random_forest_classifier.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Goals: rebuild Random Forest Classifier from scratch\n",
    "\n",
    "Use only python and some basic numpy function (numpy slicing, np.mean, np.nonzero)\n",
    "\n",
    "Produce comparable results to Sklearn Ramdon Forest on some classification dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from model.utils import draw_tree\n",
    "from model.metrics import *\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generate a random-generated dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "nrows=1000\n",
    "ncols=10\n",
    "df = pd.DataFrame(np.random.randint(1,10,(nrows,ncols)))\n",
    "df['y'] = np.random.randint(0,2,(nrows,))\n",
    "y = df.y.values\n",
    "df.drop('y',axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\qtran\\AppData\\Local\\Continuum\\Miniconda3\\envs\\fastai-cpu\\lib\\site-packages\\sklearn\\ensemble\\weight_boosting.py:29: DeprecationWarning: numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n",
      "  from numpy.core.umath_tests import inner1d\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log loss score: 0.24609764270160742\n",
      "AUC score: 0.9992256902761105\n"
     ]
    }
   ],
   "source": [
    "rf = RandomForestClassifier(n_estimators=5,bootstrap=False,min_samples_leaf=3)\n",
    "rf.fit(df,y)\n",
    "\n",
    "rf_pred = rf.predict_proba(df)[:,1]\n",
    "print(f'Log loss score: {logloss(y,rf_pred)}')\n",
    "print(f'AUC score: {metrics.roc_auc_score(y,rf_pred)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from model.random_forest import RandomForest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log loss score: 0.2061878085760237\n",
      "AUC score: 0.9726410564225689\n"
     ]
    }
   ],
   "source": [
    "tb = RandomForest(df, y, n_trees=5, sample_sz=nrows,is_reg=False)\n",
    "tb_pred = tb.predict_proba(df)\n",
    "print(f'Log loss score: {logloss(y,tb_pred)}')\n",
    "print(f'AUC score: {metrics.roc_auc_score(y,tb_pred)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Results are comparable!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Benchmarking: Wisconsin Breast Cancer prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Breast Cancer datasets is available machine learning repository maintained by the University of California, Irvine. The dataset contains 569 samples of malignant and benign tumor cells.\n",
    "\n",
    "- The first two columns in the dataset store the unique ID numbers of the samples and the corresponding diagnosis (M=malignant, B=benign), respectively.\n",
    "- The columns 3-32 contain 30 real-value features that have been computed from digitized images of the cell nuclei, which can be used to build a model to predict whether a tumor is benign or malignant.\n",
    "\n",
    "This dataset's missing values are already handled."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(569, 32)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>diagnosis</th>\n",
       "      <th>radius_mean</th>\n",
       "      <th>texture_mean</th>\n",
       "      <th>perimeter_mean</th>\n",
       "      <th>area_mean</th>\n",
       "      <th>smoothness_mean</th>\n",
       "      <th>compactness_mean</th>\n",
       "      <th>concavity_mean</th>\n",
       "      <th>concave points_mean</th>\n",
       "      <th>...</th>\n",
       "      <th>radius_worst</th>\n",
       "      <th>texture_worst</th>\n",
       "      <th>perimeter_worst</th>\n",
       "      <th>area_worst</th>\n",
       "      <th>smoothness_worst</th>\n",
       "      <th>compactness_worst</th>\n",
       "      <th>concavity_worst</th>\n",
       "      <th>concave points_worst</th>\n",
       "      <th>symmetry_worst</th>\n",
       "      <th>fractal_dimension_worst</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>842302</td>\n",
       "      <td>M</td>\n",
       "      <td>17.99</td>\n",
       "      <td>10.38</td>\n",
       "      <td>122.80</td>\n",
       "      <td>1001.0</td>\n",
       "      <td>0.11840</td>\n",
       "      <td>0.27760</td>\n",
       "      <td>0.3001</td>\n",
       "      <td>0.14710</td>\n",
       "      <td>...</td>\n",
       "      <td>25.38</td>\n",
       "      <td>17.33</td>\n",
       "      <td>184.60</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>0.1622</td>\n",
       "      <td>0.6656</td>\n",
       "      <td>0.7119</td>\n",
       "      <td>0.2654</td>\n",
       "      <td>0.4601</td>\n",
       "      <td>0.11890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>842517</td>\n",
       "      <td>M</td>\n",
       "      <td>20.57</td>\n",
       "      <td>17.77</td>\n",
       "      <td>132.90</td>\n",
       "      <td>1326.0</td>\n",
       "      <td>0.08474</td>\n",
       "      <td>0.07864</td>\n",
       "      <td>0.0869</td>\n",
       "      <td>0.07017</td>\n",
       "      <td>...</td>\n",
       "      <td>24.99</td>\n",
       "      <td>23.41</td>\n",
       "      <td>158.80</td>\n",
       "      <td>1956.0</td>\n",
       "      <td>0.1238</td>\n",
       "      <td>0.1866</td>\n",
       "      <td>0.2416</td>\n",
       "      <td>0.1860</td>\n",
       "      <td>0.2750</td>\n",
       "      <td>0.08902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>84300903</td>\n",
       "      <td>M</td>\n",
       "      <td>19.69</td>\n",
       "      <td>21.25</td>\n",
       "      <td>130.00</td>\n",
       "      <td>1203.0</td>\n",
       "      <td>0.10960</td>\n",
       "      <td>0.15990</td>\n",
       "      <td>0.1974</td>\n",
       "      <td>0.12790</td>\n",
       "      <td>...</td>\n",
       "      <td>23.57</td>\n",
       "      <td>25.53</td>\n",
       "      <td>152.50</td>\n",
       "      <td>1709.0</td>\n",
       "      <td>0.1444</td>\n",
       "      <td>0.4245</td>\n",
       "      <td>0.4504</td>\n",
       "      <td>0.2430</td>\n",
       "      <td>0.3613</td>\n",
       "      <td>0.08758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>84348301</td>\n",
       "      <td>M</td>\n",
       "      <td>11.42</td>\n",
       "      <td>20.38</td>\n",
       "      <td>77.58</td>\n",
       "      <td>386.1</td>\n",
       "      <td>0.14250</td>\n",
       "      <td>0.28390</td>\n",
       "      <td>0.2414</td>\n",
       "      <td>0.10520</td>\n",
       "      <td>...</td>\n",
       "      <td>14.91</td>\n",
       "      <td>26.50</td>\n",
       "      <td>98.87</td>\n",
       "      <td>567.7</td>\n",
       "      <td>0.2098</td>\n",
       "      <td>0.8663</td>\n",
       "      <td>0.6869</td>\n",
       "      <td>0.2575</td>\n",
       "      <td>0.6638</td>\n",
       "      <td>0.17300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>84358402</td>\n",
       "      <td>M</td>\n",
       "      <td>20.29</td>\n",
       "      <td>14.34</td>\n",
       "      <td>135.10</td>\n",
       "      <td>1297.0</td>\n",
       "      <td>0.10030</td>\n",
       "      <td>0.13280</td>\n",
       "      <td>0.1980</td>\n",
       "      <td>0.10430</td>\n",
       "      <td>...</td>\n",
       "      <td>22.54</td>\n",
       "      <td>16.67</td>\n",
       "      <td>152.20</td>\n",
       "      <td>1575.0</td>\n",
       "      <td>0.1374</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.4000</td>\n",
       "      <td>0.1625</td>\n",
       "      <td>0.2364</td>\n",
       "      <td>0.07678</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         id diagnosis  radius_mean  texture_mean  perimeter_mean  area_mean  \\\n",
       "0    842302         M        17.99         10.38          122.80     1001.0   \n",
       "1    842517         M        20.57         17.77          132.90     1326.0   \n",
       "2  84300903         M        19.69         21.25          130.00     1203.0   \n",
       "3  84348301         M        11.42         20.38           77.58      386.1   \n",
       "4  84358402         M        20.29         14.34          135.10     1297.0   \n",
       "\n",
       "   smoothness_mean  compactness_mean  concavity_mean  concave points_mean  \\\n",
       "0          0.11840           0.27760          0.3001              0.14710   \n",
       "1          0.08474           0.07864          0.0869              0.07017   \n",
       "2          0.10960           0.15990          0.1974              0.12790   \n",
       "3          0.14250           0.28390          0.2414              0.10520   \n",
       "4          0.10030           0.13280          0.1980              0.10430   \n",
       "\n",
       "            ...             radius_worst  texture_worst  perimeter_worst  \\\n",
       "0           ...                    25.38          17.33           184.60   \n",
       "1           ...                    24.99          23.41           158.80   \n",
       "2           ...                    23.57          25.53           152.50   \n",
       "3           ...                    14.91          26.50            98.87   \n",
       "4           ...                    22.54          16.67           152.20   \n",
       "\n",
       "   area_worst  smoothness_worst  compactness_worst  concavity_worst  \\\n",
       "0      2019.0            0.1622             0.6656           0.7119   \n",
       "1      1956.0            0.1238             0.1866           0.2416   \n",
       "2      1709.0            0.1444             0.4245           0.4504   \n",
       "3       567.7            0.2098             0.8663           0.6869   \n",
       "4      1575.0            0.1374             0.2050           0.4000   \n",
       "\n",
       "   concave points_worst  symmetry_worst  fractal_dimension_worst  \n",
       "0                0.2654          0.4601                  0.11890  \n",
       "1                0.1860          0.2750                  0.08902  \n",
       "2                0.2430          0.3613                  0.08758  \n",
       "3                0.2575          0.6638                  0.17300  \n",
       "4                0.1625          0.2364                  0.07678  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"data/cancer.csv\")\n",
    "print(data.shape)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# some preprocessing\n",
    "target = data['diagnosis'].apply(lambda row: 1 if row=='M' else 0)\n",
    "\n",
    "data.drop(['id','diagnosis'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "def get_train_val(X,y):\n",
    "    X_train,X_val,y_train,y_val =  train_test_split(X, y, test_size=0.15, random_state=42)\n",
    "    return X_train.reset_index(drop=True),X_val.reset_index(drop=True),y_train,y_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,X_val,y_train,y_val = get_train_val(data,target)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### sklearn Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log loss train score: 0.07273869687440554\n",
      "Log loss val score: 0.13482735377469804\n",
      "AUC train score: 0.9980748074807481\n",
      "AUC val score: 0.9901620370370371\n"
     ]
    }
   ],
   "source": [
    "rf = RandomForestClassifier(n_estimators=30,max_depth=4,bootstrap=False,min_samples_leaf=5)\n",
    "rf.fit(X_train,y_train)\n",
    "\n",
    "rf_train_pred = rf.predict_proba(X_train)[:,1]\n",
    "rf_val_pred = rf.predict_proba(X_val)[:,1]\n",
    "\n",
    "print(f'Log loss train score: {logloss(y_train,rf_train_pred)}')\n",
    "print(f'Log loss val score: {logloss(y_val,rf_val_pred)}')\n",
    "print(f'AUC train score: {metrics.roc_auc_score(y_train,rf_train_pred)}')\n",
    "print(f'AUC val score: {metrics.roc_auc_score(y_val,rf_val_pred)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,15,'Predicted label')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "# Compute confusion matrix for a model\n",
    "cm = confusion_matrix(y_val, rf.predict(X_val))\n",
    "# view with a heatmap\n",
    "sns.heatmap(cm, annot=True, cmap='Blues', xticklabels=['no', 'yes'], yticklabels=['no', 'yes'])\n",
    "plt.ylabel('True label')\n",
    "plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 3.55 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "tb = RandomForest(X_train, y_train, n_trees=30, sample_sz=100,min_leaf=5,is_reg=False,max_features=0.9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log loss train score: 0.12035922333936774\n",
      "Log loss val score: 0.11296411211816172\n",
      "AUC train score: 0.9934543454345435\n",
      "AUC val score: 0.9947916666666667\n"
     ]
    }
   ],
   "source": [
    "tb_train_pred = tb.predict_proba(X_train)\n",
    "tb_val_pred = tb.predict_proba(X_val)\n",
    "print(f'Log loss train score: {logloss(y_train,tb_train_pred)}')\n",
    "print(f'Log loss val score: {logloss(y_val,tb_val_pred)}')\n",
    "print(f'AUC train score: {metrics.roc_auc_score(y_train,tb_train_pred)}')\n",
    "print(f'AUC val score: {metrics.roc_auc_score(y_val,tb_val_pred)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Los loss and ROC AUC score is a little bit better than Sklearn RF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,15,'Predicted label')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute confusion matrix for a model\n",
    "cm = confusion_matrix(y_val, tb_val_pred>=0.5)\n",
    "\n",
    "# view with a heatmap\n",
    "sns.heatmap(cm, annot=True, cmap='Blues', xticklabels=['no', 'yes'], yticklabels=['no', 'yes'])\n",
    "plt.ylabel('True label')\n",
    "plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "My RF classifier is able to draw a comparable result to sklearn Random Forest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Calculate feature importance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Feature importance will be calculated by using **permutation importance**. There is an argument which is mentioned in this [doc](http://parrt.cs.usfca.edu/doc/rf-importance/index.html#3) that feature importance calculated in sklearn is misleading. SKlearn used **mean decrease in impurity (gini improtance)**, which does not always give an accurate picture of importance. \n",
    "\n",
    "On the other hand, **permutation importance** is calculated as follows: Record a baseline accuracy (classifier) or R2 score (regressor) by passing a validation set or the out-of-bag (OOB) samples through the Random Forest. Permute the column values of a single predictor feature and then pass all test samples back through the Random Forest and recompute the accuracy or R2. The importance of that feature is the difference between the baseline and the drop in overall accuracy or R2 caused by permuting the column. This is more expensive to calculate, but results are more reliable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from model.utils import permutation_importances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature importance from sklearn RF model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def metric(rf,X_val,y_val):\n",
    "    y_val_pred = rf.predict_proba(X_val)\n",
    "    if len(y_val_pred.shape) ==2:\n",
    "        y_val_pred = y_val_pred[:,1]\n",
    "    return logloss(y_val,y_val_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fea_imp_df = permutation_importances(rf,X_val,y_val,metric)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature importance from 'scratch' RF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABP0AAAJCCAYAAAC777L3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xu0XVV99//3hxiLXIRqUx60QChVqSBEcoBGBQNSH21aRBtFpVqQAcVaUccDNq2PPkpFg1hRtLYNiikXK1UoUmIJNgIJICEnkCugVIg/rbQEL0iQa/L9/bFnymk85+TCSfY5O+/XGBln7bnmmvO79s5fnzHnWqkqJEmSJEmSJPWOHbpdgCRJkiRJkqSRZegnSZIkSZIk9RhDP0mSJEmSJKnHGPpJkiRJkiRJPcbQT5IkSZIkSeoxhn6SJEmSJElSjzH0kyRJkiRJknqMoZ8kSZIkSZLUYwz9JEmSJEmSpB7zjG4XoN71a7/2azVx4sRulyFJkiRJktQzFi9e/EBVTdhYP0M/bTUTJ06kv7+/22VIkiRJkiT1jCTf35R+bu+VJEmSJEmSeoyhnyRJkiRJktRjDP0kSZIkSZKkHuMz/aQtMHHGnG6XIEmSJEmSNtGqmdO6XcI250o/SZIkSZIkqccY+m1jSZ6X5Gub0O8vt0U9kiRJkiRJ6j2GfttYVf2oqqZvQldDP0mSJEmSJG2RURf6JXl7kmVJlia5uLXtk2Rea5+XZO/WPjvJ+UluTnJPkukDxnl/kuVtnJmt7ZQki1rb5Ul2SrJbklVJdmh9dkrygyTjk+yX5Joki5MsSLL/IPV+OMnFSb6V5O4kp7T2JDk3yYpWx/GtfWKSFe34xCRXtDnuTvKJ1j4TeFaSJUkuTbJzkjmt7hXrxxri+1uV5GNJvp2kP8khSeYm+V6S0wb0O7N9F8uSfGRA+5XtflcmOXVA+5okZ7cabkmyxxDzn9rm7V+9evUm/OKSJEmSJEkaaaMq9EtyAPAB4OiqOhh4Tzv1OeCiqjoIuBQ4f8BlewKvAH4fWB/uvRY4Dji8jfOJ1veKqjq0td0JnFxVDwJLgVe2Pn8AzK2qJ4BZwLurajJwBvD5IUo/CJgGTAE+lOR5wBuAScDBwDHAuUn2HOTaScDxwEuA45PsVVUzgEeqalJVnQC8BvhRVR1cVQcC1wz/TfKDqpoCLABmA9OB3wHOat/Pq4EXAIe1+ScnObJd+452v33A6Ume29p3Bm5p39184JTBJq6qWVXVV1V9EyZM2EiZkiRJkiRJ2hpGVegHHA18raoeAKiqn7T2KcCX2/HFdEK+9a6sqnVVdQewfvXZMcCXquoXG4xzYFuxtxw4ATigtV9GJ3gDeDNwWZJdgJcBX02yBPh7OgHjYL5eVY+0uq+jE6a9AvjHqlpbVf8F3AAcOsi186rqwap6FLgD2GeQPsuBY5Kck+SIFlQO56oB1y2sqoeqajXwaJLdgVe3f7cDtwH70wkBoRP0LQVuAfYa0P44cHU7XgxM3EgNkiRJkiRJ6pJndLuADQSoTeg3sM9jG1w/3DizgeOqammSE4Gprf0q4ONJngNMBr5FZ2Xbz6pq0mbWs/5zBus4iIH1r2WQ36SqvptkMvB7rc5rq+qsTRhz3Qbjr2vjB/h4Vf39wIuSTKUTmE6pql8kuR7YsZ1+oqrW3+egdUqSJEmSJGl0GG0r/eYBb1q/pbSFcAA301mBB50VejduZJxrgXck2WmDcXYF7ksyvo0DQFWtAW4FPgNc3Vbn/Ry4N8kb2xhJcvAQ870uyY6t7qnAIjpbYI9PMi7JBODINsemeqLVSdsu/IuqugT4JHDIZowzmLl0vp9d2vjPT/LrwG7AT1vgtz+dLcGSJEmSJEkaY0bVaq2qWpnkbOCGJGvpbD89ETgduDDJmcBq4KSNjHNNkklAf5LHgW/QeRvuB4GFwPfpbH3ddcBllwFf5anVf9AJBv82yf8FxgNfofP8vw3dCswB9gb+qqp+lOSf6WxLXkpn5d/7q+o/k0zcpC+j8zzBZUluAy6i80zAdcATwDs3cYxBVdW1SX4b+HYSgDXAH9F5VuBpSZYB36GzxVeDWDVzWrdLkCRJkiRJGlKe2rGpLZHkw8Caqvpkt2sZbfr6+qq/v7/bZUiSJEmSJPWMJIurqm9j/Ubb9l5JkiRJkiRJT9Oo2t47FlXVh7sxb9s+vO8GzX9eVXO7UY8kSZIkSZJGD0O/MaqqXt/tGiRJkiRJkjQ6ub1XkiRJkiRJ6jGGfpIkSZIkSVKPMfSTJEmSJEmSeoyhnyRJkiRJktRjfJGHtAUmzpjT7RIkSZIkST1q1cxp3S5BPcCVfpIkSZIkSVKPMfR7GpKcluTtIzTWX47EOFtbkklJfq/bdUiSJEmSJGlohn5bKMkzqurvquqiERpys0O/JONGaO7Bxh5q6/ckwNBPkiRJkiRpFNuuQ78kE5PcleQfkixL8rUkOyWZnOSGJIuTzE2yZ+t/fZKPJbkBeE+SDyc5Y8C585LMT3JnkkOTXJHk7iQfHTDnHyW5NcmSJH+fZFySmcCzWtulQ/Vr7WuSnJVkITBlkHs6LMkV7fh1SR5J8swkOya5p7VPSnJLu+d/TvKrQ9zfG5OsSLK03dczgbOA41tdxw8y/6lJ+pP0r169eiR/LkmSJEmSJG2i7Tr0a14EzKqqg4CfA+8CPgtMr6rJwIXA2QP6715Vr6yqvx5krMer6kjg74Cvt7EOBE5M8twkvw0cD7y8qiYBa4ETqmoG8EhVTaqqE4bq1+bYGVhRVYdX1Y2D1HAb8NJ2fASwAjgUOBxY2NovAv683fNy4P8NcX8fAv53VR0MHFtVj7e2y1qtl204eVXNqqq+quqbMGHCIOVJkiRJkiRpa/PtvfCDqrqpHV9CZ5vtgcA3kwCMA+4b0P+Xgq4Brmp/lwMrq+o+gLbCbi/gFcBkYFEb+1nA/YOM86ph+q0FLh+qgKp6Msm/t+DwMOBTwJHtPhYk2Y1OsHdDu+QfgK8OcX83AbOT/BNwxTD3LUmSJEmSpFHE0A9qg88P0QnsfmnrbPPwMGM91v6uG3C8/vMzgAD/UFV/sZGahuv3aFWt3cj1C4DXAk8A/wbMphP6nbGR62DA/VXVaUkOB6YBS5JM2oTrJUmSJEmS1GVu74W9k6wP+N4C3AJMWN+WZHySA0ZornnA9CS/3sZ+TpJ92rknkozfhH6bYj7wXuDbVbUaeC6wP50w80Hgp0mOaH3fBtww2CBJ9quqhVX1IeABOqsVHwJ23YxaJEmSJEmStI250g/uBP44yd8Dd9N5nt9c4Py2FfYZwKeBlU93oqq6I8n/Ba5NsgOdlXjvAr4PzAKWJbmtPddvqH6bYiGwB53wD2AZcH9VrV/V+MfA3yXZCbgHOGmIcc5N8gI6Kw/nAUuB/w+YkWQJ8PHBnuu3PVg1c1q3S5AkSZIkSRpSnsqBtj9JJgJXV9WBXS6lJ/X19VV/f3+3y5AkSZIkSeoZSRZXVd/G+rm9V5IkSZIkSeox2/X23qpaRedNvWNSkn8G9t2g+c+ram436pEkSZIkSdLosF2HfmNdVb2+2zVIkiRJkiRp9HF7ryRJkiRJktRjDP0kSZIkSZKkHmPoJ0mSJEmSJPUYQz9JkiRJkiSpx/giD2kLTJwxp9slSFJXrZo5rdslSJIkSRqGK/20yZLsnuRPu12HJEmSJEmShmfoN0okGdftGtZLMtQK0N0BQz9JkiRJkqRRztBvG0lyZZLFSVYmObW1rUlyVpKFwJQkk5Pc0PrNTbJn63dKkkVJlia5PMlOQ8wxLsk96dg9ybokR7ZzC5L8VpLntFqWJbklyUHt/IeTzEpyLXBRkgOS3JpkSev7AmAmsF9rO3dbfG+SJEmSJEnafD7Tb9t5R1X9JMmzgEVJLgd2BlZU1YeSjAduAF5XVauTHA+cDbwDuKKqLgBI8lHgZOCzG05QVWuTfBd4MbAvsBg4ooWKv1FV/57ks8DtVXVckqOBi4BJbYjJwCuq6pHW7zNVdWmSZwLjgBnAgVU1acO5JUmSJEmSNHoY+m07pyd5fTveC3gBsBa4vLW9CDgQ+GYS6IRs97VzB7awb3dgF2DuMPMsAI6kE/p9HDiFTpi4qJ1/BfCHAFX1rSTPTbJbO3dVVT3Sjr8NfCDJb9AJHe9udQ2rrWI8FWDvvffeaH9JkiRJkiSNPLf3bgNJpgLHAFOq6mDgdmBH4NGqWru+G7Cyqia1fy+pqle3c7OBP6uqlwAfadcOZQFwBHAY8A06QeFUYP6AeTZU7e/D/91Q9WXgWOARYG5bFbhRVTWrqvqqqm/ChAmbcokkSZIkSZJGmKHftrEb8NOq+kWS/YHfGaTPd4AJSaYAJBmf5IB2blfgvrYF+ISNzLUQeBmwrqoeBZYAf0InDIRO+HdCm2Mq8EBV/XzDQZL8JnBPVZ0PXAUcBDzUapEkSZIkSdIoZui3bVwDPCPJMuCvgFs27FBVjwPTgXOSLKUT1r2snf4gnTDvm8Bdw01UVY8BPxgwxwI6Qd3y9vnDQF+rZSbwx0MMdTywIskSYH/goqr6MXBTkhW+yEOSJEmSJGn0SlVtvJe0Bfr6+qq/v7/bZUiSJEmSJPWMJIurqm9j/VzpJ0mSJEmSJPUY3947RiX5APDGDZq/WlVnd6MeSZIkSZIkjR6GfmNUC/cM+CRJkiRJkvRL3N4rSZIkSZIk9RhDP0mSJEmSJKnHGPpJkiRJkiRJPcbQT5IkSZIkSeoxhn6SJEmSJElSj/HtvdIWmDhjTrdL0HZk1cxp3S5BkiRJkjTGuNJPkiRJkiRJ6jGGfqNIkrOSHNOO35tkp27XtKEkxyV5cbfrkCRJkiRJ0tAM/UaRqvpQVf1b+/heoGuhX5JxQ5w6DjD0kyRJkiRJGsUM/YAkb0+yLMnSJBcn2SfJvNY2L8nerd/sJOcnuTnJPUmmDxjj/UmWtzFmtrZTkixqbZcn2SnJbklWJdmh9dkpyQ+SjG/jT09yOvA84Lok1yU5Ocl5A+Y6JcmnhriX97frSXJekm+141cluaQdv6XVuiLJOQOuXdNWGy4EpiSZmeSO9j18MsnLgGOBc5MsSbLfiP4QkiRJkiRJGhHbfeiX5ADgA8DRVXUw8B7gc8BFVXUQcClw/oBL9gReAfw+sD7cey2dFXCHtzE+0fpeUVWHtrY7gZOr6kFgKfDK1ucPgLlV9cT6CarqfOBHwFFVdRTwFeDYJONbl5OALw1xS/OBI9pxH7BLu+4VwIIkzwPOAY4GJgGHJjmu9d8ZWFFVhwN3AK8HDmjfw0er6mbgKuDMqppUVd8b5Ps8NUl/kv7Vq1cPUaIkSZIkSZK2pu0+9KMTfn2tqh4AqKqfAFOAL7fzF9MJzNa7sqrWVdUdwB6t7RjgS1X1iwFjAByYZEGS5cAJwAGt/TLg+Hb85vZ5SFX1MPAt4PeT7A+Mr6rlQ3RfDExOsivwGPBtOuHfEcAC4FDg+qpaXVVP0gk1j2zXrgUub8c/Bx4FvpDkDcAvhqtxQK2zqqqvqvomTJiwKZdIkiRJkiRphBn6QYDaSJ+B5x/b4NrhxpgN/FlVvQT4CLBja78KeG2S5wCT6QR6G/MF4ESGX+VHWzG4qvW7mU7QdxSwH53VhhnqWuDRqlrbxnkSOIxOCHgccM0m1ChJkiRJkqRRwNAP5gFvSvJcgBbE3UxnBR50VujduJExrgXesf5tu20MgF2B+9r22hPWd66qNcCtwGeAq9cHbRt4qF2//pqFwF7AW4F/3Eg984Ez2t8FwGnAkqoqYCHwyiS/1l7W8Rbghg0HSLILsFtVfYPOS0UmDVaXJEmSJEmSRp9ndLuAbquqlUnOBm5Isha4HTgduDDJmcBqOqvmhhvjmiSTgP4kjwPfAP4S+CCdkO37wHL+Z1h2GfBVYOoQw84C/jXJfe25fgD/BEyqqp9u5LYW0HlO4ber6uEkj7Y2quq+JH8BXEdn1d83qurrg4yxK/D1JDu2fu9r7V8BLmgvC5k+2HP9tgerZk7rdgmSJEmSJElDSmfxl8aCJFcD51XVvG7Xsin6+vqqv7+/22VIkiRJkiT1jCSLq6pvY/3c3jsGJNk9yXeBR8ZK4CdJkiRJkqTu2e63944FVfUz4IUD29ozCAcLAF9VVT/eJoVJkiRJkiRpVDL0G6NasDdpox0lSZIkSZK03XF7ryRJkiRJktRjDP0kSZIkSZKkHmPoJ0mSJEmSJPUYQz9JkiRJkiSpx/giD2kLTJwxp9slaJRYNXNat0uQJEmSJOmXuNJPkiRJkiRJ6jGGfl2SZGqSq9vxsUlmdLumTdHqflm365AkSZIkSdLQ3N47wpIESFWt29Rrquoq4KqtV9XmSzKuqtYOcmoqsAa4edtWJEmSJEmSpE3lSr8RkGRikjuTfB64Dfhikv4kK5N8ZEC/1yS5K8mNwBsGtJ+Y5HPteHaS6QPOrWl/90wyP8mSJCuSHDFELW9K8ql2/J4k97Tj/dq8JHlVktuTLE9yYZJfae2rknyo9XtjktOT3JFkWZKvJJkInAa8r9UxaA2SJEmSJEnqLlf6jZwXASdV1Z8meU5V/STJOGBekoOA7wIXAEcD/w5ctpnjvxWYW1Vnt3F3GqLffODMdnwE8OMkzwdeASxIsiMwG3hVVX03yUXAO4FPt2serapXACT5EbBvVT2WZPeq+lmSvwPWVNUnB5s8yanAqQB77733Zt6iJEmSJEmSRoIr/UbO96vqlnb8piS3AbcDBwAvBvYH7q2qu6uqgEs2c/xFwElJPgy8pKoeGqxTVf0nsEuSXYG9gC8DR9IJABfQCSfvrarvtkv+oZ1fb2AYuQy4NMkfAU9uSpFVNauq+qqqb8KECZt8c5IkSZIkSRo5hn4j52GAJPsCZ9BZSXcQMAfYsfWpTRjnSdrv0p4P+EyAqppPJ5z7D+DiJG8fZoxvAycB36ET9B0BTAFuArIp99FMA/4GmAwsTuLKUEmSJEmSpDHA0G/kPZtOcPZgkj2A17b2u4B9k+zXPr9liOtX0QnZAF4HjAdIsg9wf1VdAHwROGSYGubTCR7n01lteBTwWFU92OqYmOS3Wt+3ATdsOECSHYC9quo64P3A7sAuwEPArsPMLUmSJEmSpC5z5dYIq6qlSW4HVgL30FldR1U92p53NyfJA8CNwIGDDHEB8PUktwLzeGrl3VTgzCRP0Hl77nAr/RbQ2do7v6rWJvkBnbBvfR0nAV9tK/cWAX83yBjjgEuS7EZndeB57Zl+/wJ8LcnrgHdX1YJN+2Z6y6qZ07pdgiRJkiRJ0pDSebycNPL6+vqqv7+/22VIkiRJkiT1jCSLq6pvY/3c3itJkiRJkiT1GLf3jmFJFgK/skHz26pqeTfqkSRJkiRJ0uhg6DeGVdXh3a5BkiRJkiRJo4/beyVJkiRJkqQeY+gnSZIkSZIk9RhDP0mSJEmSJKnHGPpJkiRJkiRJPcYXeUhbYOKMOd0uoeetmjmt2yVIkiRJkjRmudJPkiRJkiRJ6jGGfj0iybhu1yBJkiRJkqTRwdBvjEhyZZLFSVYmObW1rUlyVpKFwJQkk5Pc0PrNTbJn63dKkkVJlia5PMlOw8zzxiQrWt/5rW1cknPbGMuS/Mk2uWlJkiRJkiRtEUO/seMdVTUZ6ANOT/JcYGdgRVUdDiwEPgtMb/0uBM5u115RVYdW1cHAncDJw8zzIeB/t77HtraTgQer6lDgUOCUJPuO8P1JkiRJkiRphPgij7Hj9CSvb8d7AS8A1gKXt7YXAQcC30wCMA64r507MMlHgd2BXYC5w8xzEzA7yT8BV7S2VwMHJZnePu/W5r93w4vbKsRTAfbee+/NvEVJkiRJkiSNBEO/MSDJVOAYYEpV/SLJ9cCOwKNVtXZ9N2BlVU0ZZIjZwHFVtTTJicDUoeaqqtOSHA5MA5YkmdTGfndVDRcWrr9+FjALoK+vrzbpBiVJkiRJkjSi3N47NuwG/LQFfvsDvzNIn+8AE5JMAUgyPskB7dyuwH1JxgMnDDdRkv2qamFVfQh4gM6qwrnAO9v1JHlhkp1H5M4kSZIkSZI04lzpNzZcA5yWZBmdcO+WDTtU1eNt++35SXaj89t+GlgJfJDOM/++DyynEwIO5dwkL6Czum8esBRYBkwEbktn7/Bq4LiRuTVJkiRJkiSNtFS5A1NbR19fX/X393e7DEmSJEmSpJ6RZHFV9W2sn9t7JUmSJEmSpB7j9t7tVJIPAG/coPmrVXV2N+qRJEmSJEnSyDH02061cM+AT5IkSZIkqQe5vVeSJEmSJEnqMYZ+kiRJkiRJUo8x9JMkSZIkSZJ6jKGfJEmSJEmS1GMM/SRJkiRJkqQe49t7pS0wccacbpfQNatmTut2CZIkSZIkaSNc6bcZkuye5E+38NqJSd460jVtS71wD5IkSZIkSdsDQ7/NszuwRaEfMBHY7MAsybgtnG+LJRlqBehEtuAeJEmSJEmStG0Z+m2emcB+SZYkOTfJmUkWJVmW5CMASQ5tn3dMsnOSlUkObNce0a59X5ITk3xu/cBJrk4ytR2vSXJWkoXAlCSTk9yQZHGSuUn2HKy4JL+eZHE7PjhJJdm7ff5ekp2S7JNkXqtx3oDzs5N8Ksl1wDlJXtlqXZLk9iS7bngPW+tLliRJkiRJ0tNj6Ld5ZgDfq6pJwDeBFwCHAZOAyUmOrKpFwFXAR4FPAJdU1Yp27YKqmlRV521knp2BFVV1OLAQ+CwwvaomAxcCZw92UVXdD+yY5NnAEUA/nZBuH+D+qvoF8Dngoqo6CLgUOH/AEC8Ejqmq/wOcAbyr3esRwCObeQ+SJEmSJEnqEl/kseVe3f7d3j7vQicEnA+cBSwCHgVO34Kx1wKXt+MXAQcC30wCMA64b5hrbwZeDhwJfAx4DRBgQTs/BXhDO76YTjC53leram07vgn4VJJLgSuq6odt/mElORU4FWDvvffeaH9JkiRJkiSNPEO/LRfg41X194Ocew6dEHA8sCPw8CB9nuR/rrTcccDxowPCtwArq2rKJta1gM7KvH2ArwN/DhRw9RD9a8Dxf9dZVTOTzAF+D7glyTGbMnlVzQJmAfT19dVGukuSJEmSJGkrcHvv5nkI2LUdzwXekWQXgCTPT/Lr7dws4IN0ts+eM8i1AKuASUl2SLIXnW3Cg/kOMCHJlDbP+CQHDFPjfOCPgLurah3wEzrB3U3t/M3Am9vxCcCNgw2SZL+qWl5V59DZJrz/IPcgSZIkSZKkUciVfpuhqn6c5KYkK4B/Bb4MfLtte10D/FGS1wBPVtWX25t3b05yNJ0VeE8mWQrMBj4N3AssB1YAtw0x5+NJpgPnJ9mNzm/2aWDlEP1XtXrmt6Ybgd+oqp+2z6cDFyY5E1gNnDTE7b43yVF0thrf0e533cB78Ll+kiRJkiRJo1Oq3IGpraOvr6/6+/u7XcZWMXHGnG6X0DWrZk7rdgmSJEmSJG23kiyuqr6N9XOln7QFDL4kSZIkSdJoZug3RiX5Gzpv6R3oM1X1pW7UI0mSJEmSpNHD0G+Mqqp3dbsGSZIkSZIkjU6+vVeSJEmSJEnqMYZ+kiRJkiRJUo8x9JMkSZIkSZJ6jKGfJEmSJEmS1GMM/SRJkiRJkqQeY+gnSZIkSZIk9ZhndLsAaSyaOGNOt0vYqlbNnNbtEiRJkiRJ0tPgSj9JkiRJkiSpxxj6jVFJpia5uh0fm2RGt2uSJEmSJEnS6OD23lEmSYBU1bpNvaaqrgKu2npVSZIkSZIkaSxxpd8okGRikjuTfB64Dfhikv4kK5N8ZEC/1yS5K8mNwBsGtJ+Y5HPteHaS6QPOrWl/90wyP8mSJCuSHDFELePaGCuSLE/yvta+X5JrkixOsiDJ/kNcf2qrvX/16tUj8O1IkiRJkiRpc7nSb/R4EXBSVf1pkudU1U+SjAPmJTkI+C5wAXA08O/AZZs5/luBuVV1dht3pyH6TQKeX1UHAiTZvbXPAk6rqruTHA58vtXyP1TVrNaXvr6+2swaJUmSJEmSNAIM/UaP71fVLe34TUlOpfP77Am8mM6qzHur6m6AJJcAp27G+IuAC5OMB66sqiVD9LsH+M0knwXmANcm2QV4GfDVzu5jAH5lM+aWJEmSJEnSNuT23tHjYYAk+wJnAK+qqoPoBG87tj6bsnLuSdrv2p4P+EyAqpoPHAn8B3BxkrcPdnFV/RQ4GLgeeBfwhTbez6pq0oB/v70lNylJkiRJkqStz9Bv9Hk2nQDwwSR7AK9t7XcB+ybZr31+yxDXrwImt+PXAeMBkuwD3F9VFwBfBA4Z7OIkvwbsUFWXAx8EDqmqnwP3Jnlj65MkB2/5LUqSJEmSJGlrcnvvKFNVS5PcDqyks9X2ptb+aNvyOyfJA8CNwIGDDHEB8PUktwLzaCsIganAmUmeANYAg670A54PfCnJ+kD4L9rfE4C/TfJ/6QSJXwGWbvGNjnGrZk7rdgmSJEmSJElDSpXvWtDW0dfXV/39/d0uQ5IkSZIkqWckWVxVfRvr5/ZeSZIkSZIkqce4vXc7lmQhv/wW3rdV1fJu1CNJkiRJkqSRYei3Hauqw7tdgyRJkiRJkkae23slSZIkSZKkHmPoJ0mSJEmSJPUYQz9JkiRJkiSpxxj6SZIkSZIkST3GF3lIW2DijDndLuFpWTVzWrdLkCRJkiRJW5Er/SRJkiRJkqQeY+g3SiQ5K8kx7fi9SXbqdk2SJEmSJEkamwz9Romq+lBV/Vv7+F7A0E+SJEmSJElbZLsP/ZK8PcmyJEuTXJxknyTzWtu8JHu3frOTnJ/k5iT3JJk+YIz3J1nexpjZ2k5Jsqi1XZ5kpyS7JVmVZIfWZ6ckP0gyvo0/PcnpwPOA65Jcl+TkJOcNmOuUJJ8a4l4mJrkryReSrEhyaZJjktyU5O4kh7V+Oye5sNV3e5LXDbh+QZLb2r+XtfapSa5P8rU2/qVJMkQNpybpT9K/evXqkfiJJEmSJEmStJm269AvyQHAB4Cjq+pg4D3A54CLquog4FLg/AGX7Am8Avh9YH2491rgOODwNsYnWt8rqurQ1nYncHJVPQgsBV7Z+vwBMLeqnlg/QVWdD/wIOKqqjgK+AhybZHzrchLwpWH6rn//AAAgAElEQVRu67eAzwAHAfsDb201nwH8ZevzAeBbVXUocBRwbpKdgfuB362qQ4DjN7j3l9JZgfhi4DeBlw82eVXNqqq+quqbMGHCMGVKkiRJkiRpa9muQz/gaOBrVfUAQFX9BJgCfLmdv5hOYLbelVW1rqruAPZobccAX6qqXwwYA+DAtmpuOXACcEBrv4xOoAbw5vZ5SFX1MPAt4PeT7A+Mr6rlw1xyb1Utr6p1wEpgXlUVsByY2Pq8GpiRZAlwPbAjsDcwHrig1fxVOgHferdW1Q/buEsGjCVJkiRJkqRR5hndLqDLAtRG+gw8/9gG1w43xmzguKpamuREYGprvwr4eJLnAJPpBHob8wU6q/TuYvhVfhvWuG7A53U89XsH+MOq+s7AC5N8GPgv4GA6gfCjQ4y7Fv/vSJIkSZIkjVrb+0q/ecCbkjwXoAVxN9NZgQedFXo3bmSMa4F3rH/bbhsDYFfgvrYt94T1natqDXArnS24V1fV2kHGfKhdv/6ahcBedLbq/uPm3OAQ5gLvXv9cviQvbe27Afe11XxvA8aNwFySJEmSJEnaxrbr1VpVtTLJ2cANSdYCtwOnAxcmORNYTecZesONcU2SSUB/kseBb9BZlfdBYCHwfTpba3cdcNlldLbPTh1i2FnAvya5rz3XD+CfgElV9dPNv9Nf8lfAp4FlLfhbRec5hZ8HLk/yRuA64OERmKsnrZo5rdslSJIkSZIkDSmdx71ptEtyNXBeVc3rdi2bqq+vr/r7+7tdhiRJkiRJUs9Isriq+jbWb3vf3jvqJdk9yXeBR8ZS4CdJkiRJkqTu2a63944FVfUz4IUD29ozCAcLAF9VVT/eJoVJkiRJkiRp1DL0G4NasDep23VIkiRJkiRpdHJ7ryRJkiRJktRjDP0kSZIkSZKkHmPoJ0mSJEmSJPUYQz9JkiRJkiSpx/giD2kLTJwxp9slbJJVM6d1uwRJkiRJktQFrvSTJEmSJEmSeoyhXxckmZrk6nZ8bJIZ3a5JkiRJkiRJvcPtvSMoSYBU1bpNvaaqrgKu2npVSZIkSZIkaXvjSr+nKcnEJHcm+TxwG/DFJP1JVib5yIB+r0lyV5IbgTcMaD8xyefa8ewk0wecW9P+7plkfpIlSVYkOWKYetYkOSfJ4iT/luSwJNcnuSfJsa3PuCTnJlmUZFmSP2ntuySZl+S2JMuTvG6De7yg3de1SZ41xPyntvvvX7169dP4ZiVJkiRJkrSlDP1GxouAi6rqpcD/qao+4CDglUkOSrIjcAHwB8ARwP/azPHfCsytqknAwcCSYfruDFxfVZOBh4CPAr8LvB44q/U5GXiwqg4FDgVOSbIv8Cjw+qo6BDgK+Ou2ehHgBcDfVNUBwM+APxxs8qqaVVV9VdU3YcKEzbxNSZIkSZIkjQS3946M71fVLe34TUlOpfPd7gm8mE64em9V3Q2Q5BLg1M0YfxFwYZLxwJVVNVzo9zhwTTteDjxWVU8kWQ5MbO2vBg4asKpwNzqh3g+BjyU5ElgHPB/Yo/W5d8C8iweMJUmSJEmSpFHGlX4j42GAtlruDOBVVXUQMAfYsfWpTRjnSdpv0lbYPROgquYDRwL/AVyc5O3DjPFEVa2fax3wWBtjHU+FvAHeXVWT2r99q+pa4ARgAjC5rSr8rwH1PzZgjrUYGEuSJEmSJI1ahn4j69l0AsAHk+wBvLa13wXsm2S/9vktQ1y/Cpjcjl8HjAdIsg9wf1VdAHwROORp1jkXeGdbOUiSFybZmc6Kv/vbysCjgH2e5jySJEmSJEnqAldrjaCqWprkdmAlcA9wU2t/tG35nZPkAeBG4MBBhrgA+HqSW4F5tBWEwFTgzCRPAGuA4Vb6bYov0Nmee1tbUbgaOA64FPiXJP10nht419Ocp2etmjmt2yVIkiRJkiQNKU/tBJVGVl9fX/X393e7DEmSJEmSpJ6RZHF7ieyw3N4rSZIkSZIk9Ri3945RSRYCv7JB89uqank36pEkSZIkSdLoYeg3RlXV4d2uQZIkSZIkSaOT23slSZIkSZKkHmPoJ0mSJEmSJPUYQz9JkiRJkiSpxxj6SZIkSZIkST3GF3lIW2DijDndLuG/rZo5rdslSJIkSZKkUcaVfiMoye5J/rTbdUiSJEmSJGn7Zug3snYHDP0kSZIkSZLUVWM29Evy9iTLkixNcnGSfZLMa23zkuzd+s1O8rdJrktyT5JXJrkwyZ1JZg8Yb02Sv05yW7t+Qms/JcmiNs/lSXZq7Xsk+efWvjTJy4CZwH5JliQ5N8nUJNcn+VqSu5JcmiTt+slJbkiyOMncJHu29tOT3NHu4yut7ZVtzCVJbk+y6xDfydQ25j8l+W6SmUlOSHJrkuVJ9mv9JrR7WdT+vby1H5bk5jbHzUle1NpPTHJFkmuS3J3kE1vlR5UkSZIkSdKIGJOhX5IDgA8AR1fVwcB7gM8BF1XVQcClwPkDLvlV4GjgfcC/AOcBBwAvSTKp9dkZuK2qDgFuAP5fa7+iqg5t89wJnNzazwduaO2HACuBGcD3qmpSVZ3Z+r0UeC/wYuA3gZcnGQ98FpheVZOBC4GzW/8ZwEvbfZzW2s4A3lVVk4AjgEeG+XrWfx8vAd4GvLCqDgO+ALy79fkMcF5VHQr8YTsHcBdwZFW9FPgQ8LEB404Cjm/jHp9kr8EmT3Jqkv4k/atXrx6mTEmSJEmSJG0tY/VFHkcDX6uqBwCq6idJpgBvaOcvBgauRvuXqqoky4H/qqrlAElWAhOBJcA64LLW/xLginZ8YJKP0tm6uwswd0ANb2/zrwUeTPKrg9R6a1X9sM23pM33M+BA4Jtt4d844L7WfxlwaZIrgStb203Ap5JcSieE/OEw382iqrqvzfc94NrWvhw4qh0fA7y4zQ3w7LZ6cDfgH5K8AChg/IBx51XVg23cO4B9gB9sOHlVzQJmAfT19dUwdUqSJEmSJGkrGauhX+iEUsMZeP6x9nfdgOP1n4f6DtZfPxs4rqqWJjkRmLo5hW4w39o2X4CVVTVlkP7TgCOBY4EPJjmgqmYmmQP8HnBLkmOq6q5NmG/g/Q681x2AKVX1P1YMJvkscF1VvT7JROD6jdyHJEmSJEmSRqExub0XmAe8KclzAZI8B7gZeHM7fwJw42aOuQMwvR2/dcD1uwL3tS25J2xQwzvb/OOSPBt4qPXfmO8AE9rqRJKMT3JAkh2AvarqOuD9tNWFSfarquVVdQ7QD+y/mfe2oWuBP1v/YcAW592A/2jHJz7NOSRJkiRJktQlYzL0q6qVdJ6Bd0OSpcCngNOBk5Iso/Msu/ds5rAPAwckWUxn6+5Zrf2DwELgm3Seebfee4Cj2pbhxcABVfVj4KYkK5KcO0z9j9MJGM9p9S8BXkZnm+8lbczb6Tx372fAe9uYS+k8z+9fN/PeNnQ60NdeFnIHTz078BPAx5Pc1GqRJEmSJEnSGJQqH7sGnbf3VtUu3a6jl/T19VV/f3+3y5AkSZIkSeoZSRZXVd/G+o3JlX6SJEmSJEmShubLGJqxtMovyUvovKF4oMeq6vBu1CNJkiRJkqTRxdBvDKqq5cCkjXaUJEmSJEnSdsntvZIkSZIkSVKPMfSTJEmSJEmSeoyhnyRJkiRJktRjDP0kSZIkSZKkHmPoJ0mSJEmSJPUY394rbYGJM+Zsk3lWzZy2TeaRJEmSJEm9xZV+kiRJkiRJUo8x9NsCSSYmeeuAzycm+Vw3a5IkSZIkSZLWM/TbMhOBt26skyRJkiRJktQNPRX6Jdk5yZwkS5OsSHJ8klVJPpbk20n6kxySZG6S7yU5rV2XJOe2a5YnOX64dmAmcESSJUne19qel+SaJHcn+cSAmtYkObvVdEuSPVr7hCSXJ1nU/r28tb+yjbskye1Jdk2yZ5L5rW1FkiOG+Q7WJDknyeIk/5bksCTXJ7knybGtz7h2X4uSLEvyJ619lyTzktzW7vd1rX1ikjuTXJBkZZJrkzxrRH88SZIkSZIkjZieCv2A1wA/qqqDq+pA4JrW/oOqmgIsAGYD04HfAc5q598ATAIOBo4Bzk2y5zDtM4AFVTWpqs5rY0wCjgdeAhyfZK/WvjNwS1UdDMwHTmntnwHOq6pDgT8EvtDazwDeVVWTgCOAR+isKpzb2g4GlgzzHewMXF9Vk4GHgI8Cvwu8fsD9ngw82OY+FDglyb7Ao8Drq+oQ4Cjgr5OkXfMC4G+q6gDgZ63mX5Lk1Bau9q9evXqYMiVJkiRJkrS19Nrbe5cDn0xyDnB1VS1omdVVA87vUlUPAQ8leTTJ7sArgH+sqrXAfyW5gU4YNlT7zweZe15VPQiQ5A5gH+AHwOPA1a3PYjoBHHRCxBc/lanx7CS7AjcBn0pyKXBFVf0wySLgwiTjgSurarjQ73GeCjuXA49V1RNJltPZlgzwauCgJNPb593ohHo/BD6W5EhgHfB8YI/W594B8y4eMNb/UFWzgFkAfX19NUydkiRJkiRJ2kp6KvSrqu8mmQz8HvDxJNe2U4+1v+sGHK///AwgDG6o9sEMHHctT323T1RVDdK+AzClqh7ZYJyZSea0e7glyTFVNb8FcdOAi5OcW1UXDVHHwPn++36ral2S9XMHeHdVzR14YZITgQnA5BYUrgJ2HOL+3N4rSZIkSZI0SvXU9t4kzwN+UVWXAJ8EDtnES+fT2ZI7LskE4Ejg1mHaHwJ2fZrlXgv82YDaJ7W/+1XV8qo6B+gH9k+yD3B/VV0AfHEz7msoc4F3tpWDJHlhkp3prPi7vwV+R9FZrShJkiRJkqQxpqdW+tF5nt65SdYBTwDvBL62Cdf9MzAFWAoU8P6q+s8kQ7X/GHgyyVI6zwj86RbUejrwN0mW0fkd5gOnAe9tgdta4A7gX4E3A2cmeQJYA7x9C+Yb6At0tufe1p7Ztxo4DrgU+Jck/XSeG3jX05ynZ62aOa3bJUiSJEmSJA0pT+0ElUZWX19f9ff3d7sMSZIkSZKknpFkcVX1baxfT23vlSRJkiRJktR723u3G0kWAr+yQfPbqmp5N+qRJEmSJEnS6GHoN0ZV1eHdrkGSJEmSJEmjk9t7JUmSJEmSpB5j6CdJkiRJkiT1GEM/SZIkSZIkqccY+kmSJEmSJEk9xhd5SFtg4ow5W3X8VTOnbdXxJUmSJElSb3OlnyRJkiRJktRjDP1GgSTjul2DJEmSJEmSeoeh3zaQ5Moki5OsTHJqa1uT5KwkC4EpSSYnuaH1m5tkz9bvlCSLkixNcnmSnYaZZ3aSv/3/2bvzaD2r+u7/7w+ByhAkYqMLfQhRWkUGCeQgDYMgoktL69BGsYIUpfJDURwKGh+wLRY0VCsKqBAtRgaHxSCCPDKUB0hEEE4gZADFivHBQjUOIKAESL6/P+595CaeieQcTnLyfq111r3vfe3he133+eu79r52kmuT3J1kvyRnJ7kzydyudq9OcmOSW5NckGRiq/+nNteSJHOSpNVfl+SUJDcnuSvJvqP5vCRJkiRJkrR2TPo9Pd5RVdOBHuCYJM8GtgCWVNWewPeB04GZrd3ZwMmt78VVtUdV7QrcCRwxxFzPAg4APgBcBpwK7ATskmRakj8FTgAOrKrdgV7gg63vGW2unYHNgL/qGnfjqnoZ8H7gn9f4SUiSJEmSJGnUeZDH0+OYJG9s5W2BPwdWAhe1uhcDOwNXt8V1E4D72rWdk5wETAImAlcOMddlVVVJFgM/r6rFAEmWAlOB/wXsCNzQ5voT4MbW9xVJPgRsDmwNLKWTOAS4uH0uaOP0q61kPBJgypQpQ4QqSZIkSZKk0WDSb5Ql2R84EJhRVb9Lch2wKfBIVa3sawYsraoZ/QwxF3hDVd2e5HBg/yGmXNE+V3WV+75vTCfZeHVV/d1qcW4KfB7oqap7kvxLi3P1cVcyyP9NVc0B5gD09PTUELFKkiRJkiRpFLi9d/RtBfymJfx2AP6inzY/BCYnmQGQZJMkO7VrWwL3JdkEOGQE4rkJ2DvJn7W5Nk/yIp5I8P2yveNv5gjMJUmSJEmSpDHgSr/RdwVwVJJFdJJ7N63eoKoeTTITOC3JVnR+l8/Q2V77UTrv/PspsJhOEnCNVdXytmLwa0me0apPqKq7knyxzbEMuGVt5pEkSZIkSdLYSZU7MDU6enp6qre3d6zDkCRJkiRJGjeSLKiqnqHaub1XkiRJkiRJGmfc3rseSnI88KbVqi+oqpPHIh5JkiRJkiStW0z6rYdacs8EnyRJkiRJkvrl9l5JkiRJkiRpnDHpJ0mSJEmSJI0zJv0kSZIkSZKkccaknyRJkiRJkjTOmPSTJEmSJEmSxhlP75XWwNRZl69Rv2WzDxrhSCRJkiRJkv6YK/0kSZIkSZKkccak3xhKclSSw0ZorP89EuNIkiRJkiRp/WfSb4wk2biqzqyqc0ZoyKec9EsyYYTmliRJkiRJ0jrEpN9aSDI1yQ+SfCXJoiQXJtk8yfQk1ydZkOTKJNu09tcl+XiS64H3JfmXJMd2XTs1ybwkdybZI8nFSX6U5KSuOQ9NcnOShUnOSjIhyWxgs1Z3/kDtWv1DST6W5PvAjAHua3aSO9o9farVTU5yUZJb2t/eo/pwJUmSJEmStMZM+q29FwNzquqlwG+Bo4HTgZlVNR04Gzi5q/2kqtqvqv69n7EeraqXA2cC32pj7QwcnuTZSV4CHAzsXVXTgJXAIVU1C/h9VU2rqkMGatfm2AJYUlV7VtV3Vw8gydbAG4Gd2j31JRw/C5xaVXsAfwt8qb+HkeTIJL1JepcvXz7005MkSZIkSdKI8/TetXdPVd3QyufR2Wa7M3B1EoAJwH1d7b8xyFiXts/FwNKqug8gyd3AtsA+wHTgljb2ZsAv+hnnlYO0WwlcNEgMvwUeAb6U5HLg263+QGDHNh7AM5NsWVUPdneuqjnAHICenp4aZB5JkiRJkiSNEpN+a2/1xNaDdBJ2/W6dBR4eZKwV7XNVV7nv+8ZAgK9U1UeGiGmwdo9U1cqBOlbV40leRidx+BbgPcABdFaFzqiq3w8xtyRJkiRJksaY23vX3pQkfQm+vwNuAib31SXZJMlOIzTXNcDMJM9pY2+dZLt27bEkmwyj3aCSTAS2qqr/A7wfmNYuXUUnAdjXblo/3SVJkiRJkrQOcKXf2rsT+PskZwE/ovM+vyuB05JsRecZfwZYurYTVdUdSU4ArkqyEfAYnff+/ZTOltpFSW5t7/UbqN1QtgS+lWRTOisGP9DqjwE+l2RRu6d5wFFre0/rq2WzDxrrECRJkiRJkgaUKl+7tqaSTAW+XVU7j3Eo66Senp7q7e0d6zAkSZIkSZLGjSQLqqpnqHZu75UkSZIkSZLGGbf3roWqWkbnpN71UpJvAi9YrfrDVXXlWMQjSZIkSZKkkWHSbwNWVW8c6xgkSZIkSZI08tzeK0mSJEmSJI0zJv0kSZIkSZKkcWZYSb8kL0pyTZIl7ftLk5wwuqFJkiRJkiRJWhPDXen3ReAjwGMAVbUIeMtoBSVJkiRJkiRpzQ33II/Nq+rmJN11j49CPNJ6Yeqsy59S+2WzDxqlSCRJkiRJkv7YcFf6/TLJ9kABJJkJ3DdqUa3HkkxK8u6xjmO0JDk8yfPGOg5JkiRJkiQNbLhJv6OBs4Adkvw38H7gqFGLav02CVivk37pGOh/43DApJ8kSZIkSdI6bMikX0v+9FTVgcBkYIeq2qeqfjrq0Q0hyWFJFiW5Pcm5SbZrB44sap9TWru5Sb6Q5NokdyfZL8nZSe5MMrdrvIeS/HuSW1v/ya3+nUluafNclGTzVv/cJN9s9bcn2QuYDWyfZGGSTybZP8l1SS5M8oMk56ftk04yPcn1SRYkuTLJNq3+mCR3tPv4eqvbr425MMltSbYc4Jl8PsnrWvmbSc5u5SOSnNTKH0yypP29v9VNbc/j88CtwLbtuS1JsjjJB9oKzx7g/BbHZiP8k0qSJEmSJGkEDJn0q6pVwHta+eGqenDUoxqGJDsBxwMHVNWuwPuAM4BzquqlwPnAaV1dngUcAHwAuAw4FdgJ2CXJtNZmC+DWqtoduB7451Z/cVXt0ea5Ezii1Z8GXN/qdweWArOAH1fVtKo6rrXbjc7qyB2BFwJ7J9kEOB2YWVXTgbOBk1v7WcBu7T76VlQeCxxdVdOAfYHfD/Bo5rXrAM9vcwLsA8xPMh14O7An8BfAO5Ps1tq8uD2/3YA/BZ5fVTtX1S7Al6vqQqAXOKTd30AxSJIkSZIkaQwNd3vv1UmOTbJtkq37/kY1sqEdAFxYVb8EqKpfAzOAr7br59JJdPW5rKoKWAz8vKoWt4TmUmBqa7MK+EYrn9fVf+ck85MsBg6hkyzsi+ELbf6VVfXAALHeXFU/a/MtbPO9GNiZzrNdCJwA/K/WfhGd1XSH8sSBKTcAn05yDDCpqgY6SGU+sG+SHYE7gJ+3FYQzgO+1e/pmS+A+BFzME0nCn1bVTa18N/DCJKcneQ3w2wHme5IkRybpTdK7fPny4XSRJEmSJEnSCBtu0u8ddN7rNw9Y0P56RyuoYQrtYJFBdF9f0T5XdZX7vg90inFf/7nAe9qKtxOBTZ9SpE+eb2WbL8DStmJuWlXtUlWvbm0OAj4HTAcWJNm4qmYD/wBsBtyUZId+A676bzqrGl9D5/eaD7wZeKit0kx//ZqHu8b5DbArcB2d3/5Lw7nRqppTVT1V1TN58uThdJEkSZIkSdIIG1bSr6pe0M/fC0c7uCFcA7w5ybMB2srD7wFvadcPAb77FMfcCJjZym/t6r8lcF/bknvIajG8q80/IckzgQdb+6H8EJicZEbrv0mSndo7FLetqmuBD9E5GGRiku3b6sRT6CRc+036NTfS2U7cl/Q7tn3S6t6QZPMkWwBv7Lr2B0n+FNioqi4CPkpn+zJP4f4kSZIkSZI0RgZa4fYkSQ7rr76qzhnZcIavqpYmORm4PslK4DbgGODsJMcBy+m8u+6peBjYKckC4AHg4Fb/UeD7wE/pbA/uS3q9D5iT5Ag6K/jeVVU3JrkhyRLgO8DlA8T/aDsY47QkW9H5LT4D3AWc1+oCnFpV9yf51ySvaPPc0cYeyHzg1VX1X0l+Cmzd6qiqW9vhJTe3tl+qqtuSTF1tjOcDX84Tp/h+pH3OBc5M8ntghu/1kyRJkiRJWvek85q7IRolp3d93RR4JZ0DL2YO0GW9lOShqpo41nGMFz09PdXbO9a7wCVJkiRJksaPJAuqqmeodsNa6VdV711t8K3oHJQhSZIkSZIkaR0zrKRfP34H/PlIBrIuWJ9W+SXZhT9OvK6oqj3HIh5JkiRJkiStO4b7Tr/LeOIk242AHYELRisoDa2qFgPTxjoOSZIkSZIkrXuGu9LvU13lx4GfVtXPRiEeSZIkSZIkSWtpo6GbAPCXVXV9+7uhqn6W5JRRjUySJEmSJEnSGhlu0u9V/dS9diQDkSRJkiRJkjQyBt3em+RdwLuBFyZZ1HVpS+CG0QxMkiRJkiRJ0poZ6p1+XwW+A3wCmNVV/2BV/XrUopIkSZIkSZK0xgZN+lXVA8ADwN8BJHkOsCkwMcnEqvp/ox+iJEmSJEmSpKdiWKf3Jvlr4NPA84BfANsBdwI7jV5o0rpr6qzLn1L7ZbMPGqVIJEmSJEmS/thwD/I4CfgL4K6qegHwSjbwd/olmZrkrV3fD09yxljG9HRIsn+SvcY6DkmSJEmSJA1suEm/x6rqV8BGSTaqqmuBaaMY1/pgKvDWoRqtr5JMGODS/oBJP0mSJEmSpHXYcJN+9yeZCMwHzk/yWeDx0QtrzSXZIsnlSW5PsiTJwUmWJfl4khuT9CbZPcmVSX6c5KjWL0k+2fosTnLwYPXAbGDfJAuTfKDVPS/JFUl+lOTfumJ6KMnJLaabkjy31U9OclGSW9rf3q1+vzbuwiS3JdkyyTZJ5rW6JUn2HeD+35zk0638viR3t/L2Sb7byq9s4y5OcnaSZ7T6ZUn+qbV7U5JjktyRZFGSryeZChwFfKDF0W8MkiRJkiRJGlvDeqcf8Hrg98D7gUOArYCPjVZQa+k1wL1VdRBAkq2AU4B7qmpGklOBucDedA4lWQqcCfwNndWLuwJ/CtySZB6dVW391c8Cjq2qv2rzHN7a7QasAH6Y5PSqugfYAripqo5vycB30tky/Vng1Kr6bpIpwJXAS4BjgaOr6oaWbH0EOBK4sqpObqvwNh/g/ucBx7XyvsCvkjwf2AeYn2TTdv+vrKq7kpwDvAv4TOvzSFXt0+7pXuAFVbUiyaSquj/JmcBDVfWp/iZPcmSLlSlTpgwQoiRJkiRJkkbTsFb6VdXDwLbA/lX1FeBLwKOjGdhaWAwcmOSUJPu2E4gBLu26/v2qerCqlgOPJJlEJyn2tapaWVU/B64H9hikvj/XVNUDVfUIcAedA0+g86y+3coL6GwNBjgQOCPJwhbfM5NsSed9iZ9OcgwwqaoeB24B3p7kX4BdqurB/gKoqv+hc7rylnR+s68CL6eTAJwPvBj4SVXd1bp8pV3v842u8iI6KzsPZZgrO6tqTlX1VFXP5MmTh9NFkiRJkiRJI2xYSb8k7wQuBM5qVc8HLhmtoNZGS2ZNp5Pc+0SSf2qXVrTPVV3lvu8bAxlgyIHq+9M97kqeWEn5WFVVP/UbATOqalr7e35LRs4G/gHYDLgpyQ5VNY9Ocu6/gXOTHDZIHDcCbwd+SCfRty8wg04ycaj7ebirfBDwOTrPc0GS4a4MlSRJkiRJ0hga7jv9jqazHfa3AFX1I+A5oxXU2kjyPOB3VXUe8Clg92F2nQccnGRCksl0Emw3D1L/ILDlWoZ7FfCertintc/tq2pxVZ0C9AI7JNkO+EVVfRH4jyHuax6dLcLzgNuAVwAr2qrHHwBTk/xZa/s2OqsXnyTJRsC27dCWDwGTgImMzH1LkiRJkiRpFA135daKqno06SwSayu+avAuY2YX4JNJVgGP0abqfkwAACAASURBVHlf3YXD6PdNOqvhbqdzbx+qqv9JMlD9r4DHk9xO5x15v1mDWI8BPpdkEZ3fYh6dgzLen+QVdFYF3gF8B3gLcFySx4CHgMFW+s2ns7V3XlWtTHIPnWQfVfVIkrcDF7Tf8RY67zRc3QTgvPZOxNB59+D9SS4DLkzyeuC9VTV/De5bkiRJkiRJoyhP7DodpFHn8In76SSa3gu8G7ijqo4f3fC0Puvp6ane3t6xDkOSJEmSJGncSLKgqnqGajfc7b2zgOV03pP3/wH/BzhhzcOTJEmSJEmSNFoG3d6bZEpV/b+qWgV8sf1pHZHk+8AzVqt+W1UtHot4JEmSJEmStG4Y6p1+l9AOjEhyUVX97eiHpOGqqj3HOgZJkiRJkiSte4ba3puu8gtHMxBJkiRJkiRJI2OopF8NUJYkSZIkSZK0jhpqe++uSX5LZ8XfZq1M+15V9cxRjU6SJEmSJEnSUzZo0q+qJjxdgUiSJEmSJEkaGUNt75UkSZIkSZK0nhlqe6+eBkneANxVVXeMdSwanqmzLh/0+rLZBz1NkUiSJEmSJP0xV/qtG94A7NjfhSQmZiVJkiRJkvSUbJBJvyRbJLk8ye1JliQ5OMk3u66/KsnFrfxQklOSLEjyn0leluS6JHcneV1rc3iSS5JcluQnSd6T5INJbktyU5KtW7vtk1zRxpqfZIckewGvAz6ZZGFrc12Sjye5Hji+jblJG+OZSZb1fe/n3q5LcmqSeUnuTLJHkouT/CjJSV3tDk1yc5vzrCQTWv0XkvQmWZrkxK72y5KcmOTWJIuT7DDiP4wkSZIkSZJGxAaZ9ANeA9xbVbtW1c7AFcBLkkxu198OfLmVtwCuq6rpwIPAScCrgDcCH+sac2fgrcDLgJOB31XVbsCNwGGtzRzgvW2sY4HPV9X3gEuB46pqWlX9uLWdVFX7VdWJwHVA337RtwAXVdVjg9zfo1X1cuBM4FvA0S2+w5M8O8lLgIOBvatqGrASOKT1Pb6qeoCXAvsleWnXuL+sqt2BL7T4JUmSJEmStA7aUJN+i4ED2wq+favqAeBc4NAkk4AZwHda20fpJAX7+l3fEm6LgaldY15bVQ9W1XLgAeCyrj5Tk0wE9gIuSLIQOAvYZpAYv9FV/hKdRCQ8OSE5kEu75l5aVfdV1QrgbmBb4JXAdOCWFssrgRe2Pm9OcitwG7ATT952fHH7XLDavf9BkiPbSsHe5cuXDxGmJEmSJEmSRsMG+b64qroryXTgL4FPJLmKTmLtMuAR4IKqerw1f6yqqpVXASvaGKtWe9/eiq7yqq7vq+g8542A+9vKuuF4uCveG5JMTbIfMKGqlgzRt3vu1ePaGAjwlar6SHenJC+gs4Jvj6r6TZK5wKb9jLuSAf53qmoOnRWN9PT0VH9tJEmSJEmSNLo2yJV+SZ5HZ/vtecCngN2r6l7gXuAEYO5Iz1lVvwV+kuRNLYYk2bVdfhDYcoghzgG+xtCr/IbjGmBmkue0WLZOsh3wTDrJxgeSPBd47QjMJUmSJEmSpKfZBpn0A3YBbm5bW4+n854+gPOBe6rqjlGa9xDgiCS3A0uB17f6rwPHtYM/th+g7/nAs+gk/tZKu78TgKuSLAKuBrapqtvpbOtdCpwN3LC2c0mSJEmSJOnplyd2rirJGcBtVfUfYx3L6pLMBF5fVW8b61iGq6enp3p7e8c6jFExddblg15fNvugQa9LkiRJkiStiSQL2iGsg9og3+nXnyQL6Gxt/cexjmV1SU6ns9X2L8c6FnWY1JMkSZIkSesyk35NVU0f6xgGUlXvXb0uyeeAvVer/mxVjcQ7/yRJkiRJkrQeM+m3nqqqo8c6BkmSJEmSJK2bNtSDPCRJkiRJkqRxy6SfJEmSJEmSNM6Y9JMkSZIkSZLGGZN+kiRJkiRJ0jhj0k+SJEmSJEkaZ0z6SZIkSZIkSePMxmMdwPoiySTgrVX1+TXoOxXYq6q+OtJxaWxMnXX5k74vm33QGEUiSZIkSZL0x1zpN3yTgHevYd+pwFufaqckE9ZwPkmSJEmSJG3ATPoN32xg+yQLk3wyyXFJbkmyKMmJAEn2aN83TbJFkqVJdm599219P5Dk8CRn9A2c5NtJ9m/lh5J8LMn3gRlJpie5PsmCJFcm2WagAJNcl+TUJPOS3NniuTjJj5Kc1NXu0CQ3t3jO6ksuJvlCkt4W94ld7ZclOTHJrUkWJ9lhhJ+tJEmSJEmSRpBJv+GbBfy4qqYBVwN/DrwMmAZMT/LyqroFuBQ4Cfg34LyqWtL6zq+qaVV16hDzbAEsqao9ge8DpwMzq2o6cDZw8hD9H62qlwNnAt8CjgZ2Bg5P8uwkLwEOBvZu97ISOKT1Pb6qeoCXAvsleWnXuL+sqt2BLwDHDhGDJEmSJEmSxpDv9Fszr25/t7XvE+kkAecBHwNuAR4BjlmDsVcCF7Xyi+kk7K5OAjABuG+I/pe2z8XA0qq6DyDJ3cC2wD7AdOCWNuZmwC9anzcnOZLO/8U2wI7Aonbt4va5APibgSZv/Y8EmDJlyhChSpIkSZIkaTSY9FszAT5RVWf1c21rOknATYBNgYf7afM4T15luWlX+ZGqWtk1z9KqmvEUYlvRPld1lfu+b9zG/EpVfaS7U5IX0FnBt0dV/SbJ3NXi6htrJYP831TVHGAOQE9PTz2FuCVJkiRJkjRC3N47fA8CW7bylcA7kkwESPL8JM9p1+YAHwXOB07ppy/AMmBako2SbEtnm3B/fghMTjKjzbNJkp3W8j6uAWb2xZtk6yTbAc+kk6B8IMlzgdeu5TySJEmSJEkaI670G6aq+lWSG5IsAb4DfBW4sW2RfQg4NMlrgMer6qvtcIzvJTkAmA88nuR2YC7wGeAndLbgLgFuHWDOR5PMBE5LshWd3+szwNK1uI87kpwAXJVkI+Ax4OiquinJbW3su4Eb1nQOSZIkSZIkja1UuQNTo6Onp6d6e3vHOoxRMXXW5U/6vmz2QWMUiSRJkiRJ2pAkWdAOYh2UK/2kNWCST5IkSZIkrctM+q2HknwO2Hu16s9W1ZfHIh5JkiRJkiStW0z6rYeq6uixjkGSJEmSJEnrLk/vlSRJkiRJksYZk36SJEmSJEnSOGPST5IkSZIkSRpnTPpJkiRJkiRJ44xJP0mSJEmSJGmcMeknSZIkSZIkjTMbj3UA0vpo6qzL/1BeNvugMYxEkiRJkiTpj7nSbxxI8oYkO451HJIkSZIkSVo3mPQbH94A9Jv0S+JqTkmSJEmSpA2MSb81kGSLJJcnuT3JkiQHJ/lm1/VXJbm4lR9KckqSBUn+M8nLklyX5O4kr2ttDk9ySZLLkvwkyXuSfDDJbUluSrJ1a7d9kivaWPOT7JBkL+B1wCeTLGxtrkvy8STXA8e3MTdpYzwzybK+7/3c2zFJ7kiyKMnXu+737CS3tJheP6oPWJIkSZIkSWvFVWBr5jXAvVV1EECSrYATk0yuquXA24Evt7ZbANdV1YdbYvAk4FV0VuZ9Bbi0tdsZ2A3YFPgv4MNVtVuSU4HDgM8Ac4CjqupHSfYEPl9VByS5FPh2VV3Y4gGYVFX7te9TgYOAS4C3ABdV1WMD3Nss4AVVtSLJpFZ3PPB/q+odre7mJP9ZVQ+v3jnJkcCRAFOmTBnu85QkSZIkSdIIcqXfmlkMHNhW8O1bVQ8A5wKHtqTYDOA7re2jwBVd/a5vCbfFwNSuMa+tqgdb0vAB4LKuPlOTTAT2Ai5IshA4C9hmkBi/0VX+Ep1EJDw5IdmfRcD5SQ4FHm91rwZmtXmvo5OY7DejV1VzqqqnqnomT548yDSSJEmSJEkaLa70WwNVdVeS6cBfAp9IchWdxNplwCPABVXVlzB7rKqqlVcBK9oYq1Z7396KrvKqru+r6PxOGwH3V9W0YYb5h1V4VXVDkqlJ9gMmVNWSQfodBLyczpbhjybZCQjwt1X1w2HOLUmSJEmSpDHkSr81kOR5wO+q6jzgU8DuVXUvcC9wAjB3pOesqt8CP0nyphZDkuzaLj8IbDnEEOcAX2OQVX5JNgK2raprgQ8Bk4CJwJXAe9P2DSfZbW3uRZIkSZIkSaPLpN+a2YXOe+0W0nnf3Umt/nzgnqq6Y5TmPQQ4IsntwFKg70CNrwPHtUM2th+g7/nAs+gk/gYyATgvyWLgNuDUqrof+FdgE2BRkiXtuyRJkiRJktZReWLnqdZWkjOA26rqP8Y6ltUlmQm8vqre9nTN2dPTU729vU/XdJIkSZIkSeNekgVV1TNUO9/pN0KSLKDzHr1/HOtYVpfkdOC1dN5BKEmSJEmSpHHOpN8IqarpYx3DQKrqvavXJfkcsPdq1Z+tqsFO9pUkSZIkSdJ6wKTfBqqqjh7rGCRJkiRJkjQ6PMhDkiRJkiRJGmdM+kmSJEmSJEnjjEk/SZIkSZIkaZwx6SdJkiRJkiSNMyb9JEmSJEmSpHHGpJ+0BqbOupypsy4f6zAkSZIkSZL6ZdJvHZNkUpJ3j3UckiRJkiRJWn+Z9Fv3TAJM+kmSJEmSJGmNbdBJvySHJVmU5PYk5ybZLsk1re6aJFNau7lJvpDk2iR3J9kvydlJ7kwyt2u8h5L8e5JbW//Jrf6dSW5p81yUZPNW/9wk32z1tyfZC5gNbJ9kYZJPJtk/yXVJLkzygyTnJ0nrPz3J9UkWJLkyyTat/pgkd7T7+Hqr26+NuTDJbUm2HOCZbJNkXmu3JMm+rf7VSW5s93ZBkomj9sNIkiRJkiRprWywSb8kOwHHAwdU1a7A+4AzgHOq6qXA+cBpXV2eBRwAfAC4DDgV2AnYJcm01mYL4Naq2h24HvjnVn9xVe3R5rkTOKLVnwZc3+p3B5YCs4AfV9W0qjqutdsNeD+wI/BCYO8kmwCnAzOrajpwNnByaz8L2K3dx1Gt7ljg6KqaBuwL/H6AR/NW4MrWbldgYZI/BU4ADmz31gt8cJDHK0mSJEmSpDG08VgHMIYOAC6sql8CVNWvk8wA/qZdPxf4t672l1VVJVkM/LyqFgMkWQpMBRYCq4BvtPbnARe38s5JTqKzdXcicGVXDIe1+VcCDyR5Vj+x3lxVP2vzLWzz3Q/sDFzdFv5NAO5r7RcB5ye5BLik1d0AfDrJ+XSSkD8b4LncApzdkoqXVNXCJPvRSTje0Ob6E+DG/jonORI4EmDKlCkDTCFJkiRJkqTRtMGu9AMC1BBtuq+vaJ+rusp93wdKnvb1nwu8p6p2AU4ENn1KkT55vpVtvgBL24rAaVW1S1W9urU5CPgcMB1YkGTjqpoN/AOwGXBTkh36DbhqHvBy4L+Bc5Mc1ua6umuuHavqiAH6z6mqnqrqmTx58lO8TUmSJEmSJI2EDTnpdw3w5iTPBkiyNfA94C3t+iHAd5/imBsBM1v5rV39twTua6vnDlkthne1+SckeSbwYGs/lB8Ck9vqRJJskmSnJBsB21bVtcCHaKsLk2xfVYur6hQ623P7Tfol2Q74RVV9EfgPOtuOb6KzpfjPWpvNk7xoGDFKkiRJkiRpDGyw23urammSk4Hrk6wEbgOOobO19ThgOfD2pzjsw8BOSRYADwAHt/qPAt8Hfgos5omk3vuAOUmOoLOC711VdWOSG5IsAb4DXD5A/I8mmQmclmQrOr/lZ4C7gPNaXYBTq+r+JP+a5BVtnjva2P3ZHzguyWPAQ8BhVbU8yeHA15I8o7U7oc0lSZIkSZKkdUyqhtrhquFK8lBVeapt09PTU729vWMdhiRJkiRJ0riRZEFV9QzVbkPe3itJkiRJkiSNSxvs9t7RsD6t8kuyC50TirutqKo9xyIeSZIkSZIkjRyTfhuoqloMTBvrOCRJkiRJkjTy3N4rSZIkSZIkjTMm/SRJkiRJkqRxxqSfJEmSJEmSNM6Y9JMkSZIkSZLGGZN+kiRJkiRJ0jhj0k+SJEmSJEkaZ0z6SZIkSZIkSeOMSb81lOSoJIeN0Fj/eyTGkSRJkiRJksCk3xpJsnFVnVlV54zQkE856ZdkwgjNLUmSJEmSpHFmg036JZma5AdJvpJkUZILk2yeZHqS65MsSHJlkm1a++uSfDzJ9cD7kvxLkmO7rp2aZF6SO5PskeTiJD9KclLXnIcmuTnJwiRnJZmQZDawWas7f6B2rf6hJB9L8n1gxgD3tazFeWOS3iS7t/v4cZKjutodl+SWdu8ndtVf0u59aZIju+ofSnJyktuT3JTkuSP6g0iSJEmSJGnEbLBJv+bFwJyqeinwW+Bo4HRgZlVNB84GTu5qP6mq9quqf+9nrEer6uXAmcC32lg7A4cneXaSlwAHA3tX1TRgJXBIVc0Cfl9V06rqkIHatTm2AJZU1Z5V9d1B7uueqpoBzAfmAjOBvwA+BpDk1cCfAy8DpgHTk7y89X1Hu/ce4Jgkz+6a+6aq2hWYB7yzv4mTHNmSjb3Lly8fJERJkiRJkiSNlo3HOoAxdk9V3dDK59HZZrszcHUSgAnAfV3tvzHIWJe2z8XA0qq6DyDJ3cC2wD7AdOCWNvZmwC/6GeeVg7RbCVw0jPvqjmViVT0IPJjkkSSTgFe3v9tau4l0koDz6CT63tjqt231vwIeBb7d6hcAr+pv4qqaA8wB6OnpqWHEKkmSJEmSpBG2oSf9Vk9KPUgnYdfv1lng4UHGWtE+V3WV+75vDAT4SlV9ZIiYBmv3SFWtHKL/cGP5RFWd9aSJk/2BA4EZVfW7JNcBm7bLj1VV3/Naif87kiRJkiRJ66wNfXvvlCR9Cb6/A24CJvfVJdkkyU4jNNc1wMwkz2ljb51ku3btsSSbDKPdSLkSeEeSiW2O57f5tgJ+0xJ+O9DZEixJkiRJkqT1zIae9LsT+Pski4Ctae/zA05JcjuwENhrJCaqqjuAE4Cr2nxXA9u0y3OARUnOH6LdiKiqq4CvAjcmWQxcCGwJXAFs3Ob9VzpJUEmSJEmSJK1n8sSOzQ1LkqnAt6tq5zEOZdzq6emp3t7esQ5DkiRJkiRp3EiyoKp6hmq3oa/0kyRJkiRJksadDfYwhqpaRuek3vVSkm8CL1it+sNVdeVYxCNJkiRJkqR1xwab9FvfVdUbxzoGSZIkSZIkrZvc3itJkiRJkiSNMyb9JEmSJEmSpHHGpJ8kSZIkSZI0zpj0kyRJkiRJksYZk36SJEmSJEnSOGPST5IkSZIkSRpnRjXpl+SYJHcmOX8txzk8yfOG0W5ukpnDHHP/JN9u5dclmbU2Ma6JJM9LcuHTPe+aas9sr7GOQ5IkSZIkSYPbeJTHfzfw2qr6SV9Fko2r6vGnOM7hwBLg3hGM7Q+q6lLg0tEYe4h57wWGlaR8OiWZUFUr+7m0P/AQ8L2nNyJJkiRJkiQ9FaO20i/JmcALgUuTPJBkTpKrgHOSTE0yP8mt7W+vrn4fSrI4ye1JZreVez3A+UkWJtksyT8luSXJkjZuhhnTa5L8IMl3gb/pqj88yRmtPDfJF5Jcm+TuJPslObutWJzb1efVSW5s8V+QZGKrX5bkxFa/OMkOrX6/Fv/CJLcl2bI9hyXt+qZJvtz63JbkFV2xXZzkiiQ/SvJvg9zfm5N8upXfl+TuVt6+3TNJXtnGX9zu6xldcf9Ta/emtkrzjiSLknw9yVTgKOAD7R72Hc4zlyRJkiRJ0tNv1Fb6VdVRSV4DvAJ4D/DXwD5V9fskmwOvqqpHkvw58DWgJ8lrgTcAe1bV75JsXVW/TvIe4Niq6gVIckZVfayVzwX+CrhssHiSbAp8ETgA+C/gG4M0f1Zr97o27t7APwC3JJkG/Aw4ATiwqh5O8mHgg8DHWv9fVtXuSd4NHNv6HgscXVU3tAThI6vNeXR7bru0ROFVSV7Urk0DdgNWAD9McnpV3dNP3POA41p5X+BXSZ4P7APMb89gLvDKqroryTnAu4DPtD6PVNU+7XndC7ygqlYkmVRV97dE7kNV9amBHlySI4EjAaZMmTJQM0mSJEmSJI2ip/Mgj0ur6vetvAnwxSSLgQuAHVv9gcCXq+p3AFX16wHGekWS77f+BwA7DWP+HYCfVNWPqqqA8wZpe1lrsxj4eVUtrqpVwFJgKvAXLeYbkiwE/h7Yrqv/xe1zQWsPcAPw6STHAJP62eK8D3AuQFX9APgp0Jf0u6aqHqiqR4A7VpvrD6rqf4CJSbYEtgW+CrycTgJwPvDi9gzual2+0q736U6ELqKzuvJQYNjbsatqTlX1VFXP5MmTh9tNkiRJkiRJI+jpTPo93FX+APBzYFc6W3f/pNUHqMEGaavVPg/MrKpd6Kze23SYMQw6dpcV7XNVV7nv+8Ytzquralr727Gqjuin/8rWnqqaTWfF32bATX3bfrsMtkW5O4Y/jDmAG4G3Az+kk+jbF5hBJ+k41Dbo7t/oIOBzwHRgQZLRfv+jJEmSJEmSRsjTmfTrthVwX1s99zZgQqu/CnhH2/5Lkq1b/YPAlq3cl+D7ZdsmO9yDMH4AvCDJ9u37361F/DcBeyf5sxbn5l1bcfuVZPu2YvAUoJfOysNu84BDWtsXAVPoJO6eqnl0thLPA26js716RVU9QOcZTO2Lm86zv76fWDcCtq2qa4EPAZOAiTz5d5AkSZIkSdI6aqySfp8H/j7JTXS2sD4MUFVX0DlFt7dtmz22tZ8LnNnqVtBZ3bcYuAS4ZTgTtq2xRwKXt8MqfrqmwVfVcjonCn8tySI6ScDVk3ire387eOR24PfAd1a7/nlgQtuy/A3g8KpasfogwzCfztbeee0E3nuA77a4H6GzCvCCNs8q4Mx+xpgAnNfa3AacWlX303m/4Rs9yEOSJEmSJGndls6r66SR19PTU729vWMdhiRJkiRJ0riRZEFV9QzVbqxW+kmSJEmSJEkaJePycIYk3wResFr1h6vqyrGIZzQk+T7wjNWq31ZVi8ciHkmSJEmSJK07xmXSr6reONYxjLaq2nOsY5AkSZIkSdK6ye29kiRJkiRJ0jhj0k+SJEmSJEkaZ0z6SZIkSZIkSeOMST9JkiRJkiRpnDHpJ0mSJEmSJI0zJv3WAUnekGTHsY5jKEmmJnnrWMchSZIkSZKkwZn0Wze8Aeg36Zdk46c5lsHmnAqY9JMkSZIkSVrHbZBJvyRbJLk8ye1JliQ5OMk3u66/KsnFrfxQklOSLEjyn0leluS6JHcneV1rc3iSS5JcluQnSd6T5INJbktyU5KtW7vtk1zRxpqfZIckewGvAz6ZZGFrc12Sjye5Hji+jblJG+OZSZb1fV/tvp6TZEEr75qkkkxp33+cZPMk2yW5Jsmi9tl3fW6STye5FjglyX4tnoXtPrYEZgP7troPjN4vJEmSJEmSpLWxQSb9gNcA91bVrlW1M3AF8JIkk9v1twNfbuUtgOuqajrwIHAS8CrgjcDHusbcmc4quJcBJwO/q6rdgBuBw1qbOcB721jHAp+vqu8BlwLHVdW0qvpxazupqvarqhOB64CDWv1bgIuq6rHVb6qqfgFsmuSZwL5AL50k3XbAL6rqd8AZwDlV9VLgfOC0riFeBBxYVf/Y4ju6qqa1sX4PzALmtzhPHeohS5IkSZIkaWxsqEm/xcCBbQXfvlX1AHAucGiSScAM4Dut7aN0koJ9/a5vCbfFdLa79rm2qh6squXAA8BlXX2mJpkI7AVckGQhcBawzSAxfqOr/CU6iUh4ckKyP98D9gZeDny8fe4LzG/XZwBfbeVzgX26+l5QVStb+Qbg00mOoZOAfHyQOf8gyZFJepP0Ll++fDhdJEmSJEmSNMKe9vfFrQuq6q4k04G/BD6R5Co6ibXLgEfoJL/6klyPVVW18ipgRRtj1WrvvlvRVV7V9X0Vnee8EXB/Wzk3HA93xXtDO0RjP2BCVS0ZpN98Okm+7YBvAR8GCvj2AO2rq9w95+wkl9N5RjclOXA4QVfVHDorGunp6akhmkuSJEmSJGkUbJAr/ZI8j8722/OATwG7V9W9wL3ACcDckZ6zqn4L/CTJm1oMSbJru/wgsOUQQ5wDfI3BV/kBzAMOBX5UVauAX9NJ3N3Qrn+PzhZhgEOA7/Y3SJLtq2pxVZ1CZ5vwDsOMU5IkSZIkSWNsg0z6AbsAN7dttsfTeU8fdN5xd09V3TFK8x4CHJHkdmAp8PpW/3XguHZgxvYD9D0feBadxN+AqmpZK85rn9+ls8LwN+37McDbkywC3ga8b4Ch3t8OObmdzvv8vgMsAh5vB6B4kIckSZIkSdI6Kk/sXFWSM4Dbquo/xjqW1SWZCby+qt421rEMV09PT/X29o51GJIkSZIkSeNGkgVV1TNUuw3ynX79SbKAzjvt/nGsY1ldktOB19LZpitJkiRJkiQNyqRfU1XTxzqGgVTVe1evS/I5Oqf0dvtsVQ31zj9JkiRJkiSNcyb91lNVdfRYxyBJkiRJkqR104Z6kIckSZIkSZI0bpn0kyRJkiRJksYZk36SJEmSJEnSOGPST5IkSZIkSRpnTPpJkiRJkiRJ44xJP0mSJEmSJGmcMeknSZIkSZIkjTMm/caRJB9LcmArvz/J5mMdkyRJkiRJkp5+Jv3Gkar6p6r6z/b1/YBJP0mSJEmSpA2QSb8RkOSwJIuS3J7k3CTbJbmm1V2TZEprNzfJaUm+l+TuJDO7xvhQksVtjNmt7p1Jbml1FyXZPMlWSZYl2ai12TzJPUk2aePPTHIM8Dzg2iTXJjkiyaldc70zyacHuJctklze5lyS5OBWPz3J9UkWJLkyyTaj90QlSZIkSZK0Nkz6raUkOwHHAwdU1a7A+4AzgHOq6qXA+cBpXV22AfYB/groS+69FngDsGcb499a24urao9WdydwRFU9ANwO7Nfa/DVwZVU91jdBVZ0G3Au8oqpeAXwdeF2STVqTtwNfHuCWXgPcW1W7VtXOwBWt3+nAzKqaDpwNnDzA8zgySW+S3uXLlw/+8CRJkiRJkjQqTPqtvQOAC6vqlwBV9WtgBvDVdv1cOkm+PpdU1aqqhWKDGAAAIABJREFUugN4bqs7EPhyVf2uawyAnZPMT7IYOATYqdV/Azi4ld/Svg+oqh4G/i/wV0l2ADapqsUDNF/8/7d37+F5VmW+x7+/AloOFQTZbhChyAhVKY2QFhERcKCIo6IjUJWt1jMquHUuD+ioo844orBFBVFx1CKiMiA6CAooclahQXqggCBQlcHRgoogp7bc+493BUJI2rRpkjZ8P9eVK++7nnW4nyeL5OXuWs8D7JfkU0n2aknGnYCdgR8nmQd8CNhmkLFOqqruqurecsstVxSWJEmSJEmSRsj6Yx3AOBCgVlKn7/H7+7VdUR9zgJdV1fwks4F9WvlZwCeTbA7sRiehtzL/AXwQuJ7BV/lRVTck2Q14URvjfOB7wKKq2mMI40iSJEmSJGmMudJv+C4ADk2yBUBLxP2Mzgo86KzQu2wlfZwPvKH3abutD4BJwO/b9trDeitX1d3AlcDngLOravkAfd7V2ve2uQJ4KvBq4NuDBZJka+CeqvomcCywK/ArYMske7Q6G7RtzZIkSZIkSVoLudJvmKpqUZJPABcnWQ5cDbwT+FqS9wJL6NxDb0V9nJukC+hJ8gDwQzqr8j4MXAH8hs6220l9mp0GnM7Dq//6Own4UZLft/v6Afwn0FVVf15BOFOBY5I8CCwF3lZVD7SHjnw+yaZ05s1ngUUrOi9JkiRJkiSNjVStbGeqxoskZwPHVdUFozFed3d39fT0jMZQkiRJkiRJjwlJrqqq7pXVc3vvY0CSzZLcANw7Wgk/SZIkSZIkjR239z4GVNVfgB37lrV7EA6UAPz7qrpjVAKTJEmSJEnSiDDp9xjVEntdYx2HJEmSJEmS1jy390qSJEmSJEnjjEk/SZIkSZIkaZwx6SdJkiRJkiSNMyb9JEmSJEmSpHHGpJ8kSZIkSZI0zpj0kyRJkiRJksYZk36SJEmSJEnSOGPSbx2UZOskZwyh3gdHIx5JkiRJkiStXUz6rYOq6raqOngIVU36SZIkSZIkPQaNy6RfktcmWZBkfpJTWtl2SS5o5Rck2baVz0ny+SQ/S3JzkoP79PO+JAtbP0e3sjcnmdvKvptkoySbJlmcZEKrs1GS3yXZIMkOSc5NclWSS5NMGSDejyY5JclPk9yY5M2tPEmOSXJNi2NWK5+c5Jr2enaSM9sYNyb5dCs/GtgwybwkpybZOMk5Le5revsa5PodneTadq2ObWVbtvOd2772HKTtW5L0JOlZsmTJavz0JEmSJEmSNFzrj3UAa1qSZwH/DOxZVbcn2bwdOgH4RlWdnOQNwOeBl7VjWwHPA6YAZwFnJDmwHd+9qu7p08+ZVfWVNta/AW+squOTzAf2Bi4EXgKcV1VLk5wEHF5VNybZHTgReMEAoe8CPAfYGLg6yTnAHkAXMA14EjA3ySUDtO0Cng3cD/wqyfFVdVSSI6qqq8X6CuC2qvqH9n7TQa7f5sDLgSlVVUk2a4c+BxxXVZe1hOl5wDP6t6+qk4CTALq7u2ugMSRJkiRJkjSyxl3Sj05C7Yyquh2gqv7UyvcA/rG9PgX4dJ8236+qB4Frkzy5le0HfL2q7unXz84t2bcZsAmd5BfAacAsOkm/VwInJtkEeC5wepLesR4/SNz/VVX3AvcmuRCYQScR+e2qWg78IcnFwHRgQb+2F1TVnQBJrgW2A37Xr85C4NgknwLOrqpLB4njr8B9wH+0xOPZfa7HM/ucxxOSTKqquwbpR5IkSZIkSWNkPCb9AgxlhVnfOvf3a7+ifuYAL6uq+UlmA/u08rOAT7aVcrsBP6Wzau8vvavtViGe3vcZqOIA+sa/nAF+rlV1Q5LdgBe1OM+vqo8PUG9ZkhnA39NJXh5BJ5E6AdijJSYlSZIkSZK0FhuP9/S7ADg0yRbw0HZVgJ/RSWIBHAZctpJ+zgfekGSjfv1MAn6fZIPWDwBVdTdwJZ1tsGdX1fKq+itwS5JDWh9JMm2Q8Q5KMrHFvQ8wF7gEmJVkvSRbAs9vYwzV0hYnSbYG7qmqbwLHArsO1KCtTty0qn4IvIvO1uHe63FEn3pDSWRKkiRJkiRpDIy7lX5VtSjJJ4CLkywHrgZmA+8EvpbkvcAS4PUr6efcltjqSfIA8EM6T8P9MHAF8Bs6W2Yn9Wl2GnA6D6/+g05i8ItJPgRsAHwHmD/AkFcC5wDbAv9aVbcl+R6dbcnz6az8e19V/U+SyUO6GJ176y1I8kvgG8AxSR4ElgJvG6TNJOC/kkyks9Lw3a38ncAXkiygM28uAQ4fYhySJEmSJEkaRanyWQtjLclHgbur6tixjmVN6u7urp6enrEOQ5IkSZIkadxIclVVda+s3njc3itJkiRJkiQ9po277b3roqr66FiM27YPb9+v+P1Vdd5A9SVJkiRJkrRuMOn3GFZVLx/rGCRJkiRJkrTmub1XkiRJkiRJGmdM+kmSJEmSJEnjjEk/SZIkSZIkaZwx6SdJkiRJkiSNMyb9JEmSJEmSpHHGpJ8kSZIkSZI0zqzVSb8k70xyXZJTh9nP7CRbD6HenCQHD7HPfZKc3V6/NMlRw4lxdSTZOskZoz2uJEmSJEmS1m7rj3UAK/F24MCquqW3IMn6VbVsFfuZDVwD3LYGY3tIVZ0FnDUSfa9k3NuAISUpJUmSJEmS9Nix1q70S/Il4GnAWUnuTHJSkvOBbySZnOTSJL9sX8/t0+59SRYmmZ/k6LZyrxs4Ncm8JBsm+UiSuUmuaf1miDG9MMn1SS4D/rFP+ewkJ7TXc5J8McmFSW5OsneSr7UVi3P6tJmZ5Oct/tOTbNLKFyf5WCtfmGRKK9+7xT8vydVJJrXrcE07PjHJ11ubq5Ps2ye2M5Ocm+TGJJ9ewfmt1+K/pvXz7la+Q2t/VbvuU1bQx1uS9CTpWbJkyVAuqyRJkiRJktawtTbpV1WH01mZty9wHLAbcFBVvRr4I7B/Ve0KzAI+D5DkQOBlwO5VNQ34dFWdAfQAh1VVV1XdC5xQVdOramdgQ+DFK4snyUTgK8BLgL2A/72C6k8EXgC8G/hBi/9ZwNQkXUmeBHwI2K+dQw/wT33a397Kvwi8p5W9B3hHVXW18e/tN+Y72nWbCrwKOLnFDNDVrtNUYFaSpw4SdxfwlKraufXz9VZ+EnBkVe3W4jhxsBOvqpOqqruqurfccsvBqkmSJEmSJGkEre3be/s6qyXsADYATkjSBSwHdmzl+wFfr6p7AKrqT4P0tW+S9wEbAZsDi+gk51ZkCnBLVd0IkOSbwFsGqfuDqqokC4E/VNXC1mYRMBnYBngmcHlbZPg44Od92p/Zvl/FwysKLwc+0+5veGZV3dpvgeLzgOPbeV+f5Dc8fF0uqKo7WwzXAtsBvxsg7puBpyU5HjgHOL+tQHwucHqf8R4/yHlLkiRJkiRpLbAuJf3+1uf1u4E/ANPorFa8r5UHqBV10la/nQh0V9XvknwUmLiiNn2ssO8+7m/fH+zzuvf9+nQSlT+uqletpP3yVp+qOjrJOcCLgF8k2Y+Hzxs6576yeB7RZ39V9eck04AD6KwcPBR4F/CXtsJQkiRJkiRJ64C1dnvvSmwK/L6qHgReA6zXys8H3pBkI4Akm7fyu4BJ7XVvgu/2toptqA/CuB7YPskO7f1gCbuh+AWwZ5K/a3FulGTHFTVIskNVLayqT9HZDtz/vnqXAIe1ujsC2wK/WpWg2rbjCVX1XeDDwK5V9VfgliSHtDppiUFJkiRJkiStpdbVpN+JwOuS/ILOFta/AVTVuXSeotuTZB4P3w9vDvClVnY/nXvzLQS+D8wdyoBVdR+d7bzntAd5/GZ1g6+qJXSeKPztJAvoJAEHfThG8672gI35dO7n96N+x08E1mtbik8DZlfV/f07WYmnABe16zQH+EArPwx4Yxt7EXDQKvYrSZIkSZKkUZSqoe5YlVZNd3d39fT0jHUYkiRJkiRJ40aSq6qqe2X11tWVfpIkSZIkSZIGsS49yGPUJPkesH2/4vdX1XljEc9ISHIFj34K72t6nzQsSZIkSZKkdZdJvwFU1cvHOoaRVlW7j3UMkiRJkiRJGhlu75UkSZIkSZLGGZN+kiRJkiRJ0jhj0k+SJEmSJEkaZ0z6SZIkSZIkSeOMST9JkiRJkiRpnDHpJ0mSJEmSJI0zJv3GQJKtk5wxhHofHI14VkWSriQvGus4JEmSJEmSNDiTfmOgqm6rqoOHUHXMkn5J1h/kUBdg0k+SJEmSJGkttlYm/ZK8NsmCJPOTnNLKtktyQSu/IMm2rXxOks8n+VmSm5Mc3Kef9yVZ2Po5upW9OcncVvbdJBsl2TTJ4iQTWp2NkvwuyQZJdkhybpKrklyaZMoA8X40ySlJfprkxiRvbuVJckySa1ocs1r55CTXtNezk5zZxrgxyadb+dHAhknmJTk1ycZJzmlxX9Pb1wCxzEhyZnt9UJJ7kzwuycQkN7fyriS/aNfye0me2MovSvLvSS4G/m+SQ9pY85NckuRxwMeBWS2uR8WQ5C1JepL0LFmyZHV+/JIkSZIkSRqmwVZzjZkkzwL+Gdizqm5Psnk7dALwjao6OckbgM8DL2vHtgKeB0wBzgLOSHJgO757Vd3Tp58zq+orbax/A95YVccnmQ/sDVwIvAQ4r6qWJjkJOLyqbkyyO3Ai8IIBQt8FeA6wMXB1knOAPeisjJsGPAmYm+SSAdp2Ac8G7gd+leT4qjoqyRFV1dVifQVwW1X9Q3u/6SCX8JetL4C9gGuA6XR+1le08m8AR1bVxUk+DvwL8K52bLOq2ruNsRA4oKr+O8lmVfVAko8A3VV1xECDV9VJwEkA3d3dNUiMkiRJkiRJGkFr40q/FwBnVNXtAFX1p1a+B/Ct9voUOkm+Xt+vqger6lrgya1sP+DrVXVPv352biv2FgKHAc9q5acBvSvXXgmclmQT4LnA6UnmAV+mk2AcyH9V1b0t7guBGS3Gb1fV8qr6A3AxnQRcfxdU1Z1VdR9wLbDdAHUWAvsl+VSSvarqzoGCqKplwK+TPKPF8Bng+XQSgJe2ZOFmVXVxa3JyO97rtD6vLwfmtJWL6w1y3pIkSZIkSVrLrI1JvwBDWSHWt879/dqvqJ85wBFVNRX4GDCxlZ8FHNhWBO4G/JTO9flLVXX1+XrGEOLpfZ+BKg6gb/zLGWAFZlXd0OJaCHyyrbgbzKXAgcBS4Cd0ko/PAwZaZdjf3/qMeTjwIeCpwLwkWwyhvSRJkiRJksbY2pj0uwA4tDfB1Gdb7s/orMCDzgq9y1bSz/nAG5Js1K+fScDvk2zQ+gGgqu4GrgQ+B5zdVuf9FbglySGtjySZNsh4B7X75m0B7APMpZNkm5VkvSRb0llRd+VQLkKztMVJkq2Be6rqm8CxwK4raHcJne26P6+qJcAWdLY+L2orBP+cZK9W9zV0ViA+SpIdquqKqvoIcDud5N9ddK6hJEmSJEmS1lJr3T39qmpRkk8AFydZDlwNzAbeCXwtyXuBJcDrV9LPuUm6gJ4kDwA/pPM03A/Tubfdb+ismuubwDoNOJ1O0q7XYcAXk3wI2AD4DjB/gCGvBM4BtgX+tapuS/I9OtuS59NZ+fe+qvqfJJOHdDE698ZbkOSXdO7Dd0ySB+ms4HvbCtpdQWebc+/KvgXAH6uqdzXi64AvtYTozQx+LY9J8nQ6KxYvaOfxW+Cott35k1V12iBtJUmSJEmSNEbycB5IqyvJR4G7q+rYsY5lbdLd3V09PT1jHYYkSZIkSdK4keSqqupeWb21cXuvJEmSJEmSpGFY67b3rouq6qNjMW7bPrx9v+L3V9V5YxGPJEmSJEmS1g4m/dZhVfXysY5BkiRJkiRJax+390qSJEmSJEnjjEk/SZIkSZIkaZwx6SdJkiRJkiSNMyb9JEmSJEmSpHHGpJ8kSZIkSZI0zpj0kyRJkiRJksYZk35jJMnkJK/u8352khPGMiZJkiRJkiSNDyb9xs5k4NUrqyRJkiRJkiStKpN+/STZOMk5SeYnuSbJrCSLk/x7kp8n6Umya5LzktyU5PDWLkmOaW0WJpm1onLgaGCvJPOSvLuVbZ3k3CQ3Jvl0n5juTvKJFtMvkjy5lW+Z5LtJ5ravPVv53q3feUmuTjIpyVZJLmll1yTZa5DzXy/JnD7xvruV79BiuyrJpUmmjMgPQJIkSZIkScNm0u/RXgjcVlXTqmpn4NxW/ruq2gO4FJgDHAw8B/h4O/6PQBcwDdgPOCbJVisoPwq4tKq6quq41kcXMAuYCsxK8tRWvjHwi6qaBlwCvLmVfw44rqqmA68A/qOVvwd4R1V1AXsB99JZVXheK5sGzBvk/LuAp1TVzlU1Ffh6Kz8JOLKqdmv9nzhQ4yRvaYnRniVLlgwyhCRJkiRJkkbS+mMdwFpoIXBskk8BZ1fVpUkAzupzfJOqugu4K8l9STYDngd8u6qWA39IcjEwfQXlfx1g7Auq6k6AJNcC2wG/Ax4Azm51rgL2b6/3A57Z4gN4QpJJwOXAZ5KcCpxZVbcmmQt8LckGwPerarCk383A05IcD5wDnJ9kE+C5wOl9xnr8QI2r6iQ6CUK6u7trkDEkSZIkSZI0glzp109V3QDsRie598kkH2mH7m/fH+zzuvf9+kAY2GDlA+nb73IeTsouraoaoHwCsEdbLdhVVU+pqruq6mjgTcCGwC+STKmqS4DnA/8NnJLktQMFUFV/prMS8CLgHXRWD04A/tJnnK6qesYqnJckSZIkSZJGkUm/fpJsDdxTVd8EjgV2HWLTS+hsyV0vyZZ0EmxXrqD8LmDSMMM9HziiT+xd7fsOVbWwqj4F9ABTkmwH/LGqvgJ8dbDzSvIkYEJVfRf4MLBrVf0VuCXJIa1OkkwbZuySJEmSJEkaIW7vfbSpdO679yCwFHgbcMYQ2n0P2AOYDxTwvqr6nySDld8BLEsyn849Av+8GrG+E/hCkgV0fpaXAIcD70qyL51VgdcCPwJeCbw3yVLgbmDAlX7AU4CvJ+lNCH+gfT8M+GKSDwEbAN9p5yRJkiRJkqS1TB7eNSqtWd3d3dXT0zPWYUiSJEmSJI0bSa6qqu6V1XN7ryRJkiRJkjTOuL33MSzJFTz6KbyvqaqFYxGPJEmSJEmS1gyTfo9hVbX7WMcgSZIkSZKkNc/tvZIkSZIkSdI4Y9JPkiRJkiRJGmdM+kmSJEmSJEnjjEk/SZIkSZIkaZwx6SetoslHnTPWIUiSJEmSJK2QSb9RlGSzJG9fzbaTk7x6TcckSZIkSZKk8cek3+jaDFitpB8wGVjlpF+S9VZzPEmSJEmSJK2jTPqNrqOBHZLMS3JMkvcmmZtkQZKPASSZ3t5PTLJxkkVJdm5t92pt351kdpITejtOcnaSfdrru5N8PMkVwB5JdktycZKrkpyXZKvBAkzyziTXthi+08o2TvK1FuvVSQ4awWskSZIkSZKkYVp/rAN4jDkK2LmqupLMBA4GZgABzkry/Kq6JMlZwL8BGwLfrKprkhwFvKeqXgyQZPYKxtkYuKaqPpJkA+Bi4KCqWpJkFvAJ4A0riHH7qro/yWat7J+Bn1bVG1rZlUl+UlV/G8a1kCRJkiRJ0ggx6Td2Zravq9v7TYCnA5cAHwfmAvcB71yNvpcD322vdwJ2Bn6cBGA94PcraLsAODXJ94Hv94n1pUne095PBLYFruvfOMlbgLcAbLvttqsRuiRJkiRJkobLpN/YCfDJqvryAMc2p5ME3IBOgm2gFXXLeOT27Il9Xt9XVcv7jLOoqvYYYlz/ADwfeCnw4STPan28oqp+tbLGVXUScBJAd3d3DXFMSZIkSZIkrUHe02903QVMaq/PA96QZBOAJE9J8r/asZOADwOnAp8aoC3AYqAryYQkT6WzTXggvwK2TLJHG2eDlsh7lCQTgKdW1YXA++g8eGSTFuuRaUsFkzx7lc5akiRJkiRJo8qVfqOoqu5IcnmSa4AfAd8Cft5yaXcD/yfJC4FlVfWt9uTdnyV5AXApsCzJfGAO8FngFmAhcA3wy0HGfCDJwcDnk2xK52f+WWDRANXXA77Z6gU4rqr+kuRfW5sFLfG3GHjx8K+IJEmSJEmSRkKq3IGpkdHd3V09PT1jHYYkSZIkSVoDli5dyq233sp999031qE8JkycOJFtttmGDTbY4BHlSa6qqu6VtXelnyRJkiRJklbq1ltvZdKkSUyePJm2a1EjpKq44447uPXWW9l+++1Xqw/v6fcYleQLSeb1+3r9WMclSZIkSZLWTvfddx9bbLGFCb9RkIQttthiWKsqXen3GFVV7xjrGCRJkiRJ0rrFhN/oGe61dqWfJEmSJEmS1gmbbLLJqI63ePFivvWtb43qmGuKK/0kSZIkSZK0yiYfdc4a7W/x0f+wRvsbrmXLlj2U9Hv1q1891uGsMlf6SZIkSZIkaZ1y0UUXsffee3PooYey4447ctRRR3HqqacyY8YMpk6dyk033QTA7NmzOfzww9lrr73YcccdOfvss4HO/Qlf//rXM3XqVJ797Gdz4YUXAjBnzhwOOeQQXvKSlzBz5kyOOuooLr30Urq6ujjuuONYvHgxe+21F7vuuiu77rorP/vZzx6KZ5999uHggw9mypQpHHbYYVQVAHPnzuW5z30u06ZNY8aMGdx1110sX76c9773vUyfPp1ddtmFL3/5y2v8GrnST5IkSZIkSeuc+fPnc91117H55pvztKc9jTe96U1ceeWVfO5zn+P444/ns5/9LNDZonvxxRdz0003se+++/LrX/+aL3zhCwAsXLiQ66+/npkzZ3LDDTcA8POf/5wFCxaw+eabc9FFF3Hsscc+lCy85557+PGPf8zEiRO58cYbedWrXkVPTw8AV199NYsWLWLrrbdmzz335PLLL2fGjBnMmjWL0047jenTp/PXv/6VDTfckK9+9atsuummzJ07l/vvv58999yTmTNnrvaTegdi0k+SJEmSJEnrnOnTp7PVVlsBsMMOOzBz5kwApk6d+tDKPYBDDz2UCRMm8PSnP52nPe1pXH/99Vx22WUceeSRAEyZMoXtttvuoaTf/vvvz+abbz7gmEuXLuWII45g3rx5rLfeeg+1AZgxYwbbbLMNAF1dXSxevJhNN92UrbbaiunTpwPwhCc8AYDzzz+fBQsWcMYZZwBw5513cuONN5r0k0bTQPcoWNvuMyBJkiRJ0mPN4x//+IdeT5gw4aH3EyZMYNmyZQ8d6/8U3CQPbb0dyMYbbzzoseOOO44nP/nJzJ8/nwcffJCJEycOGM96663HsmXLqKoBn8JbVRx//PEccMABKzjD4fGefpIkSZIkSRq3Tj/9dB588EFuuukmbr75ZnbaaSee//znc+qppwJwww038Nvf/paddtrpUW0nTZrEXXfd9dD7O++8k6222ooJEyZwyimnsHz58hWOPWXKFG677Tbmzp0LwF133cWyZcs44IAD+OIXv8jSpUsfiuFvf/vbmjplYASTfknemeS6JKcOs5/ZSbYeQr05SQ4eYp/7JDm7vX5pkqOGE+PqSLJ1kjNGe1xJkiRJkqTHkp122om9996bAw88kC996UtMnDiRt7/97SxfvpypU6cya9Ys5syZ84iVer122WUX1l9/faZNm8Zxxx3H29/+dk4++WSe85zncMMNN6xwVSDA4x73OE477TSOPPJIpk2bxv777899993Hm970Jp75zGey6667svPOO/PWt771EasT14SsaDnjsDpOrgcOrKpb+pStX1WrdAZJLgLeU1U9K6k3Bzi7qlaaSEuyT+vzxasSi1ZNd3d39d7Mcl3m9l5JkiRJkuC6667jGc94xliHsUpmz57Ni1/8Yg4+eEjrxNY6A13zJFdVVffK2o7ISr8kXwKeBpyV5M4kJyU5H/hGkslJLk3yy/b13D7t3pdkYZL5SY5uK/e6gVOTzEuyYZKPJJmb5JrW76M3Rg8c0wuTXJ/kMuAf+5TPTnJCez0nyReTXJjk5iR7J/laW7E4p0+bmUl+3uI/PckmrXxxko+18oVJprTyvVv885JcnWRSuw7XtOMTk3y9tbk6yb59YjszyblJbkzy6ZWc491JPpXkqiQ/STIjyUXtXF7a6qyX5Jh2DRckeWsr3yTJBX1iP6iVT27n/5Uki5Kcn2TDoVxzSZIkSZIkjY0RSfpV1eHAbcC+wHHAbsBBVfVq4I/A/lW1KzAL+DxAkgOBlwG7V9U04NNt1V4PcFhVdVXVvcAJVTW9qnYGNgRWulovyUTgK8BLgL2A/72C6k8EXgC8G/hBi/9ZwNQkXUmeBHwI2K+dQw/wT33a397Kvwi8p5W9B3hHVXW18e/tN+Y72nWbCrwKOLnFDNDVrtNUYFaSp64g9o2Bi6pqN+Au4N+A/YGXAx9vdd4I3FlV04HpwJuTbA/cB7y8xb4v8P/6JFSfDnyhqp4F/AV4xWABJHlLkp4kPUuWLFlBqJIkSZIkSSNrzpw56+wqv+EarQd5nNUSdgAbAF9JshA4HXhmK98P+HpV3QNQVX8apK99k1zR2r+ATkJuZaYAt1TVjdXZz/zNFdT9QauzEPhDVS2sqgeBRcBk4Dkt5suTzANeB2zXp/2Z7ftVrT7A5cBnkrwT2GyALc7PA04BqKrrgd8AO7ZjF1TVnVV1H3Btv7H6ewA4t71eCFxcVUvb695YZgKvbbFfAWxBJ6kX4N+TLAB+AjwFeHJrc0tVzRvgvB6lqk6qqu6q6t5yyy1XEKokSZIkSZJGyvqjNE7fx4+8G/gDMI1O0vG+Vh5ghTcYbKvfTgS6q+p3ST4KTFxRmz6GevPC+9v3B/u87n2/PrAc+HFVvWol7Ze3+lTV0UnOAV4E/CLJfjx83tA595XF84g+B7G0Hr5J40PxV9WDSXrbBTiyqs7r2zDJbGBLYLeqWppkMQ9f2/4xuL1XkiRJkqTHoKpiiHda0zAN9zkco7XSr69Ngd+31XOvAdZr5ecDb0iyEUCSzVsszg17AAAINklEQVT5XcCk9ro3CXV7u4/eUNdnXg9sn2SH9n6whN1Q/ALYM8nftTg3SrLjihok2aGtGPwUne3AU/pVuQQ4rNXdEdgW+NUwYlyR84C3Jdmgd7wkG9P5ufyxJfz2ZcUrCiVJkiRJ0mPMxIkTueOOO4adjNLKVRV33HEHEycOda3bo43WSr++TgS+m+QQ4ELaKsCqOjdJF9CT5AHgh8AHgTnAl5LcC+xB5958C4HFwNyhDFhV9yV5C3BOktuBy4CdVyf4qlrSVsV9O0nvs5w/BNywgmbvaom05XS26P4I2KrP8RPpnONCYBkwu6ruH6HM+X/Q2Z77y3bPviV07qV4KvCDJD3APDqJUuGTeiVJkiRJAthmm2249dZb8R7+o2PixIlss802q90+Zmc1Urq7u6unp2esw5AkSZIkSRo3klxVVd0rqzcW23slSZIkSZIkjaCx2N47opJ8D9i+X/H7+z+4Yl2W5Arg8f2KX1NVC8ciHkmSJEmSJK1dxl3Sr6pePtYxjLSq2n2sY5AkSZIkSdLay3v6acQkWQL8ZqzjGCFPAm4f6yC0VnAuqJdzQX05H9TLuaBezgX1ci6ol3NBfa3KfNiuqrZcWSWTftJqSNIzlJtmavxzLqiXc0F9OR/Uy7mgXs4F9XIuqJdzQX2NxHzwQR6SJEmSJEnSOGPST5IkSZIkSRpnTPpJq+eksQ5Aaw3ngno5F9SX80G9nAvq5VxQL+eCejkX1Ncanw/e00+SJEmSJEkaZ1zpJ0mSJEmSJI0zJv2kfpK8MMmvkvw6yVEDHH98ktPa8SuSTO5z7AOt/FdJDhjNuLXmre5cSLJ/kquSLGzfXzDasWvNGs7vhXZ82yR3J3nPaMWskTHMvxG7JPl5kkXt98PE0Yxda9Yw/kZskOTkNgeuS/KB0Y5da94Q5sPzk/wyybIkB/c79rokN7av141e1BoJqzsXknT1+RuxIMms0Y1ca9pwfi+0409I8t9JThidiDVShvk3Ytsk57fPDNf2//+MlTHpJ/WRZD3gC8CBwDOBVyV5Zr9qbwT+XFV/BxwHfKq1fSbwSuBZwAuBE1t/WgcNZy4AtwMvqaqpwOuAU0Ynao2EYc6FXscBPxrpWDWyhvk3Yn3gm8DhVfUsYB9g6SiFrjVsmL8XDgEe3/5G7Aa8dVU/wGvtMsT58FtgNvCtfm03B/4F2B2YAfxLkieOdMwaGcOZC8A9wGvb34gXAp9NstnIRqyRMsy50OtfgYtHKkaNjjUwF74BHFNVz6Dzd+KPqzK+ST/pkWYAv66qm6vqAeA7wEH96hwEnNxenwH8fZK08u9U1f1VdQvw69af1k2rPReq6uqquq2VLwImJnn8qEStkTCc3wskeRlwM525oHXbcObCTGBBVc0HqKo7qmr5KMWtNW84c6GAjVsieEPgAeCvoxO2RshK50NVLa6qBcCD/doeAPy4qv5UVX8Gfkwn4aN102rPhaq6oapubK9vo/M/9luOTtgaAcP5vUCS3YAnA+ePRrAaUas9F1pycP2q+nGrd3dV3bMqg5v0kx7pKcDv+ry/tZUNWKeqlgF3AlsMsa3WHcOZC329Ari6qu4foTg18lZ7LiTZGHg/8LFRiFMjbzi/F3YEKsl5bfvG+0YhXo2c4cyFM4C/Ab+n8y/7x1bVn0Y6YI2o4XwG9PPj+LJGfp5JZgCPA25aQ3Fp9K32XEgyAfh/wHtHIC6NvuH8XtgR+EuSM5NcneSYVd1NuP6qVJYeAzJAWf9HXA9WZyhtte4YzlzoHEyeRWc718w1GJdG33DmwseA46rq7rbwT+u24cyF9YHnAdPpbOG6IMlVVXXBmg1Ro2Q4c2EGsBzYGngicGmSn1TVzWs2RI2i4XwG9PPj+DLsn2eSrejcGuZ1VfWoFWBaZwxnLrwd+GFV/c7Pj+PCcObC+sBewLPp/EPhaXS2AX91qIO70k96pFuBp/Z5vw1w22B12tacTYE/DbGt1h3DmQsk2Qb4Hp17s/ivtOu24cyF3YFPJ1kMvAv4YJIjRjpgjZjh/o24uKpub9syfgjsOuIRa6QMZy68Gji3qpZW1R+By4HuEY9YI2k4nwH9/Di+DOvnmeQJwDnAh6rqF2s4No2u4cyFPYAj2ufHY4HXJjl6zYanUTTcvxFXt63By4Dvs4qfH036SY80F3h6ku2TPI7OgznO6lfnLDoPZwA4GPhpVVUrf2U6T+vbHng6cOUoxa01b7XnQrvp8jnAB6rq8lGLWCNltedCVe1VVZOrajLwWeDfq8onsK27hvM34jxglyQbtQTQ3sC1oxS31rzhzIXfAi9Ix8bAc4DrRylujYyhzIfBnAfMTPLE9gCPma1M66bVngut/veAb1TV6SMYo0bHas+FqjqsqrZtnx/fQ2dOPOqJr1pnDOdvxFzgiUl67+/5Albx86NJP6mPlj0/gs6HreuA/6yqRUk+nuSlrdpX6dyr69fAPwFHtbaLgP+k8x/hucA7vEn7ums4c6G1+zvgw0nmta//NcqnoDVkmHNB48gw/0b8GfgMnQ9v84BfVtU5o30OWjOG+XvhC8AmwDV05sPX2827tY4aynxIMj3JrXSe3vzlJIta2z/ReULn3Pb1ce/xuO4azlwADgWeD8zu8/mxawxOQ2vAMOeCxpFh/o1YTifxe0GShXS2Cn9lVcZP5x8cJUmSJEmSJI0XrvSTJEmSJEmSxhmTfpIkSZIkSdI4Y9JPkiRJkiRJGmdM+kmSJEmSJEnjjEk/SZIkSZIkaZwx6SdJkiRJkiSNMyb9JEmSJEmSpHHGpJ8kSZIkSZI0zvx/klARczrxwGwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fea_imp_df = permutation_importances(tb,X_val,y_val,metric)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even with different implementation, my Random Forest classifier is consistent with Sklearn RF in its first 10 important features."
   ]
  }
 ],
 "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": "3.6.6"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}