XGBoost on Ontario Road Quality.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><h1> eXtreme Gradient Boosting as a Multiclass Classifier on Ontario Road Quality with Random Search Hyperparameter Optimization</h1></center>\n",
    "\n",
    "<center><h4> Nyasha M, 30 Apr 2021 </h4></center><p>&nbsp;</p>\n",
    "\n",
    "**Background:** This notebook used eXtreme Gradient Boosting (XGBoost) on the *Pavement condition for provincial highways* dataset from the government of Canada (available here: https://open.canada.ca/data/en/dataset/01dfc06c-6ef0-4fd0-a010-8ba20e3c9d10). The dataset contains the readings from an Automatic Road Analyzer (ARAN) that was used on various sections of road in Ontario (e.g., freeway, local road, arterial). \n",
    "\n",
    "Each section of road pavement was automatically evaluated by the ARAN based on its distress (Distress Manifestation Index, DMI), wheel track rutting (in mm), and roughness (International Roughness Index, IRI). These indices were then combined into a Pavement Condition Index (PCI) for an overall rating of road pavement condition.\n",
    "\n",
    "**Objective:** I wanted to determine what features might predict for good or bad quality roads in Ontario. I also wanted to see if my final model could be simplified in some by selecting only those features which contributed significantly to--or improved--the model's balanced accuracy.\n",
    "\n",
    "**Methods:** The PCI scores from the dataset were changed into a multiclass feature, based on the category that a particular PCI score fell into (refer to the table below). I then used an XGBoost Multiclass Classifier to predict what class (i.e., road pavement quality category) different sections of roads in Ontario fell into. \n",
    "\n",
    "My preliminary model was followed by hyperparameter optimization via a random search, which was validated by using k-folds cross-validation, given the imbalanced frequency of road quality classes in my dataset. I used these optimized parameters to rerun another XGBoost model and compare it to my old one. I then computed the feature importance (f1-score) and balanced accuracy in a stepwise manner as I tested the incremental removal of features (from least to greatest importance) from my model. This final step was done to see if my final model could be simplified in some way without significantly sacrificing its balanced accuracy.\n",
    "\n",
    "| Category      | PCI score |\n",
    "| :---------- | :----------- |\n",
    "| **Good**      | 85 - 100       |\n",
    "| **Satisfactory**   | 70 - 85        |\n",
    "| **Fair**      | 55 - 70       |\n",
    "| **Poor**   | 40 - 55        |\n",
    "| **Very poor**      | 25 - 40       |\n",
    "| **Serious**   | 10 - 25        |\n",
    "| **Failed**      | 0 - 10       |\n",
    "\n",
    "Refer to the guide here for more information on the PCI and its categories: https://faapaveair.faa.gov/Help/PavementConditionIndexPCI.html.\n",
    "\n",
    "### Package Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings;\n",
    "warnings.filterwarnings('ignore');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import scipy as sc\n",
    "\n",
    "# Model preprocessing\n",
    "import sklearn.preprocessing as preprocess\n",
    "from sklearn.model_selection import train_test_split\n",
    "import itertools\n",
    "\n",
    "# Modeling and model metrics\n",
    "import xgboost\n",
    "from xgboost import XGBClassifier, plot_tree\n",
    "from sklearn.compose import make_column_transformer\n",
    "from sklearn.metrics import balanced_accuracy_score, cohen_kappa_score, classification_report, plot_confusion_matrix, make_scorer \n",
    "from sklearn.model_selection import StratifiedKFold, RandomizedSearchCV, GridSearchCV, cross_val_score\n",
    "from sklearn.feature_selection import SelectFromModel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Change plotting style.\n",
    "#plt.style.use('default')\n",
    "plt.style.use('bmh') "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Exploration \n",
    "Note: Univariate plotting was not done on all variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SectionID</th>\n",
       "      <th>Highway</th>\n",
       "      <th>DIR</th>\n",
       "      <th>LHRS</th>\n",
       "      <th>Offset</th>\n",
       "      <th>FROM_Distance</th>\n",
       "      <th>TO_Distance</th>\n",
       "      <th>PCI</th>\n",
       "      <th>DMI</th>\n",
       "      <th>IRI</th>\n",
       "      <th>Pave_Type</th>\n",
       "      <th>Pavement_Section_From</th>\n",
       "      <th>Pavement_Section_To</th>\n",
       "      <th>FUNC_CLASS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>QEW</td>\n",
       "      <td>E</td>\n",
       "      <td>10004.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.23</td>\n",
       "      <td>4.658</td>\n",
       "      <td>86.53</td>\n",
       "      <td>9.10</td>\n",
       "      <td>1.22</td>\n",
       "      <td>AC</td>\n",
       "      <td>FORT ERIE - GILMORE RD IC-5</td>\n",
       "      <td>FORT ERIE - CENTRAL AV IC</td>\n",
       "      <td>FWY</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>QEW</td>\n",
       "      <td>W</td>\n",
       "      <td>10004.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.23</td>\n",
       "      <td>4.658</td>\n",
       "      <td>84.26</td>\n",
       "      <td>8.74</td>\n",
       "      <td>1.09</td>\n",
       "      <td>AC</td>\n",
       "      <td>FORT ERIE - CENTRAL AV IC</td>\n",
       "      <td>FORT ERIE - GILMORE RD IC-5</td>\n",
       "      <td>FWY</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SectionID Highway   DIR     LHRS  Offset   FROM_Distance  TO_Distance  \\\n",
       "0        NaN      NaN  NaN      NaN      NaN            NaN          NaN   \n",
       "1        NaN      QEW    E  10004.0      0.0           0.23        4.658   \n",
       "2        NaN      QEW    W  10004.0      0.0           0.23        4.658   \n",
       "\n",
       "     PCI   DMI   IRI Pave_Type        Pavement_Section_From  \\\n",
       "0    NaN   NaN   NaN       NaN                          NaN   \n",
       "1  86.53  9.10  1.22        AC  FORT ERIE - GILMORE RD IC-5   \n",
       "2  84.26  8.74  1.09        AC    FORT ERIE - CENTRAL AV IC   \n",
       "\n",
       "           Pavement_Section_To FUNC_CLASS  \n",
       "0                          NaN        NaN  \n",
       "1    FORT ERIE - CENTRAL AV IC        FWY  \n",
       "2  FORT ERIE - GILMORE RD IC-5        FWY  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"open_data_submission_2017_pavement_final.csv\")\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x360 with 8 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.loc[:,'LHRS':'IRI'].hist(figsize=(13,5), color='khaki', edgecolor='black', layout=(2,4)), plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check if our outcome variable of interest--PCI--might have any missing values. Let's also explore how some of the categorical data looks like."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    1828\n",
       "True        1\n",
       "Name: PCI, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.PCI.isna().value_counts() # only one missing value under the PCI column."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now the categorical features:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of unique observations in Pavement_Section:  1485\n",
      "Number of unique observations in Highway:  243\n",
      "\n",
      "DIR, Pave_Type, and FUNC_CLASS: \n",
      "------------------------------------ \n",
      "B    1175\n",
      "W     187\n",
      "E     186\n",
      "N     140\n",
      "S     140\n",
      "Name: DIR, dtype: int64 \n",
      "\n",
      "AC    1422\n",
      "ST     287\n",
      "CO      85\n",
      "PC      34\n",
      "Name: Pave_Type, dtype: int64 \n",
      "\n",
      "ART    639\n",
      "FWY    626\n",
      "LOC    417\n",
      "COL    146\n",
      "Name: FUNC_CLASS, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\"Number of unique observations in Pavement_Section: \", len(df.Pavement_Section_From.unique()))\n",
    "print(\"Number of unique observations in Highway: \", len(df['Highway '].unique()))\n",
    "\n",
    "print(f\"\\nDIR, Pave_Type, and FUNC_CLASS: \\n------------------------------------\",\n",
    "      f\"\\n{df.DIR.value_counts()} \\n\\n{df.Pave_Type.value_counts()} \\n\\n{df.FUNC_CLASS.value_counts()}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to run XGBoost as a classifier, we need our outcome to be continuous. We were thinking of dividing PCI into categories based on the PCI score--what might be some plausible categories/divisions? \n",
    "\n",
    "**Note:** I tested different cut-offs ahead of time and decided to group any PCI value that was equivalent to 'very poor' or worse together, due to the very small number of PCI scores that fell into the categories of 'failure', 'serious', or 'very poor' road conditions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>&lt; poor</th>\n",
       "      <th>poor</th>\n",
       "      <th>fair</th>\n",
       "      <th>satisfactory</th>\n",
       "      <th>good</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>counts:</td>\n",
       "      <td>27</td>\n",
       "      <td>107</td>\n",
       "      <td>281</td>\n",
       "      <td>702</td>\n",
       "      <td>711</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         < poor  poor  fair  satisfactory  good\n",
       "counts:      27   107   281           702   711"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "PCI_group_counts = [len(df.PCI[df.PCI<=40]), len(df.PCI[(df.PCI<=55) & (df.PCI>40)]), len(df.PCI[(df.PCI<=70) & (df.PCI>55)]),\n",
    "                      len(df.PCI[(df.PCI<=85) & (df.PCI>70)]), len(df.PCI[(df.PCI<=100) & (df.PCI>85)])]\n",
    "pd.DataFrame(data=[PCI_group_counts], columns=['< poor', 'poor', 'fair', 'satisfactory', 'good'], index=['counts:'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data preprocessing\n",
    "Let's first create a variable called `road_status` to categorize the pavement quality of each road, based on its PCI score. \n",
    "\n",
    "We'll also remove some arbitrary features as well as features that have a massive number of unique identifiers. These would be useless/impractical to turn into value labels for our model. Finally, we'll then remove `PCI` given that its information would be stored (in the form of categories) in `road_status`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['road_status'] = np.nan\n",
    "df.loc[df.PCI<=40, 'road_status'] = 'very poor'\n",
    "df.loc[(df.PCI<=55) & (df.PCI>40), 'road_status'] = 'poor'\n",
    "df.loc[(df.PCI<=70) & (df.PCI>55), 'road_status'] = 'fair'\n",
    "df.loc[(df.PCI<=85) & (df.PCI>70), 'road_status'] = 'satisfactory'\n",
    "df.loc[(df.PCI<=100) & (df.PCI>85), 'road_status'] = 'very good'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 'SectionID' unlikely to be useful. 'Highway' has too many unique values for encoding (see above).\n",
    "drop_col = ['SectionID', 'Pavement_Section_From', 'Pavement_Section_To', 'PCI', 'Highway ']\n",
    "df = df.drop(drop_col, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Are there any missing values in the dataset and if so, where are they? OneHotEncoding won't work with NaN, so we must remove these."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>DIR</th>\n",
       "      <th>LHRS</th>\n",
       "      <th>Offset</th>\n",
       "      <th>FROM_Distance</th>\n",
       "      <th>TO_Distance</th>\n",
       "      <th>DMI</th>\n",
       "      <th>IRI</th>\n",
       "      <th>Pave_Type</th>\n",
       "      <th>FUNC_CLASS</th>\n",
       "      <th>road_status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   DIR  LHRS  Offset   FROM_Distance  TO_Distance  DMI  IRI Pave_Type  \\\n",
       "0  NaN   NaN      NaN            NaN          NaN  NaN  NaN       NaN   \n",
       "\n",
       "  FUNC_CLASS road_status  \n",
       "0        NaN         NaN  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.isna().any(axis=1)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Remove NaN row and then run OneHotEncoding."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df.isna().any(axis=1)==False]\n",
    "df = df.reset_index(drop=True) # Reset index so it starts back at 0, for (later) proper concatentation of dataframes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "enc = preprocess.OneHotEncoder()\n",
    "trans = enc.fit_transform(df[['DIR', 'Pave_Type', 'FUNC_CLASS']]).toarray()\n",
    "trans_df = pd.DataFrame(trans, columns = enc.get_feature_names(['DIR', 'Pave_Type', 'FUNC_CLASS']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Concatenate the old and HotEncoded-dataframes together."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_merged = pd.concat([df, trans_df], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, convert our categorical outcome into numbers instead of strings, then drop the old categories."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>LHRS</th>\n",
       "      <th>Offset</th>\n",
       "      <th>FROM_Distance</th>\n",
       "      <th>TO_Distance</th>\n",
       "      <th>DMI</th>\n",
       "      <th>IRI</th>\n",
       "      <th>road_status</th>\n",
       "      <th>DIR_B</th>\n",
       "      <th>DIR_E</th>\n",
       "      <th>DIR_N</th>\n",
       "      <th>DIR_S</th>\n",
       "      <th>DIR_W</th>\n",
       "      <th>Pave_Type_AC</th>\n",
       "      <th>Pave_Type_CO</th>\n",
       "      <th>Pave_Type_PC</th>\n",
       "      <th>Pave_Type_ST</th>\n",
       "      <th>FUNC_CLASS_ART</th>\n",
       "      <th>FUNC_CLASS_COL</th>\n",
       "      <th>FUNC_CLASS_FWY</th>\n",
       "      <th>FUNC_CLASS_LOC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>10004.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.230</td>\n",
       "      <td>4.658</td>\n",
       "      <td>9.10</td>\n",
       "      <td>1.22</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>10004.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.230</td>\n",
       "      <td>4.658</td>\n",
       "      <td>8.74</td>\n",
       "      <td>1.09</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>10014.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.658</td>\n",
       "      <td>13.227</td>\n",
       "      <td>7.09</td>\n",
       "      <td>1.12</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>10014.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.658</td>\n",
       "      <td>13.227</td>\n",
       "      <td>7.11</td>\n",
       "      <td>1.05</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      LHRS  Offset   FROM_Distance  TO_Distance   DMI   IRI  road_status  \\\n",
       "0  10004.0      0.0          0.230        4.658  9.10  1.22            3   \n",
       "1  10004.0      0.0          0.230        4.658  8.74  1.09            2   \n",
       "2  10014.0      0.0          4.658       13.227  7.09  1.12            0   \n",
       "3  10014.0      0.0          4.658       13.227  7.11  1.05            0   \n",
       "\n",
       "   DIR_B  DIR_E  DIR_N  DIR_S  DIR_W  Pave_Type_AC  Pave_Type_CO  \\\n",
       "0    0.0    1.0    0.0    0.0    0.0           1.0           0.0   \n",
       "1    0.0    0.0    0.0    0.0    1.0           1.0           0.0   \n",
       "2    0.0    1.0    0.0    0.0    0.0           1.0           0.0   \n",
       "3    0.0    0.0    0.0    0.0    1.0           1.0           0.0   \n",
       "\n",
       "   Pave_Type_PC  Pave_Type_ST  FUNC_CLASS_ART  FUNC_CLASS_COL  FUNC_CLASS_FWY  \\\n",
       "0           0.0           0.0             0.0             0.0             1.0   \n",
       "1           0.0           0.0             0.0             0.0             1.0   \n",
       "2           0.0           0.0             0.0             0.0             1.0   \n",
       "3           0.0           0.0             0.0             0.0             1.0   \n",
       "\n",
       "   FUNC_CLASS_LOC  \n",
       "0             0.0  \n",
       "1             0.0  \n",
       "2             0.0  \n",
       "3             0.0  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "le = preprocess.LabelEncoder()\n",
    "df_merged[['road_status']] = df[['road_status']].apply(le.fit_transform)\n",
    "\n",
    "df_merged = df_merged.drop(['DIR', 'Pave_Type', 'FUNC_CLASS'],axis=1)\n",
    "df_merged.head(4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train-test Splits\n",
    "Let's select the 'x' and 'y' features to put into the model. 'y' is our outcome variable--`road status`--which stores the categorized `PCI` scores."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create train and test splits of the data.\n",
    "x, y = df_merged.loc[:, df_merged.columns!='road_status'], df_merged['road_status']\n",
    "seed = 100\n",
    "test_size = 0.20\n",
    "\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=test_size, random_state=seed)\n",
    "\n",
    "eval_set = [(x_train, y_train), (x_test, y_test)] # Set our evaluation set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Running an initial XGBoost Model\n",
    "Let's run an initial model to see how well the model on the default settings performs on predicting poor/fair/good road pavement quality. We'll also set an early stopping round value to prevent overfitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {# Defaults\n",
    "    'booster':'gbtree',\n",
    "    'max_depth':6,\n",
    "    'min_child_weight': 1,\n",
    "    'eta':.3,\n",
    "    'tree_method':'exact',\n",
    "    \n",
    "    # Type of model, number of classes in outcome variable\n",
    "    'objective':'multi:softmax', 'num_class':5,\n",
    "    \n",
    "    # evaluation metrics for choosing the best model \n",
    "    'eval_metric':['mlogloss', 'merror']}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "stop_after = 40 # set number of early stopping rounds to prevent overfitting. Can adjust to test."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_orig = XGBClassifier(n_estimators=1000, **params, use_label_encoder=False)\n",
    "# Make equal to '_' to suppress output from .fit\n",
    "_ = xgb_orig.fit(x_train, y_train, early_stopping_rounds=stop_after, eval_set=eval_set, eval_metric=[\"merror\",\"mlogloss\"], verbose=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extract the model's evaluation results from XGBoost. Plot the model's multiclass log loss and misclassification error."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_results, test_results = xgb_orig.evals_result_['validation_0'], xgb_orig.evals_result_['validation_1']\n",
    "plot_eval_metrics=['mlogloss', 'merror']\n",
    "\n",
    "plt.style.use('bmh') \n",
    "fig, axs = plt.subplots(1,2, figsize=(16,4))\n",
    "for m in enumerate(plot_eval_metrics):\n",
    "    axs[m[0]].plot(range(0,len(train_results[m[1]])), train_results[m[1]], color='sienna', label='Training set')\n",
    "    axs[m[0]].plot(range(0,len(test_results[m[1]])), test_results[m[1]], color='midnightblue', label='Test set')\n",
    "\n",
    "    axs[m[0]].set_title(f'XGBoost {m[1]}'), \n",
    "    axs[m[0]].set_xlabel('gb trees'), axs[m[0]].set_ylabel(m[1])\n",
    "    axs[m[0]].legend(facecolor='white')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = [np.nan]*2 # Create array to store first round of XGBoost's results.\n",
    "cv_accuracies = [np.nan]*2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred[0] = xgb_orig.predict(x_test) # predict on test dataset.\n",
    "\n",
    "skf = StratifiedKFold(n_splits=10, shuffle = True, random_state = seed) # Do stratified\n",
    "cv_accuracies[0] = cross_val_score(xgb_orig, X=x_train, y=y_train, \n",
    "                                cv=skf.split(x_train,y_train), scoring='balanced_accuracy', n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Balanced accuracy: 89.28%\n",
      "Mean cross-validated balanced accuracy:  89.0%, SD: 4.993%\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "   very good       0.88      0.93      0.90        56\n",
      "satisfactory       0.89      0.89      0.89        18\n",
      "        fair       0.94      0.93      0.94       146\n",
      "        poor       0.96      0.96      0.96       142\n",
      "   very poor       1.00      0.75      0.86         4\n",
      "\n",
      "    accuracy                           0.94       366\n",
      "   macro avg       0.94      0.89      0.91       366\n",
      "weighted avg       0.94      0.94      0.94       366\n",
      "\n",
      "Cohen's Kappa: 0.910\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ---- Printing scores\n",
    "print(f\"Balanced accuracy: {balanced_accuracy_score(y_test, y_pred[0])*100:.2f}%\")\n",
    "print(\"Mean cross-validated balanced accuracy: \",\n",
    "      f\"{100*cv_accuracies[0].mean().round(2)}%, SD: {round(100*cv_accuracies[0].std(),3)}%\\n\")\n",
    "\n",
    "y_labs = le.inverse_transform(df_merged['road_status'].unique())\n",
    "print(classification_report(y_test, y_pred[0], target_names=y_labs, digits=2))\n",
    "\n",
    "print(f\"Cohen's Kappa: {cohen_kappa_score(y_test,y_pred[0]):.3f}\\n\")\n",
    "\n",
    "\n",
    "# ---- Plotting\n",
    "fig, axs=plt.subplots(1,2,figsize=(12,4))\n",
    "plot_confusion_matrix(xgb_orig, x_test, y_test, cmap='Blues', display_labels=y_labs, ax=axs[0], xticks_rotation=30)\n",
    "axs[0].set_title(\"Confusion Matrix of Classified Road Conditions\"), plt.grid(False)\n",
    "\n",
    "sns.kdeplot(cv_accuracies[0], ax=axs[1]), axs[1].set_ylabel('Density', fontsize=12), axs[1].set_xlabel('Balanced Accuracy')\n",
    "axs[1].set_title(\"Cross-validated balanced accuracy,\\n from 10 stratified k-folds\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plotting the importance of the features, according to the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax= plt.subplots(figsize=(12,6))\n",
    "xgboost.plot_importance(xgb_orig, ax=ax), ax.grid(alpha=.5), ax.set_facecolor('white')\n",
    "ax.tick_params(axis='both', labelsize=12)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the Average Distress Manifestation Index (DMI) and Average International Roughness Index (IRI) of the road were the best predictors of whether a road was in very poor, poor, fair, satisfactory, or very good condition in Ontario. This doesn't seem that surprising, actually, given that both make up the PCI score.\n",
    "\n",
    "| Feature   | Description |\n",
    "| :----------- | :----------- |\n",
    "| DMI      |  Average Distress Manifestation Index (DMI) of the pavement surface       |\n",
    "| IRI      |  Average International Roughness Index (IRI; m/km) of the pavement surface |\n",
    "| LHRS      |  Location reference based on Linear Highway Reference System Number |\n",
    "| To_Distance      |  End chainage in kilometre (km) of the reporting interval |\n",
    "| From_Distance      |  Start chainage in kilometre (km) of the reporting interval |\n",
    "| Offset      |  Average International Roughness Index (IRI; m/km) of the pavement surface |\n",
    "| Pave_Type      |  <dl><dt>One of the four pavement material types: </dt><dd>AC – Asphalt Concrete</dd><dd>PC – Portland Cement Concrete</dd><dd>COM – AC/PC Composite</dd><dd>ST – Surface-treated</dd></dl>|\n",
    "| DIR     |  <dl><dt>Direction; travelling direction in which the pavement condition data was collected. <br />Surveys are conducted in one direction only for undivided highways.\tOptions: </dt><dd>E – East</dd><dd>W – West</dd><dd>N – North</dd><dd>S - South</dd></dl>|\n",
    "| FUNC_CLASS      | <dl><dt>Function Class; one of the four function classes of the roadway: </dt><dd>FWY – Freeway</dd><dd>ART – Arterial Highway</dd><dd>COL – Collector</dd><dd>LOC – Local Road\tText</dd></dl>|\n",
    "\n",
    "Path that one of the decision trees in the model took:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\r\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\r\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\r\n",
       "<!-- Generated by graphviz version 2.38.0 (20140413.2041)\r\n",
       " -->\r\n",
       "<!-- Title: %3 Pages: 1 -->\r\n",
       "<svg width=\"2614pt\" height=\"566pt\"\r\n",
       " viewBox=\"0.00 0.00 2613.99 566.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\r\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(4 562)\">\r\n",
       "<title>%3</title>\r\n",
       "<polygon fill=\"white\" stroke=\"none\" points=\"-4,4 -4,-562 2609.99,-562 2609.99,4 -4,4\"/>\r\n",
       "<!-- 0 -->\r\n",
       "<g id=\"node1\" class=\"node\"><title>0</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1231.99\" cy=\"-540\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1231.99\" y=\"-536.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;8.79500008</text>\r\n",
       "</g>\r\n",
       "<!-- 1 -->\r\n",
       "<g id=\"node2\" class=\"node\"><title>1</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"689.993\" cy=\"-453\" rx=\"74.9875\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"689.993\" y=\"-449.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">IRI&lt;0.894999981</text>\r\n",
       "</g>\r\n",
       "<!-- 0&#45;&gt;1 -->\r\n",
       "<g id=\"edge1\" class=\"edge\"><title>0&#45;&gt;1</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1170.87,-529.415C1069.76,-513.558 868.217,-481.95 760.968,-465.131\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"761.312,-461.642 750.89,-463.55 760.227,-468.557 761.312,-461.642\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1028.49\" y=\"-492.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 2 -->\r\n",
       "<g id=\"node3\" class=\"node\"><title>2</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1576.99\" cy=\"-453\" rx=\"70.3881\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1576.99\" y=\"-449.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">IRI&lt;1.65499997</text>\r\n",
       "</g>\r\n",
       "<!-- 0&#45;&gt;2 -->\r\n",
       "<g id=\"edge2\" class=\"edge\"><title>0&#45;&gt;2</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1282.53,-526.549C1345.28,-511.088 1452.12,-484.765 1518.4,-468.436\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1519.4,-471.794 1528.28,-466.003 1517.73,-464.997 1519.4,-471.794\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1433.49\" y=\"-492.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 3 -->\r\n",
       "<g id=\"node4\" class=\"node\"><title>3</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"248.993\" cy=\"-366\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"248.993\" y=\"-362.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;8.44000053</text>\r\n",
       "</g>\r\n",
       "<!-- 1&#45;&gt;3 -->\r\n",
       "<g id=\"edge3\" class=\"edge\"><title>1&#45;&gt;3</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M633.692,-441.148C552.389,-425.478 402.376,-396.564 315.651,-379.848\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"316.137,-376.377 305.655,-377.921 314.812,-383.251 316.137,-376.377\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"530.493\" y=\"-405.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 4 -->\r\n",
       "<g id=\"node5\" class=\"node\"><title>4</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"689.993\" cy=\"-366\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"689.993\" y=\"-362.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;8.75500011</text>\r\n",
       "</g>\r\n",
       "<!-- 1&#45;&gt;4 -->\r\n",
       "<g id=\"edge4\" class=\"edge\"><title>1&#45;&gt;4</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M689.993,-434.799C689.993,-423.163 689.993,-407.548 689.993,-394.237\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"693.493,-394.175 689.993,-384.175 686.493,-394.175 693.493,-394.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"697.493\" y=\"-405.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 5 -->\r\n",
       "<g id=\"node18\" class=\"node\"><title>5</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1576.99\" cy=\"-366\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1576.99\" y=\"-362.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;8.93499947</text>\r\n",
       "</g>\r\n",
       "<!-- 2&#45;&gt;5 -->\r\n",
       "<g id=\"edge17\" class=\"edge\"><title>2&#45;&gt;5</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1576.99,-434.799C1576.99,-423.163 1576.99,-407.548 1576.99,-394.237\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1580.49,-394.175 1576.99,-384.175 1573.49,-394.175 1580.49,-394.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1611.49\" y=\"-405.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 6 -->\r\n",
       "<g id=\"node19\" class=\"node\"><title>6</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2027.99\" cy=\"-366\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2027.99\" y=\"-362.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;9.48500061</text>\r\n",
       "</g>\r\n",
       "<!-- 2&#45;&gt;6 -->\r\n",
       "<g id=\"edge18\" class=\"edge\"><title>2&#45;&gt;6</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1631.84,-441.662C1714.57,-426.07 1871.1,-396.569 1960.64,-379.694\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1961.51,-383.092 1970.69,-377.801 1960.21,-376.213 1961.51,-383.092\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1837.49\" y=\"-405.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 7 -->\r\n",
       "<g id=\"node6\" class=\"node\"><title>7</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"77.9931\" cy=\"-279\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"77.9931\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.142250508</text>\r\n",
       "</g>\r\n",
       "<!-- 3&#45;&gt;7 -->\r\n",
       "<g id=\"edge5\" class=\"edge\"><title>3&#45;&gt;7</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M216.567,-349.664C203.878,-343.608 189.221,-336.544 175.993,-330 156.604,-320.407 135.189,-309.516 117.336,-300.353\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"118.848,-297.195 108.355,-295.734 115.647,-303.42 118.848,-297.195\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"210.493\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 8 -->\r\n",
       "<g id=\"node7\" class=\"node\"><title>8</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"248.993\" cy=\"-279\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"248.993\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;8.64500046</text>\r\n",
       "</g>\r\n",
       "<!-- 3&#45;&gt;8 -->\r\n",
       "<g id=\"edge6\" class=\"edge\"><title>3&#45;&gt;8</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M248.993,-347.799C248.993,-336.163 248.993,-320.548 248.993,-307.237\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"252.493,-307.175 248.993,-297.175 245.493,-307.175 252.493,-307.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"256.493\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 9 -->\r\n",
       "<g id=\"node10\" class=\"node\"><title>9</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"635.993\" cy=\"-279\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"635.993\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;8.73499966</text>\r\n",
       "</g>\r\n",
       "<!-- 4&#45;&gt;9 -->\r\n",
       "<g id=\"edge9\" class=\"edge\"><title>4&#45;&gt;9</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M679.065,-347.799C671.33,-335.622 660.826,-319.089 652.128,-305.398\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"655.037,-303.449 646.72,-296.885 649.128,-307.203 655.037,-303.449\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"701.493\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 10 -->\r\n",
       "<g id=\"node11\" class=\"node\"><title>10</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"849.993\" cy=\"-279\" rx=\"70.3881\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"849.993\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">IRI&lt;1.03999996</text>\r\n",
       "</g>\r\n",
       "<!-- 4&#45;&gt;10 -->\r\n",
       "<g id=\"edge10\" class=\"edge\"><title>4&#45;&gt;10</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M719.34,-349.41C745.389,-335.571 783.612,-315.265 811.992,-300.188\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"813.644,-303.274 820.833,-295.491 810.359,-297.092 813.644,-303.274\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"787.493\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 15 -->\r\n",
       "<g id=\"node8\" class=\"node\"><title>15</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"79.9931\" cy=\"-192\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"79.9931\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.111110717</text>\r\n",
       "</g>\r\n",
       "<!-- 8&#45;&gt;15 -->\r\n",
       "<g id=\"edge7\" class=\"edge\"><title>8&#45;&gt;15</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M216.538,-262.722C203.845,-256.673 189.193,-249.6 175.993,-243 157.131,-233.57 136.359,-222.799 118.975,-213.676\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"120.18,-210.354 109.7,-208.795 116.92,-216.549 120.18,-210.354\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"210.493\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 16 -->\r\n",
       "<g id=\"node9\" class=\"node\"><title>16</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"248.993\" cy=\"-192\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"248.993\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.361243784</text>\r\n",
       "</g>\r\n",
       "<!-- 8&#45;&gt;16 -->\r\n",
       "<g id=\"edge8\" class=\"edge\"><title>8&#45;&gt;16</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M248.993,-260.799C248.993,-249.163 248.993,-233.548 248.993,-220.237\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"252.493,-220.175 248.993,-210.175 245.493,-220.175 252.493,-220.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"256.493\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 17 -->\r\n",
       "<g id=\"node12\" class=\"node\"><title>17</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"419.993\" cy=\"-192\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"419.993\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.177018344</text>\r\n",
       "</g>\r\n",
       "<!-- 9&#45;&gt;17 -->\r\n",
       "<g id=\"edge11\" class=\"edge\"><title>9&#45;&gt;17</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M598.369,-263.194C561.932,-248.856 506.566,-227.068 467.192,-211.574\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"468.135,-208.184 457.548,-207.779 465.572,-214.697 468.135,-208.184\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"575.493\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 18 -->\r\n",
       "<g id=\"node13\" class=\"node\"><title>18</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"635.993\" cy=\"-192\" rx=\"120.479\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"635.993\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">FROM_Distance&lt;356.510498</text>\r\n",
       "</g>\r\n",
       "<!-- 9&#45;&gt;18 -->\r\n",
       "<g id=\"edge12\" class=\"edge\"><title>9&#45;&gt;18</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M635.993,-260.799C635.993,-249.163 635.993,-233.548 635.993,-220.237\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"639.493,-220.175 635.993,-210.175 632.493,-220.175 639.493,-220.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"643.493\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 19 -->\r\n",
       "<g id=\"node16\" class=\"node\"><title>19</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"849.993\" cy=\"-192\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"849.993\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.279655546</text>\r\n",
       "</g>\r\n",
       "<!-- 10&#45;&gt;19 -->\r\n",
       "<g id=\"edge15\" class=\"edge\"><title>10&#45;&gt;19</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M849.993,-260.799C849.993,-249.163 849.993,-233.548 849.993,-220.237\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"853.493,-220.175 849.993,-210.175 846.493,-220.175 853.493,-220.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"884.493\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 20 -->\r\n",
       "<g id=\"node17\" class=\"node\"><title>20</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1020.99\" cy=\"-192\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1020.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.145590574</text>\r\n",
       "</g>\r\n",
       "<!-- 10&#45;&gt;20 -->\r\n",
       "<g id=\"edge16\" class=\"edge\"><title>10&#45;&gt;20</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M882.057,-262.837C894.829,-256.743 909.64,-249.606 922.993,-243 942.383,-233.407 963.797,-222.516 981.65,-213.353\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"983.339,-216.42 990.632,-208.734 980.138,-210.195 983.339,-216.42\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"960.493\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 29 -->\r\n",
       "<g id=\"node14\" class=\"node\"><title>29</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"547.993\" cy=\"-105\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"547.993\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.124613091</text>\r\n",
       "</g>\r\n",
       "<!-- 18&#45;&gt;29 -->\r\n",
       "<g id=\"edge13\" class=\"edge\"><title>18&#45;&gt;29</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M618.185,-173.799C605.013,-161.076 586.916,-143.596 572.397,-129.572\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"574.8,-127.027 565.176,-122.597 569.937,-132.062 574.8,-127.027\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"632.493\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 30 -->\r\n",
       "<g id=\"node15\" class=\"node\"><title>30</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"722.993\" cy=\"-105\" rx=\"79.0865\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"722.993\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.0332168303</text>\r\n",
       "</g>\r\n",
       "<!-- 18&#45;&gt;30 -->\r\n",
       "<g id=\"edge14\" class=\"edge\"><title>18&#45;&gt;30</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M653.599,-173.799C666.502,-161.192 684.184,-143.916 698.47,-129.959\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"701.299,-132.089 706.006,-122.597 696.407,-127.082 701.299,-132.089\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"692.493\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 11 -->\r\n",
       "<g id=\"node20\" class=\"node\"><title>11</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1452.99\" cy=\"-279\" rx=\"70.3881\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1452.99\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">IRI&lt;1.13499999</text>\r\n",
       "</g>\r\n",
       "<!-- 5&#45;&gt;11 -->\r\n",
       "<g id=\"edge19\" class=\"edge\"><title>5&#45;&gt;11</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1552.87,-348.867C1544.07,-342.96 1534.07,-336.203 1524.99,-330 1511.62,-320.858 1496.92,-310.674 1484.31,-301.896\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1486.16,-298.919 1475.96,-296.073 1482.16,-304.662 1486.16,-298.919\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1559.49\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 12 -->\r\n",
       "<g id=\"node21\" class=\"node\"><title>12</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1623.99\" cy=\"-279\" rx=\"70.3881\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1623.99\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">IRI&lt;1.45500004</text>\r\n",
       "</g>\r\n",
       "<!-- 5&#45;&gt;12 -->\r\n",
       "<g id=\"edge20\" class=\"edge\"><title>5&#45;&gt;12</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1586.5,-347.799C1593.17,-335.738 1602.21,-319.403 1609.73,-305.79\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1612.88,-307.33 1614.66,-296.885 1606.75,-303.943 1612.88,-307.33\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1611.49\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 13 -->\r\n",
       "<g id=\"node32\" class=\"node\"><title>13</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2027.99\" cy=\"-279\" rx=\"61.99\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2027.99\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">LHRS&lt;61405</text>\r\n",
       "</g>\r\n",
       "<!-- 6&#45;&gt;13 -->\r\n",
       "<g id=\"edge31\" class=\"edge\"><title>6&#45;&gt;13</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M2027.99,-347.799C2027.99,-336.163 2027.99,-320.548 2027.99,-307.237\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"2031.49,-307.175 2027.99,-297.175 2024.49,-307.175 2031.49,-307.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2062.49\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 14 -->\r\n",
       "<g id=\"node33\" class=\"node\"><title>14</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2324.99\" cy=\"-279\" rx=\"70.3881\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2324.99\" y=\"-275.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">IRI&lt;3.17999983</text>\r\n",
       "</g>\r\n",
       "<!-- 6&#45;&gt;14 -->\r\n",
       "<g id=\"edge32\" class=\"edge\"><title>6&#45;&gt;14</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M2074.4,-351.719C2127.34,-336.567 2213.84,-311.81 2270.09,-295.712\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"2271.23,-299.026 2279.89,-292.91 2269.31,-292.296 2271.23,-299.026\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2201.49\" y=\"-318.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 21 -->\r\n",
       "<g id=\"node22\" class=\"node\"><title>21</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1236.99\" cy=\"-192\" rx=\"120.479\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1236.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">FROM_Distance&lt;15.8225002</text>\r\n",
       "</g>\r\n",
       "<!-- 11&#45;&gt;21 -->\r\n",
       "<g id=\"edge21\" class=\"edge\"><title>11&#45;&gt;21</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1416.35,-263.581C1380.96,-249.652 1327.05,-228.439 1287.57,-212.905\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1288.46,-209.49 1277.87,-209.085 1285.89,-216.004 1288.46,-209.49\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1392.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 22 -->\r\n",
       "<g id=\"node23\" class=\"node\"><title>22</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1452.99\" cy=\"-192\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1452.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.169456869</text>\r\n",
       "</g>\r\n",
       "<!-- 11&#45;&gt;22 -->\r\n",
       "<g id=\"edge22\" class=\"edge\"><title>11&#45;&gt;22</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1452.99,-260.799C1452.99,-249.163 1452.99,-233.548 1452.99,-220.237\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1456.49,-220.175 1452.99,-210.175 1449.49,-220.175 1456.49,-220.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1460.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 23 -->\r\n",
       "<g id=\"node26\" class=\"node\"><title>23</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1623.99\" cy=\"-192\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1623.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.386087984</text>\r\n",
       "</g>\r\n",
       "<!-- 12&#45;&gt;23 -->\r\n",
       "<g id=\"edge25\" class=\"edge\"><title>12&#45;&gt;23</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1623.99,-260.799C1623.99,-249.163 1623.99,-233.548 1623.99,-220.237\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1627.49,-220.175 1623.99,-210.175 1620.49,-220.175 1627.49,-220.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1658.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 24 -->\r\n",
       "<g id=\"node27\" class=\"node\"><title>24</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1792.99\" cy=\"-192\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1792.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">DMI&lt;9.23500061</text>\r\n",
       "</g>\r\n",
       "<!-- 12&#45;&gt;24 -->\r\n",
       "<g id=\"edge26\" class=\"edge\"><title>12&#45;&gt;24</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1656.09,-262.895C1668.86,-256.808 1683.67,-249.662 1696.99,-243 1715.85,-233.57 1736.63,-222.799 1754.01,-213.676\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1756.07,-216.549 1763.29,-208.795 1752.81,-210.354 1756.07,-216.549\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1734.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 31 -->\r\n",
       "<g id=\"node24\" class=\"node\"><title>31</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1152.99\" cy=\"-105\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1152.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.113415144</text>\r\n",
       "</g>\r\n",
       "<!-- 21&#45;&gt;31 -->\r\n",
       "<g id=\"edge23\" class=\"edge\"><title>21&#45;&gt;31</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1219.99,-173.799C1207.54,-161.192 1190.46,-143.916 1176.67,-129.959\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1178.91,-127.25 1169.39,-122.597 1173.93,-132.17 1178.91,-127.25\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1235.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 32 -->\r\n",
       "<g id=\"node25\" class=\"node\"><title>32</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1321.99\" cy=\"-105\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1321.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.365852714</text>\r\n",
       "</g>\r\n",
       "<!-- 21&#45;&gt;32 -->\r\n",
       "<g id=\"edge24\" class=\"edge\"><title>21&#45;&gt;32</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1255.23,-174.082C1261.26,-168.397 1267.97,-161.98 1273.99,-156 1282.36,-147.692 1291.36,-138.418 1299.27,-130.134\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1301.84,-132.52 1306.19,-122.862 1296.76,-127.695 1301.84,-132.52\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1296.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 33 -->\r\n",
       "<g id=\"node28\" class=\"node\"><title>33</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1625.99\" cy=\"-105\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1625.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.130181283</text>\r\n",
       "</g>\r\n",
       "<!-- 24&#45;&gt;33 -->\r\n",
       "<g id=\"edge27\" class=\"edge\"><title>24&#45;&gt;33</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1762.36,-175.41C1735.26,-161.614 1695.53,-141.392 1665.93,-126.33\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1667.21,-123.052 1656.71,-121.635 1664.04,-129.29 1667.21,-123.052\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1754.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 34 -->\r\n",
       "<g id=\"node29\" class=\"node\"><title>34</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1808.99\" cy=\"-105\" rx=\"87.1846\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1808.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">Pave_Type_CO&lt;0.5</text>\r\n",
       "</g>\r\n",
       "<!-- 24&#45;&gt;34 -->\r\n",
       "<g id=\"edge28\" class=\"edge\"><title>24&#45;&gt;34</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1796.23,-173.799C1798.42,-162.163 1801.36,-146.548 1803.87,-133.237\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1807.35,-133.65 1805.76,-123.175 1800.47,-132.355 1807.35,-133.65\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1809.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 39 -->\r\n",
       "<g id=\"node30\" class=\"node\"><title>39</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1636.99\" cy=\"-18\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1636.99\" y=\"-14.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.368461013</text>\r\n",
       "</g>\r\n",
       "<!-- 34&#45;&gt;39 -->\r\n",
       "<g id=\"edge29\" class=\"edge\"><title>34&#45;&gt;39</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1776.34,-88.2219C1764,-82.2328 1749.83,-75.3305 1736.99,-69 1717.28,-59.2786 1695.43,-48.3725 1677.2,-39.2327\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1678.54,-35.9865 1668.03,-34.6298 1675.4,-42.243 1678.54,-35.9865\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1771.49\" y=\"-57.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 40 -->\r\n",
       "<g id=\"node31\" class=\"node\"><title>40</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1809.99\" cy=\"-18\" rx=\"79.0865\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1809.99\" y=\"-14.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.0221580528</text>\r\n",
       "</g>\r\n",
       "<!-- 34&#45;&gt;40 -->\r\n",
       "<g id=\"edge30\" class=\"edge\"><title>34&#45;&gt;40</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M1809.2,-86.799C1809.33,-75.1626 1809.52,-59.5479 1809.67,-46.2368\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"1813.17,-46.216 1809.79,-36.1754 1806.17,-46.1335 1813.17,-46.216\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1817.49\" y=\"-57.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 25 -->\r\n",
       "<g id=\"node34\" class=\"node\"><title>25</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1963.99\" cy=\"-192\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1963.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.164072752</text>\r\n",
       "</g>\r\n",
       "<!-- 13&#45;&gt;25 -->\r\n",
       "<g id=\"edge33\" class=\"edge\"><title>13&#45;&gt;25</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M2015.35,-261.207C2006.12,-248.946 1993.44,-232.114 1983,-218.239\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1985.64,-215.933 1976.83,-210.049 1980.05,-220.143 1985.64,-215.933\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2034.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 26 -->\r\n",
       "<g id=\"node35\" class=\"node\"><title>26</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2129.99\" cy=\"-192\" rx=\"70.3881\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2129.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">IRI&lt;2.49499989</text>\r\n",
       "</g>\r\n",
       "<!-- 13&#45;&gt;26 -->\r\n",
       "<g id=\"edge34\" class=\"edge\"><title>13&#45;&gt;26</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M2048.19,-261.701C2055.41,-255.836 2063.59,-249.15 2070.99,-243 2081.51,-234.268 2092.99,-224.575 2103,-216.066\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"2105.33,-218.683 2110.68,-209.536 2100.79,-213.352 2105.33,-218.683\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2094.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 27 -->\r\n",
       "<g id=\"node40\" class=\"node\"><title>27</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2324.99\" cy=\"-192\" rx=\"106.681\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2324.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">TO_Distance&lt;18.1739998</text>\r\n",
       "</g>\r\n",
       "<!-- 14&#45;&gt;27 -->\r\n",
       "<g id=\"edge39\" class=\"edge\"><title>14&#45;&gt;27</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M2324.99,-260.799C2324.99,-249.163 2324.99,-233.548 2324.99,-220.237\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"2328.49,-220.175 2324.99,-210.175 2321.49,-220.175 2328.49,-220.175\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2359.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 28 -->\r\n",
       "<g id=\"node41\" class=\"node\"><title>28</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2527.99\" cy=\"-192\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2527.99\" y=\"-188.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.105131119</text>\r\n",
       "</g>\r\n",
       "<!-- 14&#45;&gt;28 -->\r\n",
       "<g id=\"edge40\" class=\"edge\"><title>14&#45;&gt;28</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M2360.35,-263.194C2394.24,-249.004 2445.56,-227.517 2482.48,-212.056\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"2484.14,-215.155 2492.01,-208.065 2481.44,-208.699 2484.14,-215.155\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2445.49\" y=\"-231.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 35 -->\r\n",
       "<g id=\"node36\" class=\"node\"><title>35</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1983.99\" cy=\"-105\" rx=\"61.99\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1983.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">LHRS&lt;71470</text>\r\n",
       "</g>\r\n",
       "<!-- 26&#45;&gt;35 -->\r\n",
       "<g id=\"edge35\" class=\"edge\"><title>26&#45;&gt;35</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M2098.86,-175.754C2087.17,-169.826 2073.84,-162.827 2061.99,-156 2046.56,-147.108 2029.88,-136.552 2015.92,-127.436\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"2017.46,-124.26 2007.19,-121.686 2013.62,-130.107 2017.46,-124.26\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2096.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 36 -->\r\n",
       "<g id=\"node37\" class=\"node\"><title>36</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2141.99\" cy=\"-105\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2141.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.135125488</text>\r\n",
       "</g>\r\n",
       "<!-- 26&#45;&gt;36 -->\r\n",
       "<g id=\"edge36\" class=\"edge\"><title>26&#45;&gt;36</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M2132.42,-173.799C2134.06,-162.163 2136.27,-146.548 2138.15,-133.237\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"2141.64,-133.567 2139.57,-123.175 2134.7,-132.588 2141.64,-133.567\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2143.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 41 -->\r\n",
       "<g id=\"node38\" class=\"node\"><title>41</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"1982.99\" cy=\"-18\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"1982.99\" y=\"-14.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.159077048</text>\r\n",
       "</g>\r\n",
       "<!-- 35&#45;&gt;41 -->\r\n",
       "<g id=\"edge37\" class=\"edge\"><title>35&#45;&gt;41</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M1983.79,-86.799C1983.65,-75.1626 1983.47,-59.5479 1983.31,-46.2368\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"1986.81,-46.1335 1983.2,-36.1754 1979.81,-46.216 1986.81,-46.1335\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2018.49\" y=\"-57.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 42 -->\r\n",
       "<g id=\"node39\" class=\"node\"><title>42</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2153.99\" cy=\"-18\" rx=\"77.9862\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2153.99\" y=\"-14.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=&#45;0.105199307</text>\r\n",
       "</g>\r\n",
       "<!-- 35&#45;&gt;42 -->\r\n",
       "<g id=\"edge38\" class=\"edge\"><title>35&#45;&gt;42</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M2014.99,-89.3803C2028.01,-83.1797 2043.28,-75.822 2056.99,-69 2076.21,-59.4449 2097.4,-48.5579 2115.07,-39.3878\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"2116.7,-42.4847 2123.96,-34.765 2113.47,-36.2745 2116.7,-42.4847\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2093.49\" y=\"-57.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "<!-- 37 -->\r\n",
       "<g id=\"node42\" class=\"node\"><title>37</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2318.99\" cy=\"-105\" rx=\"79.0865\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2318.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.0521697439</text>\r\n",
       "</g>\r\n",
       "<!-- 27&#45;&gt;37 -->\r\n",
       "<g id=\"edge41\" class=\"edge\"><title>27&#45;&gt;37</title>\r\n",
       "<path fill=\"none\" stroke=\"#0000ff\" d=\"M2323.78,-173.799C2322.96,-162.163 2321.86,-146.548 2320.92,-133.237\"/>\r\n",
       "<polygon fill=\"#0000ff\" stroke=\"#0000ff\" points=\"2324.4,-132.904 2320.21,-123.175 2317.42,-133.397 2324.4,-132.904\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2356.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">yes, missing</text>\r\n",
       "</g>\r\n",
       "<!-- 38 -->\r\n",
       "<g id=\"node43\" class=\"node\"><title>38</title>\r\n",
       "<ellipse fill=\"none\" stroke=\"black\" cx=\"2491.99\" cy=\"-105\" rx=\"75.2868\" ry=\"18\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2491.99\" y=\"-101.3\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">leaf=0.293947279</text>\r\n",
       "</g>\r\n",
       "<!-- 27&#45;&gt;38 -->\r\n",
       "<g id=\"edge42\" class=\"edge\"><title>27&#45;&gt;38</title>\r\n",
       "<path fill=\"none\" stroke=\"#ff0000\" d=\"M2357.8,-174.702C2369.55,-168.837 2382.88,-162.151 2394.99,-156 2414.13,-146.286 2435.32,-135.381 2453,-126.24\"/>\r\n",
       "<polygon fill=\"#ff0000\" stroke=\"#ff0000\" points=\"2454.62,-129.34 2461.89,-121.636 2451.4,-123.123 2454.62,-129.34\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"2428.49\" y=\"-144.8\" font-family=\"Times New Roman,serif\" font-size=\"14.00\">no</text>\r\n",
       "</g>\r\n",
       "</g>\r\n",
       "</svg>\r\n"
      ],
      "text/plain": [
       "<graphviz.files.Source at 0x14efe1560c8>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgboost.to_graphviz(xgb_orig, num_trees=8, format='png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost Parameter Tuning and Testing\n",
    "Could our model have been improved by changing its parameters somehow? Let's test this with a list of various parameter alternatives we could have chosen. We will also use stratified k-folds sampling while running our cross-validated search to make sure that we are accounting for the imbalanced frequency of outcome classes in our dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid = {\n",
    "    'eta': [0.1, 0.3, 0.5],\n",
    "    'max_depth':[4, 5, 6],\n",
    "    'min_child_weight': [1, 1.3, 1.6],\n",
    "    'alpha':[0,1,2],\n",
    "    'n_estimators':[500, 750, 1000]}\n",
    "\n",
    "scorers = {'neg_log_loss_score': 'neg_log_loss', 'balanced_accuracy': make_scorer(balanced_accuracy_score)}\n",
    "iter_num = 80\n",
    "\n",
    "model = XGBClassifier(objective='multi:softmax')\n",
    "skf = StratifiedKFold(n_splits=10, shuffle = True, random_state = seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below are two search methods that we could use to find the ideal parameters. **Warning: `GridSearchCV` is VERY slow!**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Try grid search--warning: SLOW!.\n",
    "# gsearch = GridSearchCV(model, param_grid=param_grid, scoring=scorers, n_jobs=2,\n",
    "#                       cv=skf.split(x_train,y_train), verbose=3, \n",
    "#                                    refit='neg_log_loss_score', return_train_score=True)\n",
    "#print(gsearch.fit(x, y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[16:13:49] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.4.0/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n",
      "\n",
      "The best iteration of the model: \n",
      "\n",
      " XGBClassifier(alpha=2, base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
      "              colsample_bynode=1, colsample_bytree=1, eta=0.3, gamma=0,\n",
      "              gpu_id=-1, importance_type='gain', interaction_constraints='',\n",
      "              learning_rate=0.300000012, max_delta_step=0, max_depth=5,\n",
      "              min_child_weight=1.3, missing=nan, monotone_constraints='()',\n",
      "              n_estimators=500, n_jobs=8, num_parallel_tree=1,\n",
      "              objective='multi:softprob', random_state=0, reg_alpha=2,\n",
      "              reg_lambda=1, scale_pos_weight=None, subsample=1,\n",
      "              tree_method='exact', validate_parameters=1, verbosity=None)\n",
      "\n",
      "\n",
      "The best parameter settings for the model: \n",
      "\n",
      " {'n_estimators': 500, 'min_child_weight': 1.3, 'max_depth': 5, 'eta': 0.3, 'alpha': 2}\n"
     ]
    }
   ],
   "source": [
    "random_search = RandomizedSearchCV(model, param_distributions=param_grid, n_iter=iter_num, scoring=scorers, \n",
    "                                   n_jobs=2, cv=skf.split(x_train,y_train), verbose=False, random_state=seed,\n",
    "                                   refit='neg_log_loss_score', return_train_score=True)\n",
    "_ = random_search.fit(x_train, y_train)\n",
    "\n",
    "print(\"\\nThe best iteration of the model: \\n\\n\", random_search.best_estimator_)\n",
    "print(\"\\n\\nThe best parameter settings for the model: \\n\\n\",random_search.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Optimizing XGBoost and Re-running the Model on the Ontario Road Pavement Dataset\n",
    "Now that we have some optimized parameters, let's try re-running our first XGBoost model but with these new parameters put in place."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(alpha=2, base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1, eta=0.3,\n",
       "              eval_metric=['mlogloss', 'merror'], gamma=0, gpu_id=-1,\n",
       "              importance_type='gain', interaction_constraints='',\n",
       "              learning_rate=0.300000012, max_delta_step=0, max_depth=5,\n",
       "              min_child_weight=1.3, missing=nan, monotone_constraints='()',\n",
       "              n_estimators=500, n_jobs=8, num_class=5, num_parallel_tree=1,\n",
       "              objective='multi:softprob', random_state=0, reg_alpha=2,\n",
       "              reg_lambda=1, scale_pos_weight=None, subsample=1,\n",
       "              tree_method='exact', use_label_encoder=False, ...)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params_improved = {**params, **random_search.best_params_}\n",
    "\n",
    "eval_set = [(x_train, y_train), (x_test, y_test)]\n",
    "\n",
    "xgb_improved = XGBClassifier(**params_improved, use_label_encoder=False)\n",
    "xgb_improved.fit(x_train, y_train, early_stopping_rounds=stop_after, eval_set=eval_set, eval_metric=[\"merror\", \"mlogloss\"], verbose=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's extract the model's evaluation results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_results, test_results = xgb_improved.evals_result_['validation_0'], xgb_improved.evals_result_['validation_1']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_eval_metrics=['mlogloss', 'merror']\n",
    "\n",
    "fig, axs = plt.subplots(1,2, figsize=(16,4))\n",
    "for m in enumerate(plot_eval_metrics):\n",
    "    axs[m[0]].plot(range(0,len(train_results[m[1]])), train_results[m[1]], color='sienna', label='Training set', linewidth=1.5)\n",
    "    axs[m[0]].plot(range(0,len(test_results[m[1]])), test_results[m[1]], color='midnightblue',label='Test set', linewidth=1.5)\n",
    "\n",
    "    axs[m[0]].set_title(f'XGBoost {m[1]}'), \n",
    "    axs[m[0]].set_xlabel('gb trees'), axs[m[0]].set_ylabel(m[1])\n",
    "    axs[m[0]].legend(facecolor='white')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "y_pred[1] = xgb_improved.predict(x_test) # predict on test dataset.\n",
    "\n",
    "skf = StratifiedKFold(n_splits=10, shuffle = True, random_state = seed) # Do stratified\n",
    "cv_accuracies[1] = cross_val_score(xgb_improved, X=x_train, y=y_train, \n",
    "                                cv=skf.split(x_train,y_train), scoring='balanced_accuracy', n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Balanced accuracy: 89.14%\n",
      "Mean cross-validated balanced accuracy:  88.0%, SD: 5.956%\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "   very good       0.87      0.93      0.90        56\n",
      "satisfactory       0.89      0.89      0.89        18\n",
      "        fair       0.94      0.92      0.93       146\n",
      "        poor       0.96      0.96      0.96       142\n",
      "   very poor       1.00      0.75      0.86         4\n",
      "\n",
      "    accuracy                           0.94       366\n",
      "   macro avg       0.93      0.89      0.91       366\n",
      "weighted avg       0.94      0.94      0.94       366\n",
      "\n",
      "Cohen's Kappa: 0.906\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ---- Printing scores\n",
    "print(f\"Balanced accuracy: {balanced_accuracy_score(y_test, y_pred[1])*100:.2f}%\")\n",
    "print(\"Mean cross-validated balanced accuracy: \",\n",
    "      f\"{100*cv_accuracies[1].mean().round(2)}%, SD: {round(100*cv_accuracies[1].std(),3)}%\\n\")\n",
    "\n",
    "y_labs = le.inverse_transform(df_merged['road_status'].unique())\n",
    "print(classification_report(y_test, y_pred[1], target_names=y_labs, digits=2))\n",
    "\n",
    "print(f\"Cohen's Kappa: {cohen_kappa_score(y_test,y_pred[1]):.3f}\\n\")\n",
    "\n",
    "\n",
    "# ---- Plotting\n",
    "fig, axs=plt.subplots(1,2,figsize=(12,4))\n",
    "plot_confusion_matrix(xgb_improved, x_test, y_test, cmap='Blues', display_labels=y_labs, ax=axs[0], xticks_rotation=30)\n",
    "axs[0].tick_params(axis='both', which='major', labelsize=8)\n",
    "axs[0].set_title(\"Confusion Matrix of Classified Road Conditions\", fontsize=11), plt.grid(False)\n",
    "\n",
    "sns.kdeplot(cv_accuracies[1], ax=axs[1]), axs[1].set_ylabel('Density', fontsize=12), axs[1].set_xlabel('Balanced Accuracy')\n",
    "axs[1].set_title(\"Cross-validated balanced accuracy,\\n from 10 stratified k-folds\", fontsize=11)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax= plt.subplots(figsize=(12,6))\n",
    "xgboost.plot_importance(xgb_improved, ax=ax, importance_type ='weight'), ax.grid(alpha=.5), ax.set_facecolor('white')\n",
    "ax.tick_params(axis='both', labelsize=12)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How do these results compare to the old model?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6EAAAGoCAYAAAC3y+PLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOyde3wU5bn4v89mc08gCGKDCBSvSNB4x4rK5QBe8MLx0kJtxUt7bIs/DygeL62i9oj3u0erYlG0WKWi1doUCULVwhGsEREF8RhFVGgCCcnmutn398dMwrJkSZZMsrMzz/fz2U+y874z83yzeefZ952Zd8QYg6IoiqIoiqIoiqL0BIFkB6AoiqIoiqIoiqL4B+2EKoqiKIqiKIqiKD2GdkIVRVEURVEURVGUHkM7oYqiKIqiKIqiKEqPoZ1QRVEURVEURVEUpcfQTqiiKIqiKIqiKIrSY2gnVFEURVEURUkJRGSaiIQ7qDNERIyIjOqpuBRFSQzthCpKFxGReXayu7+dMiMiFyUpntjXjxzeT1hEpjm5zb2MY5mIPJXsOPaEiIyyP4MhyY5FURTFjYjIfiLysIiUi0iTiPxLRBaKSHE37W92nFx5ncP72Sgis53c5l7GMU9EliQ7jj0hIgPtz2B0smNRuh/thCqKM9QDvxKRQ5IdiM3bQGHM65WkRrQHRCQj2TF0F152UxRFcQIROQBYDfwA+AVwEHAm0AysFJHTumnX5eyeKx/upn11GS/nEy+7Ke2jnVBFcYZ/AO8Dd++pkojkiciDIrJZROpE5AMR+feo8udE5Lmo95fYo4KXRy17RkRe7CCeJmPMdzGvhqht/EhEykSkwR51vk9EcqPKx9tnGLeJSLWILBeR46PKy4E04Peto8f28t0uk4od2RSR0fb7M0XkHRFpAH5ulx0jIotFpNYeBX9ZRAZ34Br7N14mInNF5LcislVEqkTkv0UkICI3icgWe9v/HbNeuV3vKRHZISIVInKniASi6uSLyO/s9RtEZLWITIgqb70E7Mci8oaIhIA/YA0KAHxhly+z6x8tIn+146wVkVWxX7bsuG61/2+22fHfIyJpMfV+JSLrRKTR3t7CqLKgPer/hR33xyLyH4n8XRVFUbqRR4F0YIwx5q/GmK+MMe8ZY6YAS4F5IpIdb2URudA+49ggIv8AjujkflvayZWhqO2OF5F3RaTeztu/F5G+UeV7PIbbx/oDgZtl55nWIVF5cGCMR9sVRnHyye122UEi8ic7v2238+aITjq37mueiCwRkStF5Gs7/qdEJF1ErhCRL+1tPyFRHUQ7xz4tInfYeXKHvV52VJ10u3yzWGe114nI1Jj9GxH5fyLyBxGpBp4HNtnFb9nl5Xbd79vfB74R67vTRyLyk5jtLbPj+I2IfGfny3kS9d3GrvdDEXnf/l+ptD+/PlHlV4rIp3b5ZyJyo4gEE/nbKp1DO6GK4hwzgLNEZEx7hSIiwGvAkcAPgSLgMeAFERlnV1sKjI1abSzwL2Bc1LIxdr29wk5wjwH3AocDPwX+DXg8qloe1peCkVgj058BJVHJ9zigBfhPdo4eJ8q9wF3AMOAVETkcWA6sAI7Fcm8B3hSRrAS3fT7WF5pRwEzgBuB12+tk4BrgBhE5PWa9K4FvsPxmANOxHFt5GpgIXAQcBbwLvC4ih8Vs506szucI4L+Ac+zlx2P9rVoHHnoBLwCjgaOBvwF/lt3PqF8JfAucAPw/O6afthaKyC32Pv/H3udpQFnU+k/Z+/wPrL/3rcCdInIZiqIoScTuAJwJPGKM2dFOlTnAfsD4OOsfhXUcfQkrv94DPOhAXGOBV+1tHwGcCwwBFtn5HDo+hv871tnWe9mZK1s7Wp0lOp88KiL7Ae8AW7Hy2UhgPbBMRPZNcNvHYeXb8cBUrNz2KlbePx34if2KzRXnA33t/f8YONuOs5XbgZ9h5aoi4DnguajvOq3cjJXzjwZutH8CnIf1tzrOfp8HlGLlthHAE1iD4LHft84H9sH6PKZifWbXthaKyCV2LK/Y+xoDlGANqiPWZdPXANdj5cqrsPLmzSjOY4zRl7701YUXMA9YYv++APgACNjvDXCR/ftooAHoHbP+08Ar9u+D7XUOt99/DVwNbLHfH2yXH9JBPGGgNur1eVR5OXBFzDqn2NvtE2ebAWA78OOoZWFgWky9aUA4ZtlAe9ujo/4OBvhJO3G/ELMsE6gDzt2D7zLgqZj3ZTF1PgY+iln2IXBPzN/l7Zg6twNf278fZMd9RkydfwJP278Psev8JqbOKHv5kE78P30I3BgT159j6pQAC+zfc7EuB78mzva+D0SAw2KW3xT7d9KXvvSlr55+YQ3OGWBynPJ97PJZ9vtd8gxWp+IfMetMt9cZtYf9zraPjbUxr0y7fBlwR8w6g+ztFu9hu7HH8I3A7Jg6rXlwYMzytry6h3wyG1gZs0yAz4H/3ENc87C/q0S93wpkRC37C1DR+jewl70KLIx6v8zOS2lRy34ONNr5KMf+/Zcx+18ELI16b4C5MXV2+b7Qwf/Nq8CTMXGtianzOLAi6v1XWIMd7W0vB+v7xmkxy38KVCW7nXjxpaeXFcVZrgM+xUqST8eUHQdkAJt3DqKCvewzAGPMlyLyBTBWRFqAAqyzWzeJSBFwErDZGLOhgzj+F7g46n0YwB4lHQzcJyL3RJW3BnQQsEpEvo91tuxEoD9WJzTHXtcp3ot5fxxwkIjUxizPwup8J8KHMe+/s1+xy/rHLFsR8/5d4HoR6YV11hjg7zF1/o71d4om1q1d7M/jFqyzvt8Dgli+sX/nspj3m7E6lwDD7XUWx9nNsVif7+qY/7sg1plmRVGUZCIdlJsOyg/HOksWzTud3Pcmdr3SCKDJ/nkcMFJEprez3sFAWQLH8K7QXq48pp1cmU3iufITY0xT1PvvgPXGmMaYZcNiYzLGROePd7G+yxxov89g91y5HOsM4y7b6UyQIpKDNXB6FtYZ0gysQeq3Yqq2lysn2NvoDxxA/Fw5HOtv+CexbzGySQOyRGRfY8y/OhOv0jm0E6ooDmJ3Iu8Hfiu737cZAKrZeXlJNNFJYClWUmwB3jHG1IvI3+1lP6Bzl+LWG2M2trO89RL8q9j94A3WmVewLl2tAH6FlaSbsJJ6RxMHRNpZlh6nbijmfQCYD9zRTt3KDvYbS3PMexNnWUe3JHT05ai1TuyXpFi3eMzDGlm/FvgC64zmC+z+d26Ked9e7PG+qLXW+wHWKG9n1lEURekpPsPKHUVYZ8tiKbJ/ro+zfnvH4M7SHCdXgnXsvBMrL8XSOqg5j84dw2NpzZVtOUas+/zby0nt5cpSrLO9sVR3sN9YujNXxn4mXcmVd2Pd1nI11kB/COsS594x9ZzIlRcA7Q30b+tkrEon0U6oojjPHOBSrHsBo1mNdWYzyxizdg/rLwUewUpSpVHLxmHdExi73U5jjNkiIpuAQ40xT7ZXx77v83Csy07/Zi8byO5nDZuw76OIYiuQJiL7GWO22MuOpnOsxrrv5nNjXwOTBEbGvD8R+MYYs0NEPraXnQK8EVXnZKxLsPdEa2KM/XudAlxrjPkzgD2BwlBgT/8fsazDusx7IvBRO+Xv2z8HGWNeT2C7iqIo3Y4xZpuI/BVrhvkHze73hd4AbAHejLOJj7GuEoom9v3esBoYvodOKnTuGB4vVwIMYOc9osV0buBzNdbVVpuNMfWdqN8dHCciaVFnQ0/E8vwcq5PXCJyK9dm0ckrM+/bYU6583hjzRwCxJgw8BOv/olMYY7aKyNdYufK1dqp8jJVLhxpj3minXHEYnZhIURzGGFMD/AZrxC6apcAS4GURmSwiQ8WaDfZKEflZTL0+WDf6L41adjpWR3CvJyWyuRH4fyLyaxEpEpFDReRcEfmdXb4dazKkn4nIISJyIta9rrHJ7gtgjIgMEJF+9rL3gBrgDhE5WKxZAm/qZFy3Y13y85yIHG/PhjdGrFlhh3bBNxGKxZpF9hB7Jr+rgPsBjDGfY0188T8iMlFEDhORB7FG6fc4KzLwJdagwhki0l9EWkdv1wM/FpERYj0LbwG7J989YoypxRoRni3WDLmHiMiRInK9Xb4R69LwJ0XkJ2LNqnikiFwqIns9oKEoiuIgv8K6+mepiJwmIgeIyHEi8gesyWOm7aHDdT9wolizmx8iIpPZPf/uDTcB54jI/SJSLCIH2rHNlZ0zwXbmGP4FcJKIDBKRfnYHaiNWXpht55JRtkdnBmAfsffxioicLNYsuqNs/x844N0Z+mJNkjRMRM4EbsO6PzNkjKkDHgJuE5EL7O8CN2Cdyby9g+1WYN2XO0FEvic7Z61dj/VZHC/WJIZPYHXgE+UW4D/EmkF3mIgMF5HpItLPzqW3A7fbyw61y38kInd2sF1lL9BOqKJ0D3Ox7/NsxT67dzbwMnAf1iUlf8GaFfDzqHrfYZ3dqmHnGbY1QBXWWcKvuhKYMWY+cKG93/eAVVgTHWy2yyNYl6McaO93HvAA1uys0VwNHIOVYP9lr7sNmIJ1RnENVmf8WjqBMeYTrEtG87BmGFwHPIl1j0ZV4qZ7xcNY9/Ksxkr0j2F3Qm0ut2N7Duu+05OAScaYT/e0Ufus8PVY9wx/izWhAsAlWMfh97Bm6yvB+jwS5TfYgwtYI/CL2fUM9M9tjxux/q6lWPcM/99e7EtRFMVRjDFfYt2//r/A77By4l+x7vs70RhTsod138eaCfVHWFeDXIc1u3lXY3oL617PEViP2VqDdRytYeclq505ht+MddnoeqxcOcgYE8aaJb8/Vp5/FOv43N4tLbFxbcE681iB9X1iPdbjTQaze57uLhZi/R3ewbr8+A12zfU3YuXvB7DOMF6ENUlj7L27u2B///gV1neUTez8DjQDq9P+Flb+2mzHkBDGmKewziKfj3X/6N+xBvjDdvlt9r4ux8rx79jvyxPdl9Ixkryr3hRFUdyDWM8je8oY89tkx6IoiqIobkSsZ59uNMZc3lFdRdkTeiZUURRFURRFURRF6TG0E6ooiqIoiqIoiqL0GHo5rqIoiqIoiqIoitJj6JlQRVEURVEURVEUpcfQ54QqcVm+fLnJzs7uuKLLMcYg0plHb7kb9XAX6uEuutOjpqamYty4cft2y8YVz+GV3JkoXjmWJIofvf3oDOqdKB3lTu2EKnFJT0/n+OOPT3YYXaaiooJ+/fp1XNHlqIe7UA930Z0epaWlX3bLhhVP4pXcmSheOZYkih+9/egM6p0oHeVOvRxXiYtXRnvS09OTHYIjqIe7UA934RUPJfXxSu5MFL+2QT96+9EZ1NtptBOqxMUriTQjIyPZITiCergL9XAXXvFQUh+v5M5E8Wsb9KO3H51BvZ1GO6FKXCKRSLJDcIRQKJTsEBxBPdyFergLr3goqY9Xcmei+LUN+tHbj86g3k6jnVAlLoGAN/498vLykh2CI6iHu1APd+EVDyX18UruTBS/tkE/evvRGdTbafx5pFQ6hVeeIdvQ0JDsEBxBPdyFergLr3goqY9Xcmei+LUN+tHbj86g3k6jnVAlLl5JpOFwONkhOIJ6uAv1cBde8VBSH6/kzkTxaxv0o7cfnUG9nUY7oUpc0tLSkh2CIxQUFCQ7BEdQD3ehHu7CKx5K6uOV3Jkofm2DfvT2ozOot9NoJ1SJS0tLS7JDcISqqqpkh+AI6uEu1MNdeMVDSX28kjsTxa9t0I/efnQG9XYa7YQqcfHKNPNemVJbPdyFergLr3goqY9Xcmei+LUN+tHbj86g3k6jnVAlLl5JpMFgMNkhOIJ6uAv1cBde8VBSH6/kzkTxaxv0o7cfnUG9nUY7oUpcvPKss7q6umSH4Ajq4S7Uw114xUNJfbySOxPFr23Qj95+dAb1dhrthCpx8crkCvn5+ckOwRHUw12oh7vwioeS+ngldyaKX9ugH7396Azq7TTaCVXi4pXRXK+MXKmHu1APd+EVDyX18UruTBS/tkE/evvRGdTbabQTqsTFK88688pMherhLtTDXXjFQ0l9vJI7E8WvbdCP3n50BvV2Gu2EKnHxyiVFXnmuk3q4C/VILkOGDGHEiBEUFxdz7LHHUlBQQFlZGSNHjmxb9t577wFw9913U1xcTHFxMUVFRaSlpbFt27YkGyhexSu5M1FS9VjSVfzo7Udn8Ib3gw8+SFFREcOHD+eBBx4AYPbs2ey///5tefKNN94AoLKykjFjxjBkyBCmT5/ueCzaCU1RRKRcROpFpEZEqkTkHyJyhYgE7PJ5ImJE5OyY9R6wl0+z308TkXfa24dXRny88lwn9XAX6pF83nrrLcrKyli9ejVVVVVce+213HzzzZSVlXHrrbdy7bXXAjBr1izKysooKytjzpw5nHrqqeyzzz5Jjl5JBpo7u49UPpZ0BT96+9EZUt977dq1PPnkk7z33nt8+OGHvP7663z22WcAzJgxoy1PnnHGGQBkZWVx2223MXv27G6Jx59zDXuHs4wxS0SkN3Aq8CBwAnCJXb4BuBj4M4CIBIELgM87s/EvttVx1d1LHQ+6p+mXEaGiKfXHW9TDXahHclg1a2y7yzMzMxERduzYAUB1dTUDBgzYrd6CBQuYMmVKt8aouB7Nnd1Aqh1LnMKP3n50htT2XjVrLJ988gkjR44kJycHgFNPPZVFixbFXSc3N5dRo0axdu3abokpNf+Syi4YY6qNMX8GfghcLCJFdtFrwEki0sd+fxqwBviuc9t1PNSk0KIerkI93EWqeogIEyZM4JhjjuGJJ54gEAjwwAMPMGvWLA444ACuueYa5syZs8s6dXV1lJSUcN555yUpasVNaO50llQ9lnQVP3r70RlS37uoqIi///3vVFZWUldXxxtvvMGmTZsAeOSRRzjiiCO49NJL2b59+y7rddezj/VMqIcwxrwnIl8DJ9uLGrBGcn8EPAb8FHgW+FVnthepq2bjw9cTNkKaGIaedAbZx55FYVaEUFhoMdAr3bC1MUDfjAgBYEtjgO9lRagNW/+weUHDdw0B9suMEAEqmwL0z4ywo1lIE8gNGr5tCFCYFSEcgarmAP0yI1Q1C5kByE7bWd4UgZqw0DfDsL1JyE4zZKXRVt7QAvUtQp8MQ2WTkB80ZAQgKIZe6Yb6FqExAgXphorGAAXpEYKBneu73ckqs5xay1PRqU96hJYQu31OqeY0OKeFsur0Pf7vpYLTwXlhPq0Jdro9JdupsbGRmpoalixZwr777svmzZv54Q9/SH5+PsuWLePOO+9k/PjxlJSUcPHFF7Nw4UIKCgqoqqritddeY+TIkUQiERoaGmhoaCAcDreVZ2RkEAwGqaurIz8/n7q6Ot9eWuknNHc609YPzgvz0Y50Tzl15nNKF0N+0HjKqaPPqfV7lZecOvM5HZwXZlN9Wko6NTY2su+++zJjxgzGjh1LTk4OI0aMoLm5malTpzJr1ixCoRD33HMPV155JQ888EBbbqypqSEcDlNRUUFeXp5juVP8OotbqiMi5cDlxpglMctXYo3iHgx8DZQAd2ON5G4AhgBvAk8ZY+bZ97dcbowZFbuP+a+Vmoc+7Z7Rj54kO83qgKY66uEu1CM5tHc57uzZs8nKymLOnDlUVVUhIhhj6N27d9vluQCTJ0/mggsuYOrUqQnts7S09P1x48Yd2+XglaSjubP7SLVjiVP40duPzpDa3u3lzhtuuIGBAwfyy1/+sm1ZeXk5kyZN2uUS3CeffJIPP/yQRx55JKF9dpQ79Uyo99gfaJv20RjzjojsC/waeN0YU9/Z0+oH9s1m1awfdE+UPci2bds8MQmJergL9UgeoVCISCRCfn4+oVCIxYsXM2PGDAYMGMDy5csZPXo0S5cu5eCDD25bp7q6muXLl/Pcc88lMXLFxWju7CKpeCxxAj96+9EZvOG9detW+vfvz1dffcXLL7/MihUr+PbbbyksLARg0aJFFBUV7bJOY2Njt8SinVAPISLHYSXSd7AmWWjlOeAmYEwy4ko2Xjnbrx7uQj2Sx5YtW5g8eTIA4XCYqVOnMnbsWAoLC7nqqqsIh8NkZWXxxBNPtK2zaNEiJkyYQG5ubrLCVlyK5k5nSMVjiRP40duPzuAN7/POO4/KykrS09N59NFH6dOnDz/5yU8oKytDRBgyZAi/+93v2uoPGTKE6upqmpubeeWVV1i8eDGHH364I7FoJ9QDiEgv4BSsGf6eM8Z8FDNi+xDwNvD3RLbrlWed9erVK9khOIJ6uAv1SB5Dhw7lww8/3GVZc3Mzo0aN4v333293nWnTpjFt2rQeiE5JFTR3OksqHkucwI/efnQGb3i//fbbuy2bP39+3Prl5eU0NzeTnp7ueCw6O25q85qI1ACbgBuB+9g5xXwbxphtxphSk+AQjlcm5Kiurk52CI6gHu5CPdyFVzyUHkFzZzfg1zboR28/OoN6O42eCU1RjDFDOiiftoeyUVG/zwPmtVcvEPDGGEVWVlayQ3AE9XAX6uEuvOKhdC+aO7sPv7ZBP3r70RnU22n8eaRUFEVRFEVRFEVRkoJ2QpW4RCKRZIfgCA0NDckOwRHUw12oh7vwioeS+ngldyaKX9ugH7396Azq7TTaCVXi4pXJFXr37p3sEBxBPdyFergLr3goqY9Xcmei+LUN+tHbj86g3k6jnVAlLl6ZXCH6YfWpjHq4C/VwF17xUFIfr+TORPFrG/Sjtx+dQb2dRjuhiufp7APG3Y56uAv1cBde8VCUVMWvbdCP3n50BvV2Gu2EKnHxyiVFXnk4vXq4C/VwF17xUFIfr+TORPFrG/Sjtx+dQb2dRjuhSly8cklRTU1NskNwBPVwF+rhLrzioaQ+XsmdieLXNuhHbz86g3o7jXZClbh45Vln2dnZyQ7BEdTDXaiHu/CKh5L6eCV3Jopf26Afvf3oDOrtNP48UiqdwhiT7BAcwSvT5auHu1APd+EVDyX18UruTBS/tkE/evvRGdTbabQTqsTFK4m0sbEx2SE4gnq4C/VwjiFDhjBixAiKi4s59thj25Y//PDDHHrooQwfPpxrr70WgDfffJNjjjmGESNGcMwxx7B06VLAHR6KAt7JnYni1zboR28/OkPXve+//36GDx9OUVERU6ZM2eX5m1deeSV5eXm71H/xxRc5/PDDGT58OFOnTu3SvrtCd33ewW7ZqpJURKQcuBwYCMwF6oEI8AVwozHmdbveEHtZujEmHLsdr0yuUFBQkOwQHEE93IV6OMtbb71Fv379dnn/6quvsmbNGjIzM9m6dSsA/fr147XXXmPAgAGsXbuWiRMnsnnzZtd4KKmL5s6u4dc26EdvPzpD17w3b97MQw89xLp168jOzubCCy/khRdeYNq0aaxevZqqqqpd6n/22WfMmTOHd999lz59+rTlwGTQXZ+3dkK9zwpjzCgRCQA/A14QkYHGmKqOVvy/f9Uw/e6l3R9hNzM4p4Uv61L/S4F6uAv16DqrZo2NW/bYY49x3XXXkZmZCUD//v0BOOqoo9rqDB8+nIaGBhobG6mpqdmlE6soXcT3uTNRvHJMTBQ/evvRGfbeuzXXhcNh6uvrSU9Pp66ujgEDBtDS0sKsWbP4wx/+wKJFi9rWefLJJ/nVr35Fnz59gJ05MBlUVVV1S37Vy3F9gjEmAswHcoGDO7VOt0bUc4Q9cgm/ergL9XAOEWHChAkcc8wxPPHEEwBs2LCBt99+mxNOOIFTTz2VVatW7bben/70J4466igyMzN9e/ZJ6V78nDsTxQ3HkmTgR28/OkPXvPfff3+uueYaBg0aRGFhIb1792bChAk88sgjnH322RQWFu5Sf8OGDWzYsIGTTjqJkSNHUlJS0sXo957uyq96JtQniEgacAnQDHzZmXWaaqvZ+PD1hI2QJoahJ51B9rFnUZgVIRQWWgz0SjdsbQzQNyNCANjSGOB7WRFqw9aDbfOChu8aAuyXGSECVDYF6J8ZYUezkCaQGzR82xCgMCtCOAJVzQH6ZUaoahYyA5CdtrO8KQI1YaFvhmF7k5CdZshKo628oQXqW4Q+GYbKJiE/aMgIQHWTMDinhfoWoTECBemGisYABekRgoGd67vdqT4MhVktZETFnIpOYqBXMLLb55RqTnlBQ0bA7PF/LxWc8oKGXsFIp9uTk061tbU0NDSwbNky8vLyqKio4IILLuB73/seTU1NbNmyhddff53PPvuM888/nw8++IDc3FxqamooLy9n1qxZvPjii4TDYRoaGqiqqiIrK4va2lpyc3NpamqiubmZgoICqqqqSE9PJyMjg1AoRF5eHg0NDYTD4bbyjIwMgsEgdXV15OfnU1dX59vHbSgWfs6dibb1vKAhO814yqkzn1NNszAou8VTTh19Tq3fq7zk1JnPKS9oKMxqSdipoqICYwwLFy5kzZo19O3blylTpjB37lwWLFjAK6+8QjgcxhhDTU0NgUCAuro61q9fz8svv8zmzZs5++yzWb58Ofvttx8ADQ0N9O7dmx07diAibbkxOzubSCRCY2NjW25LS0sjJyeHmpoacnJyCIfDNDU1tZUHg8E95k5jDPX19Y7nTvHrDfReJua+lqeAWqxR3DBwsTHmRbveEPZwX8uC15eY+z5J/ZPlXrlsRD3chXp0nfYux509ezZ5eXksWbKE6667jtGjRwNw4IEHsnLlSvbdd1++/vprxo4dy+9//3tOOukkACoqKrrtctzS0tL3x40bd2zHNZVURnNn1/DKMTFR/OjtR2fo2uW4L730EiUlJcydOxeAZ599lptvvpn6+nqysrIA+Oqrrxg6dCgbN27kiiuuYOTIkUybNg2AcePGcccdd3Dcccc55tNZ9ja/dpQ79Uyo91lp39eShzXRwsnAi51Z8fv98lg1a2S3BtcT1NXVkZOTk+wwuox6uAv1cIZQKEQkEiE/P59QKMTixYu56aabyMvLY+nSpYwePZoNGzbQ1NREv379qKqq4swzz2TOnDltHVDAE5+F4ip8nzsTJdnHkmThR28/OkPXvAcNGsTKlSupq6sjOzub0tJSZs6cyZVXXtlWJy8vj40bNwJw7rnnsmDBAqZNm0ZFRQUbNmxg6NChjngkSnd91toJ9QnGmFoR+SXwuYg8bRcnyVgAACAASURBVIz5oBPr9EBk3U84vNtAdUqiHu5CPZxhy5YtTJ48uS2WqVOnctppp9HU1MSll15KUVERGRkZPPPMM4gIjzzyCBs3buS2227jtttuA2Dx4sVtI8mK4iR+zp2JkuxjSbLwo7cfnaFr3ieccALnn38+Rx99NMFgkKOOOoqf//zncetPnDiRxYsXc/jhh5OWlsbdd99N375993r/XaG7Pm/thPoIY0yliDwF3ARM7kT97g+qB2hqakp2CI6gHu5CPZxh6NChfPjhh7stz8jI4Lnnnttt+a9//Wt+/etf77a8oqKiW+JTFL/mzkRJ9rEkWfjR24/O0HXvW265hVtuuSVueW1tbdvvIsJ9993Hfffd16V9OkF3fd7aCfUfD2CN6B4B7NhTRa/MNumV51mph7tQD3fhFQ/FtfgudyaKX9ugH7396Azq7TT+u3PeBxhjhhhjlhhj5hljRsWUfW2MyTTGrDHGlBtjpL2JFQDPzAoZ+wDgVEU93IV6uAuveCjJQ3Nn1/BrG/Sjtx+dQb2dRjuhSlxEJNkhOEIw6I0T/urhLtTDXXjFQ0l9vJI7E8WvbdCP3n50BvV2Gu2EKnHxSiL1yoQl6uEu1MNdeMVDSX28kjsTxa9t0I/efnQG9XYa7YQqcYlEIskOwRGib/ROZdTDXaiHu/CKh5L6eCV3Jopf26Afvf3oDOrtNNoJVeISCHjj3yM3NzfZITiCergL9XAXXvFQUh+v5M5E8Wsb9KO3H51BvZ3Gn0dKpVN4ZZp5r0wlrh7uQj3chVc8lNTHK7kzUfzaBv3o7UdnUG+n0U6oEhevJNLm5uZkh+AI6uEu1MNdeMVDSX28kjsTxa9t0I/efnQG9XYa7YQqcfHKs8688lwn9XAX6uEuvOKhpD5eyZ2J4tc26EdvPzqDejuNdkKVuHjlWWdeea6TergL9XAXXvFQUh+v5M5E8Wsb9KO3H51BvZ1GO6FKXLwyzXx6enqyQ3AE9XAX6uEuvOKhpD5eyZ2J4tc26EdvPzqDejuNdkKVuHglkWZkZDi2rUsvvZT+/ftTVFTUtuw3v/kNRxxxBMXFxUyYMIFvvvkGgGXLltG7d2+Ki4spLi7m1ltv7dK+nfRIJurhLtRDUZzFK7kzURJtg5s2bWLMmDEMGzaM4cOH8+CDDwLxc+rdd9/dlk+LiopIS0tj27Ztjnskih+PPX50BvV2Gu2EuhARKReRf4tZNlpEvm6n7jIRuTyqTkREakWkRkTWi8glMfXPEZEyEdkhIhUiUioiQ9qLwyvPOguFQo5ta9q0aZSUlOyybNasWaxZs4aysjImTZq0S2fz5JNPpqysjLKyMm666aYu7dtJj2SiHu5CPRSvoLkzuSTaBoPBIPfeey+ffPIJK1eu5NFHH2XdunVxc+qsWbPa8umcOXM49dRT2WeffbpDJSH8eOzxozOot9MEu2WrSjL5xhgzUKyh2NOBP4vIP4wx60XkIOBZ4N+BpUAeMAFoN2N+XlnHVXcv7am4u428YITacNfGW1bNGgvAKaecQnl5+S5lvXr1avs9FAp12yh4Xl5et2y3p1EPd6EeigJo7uwyieTaVbPGUlhYSGFhIQD5+fkMGzaMzZs3c/jhh7fVi5dTFyxYwJQpU5wJvIv48djjR2dQb6fRM6EexVi8AWwDjrAXFwNfGGNK7fIaY8yfjDFftbeNNI9cUZQf7P7p8m+88UYOOOAAnn/++V3OhK5YsYIjjzyS008/nY8//rhL+2hoaOhqmK5APdyFeijKTjR37j1dybXl5eV88MEHnHDCCUD8nApQV1dHSUkJ5513XpfidQo/Hnv86Azq7TR6JtSjiEgAmAT0Azbai/8JHCYi9wN/BlYZY2rjbSMcqmbjw9cTNkKaGIaedAbZx55FYVaEUFhoMdAr3bC1MUDfjAgBYEtjgO9lRagNW1k4L2j4riHAfpkRIkBlU4D+mRF2NAtpArlBw7cNAQqzIoQjUNUcoF9mhKpmITMA2Wk7y5siUBMW+mYYtjcJ2WmGrDTayhtaoL5F6JNhqGwS8oOGjAAExTA4p4X6FqExAgXphorGAAXpEYKBnevvyamiooKsrCwAtm3bhjGGyspKRITc3FxmzJjBDTfcwN13381dd93FnDlzGDx4MGVlZfTv35+XX36Zc845h3/+8580NTVRUFBAVVUVwWCQrKwsamtryc3Npampiebm5rby9PR0MjIyCIVCbQ8LDofDbeUZGRkEg0Hq6urIz8+nrq6OlpaWtvLMzEwCgQD19fXk5+cTCoUwxtCrVy+qq6vbnBoaGujduzc7duxoc6qpqSE7O5tIJEJjY2PbNtPS0sjJyaGmpoacnBzC4XBCTqFQiMzMTEKhEHl5eTQ0NKSkU2VlJXl5ebt9TqnmVFlZSWZm5h7/91LBqbKyEqDT7SkRJ8U/aO7cmTtbyzubOw/IbqGquXNO27Zta2vrLS0tTJ48mVtvvZWcnBwqKiq45ppruOmmm7jtttu4//77ue6669ra+vPPP8/xxx/fVtfJtr43x69wOLzL94Huyp1uOia3tLRQUVHhKadk5xk3fB9w+ntbh8dbvz5U2c2ISDlwuTFmSdSy0ViXAe2IqZ4HXGGMeSqmTjbWIMPVxpgHorYzEpgJjAHygReA6e0l1GdfKzUPf5r6Q7oZAUNTpGserZfjgjViO2nSJNauXbtbvS+//JIzzzyz3bIhQ4awevVq+vXrt1cxhMNhgsHUHzdSD3ehHh1TWlr6/rhx447tlo0rjqG5M7kkkmtbc2pzczOTJk1i4sSJzJw5c7d67eXUyZMnc8EFFzB16lRnAu8iXjmGJoIfnUG9E6Wj3Om/v2Rq840xZmD0AhFZ1l4dEckE7gDGAm2J1BizErjQXvc44I/AjcD1sTs7qG82q2b9wFGBZFBRUbHXHb/O8Nlnn3HwwQcD8Oc//5nDDjsMgO+++4799tsPEeG9994jEonQt2/fvd5PVVVVt3r0FOrhLtRD8QGaO3uARHOtMYbLLruMYcOG7dIBjZdTAaqrq1m+fDnPPfecc4F3ET8ee/zoDOrtNNoJ9SjGmEYR+S9gvYica4x5pZ06q0TkZaBo9y14Z5p5J6eWnjJlCsuWLaOiooKBAwdyyy238MYbb7B+/XoCgQCDBw/m8ccfB2DhwoU89thjBINBsrOzeeGFF7r0N/XK1ODq4S7UQ1F2orlz70m0Db777rvMnz+fESNGUFxcDMDtt9/O3Llz282pAIsWLWLChAnk5uY6GntX8OOxx4/OoN5Oo51Q95IuIllR7xP+rIwxTSJyL3AT8IqIjAKGAa8aY7aKyGHA2cAz7a3vlUTq5KUTCxYs2G3ZZZdd1m7d6dOnM336dMf27ZVLQNTDXaiH4jE0dyaJRNvgqFGjaO+WsDPOOCPuOtOmTWPatGmJhtat+PHY40dnUG+n0dlx3csbQH3Ua/ZebudpYJCInAVUYSXOj0SkFigBFgF3tbeiV551VldXl+wQHEE93IV6uAuveChdRnNnkvBrG/Sjtx+dQb2dxp9depdjjBmSQN3RUb8vAwbGlNdhzfLXylmd3XZaWlpnq7qa/Pz8ZIfgCOrhLtTDXXjFQ9l7NHcmF7+2QT96+9EZ1Ntp9EyoEhevjOZ6ZeRKPdyFergLr3goqY9Xcmei+LUN+tHbj86g3k6jnVAlLl55fI9XnvOnHu5CPdyFVzyU1McruTNR/NoG/ejtR2dQb6fRTqgSF69cUlRQUJDsEBxBPdyFergLr3goqY9Xcmei+LUN+tHbj86g3k6jnVAlLl4Z8amqqkp2CI6gHu5CPdyFVzyU1McruTNR/NoG/ejtR2dQb6fRTqgSF69MM5+ZmZnsEBxBPdyFergLr3goqY9Xcmei+LUN+tHbj86g3k6jnVAlLl5JpIGAN/7N1cNdqIe78IqHkvp4JXcmil/boB+9/egM6u34drtlq4on8MoMf/X19ckOwRHUw12oh7vwioeS+ngldyaKX9ugH7396Azq7TTaCVXi4pXJFbzyXCf1cBfq4S684qGkPl7JnYni1zboR28/OoN6O412QpW4eGVyhVAolOwQHEE93IV6uAuveCipj1dyZ6L4tQ360duPzqDeTqOdUMXzdOWZbZdeein9+/enqKiobdmsWbM47LDDOOKII5g8efIus4atWbOGE088keHDhzNixAgaGhq6FHs0Xnn2nHq4C/VQFMUJYtvgpk2bGDNmDMOGDWP48OE8+OCDALz00ksMHz6cQCDA6tWr2+pXVlYyZswY8vLymD59eo/G3hX8eOzxozOot9NoJ7QHEJGPRWR0suNIFK9cUtSrV6+9XnfatGmUlJTssmz8+PGsXbuWNWvWcMghhzBnzhwAwuEwF110EY8//jgff/wxy5YtIz09vUuxR9MVDzehHu5CPRS3orkztYhtg8FgkHvvvZdPPvmElStX8uijj7Ju3TqKiop4+eWXOeWUU3apn5WVxW233cY999zTk2F3GT8ee/zoDOrtNMFu2aqLEJHaqLc5QCPQeq3MfxhjnheRw4E7gFOxOuargRuNMf/oYNtDgC+A1vPUIWAV8KAx5s3WesaY4Z2Is3Vb6caYcEf1e4L/+1cN0+9emuwwuszgnBa+rEvsS8GqWWMBOOWUUygvL9+lbMKECW2/jxw5koULFwKwePFijjjiCI488kgA+vbt24Wod6e6upp+/fo5us1koB7uQj2U9tDcufd4JXcmSnSuXTVrLIWFhRQWFgLWPWXDhg1j8+bNjB8/vt31c3NzGTVqFBs3buyxmJ3Aj8cePzqDejuN58+EGmPyWl/AV8BZUcueF5EDgXeBj4DvAwOARcBiETmxk7spsLd/JPAmsEhEpjku08N4ZX6/2nD3TZf/9NNPc/rppwOwYcMGRISJEydy9NFHc9dddzm6r6ysLEe3lyzUw12oh9Iemjv3Hq/kzkTZU64tLy/ngw8+4IQTTujBiHoGPx57/OgM6u00nj8T2glmAyuMMTdGLXtIRIYBdwKntLtWOxhjvgMeFJF04E4RedYYExGRcuByY8wSETke+B/gEKAeeN4YMxP4u72ZKvsZY+OBrcCTWAnaAH8DfmWMqQKwt/sI8FNgMFACXGyMabDLzwFuAYYC/7LXLRGR3sB9wBlY+fL3wM3GmF1mUwjXVrPx4esJGyFNDENPOoPsY8+iMCtCKCy0GOiVbtjaGKBvRoQAsKUxwPeyIm3JKC9o+K4hwH6ZESJAZVOA/pkRdjQLaQK5QcO3DQEKsyKEI1DVHKBfZoSqZiEzANlpO8ubIlATFvpmGLY3Cdlphqw02sobWqC+ReiTYahsEvKDhowANLZYI7T1LUJjBArSDRWNAQrSIwQDO9ePdmpsbCQUCmGMobm5mZaWFmprrRMDDQ0N9O7dm9mzZxOJRDj//POpqKigrq6Ot99+m7/97W8UFhYybtw4DjnkEE4//XRqamrIyckhHA7T1NREQUEBVVVVBINBsrKyqK2tJTc3l6amJpqbm9vK09PTycjIIBQKISKEw2HC4XBbeUZGBsFgkLq6OvLz86mrq6OlpaWtPDMzk0AgQH19Pfn5+W1OvXr1orq6uu3A0uq0Y8cORITc3FxqamrIzs4mEonQ2NjYts20tDRycnL22qm5uZm0tDRCoRB5eXk0NDSkpNOOHTvIysra7XNKNaft27eTlpa2x/+9VHDavn074XC40+0pESelXWajudPTubO1vLO5c0BWC3UtQt+MCJWVlW1tPRwOM3nyZG699Vays7OprKxERIhEIlRVVREKhXZp6zU1NTQ2NtLY2OhI7uzu41daWlqbU3fmTjcdk4PBIBUVFZ5ySnaeccP3Aae/t3WE+Okm2+iEFrXsO+B6Y8zvY+qOAZYA+caYujjbG0I7lwGJyFDgc+BwY8wnMYl0BfA/xpj5IpIHFBljVra3LRE5CGuE+e9AL+BPwD+NMf8Z5bMVOBdowBqVftAY87idsN8EzgdKgULb5VMReQXYAswEcoHXgbnGmN9F+y14fYm575PUP1nelctxwRrBnTRpEmvXrm1b9swzz/D4449TWlpKTk4OAC+88AIlJSXMmzcPgNtuu42srCxmzZrVdQmgoqLCE5eBqIe7UI+OKS0tfX/cuHHHdsvGUwDNnf7MnYkSezkuQHNzM5MmTWLixInMnDlzl/qjR4/mnnvu4dhjd21a8+bNY/Xq1TzyyCM9E3gX8coxNBH86AzqnSgd5U49Ewr9gG/bWf4t1uXKfYB2E+ke+Mb+uU87Zc3AQSLSzxhTAayMtxFjzEag9eaIf4nIfcDNMdUeMsZ8AyAirwHF9vLLgKej7q/ZbNfZDzgd6zKoeiAkIvcDPwd2SaRD981n1aTUv3SmubnZ0QmCSkpKuPPOO1m+fHlbBxRg4sSJ3HXXXdTV1ZGRkcHy5cuZMWOGY/vt3bu3Y9tKJurhLtRD2Us0d3o8dyZKbK41xnDZZZcxbNiw3TqgXsKPxx4/OoN6O43/hup2pwJrpDOWQqzLbbbvxTb3t39ua6fsMqzLiT4VkVUiMineRkSkv4i8ICKbRWQH8BxW4o/mu6jf64A8+/cDsEaUYxkMpAPfikiViFRhJdD+sRW9chnajh079nrdKVOmcOKJJ7J+/XoGDhzI3LlzmT59OjU1NYwfP57i4mKuuOIKAPr06cPMmTM57rjjKC4u5uijj+bMM890SqNLHm5CPdyFeih7ieZOj+fORIltg++++y7z589n6dKlFBcXU1xczBtvvMGiRYsYOHAgK1as4Mwzz2TixIlt6wwZMoSZM2cyb948Bg4cyLp163paI2H8eOzxozOot9PomVDrsqELsO7tiOZCrPtdEh3JBZiMdanP+tgCY8xnwBQRCQD/DiwUkb5Y963EMsdefoQxplJEzsW6j6UzbAIOjLO8EejnlpkEuxv7PqG9YsGCBbstu+yyy+LWv+iii7jooov2en97oisebkI93IV6KHuJ5k5lF2Lb4KhRo+I+X3Dy5MntLo+djT4V8OOxx4/OoN5Oo2dCrckHfiAi/y0i+4hIvohciTVhwX8lsiER2U9EpmNd9nO9MWa3SfJE5CIR2dcuq7IXt2BNfhDBmgihlXygFmvChf2BRG4unAtcIiLjRCQgIvuLyGHGmG+BxcC9ItLLLjtQRE6N3YBXnnWWm5ub7BAcQT3chXq4C694pBCaOz2eOxPFr23Qj95+dAb1dhrfd0Lt0dVRWLPolWPdz3IeMNEY824nN1MlIiGsqerPAC4wxjwdp+5pwMdiPYPtQeBHxpgGe9T4v4F37Ut9RmIl+aOBauAvwMsJeL0HXALcb6+/HOtyIrC+JGQA67AumVpIO5dVeeWSopqammSH4Ajq4S7Uw114xSNV0Nzp/dyZKH5tg3709qMzqLfT+Gp2XCUxVq5caUaOHJnsMLpMKBTyxOiVergL9XAX3enh99lxlcTwSu5MFK8cSxLFj95+dAb1TpSOcqfvz4Qq8fHKAEUk4o1Hh6uHu1APd+EVDyX18UruTBS/tkE/evvRGdTbabQT2gEi8mMRqW3n9XGyY+tuvJJIGxsbkx2CI6iHu1APd+EVD6+gudN/+LUN+tHbj86g3k6js+N2gDHmeeD5ZMeRDLwyuUJBQUGyQ3AE9XAX6uEuvOLhFTR3+g+/tkE/evvRGdTbafRMqBIXr0yuUFVV1XGlFEA93IV6uAuveCipj1dyZ6L4tQ360duPzqDeTqOdUCUuXnkekldGpdXDXaiHu/CKh5L6eCV3Jopf26Afvf3oDOrtNNoJVeISCHjj3yMnJyfZITiCergL9XAXXvFQUh+v5M5E8Wsb9KO3H51BvZ3Gn0dKpVN45ZIirzzXST3chXq4C694KKmPV3Jnovi1DfrR24/OoN5Oo51QJS5eGc31ysiVergL9XAXXvFQUh+v5M5E8Wsb9KO3H51BvZ3Gn0dKpVOk6jTz999/P8OHD6eoqIgpU6ZQW1tLaWkpRx99NMXFxYwaNYqNGzcmO8yECYfDyQ7BEdTDXaiHojhLqubOrtLaBjdt2sSYMWMYNmwYw4cP58EHHwRg27ZtjB8/noMPPpjx48ezfft2AJYtW0bv3r0pLi6muLiYW2+9NWkOe4Mfjz1+dAb1dhrthHYz9nPRhiY7jr0hFRPp5s2beeihh1i9ejVr166lpaWFP/7xj/ziF7/g+eefp6ysjKlTp/Lb3/422aEmTFNTU7JDcAT1cBfqobgRzZ2pR2sbDAaD3HvvvXzyySesXLmSRx99lHXr1nHHHXcwbtw4PvvsM8aNG8cdd9zRtu7JJ59MWVkZZWVl3HTTTclS2Cv8eOzxozOot9O4+jmhIlIO7AdE32AxAXgXCNnvK4DHjTF3xKw7DbgaOBDYASwCrjfGVNnls4GbgauMMQ9FrfefwP3ALcaY2XuIbTSwFKizF1UB/wDuNsasaq1njMnrhOdo4DljzMCO6vYkGyvruerupckOo9OsmjUWsEZs6uvrSU9Pp66ujoMOOggRYceOHQBUV1czYMCAZIa6V3jl+VTq4S7Uw3to7kwuqZY7nSIjYHj36nEUFhZSWFgIQH5+PsOGDWPz5s28+uqrLFu2DICLL76Y0aNHc+eddyYxYmfw47HHj86g3k6TCmdCzzLG5LW+gG/s5QX2+/OB34jI+NYVRORq4E5gFtAbGAkMBt4UkYyobW8ALo7Z30/t5Z3hGzuGfHsfnwJvi8i4hAxdSrqk3mju/vvvzzXXXMOgQYMoLCykd+/eHHfccTz11FOcccYZDBw4kPnz53PdddclO9SE8crzqdTDXaiHZ9HcmSRSMXc6QWFWZLdl5eXlfPDBB5xwwgls2bKlrXNaWFjI1q1b2+qtWLGCI488ktNPP52PP/64x2J2Aj8ee/zoDOrtNK4+E9oZjDGrReRjoBgrUfYCbgEuNcaU2NXKReRC4P+Ai4Cn7eWrgGNEZLgx5mMRGQ5k28sTicEAXwM3icg+WEn8WAARMcDBxpiNInIGcA9wANYI8/3AY8BfgUwRqbU3eQgwEHgQGAbUA38CZhpjmqK2+wusEet+wB+A6XYsiMjPgJn2djYBFxlj/ikiA4CHgVOAWuD+6NHsaJpC1Wx8+HrCRkgTw9CTziD72LMozIoQCgstBnqlG7Y2BuibESEAbGkM8L2sCLVh6zlpeUHDdw0B9suMEAEqmwL0z4ywo1lIE8gNGr5tCFCYFSEcgarmAP0yI1Q1C5kByE7bWd4UgZqw0DfDsL1JyE4zZKXRVl5dXU0oFGLhwoV88sknZGVlcfHFF/PSSy/xl7/8hZdeeomRI0dy1113cdVVV3HffffR0tJCQUEBVVVVZGZmEggEqK+vJz8/n1AohDGGXr16UV1dTVZWFgANDQ307t2bHTt2ICLk5uZSU1NDdnY2kUiExsbGtm2mpaWRk5NDTU0NOTk5hMNhmpqa2sqDwSBZWVnU1taSm5tLU1MTzc3NbeXp6elkZGQQCoVoaWmhqqqKcDjcVp6RkUEwGKSuro78/Hzq6upc79TQ0EB9fT2hUIi8vDwaGhpS0qm2trbdzymVneL976WCU21tbbc5eRHNnZo7W8sbWqC+ReiTYahsEvKDhozAzvL6FqExAgXphorGAAXpEYJR5a1OfTNaaGxsbGvrgUCAc845hzlz5hAIBDDG0Nzc3NbWASoqKjj00ENZu3Yt6enprFixgrPOOov3338/ZY5fAJWVlb46Jrd+dl5ySnaecfP3gb393tYR4uZ7F+xLii43xiyJWjYE+AJIN8aERWQk1qU9PzbGLBKR04DXgSxjTDhme88AGcaYKfYlRQcB64Dexpj/EpG7gO3AcGBjJy4p2u0yIBEZCywB8o0xoZhE+i1woTHmbRHpA3zfTm67bUtEjgHSgdVYyfCvwO+MMQ/Y5Qb4C9YXg17A+8BPjDElInIB8ABwrr3+gUAzVkJdBbwK3GFvdwnwC2PM32Id57+2xDz0aSqcLLdYNWssL730EiUlJcydOxeAZ599lnfeeYfS0lI+//xzAL766itOO+001q1bl8xwE6ahoaHt4JPKqIe7UI+OKS0tfX/cuHHHdsvGuwHNnZo7k0FeMMJbM/4NgObmZiZNmsTEiROZOXMmAIceeijLli2jsLCQb7/9ltGjR7N+/frdtjNkyBBWr15Nv379ejT+vcUrx9BE8KMzqHeidJQ7U+FM6Csi0poQlwH/af9eISKZQBZwL/CKvbwfUBGbRG2+BY6JWfYc8I6I/Br4EXASMKcL8X4DCFDAzntvWmkGDheRD40x27GSdrsYY96PelsuIr8DTsVKkK3cYd+nUyUib2GNaJcAlwN3Rd1fsxFARE4A9jXGtE49938i8iSW926J9MC+Oaya9YPOOLuGQYMGsXLlSurq6sjOzqa0tJTDDjuM6upqNmzYwCGHHMKbb77JsGHDkh1qwtTW1nri4Kce7kI9PIvmTs2dPUpFRQVgTcx02WWXMWzYsLYOKMDZZ5/NM888w3XXXcczzzzDOeecA8B3333Hfvvth4jw3nvvEYlE6Nu3b1Ic9gY/Hnv86Azq7TSp0Ak9t53RXLASpsFKrFOwRj6bsCZb6CciwXaSaaFd3oYx5isR2QjcDnxmjNnUepnIXrK/HVd7F1CfB/wauENE1gDXGWNWtLcRETkEuA/r0qQcrM/q/Zhq30X9Xge0TuRwAPB5O5sdDAwQkejY0oC324shFZ91dsIJJ3D++edz9NFHEwwGOeqoo/jFL37BYYcdxnnnnUcgEKBPnz48/fTTHW/MZeTm5iY7BEdQD3ehHp5Fc6fmzh6ltQ2+++67zJ8/nxEjRlBcXAzA7bffznXXXceFF17I3LlzGTRoEC+99BIACxcu5LHHHiMYDJKdnc0LL7xAF/+XehQ/Hnv86Azq7TSp0AmNizGmBbhXRCYDv8Qa6VwBNAL/DrzYWldEcoHTvv052gAAIABJREFUgRva2dSzWPe6XOJAWJOBfxpjYkdysUdXzxGRdGC6Hd8BWIk3lseAD4Apxpgae+bB8zsZwyasy4jaW/6FMebgzmzEzZdq74lbbrmFW265pe19dXU1kydPZvLkyUmMqus0NTWRnZ2d7DC6jHq4C/XwH5o74+Lr3NlVWtvgqFGj4v4NSktLd1s2ffp0pk+f3t3hdRt+PPb40RnU22m8Mlx3B3CtiGQZY6qxJld4WEROE5F0ewT4JawJEOa3s/4fsaavf7Gdsg4Ri/1F5Gasy3l2S9YikiEiPxaR3saYZqzJFVrv2t0C9BWR3lGr5Nt1akXkMKyJFDrLU8A1InKMHdtBIjIYeA/YISL/JSLZIpImIkUiclx7G/FKIm1ubk52CI6gHu5CPdyFVzx6GM2du6K5swv4tQ360duPzqDeTuOVTuhfsO4R+RmAMeYurGR2D1Yy+l+skcxxxpjG2JWNMfXGmCXGmPoE9ztArFn5arEmLRgBjDbGLI5T/ydY96jsAK7AmhgBY8ynwAKs+0yqxJqF7xpgKlADPImV7DuFMeYl4L+xZv2rwbrnZx979PssrPtfvsC6vOoprKn4dyMtLa2zu3Q1Xnmuk3q4C/VwF17x6GE0d+7qo7mzC/i1DfrR24/OoN5O4+rZcZXk8o9//MP84AepP7lCRUVFysyytyfUw12oh7voTo9Umx1XSS5eyZ2J4pVjSaL40duPzqDeidJR7vTKmVClG0iliQH2RHp6erJDcAT1cBfq4S684qGkPl7JnYni1zboR28/OoN6O412QveAiNwgIrXtvP6a7Nh6Aq8k0oyMjGSH4Ajq4S7Uw114xcMLaO70Ru5MFL+2QT96+9EZ1NtptBO6B4wxtxtj8tp5nZ7s2HqCSCSS7BAcIRTabbLFlEQ93IV6uAuveHgBzZ3eyJ2J4tc26EdvPzqDejuNdkKVuHjlWWd5eXkdV0oB1MNdqIe78IqHkvp4JXcmil/boB+9/egM6u00/jxSKp3CK5NWNTQ0JDsER1APd6Ee7sIrHkrq45XcmSh+bYN+9PajM6i302gnVImLVxJpOBxOdgiOoB7uQj3chVc8lNTHK7kzUfzaBv3o7UdnUG+n0U6oEhevPOvMK891Ug93oR7uwiseSurjldyZKH5tg3709qMzqLfTaCdUiUtLS0uyQ3CEqqqqZIfgCOrhLtTDXXjFQ0l9vJI7E8WvbdCP3n50BvV2Gu2EuhCx+L2IbBeR9+xlvxCRLfY09317KI6e2E3CrF+/nuLi4rZXr169eOCBB9rK77nnHkSEiooKwDtTaquHu1APd+EVD2Xv0dzpPJdeein9+/enqKiobdns2bPZf//923LwG2+80VZ2ySWXMGLECI488kiWLVuWhIh7Hj8ee/zoDOrtNNoJTQIiMk1EPhKROhH5TkQeE5Hoc92jgPHAQGPM8SKSDtwHTLCnua/cy/2OFpGvE6i/N7vpdg499FDKysooKyvj/fffJycnh8mTJwOwadMm3nzzTQYNGtRWPxgMJitUR1EPd6Ee7sIrHkp8NHf2PNOmTaOkpGS35TNmzGjLw2eccQYA8+fPB+Cjjz7izTff5Oqrr/bF42r8eOzxozOot+Pb7ZatKnERkauBa4GLgVJgf+B/gDdF5CRjTBMwGCg3xrQ+mGc/IAv4uCdj/aKilivvXtqTu+yQVbPG7vK+tLSUAw88kMGDBwNWYrzrrrs455xz2urU1dWRk5PTo3F2B+rhLtTDXXjFQ2kfzZ09S2uuPeWUUygvL+/UOh999BHjxo0DoH///hQUFLB69WqOP/747grTFfjx2ONHZ1Bvp9EzoT2IiPQCbgGuNMaUGGOajTHlwIVYyfMiEbkMeAo40b586P+zd+5xVlb1/n9/hrkxwwAKh+Qikic1AhHLUtMMxcvRg1l2g8obVufU7xzKkyjeSrCLpnlJzU5pSV5QK295QQWk0ykVsPBkmUmKgCI4yAxzn9kz398fzzPjdpg9s/fMM7PXfp71fr32i9lrPZfve4a1v/tZz3rWWga8GB6iRtKqcMjRNZK2S6qV9H+SpofnKJN0laRN4RCkH0saLqkSeBSYEB63XtKE3uJNmfu9uXfddRfz5s0D4MEHH2TixIkcdNBB79imqqoqH6FFjvdwC+/hFnHx8OyOz53uccMNNzBjxgzmz5/Pzp07AfjABz7AAw88QCqV4pVXXuHZZ59l8+bNeY508EniZ08SncF7R42/Ezq0fJigV/be9EIzq5f0KHCcmc2T1A580cyOBJA0BXgFGG1mKUknAEcB+wO1wHuBzqeGrwD2BWYCbcCdwDfN7AJJJwK3m9mkbILtaKxhw/UXkDIxTMa+R5zE8ENOZnx5Bw0p0W4wssTY3lLEmNIOioBtLUXsVd5BfSpIwiOKjTeai3hXWQcdwI7WIsaVdbCrTQwTVBYbW5uLGF/eQaoDatqKGFvWQU2bKCuC4cPerm/tCNYqqq+vp7Kykvr6eh544AEuu+wyNm3axJIlS3jooYeorq7GzKitre38/SKJ0tJSiouLaWxspKqqisbGRtrb2xk9ejQ1NTWUlZVRVFREU1MTVVVVNDQ0YGaMHDmS2tpaysvLgSCGUaNGsWvXLiRRWVlJXV0dw4cPp6Ojg5aWlq5jDhs2jIqKCurq6qioqCCVStHa2tpVX1xcTHl5eZdTa2srbW1tXfUlJSWUlpbS0NBAKpWivLycVCrVVV+ITk1NTYwdO5aGhgZGjBhBc3NzQTq99dZb7L333rv9nQrNaevWrey11169/t8rBKfXX3+dPfbYI+v2lIuTJ+/43DnA3FmXEmNKjZ2tYvgwo3wYXfXN7dDULvYoNXa0iqpio7q6uqtd1NXV0dHRQXV1NVVVVcybN4+vfOUr7LHHHpx//vksWLCAH//4xxx//PG8+OKLHHzwwUyaNInDDz+cxsZG6uvrgfzlzsH+/DIzioqKYuXU19+p83tVnJzynWdc+D4Q9fe2vlBS17PKB5K+AFxlZnv1UHc58AEzO07SmfScSEvCRHoM8GPgdGCNmXWE2wmoB2aY2T/CssOBO83s3ZJmkUMiXfbQCrv6BbdulqcPx33ggQe48cYbefzxx7uGAXUOF9iyZQsTJkxgzZo1FBcXM3bs2HyFHBnV1dXewyG8h1sMpsfKlSufnT179iGDcnBPn/jcOfSk59qNGzcyZ84cnn/++d22S6/r3gY//OEPc/PNN/O+971vSGLOF3H5DM2FJDqD986VvnKnvxM6tFQDYyUVm1n3lV/Hh/V9YmarJN0A3AhMlnQfcC5BT3EF8GzaxAgC+rVo2b7/VMXaOYf2Z9chYdmyZV1DcQ888EC2b9/eVTdlyhTWrVvH2LFjY7O4cFzWp/IebuE9PAWAz50OsXXrVsaPHw/Afffd1zVzbufdn8rKSp544gmKi4tjfwEKyfzsSaIzeO+oKeyuusLjKaAFODW9MHzm5ESCyRaywsx+aGYfAKYRDC1aSJCIm4BpZjY6fI0ysxGdu+USrMvD0BobG3niiSc49dRT+9w2Lus6eQ+38B5uERcPT4/43Jkn5s2bx+GHH86LL77IpEmTuOWWWzjvvPM48MADmTFjBk8++STXXHMNABs2bOD9738/U6dO5YorruiaLTfuJPGzJ4nO4L2jxt8JHULMrFbSYuB6Sbt45wx/W4CsPrElfZCgA+GPQAPQDLSbWYeknwLXSPoPM9suaSIw3cweA7YBYySNMrPaLM7TD8uhoaKigh07Ms+2nz6bX1lZ2RBENPh4D7fwHm4RFw/P7vjcmT+WLVu2W9nZZ5/d47b77bcfL774Yo91cSaJnz1JdAbvHTX+TugQY2bfBy4ErgJ2Ac8Am4HZZtaS5WFGAj8FdgKvAjvC4wGcD2wAng6T9QrggPDcfwOWAS9Lqulrhr+4JNKionj8N/cebuE93CIuHp6e8bnTfZLaBpPonURn8N5R4++E5gEzuwW4pZf6W4Fb095vJHg+pfP9SmBGhn2bCRL1hRnq52cbZ1wWmW5qaqKysjLfYQwY7+EW3sMt4uLhyYzPnW6T1DaYRO8kOoP3jppkXtJ7smLYsH7NyeAccVnXyXu4hfdwi7h4eAqfuOTOXElqG0yidxKdwXtHjb8I9WQkLpMrNDQ05DuESPAebuE93CIuHp7CJy65M1eS2gaT6J1EZ/DeUeMvQj2xJy5r4XoPt/AebhEXD4+nUElqG0yidxKdwXtHjb8I9WQkLkOKRo4cme8QIsF7uIX3cIu4eHgKn7jkzlxJahtMoncSncF7R42/CPVkJC5Dimpr+5xRvyDwHm7hPdwiLh6ewicuuTNXktoGk+idRGfw3lHjL0I9GYnLVNTl5eX5DiESvIdbeA+3iIuHp/CJS+7MlaS2wSR6J9EZvHfUJPOT0uPxeDwej8fj8Xg8ecFfhHoyEpe1zpqbm/MdQiR4D7fwHm4RFw9P4ROX3JkrSW2DSfROojN476jxF6GejMRlcoVRo0blO4RI8B5u4T3cIi4ensInLrkzV5LaBpPonURn8N5R4y9CPRkZqskV2tvbOfjgg5kzZw4QTAV90UUXsf/++zN16lR++MMfDuj4u3btiiLMvOM93MJ7uEVcPDyFT74nJrruuuuYPn0606ZN49prrwVg/fr1HHbYYcycOZNDDjmENWvWRH7epLbBJHon0Rm8d9QU5EWopI2SmiTVS9om6eeSRgxxDH8Jz18vqV1Sc9r7C4cylm5xvUeSSdrtyk0BXw9jb5C0RdLdkqblI9ZOrrvuOqZOndr1/tZbb2Xz5s387W9/44UXXmDu3LkDOr6kgYboBN7DLbyHW8TFYzDxubPXuAoud/bE888/z09/+lPWrFnDc889x0MPPcRLL73Eeeedx7e+9S3Wr1/PkiVLOO+88yI/d1LbYBK9k+gM3jtqigflqEPDyWa2QtJE4DHgYmDRUJ3czLqSj6TVwO1mdvNQnb8XzgDeAuZJ+oaZtaXV3QgcB3wJ+APB3/+TwL8Cf+l+oH/saOJrV64alCDXLjwGgC1btvDwww9z0UUXcfXVVwNw0003ceedd3bNMDhu3LgBnauysnJgwTqC93AL7+EWcfEYAnzu7JmCyJ29sXbhMbzwwgscdthhVFRUAPDRj36U++67D0lddzNqa2uZMGFC5OdPahtMoncSncF7R01B3glNx8xeAx4Fpks6S9ILkuokvSzp3zq3C8vnpL0vllQt6f3h+8Mk/UFSjaTnJM3qb0ySysPjTE0rGy+pUdIYSceGPdLflLRD0iuS5nbb/2pJm8Pe6h9J6nN+ZAVdFacBFwAiSJCdde8F/g34rJmtNrNWM2s0s9vM7Ps9Ha9Y1t9fQdZ8/etf5/vf//47prT/xz/+wd13380hhxzCiSeeyEsvvTSgc9TV1Q00TCfwHm7hPdwiLh5Dhc+d7zhvweXOTEyfPp3/+Z//YceOHTQ2NvLII4+wefNmrr32WhYuXMjee+/Nueeey/e+973Iz53UNphE7yQ6g/eOmkK+EwqApL2Bk4B7ge3AHOBl4CjgUUlrzeyPwDJgHvBQuOsJQLWZ/THsEX6YIAktB2YDv5b0XjN7M9eYzKxZ0j3AF4CLwuLPAY+Z2Y7wtvYkoAqYABwBPCRpnZltAK4CJgIzgHbgrvA4l/Rx6lnAu4C7gYOB04H7w7pjgY3h7yIrWupr2XD9BaRMDJOx7xEnMfyQkxlf3kFDSrQbjCwxtrcUMaa0gyJgW0sRe5V3UJ8Kbt2PKDbeaC7iXWUddAA7WosYV9ZBQ0MDDz/8MCNHjuSggw7i4Ycfpq2tjZaWFpqbmykqKmLVqlXce++9nHXWWdx///0UFxdTXl5OfX09lZWVtLa20tbWxujRo6mpqaGkpITS0lIaGhoYMWIEzc3NpFIpSkpKqK6uprS0lOLiYhobG6mqqqKxsZH29vau/cvKyigqKqKpqYmqqioaGhowM0aOHEltbW3XOknNzc2MGjWKXbt2IYnKykrq6uoYPnw4HR0dtLS0dB1z2LBhVFRUUFdXR0VFBalUitbW1q76XJyKioqoqakhlUp11ReiUyqVoqmpabe/U6E5dcbd2/+9QnBqbm6mqakp6/bkqlNzczM1NTX9+ozoyymO+Nz5DmZRILlzV5sYJqgsNrY2FzG+vINUB9S0FVFdXc0+++zDggULOProoxk1ahQHHHAA7e3tXH/99SxevJi5c+dy9913c/rpp7Ny5cpI23pzczMtLS1O5Jmh/PwqLi5mx44dsXLq6+/U+b0qTk75zjMufB+I+ntbX8gsfz12/UXSRmAskAJqCZLgN8ysqdt29wNPmtl1kt4D/Al4l5k1SroDeNHMlkg6H5huZqel7fsYcKeZLc0intV0G1Ik6QjgdmBfMzNJ64ElZnavpGOBR4DRZtYYbn8vsBa4AmgEDjCzV8O6jwA/M7P9+ojjVmCEmX0q3GcFMCFM3t8CjjOzI/vy6eSOh1bYtS8Mzs3ytQuP4YILLuC2226juLiY5uZmdu3axamnnsq6detYvnw5U6ZMwcwYPXo0tbW1/T5XXV0dVVVVEUafH7yHW3gPtxhMj5UrVz47e/bsQwbl4EOIz50Z47iVAsmdvdH5mEs6F154IZMmTeKCCy6gpqYGSZhZ15fXKInLZ0muJNE7ic7gvXOlr9xZyHdCP25mK9ILJJ0IfAvYn2CocQXwZwAz2yDpBeBkSb8BPkbQ4wmwD/BpSSenHa4EeLK/wZnZ7yWlgCMl7QQmEyT8TnZ0JtGQVwl6dvcCyoDn9PaDwH0+ESypkuAZldPDov8FthL0YN8A7ADG5+Lw7j0rWLvww7nskhPf+973uoYErV69mquuuorbb7+dRYsWsWrVKubPn89vf/tb9t9//wGdp6WlJRYfGt7DLbyHW8TFYwjwuTONQsydfbF9+3bGjRvHpk2buPfee3nqqae4/vrr+e1vf8usWbNYtWoV++3X63V5v0hqG0yidxKdwXtHTSFfhL4DSWXArwkSyQNm1hb25qYnoc5hRUXAX8PhOwCbgdvM7EsRh/ULgmFFNcA9ZtaSVjdG0vC0HujJwDpgG9BK0Ju7LYdzfRIYAfxE0k1h2WiC38cNwErgOkkHm9mfsjlgvtY6W7RoEZ///Oe55pprGDFiBDffPLA5K0aPHh1RZPnFe7iF93CLuHgMNT53xid3dvLJT36SHTt2UFJSwo033sgee+zBT3/6U772ta+RSqUoLy/nJz/5SeTnTWobTKJ3Ep3Be0dNbC5CgVKCXtA3gVTYs3s88HzaNncB3wH2BO5MK78dWCvpBIJhOCXAYcAGM9sygJhuA54lGCL02W51RcClki4BPgycCCwys3ZJNwPXSloAVBM84/I+M3u8l3OdAfwU+GZa2WTgaUlTzewFST8B7pb0JeApYBjwCWCimV3Z/YBD+SzUrFmzmDVrFhD8Z3/44Yd73yEHampqGDt2bGTHyxfewy28h1vExSMP+NxZwLmzJ373u9/tVnbkkUfy7LPPDup5k9oGk+idRGfw3lFT8LPjdmJmdcAC4B5gJ8FkBg9222YrQQL5MMEEBJ3lm4FTgAsJEvFmYCED/P2Y2UaCIU2tZvaHbtVbgAaCYT9LgS+aWec0sN8gGGK0huC5nceBjGNnJE0mmFjhWjN7I+21huCLwRnhpv8PuCl87QReIhha1eMVn2KyHlK+e6Wjwnu4hfdwi7h4DDU+d/rcGRVJbYNJ9E6iM3jvqCnIiYkKCUm/AF42s0vTyo4FbjazKfmKKxueeeYZO/TQQ/MdxoBpaWmhrKws32EMGO/hFt7DLQbTIy4TExUSPncWHnH5LMmVJHon0Rm8d670lTtjcyfURSTtS9BL/LN8x9If8j2kKCrisq6T93AL7+EWcfHw+NxZqCS1DSbRO4nO4L2jxl+E9oGk+gyvj/Sx3/eA54DvmtmmiGK5JEMsv4ni+N0pKorHf4+Kiop8hxAJ3sMtvIdbxMUjLvjcmTyS2gaT6J1EZ/DeUROniYkGBTMb0c/9LgAuyFC3ApjSj2NeBlzWn3j6Q1yGaqdSqXyHEAnewy28h1vExSMu+NyZPJLaBpPonURn8N5Rk8zuOk9WxCWRtra25juESPAebuE93CIuHp7CJy65M1eS2gaT6J1EZ/DeUeMvQj0ZicssYHFZ18l7uIX3cIu4eHgKn7jkzlxJahtMoncSncF7R42/CPVkJC6TK9TU1OQ7hEjwHm7hPdwiLh6ewicuuTNXktoGk+idRGfw3lHjL0I9GYnLWmfFxfF49Nl7uIX3cIu4eHgKn7jkzlxJahtMoncSncF7R42/CPVkJC6JtLy8PN8hRIL3cAvv4RZx8fAUPnHJnbmS1DaYRO8kOoP3jhp/EerJSEdHR75DiIT6+vp8hxAJ3sMtvIdbxMXDU/jEJXfmSlLbYBK9k+gM3jtq/EWoJyNRrnXW3NzMhz70IQ466CCmTZvGt771LQDOPPNM3v3udzNz5kxmzpzJ+vXrIztnJ5WVlZEfMx94D7fwHm4RFw9P4ZNt7syUF1955RUOPfRQ9ttvPz772c8WzIycSW2DSfROojN476gpuItQSRslNYULTW+T9HNJ/VqPbAAx/CVtset2Sc1p7y8cyljSYhot6TpJm8I4Nki6WtKYtG3OlvS8pEZJWyXdKGlUpmNGOc18WVkZq1at4rnnnmP9+vUsX76cp59+GoArr7yS9evXs379embOnBnZOTsplATeF97DLbyHW8TFY7DwuTNjTHnLnZny4vnnn88555zDSy+9xB577MEtt9wSgengk9Q2mETvJDqD946afj9pKmk40G5m+fiLnGxmKyRNBB4DLgYWDdXJzWxa58+SVgO3m9nNQ3X+7kgqB1YBbwLHA38HxgJfBQ4BHpN0PnAOcDrwJLA3cFNY9xEza+t+3I07Glhw5aoBxbZ24TGdMTJiRPB9p62tjba2tiF7bqatbTe1gsR7uIX3cItC8Mhz3gSfO99BPnPn2oXHZMyLq1at4s477wTgjDPO4NJLL+UrX/lKJM6DSSG0wcEgid5JdAbvHTVZ3wmVdJWkD4U//yvwFlAj6eRBiSwLzOw14FFguqSzJL0gqU7Sy5L+rXO7sHxO2vtiSdWS3h++P0zSHyTVSHpO0qz+xiSpPDzO1LSy8WEP6hhJx4Y90t+UtEPSK5Lmdtv/akmbw97qH4WJsjfOBPYCPmFmfzOzDjPbbmaXmtljkvYAvgV81cweN7M2M3sZ+BSwHzCvp4O2WbQXie3t7cycOZNx48Zx3HHHceihhwJw0UUXMWPGDM455xxaWloiPSfEZ10n7+EW3sMtXPRwMW+Cz51pnEmec2f3vPjP//zPjB49ums2ykmTJvHaa69lfbx84mIbHAqS6J1EZ/DeUZPLndDPA98Mf/4m8AWgFrgG+E3EcWWFpL2Bk4B7ge3AHOBl4CjgUUlrzeyPwDKCZPFQuOsJQLWZ/THsEX4YOA1YDswGfi3pvWb2Zq4xmVmzpHsIfj8XhcWfAx4zsx3h3b9JQBUwATgCeEjSOjPbAFwFTARmAO3AXeFxLunltMcCj5pZY4b6Iwj+1g90i7VO0nLgOOAXu+3VWMOG6y8gZWKYjH2POInhh5zM+PIOGlKi3WBkibG9pYgxpR0UAdtaitirvIP6VJCEq6urGTVqFLt27UISzzzzDFu2bOGss87imWee4dxzz+WAAw5g+/btnHvuuXznO99hwYIFVFRUkEqlaG1tZfTo0dTU1FBcXEx5eTn19fVUVlbS2tpKW1tbV31JSQmlpaU0NDQwYsQImpubSaVStLe3M2zYMEpLSykuLqaxsZGqqioaGxtpb2/v2r+srIyioiKampqoqqqioaEBM2PkyJHU1tZ2zQ7W3Nz8DqfKykrq6uoYPnw4HR0dtLS0dB1z2LBhVFRUUFdXN2Cn1tbWrmN01heiU0NDA+PGjdvt71RoTtXV1eyzzz69/t8rBKfXXnuNCRMmZN2eXHV69dVXGTNmTL8+I/pyGgDO5U3wuTONvOXO6urqrnaxcuVKOjo6+MQnPsExxxyDmXXlzp07d3a1g6HIMwNp62+88QaTJk1yIs8M5edXKpWipKQkVk59/Z06v1fFySnfecaF7wNRf2/rC2X77IKkWjMbpeA5ib+Z2T+F5bvMbGRWB4kASRsJhsukCJL5w8A3zKyp23b3A0+a2XWS3gP8CXiXmTVKugN40cyWhENtppvZaWn7PgbcaWZLs4hnNd2GFEk6Argd2NfMTNJ6YImZ3SvpWOARYHRn4pN0L7AWuAJoBA4ws1fDuo8APzOz/XqJ4Ung92Z2cYb6M4Fvm9mkHuquAqaZ2Ynd6+58aIVd88LAHhvuHI7bncWLF1NZWcm5557bVbZ69WquuuoqHnrooR736S+1tbWMGpXx8Z2CwXu4hfdwi8H0WLly5bOzZ88+JNf9XMmb4Tk34nNn9xjyljt7yo2LFy+moqKCK664gjfeeIPi4mKeeuopLr30Uh577LFej+cCcfksyZUkeifRGbx3rvSVO3O5E/p3SZ8H3gM8ASBpLNDU616Dw8fNbEV6gaQTCYbN7E8wzLgC+DOAmW2Q9AJwsqTfAB8DDg533Qf4dLfhUSUEz370CzP7vaQUcKSkncBkgoTfyY5uPa+vEvTs7gWUAc/p7eclsxnXswMY30t9NTBOUpGZdZ87fnxYvxv7jh3B2oWHZXH6vnnzzTcpKSlh9OjRNDU1sWLFCs4//3y2bt3K+PHjMTPuv/9+pk+fHsn50iktLY38mPnAe7iF93ALRz1cypvgc2d38po7M+XFo48+ml/96lfMnTuXpUuXcsopp2Shkn8cbYODThK9k+gM3jtqcrkI/SpwHdAGzA/LTgAejzqoXJFUBvyaYOKAB8ysLezNTU9CncOKioC/hsN3ADYDt5nZlyIO6xcEw4pqgHvMLP1hxzGShqfFKef7AAAgAElEQVT1QE8G1gHbgFaC3txtOZxrBfBNSRUZhhX9nmB40scJhl8BIKmK4G94bg/7RLrW2datWznjjDNob2+no6ODz3zmM8yZM4djjjmGN998EzNj5syZ/PjHP47snJ00NDQwfPjwyI871HgPt/AebuGoh7N5E3zuJM+5M1NefN/73sfcuXO5+OKLOfjggzn77LNzUMofjrbBQSeJ3kl0Bu8dNVlfhJrZWuDD3cruAO6IOqh+UErQC/omkAp7do8Hnk/b5i7gO8CewJ1p5bcDayWdQJCQSoDDgA1mtmUAMd0GPEswROiz3eqKgEslXULwOz0RWGRm7ZJuBq6VtICgl3Ui8D4z6+1Ly63Al4FfSTqHYIa/McBXgDXhBAuXATdKqifoqZ5EMMPfK7zz9/F2kBGuEzpjxgz+9Kc/7Va+atXAZt/Nhs7ZBwsd7+EW3sMtXPRwPG+Cz523ksfcmSkv7rvvvqxZsyarY7iEi21wKEiidxKdwXtHTU5XGZKOk3RLOCwHSYdI6vmhvyHEzOqABcA9wE6CyQwe7LbNVuApgsR1d1r5ZuAU4EKCRLwZWMgA11A1s40EQ5pazewP3aq3AA3AVmAp8EUzeyms+wbBEKM1BM/tPE4wC19v52oGjgE2EHwZqAOeBkYRPC+DmX2XYMjVNcAugt/FK8BxlmG5gGyfF3ad5ubmfIcQCd7DLbyHW7jq4WreBJ87fe6MFlfb4GCTRO8kOoP3jpqs74RK+k/ga8DNBNOTQ/Bcyw/p1tM7mJjZlAzlNwI39rHv7AzlzwAf7Wc8s3qp3gSszrDfEmBJD+XNBOu25bR2m5nVEHyZWNDLNj8BfpLDMXMJwVlSqVS+Q4gE7+EW3sMtXPRwJW+Cz529xOFzZ0S42AaHgiR6J9EZvHfU5PJM6NeB2Wa2MZwVD+BvwAHRh1X4SNqXoJf4wHzH0l+GDRuW7xAiIS7rOnkPt/AebuGoh8+bOeJzZ+HiaBscdJLonURn8N5Rk8uwmSqC4TYAnd18JQSTAcQSSfUZXh/pY7/vAc8B3zWzTRHFckmGWAZtrbkBro/nDDU1NfkOIRK8h1t4D7dw1CNxeRN87oxL7swVR9vgoJNE7yQ6g/eOmlzuhP4PwTCX76SVLWAA07G7jpn160lcM7sAuCBD3QpgSj+OeRlwWX/i6S9pU90XNHGZUtt7uIX3cAtHPRKXN8HnzrjkzlxxtA0OOkn0TqIzeO+oyeUi9D+B30j6ElAl6UWCh/RP7n03T6ESl0RaXJzLf3N38R5u4T3cwlEPnzcTSFxyZ6442gYHnSR6J9EZvHfkx81h223AB8PXPgRDjNb0sICzJyZEuU5oPmlsbKSioiLfYQwY7+EW3sMtHPXweTOBxCV35oqjbXDQSaJ3Ep3Be0dNVhehkoYB9cBoM1tDMAW6J+bEZXKFqqqqfIcQCd7DLbyHW7jm4fNmcolL7swV19rgUJFE7yQ6g/eOmqwmJjKzdt5exNmTEOLSm9vY2JjvECLBe7iF93AL1zx83kwuccmdueJaGxwqkuidRGfw3lGTy3DcO4CHJF1HsGB010JYZrYq6sA8+Scua53FZaZC7+EW3sMtHPXweTOBxCV35oqjbXDQSaJ3Ep3Be0dNLhehXwn/vbRbuQH7RhKNxyniMqQoLus6eQ+38B5u4aiHz5sJJC65M1ccbYODThK9k+gM3jtqsl4n1MzeneHlE2lMyabnY/78+YwbN47p06fvVnfVVVchierq6sEIL2visq6T93AL7+EWLnr4vJlMesudmzdv5uijj2bq1KlMmzaN6667DoC33nqL4447jv3224/jjjuOnTt3DlW4keFiGxwKkuidRGfw3lGT9UWoi0jaKKkpXHh6m6SfS+rX+mQDiOEvaYtft0tqTnt/4VDGEsYjSReHv5t6SVsk3RHWvdhLrOf1cKw+z3fmmWeyfPny3co3b97ME088weTJk6PQGhBlZWX5DiESvIdbeA+3iIvHUOBzZ4/xDEnuLC4u5gc/+AEvvPACTz/9NDfeeCN//etfufzyy5k9ezYvvfQSs2fP5vLLLx9E28EhqW0wid5JdAbvHTVZD8eVtJm051nSMbN8XmmcbGYrJE0EHgMuJlgcfEgws2mdP0taDdxuZjcP1fl7YD4wFzjGzF6WNB6YA2BmB3RuJOl/gZvN7NZMB/pHdQMLruz5saW1C48B4KijjmLjxo271Z9zzjl8//vf55RTTum3SFQUFRV0X0sX3sMtvIdbuOjhcN4Enzu7M+i5c+3CYxg/fjzjx48Hghknp06dymuvvcYDDzzA6tWrATjjjDOYNWsWV1xxRWRyQ4GLbXAoSKJ3Ep3Be0d+3By2/QJwWtrrPOA14AeDEFfOmNlrwKPAdElnSXpBUp2klyX9W+d2YfmctPfFkqolvT98f5ikP0iqkfScpFn9jUlSeXicqWll4yU1Shoj6diw1/WbknZIekXS3G77Xy1pc9hb/SNJ5X2c9oPAcjN7Ofy9bDWzn/Yn/mH9XG/7wQcfZOLEiRx00EH9O0DENDU15TuESPAebuE93MJRD6fzJvjcmcaQ586NGzfypz/9iUMPPZRt27Z1XZyOHz+e7du39+fUecXRNjjoJNE7ic7gvaMm6zuhZvbb7mVh7+Vy4LoIY+oXkvYGTgLuBbYT9GC+DBwFPCpprZn9EVgGzAMeCnc9Aag2sz+GPcIPE3xZWA7MBn4t6b1m9mauMZlZs6R7CL6IXBQWfw54zMx2hEN2JgFVwATgCIKZFNeZ2QbgKmAiMANoB+4Kj3NJL6d9GviBpDeAJ4H14VIBOdNSX8uG6y8gZWKYjH2POInhh5zM+PIO6urqKCoqoqmpiZaWFtrb29mxYwfFxcUsXryYBx98kPr6ejo6Omhra2PHjh1IorKykrq6OoYPH05HRwctLS2MHj2ampoahg0bRkVFBXV1dVRUVJBKpWhtbe2qLy4upry8nPr6eiorK2ltbaWtra2rvqSkhNLSUhoaGhgxYgTNzc2kUimGDx9OdXU1paWlFBcX09jYSFVVFY2NjbS3t3ftX1ZW1uVUVVVFQ0MDZsbIkSOpra2lvDz4DtPc3MyoUaPYtWvXkDqVlJRQU1NDKpXqqi9EJzOjqalpt79TITqlUqle/+8VglNbWxtNTU1ZtydXndra2qipqenXZ0RfTv3F9bwJPnemMei5s7q6uqtdvPnmm5x66qlcdtlltLW1AVBdXd3VLsyMtra2Ic8zA2nrbW1ttLS0OJFnhvLzq7y8PK/fcfLxmdz5vSpOTvnOMy58H4j6e1tfaCBTiUvaA9hoZqP6fZABIGkjMBZIAbUESfAbZtbUbbv7gSfN7DpJ7wH+BLzLzBrDZz5eNLMlks4HppvZaWn7PgbcaWZLs4hnNd2GFEk6Argd2NfMTNJ6YImZ3SvpWOARgsXMG8Pt7wXWAlcAjcABZvZqWPcR4Gdmtl8fcZwGnAkcBjQD3zOzq7pt0+eQomUPrbCrX+j5ZnnncFwIenPnzJnD888/z5///Gdmz55NRUUFAFu2bGHChAmsWbOGvfbaq7ewB4233nqLPffcMy/njhLv4Rbewy0G02PlypXPzp49+5AojpXvvBnGsBGfO3uKY1BzZ2febGtrY86cOZxwwgn813/9FwAHHHAAq1evZvz48WzdupVZs2bx4osv9hauc8TlsyRXkuidRGfw3rnSV+7M5ZnQJd2KKgh6Tx/NOapo+biZrUgvkHQi8C1gf4IhxxXAnwHMbIOkF4CTJf0G+BhwcLjrPsCnJZ2cdrgSgl7RfmFmv5eUAo6UtBOYTJDwO9nRmURDXiXo2d0LKAOe09uTHGQ1yMfMbgNuk1QCfDL8+U9mtjKX2PfZs4K1Cz+cyy4ceOCB7xhGNGXKFNatW8fYsWNzOk6UxGXNNu/hFt7DLVz0cDhvgs+dPZ1z0HOnmXH22WczderUrgtQgI997GMsXbqURYsWsXTpUifmU8gVF9vgUJBE7yQ6g/eOmlzWCd272/sG4GrgtujCGTiSyoBfA6cDD5hZW9ibm56EOocVFQF/DYfvAGwGbjOzL0Uc1i8IhhXVAPeYWUta3RhJw9N6oCcD64BtQCtBb+62/pzUzNqAuyQtAqYDOSXSbNY6mzdvHqtXr6a6uppJkyaxePFizj777P6EO2iMHDky3yFEgvdwC+/hFo56FETeBJ870xnM3Pn73/+e2267jQMPPJCZM2cC8N3vfpdFixbxmc98hltuuYXJkyfzy1/+sj+h5xVH2+Cgk0TvJDqD946aXC5CLzCzN7oXStoL2K08j5QS9IK+CaTCnt3jgefTtrkL+A6wJ3BnWvntwFpJJwArCHpyDwM2mNmWAcR0G/AswRChz3arKwIulXQJ8GHgRGCRmbVLuhm4VtICoJrgGZf3mdnjmU4kaT7B3+N3BF94TgQOANbkGnQ247mXLVvWa31PM+cONbW1tXm9ExsV3sMtvIdbOOpRKHkTfO4cktx55JFHZryrsHJlTte6zuFoGxx0kuidRGfw3lGTy+y4f89Q/tcoAokKM6sDFgD3ADsJJjN4sNs2W4GnCBLX3Wnlm4FTgAsJEvFmYCEDXE/VzDYSDGlqNbM/dKveQpDwtgJLgS+a2Uth3TcIhhitIXhu53Gg12dagF0EU+1vJvD/LvBlM3sq17jjMhV150PchY73cAvv4RaOehRE3gSfO/G5c8A42gYHnSR6J9EZvHfU5HIndLdnKiSNBDqiCyc3zGxKhvIbgRv72Hd2hvJngI/2M55ZvVRvAlZn2G8J0P3ZIcysmWDdtqzXbjOzXwG/ymK7I7M9psfj8Xj6hXN5E3zuzHAsnzs9Ho9nCOmzuy5cZ2sTMFzSpvQXQQ/k/YMeZYEjaV+CXuKf5TuWXOjoyOv3pMhobm7OdwiR4D3cwnu4hUsePm9Gg8+dhYVLbXAoSaJ3Ep3Be0dNNndCv0DQm/sIwRpgnRiwzcwKaw7xfiKpPkPViWb2u172+x7wH8C3zWxTRLFcApzfQ9WTZnZyD+X9IpuJiQqBUaPythJCpHgPt/AebuGYh8+bIT53JgfH2uCQkUTvJDqD946aPi9CLVxsW9LYbtOhJwozG9HP/S4ALshQtwKY0o9jXgZc1p94cmEgi7S7xK5duxgzZky+wxgw3sMtvIdbuOTh8+bb+NyZHFxqg0NJEr2T6AzeO2qyfiY0XJx6JvARgkWulVb3zcgj83giIm2tuILGe7iF93ALFz183vQkCRfb4FCQRO8kOoP3jpqsp3CT9GXg98AxBMNZDiSYge49gxKZJ+/EZUhRZWVlvkOIBO/hFt7DLVz08HkzmcQld+aKi21wKEiidxKdwXtHTS7ziJ8H/IuZfQJoCv/9FNA2KJF58k5chhTV1dXlO4RI8B5u4T3cwlEPnzcTSFxyZ6442gYHnSR6J9EZvHfU5HIROi5tEoEOSUVm9igQ2cP8HreIy1pnw4cPz3cIkeA93MJ7uIWjHj5vJpC45M5ccbQNDjpJ9E6iM3jvqMllndAtkqaEi0f/HThFUjXQOiiRefKOmeU7hEiIy3T53sMtvIdbOOrh82YCiUvuzBVH2+Cgk0TvJDqD946aXLrrvg9MDX9eAtwOrAIWRx2Uxw3ikkhbWlryHUIkeA+38B5u4aiHz5sJJC65M1ccbYODThK9k+gM3jtqsr4INbNbw2FEhP/uAexhZjcNSmSevNPb5Arz589n3LhxTJ8+vavsrbfe4rjjjmO//fbjuOOOY+fOnUMRZp+MHj063yFEgvdwC+/hFi56+LyZTDLlzp7y5qWXXsrEiROZOXMmM2fO5JFHHhmqMCPHxTY4FCTRO4nO4L2jJqcHFySNkXSapPPMrBUYKWnSoETWdywbJTVJqpe0TdLPJfVrPbIBxPCX8Pz1ktolNae9v3AoYwnjOVZSR3j+Okl/k3RGWn2ZpCWSNkhqCH+HN0ua3NPxeptc4cwzz2T58uXvKLv88suZPXs2L730ErNnz+byyy+PSm1A1NTU5DuESPAebuE93MJVD5fypmdoyJQ7e8qbAOeccw7r169n/fr1nHTSSYMd3qDhahscbJLonURn8N5Rk8sSLR8FXgQ+D1wSFu8H5LNH9+RwIez3Ax8ELh7Kk5vZNDMbEcbwO+A/Ot+b2XeHMpY0NoXxjAQuAm6RdICCRX7uBU4EPguMAmYC/0ewfMBubHyrkQ9eueodr06OOuoo9txzz3ds/8ADD3DGGcE17xlnnMH9998fuVx/iMt0+d7DLbyHW7joMdC82a2ztfP1OUlbeth2taQvhj9fKskkfTqtvjgsm5JW9iFJj0iqkfSWpDWSzsoirpGSrpW0KYxpQ/h+bFrcx/ay/7vDDtMf9VB3iqT1knZJqpa0sjNmSaMl/UzSG2FH698lnZ9FvCapx2VxJE2SdIekHWHn7BpJc7ptI0kLJD0fbrNF0i8lHdjTMTPlzp7yZpxwsQ0OBUn0TqIzeO+oyeVO6LXAZ83sX4BUWPYM8KHIo8oRM3sNeBSYLuksSS+ECeplSf/WuV1YPiftfXGY5N4fvj9M0h/ChPycpFn9jUlSeXicqWll4yU1hj3jx4aJ+pth8ntF0txu+18taXN4p/dHkspz+J2Ymf0aqCN4JukE4Gjg42b2rJmlzKzGzH5oZrf2dIx2y21x2m3btjF+/HgAxo8fz/bt23Paf7CoqKjIdwiR4D3cwnu4haMeUeTNk9M6N0cAr2e531vAEkk9fnuQdDjB86m/JVi3dAzwFYKOyoxIKgVWAtOAfyHo8PwwsIPsvU4HdgJzJZWlHfs9wC8I1lIdBbwb+BHQOSvGNcAIgpw2CvgY8I8sz9mTy57A/xJMFDUNGBue405Jn0rb9Drga8ACYE9gf+B+4F97Om6uufOGG25gxowZzJ8/35nHWPqDo21w0EmidxKdwXtHTS6z404xs5Xhz51P3bfmeIxBQdLewEkEd/q2A3OAl4GjgEclrTWzPwLLgHnAQ+GuJwDVZvZHSROBh4HTgOXAbODXkt5rZm/mGpOZNUu6B/gCwR1JgM8Bj5nZjuDGJJOAKmACcATwkKR1ZrYBuAqYCMwA2oG7wuNcQhZIKgJOJUjYfyb4cvFUeMGenUNjDRuuv4CUiWEy9j3iJFKpo6ipqaGsrIzGxkba29tpaWmhoaEBM6OtrY3a2lrKy4Pr5erqakaNGsWuXbuQRGVlJXV1dQwfPpyOjg5aWloYPXo0NTU1DBs2jIqKCurq6qioqCCVStHa2tpVX1xcTHl5OfX19VRWVtLa2kpbW1tXfUlJCaWlpTQ0NDBixAiam5tJpVK0t7czbNgwSktLKS4uprGxkaqqqq74O/cvKyujqKiIpqYmqqqqupxGjhz5Dqfm5ua8OLW2tnYdo7O+EJ0aGhoYN27cbn+nQnOqrq5mn3326fX/XiE4vfbaa0yYMCHr9uSq05YtWxgzZky/PiP6choA+cyby4HpBDloaQ/1VwJLzeyKtLJngc/0cdzTgcnA0WZWH5ZtBy7LIbbTCUYuXUqwXM2vwvKZwCtpv7M64Ndp+30QuNjMOq/U/ha++ss5QD1wtpl1XuguU/CIyg8k/ZrgAv3/AYeb2Zq0fe/IdNDecmd9fT0dHR1UV1dTVVXF3Llz+epXv8qoUaNYtGgRCxYs4KabbspbnhlIW3/jjTeYNGmSE3lmKD+/UqkUJSUlsXLq6+/U+b0qTk75zjMufB+I+ntbn5hZVi/g98AJ4c9vhf8eD6zO9hhRvoCNYfKoAV4l6C0d3sN29wNfC39+D0FSqwjf3wF8M/z5fOC2bvs+BpyRZTyrgS92KzsCeAVQ+H49cGr487EEX0Yq0ra/F7iA4A51M7BPWt1HgJf6iOFYgh7jGoJe8D8Bnwnrfg7cnsvv+LYHn7BDvr/yHa90XnnlFZs2bVrX+/33399ef/11MzN7/fXXbf/99zcXaGhoyHcIkeA93MJ7uMVgeqxYsWKd9S9PDShvhnnu2G5ls4AtPWzblYMILu5uJ7hT+DJQQnDha8AUoIKgc/PofjjdRXDxmlPcaXUfAVoIJmm6HngwrW7fMPddQzByZ0S3fW8G/gKcBeyXQ8wGvKeH8qeBxT2Uvzvc5wDg34FXc/kd9ZY7u+fNdHqrKwTi8lmSK0n0TqKzmffOlb5yZy69sd8guFP3MDBc0n8T9GCeksMxoubjZrYivUDSicC3CIbLFBEk2z8DmNkGSS8AJ0v6DUGCPjjcdR/g05LSFxEvAZ7sb3Bm9ntJKeBISTsJeo8fTttkh5k1pr1/leCu6F5AGfBceMcUINvxPZvMbEoP5Tt42zUr/nlsJWsXHp719h/72MdYunQpixYtYunSpZxySj7/a7xNKpXqe6MCwHu4hfdwC0c9osib94d5BIILzWuz3dHMHpR0EfBF4KdpVXsQ5MetOcTRyRiCO6b95QzgUTPbKelO4H8kjTOz7Wb2cvgYzH8B9wBVku4imG+hHvhPgruX/wH8RNKrwH9aOANxPxhLz7+DrWn1YzJsk5FccufWrVu7HmO577773jFzbqHhaBscdJLonURn8N5R0+czoZL2AjCzpwmGhv4F+BnBHb4PmdnaQYmsH4TPlvyaYCjru8xsNPAI77yA6xySewrwVwuGvgJsJrgTOjrtVWlmA53i9RcEw6FOA+4xs/TFdsZIGp72fjLB8z7bCO6SHpAWyygzGzWAOFYAh0uakO0OZpnXOps3bx6HH344L774IpMmTeKWW25h0aJFPPHEE+y333488cQTLFq0aADhRkdrazzWhfcebuE93MIlj4jz5sfT8sDHCZ4tLelhuxKgrYfyiwke5UifU2AnwaiZ8TnE0cmOfu5HmO8+TTiU1cyeAjYRPKpCWPa0mX3GzP6J4K7pUWH8mFmTmX3XzD5AcHF4D/DL8NnO/lCdwWV8Wn3OvplyZ09587zzzuPAAw9kxowZPPnkk1xzzTW5nMopXGqDQ0kSvZPoDN47arK5E/p3gokHMLPXJR1mZqcOSjQDp5TgDuKbQCq8K3o88HzaNncB3yGYXODOtPLbgbWSTiC4YCsBDgM2mNluMxHmwG0EvcaNBLPSplMEXCrpEoKJHU4EFplZu6SbgWslLSBIhBOB95nZ4/2M4zGCu7r3Sfp3gllxKwkukBvMbLdnhnqbDWvZsmU9lq9cubLH8nwSl3WdvIdbeA+3cMxjMPPmJmCspBHh3UEUDJnZh2A0zTswsyckbQC+mlbWKOkp4JPkPtpnBfBtSZVm1pDjvp8g+L38SNL1YdlogmdEd7vDa2ZrJd1L8Gxr97pdkr5L8AjLuwkeQcmVFcAnJS22t58JheC52M0Ef8cO4EZJh5jZumwOmil39pQ3zz777FxjdhbH2uCQkUTvJDqD946abGbH7T4MdNYgxBEJZlZHMHvdPQQ9vZ8DHuy2zVbgKYKLvrvTyjcT3B29kOAidjOwkBzXUu0hpo0Ew4FbzewP3aq3AA0EQ32WEjzP81JY9w2CLxRrgFrgcYKp/fsbhxFMVPQ4wSQQu8K4ZhLMkLgbA5yQwxnisq6T93AL7+EWjnkMWt40s00EM+xeIWlEOAJoIcEd0qcz7HYRcF63svOAMyUtlDQGQNJB4fDX3riNID/+WtJ7JRWFM75fKCl9kcuScJb3zlcxwVDcnwEHEuSemQRzJ8yUdKCkIyV9SdK4MJ73Ejw283T4/hJJH5RUGs4W/zWCORBe7CNmgNJu8QwjePZ0JMFSZnuF5fPC39fC8LGmlwjmnFgmaVbnuSXNldTjcJ+45M5ccawNDhlJ9E6iM3jvqMnmTmjmMZl5JMNzj5jZjcCNfew7O0P5M8BH+xnPrF6qNxE8y9PTfkuAJT2UNwOLwle2MawgmHQiU30Lwey62c6wm+2pnaa4OO8TOEeC93AL7+EWjnkMdt78LHA1sIEgjz8LnBTmjd2DCeYnWEPa8itm9gdJxwCLgYsltQMv0Xf+bFGwBuhi4AmC50u3AQ8QXBx38ki3XW8lmHX+YDN7I638DUnLCS5QbyW46Py2pEqCUUB3A9/vPD3BJHuTCS66/w/4V3t7lt7e+Eu3918ys5slHQlcAfyVYCTVX4HTzOyBtG0XhK8bCe667iRY2mW33A3xyZ254lgbHDKS6J1EZ/DekR83m20kHc3bPbvd32NmPd5J84CkfQnusPa4qLXLxCWRdk5nXeh4D7fwHm7hmEckebOXztbNBM9WZtrv0h7KTuqh7B0XptliZrXA18NXT/VTMux6Vobt02M7uadtwu2+DXw7uyjfsV/GZBbeWZ7Xx/5GsFboddmcLy65M1cca4NDRhK9k+gM3jtqsrkI3U4wfKaTHd3eG8G06rFFUqZe1hPN7He97Pc9gln8vh0muihiuYRgOZnuPGlmGZN3f+jo6Oh7owKgvr4+Fh8c3sMtvIdbOOaR+LyZZOKSO3PFsTY4ZCTRO4nO4L2jps+L0F56NBODmY3o534XEEya0FNdr0NneznmZeS2KHi/KSoa0OOwzlBZWZnvECLBe7iF93ALlzwKPW9KupBgfoTu/M7Mcr5zOthI+gjQ4zIt/c3fAyEuuTNXXGqDQ0kSvZPoDN47apL5SenJit6WaCkk4jKltvdwC+/hFnHxcIFwGZQRPbycuwAFMLPfZYh3yC9Aw3jycdq8k9Q2mETvJDqD944afxHqyUhcEmlbW09L5xUe3sMtvIdbxMXDU/jEJXfmSlLbYBK9k+gM3jtq/EWoJyO9rRNaSMRlXSfv4Rbewy3i4uEpfOKSO3MlqW0wid5JdAbvHTX+ItSTkbisdRaXdZ28h1t4D7eIi4en8IlL7syVpLbBJHon0Rm8d9T4i1BPRuIyzXxJSUm+Q4gE7+EW3sMt4uLhKXzikjtzJaltMIneSXQG7x01/iLUk5G4JNLS0tJ8hxAJ3sMtvIdbxMXDU/jEJXfmSlLbYBK9k+gM3uWMvXQAACAASURBVDtq/EWoJyOZ1jq75pprmDZtGtOnT2fevHk0NzcPcWS50dDQkO8QIsF7uIX3cIu4eHgKn0y5c/78+YwbN47p06d3lV166aVMnDiRmTNnMnPmTB555JGhCjNyktoGk+idRGfw3lHjL0I9GelprbPXXnuNH/7wh6xbt47nn3+e9vZ27rrrrjxElz0jRuRllv7I8R5u4T3cIi4ensIn0zqhZ555JsuXL9+t/JxzzmH9+vWsX7+ek046abDDGzSS2gaT6J1EZ/DeUeMvQh1A0kZJTZLqJNVI+oOkf5dUFNbfKunb4c9TJJmk+vC1UdKiHM9TL2mnpIcl7Z1p+5er6/nglau6Xp2kUimamppIpVI0NjYyYcKEAf4GBhfX79Rmi/dwC+/hFnHx8GRPoeTOzvx51FFHseeee0Zg7iZJbYNJ9E6iM3jvqPEXoe5wsplVAfsAlwPnA7f0sv3ocCHuTwGXSDouh/OMAMYD24DrM23Y01MtEydO5Nxzz2Xy5MmMHz+eUaNGcfzxx2d56vyQSqXyHUIkeA+38B5uERcPT84URO7sjRtuuIEZM2Ywf/58du7cmePe7pDUNphE7yQ6g/eOmuJBOaqn35hZLfCgpDeApyX9oI/t10n6CzATeCKH8zRL+hVwbaZtmutr2XD9BaRMDJNx9bB/57TTTuNXv/oVf/7zn9ljjz343Oc+x89//nNOOeUUzIyRI0dSW1tLeXl5cIzmZkaNGsWuXbuQRGVlJXV1dQwfPpyOjg5aWloYPXo0NTU1DBs2jIqKCurq6qioqCCVStHa2tpVX1xcTHl5OfX19VRWVtLa2kpbW1tXfUlJCaWlpTQ0NDBixAiam5tJpVKMGDGC6upqSktLKS4uprGxkaqqKhobG2lvb+/av6ysjKKiIpqamqiqqqKhocEpp/LycmpqakilUl31hejUGU/3v1OhObW3t5NKpXr9v1cITu3t7TQ1NWXdnlx1am9vp6ampl+fEX05edzH5dy57xEnkUodRU1NDfX19XR0dFBdXU1VVRVz587lq1/9KqNGjWLRokUsWLCAm266yYncmWtbb29vp6WlxYk8M5SfX5WVlezYsSNWTn39nTq/V8XJKd95xoXvA1F/b+sLmVm2n72eQULSRuCLZraiW/km4HvAocAWM7tY0hTgFaDEzFKSDgNWAZ83s/uyPY+kCuAmgv8Dp/e0/bKHVtjVL7x9s3ztwmP45S9/yfLly7nllqCj+Re/+AVPP/00P/rRj/phPjRUV1czduzYfIcxYLyHW3gPtxhMj5UrVz47e/bsQwbl4J5+Uyi5E4L8CbBx40bmzJnD888/v9t+vdUVAnH5LMmVJHon0Rm8d670lTv9nVC3eR3I9ABJtaQyoBz4AXB/lse8X1IKGAFsB07ItOGUMZWsXXj4O8omT57M008/TWNjI8OHD2flypUccojb383iMqW293AL7+EWcfHwRIJzuTMTW7duZfz48QDcd99975g5t9BIahtMoncSncF7R42/CHWbicBbGerGAgZ8HZgHlACtWRzz42Fv7jDgFOC3kt5nZm9037Cntc4OPfRQPvWpT/H+97+f4uJiDj74YL785S9nqZMfiovj8d/ce7iF93CLuHh4IsG53Akwb948Vq9eTXV1NZMmTWLx4sWsXr2a9evXI4kpU6bw3//931mE4iZJbYNJ9E6iM3jvyI87KEf1DBhJHyRIpP9LMKRoN8ysHfiBpE8AX6WXZ1Qy7HuvpP8GjgR+1X2bTGudLV68mMWLF2d7qrzT2NhIRUVFvsMYMN7DLbyHW8TFwzMwXM6dy5Yt263s7LPPzvbUzpPUNphE7yQ6g/eOGj87rmNIGilpDnAXcLuZ/TmL3S4HzpNUnsN5JOkUYA/ghZ62GTZsWLaHc5qqqqp8hxAJ3sMtvIdbxMXD0z987sw/SW2DSfROojN476jxF6Hu8BtJdcBm4CLgauCsLPd9GNgJfCnL89QDu4DvAGeY2V962jBTb26h0djYmO8QIsF7uIX3cIu4eHhyxudOR0hqG0yidxKdwXtHjR+O6wBmNqWP+jPTft5It2XILJjieNpAz9PD9rls7ixxWWLBe7iF93CLuHh4ssfnTrdIahtMoncSncF7R42/E+rJSFyGFI0ePTrfIUSC93AL7+EWcfHwFD5xyZ25ktQ2mETvJDqD944afxEaIyRNllSf4TU51+PFpcenpqYm3yFEgvdwC+/hFnHx8Aw9PndGQ1LbYBK9k+gM3jtq/HDcGGFmmwjWMIuETNPMFxplZWX5DiESvIdbeA+3iIuHZ+jxuTMaktoGk+idRGfw3lHj74R6MhKXRFpUFI//5t7DLbyHW8TFw1P4xCV35kpS22ASvZPoDN478uMOylE9sSAuM/w1NTXlO4RI8B5u4T3cIi4ensInLrkzV5LaBpPonURn8N5R4y9CPRmJy+QKcVnXyXu4hfdwi7h4eAqfuOTOXElqG0yidxKdwXtHjb8I9WQkLpMrNDQ05DuESPAebuE93CIuHp7CJy65M1eS2gaT6J1EZ/DeUeMvQj2xJy5rtnkPt/AebhEXD4+nUElqG0yidxKdwXtHjb8I9WSkpyFFU6ZM4cADD2TmzJkccsgheYgqd0aOHJnvECLBe7iF93CLuHh4Cp9Mw3Hnz5/PuHHjmD59elfZJZdcwowZM5g5cybHH388r7/++lCFGTlJbYNJ9E6iM3jvqPEXoZ6MZBpS9OSTT7J+/XrWrVs3xBH1j9ra2nyHEAnewy28h1vExcNT+GTKnWeeeSbLly9/R9nChQv5v//7P9avX8+cOXNYsmTJUIQ4KCS1DSbRO4nO4L2jZkguQiVtlNTUbQHoz0na0sO2qyV9Mfz5Ukkm6dNp9cVh2ZS0sg9JekRSjaS3JK2RdFYWcY2UdK2kTWFMG8L3Y9PiPraX/d8tqUPSj3qoO0XSekm7JFVLWtkZs6TRkn4m6Q1JdZL+Lun8LOI1Se/JUDdJ0h2SdkhqCH8Hc7ptI0kLJD0fbrNF0i8lHdjTMV95q5EPXrmqr7Ccp7y8PN8hRIL3cAvv4RZx8eiOz5/R5M9wX0l6WdJfe6hbLak5dKmWdK+k8WHdj9N+962S2tLeP9r9WJ25s/PVyVFHHcWee+75jm3T7zA0NDQU9PIucW2DfZFE7yQ6g/eOmqG8E3qymY3ofAHZjjl5C1giqcfxLZIOB1YBvwXeA4wBvgKc2NtBJZUCK4FpwL8AI4EPAzuAD2UZ2+nATmCupK6VXMMLxV8A3wBGAe8GfgR0ztt+DcHC2FPD+o8B/8jynD257An8L9Aa+owNz3GnpE+lbXod8DVgAbAnsD9wP/CvOZyL448/ng984AP85Cc/6W/IHo/H48kenz+jyZ9HAeOAfSV9sIf6/wh/v+8Jz3EVgJn9e9rv/rvA3Wl/j15/V9lw0UUXsffee3PHHXcU9J1Qj8fjyYXifAeQBcuB6cAXgKU91F8JLDWzK9LKngU+08dxTwcmA0ebWX1Yth24LIfYTgcuBi4FTgb+P3vvHidHWebtX9/MZDKZQxIgoBEWkPVATIAhCAgGNhBAwaAoygoqoEQXXZf1RVEQeAXkB4obJB7Y/a2iYKJkwQNyiCwkHNRAEBCUgyAIIYSTTJLJzPTMZKZn7vePqh6aTvdM96QyXV11X59PfWb6qdN9dffTdz1VTz3187C8DXjWzFaGr7uAX+Sttz9wnpltDF8/EU5j5f8A3cBpZpZL1NdK2hVYJOkXBEn1X4GDzOwPeev+tNRGBzObePq757DfslaGhob4xCc+wd13301TUxOdnZ184AMf4I1vfCPvec97yGQymBlTpkxh06ZNw2dN+vr6mDp1Kp2dnUiiubmZrq4uJk+ezNDQEJs3b2batGl0dHRQV1dHU1MTXV1dNDU1kc1m6e/vH55fX19PY2Mj3d3dNDc309/fz8DAwPD8iRMn0tDQQCaToaWlhb6+PrLZLIODg/T19dHQ0EB9fT09PT20trbS09PD4ODg8PqTJk1iwoQJ9Pb20traGjun/v5+stks2Wx2eH4tOmUyGerq6rb4nGrNqb29ncbGxhG/e7Xg9Morr1BXV1d2fYqr0yuvvEI2mx3Tb8RoTjWM58/Xcwrwa2By+P/9xRYysw5JNxDkzIrJ5c6siToZl9edzhlnnEFHRwfd3d0MDQ3R3t4+XC/OPPNMLrjgAi688EIuv/xyzj333Fjkzkrr+iuvvMLEiRNjkWfG8/crm82yefPmRDmN9jnljquS5FTtPBOH44Goj9tGxcy2+QSsAY4oKJsHrCuy7F3AwvD/C4ClBGc6nwEmEjScDdgdaAIGCRJhpTEtI0i+FcWdN+8QYDOwHfBd4Ma8eXsAfQRnbA8DWgrW/SHwGPBJ4K0VxGzAW4qUrwYuLFL+5nCdtwOnA89V8h5dc+MKe+dlK60UX/va1+xb3/pWyflxob+/v9ohRIJ7xAv3iBfb0mPFihUP2DjkymKT58/I8mcT0AkcAxwPtAMNJd67HYAVwK+LbOcCYOlI+8rlztyUz7PPPmuzZs0q+j1bs2ZNyXm1QFJ+Syoljd5pdDZz70oZLXeOZ3fcG8J7TnJnGMvGzG4EXgUWFszajqBL8UtjiGeHMa6X4xTgNxacjf0ZcLSknQDM7BmCg4SdgeuAdklXS2oJ1/03giuQnwceD++l2ZouPdMp7vJS3vyKfd+yw2TuP+vw4deZTIaurq7h/2+77bbXjfIXVzo7O6sdQiS4R7xwj3iRFI8SeP7c+vz5IYKG723AzQQN8sJbUb4jaRNBA3V6uK+KyeXO3DQSTz311PD/N954I3vuuedYdhkLEl4HS5JG7zQ6g3tHzXg2Qo8zs2nhdByQJTgzW8hEYKBI+XnAuUD+3bEbCe4TmTGGeNaPcT0kTQY+QtiV1czuBdYCJ+WWMbPVZnaCme1IcNb30DB+zKzXzC4xs/0Ikvl1wPXhvZ1job2Ey4y8+WP2zfHKK68wd+5c9tlnHw444ADe97738d73vndrNjku1PJAD/m4R7xwj3iRFI8SeP7c+vx5CnCdmWXNbDPwy7AsnzPMbCqwN0EjfZexOJbixBNP5KCDDuLJJ59kl1124aqrruLss89m9uzZ7L333tx2220sXrw4yl2OKwmvgyVJo3cancG9o6aa94SuBaZLarHwnhIFlrsBzxUubGa3S3oa+FxeWY+kewm61txZ4f5XABdLajazTIXrfpBgIIYrJX03LJtGcI/LFUViv1/SLwnuzSmc1ynpEuAcgu6zGyqMBQKX4yVdaK/dEwrBfT3PA38lONj4vqR3mllZz1YpfNbZHnvswZ/+9KcxhFddmpubqx1CJLhHvHCPeJEUjzLx/En5+VPSLsDhwAGSjg+Lm4BGSdPNrL1gu49IupggZ84xs4qe1F7qOaHXXnvtFmWnnXZaJZuONSmrg8Ok0TuNzuDeUVO154Sa2VrgPuCbklrC0fHOIjjDu7rEaucCXy4o+zJwqqSzJO0AIGkfSctGCWEJQQPtF5L2lDRB0g6SvirpmLzlJkpqzJvqCc6e/gjYi2AQhTbg3UCbpL0kzZX06Vz3Ikl7EtyXszp8fb6k/SU1SGokGLG2A3hylJgBGgriqSO4d2YKcJWkN4blJ4bv11lh1+ynCEYYvFbSvNy+JX1U0tnFdlTjA3IMk+tCXOu4R7xwj3iRFI9y8PxZcf78BMHJ2Lfn7fNtwDrgxBLrXEMwku77R3kvtiApubNS0lQH80mjdxqdwb2jpmqN0JB/JviRfxp4AZgPHGNmfcUWNrNVwB8Kyu4hOMN5OPCMpA3AfwPLR9px2B3nCIJR9W4nGLDgDwT3gdyXt+hyoDdv+kEY5xVm9nLe9CDBSISnECTE9wOPSOoOy38FXJbbPfBjgm6yLwJHAu+z10YZHInHCuL5pJmtB+YSdLV6nKCr1JnAJ8zsf/LWPQP4HvD9MMa/EZyVvqnYjiZMqPbXIxomT55c7RAiwT3ihXvEi6R4VIDnz/Lz5ynAlQX7fBn4L7bskptz7Ae+A5w/0ntRjKTkzkpJYR0E0umdRmdw76hRhb1MnBRx77332kEHHVTtMLaarq4uWltbqx3GVuMe8cI94sW29Fi5cuWD8+fPf+c22biTOJKSOyslKb8llZJG7zQ6g3tXymi5M52n65yySMoJis2bN1c7hEhwj3jhHvEiKR5O7ZOU3Fkpaa2DafROozO4d9QkuhEa3p/SXWT6TbVjK4akQ0rEW0433cgpNbhCrTFt2rRqhxAJ7hEv3CNeJMUjLnj+HDtJyZ2VktY6mEbvNDqDe0dNohuh4TDuLUWmrXkm5zbDzH5XIt6W0deOnqQMrtDR0VHtECLBPeKFe8SLpHjEBc+fYycpubNS0loH0+idRmdw76hJdCPU2TqS8jykpJyVdo944R7xIikeTu2TlNxZKWmtg2n0TqMzuHfUeCPUKUlSRvhramqqdgiR4B7xwj3iRVI8nNonKbmzUtJaB9PonUZncO+oSecvpVMWSelSlJTnOrlHvHCPeJEUD6f2SUrurJS01sE0eqfRGdw7arwR6pQkKWdzk3Lmyj3ihXvEi6R4OLVPUnJnpaS1DqbRO43O4N5Rk85fSqcskjLMfDabrXYIkeAe8cI94kVSPJzaJym5s1LSWgfT6J1GZ3DvqPFGqFOSYol0cHCQfffdlwULFlQhorHR399f7RAiwT3ihXvEi6R4OLVPqUbopz71KXbaaSdmz549XHb++eez995709bWxlFHHcWLL744XmFGTlrrYBq90+gM7h013gh1SlJsNKzFixczc+bMKkQzdpLyXCf3iBfuES+S4uHUPqVGkjz11FO59dZbX1d21lln8ec//5mHH36YBQsWcNFFF41HiNuEtNbBNHqn0RncO2q8ERoDJK2R1CupS1KHpHsknS5pQjj/akkXh//vLsnyHsS9RtLZFeznFUnNeWULJd1VbPlnXu1i/2/dMfx63bp13HLLLSxcuHBrdMedpDzXyT3ihXvEi6R4OJUxjvlzbrjtTZI2SFolaf9iy+ZyZ27Kceihh7L99tu/btkpU6YM/5/JZGr68S5prYNp9E6jM7h31NRvk606Y+FYM1shaSrwT8Bi4EDgkyWWn2ZmWUnvBO6W9KCZ3V7GfuqBfwcuGW3Bwg5FX/jCF7jssstqbnSw+vpkfM3dI164R7xIioczJrZp/pQ0BbgZ+CxwHdAAHAJsLrZ8pXeEnnvuufzkJz9h6tSp3HnnnRWuHR/SWgfT6J1GZ3DvyLe7TbbqjBkz2wTcKOllYLWkRaMs/4Ckx4A2oJxG6LeAL0u60sxGPLXR372Jp797Dvsta2Xjxo3MmDGDffbZh1tuuYVsNksmk6G3t5fW1lYymQxmxpQpU9i0aRONjY0A9PX1MXXqVDo7O5FEc3MzXV1dTJ48maGhITZv3sy0adPo6Oigrq6OpqYmurq6aGpqIpvN0t/fPzy/vr6exsZGuru7aW5upr+/n4GBgeH5EydOpKGhgUwmQ0tLC319fWSzWRobG2lvb6ehoYH6+np6enpobW2lp6eHwcHB4fUnTZrEhAkTYutUX19PR0cH2Wx2eH4tOpkZvb29W3xOtebU09NDS0vLiN+9WnDq6elh0qRJZdenuDr19PQAjOk3YjQnpzbYhvnzbeHy14ave4HbSi2cy51ZE3UyLq87nTPOOIOOjg66u7sZGhqivb19uF6ceeaZXHDBBVx44YVcfvnlnHvuubHInZXW9Z6eHiZPnhyLPDOev1+TJk1i/fr1iXIa7XPKHVclyanaeSYOxwNRH7eNhtI6iluckLQGWGhmKwrK1wKXEpzRXWdm50naHXgWmBieyX0XcAfwMTP7VTn7AT4HPB5ubyHwcTObV7j8tTevsMv/MoH7zzqcc845hyVLllBfX09fXx+dnZ186EMfYunSpVtpv+1pb29n+vTp1Q5jq3GPeOEe8WJbeqxcufLB+fPnv3ObbNzZKsYjf4ZXQp8luBq6DFhtZhtLLZ/LnTnuP+vw4f/XrFnDggULePTRR7dY77nnnuN973tf0Xm1QFJ+Syoljd5pdAb3rpTRcqffExpvXgS2LzGvXVIvcC9wJXBDBdv9v8C/SdpxpIXePL1lOHleeumlrFu3jjVr1rBs2TIOP/zwmmiAAjQ3N4++UA3gHvHCPeJFUjycyIgsf5pZJzCXoKftD4BXJd0o6Q3Fls/lztw0Ek899dTw/zfeeCN77rnniMvHmbTWwTR6p9EZ3DtqvBEab3YGNpSYNx1oAb4EzAMmlrtRM3uU4IzuiAMyJOUqeVKG1HaPeOEe8SIpHk5kRJo/zewvZnaqme0CzAbeBFxRYtmi2zjxxBM56KCDePLJJ9lll1246qqrOPvss5k9ezZ77703t912G4sXLx5VLK6ktQ6m0TuNzuDeUeP3hMaUcNS9nYHfE3Qn2gIzGwQWSfogQRfbogmxBF8D/giUvGemVCKdN28e8+bNq2BX1WVgYKDaIUSCe8QL94gXSfFwtp5tnT/N7AlJVwP/UmJ+0fWuvfbaLcpOO+20cncbe9JaB9PonUZncO+o8SuhMUPSFEkLCO47WWpmj5Sx2jcIBhtqLHc/ZvY08D/AGaWWKfWss1ojKc91co944R7xIikeztjZVvlT0p6Svihpl/D1PwAnAquLLZ+U3Fkpaa2DafROozO4d9R4IzQ+3CSpC3geOBe4nNLDyxdyC7AR+HSF+7wIKNnROymjQibluU7uES/cI14kxcMZE9s6f3YRXFG9T1KGoPH5KPDFYgsnJXdWSlrrYBq90+gM7h013h03BpjZ7qPMPzXv/zWACuYbMKvS/ZjZ88BIZ39H22RNMHFi2bfLxhr3iBfuES+S4uFUxnjkTzN7ATih3JiSkjsrJa11MI3eaXQG944avxLqlCQpibShoaHaIUSCe8QL94gXSfFwap+k5M5KSWsdTKN3Gp3BvaPGG6EJQtKukrpLTLtWur2hoaFtEea4k8lkqh1CJLhHvHCPeJEUD6c6RJk/k5I7KyWtdTCN3ml0BveOGu+OmyDMbC3BsPORMGFCMs5RtLRE9pZUFfeIF+4RL5Li4VSHKPNnUnJnpaS1DqbRO43O4N5Rk85fSqcskvKc0L6+vmqHEAnuES/cI14kxcOpfZKSOyslrXUwjd5pdAb3jhpvhDolSUoizWaz1Q4hEtwjXrhHvEiKh1P7JCV3Vkpa62AavdPoDO4dNd4IdUqSlGedJeW5Tu4RL9wjXiTFw6l9kpI7KyWtdTCN3ml0BveOGm+EOiVJyrPOkvJcJ/eIF+4RL5Li4dQ+ScmdlZLWOphG7zQ6g3tHjTdCnZIkZZj5pAyp7R7xwj3iRVI8nNonKbmzUtJaB9PonUZncO+o8UaoU5JiiXRwcJB9992XBQsWVCGisVFfn4xBoN0jXrhHvEiKh1P7FMud3/72t5k1axazZ8/mxBNPTOQAJ2mtg2n0TqMzuHfUeCPUKUmxZ50tXryYmTNnViGasdPT01PtECLBPeKFe8SLpHiMB5LWSOoNn4H5iqQfSxrXZw9IeizvOZyDkvryXn91PGMJ4zlC0lC4/y5JT0g6JW/+JEkXSXpaUiZ8D39Y7BmihbnzhRde4Dvf+Q4PPPAAjz76KIODgyxbtmwcrMaXtNbBNHqn0RncO2q8ERoT8g4KuiR1SLpH0umSJoTzr5Z0cfj/7pIsL2GvkXR2mfuZG257k6QNklZJ2r/Ysn9b38v+37pj+PW6deu45ZZbWLhwYQTG40dra2u1Q4gE94gX7hEvkuIxjhxrZi3AHGB/4Lzx3LmZzTKzljCG3wGfz702s0vGM5Y81obxTAHOBa6S9HYFlzZ/CRwN/DMwFWgD/gwcXriRXO7Mz5/ZbJbe3l6y2Sw9PT286U1vGged8SWtdTCN3ml0BveOGm+ExotjzawV2A34BvAV4KoRlp8WJswPA+dLOnKkjUuaAtwMfBfYHtgZuBDYXGz5Or1+mPkvfOELXHbZZTX3IO6knLlyj3jhHvEiKR7jjZm9APwGmC3pk5L+Ep4MfUbSv+SWC8sX5L2ul9QuaU74+l3hCc4OSX+SNG+sMUlqDLczM69shqQeSTuEVy3XSPq/ktZLelbSRwvWv1zS8+GV3islNVbwnpiZ/QLoAmYC7wEOA44zswfNLGtmHWb2HTO7unD9wty5884786UvfYldd92VGTNmMHXqVI466qiK35e4k9Y6mEbvNDqDe0dNOjs3xxwz2wTcKOllYLWkRaMs/4CkxwjOzN4+wqJvC5e/NnzdC9xWauFsZhNPf/cc9lvWysaNG5kxYwb77LMPt9xyC9lslkwmQ29vL62trWQyGcyMKVOmsGnTJhobg3zf19fH1KlT6ezsRBLNzc10dXUxefJkhoaG2Lx5M9OmTaOjo4O6ujqampro6uqiqamJbDZLf3//8Pz6+noaGxvp7u6mubmZ/v5+BgYGhudPnDiRhoYGMpkMLS0t9PX1kc1mGRwcpL29nYaGBurr6+np6aG1tZWenh4GBweH1580aRITJkyIrVN/fz+SyGazw/Nr0SmTydDY2LjF51RrThs2bKC1tXXE714tOG3YsIHGxsay61NcnTZs2ICkMf1GjOaUZCT9A3AMwZW+vwMLgGeAQ4HfSLrfzP4IXAucSHAiE4KGWbuZ/VHSzsAtwCeAW4H5wC8k7Wlmr1Yak5n1SboO+DjBFUmAk4D/NbP14T2XuwCtwJuAdwM3S3rAzJ4G/oPgJOvewCCwLNzO+WW+JxOADwEtwCPAZ4F7wwb7qORyZ9bEvj9r4vjjj2flypU88MAD7LTTTpxyyin853/+J5/61Kdi8ZscVV3fsGEDTU1NiXIq5/crm82yfv36RDmN9jnljquS5FTtPBOH44Goj9tG/a1N60OV44akNcBCM1tRUL4WuBQ4EFhnZudJ2h14FphoGNrzrwAAIABJREFUZllJ7wLuAD5mZr8aYR9TwvVuJkjKq81sY6nlf3LTSvvuE+L+sw7nnHPOYcmSJdTX19PX10dnZycf+tCHWLp06VZ5jwfZbDYRN5O7R7xwj3ixLT1Wrlz54Pz589+5TTZeBcJ8Mx3IApsIGpBfNLPeguVuAO40s8WS3gI8BLzBzHok/RR40swukvQVYLaZfSJv3f8FfmZm15QRz13AUjP7YV7Zu4GlwB5mZpIeBi4ys19KOgJYTtAbqCdc/pfA/cA3gR7g7Wb2XDjvEOBHZvbWEWI4guCkbCcwBDwHXGpm10n6MUG+/fhoLvBa7gS4/6zDuf7667n11lu56qqgY9NPfvITVq9ezZVXXlnO5mqGpPyWVEoavdPoDO5dKaPlztrqV5lOXiToOluMdkm9wL3AlcANI23IzDqBuYABPwBelXSjpDcUW/4tO0zm/rOC210uvfRS1q1bx5o1a1i2bBmHH354TTRAITnPdXKPeOEe8SIpHuPIcWY2zcx2M7PPmVmvpKMlrQ7HC+gguEI6HSC8wvgX4FhJTcD7gZ+F29oN+EjYhbYjXHcuMGOswZnZKoJG8lxJs4FdCRrLOdbnGqAhzxFcFX0jMAn4U14sNwM7lbHbteF7sr2Z7Wtm1+X2VYlLLnfm8ueuu+7K6tWr6enpwcxYuXJlzQ3wVw5prYNp9E6jM7h31HgjNP7sDGwoMW86QXehLwHzgImjbczM/mJmp5rZLsBsgqR9RbFllZBnnU2aNKnaIUSCe8QL94gXSfGoFpImAb8g6Mr6BjObRnC1MT8R5LrkfgB4PGyYAjwPLAkbcLmp2cy+sZVh/YSgS+4ngOvMLH/8gh0kTc57vSvBSdtXgH6CK6G5WKaa2dStiGMFcJCkskYTKsydBx54IB/+8IeZM2cOe+21F0NDQ3zmM5/ZinDiSVrrYBq90+gM7h013giNMeGotTsDvy+1jJkNmtkioA/4XCXbN7MngKsJGqPF9l90vXnz5nHzzTcXnRdHam0gpVK4R7xwj3iRFI8q0kBwBfFVICvpaKBw9JxlYdlnee0qKATdZo+V9B5JdeHAQPMk7bKVMS0hGHjvJIIGaT4TgAskNYSDIB0N/NzMBoEfAldI2lEBu0jampGA/he4E/iVpH1DxymSPqe8x7jkKJY7L7zwQp544gkeffRRlixZksiD2bTWwTR6p9EZ3Dvy7W6TrTpbRZjcFhAk/KVm9kgZq30D+PJIIwBK2lPSF3MHBuGAFCcCq4stX+w5obVIb2/v6AvVAO4RL9wjXiTFo1qYWRdwBnAdsJGg4XdjwTIvEdz+cTDwP3nlzxNcHf0qQSP2eeAstvIYw8zWEAwM1G9m9xTMXgdkgJeAawjGVHgqnPdFgu65fyC45/U2oOT9oGXEYQQDFd0G/JzgvtFHCAYDvKNw+aTkzkpJax1Mo3cancG9oyZ9d9fGm5skZQkGRXgcuBz4rzLXvYXgwOHTBI9gKUYXwQBHZ0qaBuTulTmr2MJ1dXXlRx5jkvJcJ/eIF+4RL5LiMR6Y2e4lyr8PfH+UdeeXKL8P+KcxxjNvhNlrgbtKrHcRcFGR8j7g7HAqN4YVwO4jzN9MMLruqCPsJiV3Vkpa62AavdPoDO4dNd4IjQmlDgry5p+a9/8aXn+fTu5M7axRtvECcEK5MSXl0QSZTCYRXZ/cI164R7xIiofzGpL2ILjCule1Y6mEpOTOSklrHUyjdxqdwb2jxrvjOoknKY8hco944R7xIikeSUNSd4npkFHWuxT4E3CJma2NKJbzS8RyUxTbTztprYNp9E6jM7h31PiV0IQhaVeCrrzFeEclyTwpXYqmTJlS7RAiwT3ihXvEi6R4JA0zaxnjeucA55SYN2LX2RG2+XXg62OJpxKSkjsrJa11MI3eaXQG944avxKaMMxsrZm1lJgqOpuclC5FmzZtqnYIkeAe8cI94kVSPJzaJym5s1LSWgfT6J1GZ3DvqPFGqFOSpAxF3dhYcsDgmsI94oV7xIukeDi1T1JyZ6WktQ6m0TuNzuDeUZPOX0rHcRzHcRzHcRynKngj1ClJUp511tfXV+0QIsE94oV7xIukeDi1T1JyZ6WktQ6m0TuNzuDeUeONUKckSRlcYerUqdUOIRLcI164R7xIiodT+yQld1ZKWutgGr3T6AzuHTXeCHVKkpTBFTo7O6sdQiS4R7xwj3iRFA+n9klK7qyUtNbBNHqn0RncO2q8EeqMyvPPP89hhx3GzJkzmTVrFosXL652SBUhqdohRIJ7xAv3iBdJ8XCSx+DgIPvuuy8LFiyodijblLTWwTR6p9EZ3Dtq/DmhTklyXYrq6+tZtGgRc+bMoauri/32248jjzySd7zjHVWOsDyam5urHUIkuEe8cI94kRQPp/Yp7I67ePFiZs6cmfirKGmtg2n0TqMzuHfUjMuVUElrJPVK6s6bTpK0rsiyd0laGP5/gSST9JG8+fVh2e55ZQdIWi6pQ9IGSX+Q9Mky4poi6QpJa8OYng5fT8+L+4gR1n+zpCFJVxaZ9wFJD0vqlNQuaWUuZknTJP1I0suSuiT9VdJXyohXks6Q9KikjKR1kq6XtFfeMgdLuiPc7iZJN0l6R978ecXe92I882oXADNmzGDOnDkAtLa2MnPmTF544YVyNhELurq6qh1CJLhHvHCPeJEUD6cy8o4vusJjgHsknS5pQjj/akkXh//vHh4/5I5D1kg6u8L95B/HfK/YsrncCbBu3TpuueUWFi5cGIFtvElrHUyjdxqdwb2jZjy74x5rZi25CXixzPU2ABdJKnqnv6SDgDuAu4G3ADsAnwWOHmmjkhqAlcAs4L3AFOBgYD1wQJmxnQxsBD4qaVLett8C/AT4IjAVeDNwJZAbMu/bQAswM5z/fuBvZexvMfDvwBnA9sDbgBuA94X7PQi4Dfg18KZwv38CVknao0ynYQZty7I1a9bw0EMPceCBB1a6uaoxefLkaocQCe4RL9wjXiTFwxkTx5pZK7Ab8A3gK8BVIyw/LTwO+TBwvqQjK9hPS970+WIL5efOL3zhC1x22WWpeHZoWutgGr3T6AzuHTW10B33VmA28HHgmiLzvwVcY2bfzCt7EDhhlO2eDOwKHGZm3WHZ34GvVxDbycB5wAXAscDPw/I24FkzWxm+7gJ+kbfe/sB5ZrYxfP1EOJVE0luBfwUOMrM/5M36ad7/lwE/MbP8mzbPk7RfGOPJ5WkFDGY20dbWRl1dHUNDQ5xwwglcf/31XHTRRUgik8nQ29tLa2srmUwGM2PKlCls2rRp+MG2fX19TJ06lc7OTiTR3NxMV1cXkydPZmhoiM2bNzNt2jQ6Ojqoq6ujqamJrq4umpqayGaz9Pf3D8+vr6+nsbGR7u5umpub6e/vZ2BgYHj+xIkTaWhoIJPJ0NLSQl9fH9lslrq6Onp7e2loaKC+vp6enh5aW1vp6elhcHBweP1JkyYxYcKE2DoBDAwMkM1mh+fXolN/fz8TJkzY4nOqNadNmzYxadKkEb97teC0fv16JkyYUHZ9iqvT+vXrGRgYGNNvxGhOTm1gZpuAGyW9DKyWtGiU5R+Q9BhBzr49qjhyubO7u5uOjg4OPvhg9tlnH/r7++nq6orF79e2qOvr16+nvr4+UU7lfE4TJkygr68vUU6jfU6546okOVU7z8TheCDq47ZRMbNtPgFrgCMKyuYB64osexewMPz/AmApwZXCZ4CJBA1nA3YHmoBBgoZkpTEtI2i8VhR33rxDgM3AdsB3gRvz5u0B9BFc8TwMaClY94fAY8AngbeWGe/pwHMjzC/5XoT7eWmk973Y9LObbrcc/f39dtRRR9miRYus1nj11VerHUIkuEe8cI94sS09VqxY8YCNQ670qfKpVJ4G1hL0iroauDgs2z08fqgPX78L6AE+ONb9FJtyufPss8+2nXfe2XbbbTd7wxveYJMnT7aPfexjllSS8ltSKWn0TqOzmXtXymi5czz7h9wQ3q/RIemGSlY0sxuBV4HCmyq2I+hS/NIY4tlhjOvlOAX4jQVXM38GHC1pJwAze4agsbczcB3QHt6X0hKu+28EVzA/Dzwe3os6YvfhMuLdntLvxUvA9LKs8thjx1YgOFFx2mmnMXPmTM4888xKN1N1pk2bVu0QIsE94oV7xIukeDiR8SJBXixGu6Re4F6CW2XKPSbJP47pkPTpYgvlcuell17KunXrWLNmDcuWLePwww9n6dKlFWrUDmmtg2n0TqMzuHfUjGcj9DgzmxZOxwFZgiubhUwEBoqUnwecCzTmlW0kuM9yxhjiWT/G9ZA0GfgIYVdYM7uX4KzrSbllzGy1mZ1gZjsSXDU9NIwfM+s1s0vMbD+CxuV1wPWSSiXMcuId6b2YAbSXqTdM7lL6qlWrWLJkCXfccQdtbW20tbWxfPnySjdXNTo6OqodQiS4R7xwj3iRFA8nMnYmGFOiGNMJxmX4EsEJ42LHIsXIP46ZZmY/KLZQWrtwp7UOptE7jc7g3lFTzTvl1wLT864OIkkEAws8V7iwmd0OPA18Lq+sh+BM5vFj2P8K4D2SxjLu8AcJBjK6Mhzh9mWChFf0nkszux/4JcG9rYXzOoFLgGaCgYRKsRLYRdI7S+wjQ/BefKTI7BPC9StC4XOB5s6di5nx5z//mYcffpiHH36YY445ptLNVY3C4fJrFfeIF+4RL5Li4Ww9kvYnyMm/L7WMmQ2a2SKCW2c+V2q5Me5/i7J58+Zx8803R7mb2JHWOphG7zQ6g3tHTdUaoWa2FrgP+KaklnB02bMIrpCuLrHaucCXC8q+DJwq6SxJOwBI2kfSslFCWAI8D/xC0p6SJkjaQdJXJeW3sCZKasyb6gm64v4I2ItgQIM24N1Am6S9JM2V9Olc91xJexLc17o6fH2+pP0lNUhqJBjxtgN4coT36ymCbkPXho9ZaQjj+WjeEPNnA6eEj3FplbRdODT9QcCF+dsrcGpUkayZlNH8mpqaqh1CJLhHvHCPeJEUD2fsKHjs2gKCMR+WmtkjZaz2DeDLYS6OhKTkzkpJax1Mo3cancG9o6bav5T/DOxEcIXzBWA+cIyZ9RVb2MxWAX8oKLsHODycnpG0AfhvYMT+oma2GTiCYFTa24HOcNvTCRrHOZYDvXnTD8I4rzCzl/OmBwlG8j2FoEH5fuARSd1h+a8IRq+FYGCEHxN0kX0ROBJ4n702Sm8pzgC+B3w/3MffCK7K3hQ6/R54D/AhgvtAnwP2BeaGjdgcOxc49QL/WLizpHQpSspzndwjXrhHvEiKhzMmbpLURXBi+VzgcoIB+crhFoLbWYre31lkP/nPCf1VsYWSkjsrJa11MI3eaXQG946acXlEi5ntXqL8eYp3H83Nv6BI2Rb9QC14ZMloA/sU2/4m4AvhVGz+7iVWLZrcCmI7doT9XgxcXF6Ur1vPCJ4VuniEZX5PcI9Lqfl3AVv2FSpCUs7mJuXMlXvEC/eIF0nxcCpjhDydm39q3v9rKMh/YV6dtbX7yScpubNS0loH0+idRmdw76hJ5y+lUxZBbq59stlstUOIBPeIF+4RL5Li4dQ+ScmdlZLWOphG7zQ6g3tHTaIboeH9nd1Fpt9UO7ZiSDqkRLyjddPdJiQlkfb391c7hEhwj3jhHvEiKR5OdZC0a6n8K2nXSraVlNxZKWmtg2n0TqMzuHfUjEt33GphZpcQjDxbE5jZ7wiGjY8FSRkFLCnPdXKPeOEe8SIpHk51CAdLjCT/JiV3Vkpa62AavdPoDO4dNYm+EupsHUkZXCEpz3Vyj3jhHvEiKR5O7ZOU3Fkpaa2DafROozO4d9R4I9QpSbFnndUi9fXJuODvHvHCPeJFUjyc2icpubNS0loH0+idRmdw76jxRqhTkqQk0sbGyB7/VlXcI164R7xIiodT+yQld1ZKWutgGr3T6AzuHTXeCHVKMjQ0VO0QIqG7uyrjOkWOe8QL94gXSfFwap+k5M5KSWsdTKN3Gp3BvaPGG6FOSZLyrLPm5uZqhxAJ7hEv3CNeJMXDqX2SkjsrJa11MI3eaXQG946adP5SOmWRG2b++eef57DDDmPmzJnMmjWLxYsXVzmyykjKkNruES/cI14kxcOpfXK5s6+vjwMOOIB99tmHWbNm8bWvfa3KkW1b0loH0+idRmdw76hJ5x22TlnkEml9fT2LFi1izpw5dHV1sd9++3HkkUfyjne8o8oRlsfAwEC1Q4gE94gX7hEvkuLh1D653Dlp0iTuuOMOWlpaGBgYYO7cuRx99NG8613vqnKE24a01sE0eqfRGdw7avxKaJWQtEZSr6QuSR2S7pF0uqQJ4fyrJV0c/r+7JMt7cPYaSWeXsY9zJC0vKHuqRNlHC9d/en0vADNmzGDOnDkAtLa2MnPmTF544YWxqo87SXmuk3vEC/eIF0nxyJGXI7rzppMkrSuy7F2SFob/XxDmi4/kza8Py3bPKztA0vIw/2yQ9AdJnywjrimSrpC0Nozp6fD19Ly4jxhh/TdLGpJ0ZZF5H5D0sKROSe2SVuZiljRN0o8kvRzmzb9K+koZ8ZqkTN572BGWPynphLzl3h0uW1jWLWkfSZskva1g2yslXVq4z1zulERLS/Do0YGBAQYGBhI9aFHS6mC5pNE7jc7g3lHjjdDqcqyZtQK7Ad8AvgJcNcLy08ysBfgwcL6kI0fZ/m+Bd0uqA5D0RmAiMKeg7C3hsq9jomyLDa5Zs4aHHnqIAw88cDS32JCU5zq5R7xwj3iRFI8CjjWzltwEvFjmehuAi3K/84VIOgi4A7ib4Pd/B+CzwNEjbVRSA7ASmAW8F5gCHAysBw4oM7aTgY3ARyVNytv2W4CfAF8EpgJvBq4EcqP8fBtoAWaG898P/K3Mfe6T9z7mjqZ+C/xT3jKHAk8UKbvHzP4E/AdwlcJWpKTTgJ2BCwt3lp87BwcHaWtrY6edduLII4+sqdxZKQmtg6OSRu80OoN7R413x40BZrYJuFHSy8BqSYtGWf4BSY8BbcDtIyx6P0Gjsw14kCCh3gnsUVD2NzPb4uCmv3sTbW1t1NXVMTQ0xAknnMD111/PRRddhCQymQy9vb20traSyWQwM6ZMmcKmTZuGh3Pu6+tj6tSpdHZ2Ionm5ma6urqYPHkyQ0NDbN68mWnTptHR0UFdXR1NTU10dXXR1NRENpulv79/eH59fT2NjY10d3fT3NxMf38/AwMDw/MnTpxIQ0MDmUyGlpYW+vr6yGazSKK9vZ2Ghgbq6+vp6emhtbWVnp4eBgcHh9efNGkSEyZMiK3T0NAQHR0dZLPZ4fm16LR582Z6e3u3+JxqzSmTyYz63asFp0wmU1F9iqtTJpMZ82/EaE41yK3AbODjwDVF5n8LuMbMvplX9iBwQpFl8zkZ2BU4zMxywyX+Hfh6BbGdDJwHXAAcC/w8LG8DnjWzleHrLuAXeevtD5xnZhvD10+E01j5LfDlvNeHAN8kaATnl+VO0F4KfAj4nKSfh8u+38z6CjdcmDtPPvlkTj75ZI477jjuu+8+Zs+eXfXfr21R1zOZTOKcyvmcJLF+/fpEOY32OeWOq5LkVO08E4fjgaiP20ZDuXsXnPFF0hpgoZmtKChfS5DsDgTWmdl5YXekZ4GJZpaV9C6Cs9gfM7NfjbKfO4Ebzezbkr4H/BH4R6A9r6zJzD5VuO7Sm1bYx48NelUNDAywYMEC3vOe93DmmWdulft409vby+TJk6sdxlbjHvHCPeLFtvRYuXLlg/Pnz3/nNtl4CYrlCEnzgKVmtkvBsneF5T+UdAHB1c3rgCuAtwMGDBBcXfw7QQPvCDO7s8KYlgGbzeyUSuLOm3cIsAJ4I3ARsJuZvT+ctwfwOPCfwI3A/XkNXST9EDiI4Irk783sqTJjNuCtZvZ0QfmuwBpgOtABvEzQK+mvwD5h2QaChuZvw3X2JbgS/CDwmJl9odg+83NnPhdeeCHNzc186UtfKif0miMpvyWVkkbvNDqDe1fKaLnTu+PGjxeB7UvMa5fUC9xL0E3phjK2dzfB1U4Izur+Lpzyy+4utuIeOzQBwSALp512GjNnzqy5BihAJpOpdgiR4B7xwj3iRVI8CrghvGezQ1I5v/fDmNmNwKvAwoJZ2xHk/pfGEM8OY1wvxynAb8KrmT8Djpa0E4CZPQPMI+jieh1BvrtaUku47r8BPwU+Dzwe3os6YvfhPP6Y9z5+J9zfWmAtQQ7cB3jKzHqBVXlljcB9uY2Y2UMEt8zMBL5aame53Pnqq68Od2Pr7e1lxYoV7LnnnmWGXHsktA6OShq90+gM7h013giNHzsTnH0txnSCe2K+RJCsJ5axvd8CcyVtB+wYnj2+Bzg4LJtNkftB4bVnna1atYolS5Zwxx130NbWRltbG8uXLy+2SizJDQxR67hHvHCPeJEUjwKOM7Np4XQckKX47/5EgiudhZwHnEvQmMqxkeA+yxljiGf9GNdD0mTgIwQNSczsXoJG4Em5ZcxstZmdYGY7EjQEDw3jx8x6zewSM9uPoDF8HXC9pFInbfOZk/c+npFX/ttwH4cSnJwF+H1e2X1mtrlgW48Ba8ysp9TOcrnzpZde4rDDDmPvvfdm//3358gjj2TBggVlhFubJLQOjkoavdPoDO4dNX5PaIyQtD9BI/T3BN1xt8DMBoFFkj4IfI6gu9VI3EswiMNnCM7wYmadkl4My140s2dL7AuAuXPnDv9fi/T19Q33oa9l3CNeuEe8SIrHKKwFpktqyXVVDQfK2Q14rnBhM7td0tMEuSJX1iPpXuB4gjECKmEFcLGkZjOr9NT4BwkGMrpS0nfDsmkE94hukcfM7H5JvyQ4UVo4r1PSJcA5BF2MS524HY3fAv9C8N79OCz7HcEV2+cocYJ2NHL5cu+99+ahhx4aY2i1R0rq4Bak0TuNzuDeUeNXQmNAOOT9AmAZwX09j5Sx2jeAL0sa8VsRdi96ADiT1870QtDQPZMRkmwtNzzzyWaz1Q4hEtwjXrhHvEiKx0iEXUjvA74pqSUcXfYsgiukq0usdi6vH4CH8PWpks6StANA+BiSZaOEsAR4HviFpD0lTZC0g6SvSjomb7mJkhrzpnqCht2PgL0IBiFqA94NtEnaS9JcSZ/Odc+VtCfBCLirw9fnS9pfUkOY9/6d4L7NJ0eJeSR+C+xLMCLuqrDsEYKG7WFsZSM0baShDhYjjd5pdAb3jhpvhFaXmyR1EST1c4HLgVGf0xZyC0G3qk+XsezdwE4EDc8cvwvLSibZurqio/vXHEl5rpN7xAv3iBdJ8SiDfyb47X4aeAGYDxxTbJRWADNbBfyhoOwe4PBwekbSBuC/gRHvswi7ph5BMCrt7UBnuO3p5N07GW6nN2/6QRjnFWb2ct70IMFIvqcQNCjfDzwiqTss/xVwWW73BFcr2wnGTjgSeF/+4EWVYmZ/JRio6SUz6wjLhkKnKQS3rlRMUnJnpaSoDr6ONHqn0RncO2p8dFynJPfcc48dfPDB1Q5jq2lvb2f69OnVDmOrcY944R7xYlt6VGN0XKd2SUrurJSk/JZUShq90+gM7l0pPjquM2bCZ3LXPA0NDdUOIRLcI164R7xIiodT+yQld1ZKWutgGr3T6AzuHTXeCK1xJO0qqbvEtOtWbjuqMKtKfX0yxt9yj3jhHvEiKR7VJry/s1g++U21YyuGpENK5cAqxlStXVeVtNbBNHqn0RncO2q8EVrjmNlaM2spMa3dmm0PDQ1FFWZV6ekpOZJ+TeEe8cI94kVSPKpN+BiUYvmk3Gdyjitm9rtSObBaMSUld1ZKWutgGr3T6AzuHTXeCHVKkpTBFVpbW6sdQiS4R7xwj3iRFA+n9klK7qyUtNbBNHqn0RncO2q8EeqUJClnc5Ny5so94oV7xIukeDi1T1JyZ6WktQ6m0TuNzuDeUeONUKckSRk5eXBwsNohRIJ7xAv3iBdJ8XBqn6TkzkpJax1Mo3cancG9o8YboU5JktKlKCnPdXKPeOEe8SIpHk7tk5TcWSlprYNp9E6jM7h31Hgj1ClJUs74dHR0VDuESHCPeOEe8SIpHk7tk5TcWSlprYNp9E6jM7h31Hgj1ClJbpj5T33qU+y0007Mnj27yhGNjUmTJlU7hEhwj3jhHvEiKR5O7ZPLnc8//zyHHXYYM2fOZNasWSxevLjKkW1b0loH0+idRmdw76jxRqhTklwiPfXUU7n11lurHM3YmTAhGV9z94gX7hEvkuLh1D653FlfX8+iRYv4y1/+wurVq/n+97/P448/XuXoth1prYNp9E6jM7h35NvdJlstgqQ1knoLHiZ9kqR1RZa9S9LC8P8LJJmkj+TNrw/Lds8rO0DSckkdkjZI+oOkT5YR1xRJV0haG8b0dPh6el7cR4yw/pslDUm6ssi8D0h6WFKnpHZJK3MxS5om6UeSXpbUJemvkr4yWrzhupL0jKQtsln43vWFLu2SfilpRjjvv/Le+35JAyM9iDw3wt+hhx7K9ttvX05osaS3t7faIUSCe8QL94gXSfFwRibvWKIrzPf3SDpd0oRw/tWSLg7/3z08VsjluTWSzi5jH+dIWl5Q9lSJso8Wrv9sezcAM2bMYM6cOUDwiIOZM2fywgsvjFU99qS1DqbRO43O4N5RM95N+mMLHiT9YpnrbQAuklT0bn9JBwF3AHcDbwF2AD4LjPhwbUkNwEpgFvBeYApwMLAeOKDM2E4GNgIflTR8vVrSW4CfAF8EpgJvBq4EcmO3fxtoAWaG898P/K3MfR4K7ATsIWn/IvM/H76/bwn38R8AZnZ63nt/CfA/Iz2IPCmDKyTluU7uES/cI14kxcMpi2PNrBXYDfgG8BXgqhGWnxbmvQ8D50s6cpTt/xZ4d+6YQ9IbgYnAnIKyt4TLvo6saYsNrlmzhoceeogDDzxwNLeaJa11MI3eaXQG946a+m2y1ei5FZgNfBy4psimO0jlAAATd0lEQVT8bwHXmNk388oeBE4YZbsnA7sCh5lZd1j2d+DrFcR2MnAecAFwLPDzsLwNeNbMVoavu4Bf5K23P3CemW0MXz8RTuVwCvBrYHL4//3FFjKzDkk3AP9a5nZfxyuvvEJbWxt1dXX09vayceNGstksHR0dTJo0iQkTJtDb20trayuZTAYzY8qUKWzatInGxkYA+vr6mDp1Kp2dnUiiubmZrq4uJk+ezNDQEJs3b2batGl0dHRQV1dHU1MTXV1dNDU1kc1m6e/vH55fX19PY2Mj3d3dNDc309/fz8DAwPD8iRMn0tDQQCaToaWlhb6+PrLZLENDQ0yYMIGGhgbq6+vp6emhtbWVnp4eBgcHh9ePu9PAwACTJ08mm80Oz69Fp56eHnbcccctPqdac9qwYQP/8A//MOJ3rxacXnrpJd74xjeWXZ/i6vTiiy+y3Xbbjek3YjQnJ56Y2SbgRkkvA6slLRpl+QckPUaQn28fYdH7CRqdbQTHEocCdwJ7FJT9zcy2OJluPR3DuXNoaIgTTjiB66+/nosuughJZDKZqv9+bYu6/ve//503velNiXIq53MaGhqirq4uUU6jfU6546okOVU7z8TheCDq47ZRMbNxmYA1wBEFZfOAdUWWvQtYGP5/AbCU4ErhM2FiqAcM2B1oAgYJGpKVxrSMoPFaUdx58w4BNgPbAd8FbsybtwfQR3DF8zCgpWDdHwKPAZ8E3lpBzE1AJ3AMcDzQDjSUeO92AFYAvy6ynQuApSPta9WqVZbj2WeftVmzZlkt0t7eXu0QIsE94oV7xItt6bFixYoHbJxypU+j5sCiORlYS9AD6mrg4rBs9/BYoT58/S6gB/hgGfu5E/g/4f/fAz4F/H8FZT8qtu7Pbrp9+LvT399vRx11lC1atMiSTlJ+Syoljd5pdDZz70oZLXeOd3fcG8J7OHJX6MrGzG4EXgUWFszajqBb8UtjiGeHMa6X4xTgNxZczfwZcLSknQDM7BmCRvbOwHVAe3ivSku47r8BPwU+Dzwe3os6YvfhkA8RNHxvA24maJC/r2CZ70jaRNBAnR7uq2KS0h13ypQp1Q4hEtwjXrhHvEiKhzNmXgRKDV7QLqkXuJfgtphyjj/uJrjaCcEJ59+FU37Z3cVW3GPHoOuamXHaaacxc+ZMzjzzzHIcapq01sE0eqfRGdw7asa7EXqcmU0Lp+OALMGVzUImAgNFys8DzgUa88o2EtxnOWMM8awf43pImgx8hKAhiZndS3Am9qTcMma22sxOMLMdCRLWoWH8mFmvmV1iZvsRNIavA66XNNoIQKcA15lZ1sw2A78My/I5w8ymAnsTNNJ3GYtj7lL6iSeeyEEHHcSTTz7JLrvswlVXjXTrTfzYtGlTtUOIBPeIF+4RL5Li4YyZnQnGjyjGdILxEb5EcHK42HFHIb8F5kraDtjRzJ4C7gEODstmU+R+UHgtd65atYolS5Zwxx130NbWRltbG8uXLy+2SiJIax1Mo3cancG9o6ba94SuBaZLarHwnkxJIhhs4LnChc3sdklPA5/LK+uRdC9B19Q7K9z/CuBiSc1mlqlw3Q8SDGR0paTvhmXTCO4RvaJI7PdL+iVB4iqc1ynpEuAcggGMiiZSSbsAhwMHSDo+LG4CGiVNN7P2gu0+Eo4S+H1Jc8yCPkTlkhuS+dprr61ktdiR6z9f67hHvHCPeJEUD6dywgH6dgZ+DxQd+cfMBoFFkj5IcAyxRZ4u4F6CQQM/A6wKt9Ep6cWw7EUze7bYirncOXfuXCpMuzVNWutgGr3T6AzuHTVVfeCNma0F7gO+KaklHF32LIIrpKtLrHYu8OWCsi8Dp0o6S9IOAJL2kbRslBCWAM8Dv5C0p6QJknaQ9FVJx+QtN1FSY95UT3D18UfAXgQDFbQB7wbaJO0laa6kT+e650rak+C+1tXh6/Ml7S+pQVIj8O9AB/DkCPF+Avgr8Pa8fb4NWAecWGKdawhG0n3/KO+F4ziO49QMCh6xtoBgfIelZvZIGat9A/hymHdLYma9wAPAmQTdcHP8PiwrehXUcRzHKY84PHX1nwkaSU8DLwDzgWPMrK/Ywma2CvhDQdk9BFcIDweekbQB+G9gxH4vYXfWIwhGpb2dYMCfPxB03bkvb9HlQG/e9IMwzivM7OW86UGCkXxPIWhQvh94RFJ3WP4r4LLc7oEfE9y3+SJwJPA+e22U3mKcAlxZsM+Xgf9iyy65Ocd+4DvA+SO9F8XIPSe01unrK/pVqjncI164R7xIiodTFjdJ6iI4iXwucDnBIH/lcAvBbTyfLmPZuwmOT36fV/a7sKxkIzQpubNS0loH0+idRmdw76hRmrqKOJVx3333WRKeaTYwMMDEieXcAhRv3CNeuEe82JYeK1eufHD+/Pnv3CYbdxJHUnJnpSTlt6RS0uidRmdw70oZLXfG4UqoE1OS8ny8zs7OaocQCe4RL9wjXiTFw6l9kpI7KyWtdTCN3ml0BveOmsQ3QsP7O7uLTL+pdmzFkHRIiXhH6qbrjEAw1lXt4x7xwj3iRVI8nPFB0q6lcq2kXasdXy2S1jqYRu80OoN7R021R8fd5pjZJcAl1Y6jXMzsdwRDyVedpDwntLm5udohRIJ7xAv3iBdJ8XDGh3BgxG2Sa5OSOyslrXUwjd5pdAb3jprEXwl1xk5SuhR1dXVVO4RIcI944R7xIikeTu2TlNxZKWmtg2n0TqMzuHfUeCPUKUnuWWe1zuTJk6sdQiS4R7xwj3iRFA+n9klK7qyUtNbBNHqn0RncO2rS+UvplEVSRk5OynD57hEv3CNeJMXDqX2SkjsrJa11MI3eaXQG944ab4Q6JUlKIt28eXO1Q4gE94gX7hEvkuLh1D5JyZ2VktY6mEbvNDqDe0eNN0KdkiRlcIVp06ZVO4RIcI944R7xIikeTu2TlNxZKWmtg2n0TqMzuHfUeCPUKUlSBlfo6OiodgiR4B7xwj3iRVI8nNonKbmzUtJaB9PonUZncO+o8UaoU5L169dXO4RIWLJkSbVDiAT3iBfuES+S4uHUPknJnZWS1jqYRu80OoN7R403Qp2SbNiwodohRMLSpUurHUIkuEe8cI94kRQPp/ZJSu6slLTWwTR6p9EZ3DtqvBHqlCQpgyskpWuUe8QL94gXSfFwap+k5M5KSWsdTKN3Gp3BvaNGaf2xdEZn+fLlXZMmTXqy2nFsLRs2bJi+/fbbt1c7jq3FPeKFe8SLbeyx2/z583fcRtt2EkZScmelJOW3pFLS6J1GZ3DvMaw6Yu70RqjjOI7jOI7jOI4zbnh3XMdxHMdxHMdxHGfc8Eao4ziO4ziO4ziOM254I9RxHMdxHMdxHMcZN7wR6jiO4ziO4ziO44wb3gh1tkDS9pJ+JSkj6TlJJ1U7pmJI+rykByRtlnR1wbz5kp6Q1CPpTkm75c2bJOlHkjolvSzpzHEP/vWxTpJ0Vfhed0l6SNLRefNryWWppJfCeP4qaWHevJrxCGN6q6Q+SUvzyk4KP6eMpBskbZ83L3b1RtJdoUN3OD2ZN6/WXD4q6S9hTH+TdEhYXlPfKye5xLHebC1Jyk9jJQm5oFLS+HsraXdJyyVtDOP/nqT6cF6bpAdD7wclteWtJ0nflLQ+nC6TpOqZlEbb6Lh5pHVHxMx88ul1E3At8D9ACzAX2ATMqnZcReL8EHAc8J/A1Xnl08OYPwI0At8CVufNvxT4HbAdMBN4GXhvFT2agQuA3QlODC0AusLXteYyC5gU/r9nGM9+teYRxnRbGNPSPLcu4NCwbvwMWJa3fOzqDXAXsLDE51QzLsCRwHPAu8I6snM41dz3yqfkTnGrNxE5JSY/bcV7UPO5oELfVP7eAsuBq0O3NwKPAGcADeH78X+ASWHZc0BDuN6/AE8Cu4Tv0+PA6dX2KeEY+XHzaOuOGE+13xCf4jWFCacfeFte2RLgG9WObYSYLy6oTJ8B7ilw6gX2DF+/AByVN//r+QkkDhPwZ+D4WnYB3g68BJxQax7AR4HrCA6+cgcelwA/y1vmH8O60hrXekPpRmhNuQD3AKcVKa+p75VPyZ3iWG+2oWvN56cKXBORCyp0TuXvLfAX4Ji8198C/n/gqNBNefPW8loj7B7gM3nzTqPMRlgVXSM7bh5t3ZEm747rFPI2YNDM/ppX9ieCM3+1wiyCmAEwswzwN2CWpO2AN+XPJ2Z+kt5A8Dk8Rg26SLpSUg/wBEEjdDk15CFpCnAR8MWCWYUOfyM82CDe9eZSSe2SVkmaF5bVjIukOuCdwI6Snpa0LuwmNZka+l45iSdW9WZbUev5qRISmAtGJeW/t4uBj0pqkrQzcDRwK4HDny1sYYX8mdfcXve+UHvesHWfbcl1R9upN0KdQloILqvns4ngDF+tMJJDS97rwnlVR9JE4KfANWb2BDXoYmafC2M4BPglsJna8vg6cJWZPV9QPppDHOvNV4A9CLoI/Tdwk6R/pLZc3gBMBD5M8J1qA/YFzqO2vldOsolbvYmcJOSnCklSLiiXNP/e3k3QcOoE1gEPADcw+mdaOH8T0BLX+0JLsDWf7Zi/894IdQrpBqYUlE0huP+hVhjJoTvvdeG8qiJpAkG3nX7g82FxTbqY2aCZ/Z7gHonPUiMe4WADRwDfLjJ7NIfY1Rszu8/Musxss5ldA6wCjqG2XHrDv981s5fMrB24nPI8IAbfKycVxK3eREqS8lM5JC0XVEAqf2/D7/f/Epw4bya4z3E74JuM/pkWzp8CdBdcOY07W/PZjvk7741Qp5C/AvWS3ppXtg9B15ta4TGCmAGQ1Exwz8ZjZraRoIvoPnnLV90vPGN2FcFZyOPNbCCcVXMuBdQTxktteMwjGHBjraSXgS8Bx0v6I1s67EEwSMFfqZ16Y4CoIZfw+7GOIPZCauV75SSfWNWbKElwfhqJeSQ7FxQlxb+32wP/AHwvPGm7HvgxQeP7MWDvgiube/Oa2+veF2rLO8fWfLYl1x11r9W+Odan+E3AMoLR3ZqBdxPTkd0IGjiNBKN2LQn/rwd2DGM+Piz7Jq8f5esbBN0utiMYwfUlqj8S638Bq4GWgvKacQF2IhjEoQWoA94DZIAP1IoH0EQwKl5u+g/g52H8uW46h4R1YymvHxExVvUGmBZ+Brl68bHw83h7DbpcBNwffse2Ixil7+u18r3yKR1T3OpNhF41n5/G4JyYXDAG91T+3gLPAGeH+XIa8CuC7ue50XH/neBkw+d5/ei4pxMMarQzwb2TjxHf0XEjP24ebd0R46n2G+JT/CaCM0I3EBywrgVOqnZMJeK8gOBsXf50QTjvCIKBcXoJRgjdPW+9ScCPwiTyCnBmlT12C2PvI+jWkJs+Vksu4Q/R3UBHGM8jwKfz5teER5Hv2NK81yeFdSID/BrYPm9erOpN+HncT9AlpoPgIPLIGnWZCFwZerwMfAdorNXvlU/JnOJWbyJySkR+iuB9qNlcMAbXVP7eEtz/ehewEWgHrgd2CuftCzwYev8R2DdvPQGXARvC6TLyRtKN08Q2Om4ead2RJoUrO47jOI7jOI7jOM42x+8JdRzHcRzHcRzHccYNb4Q6juM4juM4juM444Y3Qh3HcRzHcRzHcZxxwxuhjuM4juM4juM4zrjhjVDHcRzHcRzHcRxn3PBGqOM4juM4juM4jjNueCPUcRzH+X/t3E2IVWUcx/HvT2chNFBIEo3vYYtoH0GBuhKiQTeiUAshWrYIFTJatJAIglbSsoUK6kJxYS5UVDQMqo170cYJRaUYckTBl3+LcwYu4l3N3DPjne8HHricc57nnGdz//ye8yJJktQZQ6ikOZPkryQPk0z3tLH5vi5JkhYqa6cWI0OopLk2XlWjPe1W1xeQZKTrc0qSNAvWTi0qhlBJnUvyepJTSaaS/JvkcpIl7b7VSU4kuZfknyQH2u1LknyTZCLJ3SQHk7za7luXpJJ8luQmcL7d/n6SK+15ribZNF9zliRpNqydGiaGUEnzYTfwN7ACeAP4GqgkS4FTwASwDlgJHG377GrbZuAtYBQ48Ny4G4F3gC1JVgK/APuB5cAe4HiSFQOakyRJg2Tt1NAwhEqaayfb1dOpJCf7HPMYeBNYW1WPq+pyVRXwHjAG7K2qB1X1qKp+bft8AvxYVderahrYB+x87vGhb9t+D4FPgdNVdbqqnlXVWeBP4KMBzFmSpNmwdmpRMYRKmmvbquq1tm3rc8wPwDXgTJLrSb5qt68GJqrqyQv6jNGs8s6YAEZoVoNnTPb8Xgts7ynqU8CHNAVckqSFxNqpRcUXkCV1rqru0zxWtDvJu8CFJH/QFMI1SUZeUExv0RTHGWuAJ8AdYNXM0D37J4FDVfX5IOYgSVKXrJ0aJt4JldS5JB8n2ZAkwH/A07b9DtwGvk/ySpJlST5oux0BvkyyPsko8B1wrM/KL8BhYDzJliRL27E2JVnV53hJkhYsa6eGiSFU0nx4GzgHTAO/AT9V1cWqegqMAxuAmzQfYNjR9vkZOARcAm4Aj4Av+p2gqiaBrTQfbrhHs7q7F//3JEkvJ2unhkaa95klSZIkSRo8VzUkSZIkSZ0xhEqSJEmSOmMIlSRJkiR1xhAqSZIkSeqMIVSSJEmS1BlDqCRJkiSpM4ZQSZIkSVJnDKGSJEmSpM78D7o7UxSejP/EAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 936x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax= plt.subplots(1,2, figsize=(13,6))\n",
    "xgboost.plot_importance(xgb_improved, ax=ax[0]), ax[0].grid(alpha=.5), ax[0].set_title('New Feature Importance')\n",
    "ax[0].set_facecolor('white')\n",
    "ax[0].tick_params(axis='both', labelsize=12)\n",
    "\n",
    "# Note: I went back to set the original xgboost model to n=1000 estimators as well to ensure comparability\n",
    "xgboost.plot_importance(xgb_orig, ax=ax[1]), ax[1].grid(alpha=.5), ax[1].set_title('Old Feature Importance')\n",
    "ax[1].set_facecolor('white')\n",
    "ax[1].tick_params(axis='both', labelsize=12)\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Old balanced accuracy: 89.28%\n",
      "Mean (old) cross-validated balanced accuracy:  89.0%, SD: 4.993%\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "   very good       0.88      0.93      0.90        56\n",
      "satisfactory       0.89      0.89      0.89        18\n",
      "        fair       0.94      0.93      0.94       146\n",
      "        poor       0.96      0.96      0.96       142\n",
      "   very poor       1.00      0.75      0.86         4\n",
      "\n",
      "    accuracy                           0.94       366\n",
      "   macro avg       0.94      0.89      0.91       366\n",
      "weighted avg       0.94      0.94      0.94       366\n",
      "\n",
      "Cohen's Kappa: 0.910\n",
      "\n",
      "\n",
      "New balanced accuracy: 89.14%\n",
      "Mean (new) cross-validated balanced accuracy:  88.0%, SD: 5.956%\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "   very good       0.87      0.93      0.90        56\n",
      "satisfactory       0.89      0.89      0.89        18\n",
      "        fair       0.94      0.92      0.93       146\n",
      "        poor       0.96      0.96      0.96       142\n",
      "   very poor       1.00      0.75      0.86         4\n",
      "\n",
      "    accuracy                           0.94       366\n",
      "   macro avg       0.93      0.89      0.91       366\n",
      "weighted avg       0.94      0.94      0.94       366\n",
      "\n",
      "Cohen's Kappa: 0.906\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Old model\n",
    "print(f\"Old balanced accuracy: {balanced_accuracy_score(y_test, y_pred[0])*100:.2f}%\")\n",
    "print(\"Mean (old) cross-validated balanced accuracy: \",\n",
    "      f\"{100*cv_accuracies[0].mean().round(2)}%, SD: {round(100*cv_accuracies[0].std(),3)}%\\n\")\n",
    "\n",
    "print(classification_report(y_test, y_pred[0], target_names=y_labs, digits=2))\n",
    "print(f\"Cohen's Kappa: {cohen_kappa_score(y_test,y_pred[0]):.3f}\\n\")\n",
    "\n",
    "# New model\n",
    "print(f\"\\nNew balanced accuracy: {balanced_accuracy_score(y_test, y_pred[1])*100:.2f}%\")\n",
    "print(\"Mean (new) cross-validated balanced accuracy: \",\n",
    "      f\"{100*cv_accuracies[1].mean().round(2)}%, SD: {round(100*cv_accuracies[1].std(),3)}%\\n\")\n",
    "\n",
    "print(classification_report(y_test, y_pred[1], target_names=y_labs, digits=2))\n",
    "print(f\"Cohen's Kappa: {cohen_kappa_score(y_test,y_pred[1]):.3f}\\n\")\n",
    "\n",
    "\n",
    "# ------ plotting\n",
    "fig, axs=plt.subplots(1,1,figsize=(8,3))\n",
    "sns.set(rc={'axes.facecolor':'white'})\n",
    "sns.kdeplot(cv_accuracies[0], ax=axs, label='Old model')\n",
    "sns.kdeplot(cv_accuracies[1], ax=axs, label='New model')\n",
    "\n",
    "plt.ylabel('Density', fontsize=12), plt.xlabel('Balanced Accuracy')\n",
    "plt.title(\"Cross-validated balanced accuracy,\\n from 10 stratified k-folds\", fontsize=11)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Suggestions on the Most Important Predictors of Road Pavement Quality\n",
    "Let's now see if we could reduce our model into a more simplified version and see which features might be the most important predictors of road pavement quality."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold: 0.000, balanced accuracy: 89.77%, Features left: 19\n",
      "Threshold: 0.000, balanced accuracy: 89.77%, Features left: 19\n",
      "Threshold: 0.000, balanced accuracy: 89.77%, Features left: 17\n",
      "Threshold: 0.001, balanced accuracy: 89.77%, Features left: 16\n",
      "Threshold: 0.002, balanced accuracy: 89.77%, Features left: 15\n",
      "Threshold: 0.002, balanced accuracy: 89.77%, Features left: 14\n",
      "Threshold: 0.008, balanced accuracy: 89.77%, Features left: 13\n",
      "Threshold: 0.008, balanced accuracy: 89.77%, Features left: 12\n",
      "Threshold: 0.010, balanced accuracy: 89.77%, Features left: 11\n",
      "Threshold: 0.010, balanced accuracy: 89.77%, Features left: 10\n",
      "Threshold: 0.013, balanced accuracy: 89.77%, Features left: 9\n",
      "Threshold: 0.014, balanced accuracy: 89.91%, Features left: 8\n",
      "Threshold: 0.018, balanced accuracy: 89.91%, Features left: 7\n",
      "Threshold: 0.018, balanced accuracy: 88.66%, Features left: 6\n",
      "Threshold: 0.021, balanced accuracy: 88.66%, Features left: 5\n",
      "Threshold: 0.031, balanced accuracy: 88.80%, Features left: 4\n",
      "Threshold: 0.035, balanced accuracy: 88.80%, Features left: 3\n",
      "Threshold: 0.197, balanced accuracy: 89.77%, Features left: 2\n",
      "Threshold: 0.612, balanced accuracy: 64.10%, Features left: 1\n"
     ]
    }
   ],
   "source": [
    "thresholds = np.sort(xgb_improved.feature_importances_) # sorted from least to most important.\n",
    "select_accuracy = [np.nan]*len(thresholds)\n",
    "\n",
    "for i, thresh in zip(range(len(thresholds)), thresholds):\n",
    "    # select features using threshold\n",
    "    selection = SelectFromModel(xgb_improved, threshold=thresh, prefit=True)\n",
    "    select_x_train = selection.transform(x_train)\n",
    "    \n",
    "    # Returns boolean mask/array of the features (as names) that were removed out from the training set by SelectFromModel,\n",
    "    # based on our threshold--recall that thresh passing over the features in order of least to greatest importance.\n",
    "    feature_idx = selection.get_support() \n",
    "    # Find/select the colnames from the training set that were NOT removed according to thresh\n",
    "    feature_name = x_train.columns[feature_idx] \n",
    "    \n",
    "    # Keep track of the features excluded from the training set\n",
    "    reduced_x_train = pd.DataFrame(select_x_train, columns=feature_name)\n",
    "    # Update eval set to allow for early stopping rounds to work\n",
    "    select_eval_set = [(reduced_x_train, y_train), (x_test[x_test.columns & reduced_x_train.columns], y_test)]\n",
    "    selection_model = XGBClassifier(**params_improved, use_label_encoder=False)\n",
    "    selection_model.fit(reduced_x_train, y_train, early_stopping_rounds=stop_after, eval_set=select_eval_set, verbose=False)\n",
    "    \n",
    "    # Evaluate model\n",
    "    select_x_test = selection.transform(x_test)\n",
    "    select_pred = selection_model.predict(select_x_test)\n",
    "    select_accuracy[i] = balanced_accuracy_score(y_test, select_pred)\n",
    "    print(f\"Threshold: {thresh:.3f}, balanced accuracy: {select_accuracy[i]*100.0:.2f}%, Features left: {select_x_train.shape[1]}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's the extract feature names used during our testing process and sort them in ascending order of their feature importance. We'll use this to plot the changes in the balanced accuracy of our model from the progressive removal of features (in order of least to greatest importance)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>DIR_E</th>\n",
       "      <th>FUNC_CLASS_FWY</th>\n",
       "      <th>FUNC_CLASS_ART</th>\n",
       "      <th>DIR_B</th>\n",
       "      <th>DIR_N</th>\n",
       "      <th>FUNC_CLASS_COL</th>\n",
       "      <th>FROM_Distance</th>\n",
       "      <th>Offset</th>\n",
       "      <th>Pave_Type_ST</th>\n",
       "      <th>FUNC_CLASS_LOC</th>\n",
       "      <th>LHRS</th>\n",
       "      <th>DIR_S</th>\n",
       "      <th>TO_Distance</th>\n",
       "      <th>Pave_Type_PC</th>\n",
       "      <th>Pave_Type_AC</th>\n",
       "      <th>Pave_Type_CO</th>\n",
       "      <th>DIR_W</th>\n",
       "      <th>IRI</th>\n",
       "      <th>DMI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>0.001371</td>\n",
       "      <td>0.001523</td>\n",
       "      <td>0.002328</td>\n",
       "      <td>0.007663</td>\n",
       "      <td>0.008348</td>\n",
       "      <td>0.009651</td>\n",
       "      <td>0.009776</td>\n",
       "      <td>0.012997</td>\n",
       "      <td>0.013738</td>\n",
       "      <td>0.017868</td>\n",
       "      <td>0.018492</td>\n",
       "      <td>0.0211</td>\n",
       "      <td>0.030854</td>\n",
       "      <td>0.035115</td>\n",
       "      <td>0.197246</td>\n",
       "      <td>0.611869</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   DIR_E  FUNC_CLASS_FWY  FUNC_CLASS_ART     DIR_B     DIR_N  FUNC_CLASS_COL  \\\n",
       "0    0.0             0.0        0.000061  0.001371  0.001523        0.002328   \n",
       "\n",
       "   FROM_Distance   Offset   Pave_Type_ST  FUNC_CLASS_LOC      LHRS     DIR_S  \\\n",
       "0       0.007663  0.008348      0.009651        0.009776  0.012997  0.013738   \n",
       "\n",
       "   TO_Distance  Pave_Type_PC  Pave_Type_AC  Pave_Type_CO     DIR_W       IRI  \\\n",
       "0     0.017868      0.018492        0.0211      0.030854  0.035115  0.197246   \n",
       "\n",
       "        DMI  \n",
       "0  0.611869  "
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_dict = {}\n",
    "for col, imp in zip(x_train.columns,xgb_improved.feature_importances_):\n",
    "    feature_dict.update({col: imp})\n",
    "feature_imp_named = pd.DataFrame(data=feature_dict, index=[0]).sort_values(by=0, axis=1)\n",
    "feature_imp_named"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('default')\n",
    "fig, ax= plt.subplots(1,2, figsize=(10,5))\n",
    "xgboost.plot_importance(xgb_improved, ax=ax[0]), ax[0].grid(alpha=.5), ax[0].set_title('New Feature Importance')\n",
    "ax[0].set_yticklabels(labels=feature_imp_named.columns, fontsize=9)\n",
    "\n",
    "ax[1].plot(range(len(thresholds)), select_accuracy, marker='o', markerfacecolor='lightgreen', color='green')\n",
    "ax[1].set_xticks(range(len(thresholds)))\n",
    "ax[1].set_xticklabels(feature_imp_named.columns, rotation=45, ha='right', fontsize=9), \n",
    "\n",
    "ax[1].set_ylabel('balanced accuracy')\n",
    "ax[1].set_title('Balanced accuracy with select \\nfeatures removed, in order of their ascending importance', fontsize=9)\n",
    "ax[1].axvline(x=12, ymin=0, ymax=1, color='red', ls='--')\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusions:\n",
    "The results of the feature importance plot, as well as from the balanced accuracy plot, suggest that `DMI` and `IRI` may be the two most important features which predict for road pavement quality in Ontario. Factoring in the 'to' (and possibly, from) distances of a road may also result in slightly more accurate predictions of said road's pavement quality.\n",
    "\n",
    "**Note:** `DMI` and `IRI` were both likely strongly associated with our outcome *because* they are already constructs of our outcome--PCI--as mentioned in the introduction. Thus, our modeling was likely done on data that was already somewhat redundant. Our modeling and results could have been more meaningful--and less redundant--by *not* using features that are already constructs of our outcome. "
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}