2. EDA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import scipy.stats as stats\n",
    "from eda_helpers import normality_check, count_outliers"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exploratory Data Analysis (EDA)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Load train set and sample the dataset to a manageable size if necessary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Billitonstraat 4 3, Amsterdam</td>\n",
       "      <td>1094 BC</td>\n",
       "      <td>385000.0</td>\n",
       "      <td>49</td>\n",
       "      <td>3</td>\n",
       "      <td>4.938408</td>\n",
       "      <td>52.365552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Zuidelijke Wandelweg 151 PP, Amsterdam</td>\n",
       "      <td>1079 RK</td>\n",
       "      <td>930000.0</td>\n",
       "      <td>123</td>\n",
       "      <td>4</td>\n",
       "      <td>4.894206</td>\n",
       "      <td>52.338220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Tweede Keucheniusstraat 8 3, Amsterdam</td>\n",
       "      <td>1051 VR</td>\n",
       "      <td>500000.0</td>\n",
       "      <td>70</td>\n",
       "      <td>3</td>\n",
       "      <td>4.869195</td>\n",
       "      <td>52.380747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>De Bazelhof 24, Amsterdam</td>\n",
       "      <td>1067 LZ</td>\n",
       "      <td>400000.0</td>\n",
       "      <td>107</td>\n",
       "      <td>6</td>\n",
       "      <td>4.795810</td>\n",
       "      <td>52.376916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Bos en Lommerweg 360, Amsterdam</td>\n",
       "      <td>1061 DJ</td>\n",
       "      <td>475000.0</td>\n",
       "      <td>98</td>\n",
       "      <td>4</td>\n",
       "      <td>4.838520</td>\n",
       "      <td>52.379072</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  Address      Zip     Price  Area  Room   \n",
       "0           Billitonstraat 4 3, Amsterdam  1094 BC  385000.0    49     3  \\\n",
       "1  Zuidelijke Wandelweg 151 PP, Amsterdam  1079 RK  930000.0   123     4   \n",
       "2  Tweede Keucheniusstraat 8 3, Amsterdam  1051 VR  500000.0    70     3   \n",
       "3               De Bazelhof 24, Amsterdam  1067 LZ  400000.0   107     6   \n",
       "4         Bos en Lommerweg 360, Amsterdam  1061 DJ  475000.0    98     4   \n",
       "\n",
       "        Lon        Lat  \n",
       "0  4.938408  52.365552  \n",
       "1  4.894206  52.338220  \n",
       "2  4.869195  52.380747  \n",
       "3  4.795810  52.376916  \n",
       "4  4.838520  52.379072  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('data/data_train.csv')\n",
    "data.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. For supervised learning tasks, identify the target attribute"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "TARGET = 'Price'"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Study each attribute and its characteristics:"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. **Name**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Address', 'Zip', 'Price', 'Area', 'Room', 'Lon', 'Lat'], dtype='object')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. **Type (categorical, int/float, bounded/unbounded, text, structured, etc.);**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Price</th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>data_type</th>\n",
       "      <td>continuous</td>\n",
       "      <td>text</td>\n",
       "      <td>text</td>\n",
       "      <td>continuous</td>\n",
       "      <td>descrete</td>\n",
       "      <td>continuous</td>\n",
       "      <td>continuous</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Price Address   Zip        Area      Room         Lon   \n",
       "data_type  continuous    text  text  continuous  descrete  continuous  \\\n",
       "\n",
       "                  Lat  \n",
       "data_type  continuous  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_info = pd.DataFrame.from_dict(\n",
    "    {\n",
    "        \"Price\": \"continuous\",  # <- This is our TARGET\n",
    "        \"Address\": \"text\",      # assumption: might be non-informative, because there is also longtitude and latitude data present, which might appear to be more infirmative\n",
    "        \"Zip\": \"text\",          # because zip contains not only numbers, but also letters\n",
    "        \"Area\": \"continuous\",\n",
    "        \"Room\": \"descrete\",\n",
    "        \"Lon\": \"continuous\",\n",
    "        \"Lat\": \"continuous\",\n",
    "    },\n",
    "    orient=\"index\",\n",
    "    columns=[\"data_type\"],\n",
    ")\n",
    "display(data_info.transpose())"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. **Percentage of missing values;**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "o 3 (2.84%) missing values\n"
     ]
    }
   ],
   "source": [
    "print(\n",
    "    f\"o {data.isna().sum().sum()} ({data.isna().sum().sum()*100/data.shape[0]*data.shape[1]:.2f}%) missing values\"\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check if missing values are represented in a different way as in the description of the dataset it is stated that there are missing values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Uncomment to inspect all columns\n",
    "\n",
    "# for column in data.columns:\n",
    "#     print(f\"{column}: {data[column].unique()}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is no missing data, that is represented in a different way."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>missing values</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Address  Zip  Price  Area  Room  Lon  Lat\n",
       "missing values        0    0      3     0     0    0    0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_missing_value = data.isna().sum().to_frame(name=\"missing values\")\n",
    "display(df_missing_value.transpose())"
   ]
  },
  {
   "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>data_type</th>\n",
       "      <th>missing values</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Price</th>\n",
       "      <td>continuous</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Address</th>\n",
       "      <td>text</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Zip</th>\n",
       "      <td>text</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Area</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Room</th>\n",
       "      <td>descrete</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lon</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lat</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          data_type  missing values\n",
       "Price    continuous               3\n",
       "Address        text               0\n",
       "Zip            text               0\n",
       "Area     continuous               0\n",
       "Room       descrete               0\n",
       "Lon      continuous               0\n",
       "Lat      continuous               0"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_info = pd.concat([data_info, df_missing_value], axis=1)\n",
    "data_info"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>Zeeburgerkade 760, Amsterdam</td>\n",
       "      <td>1019 HT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>107</td>\n",
       "      <td>3</td>\n",
       "      <td>4.945022</td>\n",
       "      <td>52.369244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>418</th>\n",
       "      <td>Suikerplein 16, Amsterdam</td>\n",
       "      <td>1013 CK</td>\n",
       "      <td>NaN</td>\n",
       "      <td>81</td>\n",
       "      <td>3</td>\n",
       "      <td>4.880976</td>\n",
       "      <td>52.389623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>529</th>\n",
       "      <td>Haarlemmerweg 705, Amsterdam</td>\n",
       "      <td>1067 HP</td>\n",
       "      <td>NaN</td>\n",
       "      <td>366</td>\n",
       "      <td>12</td>\n",
       "      <td>4.787874</td>\n",
       "      <td>52.383877</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          Address      Zip  Price  Area  Room       Lon   \n",
       "110  Zeeburgerkade 760, Amsterdam  1019 HT    NaN   107     3  4.945022  \\\n",
       "418     Suikerplein 16, Amsterdam  1013 CK    NaN    81     3  4.880976   \n",
       "529  Haarlemmerweg 705, Amsterdam  1067 HP    NaN   366    12  4.787874   \n",
       "\n",
       "           Lat  \n",
       "110  52.369244  \n",
       "418  52.389623  \n",
       "529  52.383877  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[data.isna().any(axis=1)]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. **Check if there are duplicates**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "o 1 (0.001%) duplicate values\n"
     ]
    }
   ],
   "source": [
    "print(\n",
    "    f\"o {data.duplicated().sum()} ({data.duplicated().sum()/len(data):.3f}%) duplicate values\"\n",
    ")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. **Check outliers**"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Let's check the **distribution** of the **numerical** (continuous and descrete) independent variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Price', 'Area', 'Room', 'Lon', 'Lat']"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list_numerical = []\n",
    "for col in data.columns:\n",
    "    if data_info['data_type'][col] == \"continuous\" or data_info['data_type'][col] == \"descrete\":\n",
    "        list_numerical.append(col)\n",
    "\n",
    "list_numerical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x300 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 5, figsize=(18, 3))\n",
    "\n",
    "for i, col in enumerate(list_numerical):\n",
    "    plot = sns.histplot(ax=axes.flatten()[i], x = data[col], bins = 30)\n",
    "\n",
    "plt.tight_layout() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>data_type</th>\n",
       "      <th>missing values</th>\n",
       "      <th>distribution</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Price</th>\n",
       "      <td>continuous</td>\n",
       "      <td>3</td>\n",
       "      <td>not normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Address</th>\n",
       "      <td>text</td>\n",
       "      <td>0</td>\n",
       "      <td>not applicable</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Zip</th>\n",
       "      <td>text</td>\n",
       "      <td>0</td>\n",
       "      <td>not applicable</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Area</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "      <td>not normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Room</th>\n",
       "      <td>descrete</td>\n",
       "      <td>0</td>\n",
       "      <td>not normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lon</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "      <td>normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lat</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "      <td>not normal</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          data_type  missing values    distribution\n",
       "Price    continuous               3      not normal\n",
       "Address        text               0  not applicable\n",
       "Zip            text               0  not applicable\n",
       "Area     continuous               0      not normal\n",
       "Room       descrete               0      not normal\n",
       "Lon      continuous               0          normal\n",
       "Lat      continuous               0      not normal"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a 'Distribution' column with default value\n",
    "data_info['distribution'] = 'not applicable'\n",
    "normality_check(data, data_info, data.columns)\n",
    "data_info\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>data_type</th>\n",
       "      <th>missing values</th>\n",
       "      <th>distribution</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Price</th>\n",
       "      <td>continuous</td>\n",
       "      <td>3</td>\n",
       "      <td>right_skewed</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Address</th>\n",
       "      <td>text</td>\n",
       "      <td>0</td>\n",
       "      <td>not applicable</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Zip</th>\n",
       "      <td>text</td>\n",
       "      <td>0</td>\n",
       "      <td>not applicable</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Area</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "      <td>right_skewed</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Room</th>\n",
       "      <td>descrete</td>\n",
       "      <td>0</td>\n",
       "      <td>right_skewed</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lon</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "      <td>normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lat</th>\n",
       "      <td>continuous</td>\n",
       "      <td>0</td>\n",
       "      <td>heavy_tailed</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          data_type  missing values    distribution\n",
       "Price    continuous               3    right_skewed\n",
       "Address        text               0  not applicable\n",
       "Zip            text               0  not applicable\n",
       "Area     continuous               0    right_skewed\n",
       "Room       descrete               0    right_skewed\n",
       "Lon      continuous               0          normal\n",
       "Lat      continuous               0    heavy_tailed"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Specify the type on non-Gaussian distributions:\n",
    "\n",
    "data_info.loc[['Price', 'Area', 'Room'], 'distribution'] = 'right_skewed'\n",
    "data_info.loc['Lat', 'distribution'] = 'heavy_tailed'\n",
    "\n",
    "data_info"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Price, Area, and Room are right scewed destributions, that is why simple ± 3std rule can not be applied to detect outliers. Instead, we can apply Q3 +1.5(IQR) rule. Here we will use a custom function, that defines a type of distribution and applies correct formula to detect outliers:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Impossible to define outliers for Address data: data is not in [continuous, descrete]\n",
      "Impossible to define outliers for Zip data: data is not in [continuous, descrete]\n",
      "o 61 datapoints with Price > 1196250.0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Prinsengracht 416 A, Amsterdam</td>\n",
       "      <td>1016 JC</td>\n",
       "      <td>1275000.0</td>\n",
       "      <td>165</td>\n",
       "      <td>5</td>\n",
       "      <td>4.882122</td>\n",
       "      <td>52.366395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Noordermarkt 32 A, Amsterdam</td>\n",
       "      <td>1015 MZ</td>\n",
       "      <td>1200000.0</td>\n",
       "      <td>105</td>\n",
       "      <td>4</td>\n",
       "      <td>4.886034</td>\n",
       "      <td>52.379546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Singel 318, Amsterdam</td>\n",
       "      <td>1016 AE</td>\n",
       "      <td>4900000.0</td>\n",
       "      <td>623</td>\n",
       "      <td>13</td>\n",
       "      <td>4.887869</td>\n",
       "      <td>52.371039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Pieter Cornelisz. Hooftstraat 169 B, Amsterdam</td>\n",
       "      <td>1071 BW</td>\n",
       "      <td>1695000.0</td>\n",
       "      <td>117</td>\n",
       "      <td>3</td>\n",
       "      <td>4.876144</td>\n",
       "      <td>52.358928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>Reguliersgracht 87, Amsterdam</td>\n",
       "      <td>1017 LN</td>\n",
       "      <td>1450000.0</td>\n",
       "      <td>124</td>\n",
       "      <td>5</td>\n",
       "      <td>4.896425</td>\n",
       "      <td>52.360892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>683</th>\n",
       "      <td>Colenbranderhof 30, Amsterdam</td>\n",
       "      <td>1065 BG</td>\n",
       "      <td>1650000.0</td>\n",
       "      <td>235</td>\n",
       "      <td>7</td>\n",
       "      <td>4.820848</td>\n",
       "      <td>52.358631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>686</th>\n",
       "      <td>Valeriusstraat 148 Hs, Amsterdam</td>\n",
       "      <td>1075 GE</td>\n",
       "      <td>1775000.0</td>\n",
       "      <td>177</td>\n",
       "      <td>9</td>\n",
       "      <td>4.865595</td>\n",
       "      <td>52.353408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>687</th>\n",
       "      <td>Keizersgracht 171 F, Amsterdam</td>\n",
       "      <td>1016 DP</td>\n",
       "      <td>1625000.0</td>\n",
       "      <td>199</td>\n",
       "      <td>6</td>\n",
       "      <td>4.885830</td>\n",
       "      <td>52.374742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>688</th>\n",
       "      <td>Prinseneiland 545, Amsterdam</td>\n",
       "      <td>1013 EC</td>\n",
       "      <td>1950000.0</td>\n",
       "      <td>258</td>\n",
       "      <td>4</td>\n",
       "      <td>4.887444</td>\n",
       "      <td>52.385346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>690</th>\n",
       "      <td>James Bradleystraat 11, Amsterdam</td>\n",
       "      <td>1086 ZM</td>\n",
       "      <td>1200000.0</td>\n",
       "      <td>231</td>\n",
       "      <td>8</td>\n",
       "      <td>4.979592</td>\n",
       "      <td>52.360578</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>61 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            Address      Zip      Price  Area   \n",
       "7                    Prinsengracht 416 A, Amsterdam  1016 JC  1275000.0   165  \\\n",
       "9                      Noordermarkt 32 A, Amsterdam  1015 MZ  1200000.0   105   \n",
       "12                            Singel 318, Amsterdam  1016 AE  4900000.0   623   \n",
       "22   Pieter Cornelisz. Hooftstraat 169 B, Amsterdam  1071 BW  1695000.0   117   \n",
       "33                    Reguliersgracht 87, Amsterdam  1017 LN  1450000.0   124   \n",
       "..                                              ...      ...        ...   ...   \n",
       "683                   Colenbranderhof 30, Amsterdam  1065 BG  1650000.0   235   \n",
       "686                Valeriusstraat 148 Hs, Amsterdam  1075 GE  1775000.0   177   \n",
       "687                  Keizersgracht 171 F, Amsterdam  1016 DP  1625000.0   199   \n",
       "688                    Prinseneiland 545, Amsterdam  1013 EC  1950000.0   258   \n",
       "690               James Bradleystraat 11, Amsterdam  1086 ZM  1200000.0   231   \n",
       "\n",
       "     Room       Lon        Lat  \n",
       "7       5  4.882122  52.366395  \n",
       "9       4  4.886034  52.379546  \n",
       "12     13  4.887869  52.371039  \n",
       "22      3  4.876144  52.358928  \n",
       "33      5  4.896425  52.360892  \n",
       "..    ...       ...        ...  \n",
       "683     7  4.820848  52.358631  \n",
       "686     9  4.865595  52.353408  \n",
       "687     6  4.885830  52.374742  \n",
       "688     4  4.887444  52.385346  \n",
       "690     8  4.979592  52.360578  \n",
       "\n",
       "[61 rows x 7 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "o 41 datapoints with Area > 190.0\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Noorderakerweg 124, Amsterdam</td>\n",
       "      <td>1069 LW</td>\n",
       "      <td>950000.0</td>\n",
       "      <td>226</td>\n",
       "      <td>7</td>\n",
       "      <td>4.776234</td>\n",
       "      <td>52.349215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Singel 318, Amsterdam</td>\n",
       "      <td>1016 AE</td>\n",
       "      <td>4900000.0</td>\n",
       "      <td>623</td>\n",
       "      <td>13</td>\n",
       "      <td>4.887869</td>\n",
       "      <td>52.371039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>Ruysdaelkade 109, Amsterdam</td>\n",
       "      <td>1072 AN</td>\n",
       "      <td>1650000.0</td>\n",
       "      <td>232</td>\n",
       "      <td>9</td>\n",
       "      <td>4.887183</td>\n",
       "      <td>52.355020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>Gerrit van der Veenstraat 145 B*, Amsterdam</td>\n",
       "      <td>1077 DZ</td>\n",
       "      <td>1625000.0</td>\n",
       "      <td>257</td>\n",
       "      <td>9</td>\n",
       "      <td>4.869761</td>\n",
       "      <td>52.348636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>Utrechtsedwarsstraat 4, Amsterdam</td>\n",
       "      <td>1017 WG</td>\n",
       "      <td>1700000.0</td>\n",
       "      <td>194</td>\n",
       "      <td>6</td>\n",
       "      <td>4.896542</td>\n",
       "      <td>52.360937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>Willem Augustinstraat 179, Amsterdam</td>\n",
       "      <td>1061 MJ</td>\n",
       "      <td>975000.0</td>\n",
       "      <td>221</td>\n",
       "      <td>5</td>\n",
       "      <td>4.838629</td>\n",
       "      <td>52.374910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>Borneokade 70, Amsterdam</td>\n",
       "      <td>1019 AW</td>\n",
       "      <td>1750000.0</td>\n",
       "      <td>246</td>\n",
       "      <td>8</td>\n",
       "      <td>4.909819</td>\n",
       "      <td>52.368653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>Sophialaan 45, Amsterdam</td>\n",
       "      <td>1075 BM</td>\n",
       "      <td>3500000.0</td>\n",
       "      <td>374</td>\n",
       "      <td>7</td>\n",
       "      <td>4.857608</td>\n",
       "      <td>52.352366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>Vondelstraat 51 hs, Amsterdam</td>\n",
       "      <td>1054 GJ</td>\n",
       "      <td>3925000.0</td>\n",
       "      <td>319</td>\n",
       "      <td>7</td>\n",
       "      <td>4.875471</td>\n",
       "      <td>52.361571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>Singel 146 A, Amsterdam</td>\n",
       "      <td>1015 AG</td>\n",
       "      <td>3500000.0</td>\n",
       "      <td>348</td>\n",
       "      <td>8</td>\n",
       "      <td>4.889816</td>\n",
       "      <td>52.375532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>172</th>\n",
       "      <td>Valeriusstraat 193, Amsterdam</td>\n",
       "      <td>1075 EW</td>\n",
       "      <td>1698000.0</td>\n",
       "      <td>205</td>\n",
       "      <td>5</td>\n",
       "      <td>4.860637</td>\n",
       "      <td>52.351894</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>Broekergouw 4, Amsterdam</td>\n",
       "      <td>1027 AH</td>\n",
       "      <td>1695000.0</td>\n",
       "      <td>467</td>\n",
       "      <td>14</td>\n",
       "      <td>4.970830</td>\n",
       "      <td>52.412038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>Jacob van Lennepstraat 11 I/ II, Amsterdam</td>\n",
       "      <td>1053 HA</td>\n",
       "      <td>1095000.0</td>\n",
       "      <td>204</td>\n",
       "      <td>6</td>\n",
       "      <td>4.876183</td>\n",
       "      <td>52.366972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>Levantplein 9, Amsterdam</td>\n",
       "      <td>1019 MA</td>\n",
       "      <td>1450000.0</td>\n",
       "      <td>273</td>\n",
       "      <td>5</td>\n",
       "      <td>4.942112</td>\n",
       "      <td>52.376819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>203</th>\n",
       "      <td>Prinsengracht 483 A, Amsterdam</td>\n",
       "      <td>1016 HP</td>\n",
       "      <td>1750000.0</td>\n",
       "      <td>212</td>\n",
       "      <td>4</td>\n",
       "      <td>4.883097</td>\n",
       "      <td>52.369689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>207</th>\n",
       "      <td>Olympiakade 12, Amsterdam</td>\n",
       "      <td>1076 XP</td>\n",
       "      <td>2750000.0</td>\n",
       "      <td>309</td>\n",
       "      <td>8</td>\n",
       "      <td>4.862413</td>\n",
       "      <td>52.349528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>258</th>\n",
       "      <td>Westeinde 16, Amsterdam</td>\n",
       "      <td>1017 ZP</td>\n",
       "      <td>4550000.0</td>\n",
       "      <td>497</td>\n",
       "      <td>13</td>\n",
       "      <td>4.898620</td>\n",
       "      <td>52.358798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>278</th>\n",
       "      <td>Keizersgracht 544, Amsterdam</td>\n",
       "      <td>1017 EL</td>\n",
       "      <td>5950000.0</td>\n",
       "      <td>394</td>\n",
       "      <td>10</td>\n",
       "      <td>4.887378</td>\n",
       "      <td>52.365078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>Raphaëlstraat 5, Amsterdam</td>\n",
       "      <td>1077 PN</td>\n",
       "      <td>1700000.0</td>\n",
       "      <td>203</td>\n",
       "      <td>5</td>\n",
       "      <td>4.869966</td>\n",
       "      <td>52.349904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>323</th>\n",
       "      <td>Kromme Waal 18, Amsterdam</td>\n",
       "      <td>1011 BS</td>\n",
       "      <td>1500000.0</td>\n",
       "      <td>194</td>\n",
       "      <td>9</td>\n",
       "      <td>4.902614</td>\n",
       "      <td>52.374848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>368</th>\n",
       "      <td>Tweede Jan van der Heijdenstraat 13, Amsterdam</td>\n",
       "      <td>1073 VE</td>\n",
       "      <td>2650000.0</td>\n",
       "      <td>387</td>\n",
       "      <td>10</td>\n",
       "      <td>4.899350</td>\n",
       "      <td>52.355179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>382</th>\n",
       "      <td>Klaprozenweg 73 A, Amsterdam</td>\n",
       "      <td>1032 KK</td>\n",
       "      <td>1100011.0</td>\n",
       "      <td>207</td>\n",
       "      <td>5</td>\n",
       "      <td>4.900480</td>\n",
       "      <td>52.400948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>393</th>\n",
       "      <td>Uilengouw 2, Amsterdam</td>\n",
       "      <td>1027 EZ</td>\n",
       "      <td>899000.0</td>\n",
       "      <td>258</td>\n",
       "      <td>1</td>\n",
       "      <td>4.967897</td>\n",
       "      <td>52.410672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>411</th>\n",
       "      <td>Nieuwe Leliestraat 11, Amsterdam</td>\n",
       "      <td>1015 SH</td>\n",
       "      <td>2450000.0</td>\n",
       "      <td>259</td>\n",
       "      <td>8</td>\n",
       "      <td>4.883278</td>\n",
       "      <td>52.375756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>422</th>\n",
       "      <td>IJsselmeerstraat 52, Amsterdam</td>\n",
       "      <td>1024 ML</td>\n",
       "      <td>799000.0</td>\n",
       "      <td>230</td>\n",
       "      <td>8</td>\n",
       "      <td>4.961381</td>\n",
       "      <td>52.389087</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>423</th>\n",
       "      <td>Dorpsstraat Holysloot 62, Amsterdam</td>\n",
       "      <td>1028 BE</td>\n",
       "      <td>1495000.0</td>\n",
       "      <td>214</td>\n",
       "      <td>5</td>\n",
       "      <td>5.029122</td>\n",
       "      <td>52.413550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>431</th>\n",
       "      <td>Bosplaat 80, Amsterdam</td>\n",
       "      <td>1025 AT</td>\n",
       "      <td>1000000.0</td>\n",
       "      <td>220</td>\n",
       "      <td>7</td>\n",
       "      <td>4.948314</td>\n",
       "      <td>52.400138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>462</th>\n",
       "      <td>Geschutswerf 81, Amsterdam</td>\n",
       "      <td>1018 AW</td>\n",
       "      <td>1475000.0</td>\n",
       "      <td>220</td>\n",
       "      <td>5</td>\n",
       "      <td>4.920481</td>\n",
       "      <td>52.366984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466</th>\n",
       "      <td>Badhuiskade 353, Amsterdam</td>\n",
       "      <td>1031 KV</td>\n",
       "      <td>3680000.0</td>\n",
       "      <td>374</td>\n",
       "      <td>4</td>\n",
       "      <td>4.905695</td>\n",
       "      <td>52.385321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>517</th>\n",
       "      <td>Plantage Muidergracht 105 A, Amsterdam</td>\n",
       "      <td>1018 TP</td>\n",
       "      <td>2250000.0</td>\n",
       "      <td>250</td>\n",
       "      <td>8</td>\n",
       "      <td>4.915246</td>\n",
       "      <td>52.364132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>529</th>\n",
       "      <td>Haarlemmerweg 705, Amsterdam</td>\n",
       "      <td>1067 HP</td>\n",
       "      <td>NaN</td>\n",
       "      <td>366</td>\n",
       "      <td>12</td>\n",
       "      <td>4.787874</td>\n",
       "      <td>52.383877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>537</th>\n",
       "      <td>Parosstraat 6, Amsterdam</td>\n",
       "      <td>1060 LP</td>\n",
       "      <td>1500000.0</td>\n",
       "      <td>255</td>\n",
       "      <td>7</td>\n",
       "      <td>4.790343</td>\n",
       "      <td>52.345566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>552</th>\n",
       "      <td>Den Texstraat 35 2, Amsterdam</td>\n",
       "      <td>1017 XX</td>\n",
       "      <td>1320000.0</td>\n",
       "      <td>230</td>\n",
       "      <td>8</td>\n",
       "      <td>4.894964</td>\n",
       "      <td>52.358917</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>625</th>\n",
       "      <td>Herengracht 116 F+K, Amsterdam</td>\n",
       "      <td>1015 BT</td>\n",
       "      <td>1595000.0</td>\n",
       "      <td>199</td>\n",
       "      <td>6</td>\n",
       "      <td>4.888459</td>\n",
       "      <td>52.375994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>635</th>\n",
       "      <td>Bellamystraat 13 15, Amsterdam</td>\n",
       "      <td>1053 BE</td>\n",
       "      <td>1250000.0</td>\n",
       "      <td>202</td>\n",
       "      <td>6</td>\n",
       "      <td>4.866447</td>\n",
       "      <td>52.367118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>649</th>\n",
       "      <td>Oudezijds Voorburgwal 126 A+B, Amsterdam</td>\n",
       "      <td>1012 GH</td>\n",
       "      <td>1400000.0</td>\n",
       "      <td>213</td>\n",
       "      <td>7</td>\n",
       "      <td>4.897256</td>\n",
       "      <td>52.373343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>672</th>\n",
       "      <td>Baarsjesweg 313 H, Amsterdam</td>\n",
       "      <td>1058 AH</td>\n",
       "      <td>1499000.0</td>\n",
       "      <td>245</td>\n",
       "      <td>8</td>\n",
       "      <td>4.854662</td>\n",
       "      <td>52.358204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>683</th>\n",
       "      <td>Colenbranderhof 30, Amsterdam</td>\n",
       "      <td>1065 BG</td>\n",
       "      <td>1650000.0</td>\n",
       "      <td>235</td>\n",
       "      <td>7</td>\n",
       "      <td>4.820848</td>\n",
       "      <td>52.358631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>687</th>\n",
       "      <td>Keizersgracht 171 F, Amsterdam</td>\n",
       "      <td>1016 DP</td>\n",
       "      <td>1625000.0</td>\n",
       "      <td>199</td>\n",
       "      <td>6</td>\n",
       "      <td>4.885830</td>\n",
       "      <td>52.374742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>688</th>\n",
       "      <td>Prinseneiland 545, Amsterdam</td>\n",
       "      <td>1013 EC</td>\n",
       "      <td>1950000.0</td>\n",
       "      <td>258</td>\n",
       "      <td>4</td>\n",
       "      <td>4.887444</td>\n",
       "      <td>52.385346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>690</th>\n",
       "      <td>James Bradleystraat 11, Amsterdam</td>\n",
       "      <td>1086 ZM</td>\n",
       "      <td>1200000.0</td>\n",
       "      <td>231</td>\n",
       "      <td>8</td>\n",
       "      <td>4.979592</td>\n",
       "      <td>52.360578</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            Address      Zip      Price  Area   \n",
       "11                    Noorderakerweg 124, Amsterdam  1069 LW   950000.0   226  \\\n",
       "12                            Singel 318, Amsterdam  1016 AE  4900000.0   623   \n",
       "35                      Ruysdaelkade 109, Amsterdam  1072 AN  1650000.0   232   \n",
       "39      Gerrit van der Veenstraat 145 B*, Amsterdam  1077 DZ  1625000.0   257   \n",
       "46                Utrechtsedwarsstraat 4, Amsterdam  1017 WG  1700000.0   194   \n",
       "60             Willem Augustinstraat 179, Amsterdam  1061 MJ   975000.0   221   \n",
       "74                         Borneokade 70, Amsterdam  1019 AW  1750000.0   246   \n",
       "85                         Sophialaan 45, Amsterdam  1075 BM  3500000.0   374   \n",
       "136                   Vondelstraat 51 hs, Amsterdam  1054 GJ  3925000.0   319   \n",
       "144                         Singel 146 A, Amsterdam  1015 AG  3500000.0   348   \n",
       "172                   Valeriusstraat 193, Amsterdam  1075 EW  1698000.0   205   \n",
       "173                        Broekergouw 4, Amsterdam  1027 AH  1695000.0   467   \n",
       "177      Jacob van Lennepstraat 11 I/ II, Amsterdam  1053 HA  1095000.0   204   \n",
       "199                        Levantplein 9, Amsterdam  1019 MA  1450000.0   273   \n",
       "203                  Prinsengracht 483 A, Amsterdam  1016 HP  1750000.0   212   \n",
       "207                       Olympiakade 12, Amsterdam  1076 XP  2750000.0   309   \n",
       "258                         Westeinde 16, Amsterdam  1017 ZP  4550000.0   497   \n",
       "278                    Keizersgracht 544, Amsterdam  1017 EL  5950000.0   394   \n",
       "316                      Raphaëlstraat 5, Amsterdam  1077 PN  1700000.0   203   \n",
       "323                       Kromme Waal 18, Amsterdam  1011 BS  1500000.0   194   \n",
       "368  Tweede Jan van der Heijdenstraat 13, Amsterdam  1073 VE  2650000.0   387   \n",
       "382                    Klaprozenweg 73 A, Amsterdam  1032 KK  1100011.0   207   \n",
       "393                          Uilengouw 2, Amsterdam  1027 EZ   899000.0   258   \n",
       "411                Nieuwe Leliestraat 11, Amsterdam  1015 SH  2450000.0   259   \n",
       "422                  IJsselmeerstraat 52, Amsterdam  1024 ML   799000.0   230   \n",
       "423             Dorpsstraat Holysloot 62, Amsterdam  1028 BE  1495000.0   214   \n",
       "431                          Bosplaat 80, Amsterdam  1025 AT  1000000.0   220   \n",
       "462                      Geschutswerf 81, Amsterdam  1018 AW  1475000.0   220   \n",
       "466                      Badhuiskade 353, Amsterdam  1031 KV  3680000.0   374   \n",
       "517          Plantage Muidergracht 105 A, Amsterdam  1018 TP  2250000.0   250   \n",
       "529                    Haarlemmerweg 705, Amsterdam  1067 HP        NaN   366   \n",
       "537                        Parosstraat 6, Amsterdam  1060 LP  1500000.0   255   \n",
       "552                   Den Texstraat 35 2, Amsterdam  1017 XX  1320000.0   230   \n",
       "625                  Herengracht 116 F+K, Amsterdam  1015 BT  1595000.0   199   \n",
       "635                  Bellamystraat 13 15, Amsterdam  1053 BE  1250000.0   202   \n",
       "649        Oudezijds Voorburgwal 126 A+B, Amsterdam  1012 GH  1400000.0   213   \n",
       "672                    Baarsjesweg 313 H, Amsterdam  1058 AH  1499000.0   245   \n",
       "683                   Colenbranderhof 30, Amsterdam  1065 BG  1650000.0   235   \n",
       "687                  Keizersgracht 171 F, Amsterdam  1016 DP  1625000.0   199   \n",
       "688                    Prinseneiland 545, Amsterdam  1013 EC  1950000.0   258   \n",
       "690               James Bradleystraat 11, Amsterdam  1086 ZM  1200000.0   231   \n",
       "\n",
       "     Room       Lon        Lat  \n",
       "11      7  4.776234  52.349215  \n",
       "12     13  4.887869  52.371039  \n",
       "35      9  4.887183  52.355020  \n",
       "39      9  4.869761  52.348636  \n",
       "46      6  4.896542  52.360937  \n",
       "60      5  4.838629  52.374910  \n",
       "74      8  4.909819  52.368653  \n",
       "85      7  4.857608  52.352366  \n",
       "136     7  4.875471  52.361571  \n",
       "144     8  4.889816  52.375532  \n",
       "172     5  4.860637  52.351894  \n",
       "173    14  4.970830  52.412038  \n",
       "177     6  4.876183  52.366972  \n",
       "199     5  4.942112  52.376819  \n",
       "203     4  4.883097  52.369689  \n",
       "207     8  4.862413  52.349528  \n",
       "258    13  4.898620  52.358798  \n",
       "278    10  4.887378  52.365078  \n",
       "316     5  4.869966  52.349904  \n",
       "323     9  4.902614  52.374848  \n",
       "368    10  4.899350  52.355179  \n",
       "382     5  4.900480  52.400948  \n",
       "393     1  4.967897  52.410672  \n",
       "411     8  4.883278  52.375756  \n",
       "422     8  4.961381  52.389087  \n",
       "423     5  5.029122  52.413550  \n",
       "431     7  4.948314  52.400138  \n",
       "462     5  4.920481  52.366984  \n",
       "466     4  4.905695  52.385321  \n",
       "517     8  4.915246  52.364132  \n",
       "529    12  4.787874  52.383877  \n",
       "537     7  4.790343  52.345566  \n",
       "552     8  4.894964  52.358917  \n",
       "625     6  4.888459  52.375994  \n",
       "635     6  4.866447  52.367118  \n",
       "649     7  4.897256  52.373343  \n",
       "672     8  4.854662  52.358204  \n",
       "683     7  4.820848  52.358631  \n",
       "687     6  4.885830  52.374742  \n",
       "688     4  4.887444  52.385346  \n",
       "690     8  4.979592  52.360578  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "o 66 datapoints with Room > 5.5\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>De Bazelhof 24, Amsterdam</td>\n",
       "      <td>1067 LZ</td>\n",
       "      <td>400000.0</td>\n",
       "      <td>107</td>\n",
       "      <td>6</td>\n",
       "      <td>4.795810</td>\n",
       "      <td>52.376916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Noorderakerweg 124, Amsterdam</td>\n",
       "      <td>1069 LW</td>\n",
       "      <td>950000.0</td>\n",
       "      <td>226</td>\n",
       "      <td>7</td>\n",
       "      <td>4.776234</td>\n",
       "      <td>52.349215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Singel 318, Amsterdam</td>\n",
       "      <td>1016 AE</td>\n",
       "      <td>4900000.0</td>\n",
       "      <td>623</td>\n",
       "      <td>13</td>\n",
       "      <td>4.887869</td>\n",
       "      <td>52.371039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>Ruysdaelkade 109, Amsterdam</td>\n",
       "      <td>1072 AN</td>\n",
       "      <td>1650000.0</td>\n",
       "      <td>232</td>\n",
       "      <td>9</td>\n",
       "      <td>4.887183</td>\n",
       "      <td>52.355020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>Gerrit van der Veenstraat 145 B*, Amsterdam</td>\n",
       "      <td>1077 DZ</td>\n",
       "      <td>1625000.0</td>\n",
       "      <td>257</td>\n",
       "      <td>9</td>\n",
       "      <td>4.869761</td>\n",
       "      <td>52.348636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>672</th>\n",
       "      <td>Baarsjesweg 313 H, Amsterdam</td>\n",
       "      <td>1058 AH</td>\n",
       "      <td>1499000.0</td>\n",
       "      <td>245</td>\n",
       "      <td>8</td>\n",
       "      <td>4.854662</td>\n",
       "      <td>52.358204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>683</th>\n",
       "      <td>Colenbranderhof 30, Amsterdam</td>\n",
       "      <td>1065 BG</td>\n",
       "      <td>1650000.0</td>\n",
       "      <td>235</td>\n",
       "      <td>7</td>\n",
       "      <td>4.820848</td>\n",
       "      <td>52.358631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>686</th>\n",
       "      <td>Valeriusstraat 148 Hs, Amsterdam</td>\n",
       "      <td>1075 GE</td>\n",
       "      <td>1775000.0</td>\n",
       "      <td>177</td>\n",
       "      <td>9</td>\n",
       "      <td>4.865595</td>\n",
       "      <td>52.353408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>687</th>\n",
       "      <td>Keizersgracht 171 F, Amsterdam</td>\n",
       "      <td>1016 DP</td>\n",
       "      <td>1625000.0</td>\n",
       "      <td>199</td>\n",
       "      <td>6</td>\n",
       "      <td>4.885830</td>\n",
       "      <td>52.374742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>690</th>\n",
       "      <td>James Bradleystraat 11, Amsterdam</td>\n",
       "      <td>1086 ZM</td>\n",
       "      <td>1200000.0</td>\n",
       "      <td>231</td>\n",
       "      <td>8</td>\n",
       "      <td>4.979592</td>\n",
       "      <td>52.360578</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>66 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         Address      Zip      Price  Area   \n",
       "3                      De Bazelhof 24, Amsterdam  1067 LZ   400000.0   107  \\\n",
       "11                 Noorderakerweg 124, Amsterdam  1069 LW   950000.0   226   \n",
       "12                         Singel 318, Amsterdam  1016 AE  4900000.0   623   \n",
       "35                   Ruysdaelkade 109, Amsterdam  1072 AN  1650000.0   232   \n",
       "39   Gerrit van der Veenstraat 145 B*, Amsterdam  1077 DZ  1625000.0   257   \n",
       "..                                           ...      ...        ...   ...   \n",
       "672                 Baarsjesweg 313 H, Amsterdam  1058 AH  1499000.0   245   \n",
       "683                Colenbranderhof 30, Amsterdam  1065 BG  1650000.0   235   \n",
       "686             Valeriusstraat 148 Hs, Amsterdam  1075 GE  1775000.0   177   \n",
       "687               Keizersgracht 171 F, Amsterdam  1016 DP  1625000.0   199   \n",
       "690            James Bradleystraat 11, Amsterdam  1086 ZM  1200000.0   231   \n",
       "\n",
       "     Room       Lon        Lat  \n",
       "3       6  4.795810  52.376916  \n",
       "11      7  4.776234  52.349215  \n",
       "12     13  4.887869  52.371039  \n",
       "35      9  4.887183  52.355020  \n",
       "39      9  4.869761  52.348636  \n",
       "..    ...       ...        ...  \n",
       "672     8  4.854662  52.358204  \n",
       "683     7  4.820848  52.358631  \n",
       "686     9  4.865595  52.353408  \n",
       "687     6  4.885830  52.374742  \n",
       "690     8  4.979592  52.360578  \n",
       "\n",
       "[66 rows x 7 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "o 1 datapoints with Lon < 4.730323562198041\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>504</th>\n",
       "      <td>Lijnbaansstraat 7, Amsterdam</td>\n",
       "      <td>1016 ST</td>\n",
       "      <td>950000.0</td>\n",
       "      <td>133</td>\n",
       "      <td>7</td>\n",
       "      <td>4.644819</td>\n",
       "      <td>52.37806</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          Address      Zip     Price  Area  Room       Lon   \n",
       "504  Lijnbaansstraat 7, Amsterdam  1016 ST  950000.0   133     7  4.644819  \\\n",
       "\n",
       "          Lat  \n",
       "504  52.37806  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "o No outliers in Lat\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Price</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [Address, Zip, Price, Area, Room, Lon, Lat]\n",
       "Index: []"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>lower_threshold</th>\n",
       "      <th>upper_threshold</th>\n",
       "      <th>n_outliers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Address</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Zip</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Price</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1196250.0</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Area</th>\n",
       "      <td>NaN</td>\n",
       "      <td>190.0</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Room</th>\n",
       "      <td>NaN</td>\n",
       "      <td>5.5</td>\n",
       "      <td>66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lon</th>\n",
       "      <td>4.730324</td>\n",
       "      <td>5.048645</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lat</th>\n",
       "      <td>52.288533</td>\n",
       "      <td>52.43702</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        lower_threshold upper_threshold n_outliers\n",
       "Address             NaN             NaN        NaN\n",
       "Zip                 NaN             NaN        NaN\n",
       "Price               NaN       1196250.0         61\n",
       "Area                NaN           190.0         41\n",
       "Room                NaN             5.5         66\n",
       "Lon            4.730324        5.048645          1\n",
       "Lat           52.288533        52.43702          0"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count_outliers(data, data_info, data.columns, show_details = True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- There is one data point in Area feature, that seems to be a data mistake: in row 393 there is only 1 room per 258 sq.m. which is too much for a house (it is predefined that this dataset is about residential houses, not warehouses, shops, etc.)\n",
    "- All other outliers are just big expensive houses.\n",
    "- There is an outlier below 3std in Longtitude. Lel's plot Lat vs Lon and see what the map of houses looks like:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_info['n outliers'] = 0\n",
    "data_info.loc['Area', 'n outliers'] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Lat', ylabel='Lon'>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAGwCAYAAABB4NqyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACqV0lEQVR4nOydeXwTdf7/X7nTtE1bEo4ibSmka4G2UKxcaVEB1wMQkfVAfrsc6uoXEFxdF6scImjF9QR0d2FB9LsKeyDgtX69doECCkKRcogtIEUolJYmIc19/P5oZ5gkM5OkTS/6fj4efUAmM5PPTCbzec37lPj9fj8IgiAIgiC6ENL2HgBBEARBEERbQwKIIAiCIIguBwkggiAIgiC6HCSACIIgCILocpAAIgiCIAiiy0ECiCAIgiCILgcJIIIgCIIguhzy9h5AR8Tn8+HcuXNITEyERCJp7+EQBEEQBBEBfr8fly9fRu/evSGVitt4SADxcO7cOaSlpbX3MAiCIAiCaAZnzpxBnz59RNchAcRDYmIigMYTqNVq23k0BEEQBEFEgsViQVpaGjuPi0ECiAfG7aXVakkAEQRBEEQnI5LwFQqCJgiCIAiiy0ECiCAIgiCILgcJIIIgCIIguhwkgAiCIAiC6HKQACIIgiAIostBAoggCIIgiC4HCSCCIAiCILocJIAIgiAIguhykAAiCIIgCKLLQQKIIAiCIIguB7XCIAiCIIirHLPNhVqrCxaHG9o4BfTxSiRplO09rHaFBBBBEARBXMWcM9mxYPMh7KyoZZeNztLjxSl56J0c144ja1/IBUYQBEEQVylmmytE/ADAjopaPLX5EMw2VzuNrP0hAUQQBEEQVym1VleI+GHYUVGLWisJIIIgCIIgrjIsDrfo+5fDvH81QwKIIAiCIK5StGqF6PuJYd6/miEBRBAEQRBXKfoEJUZn6XnfG52lhz6h62aCkQAiCIIgiKuUJI0SL07JCxFBo7P0WDElr0unwlMaPEEQBEFcxfROjsOqqfmotbpw2eFGoloBfQLVASIBRBAEQRBXOUkaEjzBkAuMIAiCIIguB1mAiLBQCXWirbhgcaC+wQWLwwNtnBwpGiV6atXtPSyCIK5CSAARopwz2bF422Fkp2qRn5aMarMDNRoF0rtpcE2Kpr2HR1xFVNU1oHhLOXZV1rHLCg06vDA5F+m6+HYcGUEQVyMkgAhBzDYXFm87jPuGpePtXaew+utK9r1Cgw4v3pWHPt1IBBEt54LFESJ+AKC0sg5PbynHK/cMIUsQQRAxhWKACEFqrS5kp2rx9q5TvBNT8ZbyLt1Hhogd9Q2ukGuMobSyDvUNdJ0RBBFbSAARglgcbuSnJQtOTDu7eB8ZomWYbS6cqLGirKoePr8fc8cYoFHKeNe1ODxtPDqCIK52yAVGCKJVK1Btdoiu05X7yBDN55zJHtKh2mjQYeXUfMzbWAabyxuwvlZNtyqCIGILWYAIQfQJSiTHUR8ZIraYba4Q8QMAuyrr8PauU5hVmBmwvNCgQ7xShrKqepy4aCW3K0EQMYEeqwhBkjRKZOg0KDToUMrjBuvqfWSI5lFrdYWIH4ZdlXWYZbwigAoNOjx7Rw4m/2k3624dnaXHi1Py0Ds5rk3GSxDE1QlZgAhRrknR4MW78lBEfWSIGGEJ4zbVqhX4x8Mj8dn8Iswdk4X71uwJiDXbUVGLpzYfIksQQRAtgixARFj6dNNgNfWRIWKENozbtFu8Ev17JOBEjRX3rfmGd50dTQH4dA0SBNFcSAAREUF9ZIhYoU9QYnSWHjt43GBct2o4SxEF4BME0RLIBUYQRJuSpFHixSl5GB3GrRrOUkQB+ARBtASyABEE0eb0To7DqjBu1UgtRQRBEM2BBBBBEO1COLcqYyl6avOhABEkFIBPTXsJgogGEkAEQXRYGEuRyeaGze2BRCIB/MBZkx0muwsapRzJcQrYXF78Iai2EKXLEwQhRrvGAD377LOQSCQBf7169RLdZvv27bjuuuugVqvRr18//PnPfw5ZZ/PmzRg4cCBUKhUGDhyILVu2tNYhEATRyjS4vFj2yVGcNTmw7OOjuPWNnfjVn/fgtjdK8fSWchw7fxln6m3Yf7o+YDtKlycIQox2twANGjQIX375JftaJuPvBQQAp06dwu23346HHnoIf/vb37Br1y7Mnj0b3bt3x5QpUwAAe/bswb333otly5Zh8uTJ2LJlC+655x6UlpZi+PDhrX48BEHEDqZq9OC0ZN6mvMzrSYN7Y9XUfACA0+ODWiHDgap6rC89RenyBEHwIvH7/f72+vBnn30WW7duxcGDByNaf8GCBfjwww9x7NgxdtkjjzyC77//Hnv27AEA3HvvvbBYLPj3v//NrnPrrbciJSUFGzdu5N2v0+mE0+lkX1ssFqSlpcFsNkOr1TbjyAiCiAUnaqwY++p2rJtegAfe+U5wvY8eNeKlf/+AnRyBZDToMNOYiR4JSuSlpbTFcAmCaGcsFguSkpIimr/bPQ2+oqICvXv3RmZmJu677z6cPHlScN09e/bgl7/8ZcCyW265Bd999x3cbrfoOrt37xbcb0lJCZKSkti/tLS0FhwRQRCxgqkF5PT4RNersTgDxA9wpbdYUpyw9cdsc+HkRSt+vHAZP5y34MDpS9RvjCC6CO3qAhs+fDjeffdd/OIXv8CFCxewfPlyjBo1CkeOHIFOpwtZ//z58+jZs2fAsp49e8Lj8aC2thapqamC65w/f15wHMXFxXj88cfZ14wFqDNAmS/E1QxTC0glb96z2q7KOri8/OLpnMmOxdsO475h6SHuNQqg7nzQvZCIlnYVQLfddhv7/9zcXIwcORL9+/fHO++8EyBIuEgkkoDXjAePu5xvneBlXFQqFVQqVdTjjwUt+dGeM9lDumrTjZu4mmBqAZWdMcFo0IXEAAFAkUGHsjMmwX2Y7W6YbYFxQOFii5gA6lVT82kS7QTQvZBoDu3uAuMSHx+P3NxcVFRU8L7fq1evEEtOTU0N5HI5azESWifYKtQROGeyY+7GMox9dTsmv7UbY1/Zjkc3luGcyR6wntnmwokaK8qq6lnzPHMDD+6q3RUzX/jOD3F1wNQCOl5twUxjJoyGQMuw0aDDkjsGYX3pKcF9mO3ukN8V05E+Py2ZV1QBV/qNER0buhcSzaXds8C4OJ1OHDt2DEVFRbzvjxw5Eh999FHAss8//xwFBQVQKBTsOl988QV+97vfBawzatSo1ht4MxD70S7YfAjL78wRrW/yzPgBIdty99FVMl/oye/qp3dyHF6+ezDqGlx4duIgeH1+WJ0eaJQyxCvliFPKUJCRwlsx2thkHQq26EQaW1Rvc6Gsqp5cKh0YRszy0ZXuhUT0tKsA+v3vf4+JEyciPT0dNTU1WL58OSwWC6ZPnw6gMTbn7NmzePfddwE0ZnytXr0ajz/+OB566CHs2bMH69atC8jumj9/PkaPHo0VK1Zg0qRJ2LZtG7788kuUlpa2yzEKUWt1Yf/peswdY0B+WnJI6m5ljRXv7P4Js28y8NY3mV5vF9hzI12hUWS4Jz9yX1w9hKsazVcxmskCm7exDEDgZBhpbJHZ7mazz0hYd0yoaS7RXNpVAP3888+YOnUqamtr0b17d4wYMQLffPMNMjIyAADV1dWoqqpi18/MzMSnn36K3/3ud3jzzTfRu3dvrFy5kq0BBACjRo3Cpk2bsHDhQixatAj9+/fH3//+9w5XA8jqdGPl1Hy8vesUVn9dyS43GnRYOTUfHp8fOypq4fX7MaswM2CdSOgKjSLpyY9gYCpGnzXZ8VOdDSq5FGVnTJi3sQw2l5ddj5kMI4ktMgbFFpGw7phQ01yiubSrANq0aZPo+xs2bAhZdsMNN+DAgQOi2/3qV7/Cr371q5YMrdVJjlPipf87LljYbcGt2ezrWcbMkO3LzphQlKUPEQAapQyLJgyEz+8PMN0DuOoyJOjJj+CSpFGi1urC7PeE7w/MZMjEFi3Zdhgzm35fu3hqCDHWIwYS1h0PappLNJcOFQPUlXB5fYLBl7sq6+DixCbwxSmsLz2FT+cVYfG2w+wPX6OUYf2M6/Hm15Uo/qCcXbcoS485Nxkwa8M+9mn4ajDnd9MosW56QYj7kDlGevLrekQzGfLFFtlcXqgVUnx6+HyI9YhBTFhTKnbbE23TXIJgIAHUTlidHtH3L1qvVKbmi1MoyEhBikaBVVPzUWt14bLDjRSNEgu3HsbOysCb/86KWviCXGmd3Zx/zmQPOVbGfThvYxkKMlLoya8LEu1kyBdbdKLGKupyFhLWFJDffjAuUOZemKhWQJ9A4pMQhwRQOxHOb62UNYqeoiw9ai47A94Lvpkz/56osYaIHwY+V1pnNeezwc9Bx8pY1BZNGIgbf9G90x0XERtaOhk2x6VCAfntT7hAeYIIhgRQOxF8k9UoZZhVmIn8tGQAjTfxrx6/AWq5FAlqOYb17Rb2Zh4uJobPldYZ42TEgp93Vdbh2YmDkJocR+6ILkxLJsPmuFQoIJ8gOh8kgNoJ7k32u9P1ghlhc28y4JrkOPTvkRB2n+GsSnyutM4YJ2N1ugXLB9hcXthdHnJHEC0iWisSBeQTROeDBFA7wtxkTTY3Fm4tF8wIm5CbCrlMGnbiFjPdB6f0Ah0/Q0LIgpMUp0RZVT1v+YB5G8uQFKcUdUf88e7BsDo8ZBkieAm+7jL18WGvD0rFJojOBwmgdoZJ3Q3uZM3AxO6Y7e6wAoixKgVP/kVZesy+0YAH3tnHLuvoGRJCFpySu3KxaOthQbG4aMJAuLw+UXfEiRor7v/rtwH7JcsQAVxpkJqdqkV+WjKqzQ7UaBRI76bBNSkawe0oFZsgOh8kgDoAkcTuRGpClwC4PacXZozqC6fHB5VcijqrE4lqGVbfPxQOtxf99PFITVJHLX7aKqZGLKD0dJ1NNNB78YSBYTPszEHnkgJVCaDxuuN2h+daGAsNOrx4Vx76dOMXQZSKTRCdDxJAbYiQgIgkdidRrQjpaM23/+C+YQxGgw756Sk4dMbUrIm+LWNqxAJKTXZxIWh3ecOezz4pcdAoZQE1XsIFqlJA9dVPrdWF7FQtb3f40so6FG8px2qR3w6lYhNE54IEUBshJiD0CUreqs5Ao3CpsTjQPVGFRzeW4YXJuXB5fTDb3UiKUyBeJWfjWeKUMgxOS8b+0/UhBdx2VdZhzo0GTBuW3izLj1jj1pfvHoyeWjW7bkuFgphFLFzvJmbSEYuFKv/ZzNteRMjKxvfd3TygB569YxAcbh+JoqsEi8ON/LRkwRpAOyPI5qJUbILoPIjPJkRMCFcjBABWTMlDUZY+4P3GLLAsjOyvx5z3DmBHRS2WfHgYp2obYLa54fb68Pt/HMTYV7dj8lu7cevrO1FW1ZhRplHKQsahVsiQ2gxrjZhFZmdTTM05kx3VJjvmbixjxzP2le14dGMZzpnEG7cGI2bBYVqA8MHEWiRplHhuUg6MBl3A+0x7g+WfHGPLDXDhC1Tl++40ShnuHZaOP2w+hImrS/HVDzX4qbYB352ux4/nL8Nsc0V4pERHQqtWhO0OT9lcBHH1QBagNiCSGiH9eyRg9dR8nLc4YLF7oFHKEKeUweby4NfrvsWZejs0ShmmDs/Ahl2nMDg9BWWl9YLBwHwWjqS45mWihItRMtndeGrzIdyWmxqTQnBiFpzj1RaUTM7F01vKRWMtzHYX8tNTMMuYycZCcZtjBk90QoGqzHfHrdOkkEmhkEnxYGEm+qRo8NxHRwLONQVVd070CUpcsFA2F0F0FUgAtQGR1ghhzOdlVfUYv6o0ZL1ZhZlsfMIMo3CHeL6qzy3JRIkkRmlHRS2mj+rL+360heDEAkqfm5SDVIFYC6CxGjbjDgSARwX6OXFdaYw7q9bqwsnahgB3lsXhhkYp463TVGTQYfZNBuyvMoUcLwVVdz6SNEpk6DQoNOhQypOVSdlcBHF1QQKoDYi2RojQ+tz4hHCmeu77Lc1EibS+kNiYonUdhAsoDY614IvTKeTUBuKKoNFZehi6J2Dr7FHQximglEnx1AflvPFZSXGKAOHJZWdlHXzgt7ZR9d/OyTUpGrx4Vx6Kt4ReD5TNRRBXFySA2oBoaoSYbS74/H6sm14AiUQSUOGYKzDCBQP308dj6+xRIcKhOUHKQvWFmJiaeRvLwo6pOa6DSANKhWKsSnncgcxElpochwzEw2xzYe7GMtGiiaP66aKytjFQvEjnpE83DVZTNhfRiaGs1cggAdQGRFojhM+KEVjh+IqIKDtjgtGgC7FKAI2FD/nq/LQklb13chxevnswTtRYYbK7Q2JqRvM0beV+Rmu6Dkw2N2aM6oupw9JD2mKUVtZh4fiBGJfdg3ciCxefZXV4oAwjNoUsXxQv0nnpDNlcZpsLNZedMNndiFfKEK+SIzlO0eHHTbQu1AYockgAtRG9k+Pwx7sHo97mwmW7BwlqOVRyKc6Z7bA0PWUu3nY4ZDLeVVkHqUSCf88rgkwmYeMT1peewsqp+ew6DEaDDnNuMoR8fiy6VffUquH1+QWFHPP/tiwE9/MlGxZuLQ+opF1k0GHrHCP+fbgaf9l+Eg63F0PSU3i3jyQ+KyXM2PksXxQvQrQm50x2LPjXoYCioEUGHZbcMQgNTg96i1StJq4uuNaeBJUc352ux/7T9QHrUFwiPySA2ohqkx2nL9mwZscJ3D88A69/9WOAcCky6DHd2Be7T9SFBO3urKiFx+dHhj4+ID5h3sYyzCrMxOwbDZBJJbC5vDhQVY9ZG/bho7mFUVk6Io1XCReb05aF4M7W27Dgg0O8sTlLPzqC8bmpWDk1H1qR7LdI4rPEXJhFBj0uWBwByyhehGgtzDYX2zswuH3Ozso6LPnwCCbkpqIoq7tg1Wri6iGc1yCaYq9dERJAbYDZ5sJ/f7yIjw+dQ356ikBAbS188PMG1AJX4kmY+ISzJjt+qrNBJZdiz8k61uUTvD5DLLtVi7kH2sp1YLa5cLrOxusCBK7E5ry96xReuWeI4H4iic8ScmEWGnSYWdgX5WfNWDe9AEBjlele2ujbjBBEOJjJbsaovmF7B4arWt1SKMak/RGy6ouVQqG4xEBIALUBtVYXeiSq2JtTcwJqufEkTAPV2e8dEPzMSDPLhNbv6NRaXWHbYjg9PpRW1sHq8KCnln+dSOOz+CxfCWo5GpwedNMoKVCWaFW4k93UYemi6zo9voiqVjcXijHpGIhZ9YXmks52n29tSAC1ARaHmw2UjSZ9nYEvniTa7tNXW7dqi8MdNhOOeT/cU0+kPZw6Q2AscXXCnexidd03h1jEEhKxIZIm2lw6432+taFWGG2AVq1gb0rhbl7JQfEqQvEkjOVidFBbiFit39HRqhVsJhwf3PpEkTz1JGmU6N8jAUPSU9C/R0KnOx/E1Q13sis7Y0JRjK77aIkklpBoGyIpUMvQWe/zrQ1ZgNoAfYISe3+6xN6chNLXCw069NXH46vHb4goiDja7tNXU7dqfYISx6stmGnMhBQIiIng1ieK9KmHYhqIjgx3sltfegqr788HJBLBulyt9bQfzupQb3PBbIuN641+k+KEs+ozxV47832+tZH4/X5/ew+io2GxWJCUlASz2QytViB4JEqCs8CCA6ELDTq8MDkX6br4mHxeV+CcyY4l2w5j0DVJGDugB2osjXWIys6YsGlvFZ745bUYmp4Mm8sregOlmAaio2O2ufDoxjJ2stMoZZhzU3/cdG1PVJsbmw2XnTFhfekpFGSksMU+Y82JGivGvrpd8P110wvwzu6fWvzbod+kMIwwtDrdSIpTYvG2w7zxi63x/XcGopm/SQDx0BoCCLiSwmpzeyCVSODzA1anB4lqOVI0SvTUqmP2WV0F5mbQ0HQzcHl9sLs80MYpsWjr4YA6KXw3UKFK0Mz6FNNAdBTOmewhwfpMHzuH29cmVt1gIcbFaNAhPz0Fq7+ubNFvpy1+k53VuhQsDDVKGRZNGIih6cmwu7xk7QEJoBbTWgKoq3LB4kB9gwsWhwfauNYXe9HcQMM90X71+A3o3yOh1cZKENHATNzt6cLmE2Jc9xtTjqO5v53W/k12VusSPaxFRjTzN8UAEa3KmboG7KysRU+tGk6PD1anB2Wn61Fo0EMbp4jpUxgzOTg93oiLPsayPhJBtDYdIRORiSXk1iLjtsVhaO5vpzV/k505iy1WxWyJK5AAIlqNGosDP5vs+KS8OqRdR2b3BHxzsg5Pbi5nl7fkKeycyY7F2w4jO1WLwv560XW5N9CrrT4SQbQFzalFFimt+ZvszCKCHtZiD6XBE61Gg9OD1f+pDMl421VZh1VfV+Da1EDzJPMUZrZFl0prtrmweNth3DcsHWVV9WhweUTX595AmUwKPqhuBtGamG0unKixoqyqHicuWqO+7tub1vrttHS/ZpsLFRcuY99Pl3D0nBmn6xrYc9uZRQQ9rMUesgARrYbN7RVtVcFHc57Caq0uZKdq2cy6/PQUwVIDwTfQSCtBE0Qs6axxKFxa67fTkv3yNYk1GnR4dEwWMrppOrWIuNqK2XYESAARrYbN6RV/38X/frRPYWa7C/lpyWyLkfWlp7Byaj6AQKEldAO9muojtTWdNZumPenMcSjBtNZvpzn7NdtcIeIHuHIPmJDXG2Oye8RMRLT1tU8Pa7GHBBDRaiSJdGEHALlUwrs8mqewcyY7HG5fQNl3m8uLeRvLMKswE7OMmXB6fMjQadAnOa7dm7heTVwNVoz2oDPHofDRWr+daPdba3WFiB+GXZV1ePr2ATDZXHjxrlws+fAIvjhWw74fiYjgCp54pRz7q+qx7OOj7INcW1z79LAWWzpMDFBJSQkkEgkee+wx0fXefPNNDBgwAHFxcbj22mvx7rvvBry/YcMGSCSSkD+Hw9GKoyf46KlVoUjAl19k0KOU52YVzVMY8yS9+2RdSAsRm8uL1V9X4oF3vsPs9w5AAtBNIoaEs2J0tniWtqQzx6F0ZMKd19N1Ntzy+k4s+KAcSyYOwn+euAFf/G40PptfhMfGZcHm9gZct9wYrR/PX8anh89j4upSTH5rN375+g58fOgcVk7Nh0YpA9B21z617YkdHcICtG/fPqxZswZ5eXmi6/3pT39CcXEx1q5di+uvvx579+7FQw89hJSUFEycOJFdT6vV4vjx4wHbqtVUZLCtSdIosWJKXshEWWTQYdHEgXjx38cC1o/WlMs8Se8/XY8p+deg0KBDKU/cj9GgQ7yyQ1zqHZZozflXmxWjLenMcSixJpZupEh7Y+2sqMWzHx3BsxMHYfGH5bwWTAmAPwTdt4wGHVZOzWfT/RnX2qzCTNb9Ttd+56LdZwWr1Ypp06Zh7dq1WL58uei6//u//4uHH34Y9957LwCgX79++Oabb7BixYoAASSRSNCrV69WHTcRGb2T47C6yWRrsrvgdPuw+2Qd7l/7De4blo4HC/tBpZAiOU4ZtSmXeeKzubz495FqPH37AJR8+gNvAGSyJraTytUU+9IcVxZZMZpPRwlmbe9rONYuVH2CEkVZel5hzm0SCwDZqVoUf3AooIcgcMWKc1tuash+dlXWQSWXYlVTfKHT44NaIYM+QYn1pacANIohp8eLsqr6Tn9f6Aq0uwCaM2cOxo8fj3HjxoUVQE6nM8SSExcXh71798LtdkOhaJzkrFYrMjIy4PV6MWTIECxbtgz5+fmi+3U6nexri8XSgiMiguH68s02F3pq1RiX3aPF/mvuE9/qr09gQK8k3JbbCzOMfeH0+KCSS1Fz2YnMbhoAjRVmY3GzP1tvw+k6G0x2N9QKGb76oQY/VFtQfPsASADoOtFNr7kBuWTFaD6xDmZtjpD5+ZItRAAUZelRMjkXfZp+L61JawSC21xezLnJAJ/fH1J3jKlSzcBNmghmR0Utpo/qG7Jco5Th/uEZ2LDrVOB5M+iw+v58SCDBX0tPBuyXYuI6Nu0qgDZt2oQDBw5g3759Ea1/yy234K9//SvuvPNODB06FPv378f69evhdrtRW1uL1NRUZGdnY8OGDcjNzYXFYsEbb7wBo9GI77//HllZWbz7LSkpwdKlS2N5aIQAsQyY5D5J21xezHn/AGYVZqKnVg2VXIr0bhoUZKSgweXFE//8HtmpWuSnJaPa7ECNRoH0bhpckxL+Zs+dYDRKOfafvoTlnxxjgx+ZG2zJJ0eR0ycZB07Xd5qbXnNdWR3FitEWcL//BJUcSpkUJrsLCermC2mhYFYgOqHeHCvK2XobnvrgUIi7eGdFLZ764BBWTMmL6HfREmLtQjXbXPjD5kPYf7oeswozsXD8QJjtbjQ4PbxVqrlJE3zwvT+rMDOkiTUAVgzdlpvKvqdRyjCrMBP5ack4Vm3BZYcbydTvscPRbgLozJkzmD9/Pj7//POI43MWLVqE8+fPY8SIEfD7/ejZsydmzJiBl156CTJZYyDaiBEjMGLECHYbo9GIoUOHYtWqVVi5ciXvfouLi/H444+zry0WC9LS0lpwdERbEPwkzQQ+c7shc4skvr3rVMDTWaFBhxfuyoPH64PZzj/h8E0wQrEA+ekpGDugB9bsONlp0pmb48piBMG8sVn4nxv7Y9eJOqwvPQWby9tqKbnt5a7hVhhnxHOyRoEElRz3/mUPBqZqw4pdobEHPwxEK2aaY0Ux21w4XWfjjZUDgNLKOpyusyFBJW/V8xtrFypXUK3+uhK/HNgTpy5a8XFQFXqG4KSJYJh4IS5iVqOdlXWYYcwE0Ch+Vk7N57/fTM5Fui4+4uMiWpd2E0D79+9HTU0NrrvuOnaZ1+vFjh07sHr1ajidTlbUMMTFxWH9+vX4y1/+ggsXLiA1NRVr1qxBYmIi9Hr+bCOpVIrrr78eFRUVgmNRqVRQqVSxOTCiTeE+SXM7wp+3OGBze+Hz+QOKJHIprazD0x8cwpCmDtZA4IQjNMHwBT/uqqzDLGMmaixOdnlHDIYMnowTVOK3gGBXFt8kXZSlx0ePFraa+6+90u3FxHORQY+ND43AOZMD23+8iNtzevEed6Rjb46YidaKYra5UG12wGQXFxcmu7vVr91Yu1CDBdWc9w7gf2cNQ199o9jYFeCy0iNRLReMFxqdpUfNZWfI8kitRkKWotLKOjy9pRyv3DOELEEdhHZLgx87dizKy8tx8OBB9q+goADTpk3DwYMHQ8QPF4VCgT59+kAmk2HTpk2YMGECpFL+Q/H7/Th48CBSU1Nb61CIdoZJC9UnqrFw22Hc/NoOTH5rN8a+sh1Vl2zIT0sWrDxdWlmH/LRk9jU3lVVsgtkVtB1w5QbILO9ogcDnTHbM3ViGsa9uZ8/Pd6frBUsVBLuyhCbpnRW1WPrhkVYRP7FMt4+29URwhXEuOysbM4n2nb6Ejw+dQ72N31IW6dgjETPBRGNFYb77k7UNvNYNLiq5tNWv3Vi30eAKKo1Shkn516DG6sTeny5hwa3Z+HCuEZt+OwL/emQkruubgpkb9mHJxEEhY2AsmDf+onvIe5FajcLdb+obqERER6HdLECJiYnIyckJWBYfHw+dTscuLy4uxtmzZ9laPz/++CP27t2L4cOHo76+Hq+++ioOHz6Md955h93H0qVLMWLECGRlZcFisWDlypU4ePAg3nzzzbY7OKLNEZpsgOj9/cyEE26CCd4uKU6BPSfrMLCpx1lHCgQWOj/LPj6KddMLAL8/ILCTz5XVWqnvYu6tWH1mczPdxNwejNVv9deVWLTtMFYHWWmiGXtzXELhrHfxTe9zv/sZo/qi7IwJRQY9b9FAJltq0uDeovtuKbEOBGcE1Xen61n3U35aMl77ogKvfcFv/ff6fKJFBYPfS1DLRePeGKtRuPuNxSHeq5BoO9o9C0yM6upqVFVVsa+9Xi9eeeUVHD9+HAqFAjfddBN2796Nvn37suuYTCb89re/xfnz55GUlIT8/Hzs2LEDw4YNa4cjINoKocmm7IwJI/vpRLfleyK+7HBHXFcEaPTvJ6rlWF96Cqum5ne4QGCh82NzefHAO99h029HYEbTDTy9mwY9ElUhk1BrpL6HEyax+MyWZLpVm8ULqDKT3U6OoGEEXV2DC+tnXI8DVfVsjJTQ2JvjElLKpII974wGHZSyxuuT+92XnTHhh3NmPHHLtfDDHxALxATzb/r2NO4e2kd0PLEgllWNGUG1/ceLEfUELDTokMwThxW8z+D3xEQb8/9wFjatukNPu12KDvVN/Pe//w14vWHDhoDXAwYMQFlZGcR47bXX8Nprr8V4ZERHR2iiXF96Cr8c2FO0SCK3PggDczMWeuLjbldk0OP3t1yLWRv2IT89GTWXnR2uN4+YkLC5vPi53o7Z7x0AAHz1+A1tkvoeiTCJxWe2JNPtgiVyEXzZ4Y4oaJ5v7M3JqjPZXZjZFHjLl/ZttrsAxAd89+tLT2HrbCNe+/I4/nBrNv7H4YHZ7oZKLkXZGRPe//Y0nrptALttaxPLrNDeyXEoyEhB8QflAIR7AhYZdHh+ci4bh3PB4kB9gwsWhwfaODlSRLK1wom2VVPzUW9zCVrYCg06pMR3nPtCV6dDCSCCaC5CE6XN5cWsDfvwj4dHYsm2IwE3pUKDDjOC6oMAVyYcMTP90kmDcLrWhnXTC1B2xoSpa7/BdRkpWDYpBykaRYcSP0DkVXIBYatKrFPfIxEmLf1Ms82FS2FifYSON0mjRIZOE7F4jlfJIw6aDx672LX2UpN1ITg9PkGlwNS13wb0vGOEzLyNZfhobiGAwO/e5vLiVF0D+vdIxCWrC0kaBbRxctTb3BiangIAOFXbAEP3BNFz1lGxOq+4l4R6AuoTVKzAqaprQPGW8gCBVGjQYfmduYK/4+C6ZrVWF07WNrDfS199Ap6fnIOnt5QHXDdMFhgFQHccSAARVwViE+XAVC32n67H4PRkzB+XBbPdDYlEAn2CEq9/+WPAU3lwDILYE183jRK1Vhd08UpMHnJNh25KGKk1CxC2qsQ6biMS91b/HgnN/kzGGjODp6gdF6HjNdtccHl8WHZnDhZvOxJi1eEW1xudpYdSJhUNmp/VZK0RGrvQtSZUxyqtmwbG/jreGCWuwAr+7uVSiWBcEwC8/+DwDuW+jYZgoW9zebG+9BRbk8fm9KJe2iiIJUCI+AEaA5UXbi3H3DFZSO+mEYwRE3Pfpuvi8co9Q65YltRypMRTHaCOBgkg4qog3OQMAB9tPsSaxdeXnkRZlQmzCjMxbXgGAKBPShx6adUhE5OQmb4zdZAXOj98E7nY5BfLuI1I3VvN+Uyue21wWrJgLMjoLD0S1HLWusIUOqxrcMLp8WHXiTps2luF+4alY/aN/aGQSeHyNLZzYVxazDVWc1k8XkijlGPd9AIYuicgVWBSDb6mwtWxen5yLgCIdjYP/u7LzphEY2MydJpOc10HEyz2xGryLJuUg7IqE+9+SivrsOC2bNF6SuHctz21ahI8HRyJ3+/3t/cgOhoWiwVJSUkwm83QarXtPRwiChiTNN9EybzHrRfU4PS0uCVHZ4I5B9y+bMFFDIUm59YYy6MbywTdWy0pJHmixoqxr24HEDgJ7grKdFt+Zw6e+/govuQICK4ozE9PZv9vc3lx84AeWD45F1aHJ+Qaq7hwGTe/tkNwTB/ONaLW6kJaShyyeiZGfBwfHTqH7366xOuGK8rS4+W7B4eMB0BIZt2VZS4kqhRY+tGRdmuF0ZqcM9lZsTd3jAFlVfW8Yq/IoMfgdOEsv40PjcDUtd/gq8dvQP8egS5B7vXFB982RNsQzfxNFiDiqiLarI6uRmv1ZWvuWGLpUuPCda/xxYL01WmQqJJj6UdH8OUPFwO25YvZYf7/xbEaPHWbJ2RyqzbZ4fH5ReOFPj96Aau/rkRR0/FFUsTR6nRj7IAeeP1L/lTunRW1sDoCxyPmmmHWM9tcWH5nLhpcHthcXiTFKXgz/zojXIuhw+MVqd5cixnGvoL7SVA31qLjixFri2bA7d2stitAAogguigdQRDG0qXGhS8WhDsRfvJoIarq7SHih4Ebs8P9PxA6uZltLvz3x4v44uh5zDBmwg+EBNU+NykH9/xlD4Cmnls8rhW+CS8lTomKi1bRY+WOJ9KU/47w3ceCcG1G9p661Kz9Fhp0qLE0loXgixFr7WbA7VX9vKtBAoggiHalNSZjsaDvIoMO3/9sQnKYz+QWtOP+P3hyM9ncyLlGi6Q4BRRSKWYZM/FAYT843F42K+uCxYH7hqWzIiw4/f6cyY4F/zoUkKXIZBvKpBLRcXLH05yU/2jSwDsSkYiEcDV3rkmJC7HaFRp0WDxxEO5f+w1uHtAjIEaMEVmt2Qy4uXWriOghAUQQxFWHkHutKEuPRRMG4s43d2FVU40YIbilAZj/B09u50x2LNxaHhBLw8QQ/f6f37MZhgObMri4MJYbs80VIn6Axglv8bYjWDxxoKBrrShoPNG6ZoTSwDt6085IRUJKvFLw3BUadEjRKPHKPUNwqcEFs92NeJUMNRYn7l/7DfLTkrFowkD8/p/fh4isFVPyWs1921oV14lQSAARBHHVYba5YHd58di4LDw9fgBkEglkUgk8Pj9O1TbA5vKKZkNxSwMw/w+e3NhJOGh7vhgilVwa0iKBsdzUXHbyFs0DmipMX3YKutaW35kTMBlG45q5YHEIpoF39KadkYqEnlo1XpicG7YmT0+tGmZbY/Xu9G4y/O2B4WhweXDWZMfgtGTsP13PitkdFbVY0CSyWsN92xbxRUQjJIAIgui0BLtvummU8Pj8gq4Rh9sLeZNLSahSMDcLrChLj6V3DAIAPFSYGXGvL27cECOguBYgriUpXHf2epsbv//n97wFD93eQFEVjWumvsEVtmlnRxVA0YiESGvyJGkaay5FUsmbEVn9eyTE3BrT2vFFxBVIABEE0WzaM1OFz31TMjkHn5ZXh1hlGNfIH+8ejEsNLrwwOQc9tWq4vD48fvO1WHCrBOctDvRIVCM5Tg6L3Y2P5haKPtFbHG5olDLMKszE0PQUyKUSdItXwu31wWJvFGRjsrvj/uEZeP/b0+x2wZakeKVM9DjVCllIEDfDuOweAa+jyawL15SzIzftjFYkRFKTR8itxmfRA1rPEtOa8UVEICSACIJoFu2ZqSLkvumhVYeIH4YdFbUw2VwYntkNz2wJjdt5dEwWeiWq0CvCsSfFKXiL7DEWpDf/U4l5Y3+Bv+44gcUTBsFid/FWDE9UyUXjVMRKtSWqFbwiNBLXTLgA4Y7ctLM1REKkFj2G1rLEtGZ5CCKQjnuFEwTRYWnvTBUh901wnE0wdVYX/vTfE7xxOzKJJGxgNJd4lTykuCKzLwDIT0/BK58fx8t3D26yPoQGFZ8z2XHWZOeN8TE29aoTshCNztJDrZBi7sYy0Zo/QogFCI/N7g5tnCIk+6mjTL6tIRLCudW411ZrW2JaqzwEEQgJIIIgwhJsZZBLJNh/up533bbIVBFyz3Azt/hIjJMLBhxHO26TTTiGhrEYrP66ElaHBz15CtIyInLqsHTBGJ95G8vwvw8MC7F2jM7S44XJuVjy4ZFmi1ChAOGx2d2xeOKgDl+HJtYiIdKGwW1liblaajV1ZEgAEQQhCp+rqyhLHxIYyqW1M1WE3DPhMrs8XvHOP5GO22xz4ed6u+g6jMVAaJ+My2XGqL6CMT4AkBzH79Kqa3AFtPDgEqmY4wsQ1sYpOk0dmliKBNHaUVl6pHfT4KvHb2iWyGpJrBx326Q4BeJVclgdng5pmetskAAiiC5EtDdiIVfXzopa+Pz+kMBQhtaKjzDbXDDZ3FDKpCgy6FlrDhOMXJCegjvyemPZJ0d5u7db7OKBvZGOu9bqCrsOYzEITj1nxEaiurE56omLlwOOhQtT54dvoj9Z2yD6+SZ7+DECoQHCJ2qsrV6HRug6bM+g+nButeb2yGtJrBx3W7Gedh3JMteZIAFEEF2E5tyIow0MZfbZGvER50121NvduNTQOEEuuC0bt/1swqtf/IgXp+SxwciMGHrkhv6QSSWwubw4UFWPpzYfwv/OGiYY9xLNuK3ORquOkHBhUt8LDTokNFmrhIoOPjcpB4P7JMMHf0gM0JybDIJjCOeycbp9MNuiFyutXYdG6Drka0zb1pN7rN1q0cbKcQVggkqO707Xs67mWYWZvDFnHdEy11kgAUQQXYDmBi2HmwyDaa34CLPNhUs2F57/9FjABFBk0OGfj4zE8x8fZZcz7qTVX1fCaNAhPz0Fq7+uRMnkXLzy+XHBooIvTM6NeNxJcUq88VUFphv78gqXmcZMvP/tacwwZqLB6REtOrhk2xFc1zcF+ekpITFAszbsw0dzC3nHpU9QoihLzytQjQYddp+sQ0+tOurvorXq0JhtLtRcduJMvQ2PjfsFnrzlWlSbHVDKpPD5/ThrsuM3I/vi/uEZOFBVj/Wlp9plco+lW81kc2PGqL6YOiwdaoWMPS6by4vvTtfDZHOj1uqC1elGUpwSi7YeDhDU3BpE+WnCneupQnTzIAFEEF2A5pbXDzcZMnERrZ2pYrG7URIkfgBgZ2UdFm87gsFpybyNTXdV1uHp2wYgPy0Z1yTHoXhLOXafvMQbcOzyimeQMZhtLnai+qZpXw8U9oNcJoFOo4REAjjcXgzsnYR5G8vw/oPD4fL4BAOmma7kD7zzHe/7QhaXJI0Sz94xCIu3HRYs5BhcJygSWiPFnM/qw4xzzvsHMDQ9GTOaxmxzeQMm/taY3NvC1SbUJmXl1Hw8tfkQXpySx74/d4wBZVX1ghmFswozw2Y4UoXo6CEBRBBdgOa6NcJNhj0SVa3+1HnOZMeZertgfR8mkFiIOpsLb+86hbk3ZQEI7QzPEKlYqLW62Kd0m8uL9aWngEIgPy0Zpy/ZkKmPx46KWvZJP1GtQF2DeDyO2OQmZnGRALyWI0ZINMdaE+sU80gLDPo5/w9+L5aTe1vUrwrXJmXFlDys57izxKw7Qq7mYKhCdPSQACKILkBz3RrtXZSNmyouhpiAUEglWDxxEM6ZxLO2Ip1AuGKSG5gaXAxx5dR8/H1vFfQJypCWFcEkx/F/djiLiy5eiUNnTLyTZ0tisWIZCxNpHFnwRM99He3kLhZk3RYZbuGO+anbsgOsPeGsO06PD0erLYIZjlQhunmQACKILkBL3BrMZGiyudHg8qDB5UVynAKaMC0cYgE3VVyMJAEBUZSlR5/kOByvsaJHogpFBh2vJakoSw+5TBJR0DBXTAoFpu6qrIMEYEWiw+MTrfbcPVEVMrmNztLjpSl5ACBYkLA1BWqsYmEiLTCoUcrQLV6JddML4PT4oFbIkKJRYNyAHhFP7mabC+ctDvxcb4dEImFjbgoyUvDilDzYXd5muYKjdZmFO+YGZ2DpiHD1q1RyKdu7TgIEXEdUIbr5kAAiiC5ASyfKBpcXC7cdbpPCeNzJxuNrrNsjVt+n0KBDekocb7HAFVPyYHd78eA737HWGh8QFEitx/RRfXHbGzvZiVLsmLhiUsx1UVpZB4e7cXIX60q+eOIgzNqwD5Pyr8Gi8QPhcHtZi0uDyytY6ZkZY0evGhxJgUHmu3n18+MBArXIoMeSpma04ThnsmPBvw4JBhE/tfkQHhuXJboPPldbc1xm4Y452OIndn2PztLD0D0B7z84HNo4BV65ZwisDk+H/K47GxK/WKOZLorFYkFSUhLMZjO0Wp4SrgTRSWHERSQ3T2Zdr9+PZR8d4bWcjM7SxzRDJ3iyWTe9AA9wxEuwtaUoS4+Sybno000jeGxmmwuPNgXTMiny+WnJcHp8SIpTwOH24lFOQcdIjumcyY6nNh/CfcPSMfu9A4LrbZ09CkPSUwK2Y+oAJahlqLE4sWDzIdRaXSGfa7a5QsQPQ6zPe2vCPf/BMFl6AHiDgJl1XrgzFxn60FYi3M8QOlfcTMDP5hfh1jd2Cu7nq8dvCGgh0tzvQOyYR2fp8ce7B+PJf37Pvi9W46clNYi6ItHM32QBIoguRKRuDa4QWTe9QLTBaKwydPjiM7hPxvM2lgVkbyXHKdC/RwJbxE/o2JI0SpTclYvTdTaY7G42HfnoOTPuH54RUs06+Jj43B+M1aXa7BA9puDYFcZisOKzHwImv5K7cpGfnozjF6zQxsmRolGiwelp9YKEbYGQ9ZGbrbb2NwWiQcANLvEClpHGGcmkkqhcwZFmT/JdIyum5GGBgMW1p1YdcE5sLi+e2nwIq+8fioXjB+Kyw8NeB+G62BPNhwQQQRABBAuRaNNvm5tizDfZMHEPQONExkyS3Ikk3LFwY0KOVluwvvQUrstIwRO/vBYPvrNPtJVHJO4PvglVo5Rh0YSB8Pn9KKuqDzgPXJdVg9ONxDgFFm89jOIPytntGdePRinjHR93jJ2BYDddvEoOpUwKs92Fj+YWot4mniUndA4YIo0zkkklUbmCw8fyuEWvETHXZLxShmWTctDg8sDh9iIlXonFW48EuPCoynPrQgKIIIgAgoVI2AajHCtHS1KM+SYbm8vLWn6euX0AXB5fxHEPP1+yofiDQ7x1WOZtLMMr/3cca39TgHqbO6BAHXNMkWYMBU+o+gQl3p01DCWfHgsQNdzzwFirLlgcePwfB3nqG9Vi6YdHBFuNMGOMlvZuNRH6WY1urYoLl9klwW5KtUIWNgg6kjij0Vl66JqON9KYqXD7TYpThr1GuC41huDfiVAdIKry3LqQACIIIoBgIRIuQJOZnFqaYiw02TB1eyYPuQYDeyexy8Um87P1oeIHCK09wxQgLDLosHWOEf8+XI3DP5uhT1BG7P4Itugka5S8VZ/5zkN9g3BH+Z2VtXjkxv68AqgoS48EtRznTHaY7W5Y7I2NMrVxCkGh2Rb1b5pLj0QVirL02H+6nresQLhximU5Gg061Fx2Blh4InUFh8uedHl9Ubsp+X4nVOW5fRB/tCMIossRLETWl57CTGMmjAZdwPJgt0EkgkEMZrLhIzg+45zJjrkbyzD21e2Y/NZujH1lOx7dWNYoCGwuXLS6BOOWdlXWIT8tGcAV18jOyjos/egIeiSosGjCQCRplFEVj0zSKNG/RwIydPH4ud4uKGqCz4PFIR7bIpNKQs670aDDQ0X94HB58eS/vsdtb+zEvWu+wa1v7MQf/vU9fqptwJGz9Thx0Qpzk2spnDg1h3FBxQKzzYUTNVaUVQWODWg8fyum5GHRhIGi/a6ExslY4oKvn6IsPZ6/Mxe35/RqViCx0H6Za9/qFP/++NyUfL8TqvLcPpAFiCA6OG3ttgh+6uW6oebcaIBaIUNSXKjboKVNNIPdSYwrZFQ/HVRyKWo51ZQXbD6E/afrMXeMIcBVsv3HizD218ERJmaEmXC47j0mWHbh1sNYNTU/rPtDrZCFxPfUWl0w2SM/D1q1+C3Y5vJiljETs280wGx3s5Weu8Ur8PRW/t5iC7eWY8Gt2dhSdhaj+umglEuRoJZjcFoy9p+uD4mnaQsLQyTWp97JcSjISAlwG0YzztYqB8Dst+ayE2a7GxqlDPEqOTRKWbMKjPL9TqJxMxOxgwQQQXRg2sNtwRfXYnN58f0ZE341tA9SNIpmxUtEchNnJpu6Bhf8AJ7ddjjEFfLM+AGCrpIigx4j++mgjRMv0qiSSxuLJKbE4U//byhU8sbMMI/Pz060Yu6PQoMOH5dXBwRlvzglDxaHO6rJLCVeKVgg0WjQ4UBVPfLTkjHtr98GvHd7bqqglam0sg7PyKQY2U/HZr3998eLOHrOzMY/BYug1rQwROMabY5FhUssG5lyaXB58dzHR0N+hyV35UZdYJTvdxKpm5mILSSACKKDYra5sHjbYQxOS8aMUX1ZK8eBqnos2XYYL989uNWe2hkhwmRQAY036dtXChcLjFUTTeaY5m4sC3Fj7aioxQyTXbAC887KWizcWo4XJudiTHZ3fM3TIJWJCZl9owGT39rNigGjQYc78npDn6CEz+9HrdWFeWOz8D839seuE3VskHShQcc27uSO66nNh7BowkB89UON4GRWFHQeemrVKJmci6r6xlgevhT9l+8eHLKfy2GsTBa7G9P++i00ShkevqEfbs/phdtyesHq8OAfD4/El8cuYM2OkwFB361FNI14W6sTfUsQE3DPfniELXAZaYFRvt8JVXluH0gAEUQHpa7BhfuGpfP2mZppzERdQ+sHRi7/5FjEQc3hqk0Dwi0dghGbNLVqRdgKzJdsbjx5SzamDc8IaIlwXXoylk7Kwf6fLuGBoBT4XZV1eOmzY/jr9OvxXFDhx6IsPT56tBAerw8fHarmtaLsqKiFUibF8WoLZnL6WzEUGnQomZwbcMznTHY8vTWwwnaRQY8nbrmSos/X5iNRoPUH932NUoY37x8KtUKK5z89FlL9evX9+Zj7fhkKMlJa1cIg5BplXJxOj5d1JSao5THvRN9SxK7FL47V4Pe3XIuFTSUPbE4vr3uYi5CF9f1vT+MPt2ZjltUFv9+PPilx6NVU5uFEjRVmuwsalRxSiQRyqYTNaGtP2jOrMBZ0GAFUUlKCp59+GvPnz8frr78uuN6bb76J1atX46effkJ6ejqeeeYZ/OY3vwlYZ/PmzVi0aBFOnDiB/v374/nnn8fkyZNb+QgIIrZ4fH7BPlMA8OzEyFoENJdontwZhOIwGpweHDlnYV0yX/1Qg+PVFiydlMPryhOLJ1KGcTEBQJ3ViQfe+Y59XWTQYetsIw6eqYfd7cGTm/njTLJ7J+GV//shxPK0s6IxLX3hhIGCwgsAzHYXlk7KwZJthwO6tKdoFEjrpoHL48OB05egUcmhkkuxeOvh0M9qqgOz9jcFuNTggsPtDbEoebzivcU8Xh9mFWai2mzHJ+XVvJYywI9FEwbixl90b9VJi8+qI9RE9uYBPbD8zhws3Hq4XZrv8hEutu3ExQa2GjjjChUqjsiMnxtXVHXJBqDRwnrfmm9YYX3zgB5YMnEQireUB/wOmQegkk+PCf5+2oKOnFUYKR1CAO3btw9r1qxBXl6e6Hp/+tOfUFxcjLVr1+L666/H3r178dBDDyElJQUTJ04EAOzZswf33nsvli1bhsmTJ2PLli245557UFpaiuHDh7fF4RBETPD5/IJxHrsq6+D1tW4Xm+YGNQfHYfx8yYYFHxwKOBbmJi7kyhNzhVSbHeidHF113J2VdVj28VHMMPZF1SXhrvDh0pF9Yc55vKoxDf3luwezIlAbp4BSJkXxB4ET2XsPDhfMVNtZWcum6DNigesemfPeAfzvA8OxcGs5DlSZ2Lo5ANAnJQ7/PX4RBekp8PqFr6GdlXVYPHFQq7dZ4HP5CLkwvzhWAwD4492DO0y/q0hqDDEw1tGSu3LxVND3HSwOmKB5rlDncm2qNqCUA7c+EgA8cqMB23+8iNtzerXpuTHbXDDb3Nh1orFJ8dRh6azrdn3pqU5Vt6jdBZDVasW0adOwdu1aLF++XHTd//3f/8XDDz+Me++9FwDQr18/fPPNN1ixYgUrgF5//XXcfPPNKC4uBgAUFxdj+/bteP3117Fx48bWPRiCiCG2MOX/w1XHbSmxiMcw21woDhI/wBUrVn56Cq8lSSieSKOUIS0lDna3F+umFwS4t7ixPGVnTCFjYUSFUiYN2B+36F68KnxWViQuGq4IFOonZQ4Tx8NkqtlcXvx9bxVWTMmDw+1jRUGKRoE//mowHB4vFm89HBQMrsON1/bAyVqr6GeEE7mxgM/lIyY0vzhWg6du8/AWEGwPwtUYCr7WdlTU4nSdLSLXsdj5554jIYuZ0aDDyH66VhMbwVYstVyKFz49hvljf4GPgyyL3CKjnaVuUbsLoDlz5mD8+PEYN25cWAHkdDqhVgc++cXFxWHv3r1wu91QKBTYs2cPfve73wWsc8stt4i61ZxOJ5xOJ/vaYrFEfyAEEWOS4sRvIHyxIbEkFkHNtWHq8cwyZvJakvgmTY1ShnXTC1Dy6THB6s7XpSdjelCAMhenx4ej1ZbGiavKhDfvH4pq8xWLUDgBlBSniKqVAnsOeM5huGyxfvp4bJ09itcCwkxMPr8/JF4JaLTuLP/kKH437hein+F0+2C2tf5kFewadYexpHWkujeR9DILRqgUQjRB39zaQEIWs7IqE/acrIPb64PV6YlpHA6fi6vQoMMfbs3Gso+PCD7UzCrk/013RNpVAG3atAkHDhzAvn37Ilr/lltuwV//+lfceeedGDp0KPbv34/169fD7XajtrYWqampOH/+PHr27BmwXc+ePXH+/HnB/ZaUlGDp0qUtOhaCiDWxyqpqLuGCmiO5yUbSo0nIksRMmmdNdvxUZ0OflDi89NkPIXEvuyrrIAWw6bcjACAgjiKY5DgFm3EzabATaoU0IEZm7hgDb2yNRinDovED4Pb5UHvJiadvH4DlShmsDhfiVUq4vD6ctzhgc3uhj79SGNLicMMjMNmHS31OTVLznmPuxPSvR0YKu9EqarH0jkEoMugD+ksxGA067D5Zh55a/s+JNVyr2IkacctUR6t7Eyzg1AoZPi7nD4YHxMUtVxyI/caTOQ84fBYzrlVIqOVKcxHKfCutrMP/ODxhH2o62vcnRLtVgj5z5gzmz5+Pv/3tbyFWHSEWLVqE2267DSNGjIBCocCkSZMwY8YMAIBMdqXuh0QiCdjO7/eHLONSXFwMs9nM/p05cyb6AyKIGBOuCm00k9YFiwM/VFuw99Ql/HDeggsW8S7mDMyNf/vvb8Sn8wrxz0dGYtGEgdAoxevsMIRzoyU3ZczwwVg5AGD2ewdw8bKTN+gXaLR4SCUSSCDBdenJvOsYDTqkpcTho7mFSI5ToKBvN6z+T2WAAFlfegozjJko5FRf1ihlWD/9enxaXo1bX79SdfnpLeWIVytRbbajosaKepsbW8rO4ol/fo9j5y9j4upSTH5rNywC1gCmwnaRIfLvN3hiEhJXDFV1NiydNIi3mvRMYybWl54SfVpv7nUTjmiqfncUmGrfQ9JTkJqkxqEzJl7xU5Sl53XBMnDFgdhvPEOnYZfzVYoWsgrForq3WAJEONctgA75/fHRbhag/fv3o6amBtdddx27zOv1YseOHVi9ejWcTmeAqAEa3V3r16/HX/7yF1y4cAGpqalYs2YNEhMTodc3Xii9evUKsfbU1NSEWIW4qFQqqFSqGB4dQcSGWFS3raprCOlNVWjQ4YXJuUjXxYfdvsHlxcJth5uV7RGumGCGThPWyjF3jAFGgy5su4ALFgce3ViGlVPz4QOC0r51mDMmC9q4xiKOJ2qssLm8IZNHQPPV8QNxqrYB/bvHY/nHR0Oeeksr6/DMlnIMSU9hn84ZYbFmxwm231jZGROKsvQhEwqT+jzD2BdPjx8Auyt8CjV3YtIoZUgM47Lz+v1weX0BGWlMNWnGesFMyMHxHkqpBEs+OhJQSyma64ZBKBuqpdbF9kRs/C9MzsXSj47wbscn7sR+48xn8FmUmtM/LNK0dTHLbTjX7TVh7gkdKXW+3QTQ2LFjUV4emIo6c+ZMZGdnY8GCBSHih4tCoUCfPn0ANLrRJkyYAKm08UsZOXIkvvjii4A4oM8//xyjRo1qhaMgiNanJdVtL1gcvI05Syvr8PSWcrxyzxD01ApbYFva4DRJo8QLk3NDUnmLsvQomZyLa1I0YT+TcVnFKcStTj20KuSnJ7MCZlZTLZ4eWhV+qL6Mvt2uiC19ghLHL/C7YZjmq6OzumP2ewfw0aNGQZN/aWUdW/MHCAzuZrJ11peewodzC7Hkw8MhQaP3D8/AoxvLsPr+fGT30oYVlNyJaVZhJiSSRnHHNz4mQFchl6Ksqp53smQmZKF4jxnGTHxz8hJr6Yj0umE4Z7Jj8bbDyE7VIj8tGdVmB2o0CqR30+CaFE2rtK5oK8SEy9JJOXB6Ihd3Qr9x5jNMNneIiI62f1g0aetiltuyMyZRt+rH5dU4dMbEu99w10Nb024CKDExETk5OQHL4uPjodPp2OXFxcU4e/Ys3n33XQDAjz/+iL1792L48OGor6/Hq6++isOHD+Odd95h9zF//nyMHj0aK1aswKRJk7Bt2zZ8+eWXKC0tbbuDI4gOgli38dLKOtQ3uEQnsubUAuJyzmTHsx8dCahmnRynQIbuyg0v+IlQLpVgRL9umD82Cx6vH/EqWaML2w/RG+9Xx2owIa83Fk8YiAanFxqlDDKpBDKpBL8c2DOkaGNimD5cCapGwWV1RNZXjIGJg+BmcVnsLlErTPdEVdjq3mabC3EKGd6aNhRqhQz6BCX+c7wGiycOwrMfHeEtMzBvYxluurY7lkwYhGUfhxZ3XHLHIJjtbjyz5XDIeS2trIMfYC1Z3OXhrhtmvIu3HeYt5llo0OHFu/LQpxu/BbCzEE64xELcMZ+xIsjiFE3LFbEHmQWbD2HZpEFI4RyLmOX22DkznrjlWvjgF7zmbC5vyANSpNdDW9LuWWBiVFdXo6qqin3t9Xrxyiuv4Pjx41AoFLjpppuwe/du9O3bl11n1KhR2LRpExYuXIhFixahf//++Pvf/041gIguSbhu4+Hfj64WEFfMJKjk+O50PY6fv4ypw9LRQ6uCzemFPkEFk82Ni5cvIVmjwqKtgZNvUZYec27sj9+s38taHooMOjw7aRCWThqEJR8eCbEmLb2jsSikWHXcYKEVp5AKCqoigx6+Jl2ToA7fVywYRuQw7KioxfdnTLxWGKNBh/Kfzbg2VSvotmBakkgkEhyttrBVrWcV9oPL48WE3FRecXVdegp+PH8Zyz45hlmFmZhZ2A/xShm8Pj92n6zDxFWlWDU1n/ccAFfEXDDhrhugUTxnp2p541RKK+tQvKUcqztJvZjmEOu+ZMGiKkUTeZKE2IPMzopanLjYgA27j2BFk9VGzMW3eOIgVNXa8PtfXotnxsvwU21DiKAHQh+QOuL10KEE0H//+9+A1xs2bAh4PWDAAJSV8ae3cvnVr36FX/3qVzEcGUG0HbH0kYfrNh7+/chrAfGZ2I0GHd59YBiWfXwU35y8hJVT87Hy6wqUVZmwamo+Glw+TB2ejpmFmWw9n50VtfD5/fjt6H7w+PxsjZ4LFifkUuCFO3Pg8vojfrIOFhDM5yy9YxDmjjEA8AdaRgw6zB1jwLFqMwCgxuIUbVjKDXhl6gpldNOgrsGF9TOux4Gqevztm9PY9NsRWPKhsKVm1dR8frfFvw4FCBRu2r+v9CSMBh1G9e+OV/7vOPZX1bN1jVZNzUevJDW+OHqBDVrXJyjxx88CK12Hc6Uw73NrJkkAnLhoFb02LQ63aJzKzjboQn+1ESyqIo2jiiQbc2eQW1vIivVTXQN+/fZeAMBb04ayVbD54F7PHfF66FACiCA6Im0ZtBfr8vJi3cYLDTqkxIsfR6Sp+EIm9l2VdVi09TCGpKdgYO8kvL3rFMqqTFg5NR8bdp0SrOezq7IOT92WjRf//UNIt/c5YwwY0CsxpFDeBYsD9Q0uWBweaOPk6KZRwuPzCwqIpzYfwlvThuL2vN6YwbGe1Fx2IqObBpn6eHw2vwg2lwfL78zFwq3lAecxuCmqULG6IoMe784ahgaHm3WDeXx+XJOshkImhdPjw/oZ10OrlgfUIWLPaWXoOQWuuKZmGTNx/9pv8PAN/bBowkAs+/hISLG89x4cgRWfHUN+WnLAOdcoZeiTEod10wsCmu1yC0uq5FLoE5RYP+N6HDnbKAovWp247PRg76lLuPEX3XmrSWvVClSbHezncAtOMp/T4Owc9WI6KpG62iKtZh1steGzYiVYXSHbCcF9QOJeD0K0df0gEkAEIQKfICnK0uPZOwZBAnGXS7S0NOCYj55aNdutOnjyfmFybtg4jkizdcRM7NxA4dVfV2LuGINojzNmYq+xOHl7WPngx/N35gScC75Mt5LJufi0/FxIgDCzzn3D0jHj7X3497wieHxXLEoFGSnsvntwzs9LvxoMs93d2N5C3dhs9LmPj7JCQaxDvf/TY1g2KQerv65km5RednhC0vCLms5r7+Q40XMa7Jqyubxwe/1YKlCgbvnHR3Bd324B1pyHb+iH2walhsQGcYXo0PRklJ81Y8PM63HZ4eGt/pupj4dGKeOt5n3BohCtYvyroX14jy8WdKRso9YkEldbNNWsw4kQ7r7C1bLiuuGY60GMtq4fRAKIIAQQEiQ7K2qxuKnZpVC2Q3NoacCxEOm6eLxyz5Ar1hG1HCnxyoiyeIDInjIjMbEziJnBhWJOgtdpcHnZCsZCmW49tKqwBdtsLi/qbS4MSU8R/Uyg8Tww3/PpugY8+9ERDOydhKnD0uH0+JCpawzgZGJxuNaU0so6mB0ejM7SIy8tWbhJaVNQ6qIJAyNulZEUp8Bn84vg8/uF3QuVdfjDbdmou+xiBUmNxSEomABg4fgBGNQ7CTIpIJVIQsQad93lk3JwsraBFRoOjw/1DS4oZFJsfmRkSDd6ZtvF2w63St+oq6FRZ0vgE3982Zh81azDiRDuQxGToQkElp3gc8MlaZTI0GkELdLtUf+JBBBBCBDJE/jqrytj1vzPbBcvXBZJATIhemrVEQsePsI9ZUbTMDKSmJMigX5eDJcdHjy6sQwvTsmDxeHmfQKNNLYl2qfOcyY7nm4SXEyNHI1Shi2zR4Wkm3OtKdVmO56blINzJjvs7tAaRAw7K2pxpqlDuBgquRRGgw6nahsw+hfd2a7iQtRYGtv9LBw/AG/vOoVZxkzRZru/G/cLTF37Ddb+pgCJarnoupedHkx+azf0CUps+u3IgJT/ddMLBAtYtkTYC9EaltTOhJDV+tExBsw09sUTN1+LuobGayE4cDlSEcJ9KGpwuvHCnblweX1ocHpE4/JkEgmevn0ASj79ISTxoT3qP5EAIggBIrVqxOomrlGK/xwjrb7cHoQreMiIGaNBFzZuIDlOgSV3DMIdq3cJriOXStgJ7dGxWbzrhPsclVwqesMXcqGY7aGCa1ZhJpZ/fFTUraeUSXGpwQmlXCrYK4qB27OMT3gYDTpcsDgwb0wWeiep8ft/fo8Zo/qK7hNorJKf2ycJT285jGnDM0TXrbnsxHXpyZBJJLA5xUsBMBPoiil5IfWOoq1X01Jay5LaGRCzWvv8fgzL7Ia8a5KhUcoE3a+Rnptos9zMNhee3HwI+083BurPMPYNiLtrj/sbCSCCECAaq0YsbuJSqUR0wpNJhdu5tDdisULL7szBcx8dwZ6mLLALFofgcRYZ9JDLJNAoZLguI4V3Iisy6FHa9PS4o6IWxbcP4B2TWHyC0aBDzWWn4A1fzIVi5fmuw7n1Zt9owJ6TdbhlYC9YHO6w4qxHogqLt1XhxSl57D4Yigx6PDdpEBweL5LjlPhD0zgHpyWHLYo4MFWLn+sbm7+GG0NatzgsuHUAnF4v/OIdN6BVK/DWtKFITY4LOd/RBMpyMdtcqGtwwePzw+f3w+b0IEmjDBvLY3G42Rinm67tAaBRoClkUuysuHhVB15HYrWe8/4BPHxDPyy4NRtAY0PcZI0CPRJVUQvDaOKsuGPj+60M69uNLEAE0VGIJnAwFsF7CqkET96SDSmOh2QtzTRmdmgBBIjHCr1wVx7qG1ywOj0w6OMxsr8Oi7cdCYlHmDPGgPQUDXolx+H5O3PwzJbykBT1mYV9Mff9KzELbq+PV+isLz2FddMLIAVCCgAum5SDFI0ion5bDEzRuGduH8CmtzPZUuGsHDKpBMfOmXHLoJ7Yc7IOvZPUIiJQhx8vXGYF5X3D0rHg1mz8XG9n661csDjQVxcPp8eHGaP6YuqwdMQr5bgr/xos2hZacZqbas8QTiB+dawGt+X0wqGzlyGXSoRrJmXp8fnR83j9ywpsfGhEyPvRBMoyoscPoOSTo7hveEZIYHm4WJ6kOAXevH8o1AopVnz2Q4h4vHPINbzbXQ1EYrW2ubx47YsKvPZFBQBg6+xRyOqZGPVnRRtnFW1NsbaABBBBCCBk1WAmlKc2H8LcMQaM6qeD2e4KWxclHAqZFCu/+hGD05NZ83BSnAKJajnWbD+B5yfnxurQWg0hszhfDNLqqfmoueyE2d74xB6vlCOZI0rSdfF4+Z4huNTgwilOsbW57wd24PZ4fXigsDFwmjvZ5acnw+H24bq+3dg096Q4BTL18aKBsOGKxp012fHAO98FxPeEs3L4/H5MHZ6BBqcX60tP4c37h2LuTYaQMRdl6fHkLddi1oZ9uLZXIn49MgNurx9ML2emqbNGKYMPwMKtgQJx3IAeWHj7AFRbnHC4vUFFEZPRPVEFr8+PIoNeMIC1yKDDdGMmjpwzY/nHR7G/ytQ4Xr6aSVl6PFCYydaC4SsaGWmgLDOhDk5LRllVPfLTU0LEj0YpQ15aMn6qbcB5s53XIhSvkgsHmVfWYtG2w1dtAcZorNYMzXl4a06cVTQ1xdoKEkAEIQLXqmGyu+D2+NDg8kImlWD11KHwwY/dJ+pYS0BzM03MNhebocFtPgk0xtAEu2quhhTfSGIIemrVUMulKPn0mKAl7ofzl5GaFIfr+3bDLGMmXF4f+qTEofxnM+a8f4AVS0zqfzT9tvhgrD3c+J6yMybB7JYigx72piarm347AjaXF3PeP4A5N/XH0jsGwe72osHphVwqQWllLV7/8kdWeC+ZMAhLeer63HNdn6YaQYGf9+WxGjjdPtyem4riLeUB20w3ZuK+Nd8AQFPDWD9v37SvjtVg3sYyrP1NAWslYNwmT96ajSfR6FJK1ijw2eHzmP3elXPMVzSSaTC7cPwA/G7cL2C2u3FNShxSNFcyEZkJdf/pejx5y7XIT2sUa9zjFkqnD/7NWR0e9NSqRYPMz5rsqG1wdcrfjRjRWK2B5mdeNSfOKtKaYm0JCSCCCAN3ov75kg3FHxwSrJvS3EyTcHV0HO4rLpa2TPEVE1ptJcLCWeLmbSxDfnoy8tNT8MA73wG4Unhv8/+MwuUoU/+jeYreVVmHp28fAJlUgnuu64Nnth4OcetNN/bFo5xMG8YdZHf7QipDMzg9PqyYkodlAmnqP9fbha1UlbV44pfXYuscI7w+H2RSKaQSoNrswJ+mXQePzwe5VIoFt2ZDKZOiwemBy+vDriAhzz3OYLcJAPzj4ZF4/cuKgM9esPkQ3ntwBJZ/fDTAXZafnoweWjXb3mTd9AK88Mkx9pqttbqw/3Q9Vk7NZytVvzVtaMC+heosBf/mLA53WJfkT3U2zH7vwFWXGi/0WynK0mPOTQbM2rCPXSbWnDUczXFnRVpTrC0hAUQQEcJaacIU8GtOpkmkN5S2TPEVElorpuTBD0Qsws6Z7LA43LDYPUhQyRCnkEGtkPFWDxaid3Iclk3KQeVFK28zUb7CgIfOmPBQYSaSUsXPR7CQS1DLo3qKdrobXWv1NheWTRoEk92DarMdSlngGG/L6QkpJFg0fiDOmuzooVWLBk4/dVu2YB2jcFlk58x29O8eD61ahT80VcLmWlCCXVEvTM7FXflqjMvuwcZu1VrFyzLwtVGptbow7a/f4J2ZwzDD4uD9roBGgce9Zi0Od4jACXbXFDTVauKrs8T9zWnVClxqEB87t/Lx1ZYaLxSLBwAfzS1scXNWoPnurFg2iY0FJIAIIkKiqcwbbUBfpDeUtkrxFRNa//3xIj49VB0SEBs8mZhtLtTb3FgUFKdiNOjw6E1ZcHt9SNfFRzymSzYXa+Hhg/vUH+lTJZ/Iu3lADyy/MwcLtx4WtDhxcbi9uOtPu9nX3KdtZsK/Pacnnrw1Gwu3HkZZlQmzCjNR2F8vOrYGkdTzcDFHakVjLM6p2ga215rf78d6AQvK01vKsWpqPvp1v9JexOHx4f0Hh8Nkd4e0yBidpUdKPL9Lo9Fd7Bb9rnokqqBRythrVqtWhGTScYOnNUoZUpPVKCsVrrPE/Ob0CUrs/emSaJA5V8RejanxQu7lWB1jS9xZsW4S2xJIABFEhERT7TjagL5IbyhtlUkhJrR6JKoEu4czk0mDy4uqSza8+XWFoMXs0TEG2N1e2FzeiFxo4URiP308ts4eFfFTpZDI++JYDQDgj3cPhsXuxqWmVOw9J+sCrBhAY1zR7pOh1ZwB4N/zilBvc0Ebp4BSLkXx5kNsH7S3d51Cflqy6Pi0ccLHW3bGhKIsvWCZgFStGpcaXNh98opb670HhwvGxeyoqMXPJjuqLY4rPdR4GtuunJqPv++twnOTcuD3+fHoGANuy+mFHlo1a5Wpszb2UhMan7Epy40rXDL18fipriFgPW7wdH56Stg6S8xvLkmjxI2/6I5MfXzAOsy5Cc4iBNonA6kz0xHdWc2BBBBBREiksSGRBvRxXS9JcQq2Z5fYDYVvDEy8S0F6CtRKGX44b4m4ZooQYkIrXHyF2e7Gsx8dwR9uvVa0FcXsGw245fWd7LJw8RjhRGJqkjpmcVdfHKvBU7d5IJVIYHd54fH7UVZVHyB+irL0mD6qb4hFCGgUQS6vD0PSU3DOZMdPtQ3YWVkX0ActPz1F1Epx3mwXfP94tQXLm8oEBPd4m27siyl/3g2byxtgIQlXSfx0U1xMyeQcfFpezStcpRIJXr57MNRyKZ7eUo75434RsK5GKcP66QV49sPDmD6qL3x+v2BKfn56MitckjRK9EkJ/N6Z4OlZhZm4ZVBPUXfhozcZoJRJcPScGQ0uL5LjFOinj8cLk3PR4PSgwemFRAKUVtaGZBEC7ZOB1NnpaO6s5kACiCAiJJIMi5a6XkruyoXT44PX54fX54fN5YHNfaXvVfAYmLiO9789jSFpyVgW9JTc3CBPMbEXzv2iUcqws6IWs280iK4XPCGHi8eI9VNnJNY0PwCb24vF2w5j/YzGhqBme2Mhw27xSkz767chkyn3+Bgr09Rh6QACCyYKpYczGVt/+NchvDglDyq5FAN7J7Gd1FM0CqR30yBBJccz4wfiUoMLXr8fMokkxErFtZBEUhkbaGwAKyRcd1bUwurwoEECPHxDf1y87MTU4Rl4sKg/PD4feiSq8dJnx7Czsg67T15iM8yYEgQOt5cNCN9VWYc5NxrYh4VeWnXI78vm8mL115Wi1jKNUgZ9ohpPBfWDYyobD+ydBLPNhUebkhSCaa8MpKuBjuTOag4kgAgiApgCbUvuGIRnPwws4FeUpcfSOwYBQGPQbQtdL0smDkLxh+WCAcZcEcAEjvLVTAGaH+QpJvZqLjtFLTFKmRRvTRuKJE30NUl2VNSi2uIAwB+vEMunzkjjri41uHDfsPSQonrrphcIih+gcWJmrExMmwqu9Yxr4WBEQp+UOHx+9AIrYp7afAh/nX49Xvm/H0JSv1+YnAu5DEhUy6GQSXHbGzuDhwDgSnxauMKHzDGFa8ly2eFGYpyCzdbiBlfPMmayy2YVZrKiTa2QYc/JOhSkpwScM5VCyn53YgI32DrEZVZhJpYGtd8AGsUa99q/Glw2RGwhAUQQYeBaa5gb+//c0B8qhRTJccqoJ2Ax18u1qdqQNHsgUMhwRYDT0/iEzDRm5aO5QZ7PjB+A6fV2SCQSNgD2uowUDM/shpH9dFi8LTBIuChLj9k3GTBhdSmb6hyuNQMfJy82BKRIBxOrp85I4672/nQJI/vpQs6vmKAYk90diSo5Ljs9eGvaUPRIVKPIoA8RfYyFg2Hd9IKA1/cNS8cf/+8HfmH7wSEMSU/B6q8rQ1LGg3F6fKzFSQpJSKXxJRMH4cV/H8PXP1zEuukFovtSK2W4YHZg6vAMzCzsB7/fj/e+PY1dlXWYNjxDsF6P0aDDxLxUaJQyVgQlxwV+j2IZTELfFd93wz1PzLV/NbhsiNhCAoggRAi21jAT1uqvKzE6S9+s9Fkx14tYTynuzZz5K6uqBxDbhpNC3aQ/nGPEJ4erMWFVKQBg0YSBeGb8ADQ4vdAoZXB5fahvaExnXl96Cgs2H8LGh0bguY8Ca8IUGfSYbuSPnQEaLUNtkZ4cqVXgRoFO60IurDHZ3fH07QPx9JbD7HFrlDKsm16An2ptom0hai47A5aJXQ+llXWY2ZR5GIl7y+byYtPeKpRMyYXT7cOlBhd7LTLiBxAXdkUGPcpO16N4y+GAZdONffHNyUtQyaWC9Xp2VdZh+cfH2HIRQq4nIYErVN9GJhFvEcO99ju7y4aILSSACEKE1kg7F3O9RCtkmH01t+FkMGLdpJd8dAT5HBfGso+PYv2M6/Hm15UhFgUm8Hbq2m/w8t2D8cyEAbhs90CjkiFRJcdzHx/ldR9xLUNtkZ4ciVUgNTkOVqeHfc1177i8Pjx+87VYeoccJy42QC6VIF4lx9IPjwScE5vLiwfe+Q7LJw3CC3fm4qzJHpBefrzagucm5QAItHSEux6Y98O5t5j37xuWjiXbDmPJxEFI12nwxD8OYqYxM6D6uFiLjNk3GfDAO/sC9r+zshY++NmK2GIWmZ2VtZhh7Nss1xPfdyWXSnCitkF0OwpwJoQgAUQQIsQi7TyaQnvJIqnPQOjNnHHjRNNwUoxoah3NKszEqq8rRFOTV39diRlv78O/HhmJX/RMwE91Dbj1jZ1YOTUfdrdXMEOIQej8xrIKdSRWgR6JKozO0uO7pmrFwe6doiw9nr1jECQAvD5/gOWHK5bSUjTYfaIWyz45xgrAoiw9SibnsoUhX5zS2DjW4vBEHEclLFr0WDRxINvBnoktcnrK8do9Q/DC5FxU1FgD9hkcm6RRypGglsHvB+5b8w2vcGWujUc3loWtb5QUp2i2ZS/4uzLbXLhgccTk2ie6HiSACEKEljbwi6bQ3ugsPTJ0mqgKjDFunCXbDrPuELGGk+GIptaRmHsmWCylNE1cCVZXwAT71G3ZOHPJzlstGOA/v23ZCoSBOc/bf7zI697ZWVGLpR8ewaqp+Thx0QqNUobfju6HsQN6oMbihEQiwZFzFjzxj++Rn57MWshsLi92cgoRNri8WPCv79m4qbljDII9xrjWMqFzesHiwNl6e0BPNKDRfVbX4EJ2qjagzQoDNzbp/QeH4/+OXMDAVK1o0DfTadzmFl4HuHItxALRmj8U4EyEgQQQQYjQkoqnkRTaszo8Ia6XaLNVeifH4eW7B6OuwYVnJw5qSp/3Iiku+iDPaPpgReqe4Z4n7vlkJtiyqvqIn97bshVIML2T41CQkYLiD640GQ3Odqq2OJASr8Sb9w/FutKTAb2yuK5B4BRrIWPGX3PZiRWf/YDB6Sls9/p4pRzjBvQE8ENIvZ8ZQdYym8uL78+YcOeQ3ujeVGm57IwJyznWJi4WR6Nbr6dWJXiNFxp06JMSh/Wlp7CqycIkRFKT9fJAVX2bWmRSk+OgUcrwwp25aHB52Gu/R6KKxA8hCgkgguDA51ppbvpsJIX2+vdICHmvOdkqbZEZFZy5FUk7BiZdu67BhZO1DdDGKVByVy6e/fAIvjhWg017q/DX6ddDiuMBMTNC57etWoEIERwLJOQOm3NjfxyoMgVsG+wa5FrIgMb+XvcNSw/Z35js7vjDrdl4RibFT3UNyNTHI1Elx5IPj7DCRqOUYdH4Acjpk4SKGitUchm6aZRsFWg+mF5eYg00SybnIlEtR0FGinhwdJYemfp4fPX4DbjscGNK/jVYHFQuojVTzim4mWgOJIAIogmx5p/NSZ9tSfxQe93Qo+kmfcHiQJFBHyBcGIvIyH46qORSLBw/EHtO1gVM1kVZejw3aRAWjx8It9+PFz45isHpyZhh7Aunx4fkOAUydBo2JoYRpZcdLoQxOkWc7cbs02x3QaOSQyqRQC6VQBdFOw6hbKedFbXw+f0BFh4Grmsw2IIWr5Tj9S9/DNnf1z9chNPjw4Jbs2GyuZGqbax4/crdg3HB4sRlpxspGiWWbDsckp21bnoBHnjnuwARxIglqUSCsqp6VugLXeNmmwsLJwyAx+vH+NxUXGpwweXxYX9TaYSCjBSsmJIX0tx2NaWcEx0cEkAEAXHXyoIm1wqftUaMlsYPtRditVg++J9ROFnbAJVcisPnzPj9LdfCDz9KgwriBdd/ef+hETA1uGBzexuL4p2og6FHAlZ91ZhB9iUnCwkAW2KgweXFgs2HsL8p+DiuqcmnEJGcUz6hywRgl3x6DEsn5YTEEnEF08aHhmPXiToMbarBw0dwDBQXRvhwLWhGgw5KuUSwV9euyjooZFKMyOzGigir04NaqxNxShmWbDscUm9pZ2UtIAEWjR/ACiMmHf+t/1QGiCUmhir4Gj9nsmPxtsOsZSo4wPqTeYXoJtJ4kwQP0ZEhAUQQaB3XSoJaLtpNuyNnpwhNXg6PDxs/Pcaeq8HXJGNIegpmGjPRLV6JVz8/zpsVJsEPbNE+oHHCXzYpB/ub6hgFw8TEPPfxUeysqGV7aIn1z4rknAoJ3V2VdVDJpZg2PAM/1TbgvNnO9lJjRBh3m0KDLmy2k1CMlEouDXAnMuLrQlMFbCFO1TZg494qvDg5F34ACz44hF2VdVg3vUC0dcUztw/AR48aYXV40T1R2ZSiH75i+AWLAwv+9T0GC1QZ31lZi0VbD+OVe4YgSXTkBNExIQFEEGi+u0ooHVvIysDtpt3Zno4Za8D0UX0Bf+MEaHNfyRYSm4i5RfuARsGxZNsRXjcRg9nuZs8fk3HGdFNn9sEQScaP2eZCtdnBK3Q1ShnuH56BDbtOBRxDyV25+PTQuZDjKm1q5ioGX4yU0aBDzWUnlkwchFO1DVg3vQBlZ0x4avMhvD3z+rD721lRix0VF/FpeTV7/OGC0U12N+5b8w0A8e+IK/S5DVxniFQZL62sQ32DCz21atExEERHhAQQQaB57iqhmKGSu3Lx1AflvFYGppt2Z5swuJaT3Sfq8N6DwzHD2BfdE1XsOpFmhTEwRfGE0Cgb3V36BCXSu2nwybxCWB0eaOMUeP7OXNTbnDhvcUIpkyItRRMSg8KF+a6YpqTBCMXz9EhUCQqG3SfrUJSl5xVURQZ9iEWnKEuPZZNyIAVw68qdAXE5c8cYUP6zWbQ7PGMxCm5WGi4YXS69Uik5kkKbwQ1cw23DZJMRRGeDBBBBIPp0d7GYodN1NkF3GtNNu6c2dmNvC7guQpvLi0sNLjzwzneYO8bATtqRdhuPhNFZesQr5dAnKPH+QyPw7EdHAoRBoUGHZ+8YhMt2Lx55bz/++psCJJjkvHWAuN/VjFF9eRt16hMaM6aCEZv815eewtbZRiz7ONClZDToMLOwL8rPmvHeg8MhAaBLUKJXU/Cy2eZCQUZKwLWWn5aMRzeWCXaHXzRxEO58cxfvmMK1rijlBKmHzdxTymCyubH/dD3bwDXcNkw2GUF0NiK/IxHEVQyT/TQ6KzCuoznp2CZ7y6tHdzSCXYTMpLu+9BRmGjMD2i3wIdT8tJdWLXjOkzUKrL5/KJYGiR+g0fXy7IdHoFRIsHD8AOw+WYenNh+C2eYKWI9xe00dlo71M66HTCLB2zOuR1lVPR545zvMfu8AZm3Yhz9+9gNWTs1nrU4MYpO/zeXFOZMd1/Xtho8eNWLd9AKsm16A/PQUzH2/DN/9VI/eyWpck6yGw+1FbYMLZpuL91pjigjO21iG/PSUxkDlaUPZ/Z2qbWAtRsFj4n4HXIqy9JhZ2Bdrdpxkl4X7jj4+VI2F2w5j9f35kEqaen+JbFNo0CElvnO5cgmCgaQ7QTQRTf0dsZihWPXl6kgEuwivtF44xVYgLkhPwR15vbHsk6MhAcPBRfuY5Skaheg5T4pTCGZGlVbWYcFt2cjtk8QW++MGq/O5KF+YnIN/c+JnGHZW1sEHhMQklZ0xiXa0/66qHqu/rsSaHSdZq9LAVC02/88oxCtleO6jIwEZbtyK1X+8ezDb8kKrlmPuGAPWl57ijbfhdmivsTgCxsStAj37RgMUMimsTg/66TV4PqgIIvO9SZrOH/dYmDYkNpcX8AMTB6diVmEmLl524I7BvbHs48Dvtcigw/LJuZ3OnUsQDCSACIJDpKm7YjFDZWdMgrEhHT37S4hgFyF30p1zowFqhYytPB1c/0WtkOJZTh0g4EqRvd4pGgAIOedscHkYa5rV4YXJ7mL3zVjXhFyUPYPiZ7jwpa6vLz2FbXOMIS64IoMO0zmijts6AgD+8fAImO0uTB2egf83si+kEgm6xSvh9vpwtr4BHq8Pz2w9HCIUuS0yGIIzxgoyu6Ff9wRAIglwS35fZcLNA3rigXf2YWCqFqum5mPppBw4PYcCvre/763CS1PyYHZ4cKqppAETiM11DfbVaeD2+fG3b35C90Q1bsvphRmjGms1qeRS1Fic8Pv8qKi5DJkkfA0lguhokAAiiGYgFjN0vNqCksm5eHpLedTVozsqfAUSbS4vDp0xYdqw9JAA5OBjfPnuwawoilfJoZRJUdfghNPrC2lkyrXccC0ffCSoZbC5rljcGOuakIsyXEBvMPnpyahvcCE/PQWzmtpTqORSdE9UCTYGBQCb0wuP3x8SWG006LDg1mw8vaU8pL8X85prhRqX3R3PTBgIq8ODwv56xCllqG9woW83DVZPzcd5iwM/19sBNArvqWu/YQsTNop5CNZ0Onb+Mma/dwAAf1XruWMMKKuqR356Cv5aelKw0/z43FSkJsVh/+l65PVJgs3lRULTd2yyu5CgblmzWoJoLTqMACopKcHTTz+N+fPn4/XXXxdc77333sNLL72EiooKJCUl4dZbb8XLL78Mna7RR71hwwbMnDkzZDu73Q61mky1RGwQ69n13KQcpDajnUVHpzktOhgYy5pYI9P4pgDchVvLWStNuABfmUSCZI0Sc8cYcLzawk7uQi7KcO7JpDgF/vXISJibLE9lZ0ywBll3gEZxkJ+eLCgKvDziB2i0Ml12eHibmwKMW28ACg16+P1+xCnlWLj1cEiD2xdZgdMYXF1rdUEXr8TkIdeEfCd8Vs0TQR3g+bLgmNIDs0TS4HdV1uGhwn5QKaTYdvAsnvzXoYDzMNOYialrv8V1GSlYNmlQTBuhEkRL6RACaN++fVizZg3y8vJE1ystLcVvfvMbvPbaa5g4cSLOnj2LRx55BA8++CC2bNnCrqfVanH8+PGAbUn8ELEmnCC4GivhtuSYwlXbvj03NSTt/EqsEUKywKYb++JXf94Dm8uLQoMOL0zOZccm5KIUE1RGgw7//fEi+5mMIOBmugWPSyiWRiIRrupsDuPWO3PJht//83usmpqPt/5TGbKf4KKF4b4TvlpVVqc74FwwYocLYy0LZzVL0ijwskABTOCKRWvR1iO4Y0hvGA163mw9gmhr2l0AWa1WTJs2DWvXrsXy5ctF1/3mm2/Qt29fzJs3DwCQmZmJhx9+GC+99FLAehKJBL169Wq1MRMEw9UocloLscw5JkXd4/M3Wlc4KerlZ024vm83zDJmIlGtgNvrw76fLuHIOTNWTc1n19t1og5JcYrGCs4CLsr1pafYbLDgXmezbzTggXf2BcQ3zTJmQiaRYPKQa/Dsh4cDAo837a3CM+MHotrsQJxCBpvLg7IzJszbWIaX7x4seB4iKRfAuNYiKVooVIwTEK5V9dykHGzaW4UXpzQ+dPKJHGacYesMycRbeDBxVTsrazF/XFZIxWmCaC/aXQDNmTMH48ePx7hx48IKoFGjRuGZZ57Bp59+ittuuw01NTX417/+hfHjxwesZ7VakZGRAa/XiyFDhmDZsmXIz88X3K/T6YTT6WRfWyyWlh0UQRAhCLmlmLo8PbUqqBVy/G3PTyG9xGYaM/HoxjL8/eGRmPGXvayF5rUvKgLWG9lPx4pSPhdlQUYK+nbThFju5DIJTl5sYN1aTFAz89n3rtmD+4alYwYnDsjQPQHJGgWUMimcHi+mrv2G/Rwx0dCYWRbYRJZ7DEzAcyRFC8O5FIUsbou3HcaCW7Nx5JwZC27NhlwaOl7GQhTOalbfIG7R4h6Hx+dvdmsZgog17SqANm3ahAMHDmDfvn3hV0ajAHrvvfdw7733wuFwwOPx4I477sCqVavYdbKzs7Fhwwbk5ubCYrHgjTfegNFoxPfff4+srCze/ZaUlGDp0qUxOSaCIPjhc0txg28BoKyqXtCVsmj8ANhdXsGqzbsq67Bo22GsbrIuROKiZDDbXHj/29MBwc49ElX48cJlNiuLK8qYZq2M2DLbXAEWJzHRcPScGU9wmsgyBJcLSI4TL5cQr5ILCpynNh/CogkDRfvbPTN+ALYePIvXvqgQdfNt/PY028YkOKB7pjETHp9fdJxcMcjUWeqMtbCIq492E0BnzpzB/Pnz8fnnn0ccn3P06FHMmzcPixcvxi233ILq6mo8+eSTeOSRR7Bu3ToAwIgRIzBixAh2G6PRiKFDh2LVqlVYuXIl736Li4vx+OOPs68tFgvS0tJacHQEQQTD55biiplwwbZP3ZaNyw4Pb7wKw84g60KkLsokjRJLJ+Xgqc2H2H1rlDKsn3E9rstICRASNw/ogWfvGIRaqwsnaxuQoJJDJZPi2TsGYcmHR7CzolY0Ruj+4RlYs70SJZNz4fL6WXGWoJajwenB+w8OZ1+LVSdXyqSiAidcQc6f6+2soOGLtbK5vHj/29N4bNwvUN/gxpO3XIuF4xuz0BweHw5U1WPexjIsHD8gIosWt5hiZ6yFRVx9tJsA2r9/P2pqanDdddexy7xeL3bs2IHVq1fD6XRCJgusylpSUgKj0Ygnn3wSAJCXl4f4+HgUFRVh+fLlSE1NDfkcqVSK66+/HhUVFSHvMahUKqhUKsH3CYJoOXxuKa6YCefy+bnejiPnLMhPSxZdr7nWhd7JcXhxSh7MdjcsdjeS4hTQxikC6hpp4xpdXsG93owGHR4s7IfrMlLwPzf0h0ohRYpGiVfuGQKrwwOT3QWNUg4JAIkEWHJHTkQFBIUyDVdMyUPNZfHu8fFBVa3F4MY9PVDYD4lqOdweH3afrGPT/Udn6fHSlDz0TmrMOkuOU2BCbirkMglG9ddj8bbDAePkFlc0GnSYe1MW/nO8BkWdtBYWcfXRbAH01Vdf4auvvkJNTQ18vsAb1/r168NuP3bsWJSXlwcsmzlzJrKzs7FgwYIQ8QMANpsNcnngkJn1/H5+M6zf78fBgweRm5sbdkwEQbQuwW4pt9fPxgCldRPPDFLKpFhfegof/M8o0fWaa12oqmtA8ZbykGyzFybnon+PBACNrrK5G8t4G90CQH56Cu7/67cBLrKW9H0Tc+MJ1SBiiFcKW5CKsvQhrUlsLm9jP7RC4JaBPeH3A7cO6oWx2T2gUcrYXmZAaJ0n4Eq9oXqbC3KpFFIpUG12YNXUfFywOOBwe1F+xoxlk3Io/ofoEDRLAC1duhTPPfccCgoKkJqaColEEn6jIBITE5GTkxOwLD4+Hjqdjl1eXFyMs2fP4t133wUATJw4EQ899BD+9Kc/sS6wxx57DMOGDUPv3r3ZsY0YMQJZWVmwWCxYuXIlDh48iDfffLM5h0oQRIzhuqWOVZsDYoDEgm3Lzphgc3mhUkhjXmn7gsURIn6ARvfV01vK8co9Q9Czqd6OkNuJm/EUy0BfITdeuAa+yRqFaK2q21fuDNiGrxgi0CQC7wpfwJMbD/X50Qvo1z0eKRolElSNTWrtLg9uyemFFA25v4iOQbME0J///Gds2LABv/71r2M9ngCqq6tRVVXFvp4xYwYuX76M1atX44knnkBycjLGjBmDFStWsOuYTCb89re/xfnz55GUlIT8/Hzs2LEDw4YNa9WxEgQRPfFKORsDVFZlEuyGzrhSRmfp0U2jxAoR11BzREd9g0u051h9gws9tWrRHnBAoBuvtQN9xYpxrmhKb3e4vVg4YSB8fj9sTi/brsTh8eGvvymAye6GWiHDgap6KGQS3uDy0so6PLOlnA0uj2RcRoM+pt8PQbQGEr+Q70gEnU6HvXv3on///q0xpnbHYrEgKSkJZrMZWm0L7NcEQYjy4/nL+OXrO9jXjDuMqQOU3k2Dz46cx/rSU2yLB6btBlP/JhaVtveeuoR7/rJH8P1/PDwS1/ZMQLXZgVvf2Cm43voZ1+NAVT3y05KhjVOgW7yy1dtA8J0Hm8uLP/Ckx780JQ8enx/FHxwKqDFkNOiwZMIg3PnWLkHX2leP38C6AiMZj9nugkYlh0wigUxKvcKItiGa+btZFqAHH3wQ77//PhYtWtSsARIEQQBAg8sT8Do43fxfj4zELwf0iLjFQ3PRqsVvhYlqOeZuLMPgtGSMye6Ogb2TAoo1Hqiqx9FzZugTlChr6hDPwO0A3xoEnwezzRUifgDgu9P1qLe7sfyTo7wlBJZ9cjSgD1kwkVi0xOoSkfghOhrNEkAOhwNr1qzBl19+iby8PCgUgT7dV199NSaDIwji6kaoZQWDy+NDVg9tq0+eKfFKFBp0vD26Cg06mO1u7KyoxbFqC957cASe+/hISJzMc5Ny8Mf/+yFEXHx3uh7bf7yIgowUWJ2ekGrNsUYoTmlWYSYuibj6mGrcQsSrxKcLsVYnVP2Z6Ig0SwAdOnQIQ4YMAQAcPnw44L3mBEQTBNE10ScoBQOajQYddp+sQ09O9lFr0VOrxguTc0O6tBcadHh+ci6m/Gk3AOC+Yel47uMjvHEySz48gsFpyQAusMu5gcXFH1zJem1Nq5BQnFJ+WnLYPmRCGA06KGXiLTHEAsSp+jPREWmWAPrPf/4T63EQBNEFSdIo8ewdg7B422HBwOdx2T3aZCzpuni8cs8Q1De4YHF4oFXLkRKvRI3FjlqrCwDCFmEMtqAIVa1uTauIkFWNaeEhRmpSXEhRQ+a7MNtdAOIFtw0XIE7Vn4mORosLIf7888+QSCS45pprYjEegiC6GBIgoAWFSi5lm4raXN6wdX3EmoFGS0+tOqRAodVxJU5JqFgjE7zdPVGFt6YNZeOCCtJTBAVTtFYRoeM021ww2dxocHnQ4GrM9Prjr/Kw5MMjAQHNzHkVKjVQlKXH50fPY3B6MmYY+4Z8Fx/NLRQdHyO8ggPZmXOhDdPagyDammYJIJ/Ph+XLl+OVV16B1WoF0FjX54knnsAzzzwDKU9jPYIgCD508UocOmPiFQrh6vqIBd3Gyr3ErbfDZ0ERqp9TZNDjziHXQKOUCWZWme1unKixhhVvfMd584AeWDxhIM6aHVj1dUWAqCnK0uPtGddj5oZ97GfXXHbih2oLb18vxtW37KMj+OJYTcjnR1JfSZ+gxM0DeuDeYem8tYTuK6D2QkTHollp8MXFxVi3bh2WLl0Ko9EIv9+PXbt24dlnn8VDDz2E559/vjXG2mZQGjxBtC3nTHbBujGpAkJGqCozs20s3UvM+PLSkkMats4dY+Bt4go0iqDB6cJus/cfHI77//ptwLiDxRv3OLnWFX2CCkfPmfFxebWgRWf5pBzU21xsenyDy4sl2w7j2lQta6FJjlMgQ6fBNSmaZn0PXH6ut6F4c2CKPXc/FAhNtDbRzN/NEkC9e/fGn//8Z9xxxx0By7dt24bZs2fj7Nmz0e6yQ0ECiCDanmjr+pyosWLsq9sF34+0bk0046trcMEP4NmmpqcAsG56AR545zvB7d57cDimcUQOQ6FBhyE8LrJgocAcJ9fSVFZlwt8fHoEai1P0s/nOQbjz3Nz6SudMdvxU2xAg6CIZD0HEklavA3Tp0iVkZ2eHLM/OzsalS5eas0uCILo40db1aeugW+74Xr57MM6bG/tbhUMmlYTE3RRl6TF9VF/M21gWsn5wbBBznNyA6rljDKixOMM2kOU7B+HOc3PqKzEp8FOHpUc9HoJoL5olgAYPHozVq1dj5cqVActXr16NvLy8mAyMIIiOSSyDjlsyhjiFeLdzJni6NcZrdXggkQArv65g+38JYXN52SDvRLUCunglvH4/7nxTuOoyVygwwcXcDLT8tGQACJvV1dzGsNHCpMCL1RFqy/EQRCQ0SwC99NJLGD9+PL788kuMHDkSEokEu3fvxpkzZ/Dpp5/GeowEQXQQ2iLoONIxDE5LFsxoYoJ2YzFePgFldbohkUiwq7IO+ekpok1cD3AqQzMuoBM1VtFu7lyhwARhc609To8PR6st6J2kDnsO2gLGSiWWZdaW4yGISGhWutYNN9yAH3/8EZMnT4bJZMKlS5dw11134ciRI3j77bdjPUaCIDoA4Sr9mm2uNh3D+tJTmGnMhNGgC1iH2wy0peM9Z7Jj7sYyjH11Oya/tRtjX9mORzeWISlOyQoYoXEUNdXPWV96ih0XIwAYUcNHsFBgmp4mc9LIVXIp1peeQmpSHObeZAj97FZoPHrB4sAP1RbsPXUJP5y34ILFwb7HWKkEzwU1QiU6IM0Kghbi+++/x9ChQ+H1hveLd2QoCJogQmnroONIxhBcc6afPh6pSY2Vo1s63nBZZgtuzcb4VaW841DJpeieqMJ9a76BzeXlzaSKNuPqgsWB3//ze+ysqGUzz8qqTHj4hn646drGYpE2lxdquQw9tKqYWuSq6hpQvKU8JHX+hcm5SNfFw2xz4dGNZdgRlKnm9PiQFKeATCpBRjdNRJlkBNESWj0ImiCIrkdHqPQbPIbg5qlbZ48KCR4WItx4w7V2WDpJxlZNDh4Hk4L+/oPDBTOpeifHYdXU/Igzrnpq1VgxJQ9PbT6E9aWnsHJqPoBTeO2LCrz2RQWAK6IkluLngsURIn6AxvYfT28pxyv3DEFPrRovNo1tR0Utey64Fb0LMlIoDZ7oUJAAIggiIsI1Lm2LANdoxhBuXW2cQjRAOqyAsrtRclcuireUBwiloiw9Xmqy4mSItI4Aos+44oqmBqcbz9+ZA4fHB4vdA22cHFq1Ag63F2VNlZdjEfBdL9JAtbSyDvUNLvTUqtE7OQ7LJuWg8qKVt6I39QMjOhokgAiCiAhuReRg2irANZoxiK1784AeUMqkIS4uboB0OAEVr1KgTzcNVkdhxYkFQqKpNQLUzTYXvH5/QHuP9aWnAgK4LQ43K7jMNpdoXSJKgyc6ElEJoLvuukv0fZPJ1JKxEATRgUnSKLH8zhzejunL78xpkyd7JiBYKHaGOwaxdZ+9YxCe+qBcMEB61dT8iMVWc+rmxJpwAerNcT3xCSqjQYeVU/NZqw4AxCllmLhqF4DGytZiUBo80ZGISgAlJSWFff83v/lNiwZEEETHxGxz4bmPj2JIegpmNjUu1Shk8Pr9OGuyo9bqRJJG2ep1gaKJnRFaN1x8T63Vhf49EiIWW+1NJMcTzXiFBBXjCptVmInVX1ei0KBDjcXJvr/v9CWUTM5BD606oBHq+tJTKMhIoTR4okMRlQCiFHeC6LrUWl348lgNvmxqlsltzcCNEWmLukDRWF341j1Z2yC6DeOqiTZQub2IdYC6mKDaVVmHWcZMFBp0WDxxEO5f+w2Axush95okvF16KqAXmNGgw/oZ16NvN02HO29E14bathMEERHBkyy3NQOXtqwL1FyiCaZO0ijRv0cChqSnoH+PhA45icc6QD2coNKqFZhpzMT9a79BrbXxe55VmIl1QeIHaBRMb/2nEnFK8crdBNHWkAAiCCIigifZ/LRkwewgxu3SUYmmEGFnINbHE05QqZVSPPDOdwHfcWe+HoiuCQkggiAiIniSbU4jzo4CEyAdLBo6YnxPJMT6eMQElbEp7ie42rPT44NGKcPcMQasm16At6YNxfoZ12PuGAM0SlmHvh6IrgmlwRMEERHBWVUtbcTZmk1VI9l3Z4nviZRYHo9YBt3smwyY+/4BvNjUboSx+mgUV2LCuEUhmcwxbRxlgBEdi5i2wrhaoFYYBCEMIy58fj+e+/ioYKsIsdTr1mpSmqRR4my9DafrbDDZ3WwW0vFqC5ZOymmzhq1XC8w55goqm8uLP2w+hP2n69mWFwDQv3s8ntl6mNcNVmjQsRWjCaI1iWb+JgHEAwkggoiMaPtZAeF7bEVSs4ZPQN08oAcWThiIZ4LqFDHtGP6+twov3z24WRaR1rRWdQSiPb7TtQ0hFZ8LMlIw4+19gtu0Ra84gqBeYARBtAm9k+Pw4pQ8mO1uWOxuJMUpoI1TiDa9bGnNGqEaNdemavE0T88q5rXRoIPF7ka12QGLo7F1RIpGGdYqcc5kx+Jth5GdqkV+WjKqzQ7UaBRI76bBNSka0W07A82xxiVrFHhn908BwvetaUNFP4digIiOBgkggiCahdnmwnmLAz/X2yGRSNiCd0PTk9ku4Xy0pGaN2eZCtdmBqcPSMdOYGdCaIT8tOSD2hMvx85ex/M4cPCXS0Vzo857/5CjmjsnCK//3Q2DDU4Mey+4chDilHFaHh7WeJKjkaHB6YLZ3fGtRcytI88UItTQmjCDaGhJABEFEzTmTHQv+dQg7K/nbJHC7hAfT3Jo14VoziGWlrZiSh4U88SmllXVY+tERPDcpB5cdnhDLUF2DCw/f0B8rPvshZNv9VfWoNjvw1n8qA2rfFBp0mNHUAd3m8rZJYcjm0hJrXHDQdYqm/XvFEUQ0kAAiCCIqWKtBpXibBKZLeDDNaaoaSWsGMQtED62KNzhXo5Rh6vAMPLX5UIiIeWFyLrw+Py47PLzbzirMxOr/VPKKKj+unIeW9ONqbVpaQTq4ynZnaR1CEAAJIIIgoiSSNgkAYHF4eNeJpqFpNJ95/IIFm347HPEqOawOLxLVclywOLBg8yE0OLy82wpVsy6trMPTW8qxcMJAmO38IkDM5barsg4Lbs3G0PQU1k0XbT+utoBrjdMoZWxWF9PHKyXK8V5tpQWIqxsSQARBCMKXHRTOasC4ouJVMpy4aI2qBg8AnKixhmQjhftMt9ePXw5MxZIPD4fE+Gz67Uh4vPzuMTERU1pZh8t2j6BlKVwhyJ/r7Zj93gHWTdfg7HhBwIw17rvT9Xjz/qGoNtsD3t99sg5yqQTXdNPggsWB+gZXiJuQ7xqhbC+iM0ACiCAIXoSyg54ZP0B0O5VcCqNBh38fPo/VX1cKxsAEu0/EspHCxQ1l6jUh4gdoFDHPfngYK6bkodCgC0iPB8KLGI1Khu0VF2E06EL2HS7ol3mf2e6FO3NF128PGGvcrspaqBVSfFJeHXCcRoMO/fTx8Pr9KOYJIH9+ci5e+uwHfFJ+nl3ekWOeCIILtcIgiE6G2ebCiRoryqrqceKitVWajoplBx2oMom2SbhgcWCmMRPrS0+x24RrjhouGylBLRftdaWUSQX7UJVW1sHq8uCFybkoDGrfkBSmOrFGIcPxagtmGjNRFLTtBYsDRQbh81B2xsS+3lVZB5eAFaq96Z0ch+v7pvDGM+2qrMPqryuwq7KWV1w+s6Ucdw3tE7C8MzTDJQigAwmgkpISSCQSPPbYY6Lrvffeexg8eDA0Gg1SU1Mxc+ZM1NUF/jA3b96MgQMHQqVSYeDAgdiyZUsrjpwg2o5zJjvmbizD2Fe3Y/JbuzH2le14dGMZzpns4TeOArGYm2UfH8Vzk3JCBElRlh5LJg7CRauTzYBiCNcMM1w2UoPTI9rrqq5BfLI12zxI18XjlXuG4LP5RfjHwyPx2fwi9NVpQkQRQ6FBh3iVHEsn5eDve6twXd9u+OhRI9ZNL8C66QW4aHViZmHfkO2ZwouMAGRocPLHRMWa5ghkm8srKCB3Vtahh0CtpNLKOvTQqkKWU/NTojPQIVxg+/btw5o1a5CXlye6XmlpKX7zm9/gtddew8SJE3H27Fk88sgjePDBB1mRs2fPHtx7771YtmwZJk+ejC1btuCee+5BaWkphg8f3haHQxCtQnNrtjQHsZgbm8sLi90VEsPj9ftx55u7AoQPF7GMonAxPha7G/26JwjGDcWrxG9lWnXj+z216pDMtBcm5+LpoOrRTBYYM/G/fPdg9nP76ePh9PiQFKdAskaBV+4ZAqvDg3qbC2a7G2VnTCECEGibOjjNbTHSIPCdMYi5Cq0CAeZU+JDo6LS7ALJarZg2bRrWrl2L5cuXi677zTffoG/fvpg3bx4AIDMzEw8//DBeeukldp3XX38dN998M4qLiwEAxcXF2L59O15//XVs3LiRd79OpxNOp5N9bbFYWnpYBBFzWlpBORrCxdzIZVLUNrigT7gS8HqixioofgBxARDu8+JV8oDg6Ex9PJI0SnbCnzGqL2+MD9BYsFAqgWBANmMZYgN81XKkxAdWiA6OVwqmp7ZRoD66sazd6uC0RCAnh3EFisU7JahlvMup8CHR0Wl3F9icOXMwfvx4jBs3Luy6o0aNws8//4xPP/0Ufr8fFy5cwL/+9S+MHz+eXWfPnj345S9/GbDdLbfcgt27dwvut6SkBElJSexfWlpa8w+IIFqJltZsiQYmO4gPo0GHTw+fD3G/iW3DFQAXLA78UG3B3lOX8MN5Cy5YHKLbFjVlKQW7/X6+ZMPibYexs6IWCzYfwqIJg0LcUYUGHaYb++LOMO7Cnlo1slO1GJbZDdmp2mY17WQCioXcdK2dCh6JQGYIdpPFq+QoEjn/NRYH73uFBh1qLE5olDLMHWPAuukFeGvaUGx8aDgS1O3+fE0QorTrFbpp0yYcOHAA+/YJN9DjMmrUKLz33nu499574XA44PF4cMcdd2DVqlXsOufPn0fPnj0DtuvZsyfOnz8fvDuW4uJiPP744+xri8VCIojocDS3gnJzEKrVw8S3zNtYBiDUuhCuvk9VXQNvNtELk3Px0pQ8/IFn29k3GTBrQ+A9YkdFLYq3lGNwWjK+PFaDWqsL0/76DVZMycOC27Jhd/ng8fqw+2RdgDuqtYsStmcdnHACua7BBVy0Qi2XYsmHR/DlsRr2vU0PDccDhY31m7giqsigw7yxWeiRoAqxsDFZYG98+SNWTs3H27tOBZQUoGwwoqPTbgLozJkzmD9/Pj7//HOo1ZE9bR09ehTz5s3D4sWLccstt6C6uhpPPvkkHnnkEaxbt45dTyKRBGzn9/tDlnFRqVRQqUID+QiiI9GcCsotgTuZi8W3cN1vYgLggsURIn6AK0UHX7lnSMi2cqkEt63cyeta21lRixmj+rKva60uPPDOdwCAddML2P8HE2t3YTDh3GWtRUKYOKjLDjfu+csetlXH7hN17Hm9ZHPj9//8Hr8d3Q/zx2bB4/NDo2x0bf3neA0K0lLwh1uzsUDSGPOjVcuRrFHA6fHhkRv6Y9nHR0O+145cAZsggHYUQPv370dNTQ2uu+46dpnX68WOHTuwevVqOJ1OyGSBvuWSkhIYjUY8+eSTAIC8vDzEx8ejqKgIy5cvR2pqKnr16hVi7ampqQmxChFEZ6M5FZRj8ZlJGiXKquoFBQUQ6H4TEgD1DS7RVPX6BheyU7UB25ZV1YvGFTk9Pt4KxnKp8ANP8HivFpQyKW+9IiAwLT+4VQfQGONjc3nx+pcVeP3LipDtP3rUiHv/8g1sLi9uHtADiyYMxFMflGNnRS3WTS8IaCPCpbXFJkG0hHYTQGPHjkV5eXnAspkzZyI7OxsLFiwIET8AYLPZIJcHDplZz+/3AwBGjhyJL774Ar/73e/YdT7//HOMGjUq1odAEG1Oe7lYYuF+E2qNIfZ+uM/tplHwul/ee3A4rzBi2lJcjQG6JrsLM5vakAQXM+S6LZn3mZYlAFB2xiQonoqy9NCqFXj/weFIVCuQoJbj9//8nnWVhSsmeTWKTeLqoN0EUGJiInJycgKWxcfHQ6fTscuLi4tx9uxZvPvuuwCAiRMn4qGHHsKf/vQn1gX22GOPYdiwYejduzcAYP78+Rg9ejRWrFiBSZMmYdu2bfjyyy9RWlratgdIEK1Ee7hYuC0TgkUFE8TM1xKBO05tmKBYvvfDuf36dNNg8bbDyE9PwSxjJjw+P3onxUEq9WPrbCOWfXwkQBgZDTqsn3H9VdmZPEGlwNS132JWYSZmGTOhUcphc3kE0/K5wmV96SmsnJoPqUQSkkK/YkoeUpPjkKGLB9CY7cddJ1xF7KtRbBJXBx06TL+6uhpVVVXs6xkzZuDy5ctYvXo1nnjiCSQnJ2PMmDFYsWIFu86oUaOwadMmLFy4EIsWLUL//v3x97//nWoAEUQLSNIosWJKHk5fsmHV1xUBomLcgB4w9tfhma2HRevPpMQrBVPVCw06pMSHipJwbj+724v7h2fg7V2n2En8xc+OIT89BWVV9axFg2sNanB6UHPZye7/akGfoERBRgr73YjFQQGBwsXm8uLve6vw8t2D0eD0wOvzw+vzw+bywOb2ssUUa62ukKKTYtajtkj/J4jmIvEzviOCxWKxICkpCWazGVqttr2HQxAdArPNhbnvl2FnZaA1Zu4YQ4DY4DI6Sx8QBFtV1xBSdLDIoMfyyTnw+vww2/mtR4x1Kdjt9+OFy1j60RHsqqwLGAd38tcoZaybjDtGrkALZ73qLJwz2VmxKPa9FBp0GJJ+RSxxLT18xRSLsvSY05SNt2pqfoCwEjq/RVl6PHvHIEgA6Drp+SQ6H9HM3ySAeCABRBChnKixYuyr20OWh7M0fPX4DWyxRLPNhUsNLlhdHjQ4vVDIJFDKZHjty+P4+oeL7DaRplD/UG3BrW/sDBnHW9OGYvZ7BwCEF2gld+WyAb3Rfn5HhBFzDU43tHFKLN52OMR69sLkXLi8PljsgYLSbHNh7sYy3npCRoMO+ekpABByPhkLm7G/DgqZFC5PYwmC9aWnYHN5O/X5JDoX0czfHdoFRhBEx0Gozky4IFiz3Y2TF63wA3h22+GAjKEigw6zbzLgm5OXAraJNIXa5vKwk2/3RBXemjYUaoUM+gQlNEoZbC4v8tOSA1x2wZ9zus7WJu1F2orgGLFogubFiikygdOPbizDyqn57DKg0YV26IwJU69PCxGTQOc+n8TVCwkggiAiQigji4klEcq6cnm8+KCshtcKs7OyDj4EpmQzRJJCnaxR8maBFRl0rEUonEAz2fmF3dWSwh1N0Hy4YopOjw82lxfzNpaxwdaJagV08UroE5Rt2q6FIFoKCSCCICJCKCOr7IwJY7O7Y2pTMDJXiBQadBg3oCeGcuJNgtlVWYeHCvth7hhDiHhqcIpPyPEqeUjsCYAmK5MEC8cPCJulJPZ+V0vhDld2gDlXNpeX/T65Ls6TtQ2i23e180l0bNq9FxhBEJ0DoV5Xx6steHbiIF4hUlpZhz/+3w+ihQk1ShlSk9VsscXZ7x3ArA37UFZVD22cuLXA6vAIFlfcWVmLX/RMRGqSWrTPFVMgkI+ulsIdrgdc8LkKzvJqy3YtBNFSSAARBBExTCHGrx6/AVtnj8JXj9+Al+8eDJfXLyhEdlXWISVeeOKbVZiJ5TytFHZV1mHxtsNsCjYf4Vw2AHBNchxWCDQpLZmci+PVFt7tumIKt5DILcrS49ExWVhfeopdxleBPNKGuATRESAXGEEQUcEXUxLO9eHx+gVrxYzqpxMNUhaLGwlncUhpGmuSRjgYeOmkHDg9bddepKMjVG0cAD6aWygaTN0e7VoIormQACIIosWEEyJmm5u3TUORQQdFmBgdsbiRaBrECgUDt2cH946K0LmK5JzQ+SQ6CySACIJoMWJCpNCgw3dNPbhmFWZizo0GqBRSJKjkOHC6HpfD9AgTixuJlcWhvTq4X63Q+SQ6AySACIJoMYwQCa4gbDTo8PTtAyCXSTAuu0eAO+WJf3yP+4an47zZLtqIUy6TwGwTdoORxYEgiOZAlaB5oErQBNE8zDYXai47Yba7oVHKEK+UI1mjCBEjx89bcMvrO6FRyvDwDf1w26BULPv4SECRRG4X84KMFKokTBBEWKgVRgshAUQQrYfZ5sJ3p+tD+knNKszEmGu7Qy6TQiIB6hvc8Pr82N/kPivISKFKwgRBiEKtMAiC6LDUWkPT2m0uL9aXnsIdeb2x9OMjAe4wo0GHlVPzMW9jGVUSJggiZpAAIgiiTbE43Cg7YwqJ+5lVmIllQeIHuJI1Nqswk80Iu1q6txME0X6QACIIok3RqhVYX3oqpKGmWNNSphGnNk6BcyZ7SLA1dRsnCCJaqBI0QRBtij5BiYKMFMzbWIb89BSsm16At6YNRbwy/PNYvEoeIn6AK93GxapGEwRBcCELEEEQbQq3dg/X4vP+g8NFt+uTEgerw0PdxgmCiAkkgAiCaHP4avckqOWiVZ17adXUbZwgiJhBLjCCINqFJI0S/XskYEh6Cvr3SEBPrZq3ESe3qjN1GycIIlaQBYggiA5DuKrO0fT+IgiCEIMsQARBdCiCLUPcmB4mfkjMSkQQBBEJZAEiCKJTQb2/CIKIBSSACILodFC3cYIgWgq5wAiCIAiC6HKQACIIgiAIostBAoggCIIgiC4HCSCCIAiCILocJIAIgiAIguhykAAiCIIgCKLLQQKIIAiCIIguBwkggiAIgiC6HCSACIIgCILocnQYAVRSUgKJRILHHntMcJ0ZM2ZAIpGE/A0aNIhdZ8OGDbzrOByONjgKgiAIgiA6Ax2iFca+ffuwZs0a5OXlia73xhtv4MUXX2RfezweDB48GHfffXfAelqtFsePHw9YplarYzdggiAIgiA6Ne1uAbJarZg2bRrWrl2LlJQU0XWTkpLQq1cv9u+7775DfX09Zs6cGbCeRCIJWK9Xr16teQgEQRAEQXQy2l0AzZkzB+PHj8e4ceOi3nbdunUYN24cMjIyApZbrVZkZGSgT58+mDBhAsrKykT343Q6YbFYAv4IgiAIgrh6aVcBtGnTJhw4cAAlJSVRb1tdXY1///vfePDBBwOWZ2dnY8OGDfjwww+xceNGqNVqGI1GVFRUCO6rpKQESUlJ7F9aWlrU4yEIgiAIovMg8fv9/vb44DNnzqCgoACff/45Bg8eDAC48cYbMWTIELz++uthty8pKcErr7yCc+fOQalUCq7n8/kwdOhQjB49GitXruRdx+l0wul0sq8tFgvS0tJgNpuh1WqjOzCCIAiCINoFi8WCpKSkiObvdguC3r9/P2pqanDdddexy7xeL3bs2IHVq1fD6XRCJpPxbuv3+7F+/Xr8+te/FhU/ACCVSnH99deLWoBUKhVUKlXzDoQgCIIgiE5HuwmgsWPHory8PGDZzJkzkZ2djQULFgiKHwDYvn07Kisr8cADD4T9HL/fj4MHDyI3N7fFYyYIgiAI4uqg3QRQYmIicnJyApbFx8dDp9Oxy4uLi3H27Fm8++67AeutW7cOw4cPD9keAJYuXYoRI0YgKysLFosFK1euxMGDB/Hmm2+23sEQBEEQBNGp6BB1gISorq5GVVVVwDKz2YzNmzfjjTfe4N3GZDLht7/9Lc6fP4+kpCTk5+djx44dGDZsWFsMmSAIgiCITkC7BUF3ZKIJoiIIgiAIomMQzfzd7nWACIIgCIIg2hoSQARBEARBdDlIABEEQRAE0eUgAUQQBEEQRJeDBBBBEARBEF0OEkAEQRAEQXQ5SAARBEEQBNHlIAFEEARBEESXgwQQQRAEQRBdDhJABEEQBEF0OUgAEQRBEATR5SABRBAEQRBEl4MEEEEQBEEQXQ4SQARBEARBdDlIABEEQRAE0eUgAUQQBEEQRJeDBBBBEARBEF0OEkAEQRAEQXQ5SAARBEEQBNHlIAFEEARBEESXgwQQQRAEQRBdDhJABEEQBEF0OUgAEQRBEATR5SABRBAEQRBEl4MEEEEQBEEQXQ4SQARBEARBdDlIABEEQRAE0eUgAUQQBEEQRJeDBBBBEARBEF0OEkAEQRAEQXQ5SAARBEEQBNHlIAFEEARBEESXgwQQQRAEQRBdjg4jgEpKSiCRSPDYY48JrjNjxgxIJJKQv0GDBgWst3nzZgwcOBAqlQoDBw7Eli1bWnn0BEEQBEF0JjqEANq3bx/WrFmDvLw80fXeeOMNVFdXs39nzpxBt27dcPfdd7Pr7NmzB/feey9+/etf4/vvv8evf/1r3HPPPfj2229b+zAIgiAIgugktLsAslqtmDZtGtauXYuUlBTRdZOSktCrVy/277vvvkN9fT1mzpzJrvP666/j5ptvRnFxMbKzs1FcXIyxY8fi9ddfb+UjIQiCIAiis9DuAmjOnDkYP348xo0bF/W269atw7hx45CRkcEu27NnD375y18GrHfLLbdg9+7dgvtxOp2wWCwBfwRBEARBXL3I2/PDN23ahAMHDmDfvn1Rb1tdXY1///vfeP/99wOWnz9/Hj179gxY1rNnT5w/f15wXyUlJVi6dGnUYyAIgiAIonPSbhagM2fOYP78+fjb3/4GtVod9fYbNmxAcnIy7rzzzpD3JBJJwGu/3x+yjEtxcTHMZjP7d+bMmajHQxAEQRBE56HdLED79+9HTU0NrrvuOnaZ1+vFjh07sHr1ajidTshkMt5t/X4/1q9fj1//+tdQKpUB7/Xq1SvE2lNTUxNiFeKiUqmgUqlacDQEQRAEQXQm2s0CNHbsWJSXl+PgwYPsX0FBAaZNm4aDBw8Kih8A2L59OyorK/HAAw+EvDdy5Eh88cUXAcs+//xzjBo1KubHQBAEQRBE56TdLECJiYnIyckJWBYfHw+dTscuLy4uxtmzZ/Huu+8GrLdu3ToMHz48ZHsAmD9/PkaPHo0VK1Zg0qRJ2LZtG7788kuUlpa23sEQBEEQBNGpaPcsMDGqq6tRVVUVsMxsNmPz5s281h8AGDVqFDZt2oS3334beXl52LBhA/7+979j+PDhbTFkgiAIgiA6ARK/3+9v70F0NCwWC5KSkmA2m6HVatt7OARBEARBREA083eHtgARBEEQBEG0BiSACIIgCILocpAAIgiCIAiiy0ECiCAIgiCILgcJIIIgCIIguhwkgAiCIAiC6HKQACIIgiAIostBAoggCIIgiC4HCSCCIAiCILocJIAIgiAIguhykAAiCIIgCKLLQQKIIAiCIIguBwkggiAIgiC6HCSACIIgCILocpAAIgiCIAiiy0ECiCAIgiCILgcJIIIgCIIguhwkgAiCIAiC6HKQACIIgiAIostBAoggCIIgiC4HCSCCIAiCILocJIAIgiAIguhykAAiCIIgCKLLQQKIIAiCIIguBwkggiAIgiC6HCSACIIgCILocpAAIgiCIAiiy0ECiCAIgiCILgcJIIIgCIIguhwkgAiCIAiC6HKQACIIgiAIostBAoggCIIgiC4HCSCCIAiCILocHUYAlZSUQCKR4LHHHhNdz+l04plnnkFGRgZUKhX69++P9evXs+9v2LABEokk5M/hcLTyERAEQRAE0VmQt/cAAGDfvn1Ys2YN8vLywq57zz334MKFC1i3bh0MBgNqamrg8XgC1tFqtTh+/HjAMrVaHdMxEwRBEATReWl3AWS1WjFt2jSsXbsWy5cvF133s88+w/bt23Hy5El069YNANC3b9+Q9SQSCXr16hXxGJxOJ5xOJ/vaYrFEvC1BEARBEJ2PdneBzZkzB+PHj8e4cePCrvvhhx+ioKAAL730Eq655hr84he/wO9//3vY7faA9axWKzIyMtCnTx9MmDABZWVlovstKSlBUlIS+5eWltaiYyIIgiAIomPTrhagTZs24cCBA9i3b19E6588eRKlpaVQq9XYsmULamtrMXv2bFy6dImNA8rOzsaGDRuQm5sLi8WCN954A0ajEd9//z2ysrJ491tcXIzHH3+cfW2xWEgEEQTRZTDbXKi1umBxuKGNU0Afr0SSRtnewyKIVqXdBNCZM2cwf/58fP755xHH5/h8PkgkErz33ntISkoCALz66qv41a9+hTfffBNxcXEYMWIERowYwW5jNBoxdOhQrFq1CitXruTdr0qlgkqlavlBEQRBdDLOmexYsPkQdlbUsstGZ+nx4pQ89E6Oa8eREUTr0m4usP3796OmpgbXXXcd5HI55HI5tm/fjpUrV0Iul8Pr9YZsk5qaimuuuYYVPwAwYMAA+P1+/Pzzz7yfI5VKcf3116OioqLVjoUgCKIzYra5QsQPAOyoqMVTmw/BbHO108gIovVpNwE0duxYlJeX4+DBg+xfQUEBpk2bhoMHD0Imk4VsYzQace7cOVitVnbZjz/+CKlUij59+vB+jt/vx8GDB5Gamtpqx0IQBNEZqbW6QsQPw46KWtRaSQARVy/tJoASExORk5MT8BcfHw+dTof/397dx1RV/3EAfx8ucJErXBJF5RmXxRaCBFk+hGsx2iSV2bJ+YYCyNTENtkptumE+4O+3orFM5iKn9DAoi7ZqrkKETK1Y5M0MUwhQpzR74ikVuPL5/eGP+/Pw/HDvPVfP+7XdzXvP9x6/n/fu/d4P95wDUVFRAG6cm5OWlmZ7zlNPPQV/f3+sXLkStbW1OHLkCF588UWsWrUKEybc+Kr25ZdfxhdffIGGhgZYLBZkZmbCYrFg9erVmtRJROSq2q51D7m9fZjtRLcyzS+DH0pzczPOnz9vuz9x4kSUl5dj3bp1iI+Ph7+/P5YvX666fL6lpQXPPPMMfvvtN5jNZsTGxuLIkSOYM2eOFiUQEbksXy+PIbf7DLOd6FamiIhoPQlX09bWBrPZjNbWVvj6+mo9HSIih2i90oV1JSdwZIDDYAkzJ2PXv2J5NRjdUkbz+a357wEiIiJtmL098e/HopEwc7Lq8YSZk/Gfx6LZ/NBtzaUPgRERkWMF+k3Arn/F4o+OLrRf64aPlwcmT+TvAaLbHxsgIiKdM3uz4SH94SEwIiIi0h02QERERKQ7bICIiIhId9gAERERke6wASIiIiLdYQNEREREusMGiIiIiHSHDRARERHpDhsgIiIi0h02QERERKQ7/FMYAxARADf+qiwRERHdGno/t3s/x4fCBmgA7e3tAICQkBCNZ0JERESj1d7eDrPZPOQYRUbSJulMT08PLl26BB8fHyiKovV0xq2trQ0hISG4cOECfH19tZ6O5piHGvNQYx5qzKM/ZqLmSnmICNrb2xEYGAg3t6HP8uE3QANwc3NDcHCw1tOwO19fX81fnK6EeagxDzXmocY8+mMmaq6Sx3Df/PTiSdBERESkO2yAiIiISHfYAOmA0WhEbm4ujEaj1lNxCcxDjXmoMQ815tEfM1G7VfPgSdBERESkO/wGiIiIiHSHDRARERHpDhsgIiIi0h02QERERKQ7bIBc2JYtW6Aoiuo2bdo0AEB3dzc2bNiAWbNmwWQyITAwEGlpabh06dKQ+ywrK0N8fDz8/PxgMpkwe/ZsvPPOO/3GFRYWIiIiAl5eXoiLi8PXX3/tkBpHQ6s8du7cifvuuw8+Pj4ICAhASkoKzpw547A6R0rL10evnTt3QlEU5OTk2LO0MdEyj4sXL2LFihXw9/eHt7c3Zs+ejZqaGofUORpaZWK1WrF582ZERERgwoQJmDFjBrZu3Yqenh6H1ToSjsjjZqWlpVAUBSkpKf226WVNvdlgebjMmirksnJzc+Wee+6R5uZm2+3y5csiItLS0iKJiYny/vvvyy+//CLffPON3H///RIXFzfkPisrK6WsrExqa2ulvr5eCgoKxGAwyOeff24bU1paKh4eHlJUVCS1tbWSnZ0tJpNJzp0759B6h6NVHo888ojs27dPTp06JRaLRZKTkyU0NFQ6OjocWu9wtMqjV3V1tYSHh0t0dLRkZ2c7osRR0SqPv/76S8LCwiQjI0O+++47aWxslEOHDkl9fb1D6x0JrTLZvn27+Pv7y2effSaNjY1y4MABmThxohQUFDi03uE4Io9eTU1NEhQUJA8++KAsXbpUtU1Pa2qvofJwlTWVDZALy83NlZiYmBGPr66uFgCjflPFxsbK5s2bbffnzJkjq1evVo2JjIyUjRs3jmq/9qZVHn1dvnxZAMhXX301qv3am5Z5tLe3y8yZM6W8vFwWLlzoMg2QFnls2LBBFixYMKp9OItWmSQnJ8uqVatUY5YtWyYrVqwY1X7tzVF5WK1WmT9/vrz11luSnp7e7wNfb2vqcHn0pdWaykNgLq6urg6BgYGIiIjAk08+iYaGhkHHtra2QlEU+Pn5jWjfIoKKigqcOXMGCQkJAICuri7U1NQgKSlJNTYpKQnHjx8fcx324uw8BtsvAEyaNGlUc3cErfJ49tlnkZycjMTExPFM3+60yOOTTz5BfHw8Hn/8cQQEBCA2NhZFRUXjLcVutMhkwYIFqKiowNmzZwEAP/74I44ePYpFixaNqxZ7cEQeW7duxZQpU5CZmdlvmx7X1KHyGGy/gAZrqlPbLRqVgwcPyocffignT560/aQ9depU+eOPP/qNvXr1qsTFxUlqauqw+21paRGTySTu7u5iNBpl7969tm0XL14UAHLs2DHVc3bs2CF33XXX+IsaBy3y6Kunp0cWL17sEj/xa5VHSUmJREVFydWrV0VEXOYbIK3yMBqNYjQa5aWXXpIffvhB9uzZI15eXlJcXGy32sZKq0x6enpk48aNoiiKuLu7i6IokpeXZ7e6xsoReRw9elSCgoLk999/FxHp942H3tbU4fLoS8s1lQ3QLaSjo0OmTp0q+fn5qse7urpk6dKlEhsbK62trcPu5/r161JXVycnTpyQV199Vcxms1RWVorI/9+sx48fVz1n+/btcvfdd9utFntwRh59rVmzRsLCwuTChQv2KMGunJHH+fPnJSAgQCwWi228qzRAfTnr9eHh4SFz585VPWfdunXywAMP2KUOe3JWJiUlJRIcHCwlJSVy8uRJefvtt2XSpEmyf/9+e5c0LuPNo62tTcLDw+XgwYO2xwZrgPSwpo4kj760XFPZAN1iEhMTVceSu7q6JCUlRaKjowfs2kciMzNTkpKSRESks7NTDAaDlJWVqcY899xzkpCQMPaJO4ij87jZ2rVrJTg4WBoaGsY8X0dzdB4ff/yxABCDwWC7ARBFUcRgMIjVarVLHfbijNdHaGioZGZmqsYUFhZKYGDg2CbtYM7IJDg4WN544w3VmG3btrncB77I+PI4ceJEv/eDoii290N9fb2u1tSR5HEzrddUngN0C+ns7MTp06cxffp0ADcuU1y+fDnq6upw6NAh+Pv7j2m/IoLOzk4AgKenJ+Li4lBeXq4aU15ejnnz5o2vADtzRh6999euXYuysjIcPnwYERERdpm/vTkjj4cffhg//fQTLBaL7RYfH4/U1FRYLBYYDAa71TNeznp9zJ8/v98lvGfPnkVYWNjYJ+8gzsrkypUrcHNTf7wYDAbNL4Pva7x5REZG9ns/LFmyBA899BAsFgtCQkJ0taaOJA/AhdZUTdouGpHnn39eqqqqpKGhQb799lt59NFHxcfHR5qamqS7u1uWLFkiwcHBYrFYVJcxdnZ22vbx9NNPq640yMvLky+//FJ+/fVXOX36tOTn54u7u7sUFRXZxvResrl3716pra2VnJwcMZlM0tTU5NT6+9Iqj6ysLDGbzVJVVaXa75UrV5xaf19a5dGXqxwC0yqP6upqcXd3lx07dkhdXZ2899574u3tLe+++65T6x+IVpmkp6dLUFCQ7TL4srIymTx5sqxfv96p9ffliDz6GuiQj57W1L4GysNV1lQ2QC7siSeekOnTp4uHh4cEBgbKsmXL5OeffxYRkcbGRgEw4O3mY/ELFy6U9PR02/1NmzbJnXfeKV5eXnLHHXfI3LlzpbS0tN//vXv3bgkLCxNPT0+59957Nb/kW0S7PAbb7759+5xQ9eC0fH3czFUaIC3z+PTTTyUqKkqMRqNERkbKm2++6ehyR0SrTNra2iQ7O1tCQ0PFy8tLZsyYIZs2bVJ9cGrBEXn0Ndg5L3pZU/saKA9XWVOV/02GiIiISDd4DhARERHpDhsgIiIi0h02QERERKQ7bICIiIhId9gAERERke6wASIiIiLdYQNEREREusMGiIiIiHSHDRARERHpDhsgIrqtZGRkICUlZUzP3b9/P/z8/Ow6HyJyTWyAiIiISHfYABGRbrz22muYNWsWTCYTQkJCsGbNGnR0dAAAqqqqsHLlSrS2tkJRFCiKgi1btmg7YSJyGDZARKQbbm5ueP3113Hq1CkUFxfj8OHDWL9+PQBg3rx5KCgogK+vL5qbm9Hc3IwXXnhB4xkTkaO4az0BIiJnycnJsf07IiIC27ZtQ1ZWFgoLC+Hp6Qmz2QxFUTBt2jTtJklETsEGiIh0o7KyEnl5eaitrUVbWxusViuuXbuGf/75ByaTSevpEZET8RAYEenCuXPnsGjRIkRFReGjjz5CTU0Ndu/eDQDo7u7WeHZE5Gz8BoiIdOH777+H1WpFfn4+3Nxu/Oz3wQcfqMZ4enri+vXrWkyPiJyMDRAR3XZaW1thsVhUj02ZMgVWqxW7du3C4sWLcezYMezZs0c1Jjw8HB0dHaioqEBMTAy8vb3h7e3txJkTkbMoIiJaT4KIyF4yMjJQXFzc7/H09HTExMTglVdeQUtLCxISEpCamoq0tDT8/ffftl+AmJWVhQMHDuDPP/9Ebm4uL4Unuk2xASIiIiLd4UnQREREpDtsgIiIiEh32AARERGR7rABIiIiIt1hA0RERES6wwaIiIiIdIcNEBEREekOGyAiIiLSHTZAREREpDtsgIiIiEh32AARERGR7vwXNX6IQ1FRm/IAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(x=data['Lat'], y=data['Lon'])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The point at the bottom of the plot is a longitude outlier. \n",
    "\n",
    "There is one more point, that was not included nither to longitude nor to latitude outliers, looks like another outlier though: the one in the top right corner. Longitude might be close to upper 3std border, latitude fits inside 3std. Lets check this point:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>Lat</th>\n",
       "      <th>Lon</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>52.351935</td>\n",
       "      <td>5.009082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>52.305075</td>\n",
       "      <td>5.016522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>240</th>\n",
       "      <td>52.350513</td>\n",
       "      <td>5.016555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>270</th>\n",
       "      <td>52.350513</td>\n",
       "      <td>5.016555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>280</th>\n",
       "      <td>52.349139</td>\n",
       "      <td>5.012134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>378</th>\n",
       "      <td>52.305296</td>\n",
       "      <td>5.015376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>389</th>\n",
       "      <td>52.354845</td>\n",
       "      <td>5.001384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>423</th>\n",
       "      <td>52.413550</td>\n",
       "      <td>5.029122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463</th>\n",
       "      <td>52.353136</td>\n",
       "      <td>5.000973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>523</th>\n",
       "      <td>52.352324</td>\n",
       "      <td>5.000011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>662</th>\n",
       "      <td>52.352820</td>\n",
       "      <td>5.000330</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Lat       Lon\n",
       "38   52.351935  5.009082\n",
       "108  52.305075  5.016522\n",
       "240  52.350513  5.016555\n",
       "270  52.350513  5.016555\n",
       "280  52.349139  5.012134\n",
       "378  52.305296  5.015376\n",
       "389  52.354845  5.001384\n",
       "423  52.413550  5.029122\n",
       "463  52.353136  5.000973\n",
       "523  52.352324  5.000011\n",
       "662  52.352820  5.000330"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[lambda x: x[\"Lon\"] > 5, (\"Lat\", 'Lon')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.6338021298188536"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# how far is the point (423) from the mean latitude (based on formula x = mean + n*std)\n",
    "\n",
    "n = (5.029122 - data['Lon'].mean()) / np.std(data['Lon'])\n",
    "n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The value is close to 3, which means that this point is in 5% minority of data, that lies behind ±2std. Can easily be classified as outlier counting the fact, that also latitude is the highest for this longitude sector (>5)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>Price</th>\n",
       "      <th>Address</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Area</th>\n",
       "      <th>Room</th>\n",
       "      <th>Lon</th>\n",
       "      <th>Lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>data_type</th>\n",
       "      <td>continuous</td>\n",
       "      <td>text</td>\n",
       "      <td>text</td>\n",
       "      <td>continuous</td>\n",
       "      <td>descrete</td>\n",
       "      <td>continuous</td>\n",
       "      <td>continuous</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>missing values</th>\n",
       "      <td>3</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>distribution</th>\n",
       "      <td>right_skewed</td>\n",
       "      <td>not applicable</td>\n",
       "      <td>not applicable</td>\n",
       "      <td>right_skewed</td>\n",
       "      <td>right_skewed</td>\n",
       "      <td>normal</td>\n",
       "      <td>heavy_tailed</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n outliers</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Price         Address             Zip          Area   \n",
       "data_type         continuous            text            text    continuous  \\\n",
       "missing values             3               0               0             0   \n",
       "distribution    right_skewed  not applicable  not applicable  right_skewed   \n",
       "n outliers                 0               0               0             1   \n",
       "\n",
       "                        Room         Lon           Lat  \n",
       "data_type           descrete  continuous    continuous  \n",
       "missing values             0           0             0  \n",
       "distribution    right_skewed      normal  heavy_tailed  \n",
       "n outliers                 0           2             0  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_info.loc['Lon', 'n outliers'] = 2\n",
    "\n",
    "data_info.to_parquet(\"./data/data_info.parquet\")\n",
    "\n",
    "data_info.transpose()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Check text data (Address and ZIP)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Address\n",
       "Ringdijk, Amsterdam                         2\n",
       "Quashibastraat, Amsterdam                   2\n",
       "Slatuinenweg 21, Amsterdam                  1\n",
       "Tweede Oosterparkstraat 188 2, Amsterdam    1\n",
       "Spadinalaan 53, Amsterdam                   1\n",
       "                                           ..\n",
       "Hamontstraat 59, Amsterdam                  1\n",
       "Kanaalstraat 204 H, Amsterdam               1\n",
       "Paramariboplein 67 bel, Amsterdam           1\n",
       "Achillesstraat 113 2, Amsterdam             1\n",
       "Van Gentstraat 6 1, Amsterdam               1\n",
       "Name: count, Length: 737, dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Address'].value_counts()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 737 unique addresses out of 739 datapoints. That will not give a lot of information in terms of prediction. Let's check zip codes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Zip\n",
       "1075 XR    5\n",
       "1060 SC    3\n",
       "1060 RZ    3\n",
       "1011 AW    3\n",
       "1095 MD    3\n",
       "          ..\n",
       "1058 AR    1\n",
       "1076 RA    1\n",
       "1019 AL    1\n",
       "1053 TP    1\n",
       "1055 PE    1\n",
       "Name: count, Length: 679, dtype: int64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Zip'].value_counts()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It looks slightly better: there are some houses with the same zip codes and there could be dependancy between zip and price of the house. However, if we investigate the princip of zip codes formation, it is easy to discover, that main regions of the city are marked with 4 numbers, and subregions, which usually include just a couple of houses, are marked with 2 letters. When first 2 digits are 1 and 0, it means the house is in Amsterdam. Even though the dataset is about this city, there are also zip codes, which start with 11 - that is a Zuidost area, which is a bit futher from the city. The assumption is that there might be less clusters (and thus more predictable), if use only 4 digits instead of 4 digits + 2 letters. Let's try to separate them:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['Zip num'] = np.nan\n",
    "data['Zip letters'] = np.nan\n",
    "for row in range(data.shape[0]):\n",
    "    data.loc[row, 'Zip num'] = data.loc[row, 'Zip'][:-3]\n",
    "    data.loc[row, 'Zip letters'] = data.loc[row, 'Zip'][5:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Zip num\n",
       "1017    26\n",
       "1055    24\n",
       "1053    21\n",
       "1102    21\n",
       "1058    19\n",
       "        ..\n",
       "1086     3\n",
       "1014     3\n",
       "1109     2\n",
       "1028     2\n",
       "1022     1\n",
       "Name: count, Length: 70, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['Zip num'].value_counts()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The assumption is accepted: there are 9 times less clusters when using only numbers of zip code. This hypothesis will be tested when different models will be applied."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Identify the promising transformations that may be applied."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Divide Zip to numbers and letters or make a deeper research to discover another princip for clustering Zip codes.\n",
    "2. Add price per square meter, as in the world of real estate that is more important than a price of a house.\n",
    "3. Having coordinates of latitude and longitude, it is posible to define a 'center' point as a city center, where the houses are usually more expencive, and calculate distance to every datapoint as a vector.\n",
    "4. Address data in original format is not informative and needs to be changed or removed."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Summary:**\n",
    " - Target: `Price` numerical continuous variable;\n",
    " - There are 6 features: 4 continuous, 2 text;\n",
    " - There are 3 missing values in target variable;\n",
    " - There are no missing values in features;\n",
    " - `Price`, `Area` and `Room` have right-skewed distribution and 56, 33, 62 outliers respectively. However if we look closer to these outliers, we'll see that all of them are just big (and thus expencive) houses. These values will not be considered as outliers;\n",
    " - There are 2 outiers in `Longitude`, that are much lower than the mean;\n",
    " - Outliers of `Latitude` create a cluster togeter with other points and will not be considered as outliers.\n",
    " - There is one point, which is an outlier niether for longitude, not for latitude, but when plotting coordinates, we see that it is far from the main distribution of points. It will be considered an outlier;\n",
    " - Most values of both text features (`Address` and `Zip`) are unique, thus not informative for prediction;\n",
    " - When removing letters from `Zip` code, there are 70 unique values, which represent areas of the city. The assumption is they are more generalized, than original format of zip code.\n",
    " \n",
    "\n"
   ]
  }
 ],
 "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": "3.10.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}