Train.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "# Set random seed\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 836 entries, CFDB9498-809D-4232-8281-9E7CA354D59B to 59A22453-FB78-44C8-BCF1-884D88C17FB1\n",
      "Data columns (total 9 columns):\n",
      "clouds         836 non-null int64\n",
      "date           836 non-null object\n",
      "humidity       836 non-null int64\n",
      "temp           836 non-null float64\n",
      "type           836 non-null int64\n",
      "visibility     836 non-null int64\n",
      "weatherCond    836 non-null object\n",
      "windSpeed      836 non-null float64\n",
      "zipcode        836 non-null object\n",
      "dtypes: float64(2), int64(4), object(3)\n",
      "memory usage: 65.3+ KB\n"
     ]
    }
   ],
   "source": [
    "data = pd.read_csv('data.csv', index_col=0)\n",
    "data = data[data['weatherCond'].notnull()]\n",
    "\n",
    "import sklearn.utils\n",
    "data = sklearn.utils.shuffle(data)\n",
    "\n",
    "labels = np.ravel(data[['type']],order='C')\n",
    "data.drop(['userId', 'lat', 'lon', 'location', 'device', 'zip', 'pressure', 'windDeg'],axis=1, inplace=True)\n",
    "data.info()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#force df to have same # of hot / iced rows\n",
    "hot = data.loc[data.type == 1]\n",
    "iced = data.loc[data.type == 0]\n",
    "\n",
    "#remove all hot\n",
    "data.drop(data[data.type == 1].index, inplace=True)\n",
    "\n",
    "num_delete = hot.shape[0] - iced.shape[0]\n",
    "indexes_to_keep = set(range(hot.shape[0])) - set(range(num_delete))\n",
    "hot_sliced = hot.take(list(indexes_to_keep))\n",
    "\n",
    "data = pd.concat([data, hot_sliced])\n",
    "data['type'].value_counts()\n",
    "labels = np.ravel(data[['type']],order='C')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iced (312, 9)\n",
      "hot (312, 9)\n"
     ]
    }
   ],
   "source": [
    "print(\"iced\", data[data['type']==0].shape)\n",
    "print(\"hot\",data[data['type']==1].shape)\n",
    "# data.drop(['type'],axis=1, inplace=True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#round data\n",
    "def round(df):\n",
    "    return df.apply(lambda x: np.rint(x))\n",
    "\n",
    "data['temp'] = round(data['temp'])\n",
    "# data['windDeg'] = round(data['windDeg'])\n",
    "data['windSpeed'] = round(data['windSpeed'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def to_bins(col, bins, lables):\n",
    "    data[col] = pd.cut(data[col], bins=bins, labels=lables)\n",
    "    return data[col]\n",
    "\n",
    "def to_categorical(data, cat_vars):\n",
    "    for var in cat_vars:\n",
    "        cat_list='var'+'_'+var\n",
    "        cat_list = pd.get_dummies(data[var], prefix=var)\n",
    "        data1=data.join(cat_list)\n",
    "        data=data1\n",
    "    data = data.drop(cat_vars, axis=1)\n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "#convert visibilty from meters to miles\n",
    "def miles(meter):\n",
    "    return round(meter/1609.344)\n",
    "\n",
    "data['visibility'] = miles(data['visibility'])\n",
    "# bins = [0, 1, 5, 20]\n",
    "# vis_lables = [0,1,2]\n",
    "\n",
    "# data['visibility'] = to_bins(\"visibility\", bins=bins, lables=vis_lables)\n",
    "# # data = to_categorical(data, cat_vars=['visibility'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# clouds binning\n",
    "# cloud_bins = [0, 25, 50, 100]\n",
    "# cloud_labels = [0, 1, 2]\n",
    "\n",
    "# data['clouds'] = to_bins(\"clouds\", bins=cloud_bins, lables=cloud_labels)\n",
    "# data = to_categorical(data, cat_vars=['clouds'])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# humidity binning\n",
    "# data['humidity'].value_counts()\n",
    "# humidity_bins = [0, 25, 50, 100]\n",
    "# humidity_labels = [0, 1, 2]\n",
    "\n",
    "# data['humidity'] = to_bins(\"humidity\", bins=humidity_bins, lables=humidity_labels)\n",
    "# data = to_categorical(data, cat_vars=['humidity'])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [],
   "source": [
    "#feature engineering with date\n",
    "\n",
    "#make cols based on:\n",
    "#day_of_week\n",
    "#is_weekday\n",
    "#is_weekend\n",
    "#morning\n",
    "#afternoon\n",
    "#evening\n",
    "\n",
    "from datetime import datetime\n",
    "from dateutil import parser\n",
    "\n",
    "day_of_week = lambda x: parser.parse(x).weekday()\n",
    "is_weekend = lambda x: parser.parse(x).weekday() > 5\n",
    "month = lambda x: parser.parse(x).month\n",
    "\n",
    "\n",
    "data['day_of_week'] =  data['date'].map(day_of_week)\n",
    "data['is_weekend'] =  data['date'].map(is_weekend)\n",
    "data = to_categorical(data, cat_vars=['is_weekend'])\n",
    "\n",
    "\n",
    "data['month'] =  data['date'].map(month)\n",
    "\n",
    "\n",
    "def season_of_date(date_str):\n",
    "    month =  parser.parse(date_str).month\n",
    "    return (\n",
    "        \"spring\" if month >=2 and month <=5\n",
    "        else\n",
    "        \"summer\" if month >=6 and month <=8\n",
    "        else\n",
    "        \"fall\" if month >=9 and month <=11\n",
    "        else\n",
    "        \"winter\"\n",
    "    )\n",
    "\n",
    "# Assuming df has a date column of type `datetime`\n",
    "data['season'] = data['date'].map(season_of_date)\n",
    "data = to_categorical(data, cat_vars=['season'])\n",
    "\n",
    "\n",
    "def get_part_of_day(hour):\n",
    "    return (\n",
    "        \"morning\" if 5 <= hour <= 11\n",
    "        else\n",
    "        \"afternoon\" if 12 <= hour <= 17\n",
    "        else\n",
    "        \"evening\" if 18 <= hour <= 22\n",
    "        else\n",
    "        \"night\"\n",
    "    )\n",
    "\n",
    "part_of_day = lambda x: get_part_of_day(parser.parse(x).hour)\n",
    "data['part_of_day'] =  data['date'].map(part_of_day)\n",
    "data = to_categorical(data, cat_vars=['part_of_day'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "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>clouds</th>\n",
       "      <th>date</th>\n",
       "      <th>humidity</th>\n",
       "      <th>temp</th>\n",
       "      <th>type</th>\n",
       "      <th>visibility</th>\n",
       "      <th>weatherCond</th>\n",
       "      <th>windSpeed</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>...</th>\n",
       "      <th>is_weekend_True</th>\n",
       "      <th>month</th>\n",
       "      <th>season_fall</th>\n",
       "      <th>season_spring</th>\n",
       "      <th>season_summer</th>\n",
       "      <th>season_winter</th>\n",
       "      <th>part_of_day_afternoon</th>\n",
       "      <th>part_of_day_evening</th>\n",
       "      <th>part_of_day_morning</th>\n",
       "      <th>part_of_day_night</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>CFDB9498-809D-4232-8281-9E7CA354D59B</th>\n",
       "      <td>40</td>\n",
       "      <td>2018-12-01T17:27:45.102Z</td>\n",
       "      <td>55</td>\n",
       "      <td>42.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Clouds</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10923</td>\n",
       "      <td>5</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2A429A89-2491-4082-9622-D06E370D091A</th>\n",
       "      <td>20</td>\n",
       "      <td>2018-06-26T21:35:35.489Z</td>\n",
       "      <td>46</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Clouds</td>\n",
       "      <td>11.0</td>\n",
       "      <td>10923</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4F20435C-DF07-4D4E-BF43-BE54C4A967F6</th>\n",
       "      <td>40</td>\n",
       "      <td>2018-07-29T23:47:46.745Z</td>\n",
       "      <td>42</td>\n",
       "      <td>79.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Clouds</td>\n",
       "      <td>2.0</td>\n",
       "      <td>10901</td>\n",
       "      <td>6</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E</th>\n",
       "      <td>40</td>\n",
       "      <td>2018-07-14T20:26:08.310Z</td>\n",
       "      <td>40</td>\n",
       "      <td>88.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Clouds</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10954</td>\n",
       "      <td>5</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B9F9226B-3048-4954-830C-52D0524B7415</th>\n",
       "      <td>1</td>\n",
       "      <td>2018-07-02T10:49:06.384Z</td>\n",
       "      <td>94</td>\n",
       "      <td>72.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Mist</td>\n",
       "      <td>2.0</td>\n",
       "      <td>10977</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                      clouds                      date  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B      40  2018-12-01T17:27:45.102Z   \n",
       "2A429A89-2491-4082-9622-D06E370D091A      20  2018-06-26T21:35:35.489Z   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6      40  2018-07-29T23:47:46.745Z   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E      40  2018-07-14T20:26:08.310Z   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415       1  2018-07-02T10:49:06.384Z   \n",
       "\n",
       "                                      humidity  temp  type  visibility  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B        55  42.0     0        10.0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A        46  75.0     0        10.0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6        42  79.0     0        10.0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E        40  88.0     0        10.0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415        94  72.0     0         3.0   \n",
       "\n",
       "                                     weatherCond  windSpeed zipcode  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B      Clouds        4.0   10923   \n",
       "2A429A89-2491-4082-9622-D06E370D091A      Clouds       11.0   10923   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6      Clouds        2.0   10901   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E      Clouds        4.0   10954   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415        Mist        2.0   10977   \n",
       "\n",
       "                                      day_of_week        ...          \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B            5        ...           \n",
       "2A429A89-2491-4082-9622-D06E370D091A            1        ...           \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6            6        ...           \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E            5        ...           \n",
       "B9F9226B-3048-4954-830C-52D0524B7415            0        ...           \n",
       "\n",
       "                                      is_weekend_True  month  season_fall  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                0     12            0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                0      6            0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                1      7            0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                0      7            0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                0      7            0   \n",
       "\n",
       "                                      season_spring  season_summer  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B              0              0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A              0              1   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6              0              1   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E              0              1   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415              0              1   \n",
       "\n",
       "                                      season_winter  part_of_day_afternoon  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B              1                      1   \n",
       "2A429A89-2491-4082-9622-D06E370D091A              0                      0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6              0                      0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E              0                      0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415              0                      0   \n",
       "\n",
       "                                      part_of_day_evening  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                    0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                    1   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                    0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                    1   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                    0   \n",
       "\n",
       "                                      part_of_day_morning  part_of_day_night  \n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                    0                  0  \n",
       "2A429A89-2491-4082-9622-D06E370D091A                    0                  0  \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                    0                  1  \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                    0                  0  \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                    1                  0  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 252,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Clouds          241\n",
       "Clear           169\n",
       "Rain            161\n",
       "Fog              24\n",
       "Haze             14\n",
       "Snow              9\n",
       "Thunderstorm      4\n",
       "Smoke             2\n",
       "Name: weatherCond, dtype: int64"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['weatherCond']\n",
    "for index,rows in data.iterrows():\n",
    "    if (\"cloud\" in data.loc[index, 'weatherCond'].lower() or \"overcast\" in data.loc[index, 'weatherCond'].lower()):\n",
    "        data.loc[index, 'weatherCond'] = 'Clouds'\n",
    "        \n",
    "    if (\"snow\" in data.loc[index, 'weatherCond'].lower()):\n",
    "        data.loc[index, 'weatherCond'] = 'Snow'\n",
    "        \n",
    "    if (\"rain\" in data.loc[index, 'weatherCond'].lower() or \"drizzle\" in data.loc[index, 'weatherCond'].lower() or \"mist\" in data.loc[index, 'weatherCond'].lower()):\n",
    "        data.loc[index, 'weatherCond'] = 'Rain'\n",
    "        \n",
    "    if (\"none\" in data.loc[index, 'weatherCond'].lower()):\n",
    "        data.loc[index, 'weatherCond'] = 'Clear'\n",
    "    \n",
    "\n",
    "data['weatherCond'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "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>clouds</th>\n",
       "      <th>date</th>\n",
       "      <th>humidity</th>\n",
       "      <th>temp</th>\n",
       "      <th>type</th>\n",
       "      <th>visibility</th>\n",
       "      <th>windSpeed</th>\n",
       "      <th>zipcode</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>is_weekend_False</th>\n",
       "      <th>...</th>\n",
       "      <th>part_of_day_morning</th>\n",
       "      <th>part_of_day_night</th>\n",
       "      <th>weatherCond_Clear</th>\n",
       "      <th>weatherCond_Clouds</th>\n",
       "      <th>weatherCond_Fog</th>\n",
       "      <th>weatherCond_Haze</th>\n",
       "      <th>weatherCond_Rain</th>\n",
       "      <th>weatherCond_Smoke</th>\n",
       "      <th>weatherCond_Snow</th>\n",
       "      <th>weatherCond_Thunderstorm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>CFDB9498-809D-4232-8281-9E7CA354D59B</th>\n",
       "      <td>40</td>\n",
       "      <td>2018-12-01T17:27:45.102Z</td>\n",
       "      <td>55</td>\n",
       "      <td>42.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10923</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2A429A89-2491-4082-9622-D06E370D091A</th>\n",
       "      <td>20</td>\n",
       "      <td>2018-06-26T21:35:35.489Z</td>\n",
       "      <td>46</td>\n",
       "      <td>75.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>10923</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4F20435C-DF07-4D4E-BF43-BE54C4A967F6</th>\n",
       "      <td>40</td>\n",
       "      <td>2018-07-29T23:47:46.745Z</td>\n",
       "      <td>42</td>\n",
       "      <td>79.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>10901</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E</th>\n",
       "      <td>40</td>\n",
       "      <td>2018-07-14T20:26:08.310Z</td>\n",
       "      <td>40</td>\n",
       "      <td>88.0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10954</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B9F9226B-3048-4954-830C-52D0524B7415</th>\n",
       "      <td>1</td>\n",
       "      <td>2018-07-02T10:49:06.384Z</td>\n",
       "      <td>94</td>\n",
       "      <td>72.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>10977</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                      clouds                      date  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B      40  2018-12-01T17:27:45.102Z   \n",
       "2A429A89-2491-4082-9622-D06E370D091A      20  2018-06-26T21:35:35.489Z   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6      40  2018-07-29T23:47:46.745Z   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E      40  2018-07-14T20:26:08.310Z   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415       1  2018-07-02T10:49:06.384Z   \n",
       "\n",
       "                                      humidity  temp  type  visibility  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B        55  42.0     0        10.0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A        46  75.0     0        10.0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6        42  79.0     0        10.0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E        40  88.0     0        10.0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415        94  72.0     0         3.0   \n",
       "\n",
       "                                      windSpeed zipcode  day_of_week  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B        4.0   10923            5   \n",
       "2A429A89-2491-4082-9622-D06E370D091A       11.0   10923            1   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6        2.0   10901            6   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E        4.0   10954            5   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415        2.0   10977            0   \n",
       "\n",
       "                                      is_weekend_False  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                 1   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                 1   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                 0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                 1   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                 1   \n",
       "\n",
       "                                                ...             \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B            ...              \n",
       "2A429A89-2491-4082-9622-D06E370D091A            ...              \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6            ...              \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E            ...              \n",
       "B9F9226B-3048-4954-830C-52D0524B7415            ...              \n",
       "\n",
       "                                      part_of_day_morning  part_of_day_night  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                    0                  0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                    0                  0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                    0                  1   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                    0                  0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                    1                  0   \n",
       "\n",
       "                                      weatherCond_Clear  weatherCond_Clouds  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                  0                   1   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                  0                   1   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                  0                   1   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                  0                   1   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                  0                   0   \n",
       "\n",
       "                                      weatherCond_Fog  weatherCond_Haze  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                0                 0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                0                 0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                0                 0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                0                 0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                0                 0   \n",
       "\n",
       "                                      weatherCond_Rain  weatherCond_Smoke  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                 0                  0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                 0                  0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                 0                  0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                 0                  0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                 1                  0   \n",
       "\n",
       "                                      weatherCond_Snow  \\\n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                 0   \n",
       "2A429A89-2491-4082-9622-D06E370D091A                 0   \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                 0   \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                 0   \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                 0   \n",
       "\n",
       "                                      weatherCond_Thunderstorm  \n",
       "CFDB9498-809D-4232-8281-9E7CA354D59B                         0  \n",
       "2A429A89-2491-4082-9622-D06E370D091A                         0  \n",
       "4F20435C-DF07-4D4E-BF43-BE54C4A967F6                         0  \n",
       "4A1A66D3-53DE-491F-B1FC-7E12CE4FC35E                         0  \n",
       "B9F9226B-3048-4954-830C-52D0524B7415                         0  \n",
       "\n",
       "[5 rows x 28 columns]"
      ]
     },
     "execution_count": 254,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# one- hot encoding\n",
    "cat_vars=['weatherCond']\n",
    "for var in cat_vars:\n",
    "    cat_list='var'+'_'+var\n",
    "    cat_list = pd.get_dummies(data[var], prefix=var)\n",
    "    data1=data.join(cat_list)\n",
    "    data=data1\n",
    "data = data.drop(['weatherCond'], axis=1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 624 entries, CFDB9498-809D-4232-8281-9E7CA354D59B to 37ED8D81-924D-4EDE-8A5F-CB3875747474\n",
      "Data columns (total 28 columns):\n",
      "clouds                      624 non-null int64\n",
      "date                        624 non-null object\n",
      "humidity                    624 non-null int64\n",
      "temp                        624 non-null float64\n",
      "type                        624 non-null int64\n",
      "visibility                  624 non-null float64\n",
      "windSpeed                   624 non-null float64\n",
      "zipcode                     624 non-null object\n",
      "day_of_week                 624 non-null int64\n",
      "is_weekend_False            624 non-null uint8\n",
      "is_weekend_True             624 non-null uint8\n",
      "month                       624 non-null int64\n",
      "season_fall                 624 non-null uint8\n",
      "season_spring               624 non-null uint8\n",
      "season_summer               624 non-null uint8\n",
      "season_winter               624 non-null uint8\n",
      "part_of_day_afternoon       624 non-null uint8\n",
      "part_of_day_evening         624 non-null uint8\n",
      "part_of_day_morning         624 non-null uint8\n",
      "part_of_day_night           624 non-null uint8\n",
      "weatherCond_Clear           624 non-null uint8\n",
      "weatherCond_Clouds          624 non-null uint8\n",
      "weatherCond_Fog             624 non-null uint8\n",
      "weatherCond_Haze            624 non-null uint8\n",
      "weatherCond_Rain            624 non-null uint8\n",
      "weatherCond_Smoke           624 non-null uint8\n",
      "weatherCond_Snow            624 non-null uint8\n",
      "weatherCond_Thunderstorm    624 non-null uint8\n",
      "dtypes: float64(3), int64(5), object(2), uint8(18)\n",
      "memory usage: 84.6+ KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x123c72c18>"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdcAAAFuCAYAAADTQVJ8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsnXmYnFWx/z8zk5lksgcChEUIa7ELclllCciu7CIIgmwi\nCIICgsiqghdcQEFZL4iIgnJluSAgCIQ9qPBT1pTsIBBCVrJOMsvvjzrNNJ1eamZ6Jm/P1Od53md6\nuqvPOe87PV3vOae+VXUdHR0EQRAEQVA96pf2AIIgCIKgvxHONQiCIAiqTDjXIAiCIKgy4VyDIAiC\noMqEcw2CIAiCKjNoaQ8gqE2OqxvvCjO/8PIDXe21HXa+u+9l5r/nsrvx3cEuu0NlpLvvuTS57AbV\n17nshrS3uOzq2ha57ABob3WZdQwa4rJ7b1Gju+uVmha77LznM6N+hLvvUYMbXHb1bb4x7v2b51x2\n9+zl//y0NY9x2bUPW9Znd/fl7r7n7Xaiy270wqkuu5bhK7j7Hj602fcPUQbvdw7AVR1v9ri/nhIz\n1yAIgiCoMuFcaxQROUJELuphGxeJyBFVGlIQBEGv0VDnP7JALAsHQRAEmafJudWSFcK51ggi0gz8\nGlgNaAL+N++1U4GDgVbgUVU9Q0TOB6ao6lUisi5wlapOEJEDgLOBD1M7k0VkOeAP2ErGEOA4Vf1n\n351dEARBeRrqasu5xrJw7XAc8Kaqbo050gUAIrIR8CVgm3SsLSJfKNaAiDQClwA7A7sB89NLWwDT\ngT2AE4BhvXcaQRAEXafWloXDudYOAjwFoKqvALPS8+sCk1R1sap2AI8BGxS8N/dxWw6YoarTk+2T\n6fl7gSeAO4EfAO29dhZBEATdoKGuzn1kgXCutcPLwOYAIrIG8KP0/GRgSxEZJCJ1wPbAv4GFwIrJ\n5jPp51RgdFoGJtceMAF4X1V3BS7IazsIgiAT1NrMNfZca4ergetF5BGgAVveHauqz4vIH7GZZz3w\nOHAHMB74o4jsADwDoKqtInIi8BcRmQHkBH//Am4RkeOxz8QP+u60giAIKpOVGamXuig5F3SH6b/8\njuuDc9Y3b3W1d+oHz7v7Hs90l90F/2+hy27cKF8yBYDtV1vGZbd2s6/vOQ3DXXYjF05z2QHUtfoS\nU7SOWsllN6fV/6Xm/T4Z4Uz4cMkTb7v7/ubbN7nshux/ksvutjd913GbT41y2QGsPOUfLrvW1TZz\n2bU/fKO77znbHeGyG1XnO+/2pqHuvpuHDOmxZ7xw6NpuZ3XW/FeWuieOZeGMUA3dal5bt4hIU8Fz\nu4vIDenxbennRiKyfTX6DIIg6E1iWThY6qjqwRVe3z89PACYAjza64MKgiDoAbW2LBzONVtsJSL3\nY1G9VwLfA9ZV1YVpVjsZeBM4E2gBPgVcBewEfBr4hapeKSJvYlHEqwPXA/PSMRNARKYAmwFHAItE\n5Fngl6q6RXr9D8DPVPVvvX/KQRAElcnKjNRLONdssRjTn64G3FPGbhVgE8xB3gqsCawM3I455Rw/\nAc5V1QdE5AxgvdwLqvpuWiaeoqp/E5EFIrI+NpNdPRxrEARZotZmrrHnmi2eTfrTKUBhtED+J+sF\nVV2MaV1fU9VF2Ky0MDJnHSDnJJ+o0Pe12Ez2EMAXGRIEQdBHNNXXuY8sEM41WxRGwy0EVkz61U3K\n2JXiJWDr9HjzIq+30/kZ+F9gV2A/wrkGQZAxai2gKZxrtvkxtjx8D2m/tIucCpwtIg8CWxZ5/Rng\nRBHZUVUXYoFNU1V1RncHHARB0BvUmnONPdeMoKo35D1eiCWBAAtIKmRispuMZVdCVWdhQUyoau69\nrwHbFulrXPr5Z+DPeS81YMvDQRAEmaLW9lwjiUQAQIpSnqaqh3jsp86e5/rgzG7xpSn+2QobuewA\nFv7uNpfdtTv7Ej58MGisu+/l6+dXNgLaB49w2XU4vzAaWua67ABm0eyyW9jm+99f/P1j3H0fu+YJ\nLrvLDvy0y07ee8zd93ur7+CyW6nlfZfdtKG+JBtjGlpddgDtgwa77Nx/73Z/315emu9LqrJB46zK\nRonG5cf32DP+dux6bmd12LSXl7onjplrP0BEhgBfUdX/6W4bKa9wEARBJqnWzFVE6oErMPliC3CM\nqr6a9/qh2JZaG3C9ql5ZtKEKxJ5r/2Ac4J9eBEEQ1BhVjBbeFxiSynd+F/hZwes/xcpyfhY4VUTG\ndGe8MXPtH5wFrC8i5wEbAcum509Kif1fxcrLrQM8CIzCariqqh6W9K51WFKK4cDhaT83CIIgE1Qx\nUGlb4D4AVZ0kIv9V8Ppz2HdkK/a92K2905i59g8uxGQ3Q4EHVXVH4Fg6E0qMB84GtgNOwpZEtgS2\nFZHRyeY1Vd0JOB+LUg6CIMgMVaznOhKYnfd7m4jkTzRfwJQULwJ3p2DRLhPOtX+xEXCUiEzEon5z\nET3TVfXtlHhinqq+lJJVzKYz8cRD6eeTWGH2IAiCzFBfV+c+KvARkB9xWK+qrQAisjHweSx17Hhg\neRE5sFvj7c6bgsyRSwYxGbhUVScAX6IzGYRnWSNX4+qz2B1bEARBZqhrqHMfFXgC2BNARLYC8utd\nzgYWAAtUtQ2YCsSe6wBmKtCE3Y19SUSOxZY+zu9CG3uIyD6Y1vWIag8wCIKgJzQ0+WoAO7gd2EVE\nnsT2VI8UkUOA4ap6jYhcDTwuIouwXAE3dKeT0LkGpICmW1T1Pu97Wt9/xfXB6WhodLX3tb/6k0IN\nOXT/ykbAD2b5JuDLzHnT3feCZdZw2TW1L3LZLapvqmwEDHvfX0y+5cVJLrsZ2x3tsvvtP99z9/3t\njqdcdvUbTXDZfdDg0yqDvwB7c+s8l11HffXnHtNaff8P3vS4I5r8i48tTl3zqLnvuuzmjljZ3ffo\n4UN7HI70l3U+43ZWu/372aWuc41l4X6IiAwRkZDmBEHQb6hvqHMfWSCWhfsnOd2rK6mEqh7Rq6MJ\ngiDoIXX1tTUXDOfaP8npXjuADVX1RRHZA9gL259dF1ge26j/pqo+niLiTsGykjyuqt9dSmMPgiBY\ngqzMSL3U1q1A4CWnez0K+Gp67ig6Z7Lzk6b1K8CvRGQZ4PvA51R1W2BlEdmlj8ccBEFQkipGC/cJ\n4Vz7N38E9haR5YFVVPXZ9PxDAKr6IraEvBawHHBP0siuD6zZ98MNgiAoTkNTg/vIAuFc+yftmDB6\nHvAw8As+WQB9MwAR2RB4F3gDeAfYJWlkLwd8IadBEAR9QF19nfvIArHn2j+ZCjSJyMVYpqbHgePz\nXt80FVAfBnxNVT8UkUuAR0SkAXgTm/UGQRBkgvqG2poLhs61nyMim2NBS4en388HpqjqVT1p9/p/\nvO364Lw13Vf/9OxNfTUkAWYOXs5ld+7oDVx2v3rdVx8WgHrfktOUkWu57JZf/KHLbtaQ5V12AHMX\ntbnsVhrss6tf4E+t2jJ8Bbeth+ap6radMmptl12j8zt62fefrWwEdIwa52sQ+Fer77O73ljf/8Nz\nU33/XwCbjvVpbD9s8c38Rg/xL78OH9rc4+nkkzts73ZW2zzy6FKfvtbWrcAARUR2T1mXSr1+W/o5\nUUTWzXv+ROC3wIz0+5T00oEisqqILJMykwRBEGSaWgtoimXhGqBS5iRVLZqySFV/Cfyy4Lnzc49F\nZAKwN/D7Hg8yCIKgF6m1ZeFwrhkkzUR/oaqPpFqDD2Ll487H9kJHYeXlzlLV+0Vkiqrm1qZ+ICJj\ngRbgcGAD4DhVPTiv/YnAcZge9tNpVnw6sIWqzhCR44ERqhql54IgyAQN3vX8jFBbox04XEunPvVI\nzAmCyWPGYskgvkzxm6Pbkob1LuDMCv1cCDykqtcAvwNyDvgrwG+6PfogCIIqU9dQ7z6yQDZGERTy\nF2CLlNxhO6wEUk6XejVwM1bwvNjf79H0s6t1Wa8HDkvynA9U9YNujj0IgqDq1Fpu4XCuGURV24Fb\nsaXgO7CUhIjIRthy7eexme3lRd6+Rfq5HfBCha5ydWBR1beAWdgs+boenkIQBEFVqbWApnCu2eV6\nYP/0M8crwAQReRRzvucWed++aU91F+CiCn28BmwkIt9Kv1+LOWV36bkgCIK+oNaWhSOgKaOo6jtA\nTph2Q95LXyxiOy79nFCkqYnpKGW3Xt7jBuB6VfUJIIMgCPqIWgtoCucaACAiPwJ2BL7gsT9URrra\nvf7FdpfdB4PGuuwAxjmLm3uTQ5ywhq/4OsBlM//mshtb3+Kym9foTA7hu4wArNTss6tb7CvoPrlt\njLvvDZxF3VvHrOJrcL4/gcWUhsUuu5HOouqs+BmX2bIzX/G1Bwwf6kuyMXOh7/523WX9yVfq5/rC\nKFao8y2rttb7k2dUg5DiBGURkXHAuar6Daf9FFUdJyLN2B7sSpgMZwrwdVWdXo1xqer3Un/risif\nSsyCgyAIlgpZ2Uv1Es61j1HVKYDLsRZwJJa28AiAtE96LnBy9UYXBEGQTbKyl+olnGsvISLPAHsA\nM4HpwARVfVZEZgBvqeqmIvIc8AiwMdAB7APMBa7Bkj+8BgxOTX4AHCMiT6T3XA7Upb5eB57GdLAv\nAMcAI7Co32XT+09S1eeLFUUXkRUxnWsdNiMOgiDIFHX1teVca2u0tcWdwG7AtlhJt51FZH3gfix7\nEsBI4GZV3QEr/bYHsB8wRFW3wpJADAVQ1T8BFwBHp/YepDMYaRXgHFXdAhgO7At8D3hQVXcEjgWu\nLFMU/aw0jh0x6U8QBEGmqG+odx9ZIBuj6J/cBuwJ7I45r52xPL7PFNj9v/TzHWAIsA7wNwBVfTs9\nj4hsjTnLfYHlsQjiG9J731bVV9PjXPKIjYCjkiznWmAZShdF/7hP4IkenncQBEHVqW8a5D6yQDjX\nXkJVXwDWwJI63IPNKPdJj/MpLKP0ErA1gIisBKycnv8yaX81SWWeo3MGvHIKlAL4LPAiMBm4NAUm\nfQkrll6qKPrHfQKbd/OUgyAIeo26+nr3kQWy4eL7LxOB1VW1XUQewWaK8yq8505gFxF5GngLmJae\nPwv4pYj8M7UxD1siBnOyvxSRT2HO8i5sBnpdSso/Eji/TFH0C4DficjBmAMOgiDIFHUN/vqxWSCK\npfcDCqri9AnT58x3fXCmzG11tbd280J33y1NI1x2Qz/8t8uubdSK7r5PGrNFZSPgo9/c6rK78LnL\nXHYrn3Opyw5gdt1Qt62HUR3+gtx1i3y2dYsXOO38n4tHFq/kslt/Od/1WaZttstuRsMol11XGNnk\ncySDWj5yt+n9v2lq82m0FzUMrmyUGFGFYulvnvFVt7Maf/FvlrpuJxvz5xpFRIaIyJu92P5+IvKK\niJzUW32kfsaLyKTe7CMIgqAn1NfXu48sEMvC2WYv4BRVvaucUV/PWoMgCPqa0Ln2c0RkOKYJHQO8\nmp7bATgPWwkYDhwCTADWVtXvpP3NfwKbq+oS61wiMhoLOBqJ/U3OTu3sCfyXiExT1aeKvK+UlvZZ\nLEDp2DSWDuAWVb0s7cteAzRjpeyOzWuvAYtAflFVKyX9D4Ig6DPqG2vLXdXWrUA2OA54QVW3x2qr\ngiV8+EqKwL0NOBCrubpvcli7Aw8Xc6yJs4EHUpsHYskf7sKq05xezLEmSmlp/41JbA5Kr22XxiLA\nT4HL0lh/SmflnEHYTcNT4ViDIMgatVYVJxujqC3ydahPA4uxBBCXicgNWPL7RlWdg2VS2g1LXfg/\nZdpcj1TkXFXfBT7CtKyVKKWl/ROwIbAalmziQSxT09qY/vV7Sed6LpDLJP7p1OdwR79BEAR9SjjX\n/k++DnVTrCzctcCRKe/ve6S0hOn5Y4DlVfW5Mm2+jM0uEZGVsSXnign5K2hpFdO77phmqTdg2tjJ\nwBnpua9jdWHBklt8HjhMRDau1HcQBEFfEhma+j9XAWuIyOPACZjG9CbgsZT3dwRWuSY3s10LW24t\nx4+AnVIR9DuAY1XVp2ExLe2HqtqOzZSnquo8Vf0XNmN9XET+gc1a3wVOA85LutsbMYdLGu8C4Hjg\nRhHxx9kHQRD0MrWWRCJ0rr2IiNRjyRx2U1W/IK0GmD1vgeuD07x4jqu99sE+DR5AXZuvDum0xb4A\niLENPl0fwJF3vOayG/nVA11212+yk8tu9lNXuOwABs1612XXMtKnC13U5v+OqHeqC71NNnWhzNiQ\nOb6aEy0jfMH19c66pl3RmtLhK8zr/X+YttD/t1lusNO2zVcXl3p/UofBw0f1WHc67bJT3Sc79qSf\nLXWda22FXzkRkd2BVVX1mj7udzwWlbuViKwO3A78OudYReQ2LMdvPrNVdZ8S7d2CzZS/hGV3KuTx\n9Np7ec+drqpFK3ovjWQTQRAE1SArM1Iv/dK5qup9GRjDG8AmBc/t3822itZ/FZHzgUtU9arutBsE\nQVAr1NdY+sN+6VxF5AhgXWy2Nwor23aWqt5fwr5X9aIi8s0i778B268dD6wIHJH6PAELgnofX8Rw\nsfPZELgEaADGAser6pN5r38D+CrQDvxdVU8qdj6q+k53+g+CIKg2WYkC9lJbo+0aa2KOZS+soky5\nG4le04umdoq9H6xo+m5YdZpjRWQFrPLNVljUb5PjPE8RkYnpuDw9twFwqqp+DrgYkwLlcyRwoqpu\nDbwsIoPKnE8QBMFSp1pSHBGpF5GrROSp9L25Vgm7a0Sk29+D/XLmmngNuBtL5tAIlMuQfhumE307\n/TwJu/Eo1IuCyWTy9aJnYNKbXBTApzGdak4vWur98Mlarp/FHPmLqtoCICJF904LKLYs/C5wjogs\nwKKXCyMujgROS/vCT6XxlzqfIAiCpU4V91z3BYao6tYishXwM2wy8zEi8nXsO/GR7nbSn2euawMj\nVPXz2BLo5aUMe1kvWur9sGQt11eADUSkOS0tb9qtM7cbifNU9avA83TqbnN8DThOVXdIfWxT5nyC\nIAiWOlUslr4tlv0OVZ0E/Ff+iyKyDbAlnRn4ujfenrw547wCTEja0VuxbETlmEgv6EUxp1Xs/Uug\nqh9iy7FPAvdSufZrKW4CbhWRx7CMUoWai+cxXe5DwFTg6XLnEwRBsLSpos51JJBfT7AtbY0hIiti\neeJP7PF4Q+cadIfQuZYndK7lCZ1reULnuiTz//Df7pMdetCZJfsTkUuASar6x/T7f1R1lfT4JGyl\ncw4wDguGPVdVb+jqePvznusSiMgVFNeL7pFmm5lDRFbFZpKFPKKq5/X1eHIMafc5pDkNvlTFQ51f\nZACL6z1xXrD84g9cdvMa/UHZ3uLm4nSaR/3zIZddXRdugt9vWqGyEbBChy8JWJPzbw3QUedbDKtr\n9/XdUedPFPZB43Iuu+Wc593hXNjzfsYB2p1/xmHO67jC4vfdfbcM8d1UtDqdZoP3TgqoSrq3Ljjz\nCjyBBbr+Me25Pp97QVUvI8Xn5FQn3XGsMECca15SiaJ60V7sdzwpqUR321DVt7HydR8nlVDViUX6\nWo7OfdJNsEjn+cBvVfW67vYfBEGQCaoX0HQ7sIuIPInFoxwpIocAw6uZeGhAONcsJJXobdJ+7QSA\nVPHmOFWdvDTHFARBUC3qqpREIsXVHFfw9BLfld2dseYYEM41kkrIDVjJuWWBnwAHqerB6bUpqjou\nkkgEQZBpBvm2g7JCf44WLmSgJJUoxUOqug12w1CMSCIRBEFmqbWqOANi5poYKEklSqElns9FJUQS\niSAIskv1Apr6hGy4+L5hICeVAMsjDLAQW3ZGRFajs0pPJJEIgiC71Df4jwwwkGauuaQSX8JuKjxJ\nJVZX1faUWGF9VZ0H/EtEckkhBgN/ozOpxJUiMgTbtzw515CqLhCRXFKJLelMKpH//iVQ1Q9Tbssn\ngQ/pflKJfP4BzBKRp4GXsWVvyo0/CIJgaZOV5V4vkUQi6BaLZk11fXDqFi90tdfRNNTd96Bpr7vs\npi+3obtNL6M65rvs2gb7tI9e/eq3hq7nsgP4zkk+5dcPtzzdZXf5vv6+G53Sx3mtvvNuHuT/Qh20\nYIbLrn3IKJfdLOfGyNj571U2SrSN8MUkLnJqub26WYCWVl8CC2/ijuZ23/82wOCRy/Q4iUTLxN+5\nz3bwhEOjWPrSZKAnlRCRjYAxqvqoiLyJCab9/zFBEAR9RLWkOH3FgHaufZ1UohrkJ5WoAgcAU4BH\nq9ReEARB71Bjy8ID2rnWIkmzuxe2L7oi8Ass2GpDbN90OPAtTDP7Cqa3PRTYE9P3ronVeH0AOAJY\nlPS7YHuuq6fH+6lqKdlOEARB35KRQCUvtXUrEOQYoap7Yk7yeGB/zIkeA3wf2ElVtwVmYZG/AKNU\n9QvA3sB3VfVdLFL5ElXNSXyuS9HCbwK79M2pBEEQVKauvsF9ZIFwrrVJTg87C3hZVTuw5BBDMV1s\nrhTNo8AG6fE/0893gCEl2n0m/ZyS2gqCIMgG9fX+IwNkYxRBVykVNdcBrC8iw9LvO2BZpUq9p51P\nfgYidDwIgkxSN6jJfWSBcK79i1as0O/DIjIJS/d4ZRn7Z4ATRWTHvhhcEARBt6mxmWvoXINusWjG\ne64PTn2LL+/FjKG+wt0AIx7xVdCbss2RLruVmt1d81Gbbz9n9HxfnU1v7dXF3z/GZQfwk8smuey+\n8NLTLrsnXp/u7vvLm67sslt1ZKPLrgt12hk111ckfsGoVVx23iLxIxbPctmBX2O72Dnv6YqY03sp\nFzo1yKMX+OolAzSusHqPdadtLzzo/jQ0bPi50LkGQRAEQUUyEqjkJZzrAEFELsbK6J1Uotj6DcAt\nwDgsmcR3+3SAQRAE5QjnGmSUA4FP50USB0EQ1Ax1g3xbCVkhnGuVEZF1gF9jwUX1WFH0b2C1Wxsw\nXemtIrIDFnxUjyV+OAQrcfdHCgq6i8ihOBJDqOoNJcZ0LrAS8GcR2QNLPPEpLAnF/6nq2VW+DEEQ\nBNWlLhuBSl5qa7S1wS5YpZudMee5L1ZdZ1tgR+AsERmN6U+/kpI23IbNLJco6C4iy+JMDFFqQKr6\nA0y7uiuwHDApFWXfAjiueqceBEHQS9TV+48MEDPX6nMdcAZwHzAbS96wmYhMTK83AuOxMnOXichc\nYGXgCVV9UUSu5pMF3ddgycQQuwJP40sMUcgMYPMkv/kIGNy90wyCIOg7OjLiNL3U1mhrg32Ax1T1\nc1jB8SOBh9MMdSds2fc14FrgSFU9AngPqEtVagoLur9B1xJDVOIIYJaqHgr8DBgqIks9bD0IgqAs\nNTZzzcYo+hf/AH4gIg9hS65fBOaKyGNY0oaONAu9CXhMRJ4ARmB7ormC7o9ijvlcVZ1G1xJDVOJB\nYPfUx5WpT7/INAiCYGlQV+c/MkAkkQi6RctHM1wfnI5BvlXnqQv9n8MG5z/P2EG+atd1bYvcfc+u\n9xVBbx7kG2NjR6vL7ut3vuKyA/iiM5HD3etv6bIb+X9/dvf9lc18CRpkpK+9xc6i4QCD8BUDX9ju\n+9sMcRYN/2iRr1+A4U0+OYnXPzS0zHX3XbfQKRRwdj5vqK/wO8Do4UN77PFa//Oi+0ti0CobLHUP\nG3uu/QgRORaLOi7kJ8Dna7F+bRAEAZCZ5V4v4Vz7Eap6DXBNiZf9048gCIKsEc61f5BRveqQEu2+\njkUPrwm8gNV1PRfYJo3paODXqrqViDwHPAJsjAVE7YNFDf8K+C9MsrM6sJeqvtmjixgEQVAtasy5\n1tZo+5bM6VWLtZueXwU4R1W3wJzpvun5l1V1G2BBXhsjgZtVdQdMDrRH6nfZ9P6jsQQTQRAEmaGj\nrt59ZIFsjCKbXIc5wPuAE4ExdOpV72NJveoNmNNtVNUXgZxe9QrsOhfTq3alkDkl2gV4W1VfTY+f\nBCT3lhJN5Yqt5/pbD3gq9fEhMLnUGIIgCJYKIcXpN2ROr1qiXYCVRWRcevxZ4MX0uFQYY2F/LwBb\npz7GAOt4xhMEQdBn1Df4jwwQzrU0WdSrLtFuer4F+KWIPI05+Lu62O6fgWki8iQ2Y58P+HQsQRAE\nfUCtLQuHzrUfICJTVHVcZcuS718X2ERVb0l7wy8Cq6lqS6n3vDFtjuuD49X1fXTOUS47gFv3/aHL\n7tQNfFU0JreNcfctQ333HHPrh7rshrfPd9m1NA6rbJT40cNvuOza2n3/+x/t/Xl33zv960mX3Wc/\n5Ssa7tUqA7Q3+jKANs58x2XXOsqXW8Vb2Byg0anFrWst+a/XbebV+a7P/MW+MY4a4p8hjhja3GPd\n6aIZ77mdVdMyK4XONViSMnrVM1X1qV7o8h3gYhH5FhYJfUY5xxoEQdDnZGRG6iWcawapoFctZt/t\nWWt6/zxsjzkIgiCbhHPNPjWmYf14yVdEbgGuwqKU9wKasZqsv8Cc44bAaap6p4i8ikUOr4PlEx6F\nlZhTVT1MRD6FOfBmTKpzbDr3u4DpwD2q+uNuXuIgCIKq0lFfW+6qtm4FqkctaVhLMUJV9wQuBo4H\n9scc5JHp9fHA2dgNw0mYdGdLYNt0bj8FLkvn9lPgovS+ccCu4ViDIMgUNSbFqa1bgeqRuZqrJdot\nJH+TPqdVnYUli+gQkZl5fUxX1bcBRGSeqr6UHs9ONhsB3xORM1K7uciRN1TVn8k+CIKgL6hStRsR\nqccmG5/GVhqPycsTgIjshSkxWoHrVfXa7vSTDRff99SShrVRRIaLSBOdSSc87VZ6fTIWuDQBm2Xf\nmp73l/gIgiDoK6o3c90XGKKqW2OriT/LvSAijcCl2ORoB+BYEVmhO8MdqM61ljSsPwcmAf8LvNXd\nEy7CacB5IvIIcCPwXBXbDoIgqCpV1Llui61aoqqTsJzqOdYDXlXVmWkF73Fg++6Md0AuC6vqa9gF\nzueZInanlGjii0Vsfw/8vuDpG/JeX4gtNZca08IS7f4QKCnsVNX76Pyg/BPYPT0el2eT/3iTvLfv\nVqTJrUr1FQRBsNSo3l7qSGw7MEebiAxS1dYir83BgkG7zIB0rkuTpaBh7RVWavKJ+2d3+P4hjl3z\nBHff93T4LlPL8ANddhu8/7y777ZBPtVTfbMviYQ3m0xjF7abvuwslu5lsjMxBMBDn97GZXfAm3e6\n7FqXWc0jT41eAAAgAElEQVTd91uzfaECo5p9ySGWXbygshHQ1IUv/cWDmn2GTrtZC9vcfY8e4vsQ\njZz7vstuwZBV3H1Xg3aqlhfiI2wlMkd9cqzFXhuBxbV0mXCufUxXNaxdIUl1Di8VkCQi+wFPq+p7\nvdF/EARBb9FevWyCT2CqjD+KyFZA/t31y8DaIrIMMBdbEv5pdzoJ59qPUNWDK5icjO0xh3MNgqCm\nqGKi3tuBXVIu9TrgSBE5BBiuqteIyCnAX7CYpOtV9d3udBLOtQQZTTRxKSYH+l8RuQ+4X1UvEZFr\n01h/D6yLJZpowfZ4VwSOSD83AW4UkW2xCOFDsM/sLaqaK5u3bDo+r6oze3YVgyAIqoMzFXZFVLUd\nm2TkMznv9bvoevGTJRio0cIespho4nZgDxFpxurLfk5E6oDNSPVY83hLVXfDJD3HquqfMc3t4cBa\nwEFYUNd2wL4ikqsB+5CqbhOONQiCLNHR0eE+skA419Jkrlg6Fhb+mdTPn4DlMOf4lKoWfqIKC6Ln\nsyGwGpYW8UFsprp2eq1UgfUgCIKlRnuH/8gC4VxLk7lEE2k54x/A6cD9mLP9MTZjLqRYm+3Y31yx\nsnI7pvO5gU6daySRCIIgc7R1+I8sEM61NFlMNAHmSNcD/oVtuq8FPOJ875NYwoh3sBnr4yLyD2zW\n2q1N+yAIgr6g1paFo1h60C0WzZzi+uC0NY92tff6TH8643UG+WRnLcOWc9k1LfRvL9cvnFPZCJg9\n4lMuuxFtc112cxqGu+wAGpxywGZ8WuVpi/1xj+M+erWyEXDCeF+Fw/Y/3OHuu7nRV7x7n418WuUt\nVvJd89YurEN+tMi3MDSufr6vwVb//83cwcu47Ia3zHDZtQxd1t13NYqlT5k9z32hx40aFsXSe4qI\nbA/MUtUupe8TkS2B3wG3quqZDvuPS7/1Bvnjwf4uE4B5BWY1lWgiCIKgWtTaPLDmnStwFHALXc+N\nuxvwC1W9vKJl3/DxeFJw1DkptWEQBMGAJyuBSl4y41xF5AhM7jIC24/8ASbwPQGLzO0A9sMiXS8G\nFgF/xXLpfkZEXsqVWCtotxHTgK5B0qdiwUVHAYtE5D+qenuR9zVgmZQ2wAKXBqfnN0xtNKRxHo/p\nW7+mqgcmmyeAA4tlQhKRkcD/AKOx/dlfYXu4ufE05Z8TVoP1FKANeFxVvysi5wPbpH6PBq7H9lHX\nBP6mqscnmdBNWK7MQcDZqvqQiOwCXAAsxIqiH4XpX89I13QNTPd6YeHYgyAIlhZtNTZ1zVpA0zBM\nX7or5sDWx5IZbAu8RGei+SGqup2qfh+TxZxezLEmvg58qKrbYJrVC4DXsQjZS4o51sR+qZ+tgDOx\nJA9gzvbUFEV8MRZF/ACwkYiMEZENgGllUgyuhTmvXdN5nqKqf8sbz89y54Sl3/o+8Ll0DVZOzhGs\nhus2wAJgHczJbgHsKSLjsELpD6jq9pj29rqkib0G2F9Vd8ACoc5O7a0GHIAl7j+9xNiDIAiWCh0d\n/iMLZM25PqKq7ar6ATATm63+RkR+DWyMzWCha1rM9TBNKSm69yVshleJdbAkEiTH/U56/l3gHBH5\nDRZB3Jg0pjdhCSOOxDSypfgAS9pwE+bYGsvYroVpWe9J+tr188aefw1eVdU5qtoGvI/pWvPP+10s\nIfXywEd56bzytbbPq2qrqs7DHHYQBEFmaO/ocB9ZIGvOdTOAVJx2FJZu8GDgGOwLPxcBlh9yl9Nu\nluJlLNECIjIC2AhbFq7ES8DW6X0rAblSI5cB56nqV7GEz7kx/RqbIW4P3FOm3VOxpA9fwYKXikW1\n5c7pDcyp75L0qJdjtV1zNjmKfZryz3tlLAnGDGCkiKyYbLqstQ2CIFgadHThyAJZc67jRORB4M+Y\nY30MS+v3GOZci9WKehq4SETWK9HmNcCyIvI4MBH4vqpOdYzlTmC6iDyNFSyflp6/Cbg16V3XyY0p\nzQbnAA/mlS8qxl3ACalI+beAVhEZXOycsD3dS4BH0jj2oNMZVuJHwE5Ja3sHlgJxMfA14La0L7wz\nZWrFBkEQZIVay9CUGZ1rCmhaV1XL5dbNNCJyN/AtVfWJ/WoYr+bsun/4clOcMfoVf9+r7+CyW9Gp\nuWR+F8o1Dhvja3K5dVx2Te0+nWJbQ+H9V2kWtPq0lIOdgtjGthZ3316+ee+bLrv6g/Z1t7nLc766\nsxss79OvrjF4ocuuo6HJZQcwrbXcLlAnY4b4NLvevzXAyAWeOQXMH7aCy25Im3/3aPDIZXqsO31l\n6hy3s1p7+RGhc60WInIFtidZyAXA1FI6WBE5F0tnWMjPsTp+FXWwKZH+R8DPc461zHj2UNU+39MU\nkZ9jAVOlAr+CIAgyS61FC2fGuZYqs9aF93+j2PNJM1pSB6uqP8BkP4XvOxenDlZVF4jIdFX9TqXx\nLC1U9VtLewxBEATdJSvLvV4y41wrMYB0sI1YPda1sT3xs7FApF+o6o7J5m7gHEzDeiGmgX0Nkx0V\nrQ+boo2PwwLEVscih1cDvq2qfxGRL6RrOhuL1H5OVc8v8qcIgiDoc2ps4pq5gKZKDAQd7DHp9e2x\nyjy/SkvaQ0RktRTpOxYrU3ctnZrVd7Gi6FC5PmyLqu4BnAx8O90oXIYtWe9ISHGCIMgY7XS4jyxQ\nMzPXxCOp7NoHIpKvg50LrEtnwfCu6mD/CqaDTVmRuqyDFZFCHewCbJb9kap2JF3rl7EZcjkd7EbA\ndinXMFih9bHpPYcDLdhMezlgReCPqc55M+bEX6VyfdjCWq/LpXF+kJ5/DOi1PMpBEARdJWauvctA\n0MFOBm5OutY9MC3sDGzf+AvYjPn3mDToP8A+yfZC4KHURqWPYeHrU4ERIpIrI7NVhfcHQRD0KZFE\noncZCDrYq4F1kw72SeCtlLVqLlbD9eWUjakdW9b9s4g8iV2PFxzjXoLU1olYJqi/AquCsx5ZEARB\nH7C4rcN9ZIHM6FwrETrY3kVEzsT2mFvSEvb9qnpjKfvQuVZoMnSuZQmda3lC57okk96a4XZWW63W\n8/56Sq3tuXab7upOy+hgj1RVz/JxTgf7OPBQVnWw2Kx6kojMB94E/lDOeNRg3z//N9++yWX33qb+\nWgHevqeMWttn1+CfpM9u8dluN2eKy+6DRl9B9+XxFbAGGNVSWAa4OK2jii30LEl7Y7Ft++K8Ndt3\ns+AtbL6902ECPLDxNi67DT/wVaec6yxQ31Dv/x5f/tUHXHbz19+lshEwbJC/74XDfU7TGw/U0djs\n7rsaZGW510vNONeu6mALi6h7daeFRdRL6WDz7CsWUU/OcrOC5/pMBysiuwOrquo1pWxU9ZfAL/tq\nTEEQBF2hzT9JzwQ141y7QX8pot5jouh6EAS1Tsxcq8wASh5xBLAXJqlZEfgFpnPdEDhNVe8UkUOx\nZP8twCvAsVjSiKOw4LTzsAQUTwCClbc7ADgMkypdBdzMkoXVx2IRyIMxGdNOqrrWkn+NIAiCpcPi\nGkvRVCvRwgMheQTACFXdM73/eGB/zIEeKSLLYoXTd0rnPSudA8BMVd1WVR/EbhbOUdWtMf3q5gV9\nFCusfhZwR0pGcSs1cNMVBMHAoq29w31kgVpxrgOhiDp0JneYhUluOrDzHYI5zRfTWOGThc7zz3ua\nqubGVCyJRKnC6rnIkccqjDEIgqDPCZ1r7zAQkkdA+Ti9N4D1RWRY+j2/0HmlwumV+niBdE5EAokg\nCDJIW4f/yAK14lwHQvKIsqjqNGxP9WERmYTt617Z3fYKuAjYW0QexoqpRwKJIAgyRa3NXDOfRCKS\nR/Q+IrIntv/8dxHZGfieqhbT9n5Myzxf4eK6Vl8Cgvr5M112AO1DfYkcZuHT4c1u8cf4jxjsux8d\n3uiza+zw3W911Pnvgxc775m9e1PD5/gSgQB82OzTzr48bb7LbqUR/uQZdU7J56UrbOyy+8Xc5112\nXfnsfjRkrMvOmd+DxV2QpzQ7NbHemV8X5L2MGNrc46QOt7/wvttZ7bfhipFEorfpTrKGpJE9iM49\nzXxKJo8o1Mg6kkdsQ+deZ9WTR1QqkC4ib2JRxG8A14tIK/aZ+Gs1xxEEQdBTai1aOPPOtbeKqFfg\nKOAWVT2hi+/7hEa2UvKIlIBiQjfG58JbIF1VX6ZzH3k8pg8+p7fGFQRB0FWystzrJfPOtScMMI1s\nuQLp0yitY71SRFZPj/fDZDnri8i5KTtVEATBUqe9xmautRLQ1BMGika2XIH0cjrW69Ls+U3sOl0I\nvBSONQiCLBHRwtljoGhkyxVIL6djfSb9nEKnsw+CIMgUtRYtPBCca2hky+tYC99X6dyDIAj6nMVt\n7e4jCwyEL9EBr5GlazrWqUCTiFzcg/6CIAiqSm8uC4tIs4j8SUQeE5F7RKRoLUgRqReRe0XkuEpt\nZl7n2hNCI/txG13WsVZityuecH1wjvzseFd7E8aPdvc9pt6X42LQuz6d4vQVP+Pue3TbbJdda/My\nLrv6jjaX3awupPVodAoQ3bVAO/wzgfrFPjXZosZhlY2ApkVzKhslvPVXh9X7rvnJwzdy2Q29426X\nHcDJ24532TU5ha73vjrd3fcB6/lqB7c4C7AP6oLQdZkRQ3usO71y0ptuZ3X8VuO71J+InAKMVNXz\nReRgYGtVPbmI3Y+w+t43qOpV5drs19HCHkRkP2ym2oTJb7bKe82tkRWRY4GvYMuqawAzsBzB+XSl\nwPporKrNZWUKrA/C9npfxSKN3wa+rqqzC+rM5utYG4CTPGMIgiDICm29OxHcFvhxenwvRaSIIvJF\n7PvdVcKzXztXp0b2ZEyusrDI+70F1g/GIm0/p6qLk7TlUWDnlLawOywE3lfV75Qaj4jcB3wjF50s\nIt8Grsb2lPPP42MdaxAEQS1SrWo3InI08O2Cpz8AcstSc7D4nPz3bAgcggWcnuvpp2adq4g8A+yB\nRQBPByao6rMi8izwG8zBdGCz0ctKaEnHAJsAN2KzzuVE5A6snupzqvo1EfkUtsfajO3RHpvauCv1\ne08axymquhhAVd8QkU1UdXpKynA9dq07gJNU9V8i8gpL1l1txjI8jcFmo+XOfzVgXIHs5zJMH5tv\nt1F6vi6N9yhgLuaEP5XO9f9U9WwRuQFYNh2fV1V/XrcgCIJepFrOVVWvo0B9ISK3YfkQSD8LVx0P\nxwJQHwLGY/kM3lTVkrPYWg5ouhPTqG6LLXvuLCLrY07pwPT8dsC+IiIU0ZKq6p8xCcvhWAKJkZjs\nZWvgcyKyPPBTbGl2Qnp8Uep/HLCrqv4YC0B6PX9wqprbDPkplrVpe2yWnPujFqu7ehzwQrK9usL5\nr0RBhLKqtqlq4abgtcAJafz3AKdjTnWSqu6G1XXN35x/SFW3CccaBEGWWNTa7j66wRNYIh6wydIn\nJIuqerqqbpm+R2/A8hmUXR6u2ZkrcBuWHOHt9PMk7GbhT5hDezDZjQHWplNLugC7M/moSJuv55yK\niEzFdJ8bAd8TkTOw2V8utOQNVV2UHr+FOayPHZuI7AY8xyc1sf9MM2EoXnd1HSyqGVV9WkTKhbG8\nDayS/0TKHPUlVf1d3tPrAVfY/QWNwCvYfvDmIrJjug752dG7ovcNgiDoE3q5CPqVWP6Dx7GJ1iHw\ncaDTq6r6f11tsGZnrqr6Ajb72wKbkQ0H9gEmAy8CO+bdZTxHaS1pvq6z2F9vMnBGauvrWIaj3Pty\nXI857kEAIrIO8D9AG5/UxG6CJWso1Ve+DnZTOhNcFDv/d4FpIrJP3tMnp2vwCVPg8DT+04G7gSOA\nWap6KPAzYKiIFNP7BkEQZIK29g730VVUdb6qHqiq26rqTqo6JT1/SaFjVdXzK0UKQ23PXME0pqur\naruIPAKsn/YzHwQeF5HBWEakd+nUks4E/oPtu4JlLroR20stxmlY/t0h2J7oEuHZqnqLiKyY+lyE\n7cl+RVWnishpwLXpZyNwdJnzuQq4Md09TQYq1Ws7DPhVarsJy1f8tQKb41ObuT3fozGH/3sR2Tr1\n8QrF9b5BEASZoJdnrlWnX+tcg94jdK7lCZ1reULnWp7QuS7J2fe+7HZWF+yxXtRzDcqT9LOHFHnp\nTFV9qq/Hk+OevUa67N4dNqqyETCmoQsJqJz/Yh2jxlU2Apad+Yq76w/HrO2yW6al2Jb+ksxxOoSx\n8/0Fy9sH+xzXrHbf36bZWfgdoMlZ1L3VOQtpbGhy993g/LL3Fjf3Os35+37BZQew8IPnnJa+6/jK\nB3PdfXes6yvU7i3A7i1OXy1qbeY6IJxruUQRXWwnP1FEI3CWqk7swbiGAJNVdXwZs1uBLYG1Up+l\nEkV0p/91gat6s6ZsEARBNfDOqLNCzQY0dZGTMZlNtylIFDEBc7K/FRHf7WD3uRm4W1V3SCXunqay\nTCcIgqBf0ZsBTb1BpmeukSjCnShiU+ByLDp5IRbUVE/eLF1EJqXr1ZL6r6MzchkRuRDYMZ3Dn1Q1\nEvcHQZAZsuI0vWR95hqJIvyJIk5MxdCvwG4wSnEWcLOq7gjckff8odje7nYsmZ0kCIJgqdLW0eE+\nskDWnettWNaM3TGnsDOwN5YoYjUsUcSDWLq+/EQRHxcdL9Lm66o6U1XbsfJq+YkiJmJ5I1dItsUS\nRXyMiOyWJDifSBSRZ1cqUUSuYPrTlC//VjRRhIgcWmC3UuqXNI4NirSVCz/4uH9sVp3jUOym4i+A\nP3Q3CIKgD6i1ZeFMO9dIFOFOFPGeiGycHu8A/BtbHl5eRBpShZ3VC/vHZtIkPfCBwJexpeEj0pJ0\nEARBJujl9IdVJ9N7romJRKKISokivgb8MmVZagWOVtUpIvIA8Pf0ntz+7gXA71KA1hvp3FpEZAYw\nCdtzvh+bNQdBEGSCtvZsOE0vkUQi6BaLpv3H9cFpmOqr8b5oDb86asYCXxKAKXN9mReGD/Yv4Iwa\n3OCyW6bdl/xgdoMviH2kM3EGQEe97565vc53Lt5EFwCtzsWwGQt9bTZ0QUy5/KsPVjYCPlp3Z5fd\n3EW+L/OFbf4v/UtX2LiyEfCTeS+77Lz/CwDLDfV9Lrx7ll2RuQ4f2txjVewhN/7d7ax+f/jmkUQi\ny/SVPrariSI8+lgReRObfeb+82eo6v7dGX8QBMHSJit7qV7CuZanZCF1L6UKqYvIprlC6qp6DSYF\nqja7qmq3xx4EQZAVvFm9skK/dK4DXR9b4docCnyLzoT9x6b+b8SkP+8A26tqJPIPgiAz1NrMNdPR\nwj1goOtjc9wvIhPT8XkRWRb4PrCTqm6L6Vm/jjnYN1T1s8D5dEqRgiAIMkFEC2eDgV5IPccnloVF\nZHPgRVXNRds8Cuyaxn5fanuyiHzoaDsIgqDPiJlrBhjo+tgyvAGsLyK5sik5TewLeW2vSaeEKQiC\nIBNEEonsMBH4MGViegSYqqr/wmatj4vIP+icteb0sY9hM8TcfmNOH1uqOOdpwHlJf3sj5qg/gare\ngulHHxeRR4Ffk/Sx6f3fTM9fSWV97BpJH3sClfWxS5ACqM4DHk65hsemfq8DxqdxnE8PAriCIAh6\ng472DveRBULnGiAi2wDDVfV+EVkbuE9V1yz3noULFrg+OHXtvjqtda3+e4XpHc0uO68mdaZTcwkw\nZkh1taFerWlv5Ev1FruuXzTf3ebiQb6/jbcIeutgfzGrxc4vVe933oJWn92iNv/fZplm39/7O8PW\nc9mdN/NFd9/e/4d25/Wp74IGeWjzkB7rTnf8+aPuC/3wt7YPnWvQfapYSP114GYROQ9bbj6hGuML\ngiCoFrU2EQznSu0WU1fVa0TkB/kF00Vkd2z/1+1cVXUKllM4CIIgk7RlJArYS3/ec+0KtVxMPQiC\noN/T0e4/skBNz1wjWUTF63MisD8wDJgG7IcFTR2UTNYCHkjncxUW4FUPnN2TGXcQBEG1qbVl4Vqf\nuUayCFgmL1HERODHACJSj9W53VlVt8Qc++aqemU6j9MwHfApwDGYtnZ7TLL0K0e/QRAEfUZ7e4f7\nyAI1PXMlkkWAJeSfkNfn7sDBqUTfIixQaS5WdL0x2ayHOe69VXWmiGwEbCciW6ZmBonI2Fzu4yAI\ngqVNViQ2Xmp65hrJIkqTiqfvq6oHAd/Ezq8uFUG/GdPavpt3fjen89sjnd+M7vQbBEHQG7S1tbuP\nLFDrM1eIYuqleBWYJyJPpN/fx5auT8Nm479KS8dvY8XWr03XbyRwRUq+EQRBkAlqbeYaSSSCbjH/\n1h/7kkg0D6tsBDR89gB334sGj3LZvfDhApfdussOcfc9tHWuy25q+1CX3QqLfWmc5w8fV9ko0eCU\nzzctnuey66jzL3B92Nrkslu+fZbLbs7gUsnRlmTYIN+Jz1ns+86769++XZFXPvB9JgCO22pVl12T\n84/4/TEbuPu+eK6vAHuLU/LiTUICsMyIoT1O6rDZOfe5ndUzP9w9kkhkiQzrXb+BRTlPKnjp42QR\nabb8ZHq+EZs5f1lV3yjR5u7AqqmWbBAEQabJSqCSl3Cun6RPiqN3g+uB0/MDl4pQGNj0deBU4MRi\nxqp6XzfHEgRB0OfU2iprv3CuoXctymrpepTSux4CrIvt8d6MRSuvCfxNVY/vRn9BEAS9RlaSQ3ip\n6WjhPELv2ql3fVZE3sRkPReX0rsWvHcdLMhqC2BPEfFv8AVBEPQBba3t7iML9IuZK6F3hbQsLCIN\nmPRokarOTf0X1bvm8WqugLqIvJ/6D4IgyAy1Fi3cL2auoXftRFXbsOXq/UTk86X0rgVvq61PbRAE\nA472jg73kQX6hXNNTCSKo+fGsABLaXg58BqdetcH6NS7BkEQ1AxRLD0YEHz40XzXB8f7+Rrd7iue\nDTBnkK+A0fB6X8Hy+rn+IO6FI3zb0U0dviLxi+t8OzPzF/v3kZobfffMQ+ZMqWwEzBm6grtvrz7T\nW2B8xIKp7r4XDveN0yuA9NZA78p3aFNDdecz3gLxAGcM9xVgv3TB5O4OpyTNQ3peLF2+eYf7ZPXy\nfbvUn4g0Y5Ou5YE5wFdV9cMCm1OxQNB24Eeqenu5NvvLnmtZMqxfLVuvNdnsgc1464DxmEyo8BvH\nXRxdRM4HpqjqVd0YchAEwVKhl3WuxwPPq+r5SU55NnmZ+ERkdPp9LUx18U8gnCvZ1a96uBrYWFVn\nicgI4F/Al9IycxAEwYCgvXdzBm9LqigG3AucU/D6PCxYdVg6Kg4m08419KsAzAJOFpH/xYKc1lPV\nljQDXSud57JYmbgDsD3kr6rqpLSMcTDQCjyqqmfkXdu1gN9je7NvYbKgZdPLJ6nq846xBUEQ9AnV\nmrmKyNHAtwue/oBOhcccoFiO1Xew7+AG4L8r9ZP1gKbQr8KumAzoZiwY6UwRye0nLFDV3THJ0Z6q\nulca+8GpjNyXgG3SsbaIfCG9TzDHeqiqPgd8D3hQVXfEbiyudIwrCIKgz+hob3Mf5VDV61R1w/wD\nc6wjkskIbFKTzx7YhGx1YFXM52xRrp9Mz1wZ4PpVERkDrJZmnGeIyMrp3J9JJs+mn7OwOyqwWf4Q\nLPvSpNxMO0VG57J874HNZnOfwo2AnUTkoPS7P1t6EARBH1DJafaQJ4A9sQpqewCPFbw+E1vVbFHV\nDhGZBYwu12CmZ66hX2Uw8AcRyYVBvp/azslyyq2TTAa2FJFBaaa7PfDv9NrPsWWR36SkE5OBS9P5\nfwmLmguCIMgM1Zq5luBKYIMkfTwW+D6AiJwiInur6mPA34FJIvIU9l36QLkGsz5zhQFcr1VVp4jI\nScDdItKa+rxbVe8XkW3K9IGqPi8if8TuyOqBx4E7gE+n1x8QkS8CZwAXAtelaOiRwPnl2g6CIOhr\n2hcvqmzUTVR1PrbVWPj8JXmPzwPO87YZOtegW4TOtTyhcy1P6Fyrw0DSua785WvcJ/vuzcdGPdfg\nY/3sIUVecutX+5rRC31feu1DfI7wpfn+dMYb1r/rspsyeEWX3Qp1/v/DpjZnoixnCY/W+gZfv94K\n6MDCVt930BDneXfFsY+c+77Lrqmx2df3ML9jr3YST2/R8C5cHhrqfYP0/rW9YwS/0/x287ouu0vm\nV98Jl6OX91yrzoB2rllJLpEKln9ctDwvuURJx1qQXGIocLmq/q474w+CIMg6teZcMx3Q1AecjO0x\ndpuC5BITMCf7WxEZW/aNPedq4ABV3Sn1/8MkKwqCIOh39HJAU9WpyZlrJJcAyieXWB3Lkbka8G1V\n/YuI7AJcgGWpmg4chRUVuFBV/yEik4HvqeptInI/phH2rb8GQRD0Mu0ZcZpeanXmGsklyieXaFHV\nPVKf307PXwPsr6o7YFWDzsZyY+6RUjm2pOs4ChgSjjUIgizR3rrIfWSBWnWut2GC392x5BI7A3tj\nCRZWw5JLPIil88tPLvEb4IsU15a+rqozU8m6wuQSE4FzgVx0RbHkEh8jIrsl2c4nkkvk2ZVKLvG3\nZPs0nYksliA/uYSqbgxslq5FLgPT/ytoeyzwUZ7DfBS74bgLW1LeHbvp2AKbid9Vqu8gCIKlQUdb\nm/vIAjXpXCO5RJeTS0wDRiaHD7AD8O+UqWo+cBBwH5YJ62Ts5iUIgiAz1Nqea00618REBmhxdFWd\ngqWCvDtlC5kEPKuq95ew7wC+BtyWiqbvDPwwvXwnMFRVZwB/SY9fKzPOIAiCPqfWnGskkQi6xdz5\nC1wfnIZ6n2Jv0Ee+hAYA84b6gqIHOfv2jhGg1Snab2r37fssqm/ytde6wGUHUL9gdmUjYJ5TQ9qV\n61NtunLeHU7t7GLnV95iZxYJb7IJgOZBvmtZ79QgL+yCztWbXMTrEk4Z6tPDAlzV8WaPP0Sjdz7b\nfaVn/fWCgZFEIit60iLtlS1WLiKHYbPNIcD6dCbKPxR4RlV96Xp8Y5lS2F5XkkuIyKrAp1U19kuD\nIOh3dLT3aj3XqtNXUpyaLFauqr/FNKvjsZuCCXnj6Y0uC/v/RHKJCuyEVcIJ5xoEQb8jK1HAXso6\n10NzwRIAACAASURBVNCTlmWwiPweq+03HYtCPguYoqpXici6wFWqOkFEnsP2hTdO49sHmJvOeQPg\nNSxICee1mAt8FZvB/x2rcPNdYKiIPIlFCV+OBVUtxPZb6wva2BP4F7Bhau8xTN40GpMZzezBtQmC\nIKgq/U3nGnrS0gzHki5si1Wt37SM7Ujg5qQxfRe7UdgP05NuBZyJSX9y51LpWhwJnJjO62Us+vki\n4Peq+n/Aten1HYArsBuewjYA/pb+VoOB+aq6Cxa1vEM3r0kQBEGvUGtSnErLwgO6WHkFZqjqm+nx\nFDqdY47CDfVC7elKdOpa3xaR3Dg91+JI4LS0NP5Ukb5WSrpasOtyUZE2oHyx9SAIgsyQlShgL2Vn\nrqEnLUuxthdiy90An6lgnz+OlYCV0/Oea/E14Lg0M90U2IZPXuP3RGTj9HgHOoukF0YERKh4EAQ1\nQa1JcTwBTRMZoMXKu8EfgD+KyA7AMxVs7wR2EZGnsVl5Liir4rXAblweE5E52HV/GlslOCvth38N\n+GVKe9hK+esRBEGQeWotoCl0rkEQBEFQZWqyKk61qcVi5UEQBEF2iZlrEARBEFSZWs4tHARBEASZ\nJJxrEARBEFSZcK5BEARBUGXCuQZBEARBlQnnGgRBEARVJpxrEARBEFSZcK5BnyEi/7WU+j1NRJbr\n5T66m0YzE4jIqgXHisXOKWVB87Q3qvqjXDp091zycpxXYwzLVKutLvS5YmWroBSRRCLoESmH8TAs\nb/GPgB+p6oMlzE9L5QFvAm5S1Vkl2lwH+AlWZOFFrNLSWyVsPwesCUwC/q2qxWoGzwVuF5EpWMWk\n+1S1I6+NhymRZ1lVdyrR7+m56kIishFWgrEwn3TOdmWsStTyWK7o51T16R7YjcJqG39cLEJVb8x7\n/b/LnM/3ij0P3A2sAih23ecBg9J53pRnt6eIXKqqlRK4/hmrmlUREfkecDowH8tH3qGqKxWxO7fg\nqcVYIYw/5EpRJrtXsPSohXanq+qz+Q2IyNmqekHe7/+tqmf24Fy+gxXCGA0cKSL3qeopRezexf7O\nH2JpYhdiZTG/oaoP5NntAPwKaBCRW4G3VPW6wvby7Jcnr/CGqr5dxGZD4Eqs4MpNWJWwu4s0d0NK\nb3sXcJuqvlGiz1FYDvP8fv9YaowDhXCuQU+5CjgR+D5WOenHdFZL+gSqerCIjMGyYd2aqiJdq6oT\nC0xvTO09iX2p3QDsWNieiPwIcwjrYTmizwS+XKTfq4CrRGSDNMarReR6rEzhTKwMIcB5wB1YDeAt\ngC+UOe8NReQ4rJjF4Vjt4lJcA/wMOAerUvQbYKse2N2O5aMuVaBicpmxlOINYCdVnZb+Rv+D5ai+\nF/sCzjEWKwzxRuq3Q1W3KdLeDBE5GXPW7QCqen+Jvg/CKjnNrzDGT2M1jh/DrsungPexspiH5dk9\nhN2cPIYVxzgG+DVWWGRbABE5Oj2/nojsmd7XgOUmL3SuXTmXA4DtsRu49UXkoRJ2jwLnq6qKyJrY\nZ+8H2LV+IM/uh6m9P2E3r0/QWVLzE4jIFVid5vdINylYUY9CfoFV1ro2tXUvdnP1CVR1NxEZiZXI\nvElEhqpqsdKa92PFU3I1oDuAcK5LewBBzbMQm102qeokEak0o1kBKzA/FqsM9EUROUZVv5JnM09V\n702P/ywiS9z5J7ZV1e1F5GFV/Y2IFHVwIjIaOBhzgrOwYggN2BfKZ1VVk90KeXfct4vIN8ucxxHA\n70h1glW1XAGIZlV9KM2SVESKza67YlenqkeW6e/9Mq+VYgVVnQagqjPTtZghIoWVlPZytjcd2CQd\nYF+4pRzSG5jTrMRoVT0gPb5aRO5X1cNSEY581lHVv6bHE0XkHFV9UETOy7O5CbsJ/B5wYXquHZja\nw3Npw+omf5B+LyxFmWOV3OdOVV8TkVVV9VURaS2wa09/hw5VXZiKdZRiC2ANVS38my1B6qtDVT8s\n1aaI7AvsDGyJlR39S4nmZqvqEZX6HGiEcw16Sgc207xHRL5EZ/3ZJUgVgOZjd8zn5hySiBT+074j\nImdjM5DNgBYR2RWWmDEMStWDOkSkAftiK8bfsS/Tg/OXyVLJwcIxHo1VedoGWKIMh4g8RedMsRGb\nTT0sIpSYwQEsTLWHG0RkK+yGpMt2ItKUHr4uIltj9Xg7AArq9C4xe0+UcwrPiMjNWH3grYF/ishB\ndDqJHK0ULF1js+hPoKpHpuX9tZLNeyX6BWgCnheR53PjVNViub5Hi8jYNLteFhiV9oULHdiitKrw\nJPZ3bBGRzcj7vkufvTdF5HTMgeS3cWN+Y4U3MhX2Iiem4ysicimpdnQR3heRi/LGOEVEdmHJz9yr\naZl/WRH5LkWudb4ttjRbaQVghoh8HRgmIgdjN5zF+G9sRegibCZeyu4v6XrnakKjqo9WGEO/J5xr\n0FMOwu6Y7wUmYDPEUpyoqn/P/SIiO6jqI6q6W4FdB7aPumb6/QPMYRQ6h0ux0n7LYWX3Li3R749V\n9dq8fk9S1ctU9awCu0OxZeMDsS+KQ4u0Ve78SnEs8FNstn4apZeQK9kpdg3qgPy94A6s7rIZlZjV\nlnMKqnqCiOyNLbH/VlXvERHB9tvycS1di8iJwH7AMtiy/trY9kExLi41rgLOA54WkY+w5fhvAqey\n5DLpIdjfcR+sPONh2Gf0qCJt3kb5JXZE5AfY36IJc8L/BjYoNsD0mTorve/v+XvBBRyO/b13B14A\nzsdqMxfeGB2HLV8/ju2DH1OiPbAVobdE5NXcuZS44Tsam7FPA/6L4tcFVV0vxUjsBtyWloWLbVNs\nBwzG9l3BrmE416U9gKA2EZHDC57K7Xl9ioI7fxHZDlgf+LaIXJKebgBOADYsbDvNekbyyQCJJZbr\nVPVWEfkr5oTfUNXpBf1+Gdgb2FFEcnu2DanPy4q0N0VE7sQc1STsy6zQ5q3U9hLBR5SYVajqf0Tk\nUMwpbo3V4C1qRxnnraqrp743L7hJmVDMvitOQURGYNf7fWCsiByeHySVh3fp+mBsr/BBVf2FiPy9\nhB3A/8Oc9fppjD8sZqSqd4vIPdjN1NQUlHZfEbvpaVaY+/wMy9tmKKTSEjvYZ2gV7ObtEuCKQgMp\nERSXVjSKBcUtpnOlpRFb/i1Wgetq4JQUN4CI3IBtSRSj1IpFIauq6ndTe/VYMNlFRcb+GWwPdxds\nNlxqH3W4qu7s7HvAEM416C7rpZ9bYf94/7+9Mw+zrKru9tvNJCIRARWQ2CQEfxCjYtoAKsKnIjij\nRo1xiOKHoAaCiHECFYfWBOMAojKqoEwqip9PFIwxgmMcPkAEXA4IiqiICogMBrrzx9qn69xTZ9j7\n3qq+JXe9z9NPVdfdde6+t6rO2mut31rrq8Df4DeK5k35t3geaiOg8p5W43/U85B0Ci48uYE5YcY8\nJa6kh+E3unsDP0u524tqS87FjcUW+E2qet4fdTxvlkAqkSs+QtK7ccHHivQ6fgk8v2Xdz5nzTDcH\nrjCznWqP744boJfXDinLcY9w3iGFDKNQ41N46Pan6f9d47JyQ9zL0zWq6/TlpD8AnI/nsPfEPd0n\nNxelsOmhJKPZZbiSsOdx+M++VdhTEGIH+LmZ3SZp05Sr3JD5lIriTsDDsZ9Lr/kk3JttsjeuO3i6\nmf0C/x1qvt79zeyktIfmz61NHX5yOniuxv9WL+3Y4xG4Z/9kM7uhYw3Ad1N4+ULm3sPv96yfCcK4\nBmNRlSvISw2eUH1d0rycnpl9F/8DPMHMcsQ2O5rZ9sPLeA/wbDO7TF5ecAKjN9FNzOyLkpr5vrt1\nXC9LIJXI9eDABU8vS9d9pKQuNfXasK2kFXiosM71+OEk65BCnlGoWN4QlXWRG+I+Az90rEje5jk9\n19zCzN6TPr9I0tM71r0LeBlzB4AudgG2HxD2ZIXYE1dLeiHw+5T/3GzexcpFcTuY2R7p83MkfbVj\n3Q/x9/nTkp5Lu66gej9yVeLPxn8+GwOHWnfp3H64gf17Sd8H3mxmv2lZ96D0r2INo+/pTBLGNZiU\ne0nazMyuTyKTLZoLJH3czJ4O/H9J1cm6s54R+IYkVTesHq43s8vADbikppDjMNzTOb7x9a4//lyB\nFOR7cKQ1K3EBzYbApj1rAQ8/S9qx8bXqkHKimfUJhCoGjUKN70jaFbiIbg+uCnEfQC1k38H7gc/j\nHrXhatMuNpa0VQrL35vRGtU6P6mpgPsYFPZUIfZMDsQjAB/DQ7JtYqu1aEAUl7hLymHeLGljul8z\nZvbtlIY5EzeIzccrQeBpzEWPlgEjf1vp51bxFdy7317S9mZ2QstTn0xGRAH4jJm9vWv/s0oY12BS\nVuHexm+Au+MikxGSYR3xzAa4AfimpJvoN8LXSjqJOVXx8uoGYmYnmNmh6fN5NbId5AqkIN+DAw+9\nvQ8XjhzFfGMPgFytWx0+tma+UrdiL0mvwT3Y6v1Z621J2iOpNQ9JryXHKOzJaJlNmweHpFOBh9MR\nspe0FfAn6TU/D7gYNxyfwz3KNl4HfDUJlTbF62vbuFbScYyGH9uMQq6wJ7fhxCb4z3sbvHyry2DC\nnCjumXi4tU0UB15rerGk7+Kh/jd0rDsLwMwul5fGvLfnuT+JG9b7pNd0De6hVtT//m7AjXXf32Ru\nROFxkt5pw41FZoowrsFEmNnZSQRUiUzm/YFJ+iDdHYPalIqPAjY3s2bNX5MqDLYDcCN+yt66eq5a\nDrNOp7GuCaT+AhdIXdf1xMmDe3167ovpECmlte9jLt/5sp7XcxZzhfi3At/qWPcq3BB2hUePkfRw\nvAzkMfhrbjXotT0+qO/xGhoI2e+GG3XhYXrw0HVXjSTmHYn+XN5d6Nc9N+mqQ9BW6WNXXjhX2AMZ\nDSfwnPBn8QNI1eVrz3lXcjbEu4tVbILXyY5gZqdJ+ix+gJknxqtxUsoJV97oaT2vZUsze2g6cB7M\naDMKzOyN1eeSnkASkJnZpzqulxtRuCd5jUVmijCuwUQ0VZIdIpMz08eX4MKnr+Dhqy5P5vskkdLA\n098L7/B0UduDBZ4yAPIOTseR2sJJ6moLl1Vq0mLcl6WPXZ74K8wsp83eFWb2w57Hz8PVy9vgIdnq\nued5o5KONbODNFq/C9BVt9sbsjezc/Ac4uPN7DMZrwW5kvtk3Ju6h6QX2WgLwG2TkvqMrmukdaXC\nHshrOLGFmX1A0nPN7KtJYdvFWem5lwN/BvyAWuvEWooEYNceFXPFkDdapwqDb2Jmt9RSMCOkFMEO\neHnP8yU9wszaekY3IwoHtKyB/k5mM0sY12BSKpXkMjw0u3NzQZUTknSYpX68wFck/UdzbeLheH7y\n18ydhNuM0aeB18rLYj4CnGZmN1YPJrHRWxrh1mpPbSHSY8hoC5cYLDUpNe7kt9m7OXk99fzoWuNh\nZq8CXpUMxbyyFkm72lzP4urx/cjrkpQbsr8mqXbr5VSt9ZRpD7ub2TXpZ/kJRr2ul6d/Q7nzUmEP\nZDScAKjy35K2xRtptGJmD619z2bMee8VdU3CP+O/Y330eqMNPiHpdXi4ubWULLGHmT087fFovOxs\nHrWIwpZ9URxcm/Au5kqpDu19RTNCGNdgIhoezPeSmKOLu0l6FN4x6WF0iGLMbIfM5z4XOFc+8eZo\n4O2SPo6rGn/EXAOE43Kul6452BYukV1qkgRP+1ETmtj8xhmQ32YvyyNsM6yJt5GMkplVed2TMr3m\n3JD9h4BjGVb2AtxRCbTM7GdN5bWlxvdDufOasOdUPDIyJLqCvIYTh+Ch4p2AjwMvzbgu+EFkXt66\nxrKexyqyvFEAM3uvpGVmtkbSv+PCrjY2kLQ8qamriMY8VCt9klQ9R5sQ8ERcwHYB3kjmZODRg6/s\nTk4Y12AiGgrEbegucwHvDHMUc9Nu5tV6pmuOhGfpmNohaSdcqPMkvOXcI/Df6Y8CK83s4rT0u3hY\n8H7p83kF84nctnBQVmryfvx1Px2/gY+UxEjaz8w+aMPNDCpOwxWslafw/szvq2i7qf9e3q6v7jW3\niYVyQ/a/SCHaHG6Ul6xcgEcD2so9qsYlr2bUG24zXmfjecCr0//ndQyqhZq3YPTwtUUzVGtml+D5\n2EFq4fVlaQ9NT3OZvGXj8trny9LztAmlPpFy+5U3elPLc67AlfG/xRub3Iw3c/kkrh9ochYeOfo6\n3jf4rI6Xk1v6dBcz+3/p83PU3Qt8pgjjGkxKPfR5C66SHEHS+snT+RE+NaTztJzIDc9+F/cwjkyn\n+reZ2WvkE2/qnJq+/1TcAJ8CPKXles22cH1e+HHUSk3M7Ds9a68zszMk7W1mR0o6v/H483DPKJfj\nccP/H/Q3IOii7b2v6izvPfC9Vci+ChN2hYWvlPfCrSt7u/oaPxevp1yFt53sCh+/Ci8FGbrZb5Uh\nqBkMNZcK4hL1Dlu31iIDFSsYzYMbHflwaPVGf9DynGfgkYIVwJsk/QF4Gh2dnMzsHfJ+3jsCJ5uX\neLWRW/q0vqQHmNkl8vGLfX/bM0MY12AizOyNSXl4f/+vXdmy7FQ8/FYV7kPPDSVdtzM8q7lxYTfh\nHtyBSWSyIfAaM2uWK9zFzCrv7mJJf0sLZnajfLjA9bi33OpBJb6eXs/ZtN/w6qxO3vhd5fG15uDr\nu0ragRaP0to73eQ2ICjhTfiB437ApV1CrtyQPV4mpPQPeoYGmNkNkt7OnEe6KXOq6TpDQq6K70na\nxnpqgXNCzWPkzKGRf5R0aP1vwgbqa6t8eFNhX4VlE83Dx+oqypAUuxcAO1v7bGMk7YIfAu4CPCqJ\nENtC3bmlTwcDH5C0DR7R6CqlminCuAYT0aI83MPMDquvqcRDQzeWGs3wbPNGmzUuTD6VBeA6Sc/A\nyy12Ya6ko/laTsLD2l8D/kHSoy3VyjYxs5UpLP1k4POSrjWzp3a8npfjh49jgNOZ32heuAfVNK5d\nzS7qDQjuSk8Dgg7awsInMvraH2XtQ76fhEcV6qHZx9cer6IUB+ZuRhntChO9Qq4auwM/kfSr9P9O\nT7Mv1Czvt3w8cKCZ/U7Ss/Gf9/5mNi88m5g0/1jlw8+sfe0ourtwwegkqt8ALzDvu9zFKXj4uO0A\nU6dZ+tTFtmb2N9V/5NOxLu5ZPxOEcQ0mJUt5mB4/EL/p1m9kf9my9BJgO+BXeHj2V/UHLY0Lo7s0\noKIe8nspc0KUrhvPA8xs1/T50Skn1YqknfFRZZXxu7xlzUPM7Ftmdilz/VtXtlzuog6hSBfvxov6\nL6W/AUEXp7d8Lfe1/xv+M+y6MRdHKchrVwiZQi7gxWbWNaS8SV+o+Xi801JlSD+K6wqOw0PZbTTz\nj6XK2Sr/urYuWNKr6/9vof77fMOAYQX4gZl9qG9BUgi/MX3+BOC2ZohY0hPxNMHfy/t8g+eS9yWG\npYdxDSYmS3mYOASfstF6Y66Fe3dizlg9AlfZFjOkLm3hh5L+zMx+LG9o0Ney73zgCuBw667nPIq5\n/N3RZnZI4X66uAkvN9k07bFqjTdC8srmdXKy2vi9Grmv/VIz+2LXxmyuxOnZ1j7lpY3cOaSn4XnE\n++LNH7pyhUemx3PoCzXft/Z6SB75vyXRUhfN/GMpbX8/Q8Zyd3n/7GXA5rXPuzz2syWdyej81TdV\nnycP/U0pMvMaUlQhhaxX1a5zMS4Iu4W5PPJqWn4XZ5EwrsGk5CoPwRsb/NS6O/BkhXtzUSrYbwhT\n+m46uwGXS/opXrR/W/W9Leu3wMOP+0g6DO9O1ewMVA+/9t1onzHwOt5gte46eAegA+hXM8NwJ6c6\nD8VzlT+h/7V/KhmXtZ66tdev7ifpPXiY+Wzggh7PNLdd4XF4E4XH4OVcp+KHtSZrJH2SUeVzVxOJ\nvlBzV7lRX/vDf2I0/zgUXZkYM+sbyNCsawYf9Xg23b8/BwEPMrP/kdcAr8RbcX6Vub9LzOynwCmS\nPpy+tBz/PbqMIIxrMBkFykNwb+IKST9izsitDYcWhHtz91bU09jyJvFUbIYboRV4i7u2Wa5Zqknr\nL9CH+a32LjWzpuK4jVwBUFdJSxv/hHvkvYbdzA6AtbN8j8Jn7t6rY/kLyWtgsb2Z7S/vKPTppEZu\no6kW76M16iBpI9yb39dq7QHlA+U7JzuZ2YV4je24LEvPMzJ+MMMb7WNtXXPi12bWN6D+FjP7vaS/\nBH5laZKVpK5D8TvJGKk4a4RxDcYiCZmaxuOvk/Kwy0s4EC/VGfK4FhRJe+G/68vxMXWvM7PTa49v\njrd6OwwPSX8Ibwrxwg61Lvis2HOAVSmnWl1ro3RIALiPvA54We1zoFN12UVTgJTrPeYKgAbfoxq/\nMLO+6ER1vZfhQp574u0u+/LCuQ0s1pe0Je6ZbkrySlvIrgM2s1M6HvosXs5yhrwV4o/x2tFf4aVT\nI2i8sp02Tk/76j0QtnijfTR/f66TdDyjM2zrv49rJP0JXjb32fR896I7PZM1UnHWCOMajEvVYm5r\n3Ou4HngrPkC8i6uBb2YIVxaaVbjI5r24AOOjjIp6jsPVzuCG5VhcVHUM8Ni2C5rZQzqe67PMeQmn\nM1cHXP+8tA6wuT7LeyRfAATD71HFLZLOZbQ8o81g74M3ATkbOM/664BzG1gcjhvqrXHhXFcOe9I6\nYIBlZnY9PvHlvriQ6Sf18p66gcuNjmiuuX3F/+BG6zYz26kjH95G0xvto/n7U0UzulTA78B//38L\n7J1Kd86iZeJVonik4iwQxjUYi+rEL++p+ywz+5GkL+Fe3zs7vm0j5sZsVTfm3tmYC8TNeKjqdvMJ\nH82bzdZmdkzyhh4InGpetL/JGM+11kto5Enn0ZJLzSXLe6Ssk9PQe1Tx6Y6vj2Bmj5PPxn0krj7e\nsccA5TawuMHMJG93eV2PKnYh6oDXXtvMfkK7wGutgSso29kR/x15L3C8mX1D0oPJb6lYkdM6cR45\nKmDzDlUrat/zB2A3Sw0xWrzmU8gYqThrhHENJuV/zPv4YmZXSOrzSt+2jvbU5Hd4GPcESf/IfIFU\n1eB8T+BLtZv2OMa1xCvtGlvWpHkjzfUeSzy4G+l/jypaa4SbSHoarjJdiY/O68zxmTci2Qsv1fk6\nfhBo4y2StsC7WZ1Od2P67EHkE1L/uWSV7VQpA/mA8m+kr10ojXaJyKDk96zK4+aqgEdIHnydpte8\nic2VcfWNVJwpwrgGk3KVpLfiqtBd6O85u6LnscXkGbgY5jJJf4UX+tf5eXoNe+M38E3xm0RfKHMh\nqG56e3QtMB963jSIWd4jZR7cMxl9j05Ke2t6KdVQ+GV4Y4wrafTtTeyOq3kPqHuYbbnC9N5vi+e7\nb8Nv/PNmsprZk+TD2J8HfE7S5Wa2f8tzNweRH9nzurvI8QzrBq60bOd6SW/GDfLD6BFJLQBVeD9L\nBZxB8715vKR39VQBzCRhXINJ2Q8fO/d4XGDzlp61O6WPy/DJL7/Bb8CLgubme74h/b/+cN3Tewn+\nOt5qZufIp9hsSWM+ayYl4brq5lwZrO3xFo7fBB6Me0H/J5U8rKVHhNMk24NLHtVl6fO64nvES6mX\nG6X8WmuzAGvp7tR2vcTuZrZHEsScIuklbd+Y2ABPL6xHR6mM5Q8iR2kYeMtDpeUkpWU7z8H/bp6I\nNxg5svD51v6eFeRxS1XAXTS95hiW3kIY12AizPuXvjtz7WuqzyUto3tW6kJRGaWX4sb0FloUpuk1\nvL/2/6/T0WlK0kFmdmyPWrO4xq8yWPLG7Pua2e2S1gP+vfRaDZoeXGknJ+g/LKxP/0i13Outn/Kz\na9Lrbr3ZS/oCblhPBh5tZq1hYTVaNCYFe1s9LMDH5W0STwY+U4ntzOwfC19LadnOzXg7zmoO8K60\nRwC6qIvNcvO4pSrgXGJYegthXIN1RvJ0KrYBcnsNj4XNtYx7NH6zfQQ+hmukDrJWRrERcFfcKN8H\nP91v17jsP6UT+ipJI/1ezexzmTfliqahqQt+1qe7LjSLhgd3hfUPIuhixEtp1F+ujxvwsa+XeBfw\nbdwD+m+6BXGHmHc+2rzLsCaGWjSuxcx2T57cfsARqYzkZDO7Yuh7GTVwryCzbCdRjcX7KXOdzdYa\n1xJVcUEet1QF3EXz9/Z2PKd+L+BjeDqlre57pgjjGqxL6v1mb8WVhYv/pGbfBr4t6R64h/oD3JBW\nj28NIOkj+FSdn8o77Lyr5XKvxOsf781oXnDe1JcxcqknA5cmT/P+9IiAcqjVrq4HnCWpq3Y1m9yS\nk0K+hudo/wI3TFt0rNs8vTfrSfoYcJWZNYcgwECLxhZ+hreyXImPEDxa3rf578g3cNllO4mhsXjj\nqIp787hjqIC7aP4OnYAb7tfhB4RT8G5nM00Y12Bd8lZcKHRXYGPg9XjpzqIi7xL0ArxzzsdwL6ON\nP6/ym2Z2TbpRjmBm5+DioCeZdwm6J97xpk0lXZpLfW8yGtvjzdWHOjcNkVu72seIl6L8ZhOD10vC\nqfvgh4gqCrAFPsx+55bvfTM+TP1s/HfpK8yfMAT5TTaQ9FHcoH4EeG5lECV9izEMXE7ZTqJ3LN6Y\nquKiPG6XCrjEa05sbGZfkHSEmZmk1lF3s0YY12BdUgmf2gQki8nLcIXw/tY/MeQyeZ/U6uT/7Z61\nN0m6ArgBuIekF5nZf9QXlOZS5ZN2DmA0V9g1ODyH3NrVPpqGc1KDXV97D3yuaD0KsBqvmWxjtZn9\nRj7n91Y15vzWyG2yAXBi8+eW2H0By2Zgfii1Got3HXMioLZOTiWq4knzuNUeSw8Vt0raB48o7IZH\npWaeMK7BuuQ6M1vnuRgzax2O3sIBwFPxgeFn1sUpLbwZvwFfI+k+wCfwetI2cnOpH8K7Q+U02s8h\nt3a1ZIJOlsHOvN5jzWw/SS83s648a50fyttubiHvK9z1u5TbZAPgd/JWgBukfW5jZvvY6KDxhSib\nGXmfzOx+XQsblHijvXnc3D2Ocag4AM9zb4lHhfrU3jNDGNdg0Ul1jAAbypv813uadvUhngab6SPe\n/AAAGPpJREFU4PnJnwF3l/QPZtZVKnRHFdIzs58NhMJyc6m/MC8dWiiatat9rfVyJ+jkGuyc6z1T\n3pD+AEkjw8etvf3hi/GRhF/GG0i8COb1c4b8JhvgOfijgKfjYp+2CTOTls3MQz6O7gN4fe8v8D7W\nF7YsLfFGh/K4pWQdKszsajwCEdQI4xqsC6zxcanyKXykWWUQ+sKoN0o6GL/R7YHX7LZSkEu9Mnlk\ndaPwuY61OfwpsK+kp5O8MlxF20buBJ3cZhM513sO3oN4I0a9+1bMGzMc1/JQvZ8zdDTZaDHC4NGU\nMyTtbWZHSmqbNjRpuBXmh4WPwdMUF6d0QBVmb1LijfbmccfYY++holBlP3OEcQ0WHctvejBtlpvZ\nc4eXAd7S7gg8B3kZ3le1lYJc6kaA0j9oUSAXcjpeerQ7fmi4W8/arAk6ltlsIud6KeT4DUnnmlll\nuJBUWnc5YhR6ft+aRhhgtaT7A3dNYc/NW75v0nArzM9LLzOzi9N+L5LU1YSixBvNzePm7rH3UFGo\nsp85wrgGwRzfkbQrowahtcOOmd0A/HPz65I+aWZPbXz5Q2TkUlP+8X54Scp3cIM4CTeZ2dsk7WBm\nL5QPVuiiZIJOG02vp+R6KyWdht+PluHq1NycJOT32W1rYPFyPFR/DG5c2mbBDhq4MRS2d0h6Im68\n9sDbPraR7Y0O5XHH2GPuoWJQZT+LhHENgjn2xPOEFWso70C0WcvXsnKpkg7CBVWb4wZ5B8ZrwVix\nRt6Ld1P5hJ8+z7Vkgk7rc01wvZfi7/0ReKnUYjV/n2eEzWfxVvN4V3Z8X46BK1XYvhAXAf0LHgl4\nUce6bG80I49busdcr7lEZT8zhHENgoSZPWgBLtPmReXmUp+FezH/aWZHy8f5TcIbcWP9YbxJwod7\n1i7EDNRxr3eNmf1c0qZm9kV5l6MSisevFeYLBw1cicJW3mHqKuAZ6fBze1cevkBVDAN53DFUwLle\nc4nKfmYI4xrMPJKONbODJH2bRnhugdSXubnU5emxykB3hQqzMLMLJF0EbIeLkG7qWT7pDNSmgSu5\n3g2SnoJ72gfiJR3z0ORN9uuzdrPzhYUGrldhK2lP4FRJO5vZb/H5wSdKeo6Zfbl5sQJVMeTncXNL\ni3K95hKV/cwQxjUIvGYVvNfxeXhY6zN0zwstoiCXejqe01oh6TPAOZM8r6S/xUOt6wMfTc0XuqYW\nTToDtSmGKbne/vh78xrgMFKv2xZ1b1aT/UIjPJgvLDRwQ2U7bwH2TIYVM/ucpMek1/SIluvlqooh\nP4+bVVpUcKgoUdnPDGFcg5nHUm9VM9tcPkj6yXhN6LV4uGsekjbsEDvNaxafm0s1n7bzn3g7PqPW\nvm9MXo73eD0Xv6l/i+6RgFkTdAqaTWRP5DGz3+Ehc3DjWjGi7rX8Jvslk27q+cKH0p4vLDFwQ2U7\nt5vZlfVvMLPvS2prnwn53ijk53GzSosKDhUlKvuZYfm0NxAES4V043wCczf0PuP2LUnvTrWea7H2\nblDPAh4DXG9mR+M3s7bnf6WZXW5mH8NP/zlN1Pu4I3l+a8zbPnZ64mZ2WtrXKuChZnZmx9KqOcRO\nuEBmp7ZFBdfroy2XWjXZv5m5Jvv/0nju3YHD8VzvVyWtktQlTDsAjxDcDTjLzA4CkLSitmbEwNE9\nuxVcYfsO3DN8SfpYZ7mkkfuuvCVmW/MKSN6opLvLR+m1eqNVHtfMngHsBbzUuruhDe2xojpUbI0f\naI7tWPcdSbtK2kjShhqdfjWzhOcaBHOcj9+4DzezoVKSnYHHAm+QN+//CC7maMtr5uZS/0rSi/Eb\n/T8weRu5L0s6HdhW0nHMeSrzUP4EnaxmEwXX66M57q6vyX6T1kk3Zvbq+qLk1Z7d8v0fZO6QlRtu\nhWGF7Ufw0XRvTfv7U3yaTFe7xkFvtDSPm7HHilyveSFU9nc6wrgGwRxb4CKOfSQdBlxrqfl+EzNb\nLW+SsAbPGR4M7CfpDDNrnvBzc6kvwEtY7gn8jc3vJlTK+4Cn4B74fviQ7C5yG/JnNZsouF4JnU32\n6/8pNMJd1L3m3HArDE+7OVHSjbhwahvgSuCD1tILuUBVXJrHzVUBZx0qFkhlf6cjjGsQzLEZXo6x\nAldAdg4ZkHQUsC/u7f5rqhlcjufsRozrUC5VPh6t8tI2AB4E/Je8k9MkauXTcLHKPwKvxYeQP7Jj\nbe4EndzmEAsxkacZFs5psg+ZRniANVBWNlN7jqGynbMk3dfM3t51kUJvtDSPm6sCzjpUSPov5g8m\naHbCmjnCuAbBHOfiXuUq8+YCffwAWFkPAydvdp4AKuVSjwIuTyKR/wb+urZksZqer8Y95sPN7ExJ\nfR5XbkP+3OYQJRN5ctW9OU32Id8I9zJGuLVEYfs4Se80szs6Hi/xRpdLWm61mcJ9edycPRYeKqqc\n7TI8DN82i3fmCOMaBAkze0jB8i8CB8v74FY38AObHkSiN5daCU/ko+v+FR9J9zG8bGeSEX0b4Mbo\nAkmPpNsYQf4EndzmECUTeXLVvTlN9iHfCPexjPJwa4nC9p7ANZprSbimEaUo8UaL8rhDeyw9VJhZ\nfSDH9yT937bnnTVCLRwE43Fa+rg7Xh+7Rc/aF+CG6LF4LvUrHetOwG96G+Ae59ET7nE/4Ee4wb4n\n8PyetdUEndfjudn3dKzbwcwOM7NzzOxQvD51kuuVqHtzmuxDMsLAjWZ2JG5ERpC0Kn3ct+MaX6DD\nwOERgS5yFbZPBHYB/g6PXDRz+9mqYvMyqE/gedxv4uH/c8zs3WPucd6hAle7v63tYpIOqP07kv42\nmzNDeK5BMB6DTfHHyKVubGZfkHSEmZn6Z8QOYmY/wMPX4IKiPnIn6OQ2hyiZyAN56t6cJvuQZ4Sr\nWbIHS7p3/QEzO8HM3ixpr5JwayJXYXs7/VGKIm80J49bsMfSHG59XOAteNRi5gnjGgTjkdMUvzSX\nequkfYD1JO0GTGRcC8mdoJPbHCJ7Ik+uutfymuxDnhHOmSVbWjYD+WU7J+C1pq/DoxSn4A0/gDJV\ncY2hPG7uHotyuLgyfGe8RzPAfdN+Z5owrkEwHm/Ey1w+jIdeP9JcMEYu9QBcnbkl8Aomr3MtIWuC\njpmdJi/F+XO85rVrSHzJRJ5eda8Kh3LnGGGrzZLFf37bAz+ui3bGNHC5ZTuDUYpCbxSG87i5eyw9\nVHwMuDuuDofx5t3e6QjjGgQFaHQmZjV79Ba8s9MrOr6t10upMLOrU45yB+BiPFS6rsiaoKP85hAl\nE3l61b2W2WS/1Agn/gzPn1+GC8+ONLO1B6USA1eosM2NUuR6o+B53In3OMahYkszaxV4zTIhaAqC\nMnbEw6H/BfxdKmt4Gh5i62JjM/sC7kkYHTdSeQ/i9+OCkl4R0EJjZhfgRuZKXOXbdVBYhedxD8ab\nQ7S2ziu4Hvhr/iLu/VwFdBmkkSb7ePix/pxbm9drfha4X/rZ7EB/G8lDgb82s6cADwYOaVnzuBQW\n7SQpbC+UdI/0pQcC35bUVVt7AC4mGopSVN7o1yV9Tf3ThW7Hfz4nAnsDW427x2RIP2tmO5rZYwe8\n9ask/WnP4zNJGNcgKMDMbkse1chMTNzodpHrpWT1IF4M5BN0zscN4qGSjuhYOtIcgo4JKAXXgwx1\nb+IySR+WdLC8rWPXUO5eI9xgtaVaZfMBAm0/mxwDV6SwNbOr8eENj8NzxF2e6ZCquM6Q2rxojwwc\nKiT9PInCHg/8KP2/+trME2HhIBiP3JmYkJ9LXdB5roXkTtDJbQ5RMpEnt8SmPpT7LEtDuSWtsNEm\n9TmTbiqukPQO3BjtgedfmwyGWylU2Eo6Ae9dfC1puhD+ezTvuuTXPg/lcUtVwEM53JU23EJxZgnP\nNQjG4zl4M4Un4oX4bY0UgLVeyuuBVwIvo1tJeQZ+k/8LLcA810JyJ+g8EzjAfBj2+fj7MMn1ILPE\nxsxWm9nZZva2yrAmPthYmjPppmI/PCf8mPTxRWntRrU1veHWROm0mwfiNcMPM7OHdgiPoKz2eShC\nUrrHIa95nogvmCM81yAYAzP7PS5SGkSZ81yB44DPk3oQm9l3FmSzeeRO0KmaQzydJD7C2yGOe72S\nEpsuRnoQW96km2rt7fhwgSb1WbI5grRShe01wKZ4JKCPktrnoQhJ6R5LvOagQRjXIFh8noWHHP/T\nzI6W1GVovo439j+bueYP64rcCTq5zSEGrzemureN3KEAbfNhc9bmlM1kKWw111jkXsAPJFWD3rvK\nZrJrn5Pa/CDm6k3H2mONoUPFypb887Ke1zJThHENgsUnK5dqZisl7QQ8Gfi8pGvNbN4ggEUid4JO\nbnOIwevlltgsICWTeeprswxcZtlO1VhkQ+APta/35Zmzap9z8riFtbNDh4rL6BdYzTSRcw2CxScr\nlyppZ7xetgpHXt62bpGoJuhsZmZn0t0/N7c5RO71oEzd20aJRzoOuWUzMFy2cxvuqX8YN7AbARvj\nAxHmUaAqhvw87mBpUWLoUHGrmV3V9i/j2nd6wnMNgsUnN5d6Pp4LO9zMcuemLhS5E3Rym0OUTOTp\nVfdKWmVmh0vatyFkqvhCz7XrjBUWHgq3NhhS2O6G19IKD7uCHzzOa7tYgaoY8vO4uZ2chrzmk/ue\nRNKuZtZXY3ynJoxrECw+ubnULfBc5j6SDgOuNbN1FXbbD1fMnowPgW+doGNmF0i6CNgOr/W9qW1d\n7vUSQyU2g0320/piIyxp2+QdVv+XmRm1WbKFBq63bMfMzgHOkfT4zANU5Y12hrTHyOPmlBbl5HCH\n1MJvoyEgmyXCuAbBIlOQS90MF/SsADZhHSozLXOCTmoOcQR+7/iopDVmNq9+Nfd6ae2QujenyT5k\nGuH0Ov4Kf6//VdIr05fXww3CzjY6S3bQwNXIVdj+Ri0tH1vW5XijpXncrD0WHiraWOxw/ZImjGsQ\nLDIpl7oXw7nUc/F87KpUnlJ9/0apZnQpUNIcYlKWQV6T/USuEQa4B26U7p0+LsPDs+9rWZsbboXM\nPtIMDHQv9EZvA/4EOBV4Xnoty/E87i4T7LHkUNHGuN93pyCMaxAsPlm5VDN7SMdD9ZrLaXOHmd2W\nPNY1kvqaQ0xK8+Y81GQ/1whjZl8CviTp9Wb2prYnHyPcCvl1qdeZ2RmS9jazIyWd33i8xBstyuMW\n7LHkUBE0COMaBIvPpLnUpRRey24OsQhUTfZvkrQpnkNty/v1GuEGjwZajSvl4VbIr0sdavmY7Y2O\nkcft3eOYh4o2ltLv7TonjGsQLD6T5lKXUngtt9nEQtC8OY802e/xuHKNMMBGki7EBWer07WfnR4r\nDbdCfl3qUMvHUm8U8vO4Q3sc51DRRtsowpkhjGsQLD5/DLnUXHKbTQwyhro3p8k+5BthgFf1PFZs\n4HLLdszsUkl/wFthPgW4uvF4qTcKA3ncgj1mHSo0OtsYfLbxBsBtZraTmZ2Yue87JWFcg2CRWYBc\n6lIKr1XNIQ43szMlvWiCa2WrexP74X2MH4N7zq+G1kNKrhEGN0L7UPP28Bz5WAYuV2Gr/H7Tud4o\nDOdxc/eYe6jYMX3/e4Hjzewbkh4MvLRjfzNFGNcgmB65RvOy4SXrjJLmEEOUqHtzm+xDvhEG75N8\nOfAAPPd4c8v1SwxcrsI2t990ljeayB3d17vH3ENF9V5KGpltnJ575gnjGgTTY+TmJmkv/G9yOfAe\n4HVmdnqj5nLalDSH6KVE3TtAcypOrhEGWGZmL5b0AWB/oK1XcomBy1XY5s7uzfJGE1mj+wr2mHuo\nKJltPDOEcQ2CpcMq4Nm4YXg43nxhSYlCSppDFFCi7m1jkqk4t0u6Cy40W0P7PXHQwI2hsD0TDz9v\np/7Zvbne6GAed4w95h4qngO8GO/8dCmek595wrgGwfRo3uxvBn4J3G5mv5C0lFTCi0mJuncS2t7P\n9+KK2UtwY9TmueYYuFKF7fOBHwLHApeb2SUd63K90Zw8bukec73mm/H3rQpt74rnu2eamIoTBNOj\nmUu9EVcWnyXppbihnQVG1L30zCztYBLB1xo8HPy3eGj2gy1rcgxc6bSblbiHtx1wnKRPdKy7FLgQ\nuDvujb6757U8Cw/ZX29mR+NGbuw9ku81n413fHoxXtbz4p49zgzhuQbBIjOUS5W0v5mdhBvb7YGd\n8Cb235/Sltc1WepeZTTZH6DNCL8O2MXMrk2K5U/TUMUOhVsTpdNuqpaYj05f+l7HulxVMQzncUtL\ni3K95q0Km0vMBGFcg2DxGcql/jR9/B5zN1lbZ7ubPr3q3sIm+6VG+Ndmdi2Amf1S0jyRT46BG6Ns\nJ3e8YK6qGAbyuKV7zDxUAHxP0jbms3iDRBjXIFh8enOpZnZe+njKNDY3bTLUvVlN9kuNcOJ3ks7D\njdJKPAT61rSv16Y1JQYuV2Gb2xIzV1UM+XncrD0WeM27Az+RdF21VzPbpmefM0EY1yBYfKpc6vEz\nlkudlGoqzmCT/UTJpJuKunf3s441JQYuV2Gb2xIzV1VcH234JOAQSb80s6dNsMesQ4WZ3a9rT7NM\nGNcgWCQilzoxTXVvX5P9EiNc/56caEG2gSNfYdvaErOFXG80O49bsMesQ4WkB+D52G2BXwAvNLML\ne17TTBDGNQgWj1nPpS40fU326/Qa4THINnBkKmx7WmI21+V6o5Cfx81VAeceKo4B9jezi5OBr7QF\nM00Y1yBYJGY9l7oANNW9fU326+Qa4SwKDVx2XWoOBd4o5Odxc/eYe6hYZmYXA5jZRZJu79njzBDG\nNQiCqVKg7u1sst8g1wjn7i/bwBUobHPJ9UYhM4+bu8eCQ8Udkp6IN5LYg/6c9MwQxjUIgqkwhro3\np8k+5BvhXLINXGFdag653ihk5nFz91hwqHghPh/2X/AD0SSTku40hHENgmBalKp7c5rsQ74RzqXE\nwJWU7eSQqyrOzuMW7HHwUCFpczO7CniGpK3wcrPSwQt3SsK4BkEwFcZQ9+Y02Yd8I5xLtoGjrGwn\nh1xVcQm5e+w9VEjaEzhV0s5m9lt8lN2Jkp5jZl9eoL3+0RLGNQiCaZOr7s1psg/5RjiXEgNXUrYz\nSIE3WkLuHocOFW8B9kyGFTP7nKRqHOEjFmHff1SEcQ2CYNrkqnurJvu/pbvJPuQb4SwKDVxJ2c60\nyN3j0KHidjO7sv4FM/u+pNULudk/VsK4BkEwbXLVvYNN9hO5RnjBKSzbmQq5e8w4VCyXtNzM1hpT\nSevRP0x+ZoiRc0EQTJtLcEXvCnwEW9eElZEm+3hbyTYqI/xg4KHAmxd0tz0khe0TyKtLnQoLuMeP\nAGdIepCkTSX9ZfraWQuwzT96wnMNgmDa5Kp7c5rsQ8akm0WkpC51WizIHs3sxPTevgs/HF0JfNDM\nwrgSxjUIgumTq+7NabIP+UZ4MSgp25kWC7ZHMztL0n3N7O0LusM7AWFcgyCYNlnq3oI2krlGeDEo\nKduZFgu9x8dJeqeZ3THxzu5EhHENgmDaLLS6d5q9nBejLnWhWeg93hO4RtKPmZvn2pU3nxnCuAZB\nMG2mpu5daBapLnVBWYQ9PnGBr3enINTCQRBMm6mpe4MF4XZgFXAisDew1XS3szQI4xoEwbTJLbEJ\nliYn4GPrNgAuAI6e7naWBmFcgyCYNr+TdJ6k10o6m6TurRS+wZJnYzP7Ap5rNbycauaJnGsQBNNm\nmureYHJulbQPsJ6k3QjjCsCyNWvWDK8KgiAIghYkbYvPc30A3gzkn83sx9Pd1fQJ4xoEQRBMhKQt\ngbtW/zezn0xxO0uCCAsHQRAEYyPpBOBRwLX4wPs1dPeHnhnCuAZBEAST8EBgBzOLMGiNUAsHQRAE\nk3ANsOm0N7HUiJxrEARBUIykr+Eh4HvhxvWK9FC0PyTCwkEQBMF4PCt93BD4Q+3rm09hL0uOCAsH\nQRAE43AbsBHwYdzAbgRsDBw/zU0tFcJzDYIgCMZhN+AQQHgLRIDVwHlT29ESInKuQRAEwdhIeryZ\nfWba+1hqhOcaBEEQTMJvJB2PN+5fBmxjZvtMeU9TJ3KuQRAEwSS8H/gicHfgKuC6qe5miRDGNQiC\nIJiE68zsDOBGMzsS2HbK+1kShHENgiAIJmG1pPvjowJFlOIAYVyDIAiCyXg5cH/gGOB0fHD6zBNq\n4SAIgmAiJO0A7ABcAlwdfYZDLRwEQRBMgKSDgKfi4eAP4Ub2oGnuaSkQYeEgCIJgEp4FPAa43syO\nBnad8n6WBGFcgyAIgklYjjfwr0LBt01xL0uGCAsHQRAEk3AmcD6wnaTPAOdMeT9LgjCuQRAEwSQ8\nH/ghcCxwuZldMuX9LAlCLRwEQRBMhKSdgCcB+wK/NLOnTXlLUyc81yAIgmBsJO0M7AU8On3pe1Pc\nzpIhjGsQBEEwCecDVwCHx3ScOSIsHARBEIyNpPWB3YF9gF2Aa83s76e7q+kTpThBEATBJGwG3AdY\nAWyCT8aZeSIsHARBEEzCuXj5zSozu3Tam1kqRFg4CIIgCBaYCAsHQRAEwQITxjUIgiAIFpgwrkEQ\nBEGwwIRxDYIgCIIF5n8BwhQTZ5Qwg8kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x124ab9668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "corr = data.corr()\n",
    "sns.heatmap(corr, \n",
    "            xticklabels=corr.columns.values,\n",
    "            yticklabels=corr.columns.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#save the modified csv\n",
    "data.to_csv('coffeeBotData-Cleaned.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(['type', 'date','zipcode'],axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "final features\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 624 entries, CFDB9498-809D-4232-8281-9E7CA354D59B to 37ED8D81-924D-4EDE-8A5F-CB3875747474\n",
      "Data columns (total 25 columns):\n",
      "clouds                      624 non-null int64\n",
      "humidity                    624 non-null int64\n",
      "temp                        624 non-null float64\n",
      "visibility                  624 non-null float64\n",
      "windSpeed                   624 non-null float64\n",
      "day_of_week                 624 non-null int64\n",
      "is_weekend_False            624 non-null uint8\n",
      "is_weekend_True             624 non-null uint8\n",
      "month                       624 non-null int64\n",
      "season_fall                 624 non-null uint8\n",
      "season_spring               624 non-null uint8\n",
      "season_summer               624 non-null uint8\n",
      "season_winter               624 non-null uint8\n",
      "part_of_day_afternoon       624 non-null uint8\n",
      "part_of_day_evening         624 non-null uint8\n",
      "part_of_day_morning         624 non-null uint8\n",
      "part_of_day_night           624 non-null uint8\n",
      "weatherCond_Clear           624 non-null uint8\n",
      "weatherCond_Clouds          624 non-null uint8\n",
      "weatherCond_Fog             624 non-null uint8\n",
      "weatherCond_Haze            624 non-null uint8\n",
      "weatherCond_Rain            624 non-null uint8\n",
      "weatherCond_Smoke           624 non-null uint8\n",
      "weatherCond_Snow            624 non-null uint8\n",
      "weatherCond_Thunderstorm    624 non-null uint8\n",
      "dtypes: float64(3), int64(4), uint8(18)\n",
      "memory usage: 70.0+ KB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "print(\"final features\")\n",
    "print(data.info())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data (624, 25)\n",
      "labels (624,)\n"
     ]
    }
   ],
   "source": [
    "print(\"data\",data.shape)\n",
    "print(\"labels\", labels.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[False False False False False False False False False  True False False\n",
      "  True False False False False False False  True False False False  True\n",
      " False]\n",
      "[21 20 15 19 17 22  7  8 14  1  5 18  1 13  6 12  4  9 11  1  3 10 16  1\n",
      "  2]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_selection import RFE\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "model = LogisticRegression()\n",
    "rfe = RFE(model, 4)\n",
    "rfe = rfe.fit(data, labels)\n",
    "print(rfe.support_)\n",
    "print(rfe.ranking_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data#data[['lat','lon', 'temp', 'windSpeed']]\n",
    "y = labels\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 263,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic regression accuracy: 0.814\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import metrics\n",
    "logreg = LogisticRegression()\n",
    "logreg.fit(X_train, y_train)\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "print('Logistic regression accuracy: {:.3f}'.format(accuracy_score(y_test, logreg.predict(X_test))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)"
      ]
     },
     "execution_count": 264,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "reg = LinearRegression()\n",
    "reg.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic regression accuracy: 0.814\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "print('Logistic regression accuracy: {:.3f}'.format(accuracy_score(y_test, logreg.predict(X_test))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Accuracy: 0.809\n"
     ]
    }
   ],
   "source": [
    "#random forest\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "rf = RandomForestClassifier()\n",
    "rf.fit(X_train, y_train)\n",
    "print('Random Forest Accuracy: {:.3f}'.format(accuracy_score(y_test, rf.predict(X_test))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Support vector machine accuracy: 0.670\n"
     ]
    }
   ],
   "source": [
    "#SVM\n",
    "from sklearn.svm import SVC\n",
    "svc = SVC()\n",
    "svc.fit(X_train, y_train)\n",
    "print('Support vector machine accuracy: {:.3f}'.format(accuracy_score(y_test, svc.predict(X_test))))              "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6260903794678003\n",
      "0.6627145331443867\n"
     ]
    }
   ],
   "source": [
    "from sklearn import model_selection as ms\n",
    "from sklearn.model_selection import cross_val_score\n",
    "scoring = 'accuracy'\n",
    "\n",
    "\n",
    "kfoldMean = ms.cross_val_score(svc, X_train, y_train, cv = ms.KFold(), scoring=scoring).mean()\n",
    "Kfold_shuggle = ms.cross_val_score(svc, X_train, y_train, cv = ms.KFold(shuffle = True), scoring=scoring).mean()\n",
    "print(kfoldMean)\n",
    "print(Kfold_shuggle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#save mode\n",
    "import pickle\n",
    "# filename = 'svc.sav'\n",
    "filename = 'models/rf.sav'\n",
    "pickle.dump(rf, open(filename, 'wb'), protocol=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "#save mode\n",
    "import pickle\n",
    "filename = 'models/logreg.sav'\n",
    "# filename = 'rf.sav'\n",
    "pickle.dump(logreg, open(filename, 'wb'), protocol=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "file = open(filename, 'rb')\n",
    "svc = pickle.load(file)\n",
    "svc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}