9Lq0WW26w/2ytuOJDhun1p7lC7Ox+8QX7opfs4ODXCy2/e\n/Gv96B0HuOnQGHs9L6npyytOVc9uTEAvu3GSGw5keha6+tKjExzek+JOXx2Zdhw9kOHmw+O85vYD\nPWnDenn17Qe44WB9Xwwn7YnK0mrJU2a5ryLIe8LeXcO87MZJfuyOg23NW7FohDvUFJFwqO79D7JQ\nUhB2vAZgVKdiqT6OtuYEXr8JKBmPcv/b72i6Tx2yq216MXb9804xsHZRQGBXdXzjXUea7vvDzz3J\nM89dJbdSJD0Sb6j3Pzpk/7SLK0VW8uWOpWo3yuhQjPf+jL9M1OZw9ECG//pzr9iSa91z2wHuua1+\noEzEIsRjYXeGOpwIPpt/2Y1TvOzG4IN1J45Nj/Ff7v2Bzgd6iEXD/PpbX9azNqyXV714L6968d66\nbfZSl1GWVos7MgLIEAqF+OWffmmgY3/xp27ZtHbseA3AjbrxrXxl6slsRAPoFr8JKNnFzNGPv6aI\nv95/JBxmJBl169pvZgTQIJIaimM5k7AgdYCE4KSG4yyv7WwTUL+wowWAcSIBDUsf9sIE1C3GBHTF\nqQ/ULIM4KG6m6lkjAOrr/YP9Ipn732wT0KDhFajdOIGFzqSG7aUujdl2p1QC7Ud2tABYaFJ/3+Ca\ngDZQCqJbhhJRRpJRqpZZRGb9A4c3mxIcAeCLla7LPBUNoKd4BaoIgN5i1k04c3mJaCREZlQmL5vF\nzhYAnhh6vw+gpgFszlqbrfDaMzdSDthkU16YW+baUqFpvf9e154RangXCRITUG8xExmzJvZmRq8N\nOjtbAHg1gLJPA3A0go2Ega4Hrzq70Xrw6tA4FrgLevsHedEANg8xAW0e3r4V88/msqMFgLd+RrHY\n3Aew1QLA69DaiAkIan6Ax/Us0FhfxjtL3Ump9P2ACIDNwzuR2UlVQPuRHS0AvCagQrl5GOh2moA6\nJYJ14uj+NNFIbeWqRh9A7UUaFRNQT6nzAYgJqKfUaQAiADaVnS0Acl4fgF8DcExAW+gEht6agOJO\nNqUJRxQT0Nbh7c8h0QB6inciIyagzWVnC4Bs6yigYrlKKIRbj2Sr8Kq0nRLBguAtAdDoBK59HhUT\nUE8xwjYU2rgmJ9QjJqCtY8cKgFK5QnalyHjKTiP3awClUpV4NLLlVQbrooB6MHP0Vqxs0ACcgmPD\niWhP6oYINYxwHU5sbo2lQURMQFvHjh0VFnL2mrum0FJjGGhlS5PADCPJqDtj3EgimOGGAxl3AGql\nAYj5p/cY4Srmn94z4mir4VDIncAJm8POFQCO+WefUyO9WSLYVtv/wa4BMpUZIhYN90QAJeNRjuyz\nlyts9AHECYXY1AVaBpWhRIRYNOwOVkLviEbCjA7F2JVObGq5b2EHF4MzReD2T9gCoFkpiO2avf3b\nH7uRnFO9szffd4zzc8sNdv5YNMx9b3gRU1JLpeeEQiHe8YYXiXa1Sdz7upuIbcMEbdDoOAIqpULA\nR4BbgTxwn9b6tGf/24BfBcrAp7XWH+t0zlZgIoD27BomFGoWBlohs00z45sO93Yd2+v3p1suHO+v\ntij0jjtv2vxy1IPKy471rmKq0JogIvbNQEJrfRdwP/B+3/73Aa8B7gZ+TSmVCXDOpmNMQJOZJIlY\npDERrFTd8iQwQRCEfiLICHg38BCA1voxwF/0/SlgHDB2BivAOZvOfDZPKATjqQTxWKROA6hWLSpV\na1ucwIIgCP1CkBEwDWQ9n8tKKe95zwCPA08DX9Ra5wKcs+ks5PKMpxJEI2Hi0XCdD6CWBCbx24Ig\nDC5BvKA5IOX5HNZaVwGUUrcArwcOAyvAnyml3oI9+Dc9px1TU6lOhwSiXKmyuFTgpiO7mJpKMTIU\nYyGbd78/u2yHiI6OxHt2zV7Tr+3aDqQvakhf1JC+2DhBBMCjwBuAB5VSr8Se6RuywCpQ0FpbSqlZ\nYMw5500tzmnJ3NxSN21v/T2La1QtSA/HmJtbIhwKkS+W3e83/gGrUu3ZNXvJ1FSqL9u1HUhf1JC+\nqCF9UWMjgjCIAPgC8Fql1KPO53uVUm8FRrTWn1RKfRx4RClVAE4BfwJUgB/3nrPuFq4DrwMYIBEL\nU65YVKpVIuEwpcrWrwYmCILQb3QUAFprC3iXb/OMZ/8DwANNTvWfs2WYEFBTSMrY+oulKkOJmj9g\nqyuBCoIg9BM7cgps1gGYcDUAIwDsgb+0TWsBCIIg9BM7cgSsmYDsyFRT8sGEgm7HgvCCIAj9xo4c\nAWsmILuQlKsBFI0GIGGggiAIfS0AypUq3zh+mXKlYwRpHfPZNdIjcdfGbwZ6sy6wqQwqGoAgCINM\nX4+AT56Y5xNf/B4PP3Eh8DlVy+JqrlC3kpCx9dc0APEBCIIg9PUIuORUzHz27LXA52SXi1SqVt1K\nQqb+fs0H4JiAJApIEIQBpq8FQN6Zsc+cW6RqFr7tgCkD7V1JyAz0JgpInMCCIAgvEAGwki9zYW4l\n0Dn+JDCoOYEL/jBQqTcuCMIA09cjYN5Twnnm3GKgc/xJYFAb6I3zVxLBBEEQ+lwAFEpl928d0A/g\nTwIDSQQTBEFoRl+PgEYDiMfCzJxbxArgBzAmoHoNoN4EJD4AQRCEF4gAeNHhXeRWS1y+utrxnIVc\nnpFktG6934SnFhBIIpggCAK8QATALUcnANBn2/sBLMtiIZuvM/+AtxREvQYgJiBBEAaZvh4BC8UK\niViEmw6NAaA7OIKXVksUy1W3BpAh7i8FIZnAgiAI/S0A8sUyyXiEvbuGSQ/HOvoBmkUAgScM1FcM\nThLBBEEYZPpbAJQqJOIRQqEQx6bHuLZUYG5xreXxzSKAwFMKolRfDC4meQCCIAwwfT0C5osVkk4Z\nB3VoHGjvB2iWBAaNYaDFcpVoJEQ4FOp5mwVBEF4o9K0AqFoWxWKFZNyO5lHTnf0AzUJAAcLhENFI\nuC4TWOz/giAMOn07ChZLFSxwNYD9UyOEQjDb1gTUWAfIkIiFa5nA5apkAQuCMPD0rQAwIaBGAIRD\nIZLxCAVPeQg/C7k8yXiEkWTjUsfxWMSjAVQkBFQQhIGn46LwSqkQ8BHgViAP3Ke1Pu3s2wN8FrCA\nEHAb8JvAp4DPAEeAMvAOrfVMw5e3wQz0CU+yVjIeJV8sNz3esiwWcnYOQKiJbT8Ri7CaLwF2Qlhq\nONZNcwRBEHYcQabBbwYSWuu7gPuB95sdWusrWutXa61f4+x7HPgE8JNARGv9g8BvA7/XbcNqGkBN\nRrXTAFYLZdYKlQb7vyEeC7thoKVyVUJABUEYeIIIgLuBhwC01o8Bd7Y47kPAO7XWFjADRB3tIQMU\nu22YmembxVzAnsXnWwiAVhFAhngsQrFYwbIsiuWKhIAKgjDwBBkF00DW87mslKo7Tyn1RuC41vqk\ns2kZuA54FngA+GC3DTMD/VDcawKKUCxXqVYbk8EWWuQAGBKxCJbzvZYlZSAEQRA6+gCAHJDyfA5r\nrf2rtL8d+IDn87uBh7TWv6WUOgA8rJR6ida6rSYwNVW7TOJ8DoCJXSPu9vSoPbiPpocYGaq34ee/\nPwvAdQfH677HkBqJAxAfsv8fHU40Pa5f6Oe2bTXSFzWkL2pIX2ycIALgUeANwINKqVcCTzc55k6t\n9Tc8n68CJefvRec6HY3uc3NL7t+zC8sAlIsld3sYe+Z/4VKW8VSi7twzF20lJR6q/x4Xp4TE2Qt2\nHoFVrTY/rg+Ymkr1bdu2GumLGtIXNaQvamxEEAYRAF8AXquUetT5fK9S6q3AiNb6k0qpSepNRGBr\nA59SSn0ViAH3a61bB/A3Ie9GAXnKOjvmINs/UC8AWpWBMBin7/JayfksJiBBEAabjgLAceq+y7d5\nxrN/Hrjdd84K8L9upGHGCZxM1PsA7H2NjuCFbJ5YNEy6RXinCSddXrUFQEzWAhAEYcDp22mwCfdM\nxuqjgKCFAMjlmUg3zwGA2poAogEIgiDY9O0o6M8Etv+2FRZ/LkC+WGZ5rdTS/AMeDcARAFILSBCE\nQadvR0HXBxBvZgKqzwZuVQTOSzwmPgBBEAQvfTsKuj4AXyYw2OsEeFnIFYDWDmBoNAFJMThBEAad\nPhYAjSYgNwqoUC8Altfs9IJWDmBoNAHFJRNYEIQBp29HwUKpQjRi1/E3uD4Anwaw5giEoUTroCZ/\nGKj4AARBGHT6dhTMexaDMbTyAZgqn8NNykAbEnF/FJCYgARBGGz6VgAUiuW6UtDQOg9gtWALhOFE\naxOQJIIJgiDU07ejYL5YqUsCg5od3x8GuuYIgKFE61m9ObfklIQWE5AgCINO346C+WKlLgkMaj6A\nBg0g72gAyTYagM/pKwJAEIRBpy9HwVK5SqVq1UUAQRsfgGsC6qwBGOJSCkIQhAGnLwWAifJJ+JzA\n4XCIeDTcoAGsFcpEI+G2sf3+AV80AEEQBp2+HAXzBZME1jigJ+ORhjDQ1Xy5bQQQNNEARAAIgjDg\n9OUomC81loEwJOKNy0KuFsoMt8kBAIhGQnjrxEkmsCAIg05/CoAmWcCGZDxaJwAsy2KtUG6bBAYQ\nCoXqzECSCSwIwqDTl6Ngs1LQBlsDKGM5K3yVylXKFaujCQjqzUBiAhIEYdDpy1GwWSE4QzIewbJq\n8fy1CKDOAsAM+iGoKzEhCIIwiPTlKNisFLQh6VsUppYEFkADcL4vFg23XDhGEARhUOhrAdDKBwA1\nR3EtCSyIBlATAIIgCINOX46ENRNQcx8A1EJFuzEBJRzHrySBCYIgBFgUXikVAj4C3Arkgfu01qed\nfXuAzwIWtmn9NuA3tdYfV0q9B3gTEAM+orX+dNBGmTj/Vj4A7zHdmIDMwC8agCAIQjAN4M1AQmt9\nF3A/8H6zQ2t9RWv9aq31a5x9jwOfUErdA7zKOedHgOluGmUWfPEnb0FjRdBuTEDm+yQCSBAEIZgA\nuBt4CEBr/RhwZ4vjPgS8U2ttAT8BHFdK/RXw18AXu2mUse/7q4FC48Lwa91EATkmIEkCEwRBCCYA\n0kDW87mslKo7Tyn1RuC41vqks2kSuAN4C/Au4M+7aVTNCdw4qJtZ/Fqx3gfQjQlINABBEIQAPgAg\nB6Q8n8Na66rvmLcDH/B8XgC+r7UuAzNKqbxSalJrPd/uQlNT9mUs5/PBfRmSvoF999SS3fB4lKmp\nFFUnnPPgvox7fivG00MAjIzEOx673fR7+7YS6Ysa0hc1pC82ThAB8CjwBuBBpdQrgaebHHOn1vob\nns+PAP8B+COl1H5gGFsotGVuzh7cc8sFQkAuu8qSL16/mLcXgF+4usrc3BLXFtcAyK8W3PNbUS45\nZaSrVsdjt5OpqVRft28rkb6oIX1RQ/qixkYEYRAB8AXgtUqpR53P9yql3gqMaK0/qZSapN5EhNb6\nb5VSP6SU+hZ2dNAvOL6BQBSKFRLxSNNkrWSsflGYbkxAJoRUTECCIAgBBIAzcL/Lt3nGs38euL3J\nee9Zb6PsBeGbO2rdMFBPFFAo1DxnwI9JBIuKABAEQejTRLBSpWExGEPCtyrYmlMKOkhpBwkDFQRB\nqNGXI2G+WO6oAbilIAqdF4MxmDDQuISBCoIg9J8AqFYtiqVq01LQ0DwRLIj9H2oagGQCC4Ig9KEA\nqJWBaC4AopEwkXCIQrFCpVqlUKoESgIDODA1wq50guv3p3vWXkEQhBcqwUbOLaRdKWiwV/ZKxOxF\nYdackhHDyVig757MDPEHv/CDvWmoIAjCC5y+0wDaLQZjSCbsdYFX8yUAhpqUjBAEQRDa04cCoL0J\nCHA0gEpNA0gE0wAEQRCEGn0nAAoBBEAyHqVQqmkAQaOABEEQhBp9JwA6+QDAFg6lcpWlNWMCEgEg\nCILQLf0nAEoBfACOcFhcKgDBSkELgiAI9fSfAAjiA3D2XTUCQExAgiAIXdN3AsD1AbRZt9doB0YA\niAlIEAShe/pOAATRAMy+a0t5QExAgiAI66HvBEDBdQK38QE42sHVnJiABEEQ1kvfCYBaIlhnDSC7\nbC8OIyYgQRCE7ulDARDcCVy17DVmJBNYEAShe16QAsAbIpqMR4iE++42BEEQ+p6+GzmvLReIRcNt\nzTreJDEx/wiCIKyPvhMAC9k8u9LJtit8DXkEgDiABUEQ1kdfCYBCscLyWonJTLLtcV4NQEJABUEQ\n1kfH0VMpFQI+AtwK5IH7tNannX17gM8CFhACbgN+U2v9cWf/buDbwI9prWeafH0d8zk7rn8i3V4A\neH0AYgISBEFYH0FGzzcDCa31XUqpVwDvd7ahtb4CvBpAKfVK4HeATzifo8DHgNWgjVnIrgEw0UkD\niIkJSBAEYaMEMQHdDTwEoLV+DLizxXEfAt6ptbacz38AfBS4GLQxC1lbA+hkAkqKCUgQBGHDBBEA\naSDr+VxWStWdp5R6I3Bca33S+fyzwKzW+u+xTUOBCGoCkiggQRCEjRNk9MwBKc/nsNa66jvm7cAH\nPJ/vBapKqddi+wX+VCn1Jq31bLsLLeftHIBj100yNT7UtlHJuL0q2O6JEaamUm2PfSGyE+9pvUhf\n1JC+qCF9sXGCCIBHgTcADzp2/qebHHOn1vob5oPW+h7zt1LqYeDnOw3+ABdnl4iEQ1SLJebmym2P\njTvLQlbLFebmlgLcxguHqanUjrun9SJ9UUP6oob0RY2NCMIgAuALwGuVUo86n+9VSr0VGNFaf1Ip\nNUm9iciP1WZfHfPZPOOpBOFwZ6tRMh4htyImIEEQhPXScfR0nLrv8m2e8eyfB25vc/5rgjSkWKqQ\nXSly06GxIIe7jmCJAhIEQVgffZMINr8YLATUYEpCDydim9YmQRCEnUzfCIArV+10gU4RQIakY/qR\nSqCCIAjro28EwOy17jSAXekksWiYzEh8M5slCIKwY+kbA/rcNVsDmMy0D/80/PQ91/Ojtx9gOCkm\nIKM+VM4AAAUnSURBVEEQhPXQNwLgiiMAgmoAI8kYIzL4C4IgrJu+MQHNXVsjBOxKJba7KYIgCANB\n3wiA2WurjKUSRCN90yRBEIQdTd+MtguLa4HNP4IgCMLG6RsBULVgMmAIqCAIgrBx+kYAQHAHsCAI\ngrBxRAAIgiAMKH0lAMQEJAiCsHX0lQAQDUAQBGHr6C8BIBqAIAjCltE3AmBsNEE8JoXdBEEQtoq+\nEQC7dwWrASQIgiD0hr4RALerPdvdBEEQhIGibwTA2/7VTdvdBEEQhIGibwSAIAiCsLV0LAetlAoB\nHwFuBfLAfVrr086+PcBnsRd+DwG3Ab8JfMr5dwSIA7+rtf6bTWi/IAiCsE6CaABvBhJa67uA+4H3\nmx1a6yta61c7C7/fDzwOfAJ4OzCvtf5h4HXAh3veckEQBGFDBBEAdwMPAWitHwPubHHch4B3aq0t\n4PPAez3XKG2wnYIgCEKPCbIiWBrIej6XlVJhrXXVbFBKvRE4rrU+CaC1XnW2p4C/BH6rd00WBEEQ\nekEQDSAHpLzneAd/h7cDH/duUEpNA18BPqO1/tyGWikIgiD0nCAawKPAG4AHlVKvBJ5ucsydWutv\nmA+Oc/jLwC9qrR8O2JbQ1FSq81EDgvRFDemLGtIXNaQvNk7Isqy2B3iigF7qbLoXuAMY0Vp/Uik1\nCfyd1vp2zzkfAP4X4Fns6CALeJ3WutD7WxAEQRDWQ0cBIAiCIOxMJBFMEARhQBEBIAiCMKCIABAE\nQRhQRAAIgiAMKEHCQDeNdnWGBgGlVBRfzSTge8CfAFXs5Lpf3K72bQdKqd3At4EfAyoMaF8opd4D\nvAmIYb8jX2UA+8J5Rz6D/Y6UgXcwgM+FUuoVwP+ptX61UuooTe5fKfUO4H/Hrrzwu1rrv+30vdut\nAbSsMzQgeGsm/SvsmknvB/6T1voeIKyU+tfb2cCtxHnZPwasOpsGsi+UUvcAr3Leix8BDjGgfQH8\nJBDRWv8g8NvA7zFgfaGU+nXsGmsJZ1PD/Tu5V78MvAp7LPl9pVSs03dvtwAIWmdop+KtmRTBnuHc\nrrX+mrPtS9gz4UHhD4CPAhex80cGtS9+AjiulPor4K+BLzK4fTEDRB1rQQZ7djtofXES+CnP5zt8\n9/9a4OXAI1rrstY6B5yglrvVku0WAE3rDG1XY7YarfWq1nrFVzMp5DlkCfuh3/EopX4WmNVa/z21\nPvA+CwPTF8AkdrLlW4B3AX/G4PbFMnAddlLpA8AHGbB3RGv9BezJocF//2nscj3esXSZAP2y3YNt\nkDpDOxpfzaTPYtv1DClgcVsatvXcC7xWKfUwtk/oT4Epz/5B6osF4MvObG4G2z/mfZkHqS/eDTyk\ntVbUnou4Z/8g9YWh2RiRwxYE/u1t2W4B8Ci2jY82dYZ2LJ6aSb+htf6Ms/k7Sqkfdv5+HfC1pifv\nMLTW9zhrS7waeBL434AvDWJfAI9g23FRSu0HRoB/dHwDMFh9cZXazHYRO3DlOwPaF4YnmrwX/wLc\nrZSKK6UywE3A8U5ftK1RQMAXsGd9jzqf793OxmwD9wNjwHuVUv8Zu2bSrwAfchw43wce3Mb2bTf/\nB/CJQesLrfXfKqV+SCn1LWx1/13A88AnB60vgA8An1JKfRU7Iuo92AtPDWJfGBreC621pZT6IPbk\nIYTtJC52+iKpBSQIgjCgbLcJSBAEQdgmRAAIgiAMKCIABEEQBhQRAIIgCAOKCABBEIQBRQSAIAjC\ngCICQBAEYUARASAIgjCg/P96ryh41EAF4gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x117e55b50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# see how many decision trees are minimally required make the accuarcy consistent\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pylab as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "trees=range(100)\n",
    "accuracy=np.zeros(100)\n",
    "\n",
    "for i in range(len(trees)):\n",
    "   clf=RandomForestClassifier(n_estimators= i+1)\n",
    "   model=clf.fit(train_feature, train_target)\n",
    "   predictions=model.predict(test_feature)\n",
    "   accuracy[i]=sklearn.metrics.accuracy_score(test_target, predictions)\n",
    "\n",
    "plt.plot(trees,accuracy)\n",
    "\n",
    "# well, seems like more than 10 trees will have a consistent accuracy of 0.82.\n",
    "# Guess there's no need to have an ensemble of 100 trees!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}