Segmentation_Kmeans.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Customer segmentation - raw calculated RFM variables with K-Means Clustering\n",
    "\n",
    "\n",
    "The project concerns customer segmentation by combining the RFM method and K-Means clustering. In this approach I have used raw calculated RFM variables and K-Means Clustering to determine segments of clients. The dataset includes sample sales data based on retail analytics and contains three years of sales.\n",
    "\n",
    "\n",
    "**Customer segmentation**\n",
    "\n",
    "The customer segmentation is an effective method that enables us to get better know our clients and to better correspond, their various needs. Almost every company that sells products or services stores data of shopping. This type of data can be used to execute customer segmentation thus, the results of the analysis can be translated into marketing campaigns to increase sales. One of the most widely used techniques is RFM analysis, which allows us to create personalized special offers to improve sales and decrease customer retention.\n",
    "\n",
    "**RFM** stands for Recency, Frequency, Monetary Value and it is the technique of customer segmentation based on their transaction history. The RFM analysis is based on three criterias, which measure different customer characteristics:\n",
    "-  Recency: Days since last purchase/order of the client;\n",
    "- Frequency:  Total number of purchases the customer were made;\n",
    "- Monetary Value: Total money the customer spent per order.\n",
    "\n",
    "\n",
    "**Dataset**\n",
    "\n",
    "The dataset comes from https://www.kaggle.com/kyanyoga/sample-sales-data.\n",
    "\n",
    "The whole project is divided into 6 steps:\n",
    "-Importing libraries and loading data set\n",
    "- Data cleaning and preparation\n",
    "- Data analysis and visualization\n",
    "- RFM Analysis\n",
    "- Model: raw calculated RFM variables + K-Means Clustering\n",
    "- Description of the results\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import libriaries and data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "\n",
    "import datetime as dt\n",
    "import scipy.stats\n",
    "import statsmodels.formula.api as sm\n",
    "\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from mpl_toolkits.mplot3d import Axes3D"
   ]
  },
  {
   "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>ORDERNUMBER</th>\n",
       "      <th>QUANTITYORDERED</th>\n",
       "      <th>PRICEEACH</th>\n",
       "      <th>ORDERLINENUMBER</th>\n",
       "      <th>SALES</th>\n",
       "      <th>ORDERDATE</th>\n",
       "      <th>STATUS</th>\n",
       "      <th>QTR_ID</th>\n",
       "      <th>MONTH_ID</th>\n",
       "      <th>YEAR_ID</th>\n",
       "      <th>...</th>\n",
       "      <th>ADDRESSLINE1</th>\n",
       "      <th>ADDRESSLINE2</th>\n",
       "      <th>CITY</th>\n",
       "      <th>STATE</th>\n",
       "      <th>POSTALCODE</th>\n",
       "      <th>COUNTRY</th>\n",
       "      <th>TERRITORY</th>\n",
       "      <th>CONTACTLASTNAME</th>\n",
       "      <th>CONTACTFIRSTNAME</th>\n",
       "      <th>DEALSIZE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10107</td>\n",
       "      <td>30</td>\n",
       "      <td>95.70</td>\n",
       "      <td>2</td>\n",
       "      <td>2871.00</td>\n",
       "      <td>2/24/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2003</td>\n",
       "      <td>...</td>\n",
       "      <td>897 Long Airport Avenue</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NYC</td>\n",
       "      <td>NY</td>\n",
       "      <td>10022</td>\n",
       "      <td>USA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Yu</td>\n",
       "      <td>Kwai</td>\n",
       "      <td>Small</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10121</td>\n",
       "      <td>34</td>\n",
       "      <td>81.35</td>\n",
       "      <td>5</td>\n",
       "      <td>2765.90</td>\n",
       "      <td>5/7/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>2003</td>\n",
       "      <td>...</td>\n",
       "      <td>59 rue de l'Abbaye</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Reims</td>\n",
       "      <td>NaN</td>\n",
       "      <td>51100</td>\n",
       "      <td>France</td>\n",
       "      <td>EMEA</td>\n",
       "      <td>Henriot</td>\n",
       "      <td>Paul</td>\n",
       "      <td>Small</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10134</td>\n",
       "      <td>41</td>\n",
       "      <td>94.74</td>\n",
       "      <td>2</td>\n",
       "      <td>3884.34</td>\n",
       "      <td>7/1/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "      <td>2003</td>\n",
       "      <td>...</td>\n",
       "      <td>27 rue du Colonel Pierre Avia</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Paris</td>\n",
       "      <td>NaN</td>\n",
       "      <td>75508</td>\n",
       "      <td>France</td>\n",
       "      <td>EMEA</td>\n",
       "      <td>Da Cunha</td>\n",
       "      <td>Daniel</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10145</td>\n",
       "      <td>45</td>\n",
       "      <td>83.26</td>\n",
       "      <td>6</td>\n",
       "      <td>3746.70</td>\n",
       "      <td>8/25/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "      <td>2003</td>\n",
       "      <td>...</td>\n",
       "      <td>78934 Hillside Dr.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Pasadena</td>\n",
       "      <td>CA</td>\n",
       "      <td>90003</td>\n",
       "      <td>USA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Young</td>\n",
       "      <td>Julie</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10159</td>\n",
       "      <td>49</td>\n",
       "      <td>100.00</td>\n",
       "      <td>14</td>\n",
       "      <td>5205.27</td>\n",
       "      <td>10/10/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>2003</td>\n",
       "      <td>...</td>\n",
       "      <td>7734 Strong St.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>San Francisco</td>\n",
       "      <td>CA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>USA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Brown</td>\n",
       "      <td>Julie</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ORDERNUMBER  QUANTITYORDERED  PRICEEACH  ORDERLINENUMBER    SALES  \\\n",
       "0        10107               30      95.70                2  2871.00   \n",
       "1        10121               34      81.35                5  2765.90   \n",
       "2        10134               41      94.74                2  3884.34   \n",
       "3        10145               45      83.26                6  3746.70   \n",
       "4        10159               49     100.00               14  5205.27   \n",
       "\n",
       "         ORDERDATE   STATUS  QTR_ID  MONTH_ID  YEAR_ID  ...  \\\n",
       "0   2/24/2003 0:00  Shipped       1         2     2003  ...   \n",
       "1    5/7/2003 0:00  Shipped       2         5     2003  ...   \n",
       "2    7/1/2003 0:00  Shipped       3         7     2003  ...   \n",
       "3   8/25/2003 0:00  Shipped       3         8     2003  ...   \n",
       "4  10/10/2003 0:00  Shipped       4        10     2003  ...   \n",
       "\n",
       "                    ADDRESSLINE1  ADDRESSLINE2           CITY STATE  \\\n",
       "0        897 Long Airport Avenue           NaN            NYC    NY   \n",
       "1             59 rue de l'Abbaye           NaN          Reims   NaN   \n",
       "2  27 rue du Colonel Pierre Avia           NaN          Paris   NaN   \n",
       "3             78934 Hillside Dr.           NaN       Pasadena    CA   \n",
       "4                7734 Strong St.           NaN  San Francisco    CA   \n",
       "\n",
       "  POSTALCODE COUNTRY TERRITORY CONTACTLASTNAME CONTACTFIRSTNAME DEALSIZE  \n",
       "0      10022     USA       NaN              Yu             Kwai    Small  \n",
       "1      51100  France      EMEA         Henriot             Paul    Small  \n",
       "2      75508  France      EMEA        Da Cunha           Daniel   Medium  \n",
       "3      90003     USA       NaN           Young            Julie   Medium  \n",
       "4        NaN     USA       NaN           Brown            Julie   Medium  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(r'C:\\Python Scripts\\Datasets\\sales_data_sample.csv', encoding = 'unicode_escape')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**First observations:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ORDERNUMBER', 'QUANTITYORDERED', 'PRICEEACH', 'ORDERLINENUMBER',\n",
       "       'SALES', 'ORDERDATE', 'STATUS', 'QTR_ID', 'MONTH_ID', 'YEAR_ID',\n",
       "       'PRODUCTLINE', 'MSRP', 'PRODUCTCODE', 'CUSTOMERNAME', 'PHONE',\n",
       "       'ADDRESSLINE1', 'ADDRESSLINE2', 'CITY', 'STATE', 'POSTALCODE',\n",
       "       'COUNTRY', 'TERRITORY', 'CONTACTLASTNAME', 'CONTACTFIRSTNAME',\n",
       "       'DEALSIZE'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 2823 entries, 0 to 2822\n",
      "Data columns (total 25 columns):\n",
      " #   Column            Non-Null Count  Dtype  \n",
      "---  ------            --------------  -----  \n",
      " 0   ORDERNUMBER       2823 non-null   int64  \n",
      " 1   QUANTITYORDERED   2823 non-null   int64  \n",
      " 2   PRICEEACH         2823 non-null   float64\n",
      " 3   ORDERLINENUMBER   2823 non-null   int64  \n",
      " 4   SALES             2823 non-null   float64\n",
      " 5   ORDERDATE         2823 non-null   object \n",
      " 6   STATUS            2823 non-null   object \n",
      " 7   QTR_ID            2823 non-null   int64  \n",
      " 8   MONTH_ID          2823 non-null   int64  \n",
      " 9   YEAR_ID           2823 non-null   int64  \n",
      " 10  PRODUCTLINE       2823 non-null   object \n",
      " 11  MSRP              2823 non-null   int64  \n",
      " 12  PRODUCTCODE       2823 non-null   object \n",
      " 13  CUSTOMERNAME      2823 non-null   object \n",
      " 14  PHONE             2823 non-null   object \n",
      " 15  ADDRESSLINE1      2823 non-null   object \n",
      " 16  ADDRESSLINE2      302 non-null    object \n",
      " 17  CITY              2823 non-null   object \n",
      " 18  STATE             1337 non-null   object \n",
      " 19  POSTALCODE        2747 non-null   object \n",
      " 20  COUNTRY           2823 non-null   object \n",
      " 21  TERRITORY         1749 non-null   object \n",
      " 22  CONTACTLASTNAME   2823 non-null   object \n",
      " 23  CONTACTFIRSTNAME  2823 non-null   object \n",
      " 24  DEALSIZE          2823 non-null   object \n",
      "dtypes: float64(2), int64(7), object(16)\n",
      "memory usage: 551.5+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Shape of data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2823, 25)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data cleaning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "There are in data some unnecessary columns such as „ADDRESSLINE1”, „ADDRESSLINE2”, „POSTALCODE”, „TERRITORY”, „PHONE”.\n",
    "\n",
    "In the first step I remove them. Next I regroup the column 'PRODUCTINITIAL' use only three first signs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove columns\n",
    "data=['ADDRESSLINE1','ADDRESSLINE2','POSTALCODE', 'TERRITORY', 'PHONE', 'CITY' , 'STATE','CONTACTFIRSTNAME', 'CONTACTLASTNAME', 'MONTH_ID']\n",
    "df.drop(data,axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Regrouping product code\n",
    "df['PRODUCTINITIAL'] = df['PRODUCTCODE'].str[:3]\n",
    "df.drop('PRODUCTCODE',axis=1,inplace=True)"
   ]
  },
  {
   "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>ORDERNUMBER</th>\n",
       "      <th>QUANTITYORDERED</th>\n",
       "      <th>PRICEEACH</th>\n",
       "      <th>ORDERLINENUMBER</th>\n",
       "      <th>SALES</th>\n",
       "      <th>ORDERDATE</th>\n",
       "      <th>STATUS</th>\n",
       "      <th>QTR_ID</th>\n",
       "      <th>YEAR_ID</th>\n",
       "      <th>PRODUCTLINE</th>\n",
       "      <th>MSRP</th>\n",
       "      <th>CUSTOMERNAME</th>\n",
       "      <th>COUNTRY</th>\n",
       "      <th>DEALSIZE</th>\n",
       "      <th>PRODUCTINITIAL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10107</td>\n",
       "      <td>30</td>\n",
       "      <td>95.70</td>\n",
       "      <td>2</td>\n",
       "      <td>2871.00</td>\n",
       "      <td>2/24/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>1</td>\n",
       "      <td>2003</td>\n",
       "      <td>Motorcycles</td>\n",
       "      <td>95</td>\n",
       "      <td>Land of Toys Inc.</td>\n",
       "      <td>USA</td>\n",
       "      <td>Small</td>\n",
       "      <td>S10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10121</td>\n",
       "      <td>34</td>\n",
       "      <td>81.35</td>\n",
       "      <td>5</td>\n",
       "      <td>2765.90</td>\n",
       "      <td>5/7/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>2</td>\n",
       "      <td>2003</td>\n",
       "      <td>Motorcycles</td>\n",
       "      <td>95</td>\n",
       "      <td>Reims Collectables</td>\n",
       "      <td>France</td>\n",
       "      <td>Small</td>\n",
       "      <td>S10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10134</td>\n",
       "      <td>41</td>\n",
       "      <td>94.74</td>\n",
       "      <td>2</td>\n",
       "      <td>3884.34</td>\n",
       "      <td>7/1/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>3</td>\n",
       "      <td>2003</td>\n",
       "      <td>Motorcycles</td>\n",
       "      <td>95</td>\n",
       "      <td>Lyon Souveniers</td>\n",
       "      <td>France</td>\n",
       "      <td>Medium</td>\n",
       "      <td>S10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10145</td>\n",
       "      <td>45</td>\n",
       "      <td>83.26</td>\n",
       "      <td>6</td>\n",
       "      <td>3746.70</td>\n",
       "      <td>8/25/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>3</td>\n",
       "      <td>2003</td>\n",
       "      <td>Motorcycles</td>\n",
       "      <td>95</td>\n",
       "      <td>Toys4GrownUps.com</td>\n",
       "      <td>USA</td>\n",
       "      <td>Medium</td>\n",
       "      <td>S10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10159</td>\n",
       "      <td>49</td>\n",
       "      <td>100.00</td>\n",
       "      <td>14</td>\n",
       "      <td>5205.27</td>\n",
       "      <td>10/10/2003 0:00</td>\n",
       "      <td>Shipped</td>\n",
       "      <td>4</td>\n",
       "      <td>2003</td>\n",
       "      <td>Motorcycles</td>\n",
       "      <td>95</td>\n",
       "      <td>Corporate Gift Ideas Co.</td>\n",
       "      <td>USA</td>\n",
       "      <td>Medium</td>\n",
       "      <td>S10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ORDERNUMBER  QUANTITYORDERED  PRICEEACH  ORDERLINENUMBER    SALES  \\\n",
       "0        10107               30      95.70                2  2871.00   \n",
       "1        10121               34      81.35                5  2765.90   \n",
       "2        10134               41      94.74                2  3884.34   \n",
       "3        10145               45      83.26                6  3746.70   \n",
       "4        10159               49     100.00               14  5205.27   \n",
       "\n",
       "         ORDERDATE   STATUS  QTR_ID  YEAR_ID  PRODUCTLINE  MSRP  \\\n",
       "0   2/24/2003 0:00  Shipped       1     2003  Motorcycles    95   \n",
       "1    5/7/2003 0:00  Shipped       2     2003  Motorcycles    95   \n",
       "2    7/1/2003 0:00  Shipped       3     2003  Motorcycles    95   \n",
       "3   8/25/2003 0:00  Shipped       3     2003  Motorcycles    95   \n",
       "4  10/10/2003 0:00  Shipped       4     2003  Motorcycles    95   \n",
       "\n",
       "               CUSTOMERNAME COUNTRY DEALSIZE PRODUCTINITIAL  \n",
       "0         Land of Toys Inc.     USA    Small            S10  \n",
       "1        Reims Collectables  France    Small            S10  \n",
       "2           Lyon Souveniers  France   Medium            S10  \n",
       "3         Toys4GrownUps.com     USA   Medium            S10  \n",
       "4  Corporate Gift Ideas Co.     USA   Medium            S10  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Basic data analysis\n",
    "\n",
    "Checking unique values in each column:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of unique values in ORDERNUMBER: 307\n",
      "Number of unique values in QUANTITYORDERED: 58\n",
      "Number of unique values in PRICEEACH: 1016\n",
      "Number of unique values in ORDERLINENUMBER: 18\n",
      "Number of unique values in SALES: 2763\n",
      "Number of unique values in ORDERDATE: 252\n",
      "Number of unique values in STATUS: 6\n",
      "Number of unique values in QTR_ID: 4\n",
      "Number of unique values in YEAR_ID: 3\n",
      "Number of unique values in PRODUCTLINE: 7\n",
      "Number of unique values in MSRP: 80\n",
      "Number of unique values in CUSTOMERNAME: 92\n",
      "Number of unique values in COUNTRY: 19\n",
      "Number of unique values in DEALSIZE: 3\n",
      "Number of unique values in PRODUCTINITIAL: 8\n"
     ]
    }
   ],
   "source": [
    "for column in df:\n",
    "    print(f'Number of unique values in {column}:', df[column].nunique())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking the missing values in data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ORDERNUMBER        0\n",
       "QUANTITYORDERED    0\n",
       "PRICEEACH          0\n",
       "ORDERLINENUMBER    0\n",
       "SALES              0\n",
       "ORDERDATE          0\n",
       "STATUS             0\n",
       "QTR_ID             0\n",
       "YEAR_ID            0\n",
       "PRODUCTLINE        0\n",
       "MSRP               0\n",
       "CUSTOMERNAME       0\n",
       "COUNTRY            0\n",
       "DEALSIZE           0\n",
       "PRODUCTINITIAL     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking unique values in CUSTOMERNAME:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "92"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['CUSTOMERNAME'].nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking unique values in YEAR_ID:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2003, 2004, 2005], dtype=int64)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['YEAR_ID'].unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After first observations one can see that:\n",
    "- there are no duplicates;\n",
    "- no missing values;\n",
    "- in data are 92 unique clients for whom I will perform RFM analysis."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RFM Analysis\n",
    "\n",
    "For making RFM analysis I need only four columns: 'CUSTOMERNAME', 'ORDERNUMBER', 'ORDERDATE' and 'SALES':"
   ]
  },
  {
   "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>CUSTOMERNAME</th>\n",
       "      <th>ORDERNUMBER</th>\n",
       "      <th>ORDERDATE</th>\n",
       "      <th>SALES</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Land of Toys Inc.</td>\n",
       "      <td>10107</td>\n",
       "      <td>2/24/2003 0:00</td>\n",
       "      <td>2871.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Reims Collectables</td>\n",
       "      <td>10121</td>\n",
       "      <td>5/7/2003 0:00</td>\n",
       "      <td>2765.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Lyon Souveniers</td>\n",
       "      <td>10134</td>\n",
       "      <td>7/1/2003 0:00</td>\n",
       "      <td>3884.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Toys4GrownUps.com</td>\n",
       "      <td>10145</td>\n",
       "      <td>8/25/2003 0:00</td>\n",
       "      <td>3746.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Corporate Gift Ideas Co.</td>\n",
       "      <td>10159</td>\n",
       "      <td>10/10/2003 0:00</td>\n",
       "      <td>5205.27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               CUSTOMERNAME  ORDERNUMBER        ORDERDATE    SALES\n",
       "0         Land of Toys Inc.        10107   2/24/2003 0:00  2871.00\n",
       "1        Reims Collectables        10121    5/7/2003 0:00  2765.90\n",
       "2           Lyon Souveniers        10134    7/1/2003 0:00  3884.34\n",
       "3         Toys4GrownUps.com        10145   8/25/2003 0:00  3746.70\n",
       "4  Corporate Gift Ideas Co.        10159  10/10/2003 0:00  5205.27"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "col =['CUSTOMERNAME', 'ORDERNUMBER', 'ORDERDATE', 'SALES']\n",
    "RFM = df[col]\n",
    "RFM.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2823, 4)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RFM.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In RFM clustering it is usually assumed that only operations from the last 12 months are taken into account. The reason for this is simple. The older operations can say little about the current situation of the seller, buyer, and the product itself. In order to set them you need to adopt an agreed date for the grouping. In the data for \"today\" I will assume the day of the newest operation in the set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "RFM['ORDERDATE'] = pd.to_datetime(RFM['ORDERDATE'], errors='coerce')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>CUSTOMERNAME</th>\n",
       "      <th>ORDERNUMBER</th>\n",
       "      <th>ORDERDATE</th>\n",
       "      <th>SALES</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Land of Toys Inc.</td>\n",
       "      <td>10107</td>\n",
       "      <td>2003-02-24</td>\n",
       "      <td>2871.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Reims Collectables</td>\n",
       "      <td>10121</td>\n",
       "      <td>2003-05-07</td>\n",
       "      <td>2765.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Lyon Souveniers</td>\n",
       "      <td>10134</td>\n",
       "      <td>2003-07-01</td>\n",
       "      <td>3884.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Toys4GrownUps.com</td>\n",
       "      <td>10145</td>\n",
       "      <td>2003-08-25</td>\n",
       "      <td>3746.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Corporate Gift Ideas Co.</td>\n",
       "      <td>10159</td>\n",
       "      <td>2003-10-10</td>\n",
       "      <td>5205.27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               CUSTOMERNAME  ORDERNUMBER  ORDERDATE    SALES\n",
       "0         Land of Toys Inc.        10107 2003-02-24  2871.00\n",
       "1        Reims Collectables        10121 2003-05-07  2765.90\n",
       "2           Lyon Souveniers        10134 2003-07-01  3884.34\n",
       "3         Toys4GrownUps.com        10145 2003-08-25  3746.70\n",
       "4  Corporate Gift Ideas Co.        10159 2003-10-10  5205.27"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RFM.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2005-05-31 00:00:00')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RFM['ORDERDATE'].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "now = dt.datetime(2005,5,31)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Create the RFM Table:**\n",
    "\n",
    "Calculation of the \"Recency\" variable - information about how long ago the customer was shopping in the store:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfm_df = RFM.groupby('CUSTOMERNAME').agg({'ORDERDATE': lambda x: (now - x.max()).days}) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\"Frequency\" and \"Monetary value\" I made by the using groupby method:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfm_df['Frequency'] = RFM.groupby(by=['CUSTOMERNAME'])['ORDERNUMBER'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfm_df['MonetaryValue'] = RFM.groupby(by='CUSTOMERNAME').agg({'SALES': 'sum'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>ORDERDATE</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CUSTOMERNAME</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AV Stores, Co.</th>\n",
       "      <td>195</td>\n",
       "      <td>51</td>\n",
       "      <td>157807.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Alpha Cognac</th>\n",
       "      <td>64</td>\n",
       "      <td>20</td>\n",
       "      <td>70488.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Amica Models &amp; Co.</th>\n",
       "      <td>264</td>\n",
       "      <td>26</td>\n",
       "      <td>94117.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Anna's Decorations, Ltd</th>\n",
       "      <td>83</td>\n",
       "      <td>46</td>\n",
       "      <td>153996.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Atelier graphique</th>\n",
       "      <td>187</td>\n",
       "      <td>7</td>\n",
       "      <td>24179.96</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         ORDERDATE  Frequency  MonetaryValue\n",
       "CUSTOMERNAME                                                \n",
       "AV Stores, Co.                 195         51      157807.81\n",
       "Alpha Cognac                    64         20       70488.44\n",
       "Amica Models & Co.             264         26       94117.26\n",
       "Anna's Decorations, Ltd         83         46      153996.13\n",
       "Atelier graphique              187          7       24179.96"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I rename 'ORDERDATE' value to 'Recency':"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CUSTOMERNAME</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AV Stores, Co.</th>\n",
       "      <td>195</td>\n",
       "      <td>51</td>\n",
       "      <td>157807.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Alpha Cognac</th>\n",
       "      <td>64</td>\n",
       "      <td>20</td>\n",
       "      <td>70488.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Amica Models &amp; Co.</th>\n",
       "      <td>264</td>\n",
       "      <td>26</td>\n",
       "      <td>94117.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Anna's Decorations, Ltd</th>\n",
       "      <td>83</td>\n",
       "      <td>46</td>\n",
       "      <td>153996.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Atelier graphique</th>\n",
       "      <td>187</td>\n",
       "      <td>7</td>\n",
       "      <td>24179.96</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         Recency  Frequency  MonetaryValue\n",
       "CUSTOMERNAME                                              \n",
       "AV Stores, Co.               195         51      157807.81\n",
       "Alpha Cognac                  64         20       70488.44\n",
       "Amica Models & Co.           264         26       94117.26\n",
       "Anna's Decorations, Ltd       83         46      153996.13\n",
       "Atelier graphique            187          7       24179.96"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.rename(columns={'ORDERDATE': 'Recency'}, inplace=True)\n",
    "rfm_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model: raw calculated RFM variables + K-Means Clustering\n",
    "\n",
    "In this approach I have used raw calculated RFM variables (Recency, Frequency, MonetaryValue) and K-means clustering.\n",
    "\n",
    "There are many outliers in the data then before modeling I apply normalization."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Basic data analysis of RFM table\n",
    "\n",
    "Firstly I plot the data to get more insight:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x1008 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['figure.figsize'] = [16, 14]\n",
    "rfm_df.plot(kind=\"density\", subplots=True, layout=(3,3), sharex=False, sharey=False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Histograms of variables:Recency, Frequency and MonetaryValue:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n",
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n",
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x268af1fdeb8>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,10))\n",
    "# Recency variable distribution \n",
    "plt.subplot(3, 1, 1); sns.distplot(rfm_df['Recency'])\n",
    "# Frequency variable distribution \n",
    "plt.subplot(3, 1, 2); sns.distplot(rfm_df['Frequency'])\n",
    "# MonetaryValue variable distribution \n",
    "plt.subplot(3, 1, 3); sns.distplot(rfm_df['MonetaryValue'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on above charts it can be seen that the variables doesn't have a symmetric distribution. The Frequency and MonetarValue variables are skewed distributions. In addition many outliers are observed. \n",
    "\n",
    "Checking statistics:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>92.000000</td>\n",
       "      <td>92.000000</td>\n",
       "      <td>92.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>181.826087</td>\n",
       "      <td>30.684783</td>\n",
       "      <td>109050.313587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>131.420524</td>\n",
       "      <td>30.934503</td>\n",
       "      <td>110308.612808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>9129.350000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>80.250000</td>\n",
       "      <td>19.750000</td>\n",
       "      <td>70129.432500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>185.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>86522.610000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>229.250000</td>\n",
       "      <td>32.500000</td>\n",
       "      <td>120575.875000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>508.000000</td>\n",
       "      <td>259.000000</td>\n",
       "      <td>912294.110000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Recency   Frequency  MonetaryValue\n",
       "count   92.000000   92.000000      92.000000\n",
       "mean   181.826087   30.684783  109050.313587\n",
       "std    131.420524   30.934503  110308.612808\n",
       "min      0.000000    3.000000    9129.350000\n",
       "25%     80.250000   19.750000   70129.432500\n",
       "50%    185.000000   26.000000   86522.610000\n",
       "75%    229.250000   32.500000  120575.875000\n",
       "max    508.000000  259.000000  912294.110000"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From above analysis one can see that the minimum value for Recency for a specific client identifier is 0. So this transaction does not make sense and must be deleted. \n",
    "\n",
    "One can also see that there are not get constant mean and standard deviation values. \n",
    "\n",
    "Then I drop client with 0 value from Recency variable:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CUSTOMERNAME</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Euro Shopping Channel</th>\n",
       "      <td>0</td>\n",
       "      <td>259</td>\n",
       "      <td>912294.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>La Rochelle Gifts</th>\n",
       "      <td>0</td>\n",
       "      <td>53</td>\n",
       "      <td>180124.90</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Recency  Frequency  MonetaryValue\n",
       "CUSTOMERNAME                                            \n",
       "Euro Shopping Channel        0        259      912294.11\n",
       "La Rochelle Gifts            0         53      180124.90"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df[rfm_df['Recency'] == 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfm_df = rfm_df[rfm_df['Recency'] > 0]\n",
    "rfm_df.reset_index(drop=True,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw_data = rfm_df[['Recency','Frequency','MonetaryValue']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>90.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>90.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>185.866667</td>\n",
       "      <td>27.900000</td>\n",
       "      <td>99335.664889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>130.000104</td>\n",
       "      <td>19.475741</td>\n",
       "      <td>70991.470146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>9129.350000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>84.500000</td>\n",
       "      <td>19.250000</td>\n",
       "      <td>69411.417500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>187.500000</td>\n",
       "      <td>25.500000</td>\n",
       "      <td>85363.790000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>229.750000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>117934.592500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>508.000000</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>654858.060000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Recency   Frequency  MonetaryValue\n",
       "count   90.000000   90.000000      90.000000\n",
       "mean   185.866667   27.900000   99335.664889\n",
       "std    130.000104   19.475741   70991.470146\n",
       "min      1.000000    3.000000    9129.350000\n",
       "25%     84.500000   19.250000   69411.417500\n",
       "50%    187.500000   25.500000   85363.790000\n",
       "75%    229.750000   32.000000  117934.592500\n",
       "max    508.000000  180.000000  654858.060000"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this step I apply log transformation to the data and then normalize it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Unskew the data\n",
    "data_log = np.log(raw_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Initialize a standard scaler and fit it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "scaler.fit(data_log)\n",
    "\n",
    "data_sc = scaler.transform(data_log)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_norm = pd.DataFrame(data_sc, columns=raw_data.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.380566</td>\n",
       "      <td>1.417734</td>\n",
       "      <td>1.124103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.509786</td>\n",
       "      <td>-0.367558</td>\n",
       "      <td>-0.359312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.622670</td>\n",
       "      <td>0.132816</td>\n",
       "      <td>0.172800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.302040</td>\n",
       "      <td>1.220944</td>\n",
       "      <td>1.079099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.347089</td>\n",
       "      <td>-2.369750</td>\n",
       "      <td>-2.328643</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Recency  Frequency  MonetaryValue\n",
       "0  0.380566   1.417734       1.124103\n",
       "1 -0.509786  -0.367558      -0.359312\n",
       "2  0.622670   0.132816       0.172800\n",
       "3 -0.302040   1.220944       1.079099\n",
       "4  0.347089  -2.369750      -2.328643"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_norm.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking for skewness in the data now:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n",
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n",
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x268af636a20>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,10))\n",
    "# Recency variable distribution \n",
    "plt.subplot(3, 1, 1); sns.distplot(df_norm['Recency'])\n",
    "# Frequency variable distribution \n",
    "plt.subplot(3, 1, 2); sns.distplot(df_norm['Frequency'])\n",
    "# MonetaryValue variable distribution \n",
    "plt.subplot(3, 1, 3); sns.distplot(df_norm['MonetaryValue'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After applied normalization to the data one can see that skewness has been removed."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### K-Means clustering model\n",
    "\n",
    "K-means is an unsupervised machine learning algorithm, which is used for data clustering. In k-means algorithm number of clusters K is predetermined and the algorithm iteratively assigns each data point to one of the K clusters based on the feature similarity.\n",
    "\n",
    "The optimal number of clusters I will find using the elbow method:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "group = []\n",
    "for i in range(1, 15):\n",
    "    kmeans = KMeans(n_clusters=i, random_state=1)\n",
    "    kmeans.fit(df_norm)\n",
    "    group.append([i, kmeans.inertia_])\n",
    "groups = pd.DataFrame(group, columns = ['number of group', 'inertia'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>number of group</th>\n",
       "      <th>inertia</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>270.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>167.213622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>111.918503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>81.792459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>66.786975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>51.962335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>38.781916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>31.434893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>25.288073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>20.261730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>16.875419</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>15.598057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>13.534581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>12.620361</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    number of group     inertia\n",
       "0                 1  270.000000\n",
       "1                 2  167.213622\n",
       "2                 3  111.918503\n",
       "3                 4   81.792459\n",
       "4                 5   66.786975\n",
       "5                 6   51.962335\n",
       "6                 7   38.781916\n",
       "7                 8   31.434893\n",
       "8                 9   25.288073\n",
       "9                10   20.261730\n",
       "10               11   16.875419\n",
       "11               12   15.598057\n",
       "12               13   13.534581\n",
       "13               14   12.620361"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnIAAAHHCAYAAAA7wbXOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdd3hU153/8ffMaEZdQqACAokioTPCFKPBBUx3A/eeuMYbx07ibLzJ5pdNLxsnTt+Nvc4mWac4cbfjXmi2aTYYG2FsMOgI0cH0DgL13x93hIVMkYSkO+Xzep55RnPvuXe+c5LYn5x7z7mepqYmRERERCT6eN0uQEREREQ6RkFOREREJEopyImIiIhEKQU5ERERkSilICciIiISpRTkRESigDHG43YNIhJ5EtwuQETilzHmYeBzJ2tjrfWE284BEqy1Y8Of1wFvWWtv6dIi28AYczvwt/DHEdbaD4/TZiiwLPxxoLV2XTvOfwcwDPha+POPgR8BfmttfYcLP/H3TQRmAxdaa1/v7POLSOdRkBMRt+0ArnG7iE7SCHwG+FSQA246jfP+AHjrNI4XkRilICcibqu11sZKSHkLuB743nH2fRZ4HxjZrRWJSExTkBORaJZgjPkv4Hace35fBv6ftXZbcwNjTAj4KXAWEADmA9+21i4zxowAlgI3WGufCbcvA8qBB621Xw1vSwB2Ar+w1v7iJPU8BfzeGDPSWvt+ixrOBQqAh2gV5IwxQ4BfAhPCv2EO8A1rrQ3vb378Tn9jzM3Nl5rDLjLG3AucAWwB7rfW/q7FuTOBHwJXAX2B1cDvrLUPtWjjAb4JfAnogxNGHz7JbxSRCKLJDiLiOmNMwglep/pn1LXAWOALwLeAS4AXW5x3ErAQSAK+GH4VAguMMUOstR8AG4ELW5zzgvD7pBbbxgCZOEHxZN7AuVR8Q6vtNwKvh/cdZYwpBhaEa7or/Dv6Am8bYwrCzcYBW4GZ4b9b+itOOLwK53LufxtjpobPnYwTyj4H/A7n8vUC4P+MMT9scY77gJ8DT+P05wrgL6f4nSISITQiJyJu6wvUnWDfb4H/d5Jj9wEXWGv3AxhjtgHPG2Musda+BvwCWItz0359uM10oApnlO4a4FVgSotzng8sBkLGmBxr7Q5gKrDWWvvRKX5LA/BPnMur3wl/nw8n2H0LZ0SwpR8D9cAka+3ucPtp4fq+D3zRWvuWMaYG2HGcS9B3WWtfCh/3NrAHJ4hOwxmlHApMtNbODbd/LTy6+F1jzB+AGuDfcUYfv92iTSpwxyl+q4hEAI3IiYjbtuNc9jze63cnOQ7g1eYQF/YSTjC6IBxGzgKeaTmz01q7B2dkbWJ40yvAAGNMkTEmgDPC90uciQvNbS4Jn7stngKKwpd0wRnZywSeP07b83Eupe5vHoUEDuGM3l3Uhu9qDmhYaw/g9GVWeNNEYHOLENfs70AiMDr8Chyntsfb8N0iEgE0Iicibquz1i7u4LFbW36w1jYaY3bihJkegKd1mxbHZYb/fgOoxrm8agE/zojWEmBieKRrOM7IVVvMBzbjjMKV41xWfcVae8AY07ptNs7lzOONSJ5olLKlQ60+N/LJ/0HvyYl/Ozj90xxwd7Rq83EbvltEIoBG5EQkmvVs+SE8opWNMzK1F2gCeh/nuHxgF4C19gjwJk6QmwS8a609FN42Eeey6j5gXlsKstY2As8A14dH+K4BnjhB873htscbjRzTlu87id2c+LeDM3mjOcC1bpd9mt8tIt1EQU5EotnFxpjEFp+vw7nSMDscxhbjBKqjVx+MMT2Ayzh2XbZXcELcBJwAR/h9CPAvwHRrbVtGyJo9BQzEuU/OA7x2gnZzcGacfmCtXdz8Au7h2HXnGtrx3c3mAn2NMeNbbb8FZyTuHZzJD9U4S6O0dFUHvk9EXKBLqyLitoAxZuxJ9i+31u49wb5ewIvGmPuBYuBnOAFsRnj/d8J/zzLG/A/O/WDfxpnF+uMW53kF+CMwvsX2t4Ba4Dyc8NNm1tp3wk+e+B7wqLW25gRN/xMnUM00xvwe51Lp53EC6a0t2u0FRhhjzsd54kJbPAx8BXgu/CSINcAVOMH0Zy0mV/wI+LUx5jDOJeVxODNoRSQKaERORNyWg3Nf2Yle557k2D/hzEp9CieAPQZcaa1tArDWvoEzi9MLPIoT1jYA57R8jJa1djPOYr01OMuVYK2txglZDTgBp72ewrnf7oQTB6y1y3EmVxzCWUrkGaA/zrp2j7Zoeh/O5c8XcZYqOaVw/RNwJjL8AHgOZ3LDXdba77do9xucNeSmhNtOQjNWRaKGp6mp6dStRERERCTiaEROREREJEopyImIiIhEKQU5ERERkSilICciIiISpeJu+ZHy8vJEnMU2t9CxtZlEREREuosP6AO8FwqFPrWUUdwFOZwQN9/tIkRERETaYRzHLmQOxGeQ2wJQUlJCIBBwuxbXLF++nKFDh7pdRtRRv3WM+q1j1G8do37rGPVbx3R1v9XW1lJZWQnh/NJaPAa5BoBAIEBiYuKp2sa0eP/9HaV+6xj1W8eo3zpG/dYx6reO6aZ+O+7tYJrsICIiIhKlFOREREREopSCnIiIiEiUUpATERERiVIKciIiIiJRSkFOREREJEopyImIiIhEKQU5ERERkSilICciIiISpRTkRERERKKUgpyIiIhIlFKQExEREYlSCnIiIiIiUUpBros8MaOCj3cedLsMERERiWEKcl2gqamJ5+dW8cLc1W6XIiIiIjFMQa4LeDwehhfnUF6xnaamJrfLERERkRilINdFQsFctu+uZvMOXV4VERGRrqEg10VGmlwAllRsd7kSERERiVUKcl2kd69U+uakUW4V5ERERKRrKMh1oVBpLsurdlJT1+B2KSIiIhKDFOS6UMjkUVvfyPLVO90uRURERGKQglwXOqOoF4EEr+6TExERkS6hINeFEv0+hhVnU16xze1SREREJAYpyHWxsmAum3ccYuuuQ26XIiIiIjFGQa6LhYJ5ACzR7FURERHpZApyXSw/O5XevVIoX6kgJyIiIp1LQa6LeTweykwuH1btoK5ey5CIiIhI51GQ6wahYB5HahtYsXa326WIiIhIDFGQ6wbDirNJ8GkZEhEREelcCnLdIDkxgTMG9dQyJCIiItKpFOS6SZnJY/3WA+zce9jtUkRERCRGKMh1k1AwF9AyJCIiItJ5FOS6SWHvdLIzk3R5VURERDqNglw38Xg8lAXzWFq5g/qGRrfLERERkRigINeNyoK5VB+px67f43YpIiIiEgMU5LrRmYNz8Ho9urwqIiIinSKhu7/QGJMO/AS4GsgGKoCfWGtfCu//OfDt4xzqt9bWh9t8BfgG0AdYCtxjrX2vG8o/LanJfkoH9GSJ3c5tlwxxuxwRERGJcm6MyD0MXAZ8ATgTeA543hgzObx/OPAQTkg7+moR4m4HfgV8HwgBFphhjMntvp/QcWUml9Wb9rHnwBG3SxEREZEo160jcsaY3sA1wGXW2tfDm+8zxpwP3AG8CQwDXrbWbj3Bab4L/N5a+3j4nHcAq4EvAvd2Zf2dIRTM5ZFpK3nfbmfyqEK3yxEREZEo1t0jcoeAqcC8VtubgJ7GmB5AAbDieAcbY/KAwcDs5m3W2gZgPjC+KwrubAPzM+mRnki5HtclIiIip6lbR+SstQeA6S23GWPOBSYD9+CMxgHcbIz5CxAA5gDfttZuAfqG929qdeqPgbO6qOxO5fV6KDO5vLdiGw2NTfi8HrdLEhERkSjV7ZMdWjLGlALPA4uAP+HcNwewD7gW5/64+4A5xpiRQEp4f02rU9UASe357uXLl3ew6tOXFajmQHUtr8x6h37ZAdfqKC8vd+27o5n6rWPUbx2jfusY9VvHqN86xs1+cy3IGWPG44S49cCl1to6Y8wfgSettc0LrX1ojFkObASuwpnYAJDY6nSJwMH2fP/QoUNJTGx9mu4xOFjLcwunUU0WoVDQlRrKy8sJhUKufHc0U791jPqtY9RvHaN+6xj1W8d0db/V1NScdPDJlXXkjDE3A7OAcmCCtXY3gLW2qUWII7xtM7ALKAQ2hDfntzplPp++3BqxMlIDlBRkUa7nroqIiMhp6PYgZ4y5CXgEeBpnJO5Ai333G2Peb9V+IM56cx9Za3fgjMpNbLHfB4wD5nZ99Z2nLJjLqg172H+o1u1SREREJEp19/Ij/XDWiJsN/AfQyxjTvLsWeAa42xhzP/Agzkjb/cC7wKvhdr8FHjDG2PD2bwJp4fNGjVAwlydmWpZWbmf8yH5ulyMiIiJRqLtH5K7BmbAwGWem6ZYWr5estW/hLBZ8NvA+8CzO5ddLrLWNANbah3AWA74XWAwUAxdZa3d27085PcUFWaSn+LUMiYiIiHRYdy8/8gDwwCnazABmnKLNb3FG5qKWz+thZEkuS+x2Ghub8GoZEhEREWknVyY7iCNUmsveAzWs/Xif26WIiIhIFFKQc9HIEufxsEs0e1VEREQ6QEHORVkZSQzqm6n75ERERKRDFORcFgrmsnLdbg4drnO7FBEREYkyCnIuCwXzaGxs4oNVO9wuRURERKKMgpzLTP8sUpISdJ+ciIiItJuCnMsSfF5GDM6hvGI7TU1NbpcjIiIiUURBLgKEgnns3HuYDdsOnLqxiIiISJiCXAQoM+FlSDR7VURERNpBQS4C5GQlU9g7XUFORERE2kVBLkKEgnksX7OLwzX1bpciIiIiUUJBLkKETC71DY0sW73T7VJEREQkSijIRYghg3qSFPDp8qqIiIi0mYJchPAn+BhenEN5xTYtQyIiIiJtoiAXQcqCuWzdVc2WnYfcLkVERESigIJcBAkFnWVIynV5VURERNpAQS6C9O6VSn52KuUV29wuRURERKKAglyECZXmsWz1LmrqGtwuRURERCKcglyEKTO51NY18NGaXW6XIiIiIhFOQS7CDC3qhT/Bq8urIiIickoKchEmKZDAsKJsrScnIiIip6QgF4HKgrls2n6Qbbur3S5FREREIpiCXAQqM84yJEusRuVERETkxBTkIlC/3DRye6ZQvlL3yYmIiMiJKchFII/HQ8jk8mHVDurqG90uR0RERCKUglyEKgvmcrimgYp1u90uRURERCKUglyEGl6cTYLPo2VIRERE5IQU5CJUSpKfIQN76bmrIiIickIKchGszOSybst+du077HYpIiIiEoEU5CJYWTC8DIlG5UREROQ4FOQi2IA+GfTMSKJc68mJiIjIcSjIRTCPx0MomMvSyh00NGgZEhERETmWglyEKwvmcuhwHXbDHrdLERERkQijIBfhzizJxev16D45ERER+RQFuQiXluzHFGbpPjkRERH5FAW5KBAK5lK1cS97D9S4XYqIiIhEEAW5KBAK5gHwfqVG5UREROQTCnJRYFDfTDLTArpPTkRERI6hIBcFvF4PI00uS+x2Ghub3C5HREREIoSCXJQIBfPYf6iWqk173S5FREREIoSCXJQYWZKDxwNLNHtVREREwhTkokRmWiLF/XroPjkRERE5SkEuipQFc7Hrd3OgutbtUkRERCQCKMhFkVHBPBqbYGnlDrdLERERkQigIBdFBhf0IC3Zr8urIiIiAijIRRWfz8uZJTkssdtoatIyJCIiIvFOQS7KhIJ57N5fw7ot+90uRURERFymIBdlyoK5AJTr8qqIiEjcU5CLMj0zkhiYn6H75ERERERBLhqFgnmsWLuL6iN1bpciIiIiLlKQi0JlwVwaGpv4YNVOt0sRERERFynIRaHSAT1JTkzQ47pERETinIJcFEoIL0NSXqFlSEREROKZglyUKjO57NhzmE3bD7pdioiIiLhEQS5KlRktQyIiIhLvErr7C40x6cBPgKuBbKAC+Im19qXw/gHAg8B44DDwMPA9a219i3N8BfgG0AdYCtxjrX2v+36F+3J7plCQl0Z5xTaumlDkdjkiIiLiAjdG5B4GLgO+AJwJPAc8b4yZbIwJADOBJmAMcCdwB/CfzQcbY24HfgV8HwgBFphhTHiIKo6EgnksX72LIzX1p24sIiIiMadbg5wxpjdwDfA1a+3r1toqa+19wBycwHYd0B+4zVq7PDxK9y3g34wxyeHTfBf4vbX2cWvtivBx+4EvdudviQRlJpf6hkaWr9nldikiIiLigu4ekTsETAXmtdreBPQExgFLrbV7WuybA6QCZcaYPGAwMLt5p7W2AZiPcyk2rpwxqBeJAR/lK7e5XYqIiIi4oFvvkbPWHgCmt9xmjDkXmAzcA1wEbGp12Mfh9wKce+Y4QZuzOrXYKBDw+xhWlE251pMTERGJS90+2aElY0wp8DywCPgTcBXQOpXUhN+TgJRW21q2SWrPdy9fvrxdtUaqnJQaFu88xMw5i+iV3r7/OMvLy7uoqtimfusY9VvHqN86Rv3WMeq3jnGz31wLcsaY8Tghbj1wqbW2zhhzGEhs1bT580E+GZE7Xpt2Lag2dOhQEhNbnyb69Ol/kGnlb1CXkEMoNKjNx5WXlxMKhbqwstikfusY9VvHqN86Rv3WMeq3junqfqupqTnp4JMr68gZY24GZgHlwARr7e7wro1AfqvmzZ83ARtabWvZpvXl1riQn51Gn+xUFms9ORERkbjT7UHOGHMT8AjwNM5I3IEWu+cBZxpjerTYNgk4ACyx1u7AWW5kYovz+XAmSczt4tIjVsjksmz1TmrrGtwuRURERLpRt15aNcb0Ax7CmXX6H0AvY0zz7lrgBeCnwFPGmG/iLEXyc+C/rLW14Xa/BR4wxljgXeCbQFr4vHGpLJjLK2+vZcXaXZxZEnfL6YmIiMSt7h6RuwZnwsJknJmmW1q8XrLWHgGmhNs2T4D4E86TIACw1j6EsxjwvcBioBi4yFq7s5t+Q8QZVpRNgs+rx3WJiIjEme5efuQB4IFTtKkCLj5Fm9/ijMwJkJSYwNCiXpRXbOeOK9yuRkRERLqLK5MdpPOFgrls3HaA7Xuq3S5FREREuomCXIwoCz9qdokur4qIiMQNBbkYUZCXTk5WMkv0lAcREZG4oSAXIzweD2Uml6WVO6hvaHS7HBEREekGCnIxJBTM5XBNPSvX7T51YxEREYl6CnIxZMTgHHxej+6TExERiRMKcjEkJclP6cCeCnIiIiJxQkEuxpSZXNZ8vI/d+4+4XYqIiIh0MQW5GDOqNA/QMiQiIiLxQEEuxgzok0FWeqKWIREREYkDCnIxxuPxUBbM5X27nYbGJrfLERERkS6kIBeDQiaPg4frWLVhj9uliIiISBdSkItBZ5ocvB4o131yIiIiMU1BLgalpwQoKcxiid3mdikiIiLShRTkYlRZMI9VG/ey72CN26WIiIhIF1GQi1GhYC5NTfB+5Q63SxEREZEuoiAXo4r79SAjNcCSCl1eFRERiVUKcjHK6/UwsiSXJXY7jVqGREREJCYpyMWwUGku+w7WsmbzPrdLERERkS6gIBfDRpbkAlCu2asiIiIxSUEuhvVIT6S4XyblK7WenIiISCxSkItxoWAedv1uDlbXul2KiIiIdDIFuRhXFsylsQk+WLXT7VJERESkkynIxThTmEVqUgLlWoZEREQk5ijIxTifz8uZ4WVImpq0DImIiEgsUZCLA6FgLrv2HWH91gNulyIiIiKdSEEuDpQFnWVI9JQHERGR2KIgFwd6ZSYzoE8G5RVahkRERCSWKMjFiVAwlxVrd1F9pM7tUkRERKSTKMjFibJgLvUNTSyr0jIkIiIisUJBLk6UDuhFcqJPl1dFRERiiIJcnPAneBlenEO5liERERGJGQpycSQUzGX77mo27zjodikiIiLSCRTk4khZMA9Al1dFRERihIJcHMnrmUK/3DSWKMiJiIjEBAW5OFMWzGX56p3U1es+ORERkWinIBdnQiaP2vpG1m0/4nYpIiIicpoU5OLMGUW9SE32817lIbdLERERkdOkIBdnEv0+rp1UTOXHR1ixdpfb5YiIiMhpUJCLQ5ePHURqkpd/vLZSa8qJiIhEMQW5OJSUmMCEoRl8tGYX71fucLscERER6SAFuThVVpRKbs8UHnlthUblREREopSCXJxK8Hm46SJD1aZ9LFi2xe1yREREpAMU5OLYxFABBXlpPDptJQ0NjW6XIyIiIu2kIBfHfF4Pt0wpZdP2g8wu3+R2OSIiItJOCnJxbvSwPhQX9OCJmRXU1Te4XY6IiIi0g4JcnPN4PNw2tZTtew4zfeF6t8sRERGRdlCQE84syWFYUTZPv17JkZp6t8sRERGRNlKQE2dU7pJS9h6s4aX5a9wuR0RERNpIQU4ACA7oydlDevPc7FUcrK51uxwRERFpAwU5OeqWqUGqa+p5dnaV26WIiIhIGyjIyVED8zMZf2Y/Xpq/hj37j7hdjoiIiJxCpwU5Y4xCYQy4aYqhoaGRp1+vdLsUEREROYWEtjY0xniAG4GJQCLgCe/yAqnAuUCfTq5Pull+dhoXntOf6e+s48oJRfTulep2SSIiInIC7RlF+xHwKHA9cA1wGXAxcBNwOfBcp1cnrvjshSV4PR6emGndLkVEREROos0jcsBtwOPArcCPgf7W2tuNMaOA14B2/1vfGPMd4FJr7dgW234OfPs4zf3W2vpwm68A38AZAVwK3GOtfa+93y/H1yszmUvHDuLFuVVcO6mYwt4ZbpckIiIix9GeEbm+wGPW2iagHBgDYK1dDPwC+Hx7vtgYczfws+PsGg48hBPSjr5ahLjbgV8B3wdCOAFyhjEmtz3fLyd37aRiEgMJPDq9wu1SRERE5ATaMyJ3CGgK/70KGGSMSbbWHsYZFftRW05ijMkH/gRM4vijeMOAl621W09wiu8Cv7fWPh4+3x3AauCLwL1t/C1yCplpiVw9sZjHZ1RQuWEPJYVZbpckIiIirbRnRO5d4Pbw31VALTA5/LkUqGnjeULAAZyRt0UtdxhjegAFwIrjHWiMyQMGA7Obt1lrG4D5wPg2fr+00ZXjB5GRGuCRaSvdLkVERESOoz1B7j7gGmPMDGttHfAP4FFjzIs4lzpntOUk1tqXrbU3WWuP9yyoYeH3m40xq4wx640xfzfGNM+G7Rt+39TquI9xAqB0opQkP9efP5illTv4sGqH2+WIiIhIK22+tGqtnWeMOQsYEd70bzhB8DzgKZzJB6draPh9H3Atzv1x9wFzjDEjgZTw/tajfzVAUnu+aPny5adRZmwoLy8/ZZveyU1kpPj4w9OLueOiHDwezymPiXVt6Tf5NPVbx6jfOkb91jHqt45xs9/ac48c1toPgA/Cf9cAd3VyPX8EnrTW7gl//tAYsxzYCFzFJ/fUJbY6LhE42J4vGjp0KImJrU8TP8rLywmFQm1qu69xHQ8+8wGNyf04+4zeXVxZZGtPv8kn1G8do37rGPVbx6jfOqar+62mpuakg08nDXLGmM/jTDzYEf77pKy1f21/iccc3wTsabVtszFmF1AIzApvzgeWtWiWz6cvt0onOf+sQp6bXcUj01YyqjQPr1ejciIiIpHgVCNyfwbGAjvCf59ME3BaQc4Ycz8w3lo7ssW2gUA28FE4UFqcp0vMCO/3AeNwZsJKF0jwebl5SpBfP1rOvKWbmVjWz+2SREREhFMHuYHAlhZ/d7VngLvDge5BnJG2+3FmzL4abvNb4IFwoHsX+CaQhrP2nHSRsSP68s83V/H49ArGjsgnwadH64qIiLjtpP82ttaut9bWhj9OAA6Ftx3zAuqBz5xuMdbat3Ae/XU28D7wLM7iw5dYaxvDbR7CWQz4XmAxUAxcZK3debrfLyfm9Xq4dWopW3YdYta7G9wuR0RERDj1PXLNQc8D/A2YYIxZcJymk4D/xFmGpM2stbcfZ9sMTrGUibX2tzgjc9KNRpXmUTqgJ0/OtEweVUCi3+d2SSIiInHtVNfH5gF1OIv/eoC54c+tX38HPuy6MiUSeDwebruklN37j/Da22vdLkdERCTuneoeuTuBG3BC3A+Bh4HW19UagL3Ak51dnESeoUXZlJlcnnljFRef25+UJL/bJYmIiMStkwY5a+1KnEumhBcDvs9au7o7CpPIdevUUr7+u7m8MHc1N10cdLscERGRuNWeqYfn4kx4kDhXXNCD84bn88LcKvYdbOsjdkVERKSztSfIHQF2dVUhEl1unhKkpraBf765yu1SRERE4lZ7HtH1HeB+Y0wesBzY2rqBtXZNZxUmka0gL53Jowp59e21XDm+iOweyW6XJCIiEnfaMyL3F5zHZP0RmA+sOs5L4siNFxmamuDJWfbUjUVERKTTtWdE7gtdVoVEpdyeKUwdM4BX317LNROLyc9Jc7skERGRuNLmIGet/XtXFiLR6frzBzNz0Xoem17BN28d5XY5IiIicaU9I3LNT3q4HrgI5zmo9wCjgUXWWl1fi0NZ6UlcMW4Qz7yxiuvOH8zA/Ey3SxIREYkbbb5HzhiTgfOkhyeAKThhLh34HLDIGDOiSyqUiHfNpMGkJvt5ZNpKt0sRERGJK+2Z7PALIAicB/THedoDOE9+WIPzEHuJQ2nJfq6dVMx7K7axYq1WqBEREeku7Qly1wLfs9YuBJqaN1prd+GEvNGdXJtEkcvHDiIrPZF/vLaSpqamUx8gIiIip609QS4d2HSCfQeB1NMvR6JVUmICn7mghI/W7OJ9u8PtckREROJCe4LcMuC2E+y7Evjw9MuRaHbRuQPI7ZnCP6at0KiciIhIN2hPkPspcJ0xZhbwJZzLq+cbY/4I3AH8sgvqkyjiT/By88WG1Zv2sWDZFrfLERERiXltDnLW2peBG4ES4H9wJjv8Emc07gvW2ue7pEKJKhPKCijIS+fRaStpaGh0uxwREZGY1p4ROay1T1tr++PMXh0LDAX6Wmsf7oLaJAr5vB5unRpk0/aDzC4/0S2VIiIi0hnatSBwM2ttJVDZybVIjDh3aB+KC3rw+MwKJpT1xZ/gc7skERGRmNTmIGeM6QX8DrgCZ4aqp1WTJmtth4KhxBaPx8NtU0v54f8tZPrC9Vw+bpDbJYmIiMSk9gSvB4GrgaeBjYBugJITOrMkh+HF2Tz9eiUXnF1IcqIyvoiISGdrz79dLwG+Za29v+T9r6gAACAASURBVKuKkdjh8Xi49ZJSvvnAfF6ev4YbLihxuyQREZGY057JDk2AHqYpbRbs35NzzujNc7NXcbC61u1yREREYk57gtxrOPfHibTZLVNLqa6p59nZVW6XIiIiEnPac2n1GeCh8KSHt4Hq1g2stX/trMIkNgzok8GEkf14af4arhg3iKyMJLdLEhERiRntCXLPht8/E3611gQoyMmn3HixYf7SzTz1eiVfuma42+WIiIjEjPYEuYFdVoXEtPzsNC48pz8z3lnHVROK6N0r1e2SREREYsJJg5wx5h/AT6y1VcC9pzhXE/C5zipMYstnLyzhzfc28MRMy9dvLHO7HBERkZhwqhG5cUBG+O/xOGHtRE62T+Jcr8xkLh07iBfnVnHtpGIKe2ec+iARERE5qZMGOWvtwBZ/D+jyaiSmXTd5MNMXruPR6RV89/az3S5HREQk6rVn+RGR05KRGuDqicUsXLaFyg173C5HREQk6inISbe6cvwgMlIDPDJNa0uLiIicLgU56VYpSX6uP7+EpZU7+LBqh9vliIiIRDUFOel2l4wZQHZmEv94bSVNTZojIyIi0lEKctLtAn4fn70oiF2/h/dWbHO7HBERkailICeuOP+sAvKzU3lk2koaGzUqJyIi0hEKcuKKBJ+Xm6cEWbdlP/OWbna7HBERkaikICeuGTuiLwPzM3h8egX1DY1ulyMiIhJ1FOTENV6vh1unlrJl1yFmvbvB7XJERESijoKcuGpUaR6lA3ry5ExLTV2D2+WIiIhEFQU5cZXH4+G2S0rZvf8Ir7611u1yREREooqCnLhuaFE2ZcFc/vlmJdVH6twuR0REJGooyElEuHVKKQeq63hh7mq3SxEREYkaCnISEYoLenDe8HxemFvFvoM1bpcjIiISFRTkJGLcPCVITV0j//3EEurqtRyJiIjIqSjIScQoyEvn7muHU16xnV8/upgGrS0nIiJyUgpyElEuPncAd141lIXLtvBfTyyhQY/vEhEROaEEtwsQae2KcUXU1jXy91dXEEjw8dUbzsTr9bhdloiISMRRkJOIdN3kwdTWNfDETEvA7+VL1wzH41GYExERaUlBTiLWjRcZausaeHZ2FQG/j89ffobCnIiISAsKchKxPB4Pn7t0CDV1DbwwdzWJfh+3TC11uywREZGIoSAnEc3j8XDnlcOorWvkqdcrCfh93HBBidtliYiIRAQFOYl4Xq+Hu68bQW19A49MW0liwMeV44vcLktERMR1CnISFXxeD1/7zEjq6hr584vLCSR4mTpmoNtliYiIuEpBTqKGz+flGzeHqKtv5H+f/ZCA38f5ZxW6XZaIiIhrXA1yxpjvAJdaa8e22DYAeBAYDxwGHga+Z62tb9HmK8A3gD7AUuAea+173Ve5uMWf4OVbt43i3r8u4oGn3ieQ4GPcyL5ulyUiIuIK157sYIy5G/hZq20BYCbQBIwB7gTuAP6zRZvbgV8B3wdCgAVmGGNyu6VwcV3A7+N7/3I2pQN78ZvHy1m4bIvbJYmIiLii24OcMSbfGPMyThizrXZfB/QHbrPWLrfWvgR8C/g3Y0xyuM13gd9bax+31q7ACXr7gS92zy+QSJAUSOCHd5zD4IIe/OqR91i8cpvbJYmIiHQ7N0bkQsABYDiwqNW+ccBSa+2eFtvmAKlAmTEmDxgMzG7eaa1tAObjXIqVOJKS5OfHd46mf58Mfv7wu3ywaofbJYmIiHSrbg9y1tqXrbU3WWvXHGd3X2BTq20fh98Lwvs5QZuCzqtSokVasp+f3DWGPtmp3PvXRaxYu8vtkkRERLqNa/fInUAKUNNqW/PnpPB+TtAmqQvrkgiWkRrg3i+NITsziR8/9A6VG/ac+iAREZEYEGnLjxwGEltta/58MLyfE7Q52J4vWr58ebuLizXl5eVul9Cpbjgvnb/NOsL3/jCf28/PoXdWoEu+J9b6rbuo3zpG/dYx6reOUb91jJv9FmlBbiMwstW2/PD7JmBDi23LWrVpfbn1pIYOHUpiYus8GD/Ky8sJhUJul9Hpzjijmm8/OJ8n5u/lvi+fR2HvjE49f6z2W1dTv3WM+q1j1G8do37rmK7ut5qampMOPkXapdV5wJnGmB4ttk3CmRyxxFq7A2em68TmncYYH84kibndWKdEqLyeKfzsy+fh9Xj4wZ8W8PHOdg3UioiIRJVIC3IvAJuBp4wxw40xlwM/B/7LWlsbbvNb4GvGmNuNMUOAPwNpwEOuVCwRJz8njZ9+aQz1DU187w8L2La72u2SREREukREBTlr7RFgSvjjIuBP4ddPWrR5CGcx4HuBxUAxcJG1dmf3ViuRrLB3Bvd+cQyHa+r5/h/fZte+w6c+SEREJMq4eo+ctfb242yrAi4+xXG/xRmZEzmhQX0z+cldo/n+HxfwvT8s4OdfOY+sdE1uFhGR2BFRI3Iina2kMIsffeFcdu47zA//tJD9h2pPfZCIiEiUUJCTmHfGoF784F/OYfOOg/zw/xZw8HCd2yWJiIh0CgU5iQsjSnL47u1ns37Lfn780EKqjyjMiYhI9FOQk7gxqjSP/7h1FKs27uXevy7iSG292yWJiIicFgU5iSujh+Xz7zeW8dGaXdz3t3eprWtwuyQREZEOU5CTuDOhrB/33HAm71fu4Jf/WEx9Q6PbJYmIiHSIgpzEpQvO7s+XrhnOuyu28pvHymlQmBMRkSgUac9aFek2l543kLr6Bv7y0kcEErx87bNleL0et8sSERFpMwU5iWtXTSimpq6BR6dVEPD7+Mp1I/B4FOZERCQ6KMhJ3PvMBYaa2gaeeWMVAb+PO68cqjAnIiJRQUFOBLh1aim1dY28OG81gQQvn7t0iMKciIhEPAU5EcDj8XDHFWdQW9/As7OrSPT7uPHioNtliYiInJSCnEiYx+PhS1cPp7augcdnWgJ+H9dOHux2WSIiIiekICfSgtfr4as3jKSurpGHX11BwO/j8nGD3C5LRETkuBTkRFrxeT18/aYy6hoa+b8XlhHwe7n43AFulyUiIvIpWhBY5DgSfF6+eUuIUDCX3//zA2aXb3S7JBERkU9RkBM5AX+Cj+/cfjbDirL53RNLeOuDzW6XJCIicgwFOZGTSPT7+MHnz8H078lvHi1nYcUB6uob3C5LREQEUJATOaWkxAR+fOe5DCvOZsaSfdz18zeY8c566vV8VhERcZmCnEgbpCT5+cldo7l1cja9MpJ48Jml3P2rN5lTvpGGxia3yxMRkTilICfSRh6Ph6LeSfz6nnH84I5zSAr4+O3jS/jqb2bz9ocf06hAJyIi3UzLj4i0k8fj4ewhvRkVzGPBso95bHoFv/j7ewzqm8mtU0sJBXP1eC8REekWCnIiHeT1ehg7oi+jh+Uzd8kmnphZwX/++R2C/bO4ZWopIwbnuF2iiIjEOAU5kdPk83qYPKqA8SP78vq7G3hqluX7f1zA8OJsbp1aSnBAT7dLFBGRGKUgJ9JJEnxepowewORRBUxfuI5n3ljFN/9nPqNK87hlSpCifj3cLlFERGKMgpxIJwv4fVwxvoiLzunPK2+v5dk3V/G1/57LmOF9uOniIP17Z7hdooiIxAgFOZEukpSYwHWTBzN19ABenLeaF+auZuGyLUwY2Y8bLzbkZ6e5XaKIiEQ5BTmRLpaa7Oemi4NcNnYQz81exctvrWXe0s1ccFYhn7mwhNysFLdLFBGRKKUgJ9JNMlID3H7ZGVw5vohn3lzFtAXreHPxRqaM7s/155fQMyPJ7RJFRCTKKMiJdLOsjCTuumoYV08o5qnXLa8tWMfMRRu47LyBXDt5MBmpAbdLFBGRKKEgJ+KSnKxk/vX6M7lmUjFPzrQ8P7eKaQvXceX4Iq6aUERqst/tEkVEJMLpEV0iLsvPTuPfbwrx4P+bRJnJ5clZli/8bBbPvFHJkZp6t8sTEZEIphE5kQhR2DuDb3/uLFZv2suj0yv4x2sreWneGq4735n5GvD73C5RREQijEbkRCJMUb8e/OgL5/Lrr46jf590/vzicu76+etMW7iOuvpGt8sTEZEIoiAnEqGCA3ry0y+dx8++PIbcrBT+958f8OVfvsGbizfQ0NjkdnkiIhIBFOREItzw4hx++a9j+dEXziUtxc9/P/E+//rrN5m/dDONCnQiInFN98iJRAGPx8Oo0jxCwVwWLtvCYzMq+NUjixmYn8EtU0o5a0geHo/H7TJFRKSbKciJRBGPx8OY4fmcM7QP85du5vEZFdz710WUFPbg5imljCzJUaATEYkjCnIiUcjn9TCxrB9jR+Tz5uKNPDnL8qP/W8iQgT25ZUopw4qz3S5RRES6gYKcSBRL8Hm56Jz+TAr1Y+aiDTz9eiXf/cPbDC/O5uYpQYYM7OV2iSIi0oUU5ERigD/Bx6XnDeSCswuZsXAdz7y5im89+BZlJpebpwQpKcxyu0QREekCCnIiMSTR7+OK8UVcdE5/Xluwln++WcU37p/HWUPyuPniIEX9erhdooiIdCIFOZEYlJSYwDWTBjNl9ABeeWstz8+p4mv/PZfRw/pw08VBBvTJcLtEERHpBApyIjEsJcnPDReUcOl5A3lx3mpenLead5ZvYeyIvtx4kaEgL93tEkVE5DQoyInEgdRkPzddHOTycYN4fk4VL89fw9sfbGZCWT8+e5EhPzvN7RJFRKQDFORE4kh6SoDbLhnCleOLeG52Fa+8vZa572/m/FEFfOZCQ17PFLdLFBGRdlCQE4lDmWmJ/MvlZ3DVhCL++eYqpi1cx5uLN3LhOf254fwScrKS3S5RRETaQEFOJI5lZSRx51XDuHpiMU+/UcmsRet5/d0NTBndn+vPL6FnRpLbJYqIyEkoyIkI2T2SufvaEVw3aTBPvV7JawvWMfOd9Vxy3kCunTSYHumJbpcoIiLHoSAnIkfl9kzhqzecyXWTB/PkLMtL81YzfeE6Lhs7iKsnFpORGnC7RBERaUFBTkQ+pU92Kl+/sYzrzx/MEzMtz85exatvr+WK8YO4akIxacl+t0sUERHA63YBIhK5+uWm881bRvE/35hEmcnlqVmVfOGnM3lylqX6SJ3b5YmIxD2NyInIKfXvk8G3P3cWazbv4/EZFTw2vYKX5q3mmkmDuey8gSQl6h8lIiJu0D99RaTNBvXN5PufP4dVG/fw2PQK/v7qCl6YW8V1kwczdcxAEv0+t0sUEYkrCnIi0m6DC7L48Z2jqVi3m8emV/CXlz7i+TlVXH9+CRef2x9/ggKdiEh30D1yItJhwQE9ufdLY7jv7vPok53Gn55fxl33vc60heuoq290uzwRkZgXcSNyxhgDVBxn153W2j8bY84EfgecBewCHrDW/qY7axSRYw0ryubnd5/HB6t28Oj0Cv73nx/wzzdXceOFJUwKFbhdnohIzIq4IAcMB/YDptX2fcaYbOB14Hngy8DZwB+MMfustQ91b5ki0pLH4+HMklxGDM6hvGI7j82o4P6nlvL0G6s4qyjA8BENuuQqItLJIjHIDQNWWmu3tt5hjPk6UAt82VpbD6w0xhQD3wYU5EQigMfjYVRpHqFgLu9+tJXHZ1heWrSH+Stmcfm4QUwdPYC0FC0sLCLSGSLxHrnhwIoT7BsHzA+HuGZzgEHGmL5dXZiItJ3H4+GcoX343b9P4NZJ2Qzok8E/XlvJv9w7k4deWMa23dVulygiEvUidURulTHmbaAYqATutdbOBPry6ZD3cfi9ANjcbVWKSJt4PB6K+iRxw2Uh1n68jxfmrubVt9fyyltrOG9EX66eWMTggiy3yxQRiUqepqYmt2s4yhiTChwA3sW5XHoAuAX4N+Ai4I/A09ba77Y4ZhCwGphkrZ1zqu8oLy8fAKzt7NpFpO32VdezyB6kvOoQNXVNDMhNZExpGsX5SXg9HrfLExGJRANDodC61hsjakTOWnvIGJMB1Fpra8Oby40xQ4D/AA4Dia0Oa/58sD3fNXToUBITW58qfpSXlxMKhdwuI+qo3zrmeP02eRxUH6lj5qL1vDhvDY/P3UVBXhpXTShmUqifJkag/751lPqtY9RvHdPV/VZTU8Py5ctPuD+ighyAtfZ4gexD4HJgDZDfal/z501dWZeIdL6UJD9XTSjmsrGDeGvpZp6fs5r/eXopj0xbyeVjBzF1zADSNTFCROSEIirIGWNGAzOBydba91rsOgv4CHgP+IoxJqHFhIdJQOXxZrmKSHRI8HmZGCpgQlk/Ply1k+fmVvHItJU880YlF57TnyvGDaJ3r1S3yxQRiTgRFeSAxTj3rz1kjPkKzoK/XwbGAOfgTGb4D+CvxphfACHg34G73SlXRDqTx+NhREkOI0pyWLdlP8/PqWLagrW8+tYaxgzP5+qJxZQUamKEiEiziApy1to6Y8wU4BfAs0APoBy40Fr7PoAx5iLgAWAJsBX4trX2YXcqFpGuMqBPBl+/sYzbLinl5flrmL5wHW998DFnDOrFNZOKGRXMw+vVxAgRiW8RFeQArLUfA7edZP9inBE6EYkDvTKTuf2yM7jhghJmLtrAS/NXc+9fFtEv95OJEQG/JkaISHyKuCAnInI8zsSIIi4bO5C3P/iY5+ZU8eAzS3l0+kouGzuQqaMHkpGqiREiEl8U5EQkqiT4vEwo68f4kX35sGonz8+p4tFpFTzzxiouPKuQKycUaWKEiMQNBTkRiUoej4cRg3MYMTiH9Vv288Lc1Ux/Zx2vLVjL6OH5XKOJESISBxTkRCTq9e+Twb99diS3TA3yyltrmbZgLW+HJ0ZcPaGIs4b01sQIEYlJCnIiEjN6ZSbzuUuHcP35g3n93Q28OG81P/3bu/TNSePqiUVMChVoYoSIxBQFORGJOSlJfq4YX8Sl5w3k7Q+bJ0Z8wKPTKrh07EAuGaOJESISGxTkRCRm+Xxexo/sx7gz+7J89S6em1PFY9PDEyPOLuTK8UX0ydbECBGJXgpyIhLzPB4Pw4qzGVaczfqt+3lx7mpmvLOe1xaspXRAT4YVZzO8OJtg/5669CoiUUVBTkTiSv/eGdzzmZHcMrWUaQvWUV6xjWder+SpWZX4E7wE+/dkWFEvhhVnY/pn4U9QsBORyKUgJyJxqWdGEjdPCXLzlCCHDtfx0dpdLKvayYdVO3liluXxmZaA30fpgCyGFTmjeYMLsvAneN0uXUTkKAU5EYl7qcl+zh7Sm7OH9AbgYHUty9d8EuwenV4BQGLAx5DwpdhhxdkM7tcDn0/BTkTcoyAnItJKWkqAc4f24dyhfQDYf6iW5at3OsFu9U7+8dpKAJITfQwZ2IvhxdkMLcqmqG+mgp2IdCsFORGRU8hIDTBmeD5jhucDsPdADcvXOKN1y6p28rdXVgCQkpTAGYM+CXYD8zPxaSFiEelCCnIiIu3UIz2RsSP6MnZEXwD27D/CstWfBLv3VmwDnEu2Q8PBblhxNv17Z+gJEyLSqRTkREROU1ZGEuNH9mP8yH4A7Np3+Oj9dctW72TRR1sBSE8JMLTok2BXmJeOx6NgJyIdpyAnItLJemUmMzFUwMRQAQDb91SzvMWI3cJlWwDITAswtMhZw25YUTb9ctMU7ESkXRTkRES6WG5WCpNHFTJ5VCEAW3cdOibYvf3BxwBkpSceXepkeHE2TU1NbpYtIlFAQU5EpJv17pVK716pXHB2f5qamtiy6xDLqpzlTpat3sG8pZsBSE3yMuSDdygpzKKkIIvBhT1IT9EzYkXkEwpyIiIu8ng85GenkZ+dxsXnOsFu846DLFu9iwVLVrF1VzWLV26jeXAuPzuVkkIn1JUUZjEoP1OPFROJYwpyIiIRxOPx0C83nX656eQGdhEKhag+UseqjXup3LCHVRv3smz1TuYs2QRAgs/DgPxMSgqcYFdSmEXfnDTNjhWJEwpyIiIRLiXJz4jBOYwYnHN02659h6ncsIfKDU7Am12+idcWrAu3T2BwONgNLsiipLAHvTKTXapeRLqSgpyISBTqlZnM6GHJjB7mLFLc2NjEpu0HnGC3cQ+rNuzhudlVNDQ2hdsnHR2xKynsQXG/HqQk+d38CSLSCRTkRERigNfrobB3BoW9M7jgbGd2bG1dA2s27/tk5G7jnqNLn3g80C83HRMOdoMLsxjQJ4MEPWJMJKooyImIxKiA30dwQE+CA3oe3bb/UC2rNn5ySfbdFVt5/b0NTvsEL4P6ZlLS35klW1KYRe9eKVrbTiSCKciJiMSRjNQAoWAeoWAeAE1NTWzfc5jK9Xuo3LiHyg17mL5wPS/NWwM4T6MoKfxkIsXggh5kpiW6+RNEpAUFORGROObxeMjrmUJezxTGjXSeHdvQ0Mj6rQfCl2SdmbJPzbKEb7cjr2cKRf0y6d87w3n1SadPr1R8uiwr0u0U5ERE5Bg+n3OJdVDfTKaMHgDA4Zp6qjbtZVX4fru1H+9j4bItR9e38yd4KchNp3+f9HC4c0Jedo8kXZoV6UIKciIickrJiQnO48OKso9uq6lrYOO2A6zfsp/1Ww+wfut+llXtZHb5pqNtUpISWgQ7J+QV9k7X5VmRTqIgJyIiHZLo91Hcz1nKpKWD1bWs33qADVv3sy4c8t5aupnph+uOtslKT3RCXZ90BoSDXkFeOsmJ+teSSHvofzEiItKp0lICnDGoF2cM6nV0W1NTE7v3H/lUwJu+cD21dQ1H2/XulXJ01G5A+PJsfk4a/gTdfydyPApyIiLS5TweD70yk+mVmUyZyT26vaGxiW27D7F+y7EB772V22gMz65I8Hnom5P2qRG83KwUPYpM4p6CnIiIuMbn9ZCfnUZ+dhqjh/U5ur2uvoFN2w8eM4JXsWEP85ZuPtomKeCj8Oh9dxkMCE+0aGqegSESBxTkREQk4vgTfAzMz2RgfuYx26uP1LFh24GjI3jrt+7nvRXbmPXuhqNtAgke8t58k5ysZHJ6hF9ZyeT0SCEnyxkV1KVaiRUKciIiEjVSkvwE+/ck2L/nMdv3HqhhwzZn5G7ZynV4Amns2HuYNZv2sfdgzTFtPR7okZZ4TLhrDnvZPZxtmWkBLZsiUUFBTkREol6P9ER6pOcwvDiHvil7CYVCR/fV1DWwa+9hduw9zI49ze/V7Nh7mHVb9vPeym3HTLgA53Fl2a1G8rJ7HBv4kgL6V6i4T/8tFBGRmJbo95Gfk0Z+Ttpx9zc1NXGguu5ouGsZ9nbuPcz7ldvZvf8IrW+9y0gNOKEu89jA1zzCl5WepMkY0uUU5EREJK55PB4yUgNkpAYoarUmXrP6hkZ27TtyTNjbGR7l27rrEB9W7eRwTf0xx/i8Hnodc49eMhmpAZICCSQFfCSG35MTE0gM+Jztic57ot+nEChtoiAnIiJyCgk+79Fn0p7IocN17NgbDnitRvdWrNnFrn1HaGhs+4zagN9HcmKLwBdoEfgCPpISmwPhsSGw5b6k8DFHj01MIJDg1f1/MURBTkREpBOkJvtJTfYzoE/Gcfc3NDZxpKaeI7X1HKltCP/dcPRzzae2NxzTvqa2gcM19RysPvzJOcLvje0IiF4PR8NhywBYV1PNnIpy0lMDpCf7SUsJkJ7ivGekBkhL8ZOeEiAlyY9Po4URQ0FORESkG/i8nqNhrzM1NTVR39DI4Ron2NU0B7yaY0Niy/2Hm9sdbVPPvv0NVKzfzYHqOg61eJxaax4PpCb5ncAXDnrpyc7f6S0CX3qK83dGSoC0lACpyQqAXUFBTkREJIp5PB78CT78CT4yUgMdPk95efnR2b4NDY0cPFzHwcN1HKiu5WB1HfsP1XKwupYD1XVH3w9U13LgUC1bdhziQHUth47UfWpSSEupyX4n8LUIeq0DX3qLbekpAdKS/fh8WvfvRBTkRERE5Bg+n5fMtEQy0xLbdVxDYxPVR+o4cKjWCXmtQ184FDa/b9l1iIPVtRw8fPIAmBjwEUjw4k/wEfB/8h5I8OFP8BLwh9+P7m+xze8c63xucZzfe8w5P9l27HH/v727j7aqKvc4/j2cCK4kJlHx4guZ+dxKyUBvlhKI3iS1umYW5WuI14tevd6grEhAAy1Ht3FNu17MYaThUHoxIV8wFSQ0Ukm7YqOHZJgBcn3NVETOa388c8Fmsw/npXP22qv9+4xxxmLNvdbaz55jc86z5lxzzlpPIpXIiYiISK9o7NewrbWtO9qyBLAk4duWBG5u4vWmVppaWmluaWNrcyvNzW2xn7avbW2hubmVppa27duWVpqa27o1wKSSfv0aKid8KdF7zwgombaw6pTIiYiISK769WvgTalrdTiDevXara1t2xPAlh0TwKbmlPC1tO1clrbbztu2bduWVMZE0m29Gm93KZETERGRv1uNjdE9OnBA36Q8q1ev7pPrdlVtd/yKiIiISIeUyImIiIgUlBI5ERERkYJSIiciIiJSUErkRERERApKiZyIiIhIQSmRExERESkoJXIiIiIiBaVETkRERKSglMiJiIiIFFQhl+gys37AbGAqsCewEjjH3Z/INTARERGRKipqi9wsYBpwFnAY0AIsNbOBuUYlIiIiUkWFS+TMbAAwHZjj7re7+/8Bk4FhwEm5BiciIiJSRYVL5ICDgTcBy7ICd38Z+A3w4byCEhEREam2Ij4jNzJtN5SVPw3s3YXzGwGampp6M6ZC2rp1a94hFJLqrWdUbz2jeusZ1VvPqN56pi/rrSRfaaz0ehETud3StrzWtgJdeUZuOMDatWt7M6ZCWrNmTd4hFJLqrWdUbz2jeusZ1VvPqN56pkr1NhxYV15YxERuS9oOAEqb1QYAr3bh/IeAccAmoLV3QxMRERHpVY1EEvdQpReLmMitT9sRgJeUjwAe7+zksWPHbiWmKxEREREpgp1a4jJFHOzwW+BlYEJWYGaDgTHAfTnFJCIiIlJ1De3t7XnH0G1mNg/4N2AK8CTwDWB/4EB31ygGERERqQtF7FqFmBC4EbgGGAT8EpikJE5EIhn2vgAAC11JREFURETqSSFb5ERERESkmM/IiYiIiAhK5EREREQKS4mciIiISEEVdbCD9ICZ7Q5cApwADAV+D1zi7otzDaxAzOwAYl3fC9z92rzjqXVmdhrwZWA/Yh6kOe7+o3yjqm1m9kbgYuBkYE/gUeBCd38g18BqmJl9BTjO3Y8oKRsFXEWswb0FWADMdPeWPGKsRR3U2xhiJohDgGZgOTDD3ddXvEgdqlRvZa/PBOYC/avxfVOLXH1ZABwPTAUOBn4K3GJmE/MMqijMrD+wkBgpLZ0ws1OA64D5wIHAjcBNZnZ4roHVvtnE1EpnEfNj/h6408xG7vKsOmVm5wDzysreCNwFtAMfIuryTCJBFjqst72Be4i1y48AjgPeBtxlZv9Q9SBrUKV6K3v9UGBO1QJCLXJ1w8yGAZ8Ejnf3u1PxpWZ2FPEL7t7cgiuOi4FX8g6iCMysAfg6cKW7X5GK55nZOGAicH9uwdW+TwA/dPelAGY2nbj5OhxYlGdgtcTMRhA3CUey4yo/AJ8C9gU+4O5/BtaY2YXAFWY21923UKc6qbeTidbLqVlLUmpV/xOR2P2iiqHWlE7qLTtmEHGzv4L4PVcVapGrH5uBjxJfsFLtwJDqh1MsZvZh4Gzg9LxjKQgDRhGtcNu4+yR3/3ouERXHs8DxZjbKzBqJJG4r8Ei+YdWcscSN1Wjg12WvjQMeTUlcZjnRmj6mKtHVrl3V20+Az5R1B2ZzlNX734ld1VvmCuAx4IZqBQVqkasb7v4KcGdpmZkdRtw1nJ9LUAVhZm8m/mOe5+7rzSzvkIrggLQdYGa3AYcSq7DMdfcl+YVVCOcRLW9PAq1AG3CSu/8h16hqTPoeLQGo8H9yJLChrOzptN27byOrbbuqt/QdK/+efZVopavrXptOvm+Y2QnAscBBwMeqGZta5OqUmb0buIW4s5ifczi17mrgV+5+Y6dHSmZw2t4A/Bj4CLAUuNXM/jm3qIrhQGI96X8BDiOeM7whPYQuXbMb0YpZKtsfWOVYCsvMvgBMIwbbPJd3PLUqdbteA0xx9xeq/f5qkatDqZvwFuApYuRNc84h1SwzO5Xopjko71gKJlsu79vu/v3070fN7BBgBnX8rM2umNk+RPJ7jLsvS8UPm9l7iRHnx+cWXLFsAQaUlWX7r1Y5lsIxs37AZcCXiJkNrsw5pJqVngf+AbDI3e/s7Pi+oBa5OmNmJxN/RFcD4939xZxDqnVTgLcD683sVTPL/ghcZWaP5xhXrcu6tR4rK18DvKPKsRTJPwH9gYfKylcB76p+OIW1HhhRVpbtl3e5SgkzG0h07c8AznX32TmHVOv2AY4GPl/yN+J/02svmdlX+zoAJXJ1xMw+R9ztLyJa4jQCs3OnAO8mpmvJfiBGsB6bV1AF8AjxYPChZeUHAU9UP5zCyJKM0WXlo4G1VY6lyFYAB6fnWzNHEt/J3+QTUu1LLXE/IVp+T3T3/8k5pCLYSNxkjWb734hZ6bWxbE/q+oy6VuuEme0FfA9YRjSXv6Xkgc0mtcxV5u4by8tSvT3n7k9VP6JicPctZnY5cJGZPU20KE0GjgH0jFzHHgRWAteZ2TQisTuNuOMfl2dgBfMzYkLWm83si8RUJJcRXf1Nuzyzvl1A3KBOA1alaasyf6nnaVs6kkb47nBzambPpn+u04TA0ps+STwAPJEYvbWp5EcrO0ivc/e5xJ3pJcDvgE8Td/n35BpYDXP3NuDjxKSsC4jWo6OAo7WyQ9e5++vApLSbDeiaT3wXpWOnpO3V7Pg3YhMxx5zUoIb29vbOjxIRERGRmqMWOREREZGCUiInIiIiUlBK5EREREQKSomciIiISEEpkRMREREpKCVyIiIiIgWlCYFF5O+amf0RWOnup3RyaFWZ2RuIWd8/RfwuPsfdr883KhEpGiVyIiL5+ChwJvAt4HZAa/eKSLcpkRMRyceQtL3G3f+QayQiUlha2UFE+kzq1lwA9AfOIJKXh4EL3H11OmYOMBvoX7ouoZltAO529zPSfjtwLnAIcCLQBiwEZgBfA6YCA4G7gbPd/YWSGFYRS9OdQTwbvASY4e7PlLzfe4BvAuPTMcuB6e7u6fVRwJPAfwBnAe8Evuzu36nwuRuBfyXWrNwfeAG4CZiV1qFdAJxecspT7j6qgzr8R+DbwBHAa8A1wF7Afu4+oaRuLgKOIxbtvsrdv2hm7wIuTefuATwEXOTuK8o+01nufm3Je84FZrp7Q9pfkD7HtcDFwFBi6avp7v5IpbhFpDo02EFE+tr5wGHAOcDngVHAT1Oy012Xpe1JwA1EYrcaeB+RyM0DTkjbUicSycxU4EJiYfBbsxfNbH/gAWAfIgGbCowE7jezvcuudTnxbNtpRJdoJfOBK4E70ntfDfw7sMTMGlJ8l6ZjJxPPye3EzIYCK4B3pLi+AHw2/ZSbBSwl1rS9KSWmq4EDgOnAqem4e8xsYgdx78qBxGefSyShbwWWm9lePbiWiPQSda2KSF97DZjk7s0AZjaIaNk5mEg0umOdu09J17kXmALsDpzo7k3Az81sAnB42Xl/IRaefzmd+wxwi5kd6+63A3OAFuBId38xHXMH8ATR2nd2ybVudffvdhRgSqDOBGa7e7ZI+x1mtpFonfyYuy82s6w7dbW7P9HB5c4HBgNj3H1Duv4DgFc49hF3n1MSx83pM41395dS2WLiWbxvAWM6+gwd2IOon+UlcawD/pNIFEUkB2qRE5G+9mCWxCXr03ZQD661MvtH6oZ9nkiEmkqOeR7Ys+y827IkLllMJDlHp/2jiK7Ul83sDWlE6Waim/YjZdd6rJMYx6ftjWXlC4FWYEIn55eaCKzKkjgAd/8j0XpYrjyuCcDtWRKXzm1Ocb3fzPboRhwA67MkLl3r6RTH+A7PEJE+p0RORPraa2X7bWnbk98/r1Qo21y2X+nB3/8v3XH3NnZM+IYSXaDNZT+TiS7WDq9VQTaIofw9s8TzzZ2cX+qtwHMVyp+pUFYe15AKZaXHdTeR21ih7Fl2TppFpIrUtSoiecsSr0ailSyz+y6O7a4hpTupxW0okYgAvAQsI54B+1u9mLbDiK7Z7D37p/d8vhvX2gC8vUL527oYx7AK5SPS9oUUD0Tdlxpc4byhFcqGsb0ORSQHapETkbxlXZ77ZAVmNobKyURPHWNmA0r2s0l4l6X95cB7gd+6+8PZD/GM2ue6+V73pW35eZ8lEqaVdN19wAdKBxSY2XDgg10891gz29YCmBLYycTzdJupXPcNVO4ufaeZjS45bmSK456ufxwR6W1qkRORvC0B/gu41sy+QXQnziZajHrLW4BbzewKYhqNecC9xChPiCk1VgF3mdl3ie7aKUTCd+rOl+uYu//OzH4AzDKz3Ygk8X3EqNIVwG3duNx3iNGuS80sGzjxNWI6l7YOzwoXE6Nz70v1+jpwHrAfMU0J7v5nM1sJnGtm64BNxIjd4RWu1w4sNrOZRMvpLKIl87+78XlEpJepRU5EcpVGbJ5MdBf+jJge5EvAo734NvOJ+dJuJkaoLgQ+4e7tKYY1xPQkm4HrgB8B+wKfdvcf9uD9ziSS0ZOIgRXTiOlIJrl7a1cvkgYqHEkkWN8nErtFwIPAq52c+zjxmTYSc89dTyRjE919acmhpwP3p/iuJ7pzZ1a45CYiObwc+B6wFvigu3enq1hEepkmBBYRqVFm9iFgiLv/vKSsP/AnYKG7z6hSHAuI6Vs0Z5xIjVHXqohI7RoBLDKzS4ku2kHEqhKDiVY2Ealz6loVEalR7v5jYkWMjxMrUSwgfm+Pc/e1OYYmIjVCXasiIiIiBaUWOREREZGCUiInIiIiUlBK5EREREQKSomciIiISEEpkRMREREpKCVyIiIiIgX1Vze5dqt8fn5qAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,7))\n",
    "sns.set(font_scale=1.4, style=\"whitegrid\")\n",
    "sns.lineplot(data = groups, x = 'number of group', y = 'inertia').set(title = \"Elbow Method\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on the Elbow plot I will prepare model for 4 clusters."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The function of visualize grouping:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plots_model():\n",
    "    fig = plt.figure(figsize=(10, 8))\n",
    "    ax = fig.add_subplot(111, projection='3d')\n",
    "    # Group data by 'groups'\n",
    "    grouped_data = rfm_df.groupby('groups')\n",
    "    \n",
    "    # Plot each group separately\n",
    "    for i, (group_name, group_data) in enumerate(grouped_data):\n",
    "        xs = group_data['Recency']\n",
    "        ys = group_data['MonetaryValue']\n",
    "        zs = group_data['Frequency']\n",
    "        ax.scatter(xs, ys, zs, s=50, alpha=0.6, edgecolors='w', label=group_name)\n",
    "\n",
    "    ax.set_xlabel('Recency')\n",
    "    ax.set_zlabel('Frequency')\n",
    "    ax.set_ylabel('MonetaryValue')\n",
    "    plt.title('Visualization of created groups')\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  after removing the cwd from sys.path.\n"
     ]
    }
   ],
   "source": [
    "model = KMeans(n_clusters=4, init='k-means++', max_iter=300)\n",
    "groups = model.fit_predict(df_norm)\n",
    "df_norm['groups'] = groups\n",
    "rfm_df['groups'] = groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjwAAAHBCAYAAABt62HAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOy9eXxU1f3//5olk0wyWYcskARIgCCCSBDZFIEglKL0Y11RilVQQQG1D1zrhopWqYiyCFhEWwt8WhEsWkXFfuyjKoui/SpIQUHWzJZkZpLZl3t/f/A7t3cmdyaz3Dtzb3Kej0ceyiznnLn3zj2vea8qlmVZUCgUCoVCoXRj1NleAIVCoVAoFIrUUMFDoVAoFAql20MFD4VCoVAolG4PFTwUCoVCoVC6PVTwUCgUCoVC6fZQwUOhUCgUCqXbQwUPhUKhUCiUbg8VPBQKhUKhULo9VPBQKBQKhULp9lDBQ6FQKBQKpdtDBQ+FQqFQKJRuDxU8FAqFQqFQuj1U8FAoFAqFQun2UMFDoVAoFAql26PN9gIoFAqFQqEkjs/ng81mg8/nQygUyvZyMoJWq0VeXh7Ky8uRl5eX0hgqlmVZkddFoVAoFApFApxOJywWC8rLy2EwGKDVaqFSqbK9LElhWRahUAgulws2mw2VlZUoLi5Oehxq4aFQKBQKRSG0tLSgpqYG+fn52V5KxlCpVMjJyUFpaSlyc3NhNptTEjw0hodCoVAoFIUQCASg1+uzvYysodfr4ff7U3ovFTwUCoVCoSiI7u7Cikc6n50KHgqFQqFQKN0eKngoFAqFQqFkHIZhsGrVKkyYMAEXXngh5s6di5MnT0o2HxU8FAqFQqFQMs7atWuxdetWLFu2DH/5y1+g0Wgwb968lGN0uoIKHgpF4dDKEuJBjyWFkhkCgQA2bdqERYsWYeLEiTjvvPOwcuVKtLS04IMPPpBkTip4KIrh1ltvxahRo+Kq/1tuuQVNTU1gWRYPPfQQLrvssgyu8L9Ezz1nzhzceOONos/z1ltv4dlnn+X+vXr1agwePFhWxcg++OADTJkyBcOGDcO8efOyvZyYRB/LdJDjeaBQ4hFmMiv2Dx8+DI/Hg7Fjx3KPGQwGnH/++fjqq68kmZPW4aEohuuuuw5ffPEF/vGPf+DnP/95p+dNJhP27duHxYsXQ6VSYcGCBbjpppuysNLOPProo5KMu27dOowcOZL79zXXXINx48ZBq5XPV3vp0qWorq7GU089hbKysmwvJybRx5JC6Qm0Or348bQDVrsHFaX5GFhbAmOx9GnvFosFAFBZWRnxeEVFBUwmkyRzyueuSKF0weWXX46SkhLs3LlTUPC88847UKlUuOaaawAA/fv3z/AKYzN48OCMzNOnTx/06dMnI3MlisPhwKxZs3DJJZdkeykUCoVHq9OLXXtOwN5xzmp+0tyBH884MH1cf8lFj9frBQDodLqIx3U6HQKBgCRzUpcWRTHodDrMnDkT//rXv+BwODo9/8477+Cyyy7jfjFEu5UOHz6MuXPn4uKLL8aIESNw00034fPPP+eej+WGuOyyy/DQQw9x/25ra8OTTz6JyZMnY9iwYRg9ejQWLlyI06dPx1w736W1fft2DB48WPBvzpw53Ht2796N2bNno7GxEcOGDcP06dPx5ptvcs8PHjwYZ8+exbvvvssJKqHP8P777+Oaa65BY2Mjxo8fj8ceewx2uz3ic0+dOhWfffYZfvnLX+KCCy5AU1MT3njjjZifh3Dw4EHcdtttGDNmDEaOHIk77rgDR44cAQDs27ePW9f69esxePBg7Nu3T3AclmXx5ptv4oorrsDw4cMxZcoUrF27FuFwOGKNr776KsaMGYPx48fDbDYDAN5++23MnDkTw4YNw2WXXYYVK1Z0umGmciwB4Mcff8SCBQswcuRINDY2Yv78+Th+/HjE2B0dHXjssccwduxYjBw5EkuXLk3ohu1yufD4449j3LhxaGxsxJIlS/CnP/0pYv45c+bggQcewJIlS9DY2IgbbriBm/O5557D5ZdfjgsuuABXXHEF/vrXv0aM39TUhPvuuy/isS+++CLiPJBr8euvv8Yvf/lLDB8+HDNnzsT7778f8b49e/Zg1qxZaGxsxEUXXYR58+bh4MGDXX5Girz58bSDEzsEe4cfP57ufH8VG9IPK/q7EggEJKsiTQUPRVFcd911CAaDnYLavv76a5w4cQLXXXed4PtcLhduvfVW5Obm4ve//z1eeukl5OTkYP78+XGFSjQsy2L+/Pn45z//iXvuuQd/+MMfcOedd+KLL77AY489ltAYEydOxObNmyP+rr/+egDgrFOffvopFi5ciEGDBmH16tV46aWXUF1djWXLlnH+7c2bN6O8vByXXnopNm/eLDjXK6+8gt/85jc4//zzsXLlStx555348MMPcfPNN3O/sIBz5eofeeQRXHvttVi3bh2GDh2K3/3udxGCMJq9e/di1qxZ8Pv9eOqpp/DUU0/BZDJh1qxZ+PHHH3H++edz67rqqquwefNmnH/++YJjrVy5Es888wzGjRuH1atX44YbbsC6deuwcuVK7jUmkwnvvPMOli9fjvvuuw9VVVXYuHEjfvvb36KxsRFr167FnDlz8Kc//Qn3338/975Uj+XJkycxa9YsNDc34+mnn8ayZctgsVhw4403ciZ3lmVx++2348MPP8TixYuxfPlynD59Gq+//nqcK+AcCxcuxN///nfceeedeOGFF+BwOPDCCy90et37778Pv9+P1atXY/78+fD5fLjpppuwY8cO/PrXv8aaNWvQ2NiIxx57DGvWrOlyXiHuuusuTJo0CS+//DL69euH3/zmN9i9ezcA4NSpU1iwYAHq6+uxevVq/O53v0Nrayvmzp0Ll8uV0nyU7BNmWFjtHsHnrHYPGIljenr37n1uLqs1cm6rtZObSyyoS4uiKAYPHowLLrgAO3fujAgCfuedd1BRUYFJkyYJvu/YsWOw2+24/fbbuTiN4cOHY8OGDfD5fAnPb7VaodPpuM0ZAMaNG4czZ85g69atCY1hNBphNBq5f+/fvx87duzA3LlzcdVVVwEAjh49iiuuuAJLly7lXnfRRRdh7Nix2L9/P0aNGoVRo0ZBp9OhtLQUo0aN6jSP0+nEunXrcPXVV+Ppp5/mHieWpLfeegs333wzAMDj8eCll17CxIkTubn+7//+D//4xz9iuqJWrFiB6upqvP7661zM0IQJEzBt2jS89NJLWLNmDbeuqqoqwTUC56wVmzZtwqxZs7hYp4kTJ8LtdmPv3r1gGAYAEAwG8eCDD3JrdLlcWLNmDa6++mo89dRT3Puqqqpw33334ZtvvkFjY2PKx3L16tXQaDT405/+hJKSEgDnrH1Tp07FunXr8NRTT+Gzzz7DN998gzVr1mDq1KkAgMmTJ+OKK67ATz/9JPh5gXMWk7179+Kll17i3LOTJk3CzJkzcezYsYjXsiyL5557DgaDAQCwZcsWHD16FG+++SZGjx7Nfe5wOIwNGzbgpptuSjpWavbs2Vi8eDG3jquuugqvvPIKLr/8chw8eBA+nw/33HMPtxENGDAAb7/9NlwuF7cuirLQqFWoKM3HSXNHp+cqSvOhVktbzfm8886DwWDA/v37UV9fD+Dcd/r777+XLPaSCh6K4rj22muxdOlSnD59GrW1tfD7/fjggw9w0003QaPRCL5n0KBBKC8vx5133okZM2bg0ksvxZgxY/Dwww8nNXdlZSVnATCbzThx4gSOHTuGb775BuFwGOFwOOYahDhx4gQWL16McePGRVgl7rjjDgCAz+fDiRMncOLECXz33XcAOpuAY/Hvf/8bgUAAM2fOjHh89OjRqK6uxv79+znBA5wTAQS9Xo+SkhJ4PMK/AD0eD7777jvMnz8/IkC6uLgYkydPxj/+8Y+E1kjWGQwG8bOf/Szi8d/85jedXst393zzzTfwer24/PLLI1x4kydPhlqtxueff47GxsaUj+XevXsxZswYGAwGbny9Xo/x48fjs88+A3BOrGo0GkyePJl7n0ajwc9//nO88sorccfWaDScSCLvmzFjBlavXh3x2j59+kSIiv3796OyspITO4SrrroK27dvxzfffIMpU6bEnFsIIrSBc6X7p02bhlWrVsHtdmPEiBHIz8/H9ddfz313Ro0ahQceeCCpOSjyY2BtCX48E+nWKi3MxcDaEsnn1ul0+NWvfoWVK1eiV69eqKmpwYoVK1BZWYlp06ZJMicVPBTFceWVV+K5557Du+++i7vuugu7d+9GR0cHrr322pjvyc/Px9atW7FhwwZ8+OGH2LJlC3Q6HaZNm4alS5eisLAw4fnfe+89vPjiizh79ixKSkpw/vnnc/7oZOq4OJ1OzJ8/H2VlZXjxxRehVv/Xw2y32/Hkk0/i448/Bsuy6NevH2d5SHQOp9MJAOjVq1en53r16oX29vaIx8hnIKjV6phzdXR0gGVZlJeXC47d0dH5V2MsSDyR0DqFxo5+31133SX4WpIFkuqxtNvt+PDDDzF06NBOz+Xk5AA4F5BdVFTUKSuuoqIi7uew2+2C7+Nb/gjRx8XpdMY8pwCSOvaEaBcCWUd7ezv69OmDLVu24A9/+AO2bduGTZs2IT8/H1dffTUefPDBTkGnFOVgLNZj+rj+WcnSAoC7774b4XAYjz/+OLxeLy666CJs3LhRsmuKCh6K4jAYDJg+fToneHbs2IFx48ahtrY27vtqa2uxbNkysCyL//znP3j//ffx2muvobi4GI8//jjXlI64UAhut5v7/6+++gr3338/5syZg3nz5nEbxfLly3HgwIGEP0MwGMTixYvR1taGt956q5PgWrJkCY4dO4bXX38dI0aMgE6ng9fr7RSYGo/i4mIA5+JzGhoaIp6zWq248MILEx4rmsLCQqhUKthstk7PWa1WzgWUCEVFRQDOBYNHj3Ps2DE0NjbGfd/zzz+PAQMGdHq+tLQUQOrHsrCwEGPGjMFtt90W8zVlZWVwOp0IBoOcCAIQERQuRGVlpeD7Wltb474POHdeo91ewH9jIcjnBjpfy7Fibux2e4ToaWlpgVqt5s7jkCFD8OKLLyIcDuPbb7/Fjh078Oc//xl9+vSRdW0lStcYi/UwFuvBMKzkbqxoNBoN7rvvvk7B9VJBg5YpiuTaa6/F8ePH8eWXX2LPnj0xg5UJu3btwtixY2G1WqFSqTBkyBAsWbIE9fX1OHv2LABwboPm5mbufYcOHYrYJL755hswDIM777yT2yBCoRAX3Bu9wcRi6dKl+Oqrr7By5UrB9PkDBw7g8ssvx+jRo7lfO//85z8BRFol+FahaC688ELodDq8++67EY9/+eWXMJlMES6sZMnPz8ewYcOwa9euCHdSe3s7Pv3006TGHj58OHJycvDxxx9HPL5582bceeedMa0w5POZzWZccMEF3J/BYMDzzz/PiYJUj+Xo0aPx448/4rzzzosY/80338R7770HABg/fjwYhukURP/JJ5/E/cyjR48GwzBcYDBZS/QxEOLiiy+GxWLBl19+GfH4zp07odVqOSFrMBgirmXgnDtMiI8++ihiHR9++CFGjBgBvV6PN954A5MnT4bf74dGo0FjYyOefPJJ5Ofnc98divLJtNjJBtTCQ1Eko0aNQl1dHR577DEUFhbi8ssvj/v6kSNHgmVZLFiwALfddhuKi4vxr3/9Cz/88AP3C37y5Ml47rnn8Oijj+L222+H3W7HmjVrIqwVw4cPBwA8/fTTuOaaa+B0OrFlyxYuFdvr9XZpjt20aRO2bduG2bNno6ioqFNV0VGjRmH48OF47733MHToUFRVVeHrr7/Gxo0boVKp4PF4wDAMVCoVioqKcOTIEezZswdjxoyJGKekpAR33HEH1qxZg5ycHEyZMgVnzpzBqlWrUFdXx2WEpcqSJUswb9483HrrrZgzZw6CwSBeffVV+P1+LFq0KOFxysrKcPPNN+P1119HTk4Oxo8fj8OHD+O1117D/PnzodcLm9dLS0tx++23Y82aNWhvb8e4cePQ2tqKNWvWwOfzYdiwYQDQ5bEkRB/LRYsW4YYbbsDcuXMxe/Zs6PV6vP322/jwww+xfPlyAMCYMWMwadIkPPHEE2htbUVdXR22b98uaIHhc/HFF+OSSy7B448/jtbWVlRXV2Pbtm344YcfOEtjLK6++mps2bIFixYtwuLFi1FbW4tPPvkE27dvx4IFC7jrtampCevWrcOqVaswatQo7N27F7t27RIc88UXX0QgEEB9fT3++te/4vjx49i0aRMAYOzYsVixYgUWLFiA2bNnQ6fT4b333oPP58P06dPjrpVCkRUshaJQXn31VbahoYF99tlnBZ9/8MEH2QkTJnD/PnjwIHvbbbexY8aMYYcNG8bOnDmT3bZtW8R73n33XXb69Ons0KFD2RkzZrC7du1if/3rX7MPPvgg95o///nP7JQpU9hhw4axkyZNYh944AH2448/ZhsaGtjdu3cLzv2rX/2KnTVrFvf/DQ0NMf9YlmXPnDnDzp8/n73ooovYiy66iL3mmmvYHTt2sLfddhv7i1/8grXb7azD4WD/9re/sWPGjGEvvPBC9tSpU+zLL7/MNjQ0sMFgkJt7y5Yt7IwZM9ihQ4eyl1xyCfv444+zdrude37VqlWd3sOyLDthwoSIzy3E3r172dmzZ7PDhw9nL7roInb+/Pnsf/7zn4jXNDQ0sC+++GLccRiGYTdt2sROnTqVHTp0KDtt2jT29ddfZxmGibtGlmXZrVu3sldeeSU7dOhQdty4cey9997Lnjx5kns+1rG8/fbb2auuuop73a5du9gxY8awI0aMYE+fPs2yLMt+//337B133ME2NjayI0aMYK+55hr2/fffj5jf5/Oxzz77LDt+/Hj2wgsvZO+55x52/fr1MddLcDgc7EMPPcSOGjWKHTFiBHvfffexTz75JNvY2Mi9hn/d8GltbWUfeeQRdty4cdy1/Je//CXiNR6Ph33iiSfY0aNHsyNGjGAXLFjAHjhwgG1oaGD37t3LsizLvv3222xDQwO7c+dO9sorr2QvuOAC9vrrr2e/+OKLiLG++OILdvbs2eyoUaPY4cOHs9dddx37ySefxPxsFOn4/vvvs72ErJPqMVCxLO2WR6HIHZZlEQwGuUJ8gUAAKpUKLMty1h7gXIaNRqOBVquFRqOBRqOBSqXq0mpAySxnzpzBgQMHMG3atAgL1t13340zZ85g+/btGVnH9u3b8fDDD+Ojjz5Cv379MjInJT0OHz6MIUOGZHsZWSXVY0BdWhSKjGFZFqFQiIuT4Qsb8sePPWFZFuFwmHs9y7JQq9URIkitVkOtVlMRlEU0Gg0ee+wxfPLJJ7jmmmug0Wjwr3/9Cx999BHnLqNQKOJCBQ+FIkP4woVl2QgrTTyjrJA1J1oEhcNhuFwulJWVIScnh4qgLNC7d2+8+uqrWLt2LZYsWYJQKISBAwfixRdfxIwZM7K9PAqlW0JdWhSKzGAYBsFgkHNVCQkY4tJKhWAwiNOnT6N///6dxBOxBGm1WiqCKBQZQl1a1KVFoSgehmEQCoUQDocjXFZSQFxd0Y8xDAO/3w+fz8fNTWKB+C4xKoIoFIrSoIKHQskyQnE6UgqKWGPHsiYxDINwOBzRhkGtVnOWICqCKBSKEqCCh0LJEvHidORCLBFE3GqxRBAJlJbb56FQKD0XKngolAxDBAM/TidexWSxIens6bw/WRFE3GJUBFEoFCE2bNiATz/9FFu3bpVsDip4KJQMEh2QnEmhIyWxRBAQ2ZGcZVnBmKDuchwoFErybN68GStXrozZN08sqOChUDJApuN05AA/6JlARFAwGIzINFOpVJwAIi6xnnCMKBS5wDJhqNSarl8oIhaLBU888QT27duHuro6yeejgodCkRASpxMMBgHIQ+ik69JKd25AWASFQqGYIohWjaZQpCHU3gq/+ThCThu0xeXIraqHtsiYkbkPHTqEgoIC7Ny5E2vXrsXJkyclnY8KHgpFAkh2U2trKwwGA81iigNf4BDXlpAIcjqdyM/Ph16vp60zKBQRCLW3ov2bjxF2OwAAAdsp+E3HUNQ4NSOip6mpCU1NTZLPQ6CCh0IRGX6czpkzZzBkyBC6ISeJkAjyeDzQ6XS0dQaFIhJ+83FO7BDCbgf85uMZs/JkEip4KBSRiNX3Sm5k06WVLkKB3tGtMwgajYa2zqBQYsAyYYScNsHnQk4bWJaBStW9kgmo4KFQ0iRenI6SxYVSiFcw0efzRTxOW2dQKOdQqTXQFpcjYDvV6TltcXm3EzsAFTwUSsqQTTUYDMYsHCiV4KGbdHziiSDaOoNCOUduVT38pmMRbi1NQQlyq+qzuCrpoIKHQkmBROvpyNHCw++6rqSNPd3jSFtnUCiRaIuMKGqcmrUsrUxDBQ+FkgSp1NMRW/AwDAO3242cnBzk5OT0qI1Y7M9KW2dQejraIiO0RcZuGbMTDRU8FEoCpFo4UEwLD8uycLlcMJlMUKlUCIfDYBgGubm5yMvLQ6HBgNwwAzVYqDQaID8fTJy5lWbhyRS0dQalJ5JtsfPcc89JPgcVPBRKHNJt8CnW5ufz+WA2mxEIBFBVVYXc3FxO9Ph8PqgZBoEfj+Hs11/Db3cgr7QUvRpHQD9wAFittpMliG7KyUFbZ1AoyocKHgolBmL0vUrXwhMOh2G1WuFwOFBeXo6ysjKoVCpuk9VoNDAYDAhbrGg78DVywwx0hkKEfT7YvtiD3kVFaNWo4ff7kZuby1mDiMWCkjqJts5wuVzQ6XQoKCigrTMolCxCBQ+FEgXpZB4OhwGk1w4iVcHDsizsdjssFguKioowaNAgaLVa7jnyXyLEnD/8ADbMcHNqtTnQAgiePIX+kyciEAzC5/PB7/fD5XIBAI4fP468vDxOBOXm5kKn09FNOA2ERJDb7QYA5Obm0tYZFEoWoYKHQvn/kaLBZyqCx+VywWw2Q61Wo3///tDr9Z3G7DQuwwiOxbLnHtdoNCgoKEBBQQEA4Mcff0Tfvn0RDAY5EdTa2opQKERFkAQQt1a81hnAuXPLd4dREUShiAcVPJQeT7pxOmIRCARgNpvh9XpRVVWFoqKihNbBMAwKBg5E+w/HAL4IUqlgGDgwZuCyRqPhXC0EEhNERZC0CLXOADpXjaatMygU8aCCh9KjIXE6J06cQFlZGYqKikQdPxELTzgcRktLC9ra2mA0GlFTU5NUvBDLssipKIfx4lFwHDyIsMcLjT4PxUPPR27vKoRjzC+0rmhLEFkfKdbHF0FEAFERFJ9ksuFiBUfT1hkUSvpQwUPpkfDjdKS06MQTPCzLwuFwwGKxoKCgAAMHDkROTk5K8zAaDQouGIqC+v5gAkGodDlQFRQgHGddiaLRaJCfn4/8/HzuMSqCMgdtnUGhiAMVPJQeRaw4HSlbQAiN6/F4YDKZAAB9+/aNEBOpwgBAQQFQALA49ycVXYkgt9vdSQSR//Y0ESTVdUVbZ1AoyUEFD6VH0FWcTqYETzAYhMVigcvlQmVlJUpKSrK2+Yj9ebsSQR6PB21tbT1SBGXisyXTOoOKIEpPhAoeSrcnkXo6Uve8YhgGLS0taG1tRWlpKQYNGhSRupxpMrW5UREkjYUnUVLtH0ZbZ1AygcvlwqpVq7B7927Y7XbU1dVh4cKFmDJliiTzUcFD6bYwDINQKBQRpxPrBi6l4HG73TCZTNDr9RgwYAB0Op0k8yiFZEUQEUAMwyi2WKKchEMyrTO8Xi8MBgN0Oh1tnUERnYcffhhHjhzBsmXLUF1djQ8++ACLFi3Cpk2bMG7cONHno4KH0u1IpZ6OFILH6/XC4/HA5/OhuroaBoNB1PHTQW5d3BMRQYFAAM3NzREiqDtZgrJJLBFks9mg1Wq5Ipy0dQZFLGw2Gz766CNs2LAB48ePBwAsWLAAe/bswbZt26jgoVDikU49HTEFQCgUgtVqhdPphE6nQ2lpqehiR26CRQqiRdDp06dRWloKtfpcq4xYliC5iSClNmkl1xgROIBw6wzyWiKAaOsMZcIwTEaFq16vxx/+8AeMHDky4nGVSgWn0ynJnFTwUBQPMcen0/dKDAHBsixaW1ths9lQUlKCQYMGwWw2pzUmJRK1Wh3TEiRXEaRkYRot1uL1DxOqGk1bZ8ifVo8Dx+0n0eJuQ6+CMtSX9oMxv0TyeQ0GAy677LKIx/79739j7969ePTRRyWZkwoeiqIRo8EnkL7g6ejogNlsRk5ODurq6pCXlyfKuFIh13WlgpA7jNSoISLIbrcjGAxGiCDy/5nKoFIiiVinhKpG09YZyqDV48Anx/4Fh68dAHDa2Yyf2k5hyoAJGRE9fI4dO4ZFixbhwgsvxA033CDJHFTwUBSJ2H2vUhUAfr8fZrMZfr8fVVVVKCwsFPxFTMksQpYguYkgJZCqOy7R1hnkNbR1RnY4bj/JiR2Cw9eO4/aTGRU8X375JRYtWoQ+ffpgw4YNKRdg7QoqeCiKgtwwg8EgAHEafJJxmBgNOIUIh8OwWq1wOBwoLy9HbW1tVtLdKYmTjAjS6XQRFaPTEUFKjeEBxF87bZ0hHxiGQYu7TfC5FncbGJaBWiV9TM/OnTvx29/+FqNHj8aqVaskTe6ggoeiCEjtkJaWFhgMBtHTYxMVJizLwm63w2q1orCwEIMGDYJWG/trJFfBI9d1xUOK9cYSQfzsMClEkBLI1PVBW2dkB7VajV4FZTjtbO70XK+CsoyInXfffRcPPPAAZs6ciWeffVYyyw6BCh6K7OHH6ZjNZgwYMED0m1wiAoDU01Gr1ejXrx/0en1CYytNWMiZTGxuarUaer0+4vx2JYL4LjGhzVvJm3I21k5bZ2SG+tJ++KntVIRbqySvCPWl/SSf22w247HHHsOYMWNw//33w+FwcM/l5OSgpER8lxoVPBTZEitOJ5M9rwAgEAjAbDbD6/WiqqoKRUVFWUl3Fxu5rkuOdCWCfD4fHA6HoAhS6nGWm1CjrTPEx5hfgikDJmQlS+ujjz6C1+vF3r17MWHChIjnRo4cia1bt4o+JxU8FNkRL04n2VibRBESJgzDwGazoa2tDUajETU1NdTyy2EAACAASURBVFlJd5cCevNPn0RFUCgUQnNzM/R6fVxLkNyQm+ARIp4IcjqdcLvdqKysBEBbZ8TCmF8CY35JxmJ2CDfffDNuvvnmjM0HUMFDkRHkRhUMBmMWDlSr1ZJbeFiWhdPphNlsRkFBQVrtIKQSaBR5IiSCjh07BqPRiFAoFNcSpNPpZFWxWAmCR4joH0gajSZm6wwigqJrBfVEMil2sgUVPBRZkGg9HaldWh6PByaTCSzLora2FgUFBaKMKzZqtTotISVXy1N3JS8vLyIgM1F3WDZFkFIFD4FlWe7YxbIEAYgQQLR1RveGCh5KVkm2no5UGzXDMPB4PDh16hQqKytRUlIiWrq7FOulYkVZCFkqk4kJiq4YnYkNuDsInq7uJYBw1WjaOqN7QgUPJSukWjhQbBcRwzBcOwiNRoOBAwdG3ADlitfrRTgcRl5eHv31KXMSFQ5yE0FKFzzEWpwMtHVG94YKHkpGSafBJyBeDA/Lsujo6IDJZEJeXh569+4Nu90uutgR28LDzxhTq9UIBALIycnpVB+mq82PurSUQTwRRISQkAhK9DqIh9IFD9+llQ60dUb3gQoeSsYQo++VGBu1z+eDyWRCKBRCdXU1DAYDPB5PxtPdkyE6Y6yqqoqzdAUCAc4C4HQ6EQgEEqoPQ8kMYl9XiYgg/nWQqghSuiAWS/AIQVtnKBMqeCiSQzqZh8NhAOm1g0hHQIRCIVitVjidTlRUVKCsrCzixiVHwcOyLNrb22E2m6HX67mMsVAoBIZhoFarOesOIZYbhN8zimEYxW9oSkLqzU0qEaTkTZlhmLhV0KWAts6QN1TwUCRD7AafZIxkY3hYlkVbWxusViuKi4sF20HIUfAIWaISIdbmx+8Z5ff7YTKZIuJASCYRvel2D5IRQUJuUerSEgfaOkM+UMFDEZ1043TikWwMj8vlgslkglarRV1dXYQlhI+cYlqiG5Majca0j190z6izZ8+isLAQWq0WPp8PLpcLLS0tYBgmQgTl5uZSEZQmchIOsUSQkFuU1K9xOByixARlGjkd92gSaZ1B6oH16tUrwiVGSR0qeCiiwjAM2trauA1W7BtkosLE7/fDbDbD5/Ohd+/eKCwszEq6ezLjks3FYrHAYDB02Zg0XYQaZ4bDYW7ja29vh9/vB8MwEQIoLy8PWq1WtpuJnJCLiI5HLLeow+FAe3u7oCWI7xKTqwhKJUsrm0SLoFAoBKfTibKysojWGaSVRvR7lPRZswUVPBRR4MfpuFwuaLXatIv2CdGVSyscDsNms8Fut6NXr16ora1N6IacbcHj9XrR3Hyua3Hfvn0jREiscaVYl0ajQUFBQcS5IxWCiQiyWq0AELHxERFEEUZpm5FarUZOTg50Oh2qqqoAdLYEtbe3y1oEycWllSokRk/oM/CrwhP4wkcpIshiseD555/H559/jkAggIsvvhj3338/Bg0aJMl89A5FSYtMNvgk4wuNHW0dGThwYERl21THTZeuxg2FQrBYLOjo6EBFRQVKS0tld5PSarUwGAxcDBE55yQmyOFwwOfzQa1Wd4oJoiZ45RLtEoplCSIiyO/3y0oEydmllQhE8Agh9LmUJoJYlsXtt98Og8GA1157DXq9Hi+//DJuueUWfPTRR5L8YKaCh5IS8eJ0pOp3RcYm5lyC2+2GyWSCWq1Gv379IuITEkVKkRZLoJFA6pKSEgwaNEgx4kClUiEnJwc5OTkoLCwE8F8LH4k/aGtrg9/vj9gkycaX6udUgnsoGiVvuomsXc4iSGkurWjiCR4hYn1W0keMTygUynqpipaWFgwYMAB333036urqAAB33XUX/ud//gdHjx5FY2Oj6HNSwUNJmq7q6ajVaq7TudjwhUkgEIDFYoHb7UZVVRWKi4uzku6e7LhEoGk0mriB1FIi9udVqVTQ6XTQ6XSdRBBxgbS2tsLn80Gr1XaKCVKy66G7kqpYk4sIUrpLSyyxHD1GOBzu9KMRANhwGKoM/ugqLy/HypUruX+3tLTgtddeQ0VFBRoaGiSZkwoeSsIwDINQKIRwOMxZdIS+kFK7tBiGgcViQVtbG8rKylBdXZ32jU1KwUMIBoMwm83weDyoqqpCUVGRon+BdgVfBBUVFQEA17GaiKCOjg74/X5ZuECkQOkWHrEQEkEsy3IWQSlEkJKPPZC8hScZ+Pduf2srXMeOw2+1IbeiHIYB9cg1GiWZNxYPPfQQduzYAZ1Oh3Xr1knizgKo4KEkQCoNPsXsd8VfBymkV1hYyBXhE3sOMW+S5FjYbDa0tLSIJtCUikql4lKci4uLAcTe+KL7RSnRpaV0pBQMKpUqKREUXS6hq+9QT3NppYK/tRXmjz5G0O4AAHhOnYLr2DFUTZuaUdEzb948zJ49G1u2bMHChQuxefNmDBs2TPR5qOChxCTVejpSxPB4vV6YTCYEg0Hk5+ejb9++oo7PD7YW8ybp8XgQCATg8XhQX1+P3Nxc0cZOBznVHRLa+IT6RREXpl6vj2iaKedNTclWhmysPZ4IItdCoiJI6S4tqQQb/3vvOnacEzuEoN0B17HjGRU8JCvrmWeewf/7f/8Pb775Jp5//nnR56GCh9IJEuSWat8rtVotmoUnGAzCYrHA5XKhoqICarUa7e3toowdjZgigF8HSKPRoF+/fqKM21MQKpB34sQJlJaWgmEYeDwe2O12rmUGf9OTuwhSCnIRa3wRFG0VjCeClN4+RepeYGw4DL/VJvi832oDyzBQSSgYrVYr9u3bhyuvvDIi4WXgwIGwWCySzEkFDyUCuTT4ZBgGra2taGlpQWlpKZfF1N7eLom7DBBv3aTJZ69evVBRUYFTp06JtEJxUdpmQNxh/F//4XCY2/TcbjdaW1sRDocj4j+y2TJDaceYj1wEjxCJiCAA+OmnnyJEkJKC5KVyaZFrUqXRILeiHB6B+1NuRbmkYgcATCYT7rvvPvTu3RujRo0CcO4H7vfff4+JEydKMicVPBQA4va9SieGh2VZdHR0wGw2Izc3t5MbSMqU93QED7/JZ35+PlcHiFRHFRsxCg92BzQaTcxq0X6/P6JlhlChxEwcB6Uea6W5hPgiqLCwEO3t7Rg4cGDMIHm5iyCGYZKqJZYM5Jo0DKiH69ixCLdWTmkJDAPqJZmXzwUXXIAxY8bg8ccfx1NPPYWioiKsX78eDocDt9xyiyRzUsHTwyFxOiSNXIy+V6mKEp/PB7PZjGAwyLWDiCYbRQ27gjT5DIfDqKmpicgwkFOsTE8hXrVoEghLqkVHF0oUu1q0ks+9nC08XcGPORQKkhcSQdHlEtKpGSUGmQi6zjUaUTVtalaytNRqNVavXo0XXngB9957Lzo6OjBq1Chs3rwZtbW1ksxJBU8PhTSqCwaDkjT4TMbCEwqFYLVa4XQ6UVFRgbKysphrkZPgCYfDsFgscdctZ8Ej13VJQXS1aAARhRJJtWi+lYDfNywdlC4alEi8tStFBEllYYseN9doRK7RKHnMjhDFxcV4+umnMzYfFTw9ENIYkFTLzVaDT3614eLi4oSaZUqV8k7GTnTddrsdVqsVhYWFcdctV8Gj1I1MTMj1L9Qygy+C+DVkiAhSSlXsdFCy4EmlSnGyIojvEpPieshEWjqfTIudbEAFTw+CH6djs9lQVlYmeh0bIDFR4nK5YDKZoNVqk6o2nO0YHo/HA5PJBAApt7GgyJN4LTOIO6ytrS2iWnRXm56SRUNPX3tXIsjv96OlpUUyESR14cGeCBU8PQChOB2p+13FGpufrp1KteFsubRCoRDMZjNcLhcqKytRUlKS0LrlbOGR47rkBr9aNIFsesQdRjY9obowlOwglVjjiyD+XFKIICmztKjgoXQ74sXpiFkrJxqhscPhMGw2G+x2O3r16oXa2tqspbwnMzbLsmhtbYXNZkupyScVFt0P/qYXq2UGqQuj1WoRDofhcDi498gtGygWSt4YM1llWSoRpOTjL1eo4OmmdFVPJxOxMGSjdzgcsFgsMBgMXLp2qkgp1KLFSaput+gxAXrz6u4IuT8YhkFHRwfXNNXpdHItM6I3PTleG0q+ZrOdUi+GCKIuLfGhgqebkWg9HaljYVQqFdxuN1cxs2/fvhG1UtIZW2oLTyAQgNlshtfr5dLj07lBSNGyIl2o5Ul61Go1dDodtFotqqqqAPy3ZQY/MJpUi+a7w+RQLVpu12wyyHHt8URQtHtUq9UiFArB6XQiPz+/kyUonc+n9O99OuungqebkGzhQCktJcSFdubMGVRVVaG4uFi0mw/feiTFDc3hcMDlcsFoNKKmpka0X1hSd2KnKAOhlhkMw3C/+j0eD9ra2hAKhToVSsx0tWg5ioZEUUrj0FjuUb/fj1OnToFhmAgRRMSS2+1GQUFB0p+Rq7KsgGMTC6/Xm3J8HBU8CiedBp/hcFjUtZAvZ2trK1QqFerq6kQP3JTCRUSqO7vdbuTl5YnehV3smwtJi3e5XEl1j6ZknkSuU7VaLVgtmvzq51eLjg6KllIEKVnwZNullQ4kUB4AKioqAERagjo6OnD69GkYjUYUFBRwpUXUarWo9dTkBPlB39HRgZaWFlRWVqY0DhU8CiadvldixvDw2yro9XoMGDAAP/30k2RfPDHdcX6/P6ILe2lpqeip+mK6j7xeL5qbmwEABoMBgUCAC5AlsSGJuEWoS0vexGuZQYKi/X4/1zIjulCiGN89pQsepa4d6CzYoi1BgUAADocDDocD4XCYu5fzLTj8P/644XA4a73l0oHEOvXt2zeleEqACh5FQmqDEAtNKqpeLNHg9XphMpnAMAyqq6u5Im6ZDC5OBX7WWHl5OYxGI86ePSuZ6yndcUk16vb2dlRWVqKoqAjBYJC7KfJjQ6LdIpmyCEiNEgWamBtvvJYZRASRlhlCfcOyufZMoxSXViy6CljW6XSc9YcPv24UcZP6/X5899132LZtG+rr61FdXY2f/exnOO+887r8cbdhwwZ8+umn2Lp1KwDgoYcewo4dOwRfu3jxYixatAgAcOmll8Jmi+zEPnPmTLzwwgtx55MaKngUhJgNPtMVJKFQCBaLhduAS0tLI9YipxYQfFiWhdPphNls7pQ1JtWa012v3W6HxWLhqlGTDA7+8RaKDeE30Yy2CLAsC41Gg1AoJHr/KClR8iYmBdEtM/jVov1+f0S16Oi+YV2VV1Cy4FGySwtIvQYPv24UPyaosLAQzc3NOHz4MPbv34+NGzciJycHDQ0NGDZsGH79619jwIABEWNt3rwZK1euRGNjI/fYI488giVLlkS8bvXq1di9ezeuu+46AEBbWxtsNhveeOMNDBw4kHtdqlYZMVHOna4Hk2qcTjxSFTwMw3AXdElJCRoaGgRvnFJbeFIZm2+NEsoak1KkpTIu333Vv3//CDGTyPmPZxGw2+3w+Xw4ceJEROsEKUvl90QybZWKVy2aWADb2trg9/uh0Wg6xQR1l/OuZLEGiGuhUqlUqKysxMKFC9HR0QGz2Yx+/frh0KFDOHToEA4fPozm5mZO8FgsFjzxxBPYt28f6urqIsYqLCyMaOr81Vdf4a233sL69eu5uJojR45ApVJhxIgRsqtETwWPzEknTiceqYiGjo4OmEwm6HQ61NfXxw1IllI8JOuO4zcnFbJGEaS08CQDvylpvPWmArEIBAIBhMNh9OrVK8IEHl0bhAZFp0+2N17+r36hlhmxzjvDMIp0IwLn7ptKslxGI5WFiliOjEYjLrvsMlx22WWdXnPo0CEUFBRg586dWLt2LU6ePBlzrKeffhpTp07FxIkTucePHDmC6upq2YkdgAoe2cKP0xEKPkuXZEQDCewNBAJcXZpExs92DE90c9JY1ij+uFKsOZn1kiKNRUVFCTVTFWNtQiZwoarByQRFU84hV8GQyHkPh8M4deoUcnJyOhXGk7v47akuLTHGbWpqQlNTU5dj7dq1C0ePHsVLL70U8fjRo0eRm5uLu+66C99++y2MRiOuvvpqzJkzJ+vnhAoemUF88G1tbfB4POjdu7ckm0oigiQcDsNqtcLhcKC8vBxlZWUJX7DZjuFxu90wmUxQq9UJV0nOZgwPcbexLJtwkcZ0zPbx1hSranCsoOhs1opRAko5HtHnvaOjA/3794+ICeKL32h3mJw+Z3dwaWVL8CTKH//4R/zsZz/r5Pb64Ycf4HQ6MXPmTNx99904cOAAXnjhBdjtdtx7772izJ0qVPDIhOg4HeCcK0bK1O5YgocfKJuqpSFbMTzBYBBmsxlutzvpoofZsPDwRWWy7qtM3tBjBUUL1YrhW4FSzRDqLsjVwpMIxEpCziOBiF9y7km1aDlZALuD4JFi/QzDiBKnderUKfz73/8WFDCbN29GMBjkYgfPO+88uFwuvPLKK1i8eHFW48R67p1IRgjF6ZBMHKmItbm7XC6YzWao1epOgbLJjp/JGB6GYdDa2oqWlhaUlpYm3eSTjCt2McZY8LPFCgsLM+K+ElpDOgjViuGnSfMzhHpyULRSN95YoiFWtWghC2C0FShTIkjKPlSZQOoYnnTZvXs3ysvLMXr06E7PEVcpn8GDB3MB8+Xl5WnPnypU8GQRhmEQCoUE43SktJCQ8fkbHr9/VFVVFYqKitK6MWUyhieZYOpkxhWL6HF9Ph+am5tjZotlAqk2HaE0aRIc6/f7uUaaNCha3iT7PUjEAtja2opwOBwRCySVG7Q7WHikEjxi/Nj46quvMHr06E5jBQIBNDU14dZbb8W8efO4x7/99luUlJRkVewAVPBkhUTq6WRC8DAMg3A4jJaWFrS1tYnaPyoTMTyBQAAmkwl+v58TaWKMKzZkXL77qqKiAmVlZYq+KScCPziW0JOCopW68YpR/iJetWi/39/JDRpdKDGduZV63AliCROhcUndsXQ4fPgwbrjhhk6P63Q6NDU1Yf369aipqcGQIUPw+eefY+PGjXjwwQfTnjddqODJIMnU05Fa8JD1/PDDDygoKIgowCcGUq/f4XCgubkZRqMRtbW1shdpLpcLZ86cgcFgyIr7SohsxZckExTN3wRJ01iK9EglGOLVhiJB0aRadHShxGS+M0p3aUmVVi+Wq8xms3Hf3WgeffRRGI1GLF++HBaLBTU1NXj44Ydx4403pj1vumT/rtsDIDfqZOrpSCkYPB4PTCYTAKC6ujqhNPNkkSIAmPTscrlc0Ol0kjT5FHtDJVaMQCCQNfeVEHL79ZuISyQYDOLMmTOKCopWuqUhE8SqFs0Pivb5fFCpVJ3coLHOvdKPu5QxPMlYjp577jnBxw8ePBjzPTqdDvfccw/uueeepNcnNfK9U3QTUi0cKIXgCQaDsFgscLlcqKyshMlkkqw4lFqtRjAYFG08n88Hk8mEUCiEgoICGAwGWTf55Pfq0mq1qKiokI3YUQrRLpETJ06goqICDMN0CopOtm0CJT7ZFAz8atFCLTP41aL5AfHkGtBoNIoXPEpIS1ciVPBIRLp9r4iFRIwvLsMwaGlpQWtra0QGk9VqzXpxwK4Qinsxm82y63lFIFYok8nEua9Iewg5IaX7TkqICIoOiua3TZBTULRSN165rTteywziDuOf+1AohPb2duTn5yMvL09xm7zcg5aVChU8IiNW3yu1Ws1tSqneeMjmazabodfrO7mApHSbpTs2v+pwdNp2tisix8Lv96O5uRmhUAi1tbVcnIJSxYUSiNU2IV5QdG5uLvR6fbcIipYKuQkeIeIFxJ86dYpLyPD7/cjJyemUIi9nEUQFjzRQwSMSLMty7itys0j3giWiIZVx+C6g6upq7hcxH6mEAxk71U2eH2MkFPeSbC+tREl1zQzDwGq1wm63o7y8HEajMWKzoIIns3QVFO31erliedGuMLFTpJUgHIRQ6rrJuQeA8vJy7l4hJIBJywx+VqBcRJBUx59lWSp4KOkhVYPPVKwk/EaZXaU+SyUcyNiprN1isaCjowOVlZUoKSmJ2eRTDhYevgUtPz8/bqab3ARPTxNhtFJ0cihV8AD/TRIh648lgPkiSKhaNHlPNo6DFBYeclyo4KGkRLpxOl2RjGhgWRatra2w2WwoKSlJKPVZDg0+gcgmn2TtXTX5zLaFhzRUDQaDMS1oqYxLyRyJVoqOzg5KJihaqedd6YKnq3txvJYZxApot9sjrIDkv5lwhUoheMi9ngoeSlKwLAu/349gMMgVyMpWg0/gXKVhs9mMnJychBtlJjN+KiQ6tsvlgslkglarTXjtUq07EWHCMAxsNhtXIj3afRVrXDmi1M1YSmKlSPOzg5INipbr+Y9HdxA8yRKrZQYJis5k01wpBA9pm0MFDyUh+HE6JpMJOTk5kpbK7mpj9/v9MJvNXKXhwsLClDLBpKAr8ZBOK4tsWHhYluVaWHTlvkpm3GyhxM0sG8cwVnZQrKDo6MBYuZ33RFGy4BGz8aZarRasFi3kChUqlJjKOqJdcmJBLTxU8CRMdJyOVquVvBJyLMHDT9UuLy9PudJwNmJ4+CnyZWVlKbWyyLTgScZ9lcy46aLUDSkd5PCZ4wVFR3cQ12g00Gq1aG9vl6xvlBQoVagB0hXtI8RyhZJz397eDpvN1ikejBRK7Or8i9HWQwgqeKjg6ZJYcTpiF9YTIlo0sCwLu90Oq9UqSoftTMfwkADf3NzctKokZypome++6tWrF4xGY0o3UjlaeABlb2pyI1ZQtMViAcMwMYOic3NzRW3pIiZKEGZCZMM6pdVqodVqBVtm+Hw+OJ1O+Hw+AOhUKDH6Hi5lSjqg3PMqBlTwxKCrgORM9Lriz+F2u2EymaBWq9GvXz9RKiRnKoaHWEgCgQB69+6ddiuLTKSlk+KBQvWLUkFu4qIn3/QyhUajQU5ODjQaDcrKygB0Dor2+/0AkHJQtFRQl1b6xIsH8/v9MSuFq9VqSassy+HYZAsqeAQIh8MR9XSELhCNRsMFgUkFsSKdOnUqpViXrpA6hodhGJjNZtjtdvTq1Qt9+/aVdZNPsuaTJ0/C7/ejT58+ovQZk2K9LMvC7XZz8SU9+SamJBIJivb7/dBoNFmtFK1kwSO1SytV4lWLjq4UzrIsmpubI85/uiK4p7eVAKjgEYRsTvEuDqktPAzDwOv1wuv1ory8PKVYl65Qq9WSiDYS4Eu+zGJ3YpdCQJDYIoZhkJ+fL1oHdkB8YenxeHD27FkA58Q5y7JJWwjk6mbrbnQlHJIJio4ulCdljRilCx6lrF2oUrjH44HVaoXBYIDP5+OqRafbLqWnV1kGqOARJJGAMakED8uycDqdMJvN0Gq1KCkpQUVFhejzANJ8Bp/Ph+bmZm7cmpoa0W8+YgsIkn3Fr9AqJmKJC35hxqqqKs78TfoJyclCQEmdZArlSVUpWkmiIRq5uLRShWVZaLVaFBUVoaioiHuMf/47Ojq4lhnRhRJjfc+phYcKHkES+bJIIRa8Xi9MJhMYhkFtbS3n65UKMT8Dv8JzZWUlSktLcfjwYUl+VYgVwxMIBGAymeD3+9G7d28YDAYcOnRI9Jt9uoKH31esqKgIgwYNglqtRiAQAICk0qb5xdaohUd6xLqWYhXKI+c4OiiaL4RSSY9WsuCRq0srUYSEiZAIJvXgyD7hdDoFyyPk5OSk1KZow4YN+PTTT7F161busRUrVuDVV1/t9NpDhw5xwdebN2/Gpk2bYLPZMGTIEDzyyCMYPnx4KodCdKjgSRExY3iCwSAsFgtcLldES4VAICCp20wMwUMyxywWC4qLizs1+cx2RWQh+KnxRqNRVPeV2BCLGcuynYLVyXGI3pi66iXl8Xjg8XgQDodx5syZHt9GQakI1YjhB8W2t7fDarUCSD4oWumCR6lrBxJfP78COEGoPMIzzzyDY8eOob6+HoMHD8bEiRMxbNgwVFZWxhx78+bNWLlyJRobGyMeP3LkCK6//nrcfffdEY+T+8b27dvx+9//Hk8//TSGDBmCjRs34rbbbsMHH3wAo9GYzGGQBHp3SxExxALDMGhtbUVLSwtKS0s7tVSQOk4oXeHg8XjQ3NwMtVqN/v37d8ocy2ZF5Fjw3VdC2VexREQ6pLJefkPSrnqiJUJ02rTX6+VaefDbKPAtCfysEUpqZHrzFSsoWsmiQekurXRcT0LlEZYtW4Y9e/bgm2++wbfffou//e1v8Pl8qKiowOjRo/Hss89y7nyLxYInnngC+/btQ11dXafxjx49ismTJ8d0+2/YsAE33XQTZs6cCQB45plnMHXqVPzv//4vFi5cmNJnEhMqeASQ2qVFgnpJTZr6+nrughNrjkRIdXy+RaqqqgrFxcUxm3xKYeFJZd38ys69e/fmfOPRSLHmZMckKfHJVnROZV3RmyM/HogES/LjBPR6fUZ6CVHEIdWgaFI0UYl0R5dWOlRWVuKqq67CqFGjoNfrUVNTg2PHjuHgwYNcrCjh0KFDKCgowM6dO7F27VqcPHmSe47clwYOHCg4T0tLC06cOIExY8Zwj2k0Glx00UX46quvRPs86UAFT4qQG36yF6fP54PJZEIoFEKfPn3iVu2Vm+DhNygVskilO36iJCMg+Fa0RCo7Z1Pw8GOKUqnonO6a+Bkj/GBJYh7nB8sKdRSnIqgzcrSUJBIU7Xa7EQ6H4fV6JQmKlhI5HvNkkLLwIKn8PXjwYAwePLjTa5qamtDU1CT4/qNHjwIA3n33XTzyyCMIBoMYPXo0lixZgoqKClgsFgBAVVVVxPsqKirw3XffifxpUoMKnhQh1ZYTvTj5Qb2Juigy4dJKdHzS5DMnJyemRUpo/GzG8MhlzV2NyReScospEooTIL2EvF5vWnEiFPkQHRRts9mgUqmQn58vWc8oqVB6NhLDMJLE0qVr+SKCx2AwYNWqVbDZbFi5ciXmzJmDd955B16vFwA6hQnodDouwSLbUMEjQKJf3kQECcuyaGtrg9Vq7RTUK8b46ZBItlO0KyiZBqVSW3hiWD9bKAAAIABJREFU/ZILBoMwm83weDxJr1kqkRZrTLfbjebm5qREWbaJ7iUkRkfx7opSrQ0kNbqrnlFyFLtKd2lJtf50M2ZvvPFGXHHFFZxV8LzzzkNDQwMmTpyI3bt3czE/0eImEAhEXEPZhAqeGCSy8XW1oRMLg1arRV1dXcSv5ETIpksrOpNJbk0+hYKL+ZaSsrIyVFdXy2LNQhsev6YOiSnK5MYo5udMJk5EqroxFHGJJdSie0YlEhRNznmmRIhSRSZBapdWqqhUKk7sEEhWsclkwiWXXAIAsFqtEe4yq9UaNyMsk1DBkwaxKhX7/X6YzWb4fL6kLQzR42da8JCAajH6SGWyOWkq7qtExhUD/pikpo7ZbEZJSUmXcVBKJVacSKy6Md01NV6ptY6SSY2OJXaJJUioSJ6UlaKp4Ik9bjr3mmXLluGrr77CO++8wz12+vRp2O12DBw4EGVlZairq8P+/fsxYcIEAOfc3wcOHMANN9yQ9vrFoPvcWbKARqOJ2NDD4TBsNhvXOyrdWAwiGKT6Ake7hvx+P5qbmxEKhUQJmpWyfQEZm+++EqPXmJSCh19TRyiNv7sTr25Md06NV+Lmm849hy92uwp+l8Li1x1ieMReP7nPpyN4pk+fjq1bt2LZsmX41a9+BavVimeeeQbDhw/HpEmTAABz587FsmXLUFdXh+HDh+O1116D2+3GddddJ9InSQ8qeGKQjEuLXwnXYDCIlkocy3UjFmT8UCiElpYWOBwOlJeXw2g0ilYdVkoLVWtrK9ra2riMMbk2JiU3+59++kmUmjpiIJdeWkJ1Y4RS43U6HUKhEDo6OmAwGBSTGi+HY5wKUtSiilcp2u12o7W1FeFwOO2gaKVbeKRYP7kO0xE8o0aNwvr167FmzRr88pe/hE6nw5QpU3D//fdz997rr78eLpcLL7/8MhwOB4YOHYpNmzahrKxMlM+RLlTwpIFarYbH4+EyGvr27St6cFYqJcEThXwJjh07BoPBkFRAdSJItam63W4wDAOXyyV6oK/Ya25vb+esOko5vtlEKDWeVI89e/YsFw8UCoUUkxovxzV1RSZEg5DFLxwOR8R9pRIUrXTBI8X9nvzwTEbwPPfcc50emzBhAueuisXcuXMxd+7c5BaYIajgSZFAIACPx8PV04lVfC9dpLKSeL1eNDc3AwB69+7dKRhNDMReeygUgtlshsvlgkajQZ8+fUTPahJLRPBr6pSXl8PhcHSr2JRMQqrHqtVqlJeXQ6fTibIxUmKTLSGt0WhQUFCQUFB0tCWIiATq0uoMiTXt6d8HegeOQSzxws9eIgGZJSUlkq1DCtFAsoMqKirQ0tIiWSq0WOKBn9pP3Fc//fSTCCvsTLqNSVmWRUtLC1paWriaOn6/H3a7XcRVpo9SrUPke5nIxujz+TIWKBsLJVsb5LDuWEHR8SqCh0IhrjimHD5DssjFwtMdoYInQViWRXt7O8xmM5e95HQ6RWsgGguxBA9fNPCzg9ra2iTb+GJlsSUD6del0WgiUvuzXdRQiFg1dZQqLpREvGwhr9creaBsd0POQq2riuAulwutra2wWCxc8LRer0dubq4iYr+kOPZkD1Gy5UsMqOBJAK/XC5PJBIZhUFNTw/2qVKvVCAaDks4thuBxu90wmUydRINY48cinY2e774S6tclh7YVBP5ahWrqyFXwyHFNYsLPFiJkOjVezsIhHkpbNz8ouq2tDTU1NdBoNFzneDGDoqUkXkHVdCANVeXyObMFFTwxINlLfPdPaWlpxAWj0WhkbeFJJGVbSsGTytixLFHRSGnhSXTNLMvCbrfDYrF0WVNHbuKip974Ek2N76qbeHdHaYKHD1m7VEHRUkLcWVIInp50/caCCp4YtLe34+TJk3E3MqnTrlOdI7phZryKw3Ky8BD3lVqt7rIydbZdWsnU1FHqxtFTEEqNF6oSrdPpIjbFRNwjShUOSl03EH/tXcV+2e32rNaCIpYYKcbt6fE7ABU8McnPz+8y5VmOgodUSc7NzU0oZTsZi0ayJLp2viVNyH0lhFTr7krw8ItLVlZWdrL6pTJmNpDjmuRCrCrRJEbE4/Ggra1N0D0iRv0tOaBkwZOMaEglKFrKAHip+miFw2Fq4QEVPDFJ5MYlJ5cWaWfh9/u5dhaJji91NeRYJOMSikaqdcdbc3t7O0wmEwoKCpKqqUPFhfIhqfF8S160e8RisXSyDEj1i11qlCp4yPcs3Wrr8YKi+QHwqVj94iFlzTUqeKjgSQs5WHgYhoHNZkNbW1tK7SyyFcPj8XhgMpmgUqlSarOQSZcWqakTCARSarkh5Vop2UPIPRJtGQgGg2huboZer89aanwqKFnwSBGcG6tSdCJWv2SCouXaR6u7QAVPDBK5QDMleISsSPw0+fz8/JTbWWQ6hofvviKddlO5OWVC8PBjoUhNnVRvRnITPNTqJD5CloHjx4+jV69enDVIKDVer9fLKlMIUK7gyaRFLRGrX3RQNDnvsazDUgoeOVp4Mp09RgVPGmRK8ESnvvt8PphMJoTD4Yg0+VSQcuPjHx/ivrJarSgqKkq7S7jUMTz8mjrpdIwnY0qBz+cDwzCKqC3SUyHxQPzrJxwOR3QSt9lsAOSTKaRksu26iRcU7ff7u2yQK5XQZBhGlvFl5FyJ4YpMBCp40kDqbub8OYBzN0qr1QqHwyFaE0opawkR8cBvY9GvXz9RuoRLFcNDLGd2u12wpk4qiC0qw+EwLBYLnE4nJ/z41gK6WcoHofOu0Wg6pUvzXWH89gnZSo1XqoVHbutONiianF+fzwedTifa+ZabhYdhGMyfPx8zZszA+PHjUVlZyT0n5TmkgicGiRxwUutBSv8ocWmR2jSFhYWiNqGU0kpFvtgnT55My30lhNgigligWltbodPp0rZA8SGfWYwvMmlGajAYUF9fD5ZluWJ6Xq+Xa6mg1WojBFB03Ah1aWWORM55rCrRsVLjyXmVokq0VMXvMoESgsTjBUW3tbUhEAjAYrGkXApBCJZlZfUjKBQKwe/3Y/Xq1di+fTumT5+OsWPHYsCAAZKePyp40kRqwRMMBuF2uxEKhUSzjvCRwjXEsiwcDgfMZjNYlsXAgQNFb5wp5rp9Ph/Onj0LACgrK5PkfBKBkeqXmd+MtLa2FgUFBQiFQgiFQjHryAi1VOALIIr0pCoqu0qNd7vdkleJlrtwECLbLq1UIUHRxP3Zq1cvUYOi5Ra0rNPpsGbNGnz22WfYtWsX1q1bh7fffhtTpkzB9OnT0adPH9H3OoAKnrgk8iuYWGDE9o+SdgUdHR3QarWor6+X5AYktmuItOFgWRa1tbU4deqUJF3CxXDF8V2EpKaO3W6H1+sVaZWRpHKcWZZFa2srbDZbwoHTQi0VSNyI1+tFe3s7fD4fAODs2bMRGURK3CzkjljfW6EgWaEq0ekWzVOqdQdQ9tqBSNdTMkHRQiIoelw5CR4AKCoqwowZMzBjxgx8/fXX2LVrF/7+97/jr3/9K6ZPn46rrroq5WScWFDBkyYajUZUCwl/gyspKUHfvn259G0pEMulJSQeiGlcCtJxybAsyxVojK6pI6cUcq/Xi7Nnz0Kj0SRURDIe0XEjLMvihx9+QGFhYac4Ar4AogHR6SG121DIutdV0Ty9Xh/3vCpZNCjBpRWPrixU0UHRwDkvALEE8UWvSqXCjh07UFtbC6PRiN69eye0hg0bNuDTTz/F1q1buccOHTqEFStW4ODBg9BqtRg9ejQefPDBiDEvvfRSLgCfMHPmTLzwwgtdzjlixAhUV1dj0KBBWLt2Lf74xz9i69atmDRpEhYsWIDzzz8/obV3BRU8cUjUwiOW4HG5XDCZTNBqtVxrBb/fL2kmWLrrJ+4ri8UiGF8kVTxAqsKEX1NHKMMt2y0rgEjxWFVVJWrsE389AFBYWMjFEcQyoUvlMukpZHIDTqRont1uRygUEjyv6bpes41SXVqEVIKLSfxXtOhta2vD6dOn8cEHH8But0Ov12Po0KEYPnw4LrjgAowbNw6lpaURY23evBkrV65EY2Mj95jJZMItt9yCKVOmYMuWLfB6vVi+fDnmzp2LHTt2cA1bbTYb3njjDQwcOJB7b7z2QH6/H19//TW+//57fPnll/j+++/h9XoxYMAA3HzzzaiursamTZtw6623Yvny5Zg4cWJSx0UIeudKEzEETyAQgNlshtfr5aokkxuO1Knv6YxP+kkxDIO+fftGZJ0A4OorSCV4klk3v6ZOvAKN2RY8qVZzFoNEXSbZyh6iQdapIVQ0L169mJycHLAsi3A4LDs3SFcoWawB4mRTEdFbVVWFlStXgmEY7N+/H16vFz/++CO+++47bNu2DT//+c/x1FNPAQAsFgueeOIJ7Nu3D3V1dRHjvfvuu8jNzcWyZcu4+9Hy5csxadIkHDhwAJdccgmOHDkClUqFESNGxI29IZ/vu+++w6OPPgqr1Qq1Wo0BAwbg+uuvx7Rp09DQ0MC9fvjw4Zg7dy62bdtGBY8ciFUYMBEYhkFLSwtaW1thNBpRU1PT6WKXWvCkssELua9i3WTI+sXeEJOJPSI1dXQ6XZc1dbIleILBIEwmE3w+X8LVnMVKl483TqKNNYV6Sil54xELuW7A8erFuN1usCyL48ePc9l+fFeYnC0oSndpSbF+lmVRVlaGIUOG4Be/+IXgPIcOHUJBQQF27tyJtWvX4uTJk9xz06ZNw8iRIyN+fJH3Op1OAMCRI0dQXV3dZaAxeZ/NZoNGo8Hdd9+NkSNHYvDgwRGvI3tGOBxGXV0dampq0jgC/4UKnjgkcuGlEsPDjyHR6/VxN2Gpa/0kI6hYloXT6YTZbIbBYEjIAiF1gcB4kMBvt9vdyXKWzripIjQuy7JcyYGysjJB0SsnYmUPEQHkcrloIT0Fwq8Xk5ubC6/Xi/79+3Pi1uv1wul0dqoSLTdxq3SXlhTrJz/I+eNGz9HU1ISmpibB9/fv3x/9+/ePeGz9+vXIy8vD2LFjAQBHjx5Fbm4u7rrrLnz77bcwGo24+uqrMWfOnIi5yHUybtw4jBkzBqFQiEs+CQQCaGlpQV5eHsrKygAANTU1eOWVV0TbQ6jgSZNkLTB+vx8mkwnBYDChX/JSuoWAxNfPr+4s5L6KN36mLSbRTUkHDhyY8GabyaBlUpBRpVJxMVtKRK1Wd1lIz+fzZaTbtNyQq4UnHvx+VF2JW6lT41Ndu1KRwhpO7u9inZPXX38dW7duxaOPPsoJkx9++AFOpxMzZ87E3XffjQMHDuCFF16A3W7Hvffe22kMvV6Pt956Cxs2bMBtt92GWbNmIRAI4NVXX8VPP/2EJ554AvX19dzrxTomVPCkSaKCIRwOw2azwW63o7y8HGVlZQmfRKncQvyxY8FfdyrVnaUUEELr5ld1TkVEZKpHl9Vqhd1u79IlKCVSWrOECukJdZvmb5Kkp1RXa6ZISzzRICRu46XG8+s+ZcLy0h1cWlIJnnQtrAzDYMWKFdi4cSMWLlyIOXPmcM9t3rwZwWCQc5Ged955cLlceOWVV7B48WJubvL5PvnkE7z66qtoaGjgrERarRaXXHIJPv74Y/zud7/D888/zwkqsaCCJw6JurT8fn/M56PdQKnUFchEg8/omxxpsWAymRJ2Xwkh1dqjLUfJxBXFQ2rB09HRgebmZuTn52c8KDmbJBI4a7FY0q4hIycy1R9IbJK1ksRLjSd1n8SsGtzV2pXsOpVCsInRoNPv9+P+++/Hxx9/jMcffxyzZ8+OeJ5kBvIZPHgwZ+EtLy/n1kIET9++fbFixQro9XouG3Tq1KmoqKjAAw88gD179uCKK64QVQT2jLuthMTb0EkRvlhZTGLMkS5CLjO/34/m5maEQiGuqm8640sZwyOWMIseVwrMZjMCgQD69OnDWT56MkKBs3xXmM1mi9goGYZBIBCARqNRnIhQEum6hYRS48WsGizl2rONVDE86YzJMAwWL16MPXv2YPXq1bj88ssjng8EAmhqasKtt96KefPmcY9/++23KCkp4cQOH1Jyg3+vJlXj+/Xr16UhIVWo4EkTITESCoVgtVrhdDpFcVlkMjWduFrKy8thNBrTvnlIGcPDMAxOnjyJYDCYtjDjjytFjy6fz4eSkhL069dPNhYLufXT6mqjBM6JxuiYEb1eL8tf9UrdfKVYdzJVg6NdYcmcWyW7tEhfPClcWumM+cc//hH//Oc/sXTpUlx44YURxQULCwuRl5eHpqYmrF+/HjU1NRgyZAg+//xzbNy4EQ8++GDEWGQdDQ0N+OCDD3DgwAGMHTs24hzv2bMHbrc7oqGoWFDBE4dEvjj8tHR+sGxxcbFoLotMpKY7nU60tLQgPz9f1HLeUmyqDMOgra0NwWAQZWVlMBqNot0kxFwvqVPEsizy8vJQWloqG7GjFPgbZVtbG/r27QsAEUX0zGZzRG0gvV7fIwKipSJTQi1eanysKtFdBbsrOUuLHywu9rjpHJOdO3cCAJYuXfr/sffmQZKUdfr4k3Vf3dXV933MdM8MMDPIIQOuKId8RQ1U9qujuEIouIrKwrLI7iqyEDAQijuieCAszGpsECgYChqiIqL7VX9yDJeAMDPMTM90113VdVfWlZm/P9o3easqqyqzKrMqa6aeiAno6qq338rjfZ/8HM+Dm2++uex3u3btwoc//GF8+ctfxtDQEO644w4Eg0FMT0/ji1/8Ii655JKy95N5fPSjH8Xjjz+OW265BR/4wAcwPz8Pu92OlZUVfOtb38IZZ5yBLVu2lH1GDfQIT4sgbemZTAZ+vx8Gg0E8eWpBS8KTz+fBcRyi0aik8nCrUHvutBq1wWCQDJe2AjUID8/zCIfDWFtbEwu9l5eXdRVNIdDjnBqhnjaQVPs0iRa0mi5Ril6ERxno1vhGxe61WuO79ZgD2kWnlKa0vvKVr5T9/NOf/rThZywWC6655hpcc801sv7G+Pg4brrpJuzZswf33nuv6J5ut9txyimn4Itf/CKGhoZkz1kueoSnRZCQ+8rKCsbHx+F2uzUJB6tNeOhN2Wg0Ynp6uukao3pQq4anUlPH4XBg//79KsywHK0SnnQ6DZ/PB5vNVhYp0yLSpYbw4LEAqfZpYpZaL13SzQXRWkJPpEGq2L1ea3yxWITFYoHD4dBlmrMetIpO6dE4tFQqYceOHdixYwf2798vKvbPzc1h48aNmv3dHuGpg3o3PbEqIPnMpaUlzS4qNQkPLXpI0lcrKyuaPem3WsNDpwk9Ho+oqUPEGNVGs8SkVCrB7/cjm81iYmJCrEFpddwemoOUWSpJl7AsW5UuISRIzc6h3vnWBrV0n/L5PEKhEDKZDOLxeMcsUJqFVtIjekzzmUwmpNNpeL1emEwmbN68WXw4jEQiZXV8qv5d1Uc8DkAIg8Viwfz8PJaXlzVl0GoRnlqih1p3gTU7dj1hvlrt9K1CKTEh5qmBQAADAwNYWlqqubjobQM8nkhYvXQJy7JVnUN0FKjVrr9ug54iPHJBzi1pgbbZbJIWKO1ojW8WWhEejuPq2ul0Aq+99hruvvtuvPDCCzCbzTCZTLBYLLBarSKRvf/++1Wfd4/wKABx2s7lcqJVAQBNrR+A1gkJnb4aHh6uKvLVcuNrJsIjR1NHKwVqJccin8/D6/VCEISGdVt6WVR7eBO1tIFYlq0S0aMJkNxIQbeSyW4kPASNVKLrtcbTtV6dwPEU4dm9ezdeeuklvP/974fJZEI2m0Umkyn7rxYkrUd46oDc9DRhGBoaqnLa1rpYrhWDUqJRU8+zSy8RHqWaOlrVxTSKHEkVJXfSo6sH9WA0GiVF9AgJoiMFNAmq5SfVjcShmwlPPdJQqzVeiuB2QvxSq+OutxqeXC6Hp556Crfeeisuvvjitv7tHuGpA3oDlkMYtLopDAaDaLAmFyQalc/nG3p2aUl45JI1Ot0mV1OnE4XApCjZarUqat/XI+HR45z0BlobqJGfFE2AutUXDehuwqN07rUIbiut8c1Cqz1Eb4Qnm83CarVi06ZNANb3KgISnWMYRpM59whPAyQSCVmEgeM4zUKhSggJz/OIRCKIRqMYHh6uika1Or5SNIrwVM53eHhY9kKi1bylInaVXWJKC+p65OLYgVyzVJPJBI7jEI/Hu8os9XgiPJWQEr+s1xpfWevVyt/WMqWlJ8LjcrnwyU9+Ej/60Y9wyy23tLW+qEd46oBhGMzOzjbcqEjXkFaQu7GTYmqbzVYzGiUFrewfgPo1PHS0RMl8CbT2vQLeLEomYpJKnNcr0SM8xy6kCqLT6TSCwWBNs1SSCtMbupnwaKFlU6vWq5bsAf1PqUr08RDhSaVSeOGFF/DUU0+B53ksLi5iYGAALpcLbrcbDocDg4ODmJiYUP1v9whPA8jZVNtp/SAFOn1FF1MrGV+rzViKTDVq4VYytpaEh3iKcRyHubm5lsQktdhAMpkMMplM03UGvaiTdiCRApPJhPHxcQDdY5aqxyJXuWjX3OvJHlRG+eS2xmslPKhluUUzIDIj27dvx+9//3v84he/QLFYBMdx4np04okn4ic/+Qk4jlOVrPUIjwroFOGh00FSxdRKxldaI6RkbDpasra2hlAoBI/HU7eFWw603LDD4TDi8bhqnmJqzpXjOAQCAaRSKTidTqTTaeTzebHlloTZaxXSdiu6nZw1Mksl9SJ067Tdbm/7eezW49xJd/pasgf1WuMr71MtUk+kAUNPEZ7p6WnccccdZZEnlmWRTqdRKBTEdRdQ11YC6BEeVWA0GpvuopIDKcJD0lfNpoMaja8WSISnnqZOs9Bi3plMRuzcaPW40lCL8CQSCfj9fvT392Pjxo3iUyHdcksKaYmHF7246ulJrxl0E4FrlBpqZJaazWYRjUbbbpbarSktvRmH1muNJxs8rRLNcRxsNpvoGq4GyPrYqVZ7KdhsNpx44okA3oz2j4+Pw+PxVL1X7fOpn6OgU8g1EG1XhKdQKCAQCIBl2ZbSQTS0rOEB1juwlpeXMT4+joGBAdUuYq2iJiaTCRMTE6oW07U612KxKKYtSRcbz/PieaNbbsnCUavbhJAf4s7cg34g1TpNp0qIyKXWZql6Ig5y0Q2puHrnNxKJgGVZLC8vq5bqJA/iejkuhExHo1H8+Mc/xh//+EccPnwYH/rQh/D2t78dr7zyCt797ndrUr8D9AiPKmgH4eE4DuFwGJFIBENDQ5ienlbtItaihoe09Pt8PgBQzTmehhqERxAEJBIJBAIB9Pf3Y2lpCYcPH1ZphtV/q5nPkJz34OBg1Xmv9zTeSFm4WCyKvl90t4mewt/dDLUiJfXMUnO5nOpmqd0a4enWeZPzm0gk0N/fD5fLVbc1XokNit4iPAzDIJFI4D//8z/x61//Gueeey6Wl5exsrKCQqGAb33rW3j++eexe/duTQr69XEUuhytCAPKAcuy4DgOmUwGGzZsgNVqVXV8tQkbrakzPj6OSCSiyQ3XamSqUCjA5/OhWCxidnZWLEDUUtBQCWglZ6k0INGrUDIHuttkZWVFjLixLCsWWmrpL9VD66iVKqFrRSq7hpSkNLuVOOgtpaUUJEIlJ9W5traGUqlUFQWqJLlkzdHDQwwpQP7Vr36Fp59+Gtdeey0uvfRS3HHHHThw4ADe9ra34d///d/xrW99C08++STe/e53q15w3SM8DSDnBjIajZoU/ZI0BsuyAIDZ2VlNQpNqER4pTZ1CodCRlvd6EAQBkUgEkUhEUvtHK8LTjJaSXCXnZkG6TejogZT8fuXGqYcFVO9oJ3Go1Aaq7BpSEiXoVsLTDSmtemhGJboWyY3FYvjrX/+KpaUl2Gw22efznnvuwe9//3s8+OCD4murq6u49dZb8eyzz8Jms+Hiiy/GtddeW/YQ+8ADD2DPnj0Ih8M44YQTcMMNN2D79u1lY5M19YUXXsAJJ5wgqiyvrKyIxO5d73oXfvSjH4mZAbXX4R7hUQFqR0gEQRCd2Eka47XXXlNt/EqoUcNTS1NH65Z3pWNns1l4vV6YzeaaRcmdjPCQ+VksFlWLpuWCjgINDAwAQFW7bT6fh9FoFDdNu93eiwLpDLW6hmp5SdEpzW4mPN04bwKl0Qyprj9yr7700ku4//77EY1GYbVasW3bNmzfvh3bt2/HySefjMnJyarxHnjgAdx555045ZRTxNcKhQKuuOIKLCws4Ic//CFWVlbwpS99CSaTCddeey0A4Cc/+Qm+9rWv4dZbb8UJJ5yA++67D5/61Kfwy1/+EkNDQ+JYtFVTsVgU18NsNivWHfI8j3Q6XSbqqSZ6hKcB2l20nE6n4ff7YTaby9JXWtpXtEJKisUiAoFATU0dLQuilRATjuMQDAaRTCYxPj4Ot9td89x2gvAQw9REIiEeR60Xb7nfs1YNCfEgisViqruMHwvQW3u3FJklUQKWZUUvKVLMToQS5ZqldhrHQkqrVZVoQnLf+9734r3vfS8OHDiA5557DqlUCn/5y1/w0EMPIZ1O48EHH8Spp54KAAgGg7jpppvw9NNPY2FhoWzMX//61/B6vXjooYfgdruxadMmfOELX8Dtt9+Oz372s7DZbLjnnnvwsY99DBdddBEA4LbbbsMFF1yAH/7wh/j85z8vjkWuobPPPhtf/epX8cADD+DKK69EOp0Wr8dHHnkEALBx40bxO6mJ43tFUglq1PBUEoe+vr6yk62135XSseVq6rRb1FBqnsQPra+vD0tLSw3TMVrp+9QaM5VKwefzwel0YnFxUfdEga4hIZDaOEknESFB3WKtoCb0/n2logRHjx6FzWZDoVBQbJbaSRzLKa1mMTQ0hDPPPBPbtm0T/4bP5yuL8Lz66qtwOp342c9+hu985zs4cuSI+Lu9e/fihBNOEGvFAGDHjh3IZrN49dVXMTc3h+XlZezYsUP8vdFoxGmnnYa9e/eWzYVcLxdeeCGeeeYZ3H1p8WYGAAAgAElEQVT33XjxxRfx2muvwWq14sYbb8Rjjz2Gj3/849i6dSuAng6PLtGKtQSdvmpEHPRCeJRo6hDyoEW4uRGZIgrUhUJBtiEp0D5TUlpxupFfm94htXGSTiJCgmhrBbqT6FiF3iI8ckAK4Z1Op5hWqKcdo6fuvuMtpdXMmAaDAdPT02XvOe+883DeeedJfj4YDIpK4QSjo6MAgEAgIK79Uu95+eWXJcc0m834p3/6J0xPT+N3v/sdFhcXEQgEEAwG8Y//+I/45Cc/qXpjDsGxu9qoBC1TWplMBj6fDyaTqWH3VScNPgnotNDY2JgsTR2ygGqxGNUiJjSJbEaBWuuUFvHnCgQCqihOqzEnLcau7CSiiyxJFIgUY2qlJ9NpdON3qbxXa2nHkJQmbaNAE6B2n8tuJjxaPRS26qOVy+WqHhRJXSEhwfRr9HtoF/RKjI6O4jOf+Qw+85nPIBKJwGAwYHBwsOl5ykWP8KgApWSkWCwiGAwinU7LrtfQOsJT74arTAspTbsQQqX2pi61YWezWfh8PhiNxqZb+LUkPIVCAV6vFxzHYX5+viV/rm5DLWsFsnHSejKV6ZNuRDdGeAgarUcmkwl9fX1VNgrkXHbCLFVvnlFKQOqPtCA8rRwTktqkQX52OBxihEfqPbUKjx966CFwHAen0wm32w2XywW73S4WWJvN5p7woJ4h11pCqu5FLvvWOsJTKwpDDDRLpZKitBANrep4aA8wJUXJjaBV5COXy+HgwYOq+XN1O2i9kUo9GZZlqww2BUFALpeD3W7vmo2tG89xM5GGenVdlW3TNJlV0+6k2yM8enRKHx8fr+oQJudxfHxcrAUKhULYvHlz2XvGxsYkx7zvvvuQSqXK5kUefgYHB2EymfDwww9r8jDYIzwNoFZKi448NOMl1S77CnLT0VowrW7QWnVqEWJCFJ1dLpcqRb9qEx6WZREMBsHzvCbCkc1Cy5RWs5DSkyGqs+l0GuFwWFVV4R6qoRZxaMYstRWhy24mPFpFpwRBaGk9fOtb34qf/OQnSCaTYgfu008/DafTiRNPPBEWiwULCwt45plncPbZZwNYJ7rPPfccPvKRj0iO+eUvfxnZbBaZTEZUfY9Go3jllVdw6NAhvP/971fFa1EKPcKjAuoV5pZKJQQCAaTT6ZYiD1oTHpqUEE0dm82mihaMVhEenueRSqWQTqebjj5JQS0iwPM8QqEQYrEY3G43CoWCbshOt4COAoVCIczMzIBhmLqRA1IL1OkoULduwFrNW66CMC10qcQstdtTWlpFeFoZ913vehe+8Y1v4Nprr8X1118Pn8+H3bt345Of/KS4L1x++eXYtWsXFhYWsH37dtx///3IZDL48Ic/LDnmO97xjpp/74orrsDExIRm902P8MhAow2QYRiRkJAbk05fDQwMKEpfSaEdER5SW1RLU6dZqB3hoYuSiYCgmouFGoSHkEa73Y6lpSWwLItoNKrSDI9v1FIVJvUjoVCoa1qp9Yh2K0QrNUsl/6QUonuEp3rcVvYdq9WK++67D7fccgt27tyJ/v5+fOQjHynT19m5cyfS6TS++c1vIh6P46STTsKePXtqFiFXln+Q681oNGLnzp24/fbb8YlPfKLpOddDj/CoBFLHYzQaxfSVwWBoKn0lBS0JDxEaO3LkCAYHB1XvGlIzwsOyLLxeLwwGA0ZHR5HNZttSDC0XJKKXyWQwOTkpFnXqMX2kxzk1A1pwrTJyQFqpw+EwAFS1Umu5QfYiPM2hkVkq0QYiaU1yTjmO61qZA62OuVLC85WvfKXqtbm5Odx///11P3f55Zfj8ssvl/U3as0nl8vhwIEDIiHSggR259XRZsjZGEiEJBwOI5VKtVw4KzW+FgalhEDwPI+JiQlR4ltNqBHhIUrE8Xgc4+PjGBgYQCqVQiaTUWmWb6IZIkB3svX392NxcbHsxj5WyEW3oDJyQMvusyxbVT9CSFAvCtR5wlMJKYkD2iyVEFryGs/zYlqz09pAcqHXlJba8Pl8uPbaazE1NQW73S4SW4fDAa/Xi5/+9Kf4wAc+oNnf7xEeFUBHSJR2X8kF3ZGkBio1dWKxmGZPR61GeAiRcDqdWFpaEuepFYlQStBogUPadb1yzB7h6RykvKXo+hFaUK+yFqjZe1lvxEEuuuE6lUpr+nw+mM1m8DyPaDSKXC4nmqXq3fNNy6JlPZG+Uqkkugqk02nkcjnk83kUCgUYDAacc845YrpMi+PRIzwtIpvNwu/3g+M4jI2NlZmlqQm1UlqCICCRSCAQCJRp6iSTyY5aQEiBuMXncjlJJWKtu78aga7TakbgsNPoNhKm9lzpKBCJbNK1QPSmSafB9LppqgFyjLvt+xFpDbvdLmmWWun5pje1by18wMiDuJ4Iz/T0NPbs2SN6tZVKJZEEeTwezfZPgs6f6S5ALVuAYDCIVCqFsbExJJNJTW8cNQgP0dThOK4qEtEOYUO5oIkEcYtvp0+XHCKQy+Xg9XrBMIysVvNuIxd6hdYbsZSgHqkFquwiamSr0M3nu9sID1BNGmizVIJaat+Vnm/tfnDRouCaXH96IjwAxNo5qfRxqVTSdB/tER6FEAQBsVgMoVAI/f39Yvoqk8m0RSenGfA8j3A4jLW1tZqaOnrx6qKJhFyfLrVRb1z6WI6NjcHj8cjaHPRKePQ4Jz1BatOka4GIrQJJnZBNk7Tsdhtx6NY0HCCPNNTTBiJil6TDr521XVr5aAH6IDzk+/3mN7/B7t27sXXrVpHcOBwO9PX1YWhoCA6HA2azGU6nU7RcWlxcVG0ePcKjAMQ0E1ivXKdbKdslDKgUqVQKfr8fNpsNi4uLNaXdtUoPkbEbbay0Zo1cItFuwkO8zywWS91jqWTMTqJbN7ZOQ6qLSEpLxmg0wmAwIJ1Oy9aS6TS6nfA0oxBdTxuoXWapPM+rbrtBmlz0cN2R8zI/Pw+DwYDHHnsM8/PzGBkZQT6fRyAQEPW0HA6HWID+D//wD7jxxhtVm0eP8MgAx3Hw+Xx1TTO16qKix1fq1xUIBJDNZsvao+uNr9WG3GjuqVQKPp8PDoejrCi5EdpVw0MXeE9MTIidIkqhN8LTgzqgo0ADAwMA1q+ZcDiMQqFQpiXTSXNNOehmwqNWlKRWbRcd1cvn86qeTy0jPHqoUSLI5/NYXFzEeeedh0svvVR0WV9ZWcHXv/51JJNJfO5zn0N/fz8SiUSVC3ur0M+R0DESiQQA1LUtkOun1SzkEp7K+pepqSlZN5LWXl1Sm71SUlaJdtTwENuKvr6+lrrv1N5ECAlbW1srK6pV2onSI2HawGg0wmw2w2g0YmRkpK65Jr1pdnpz6mbCo+Xca2kDNTJLlWt5opVTOqBNt5NSEI2kRx99FJlMBldeeSX6+/vFOc7OzuILX/gCrrvuOhw6dAg7d+7U5Jj0CI8MDA0NNXyqV7ttXGp8Lf26tJx/5dikDioYDIpt/M3clFqmtDiOw9GjR5HL5VSxrVBzriQi5nQ6sXHjxrLOItKJQrfi1gq/d+vG1k0gx7ieuSbLsmIBLa0oTNri23meeoRHHmqdT1LgrtQsVasID+le6zTIHLLZrCgRAZSTMYvFgng8DpZlAbSuEi2FHuGRAbUMRFtBvfErNXWkUm6N0K4anlwuB5/PB0EQWlah1oLwCIKAdDqNbDaL4eHhmh1iSqHGXDmOg9/vF1WcXS6XqDhbufASAiRVVGu321WvF+ihGo3Ot1QBLVEUZlkWiUSiKmpgt9s1jQJ1M+HRorVbCYxGY13Lk3pmqVoRHj1Ed4A3ic2ZZ56Jr371q7j77rtxySWXoL+/XzxO9957L4rFIk488UQA2jyQ9QiPSmhXSotekCo1dVpJuWhdw0OnYEZHRzE4ONjyBa02SSNt+2STGRsbU23sVglPIpGoUnGuNZ7RaJRVVGswGGCxWERdEr0sjscSlFzjUorCdBt1IpFAMBgUa0zq+Uo1i25OcerNS6ue5UnlvUiU2klxtBqkVk+Eh1yf73nPe/Dyyy/jv//7v/Hkk09iw4YNMJvN2LdvH44cOYJ/+Zd/wVve8hYAPeFBXaMdbuZk02QYpq6mTjPQcv75fB6pVAoul0txd1M90MSvlQVfEAREIhFEIhGMjIzAZrOJ4Wg10cxmUiqV4PP5WkqtSRXVEh0pnud7Ngs6Rq02ahI1oH2lKo1Sm0G3RngEQeiKudcySz169CgYhpFMbTarDaQnwkNgsVjw+c9/Hps2bcLTTz+NUCiEfD6PqakpXH311bjgggs0PYc9wiMDekhpkb9RKpUQi8Xqauo0O7ba8y+VSvD7/Uin07BarZidnVV1/EoS2AxI3ZPJZMLGjRthsViQzWZVf9JVOj9BEMTuHo/HUze11sz3N5lMsFgsMBqNGBwcrGrFrTTbJPUkeltA9QwtIg50G3WlrxRdO0JIbjNienonDVLoVoVo4M0uqsHBQZjNZllmqXIeSPSmsgysRyw9Hg927tyJnTt3wuv1YnBwsIwAaoke4VEJ7SA8AHD48GHY7XZVIyWAuukherMeGBjA5OQkYrGYKmNXotlUEc/zCAaDYusjbfSqRW2QkjELhQJ8Ph9KpRLm5+drLgY04Wt1TpWtuJVmm6FQqGzRpaNAPXQWUr5SUmJ6le7iUh1E3RAlkYLe0llKQUdj5JqlAqjqCqMJjh4jPEajEaurq3j00Udx8OBBlEolXHPNNThy5AgmJyexZcsWTf9+j/CoBC1reIinFMdxGB0d1cRvRK0annw+D6/XC0EQxM1aSxVqQjSVPMnQXU5SUgNaER6g/oZS6c01MjLSsc2Hrj+gzTbpp06pSIKa9STdjk6Rh1pieuTcpVIpccOsrAXqZsLTjfMG3vS8qkdOKkktAJHUVjYn/OY3v0EikcD8/DxOOukk2fN4+umncdlll0n+bnp6Gr/97W+xe/du3HvvvVW/f/XVV2XVHR05cgT/9m//huXlZSwuLmLv3r342Mc+hscffxx/+MMfsGfPHmzevFn2nJWiR3hkoFMpLUEQEI1GEQ6HMTg4KG4sWqDV+dOWC5VFyVqqDCsZm6TYGun+aC1oKHU90USx1e41rSA3klDZEt9pbZkepM+dVAeRyWQSC2i7qY6r0x1araDZdFzlAwlpTnC5XPjTn/6En/70p8jn85icnMTJJ5+Mk08+GX/3d3+HTZs2SY53yimn4I9//GPZa/v378enP/1pfOYznwEA7Nu3Dzt37sTVV19d9j659/g3v/lNxGIx3Hnnndi+fTvOPPNMWCwWXH755XjxxRfx3e9+F1//+tc1S8X1ViKZaLSxSnVRtQJaU4eYU2azWc2iSK0QnnQ6DZ/PB6vVKplq64SoIQ26m83tdjfU/dGSoFWOW1kwrVZNlhy0+j3rRRKktGXoepJu3ZyUQM9Rh1odRLFYDKlUSrRUEAShKoKntzQJ0N0pLbXmTqKtl1xyCS655BIcOXIEoVAIoVAIL730Eh5++GE8/PDDeOyxxyQ/b7FYMDIyIv5cLBZx++2344ILLsDOnTsBrBOgc889t+x9ciEIAn73u99h165dOOuss0THdJvNhk2bNuGzn/0s7rjjDk3rjnqERyUwDNNUeqUStKZOZW1JJ+0fpFAqlRAIBJBOpzE5OSkunJXoZISHroep9D+rBS0VnGkQo1SDwSAWTHc7pCIJtLZMPB5HqVQq6yrqFp+pYx3EwdpqtWJiYgJAedpESkeGaDp1mtjpmVw2gla1NgzDYNOmTTjvvPPw0Y9+VPHn/+d//gd+vx979uwBsK467/f7mzbzJIKCpFM0Ho8DgPiA7HQ6kc1mmxpbLnqER0W0QnjoKATtwl45vlYRHiWkhBQlB4NBMWJS7zt3womdTgcODw9jeHhY9oKopYIzydc347jejainLcOyLGKxGAKBAEwmU5XbeLcfk27UtKkkDlJ1XLSOTDQaBc/zVbVA7Saw3ZzS0orwtDIuy7K45557cNlll4l6ZPv37wcA/PznP8cNN9yAYrGIM844A9dddx1GR0dlzWfLli3Ys2cP3vrWt4rp7omJCbAsiyeeeEKs39GKwPYIj0zI2QSb3djlauq0gzg0utCa0f9pd4SHuNobDAYxHdjqmGqAYRhks1kEg8GmHNe1mE87OgsrIaUtIyXGVplK6UZ02yYsZ42T0pGpp+zdDgLbzSktrchaK9mGRx99FPl8vqyImRAel8uFu+66C+FwGHfeeScuvfRSPPLIIw2j5y6XC1dccQX++Z//GZ/73OcwMzODUqmERx99FG+88QaefPJJ3HLLLQC0u296hEdFKCUk9JO+nPoNLVNajTRtiEBdNBpVXGvSrhoenucRCoUQi8UwPj7elMVG5Zhqged58DwPr9crOq5322aoFWoJI9Lu1LlcDoIgIBQKweFwdEVBbTemWZrVdOrr66sqnmVZtorANvJ3a+e89QKtyFqrhOeCCy7A4OCg+Noll1yC973vfWKkdsuWLdi0aRPe+c534oknnsBFF13UcNx3vetd+NrXvoYf/OAHeOyxxzA2NoY777wTw8PDuPrqq3HhhRdqawKryajHKZS0ppPWaCWaOlqmtMj4UmHQTCYDn88Hi8XSVK0JIRBaXMgkSkEKp202m2qRE7XmS+YmCALm5uZaNiI9HiDlTn3gwAFYLJauK6jtNqhh+VIZlSMEVioKRM5hK1GgbiY8WqW0BEFoivCsra3hxRdfxJVXXln2OsMwVSbaxLvR7/c3HPfo0aN48skn8aEPfQhnnnkm9u3bh0Qigf7+fmzbtg0ul0tz7aAe4ZEJtVrTiaZOLper2xpda/xSqST7/UpRGdngOA6BQACpVAoTExPo7+9vOmLSqiJyvbGj0ajYflmrcFrpmGrMlz5+k5OTYq2KmlBLeFDvIOeBrneiW+Lpglq6GFpKXK9d6JZjS0Oz2gkJAkvSmLlcDrFYrCwKpLSYXY8ie3KhxdzJA2YzhOf5558HwzB461vfWvb6rl27sHfvXjzyyCPiaysrK4jFYnULmcn3e/XVV7Fnzx5ceOGFGB8fx5lnnln2vlKppLmERY/wqIh6hKdSU6cZF24tU1pkfFLH06iAWinU3lyJVkg6nYbD4VBljjRanW8qlYLX6y0zdQ0Gg125CeoVtQpqWZatEtdrxmJBDXRb1KFdkRKpKBDHcWItEFFqlytp0O0RHrXnTvahZtbEv/71r5iZmamqz7zwwgvx4IMPYteuXfj4xz+OUCiE2267Ddu3b8c555xTczzy3fr7+zE2NoZAIIDx8fGq97VDr6tHeFRErZSTlKaOmuOrBYPBgHw+D7/fj2KxqIopKT22Wt4uxWIRPp8PhUIBTqcTbrdb9a6QZgkPLW44PT0tPtG2MqaW0Nt8WkFlQW09i4XKKJAW6MZj20niYDQaq6JAhUKhjAQR3ZZKYctuJjxa1PCQfaKZdTEcDlelrgDg9NNPx/e+9z18+9vfxsUXXwyLxYLzzz8f119/fd35k3W/v78fPM9j165deN/73ofR0VG43W7Y7XY4nU6YzWaMjIyoEqWvhR7hkQk5N5PRaCyL8NApjUpNnWagZfEvUV5dXV3FyMiIojZuOVBjs6etFwYHBzEzM4NAIKCpKrKSuTUSN9Qb4enWDUIuGgkjJpNJBINBkSiRTVRNYcRuO8Z66naiJQ0IaEkDWtiSYRiYTCbkcrmuE7bUIqXVSoTn1ltvrfm7s88+G2effbai8cgcDh48iNXVVaRSKbzyyitwu93iubVYLIhGo7jqqqtwxRVXgOM4TaQNeoRHRdApIaJTo1ZKiB5fbWSzWdHWYGJiAh6PR/W/0ercc7mcWPhLWy9orZkjB3TEqVFUTE+E53hELXsMlmXBsiwSiYQYRTgehRH1HimRkjQoFAqIRCKiECodBdI6iqcGtHA1b4XwqIl9+/ZBEARs2bIFp512Gm6//XY4nU5kMhlEo1Fks1lks1kxs0BqgXpdWl0Ag8EAlmWxvLysSKdGLiojSK2iUtU5Ho9rdoM0S0zodvhKj65Wxm0EuZYVsVgMwWBQjDg1sqzQG453AkZHgaSEEelaEpoAyeko0jt5kEK3zZlEgUix+uDgoHj+iMktieJV1gLpJZLF87zqhIzsE50+lz/4wQ9gt9txww03YGZmBjMzMwgGg3A4HHUbdrQ6Nz3CIxONLhzSGp3JZDA+Pl61Mas1BzUIDyn49fv9ZUW1qVSq7YrI9UAiT/Xa4TtFeAqFArxeLziOw/zfXOFbHbPd6PRiqFfUiiKQWhK6o4huie/007Qa0NP1qQQ0aZA6f7Vquehz2CkRUC3SiCRN1ul7fHV1FbOzs2L9qdFoxNe//nXMz8/jyiuvrJIr0ZqE9giPCiCaOiaTCU6nE0NDQ5r8HTVSWoVCAX6/H4VCATMzM2WaMFoq7yrZ7OnIU6N2eK3mXGu+erSs6EFbSNWSSAkjms3msihQN9oddFuEh6AeaahXy0WiQKFQqKxzjNRytSMKpFUNjx4iWLlcTiyRINfV66+/jpGREVH+o53oER6ZkDoxlZo6ABCNRjWbQyspLXqjHhoakky/6MGclESenE4nFhcXG4Z6tepckyInxOyTYRhdWVY0C73Np5tQS1eGqAtHo1GUSiVEIhE4nU6RCOlhE2qEbiU8SuZdq5aLkKBQKIRCodAWXSetCI8eIo7pdLpKaJXjOE3qROWgR3iaQC1NnUwmo6k3UbPRDLlt8e2ygJACaedmWRZTU1Nl7dytjNss6HEFQUA4HK5ZR6QEas6VZVl4vV4Ab2rNHE8FtnqClK7MkSNH4HQ6xTo0vQkjSqFbIzytRtMaRYHS6bSo66S2uvexHOFhWRZ2ux0cx4HjOFEp3W63i+ks4E2SrfW11yM8ClGPPCixlmgGJAIjd1HiOA6hUAiJREKUAG/k1dUJV3NSGOrxeDA1NaXoRtU6pUVIhclkaspWo3JMNUB7ho2OjoJhmCqxNpoA6d1z6lgFwzBwOBxifRctjEhvoJUt8Z3cqLqV8GhRByMVBaJTmTSJrawFUnIMtTjmeiI8U1NTMBqN4oNYsVjE1NRUL6WlZwiCAK/XW1dTR0vCAKwvoHIF/JLJJHw+H1wul6zUELA+/2KxqNZ0yyAViaGd1+UW/soZVy2sra2JRejNGpHSUGOupJDbarViaWkJwDqxlRJrI6kVnufL1IbJU2kvpaUtKjcyKWFE2mmcpFHaJYwoZ87dgnbMm2GYmureJApEe7zRJKiRMN+xmtIqFou466678Nhjj4lddNFoFL/+9a8RDofF4vK+vj7YbDZs3ry556WlB5CQZz1NHa0JD5lHvYuZFCXn8/kqpV85Y7ejhkcQBEQiEUQiEcXO61Ljqj3nTCaDbDarqhEp0NrxJVGdeDxeVshd6a1GF9jSzuNEZyYcDoubKolItsPDpodq0BtorWJarYURK9GthKdTBeI0iSV1KXQtEIkCVRa001EgrQiPHu7p0047DbFYDAcPHkShUECxWMTc3Bz+8Ic/4Le//S14nhfPHcuyePbZZ1WVcqlE549Il4BhGIyOjtbdsNpBeOqlhhoVJTc7thqoTBEZjcaWU0T0uGqA7g6z2WwYHh5WtVW12bnSUR250ToaJpMJfX19ZU+luVwOiUQCuVwOy8vLZZ5FrTpX97COZshDPWFEcs5q2St0as56gJ4UoiujQIIgiASIjrqS88dxnOoPbXqJ8Hz3u98Fz/MolUriv0KhgHw+L0bGiPgg8UXUEj3CoyLIDadl/lSqU4tlWfh8vqa7hwi0JjzJZBKRSES1FBEZV405E2kB0h3m9/tVX4SUEh6e5xEMBpFIJDAxMSHpbwM036FCnq4mJydRKBTE2gQpnRm73a6bDeV4Qj1hRFK3RewVlAojSqGbCY9e580wTFkqEyiXNQCA5eVlmM3msnuulYcOvRAeQsRbfbBVCz3CozIIadBqc6A3eFKUHI/HVSERWhGedDqNtbU1mEwmLC0tqRpqbTWlxXEc/H4/MpkMJicnxaeyTte3ZDIZeL1e2O32pqI6SkCnwcimKqUzQxdn6rHDSG/Q6vpRIoxIb6ByNkA9E4d66DbNIyJr4HA4EI/Hsbi4WBbxWFtba/ocAvqKeOkJPcKjAHI2Qa3TWmR8OiKhFolQe5Mn3jaZTEasOVF7425lzqSwu7+/H4uLi2WLiRaER86YcqM6WqNSZ6ZehxH5122mje1AO46HlDAix3EiAaKFEWnCKtVN1K1F7N26wdOKyITc0LV3JJJHzqHJZJIVydNLhEdv6BEelaGVEB4BwzAIhULgOE6RXo0cqEXWKp3DFxcXkUgkxPCtmmiGmJRKJfh8PuRyuSq16VbGbYRGY7YzqiNnPjSkOoxouf52mW72CJU8GI1GSWHEyggCfa5oQ95uQzdHpmoRNSlxS6lIHt3Vl0wmMTo6qpjwHDp0CO95z3uqXt+1axc+/OEP47XXXsPtt9+Ol19+GQMDA7j00ktxxRVXNPelO4ge4VEAOTeU2gafBIIgiG3SLpcLCwsLmuhOqGFd4fP5UCwWy8xTtYp8KanhqdT8IYKRtcZtF+GhozqTk5Nix44c8DwvEux2PuFKCbXRppuxWAx+v19WVEEOujHyoKdNuFYEgWyepJuIND84nc6mNGU6ASXaZHqDknnXiuSRe87v9+Oyyy4DAGzYsAGnnXYazjrrLJx88skYGxurO/a+ffvgcrnwq1/9quz1vr4+rK2t4ROf+AQuuOAC3HzzzfjLX/6Cm2++GX19fdi5c6fCb9xZ9AiPytBiY6ctDfr6+uB0OjXZ3FqZOyFkoVAIQ0NDolcKgVY1MXJreAgRK5VKsjR/tJpv5ZjpdBper1e2lQY9Ds/z4oJJlEzJ3Mn10U73dqnakkq7BaJRQqfBujEVIQd6J2lS3XuHDh2C2WyWVBbW6/ki90A3Ep5W6z3pe254eBiPPfYYnn/+efz5z6ZXqMIAACAASURBVH/G/v378cgjjyCbzWJychLXXHMNPvjBD0qOs3//fmzcuBEjIyNVv/v+978Ps9mMm2++WRRgPXLkCO69994e4TneoSbhoRV1x8bG4PF4EAwGdWHwSUOOx5SWEZ56c64kYsPDw7IWGK0jPHQLvNKojiAIYiur0WiEyWQSyQ/5LyE/HMeJmwFNhOixtIKU3QKdBtOD0J7W6KZNmFwbHo8HRqOxSlmYPl96cBkn6NboDqB+R6/H48G5556LmZkZbN68GXa7HQcOHMBLL72EDRs21Pzcvn37sHHjRsnf7d27F6effnrZfbljxw7cfffdCAaDDaNHesKxs7K0AXJuKrVqeEhRssPhKBO/08pKAVBOSnieRzgcxtramkjIah2jdnheVSKfz8Pr9UIQBCwsLJRtvK2M2yzIuaOjOvWELCtBCA3pSKEXSvL/ZCy6k498jv6ZbGjthpRSLV0HFAwGJTWBemgPaPJQS1m4lss4reTdTgLSbR1aNLQSHQQg2jls2bIFW7ZsqfuZ/fv3Y25uDh/96Edx9OhRzM/P43Of+xze/va3IxgMYnFxsez9o6OjAAC/398jPMczWq3hISaaJAxJFhp6fK3sHwjhkfPERAps5aoRt7OGRw0lZy2IpSAISKVSiMfjmJqaqjq3jT5LyIqc8L1USouO/hDiRYqP5abB1IaU0F6tFmsAogK23tIqUujGyEOjOddzGa8sXtdCGLHWnLvhepCCFtcIeeCW+yCVzWaxurqKwcFBXHfddXA6nfjZz36GT33qU9izZ48oSUGD/JzP51Wdu9boER6V0ezGLggCYrEYgsEgPB4PlpaWJG9ircUBSWSj1k3IcRwCgQBSqZSiVEy7anhIes1gMLSk5Kz2fNPpNKLRqKhFpEZURynIZ4lcAABMTEyI0R46DUbe3+rfVAqpwkxSXJvNZmu6jnc6rXIsoJnrvV7xeqUwIk2A1JQw6EZiSaB1hEcOHA4HnnvuOZjNZnG93Lp1Kw4ePIj77rsPNpsNhUKh7DPkZ62VkdVGj/AogFYprVwuB5/PJyv1omVKi4xfa+EjujV9fX2KNm1A2wgPsL7IRiIRWek1ueOqQXgIQUyn03C73WI6SQ6URnXkjBeLxRCJRDA8PCx5jEjXV600WLsJEPBmey4AzM7OlmkCpVIphMNhUc22HX5TctFtG7Fa11ktYUQSBYrH4yiVSlW1W81KGPRSWtVjAsoitVLSHJs2bcLvfvc7zMzMIBQKlf2O/Dw+Pt7CTNuPHuFRGUajsYoN1wJdAzM6OorBwcGGN65Wbe8EUm7sxWIRfr+/rm5NI2itXHzo0CFYLBbVzD7VmC+pwyKO9XK1iNSM6hAQU1lBEDA/P18z8mUwGGqmwci8lHaDqY1amkBSflO0S3xPiK0+tLo/pZS8KyUMAoEATCZTU9YK3ZzS0oLwcBwnRmfl4IUXXsDll1+OH/zgB9i+fbv4+iuvvIKlpSVs27YNDzzwQJnJ8FNPPYX5+XnJri49o0d4VIbcSEY6nYbP51PsyN0uJWegPM02ODhYV7dGybhqgejXAMDQ0FDLUR0arRAe2q6CFoeUM6aWUZ2hoSFZpJqGVDF0rW4w8v5OpMGk/KaklIZpAqS1xky3RXiA9nWV1ZIwkLJWoIuhpUhrNx5nAi1czZUej61bt2J6eho33ngj/uM//gMDAwN48MEH8cILL+Chhx7C2NgY7rvvPnzpS1/Cpz/9abzyyiv4/ve/j5tuuknVebcDPcKjAHJTWvU2dtpuYWJiQlE7spzxWwUZv5UOJyloURPj8/nEUHhfX5+qi16zHl2pVAper1exXYUWUZ1isQifzwee5zE3N9e0qSyNSgIE6DMNVktpmFhjRCIRCIJQlQbr1kiBGugkcahnrUD7uUmR1m4mPFpEp5SqLJvNZtx3333YvXs3rr76aiSTSZx00knYs2cPTjzxRADA/fffj9tuuw0XX3wxRkZGcN111+Hv//7vVZ13O9AjPCqjVg0PrfI7MDBQtRkqGV9LwgMAa2trokS50ohALSjpAKsHqaLpffv2dcT3qnJeJKozPT0taflRa0wtojrxeBzhcBiDg4NNdakpQTekwegN1ePxiBozLMuKtUC0JlCr3UV6Fx2Ugt6Ig5S1AimGpkmryWSCIAhd1cFHoGVKSwnGxsZwxx131Pz9tm3b8MMf/rDVqXUcPcKjMqRqbOhoiRyV33rQkvBks1nk83kwDNNSh5MUyELayqJKoieVRdPttIGQQj0T0krQY2oV1fH7/SiVSpidnW05MtcMuiUNRjRmSJSV53kxDUZ3F8kxa6z3d7oFeiM8lSCF6Xa7HR6PB8D69b62tgaWZWt28JlMJt1+L62KlruJ9LUTPcKjAEpTWs0UJSsZXy1wHIdQKIREIgGr1Yrh4WFNxN6aTRPR2kRS0RMtjomcbjg6PVkrqlM5JoEWUZ1EIoFQKASPx4Ph4WHdLPLdkgYzGAyS3UUkCkQ0gSoNN6XmqXfyIIVunLPZbBZTtWNjY5IdfADKonZ6Sl1qQU6UdIIeb+gRHoVo9ORPNl8izGe1WlWNlqi9uZNOIuLlRDp5tIDSSIwgCEgmk/D7/XC73TW1iToR4SFRnXrzqjUmEf5Ta1MnhJAYtnYiqqMUtdJgJOJFEyIlHSdqgu4uqme4SSIKerFaaBbdSHiActJQq4OP1AIlk0ld2Zloccx7EZ7a6BEelcHzPEqlElZWVhR7JMkB2TRbvVHoqEllJ5EerCtIwW2hUChzXZdCOwkPOW4syzbVop/P5xGLxeBwOFomwYQQBoNBDAwMYGpqqmsXukZpsGQyCYPBIKqMdyINBkgbbtJWC8FgEAaDATabTaw50YMmkBx0K+GpN28pYUTazoQ+Z53QcdIqpdWtpFtr9AiPSiCFoqRNWonztRKQRV5pJT4BXTwtpejcbNpJDuSQqcpW+JmZmYYLghYkTYrwJBIJMdq0uLgoe6EiG7fFYoHH40Emk0EkEgGwHmp3OBziYit3kSXptHw+j5mZmZbqwvQI+tgGAgFRA4qoQuspDSZltZDJZJBOpxEIBFAqlaqsFvSYcjgWCY8Uap0zkrqs1HFqVRixHno1PO1Fj/AohNRGmM/n4fP5wHEcZmdnsby8rHlXTDOEp1AowOv1guO4msXTWhZFNyJTcubXzLjNgD7PdFSnUbSpEnStjtFoxNDQkPg6WWSz2Szi8TiKxaIYjickSGrhSiaTCAQCcLvdmJycPGYXt0wmIyp7LywsVH3PRmmwTnWDWSwWGI1GRKNRzM/Pi5pAtdqr9V5Yq3e0qmVTS8eJVoYOBAItF7DXmnun29KPJ/QITwvgeR6RSATRaLTMpLKVCIwcKCUlgiAgGo0iHA5jeHi4bkGr1l5dtdqy5c5PybitgIxJojpKU0aNOrBqLbLZbBYsyyIcDovpEEKArFYrIpEIcrkcpqenu87HRi54nkcoFEIqlcLExETNYnA9d4PRUQcpTSC6vVqqsLbdjuOVc+4maKFlU8seQ8rUtpXInRbHvFe0XBs9wtMkyNOnxWKpKko2Go3gOE6zPKoSUsKyLLxeL4xGIzZs2NBQfK7dNTzE7JNhGFnzk4IWhIfjOBSLRYRCoZaiOko6sIiAYmV9SDabRTQaRS6Xg8FggMvlQj6fh9FoVOUpU09gWVZUIN+wYYNivzZAXjcYeX8n0mCV7dWVmkBSjuNapVRodDPh0XredAE7QT1hRNrUttbc1KjFlEIvwlMbPcKjEBzHwefziU+f/f39VRdsO+0faoE8JcdiMYyPj2NgYEB2Wz0pDFUbNJkSBAHhcBjRaLTlln01SRopkPX5fACAjRs3qhbVUQqDwQCr1Yp4PA6O4zAzMwOz2SxGgaLRKHieL0uDdZvwGgGJ8hHzVxL1ahVKu8HUSoMp2cikNIGkUiq011SjzVTrOesJnTIPlRJGJC3x5CGF3J9SMgYkndUjPO1Dj/AoRCgUAsMwdd3CO014aNuFpaUlRfntdtTwkKiTyWRSpWVfrRoeItqXz+cxPT2N1dVVRWRHTV0dYP08+v1+uFwubNiwQZyL1WotE14jC2wwGEQ+n4fNZhNrgBwOh+4Xv0KhAJ/PB4PBgIWFBU07TPScBqPRyGuqkuySaFAr8+xWwqNFSqsZ0GreBKQlvlLGgDZHVfO4k2tZ7/d8p9AjPAoxMTHRkBB0ivDQ9gbN+HTVG1stxGIxsCyrKOrUCK2mtOhaHY/Hg+npaTHcLOezaqslEyHIdDqNycnJuq3vtSIDxIDR5/OJYXZCgrQ2zZQL2gJjeHhYVfNXuaiVBqP/AcoJkNrkQcpritaXCYfDLevLdDPh0eu8yf1Jp6lJFCiTyYDjOBw6dEg1TzeyZumBAOoRPcKjEHIIAanhadccaIE+OfYG9VCLPNDWEM2AtOlaLBZF7vBy0ArhofV+6M4w8sRfD1pEdTKZDPx+PxwOh+IaFkA6MkA2xVQqhWAwKNaQkNbcTujEkGgax3GqGZuqhWZEETsRBZLaTCv1ZYxGY1karF7Nl56JQz10KqXVDGhhRIfDgUAggKmpKbEYOhQKicSVToPJXS/JvtApIUW9o3dUNEA7IzxKBPqUjg0AEARwySTywTDAc7COjcI4MADIXGA4jkMwGEQymYTD4YDL5VI9ZdFMDQ+tRySl91OPRGkR1ZHbmaQUdIHs4OCg2A5P6oBisRhKpZKsdni1QNrq9WaBUQuN0mBSUaBObMJS+jKFQkEkQXRnER1RIN+vWwmPXlJaSkFST1KeblJilnKEEcm12I3Hox3oER4N0A7Cw3EcotEoQqEQhoaGZAn0yR2bzJ1hGORWvQj/7/9DKZMBABhtNgy/7SzYFzc2HKvStoK4G6sNcjzkgpDEYrFYU++Hjmhp6YEFrJu2+ny+pqM6SkC3w1faJVS2w9N1QGo8MRLym81mu1osUU4aLPO3+6VYLHa0G4x0FpEicLqzKBqNltWUEJLWbcSn2+ZLUEuDp54wYi6Xk+ziIy3xSiM86XQad911F5544gnEYjEsLCzg85//PM4//3wAwO7du3HvvfdWfe7VV1/tyihS9824w5Db6aRlSovnecTjcVgsFiwsLKjqnUQTHj6bxdrevSLZAQAul8Pa3r2YGBuF4W+h9ErQtUSTk5NiyF2rlne5Ka1GUZ1a45L/ahHVCYfDSCaTGB8fF4+T1By0RC27hGw2K9Y2GY3GMgKktB2epOqcTmdZAfaxApIG4zgOkUgE2WwWExMT4iaklzRYZWcRXVOSTCbF6B+dBtO7NUY3pbRoyCVq9YQRc7kc4vE4/vznP+Mb3/gGNmzYgMXFRcRiMZxyyimi0GktfPGLX8S+ffuwa9cuTE1N4Ze//CWuuuoq7NmzB2eddRb27duHnTt34uqrry77XDeSHaBHeDSB0WhEqVRSfVwidEi8mObn51W/0WlSwufzKETXqt5TTKbAZVlJwkNauqVqibQignJIQbFYhNfrRalUkq3iTMbVIqpD680sLCxULSAcOCQLaaQKGfRZnOizuGBCezovpJ4wa7Xb0rYYUps3TerUTNXpEaT7UIrUSdUBAZ3vBqNrSsj83G63GAVqp81Cs+jmlFaz866s1fvgBz+IwcFBPPfcc3j55Zdx3XXXIZfLYXZ2Fqeeeiquv/56DA8Pl40RDofx+OOP45577sHb3vY2AMCVV16JP//5z/jxj3+Ms846C/v378e5556LkZGR1r6sTtAjPBpAi5RWNpuF1+uFxWLB6OgoCoWCJk81dIs3YzDAYLWCy2bL3sOYTDCYyy+dUqkEn88neh5JdRZpGeGpNS7tzTU0NISRkRFF+iilUknVzYiQ1ng8jvHxcclOuoJQxAuBV7E/cggcz8HIGLA0vAGnTmyFhWm/KSDdISTVDp9IJFAoFMR2eEKCSOqQRCK79amwEWgNoVrnVKtuMDVBiEOl2SatCRSLxeD3+xUJ7LVj3t0Y4VHTVsJsNuP888/H6aefjkgkgpNOOgn79u3Diy++iH379iGfz1d9xm6347/+679w6qmnlr3OMAwSiYTYCLO4uKjKHPWAY3MF0hByU1pqbex00S9pNU8kEsjlcqqMXwl67gaXC30bNyD+8itl73HOzsL4t+hOpRnp9PR0zZtYC8+reuMSfZdSqaQ49UcWUdIxRUT9WllYWZaF3++HxWLBhg0bJAkAwzDwp0J4LXRAfI0TeLwefgPjfaNY6J/WPMUlB1Lt8KQOKBqNgmVZCIIgFqprpSrbaZDIIcMwijWE5HaDARAji1qSoFqRklqaQHTETxCEqjRYu8hat15XWvloGQwGmM1mbN26FVu3bq35XpfLhXe84x1lr7344ot46qmn8OUvfxn79+8HAPz85z/HDTfcgGKxiDPOOAPXXXcdRkdHVZ13u9AjPBpArdQNYdgul6tM6FBrcUCy8DIGA9xveQuMNhvShw9D4AU4ZqbRv/UkwGwuIxRy0kTtquFpJapDbzizs7NiLUs8Hhe7mUgEQ243kyAIYipybGxMUp37zS8DHI6vSP7qcGwFC+5poPN8pwrEL8pisSCTycBms2FoaAiFQgHpdFoU7KTrgPReG9IIpNtscHBQ9NFrBfW6wdoRBVJSUyIlsCfVWt2sJpCSOfcIjzpjHjx4EFdddRVOPvlkfOQjH8HDDz8MYJ0Y3XXXXQiHw7jzzjtx6aWX4pFHHunKpoMe4VEIOTcWXS3fDIhGSS6Xw9TUVFXdg9YGn8Cbix9jt6HvlLfAtXnTuimd0wn+b1GddCaN/kE3PH3ybSu0iE7QhIc4rvM831RUh67VkepwIU+1pJuJLOYkClRZ25DL5UTxPzkRAAYMrEZpTRqryQIGDAQdMh4i3ki6BiutQvTQDq8WSNSVZVlNu82UpsEI+WlFtK5Z4lCrtZrUARFNILXdxtWsq2s3tKg9apbwPPvss7jqqqswOTmJe+65B2azGZdccgne9773ievfli1bsGnTJrzzne/EE088gYsuukjVubcDPcKjAZolJHLTQ+1oe68kJozDAQZALp/HWmwNrKmI5cIq8uE8JtgxbBycQ5+5tiIwUB6WB5oXMSTIlFhwAgfGbIDRaBTb9JU6rtNP0UBtDQupbiZCgGKxWJmqsd1uRz6fRyKRwOjoKNxut6z58DyPpaF5HIouo8i/WfhuMpiwaWhe0/PeLEqlEvx+P4rFImZnZyVJZr12eJpAatEOryZIYbLD4cDCwkJHiozrpcE4jkOpVCq71+RGgdSMlNTSBJJyG6ftMZQez26N7gDaRXiUFpT/7Gc/w5e+9CWcccYZuOuuu8QHbIZhqjztxsbGMDAwAL/fr9qc2wl9rSZdgkZdQc0Qknw+D5/PB57nG6aH2iVsSN84JC2TTCZhHLTg/1vZiwK/bjLqT4XgSwVxzsJZcBhrR1SMFiM4B/Cs7yWYjSbMuCfhsbrBQNmCVRJKOLC2jNcjbyBXzMNptGPL4CKQY8SoDidwyBZzYBgGLrOjZhqolQ4sg8EgqWpMFI15nofJZEImk4EgCLJbfIesHrxj4Uy8HHwdqXwafRYnto5vwZB1UPbc2oVUKgW/34+BgQFMTU0pWsDrtcPH43H4/X6YTKayCFCn3OHlFCZ3AlJpMGA96kNIEE0K6qXBtCQPdMSUgBDeSp8pmgA1iohqQRraBS3mrjRq9POf/xz/+q//iosuugi333572fHetWsX9u7di0ceeUR8bWVlBbFYrGsLmXuERwMosZYgRCISiWBkZERWLUC7vbpyuRy8Xi8MBgOmZ6fx/1afEckOAJR4DqtJPyKZNcz2T0qOyYPHgfgy/nDgKTj71p8gXgnuw9nzOzDXPyU72sMwDI4m/Hhq5fl1gpHPw8f64Y8F8ZHTPwCbxYZ4IYkX/a8imInAxBgx75nBSSObYKNSRXKjOkpBNmsS1SmVSshms2JxJ8dxZREMu91edb4NDIMx5zCGF94Gg4EBBMBmsOqiWJmA+H1lMhlMT0+3rPAN1G+Hz2QyiEQiZe3wahSSywHpNgOgublpqyDXcWUUqDINRkeBSNSo3dESKcJLawKRuq96CsPdHuFRe+4cx8k2Yw4EArjxxhuxY8cOXH/99YjH4+LvzGYzLrzwQjz44IPYtWsXPv7xjyMUCuG2227D9u3bcc4556g673ahR3g0QFnhb50LulnXcK0JDykuJhoqa2trGBsbg8fjQY7LI1NYb1Mv8iWssXFkiywYMFhJ+jDRNwozU31ZpQoZ/CXwOkrUvPNcAS8F/ooRxyDsdSJDNApcEfsjB1HiSshksgAEOJ1OZFkW3mQQVo8Vfzr6LILpiPiZlwJ/hSAIOG1i23oNjAa6Ovl8Hn6/HwzDYH5+XjyXlWkcKXdzuo7FZDfjQGwZR+OrYMBgfnAWGz1zuiI7tDL0wsKCZpostdrhSR2Q3+9HoVCoqgNScz5qFyZ3AnLTYKVSSfy505pAQLnCMKkFKpVKYjE0qfnqxnMCaFPDo2TMxx9/HCzL4qmnnsLZZ59d9rtTTz0VDz74IL73ve/h29/+Ni6++GJYLBacf/75uP7667s2qtYjPE2gUUqLbKK1CA/P8wgGg0gkEhgfH5dd30HQjggPy7JYXV2tMvu0mizw2NwIZ9cQTIdF8sOAgdloxuvRg9g6vKkqTZUpssiX8kDFcYvnkihwRdmEhxd4ZHMskskU7HYbbFYbOJ4DBIDj18X6aLJDcDi+gs3DG+Ew2FSN6pCOsEgkIsvxu1Y7dzabBZtn8ZL/Bfw1uB8mkwkmkwmB1PoxPn1iGxihswu7IAgIh8OIx+OYmJiQVIbWGmazGW63u0xxlm6PzuVyMJvNZVG0ZiIy7SpM7gQq02CFQgGBQAAARIsJAOJDgdFo7Jg1hpTCMEmDra2tifIcwWCwLA3WDSSo0zU8l112GS677LK67zn77LOryFA3o0d4NAIhJZUXNPGXcjgcWFxcbKook1YAVvvG5nkexWIRgUAAk5OTVWTMAANOGtuMI0kfMkVWfH3b+Ang/6YXs+CehquigNlmssBsNFWV0jhMNhgN8m7QfD6/XkPhHEG0Pybe2AwYGBgDJvvHkCtVC2wBEJ9eefCqRXVIWz6AsqiOEpB2bpfLhVQpg6A/CofdgRJXQqFQQLbE4oX8y5h3TcFcMnasjoXuNqulIdQJ0McPeLOOKpvNlrnDK2mH73RhcjuRyWTg8/ngdrvL5Bu07AZrBZXnm6Q6LRYL0uk0wuEwAFSlwfR4DjtNeI5H6GPVOgZB6njIxlAqlRAIBKr8pZoBWXDUvrjT6bS4gU9MTIhpmEoM2zz4PxvPxrDDg1Qhg+n+CZgNJuwLvwGGMYATqqNPfRYXNgzNIRqPvfk9wGDzyEYIgoCD8aOwm23w2Pqroj2kYDQcDmNkZAQnj5wEFgWsJFbBCwKsJgtOnFvCgLUPWaMZTotDjDxBWP/8UJ8HNqMFBkbdqI5UC3az4AQORb4Ik3k9ugPreq01z3MwGI3IJDIIh8OioB/ZwLWsYxEEAWtra4hGo4q6zToFUvNRmRYhdVR0O3ylnpJeC5O1AInWJRIJScsPOWmwdooi1oPRaITH44HH44EgCGVmuKlUSlQBp+0x9EDYtajhIRG5HqTR+bPehVCitkz0SQKBANxud5W/VLNQk/BwHIdAIIBUKoXJyUnE4/GGC5fDbIfFYMaAtQ+BZBCJfAoAMNk3BrupOj1lZkw4fXI7bIIF0VICRsaAxaF5GBkjHnntV2IL9kTfGN4++1axxT2fz8Pr9QIANmzYIHZ5vH32dCQLW8AWc3CZHUhHUjAxJvRbXDh96mQ8s/IC2FIOvCBg2OHBqZPbYDaYW66FIQWsPM9jbm6urOukFZTAgcd6xC5VyMBussLIGMEIwLBzCA6rDUNTb9YBkQ2ceB1poWdDvqsgCE1HsDqNRu3woVBI7A7iOA4Gg6Fma/2xAqIObTAYZFt+yOkGIz+3Mw1WSRoYhpFMG9NGm7lcTtQEIgSoE1FTrXR4eoSnNnqERyMYDAbk83kEg0GUSiXMzs6q0slCj69GHQ9JsdFqzslksuHYfWYnZgem8MzqiyhwhfXXrC6cMrkVFoN0zYTZYMb0wCRGMAKL0QSryYJf7PttWUTInwritfABvHXiZDGqMzo6WhVFsRosGLENArb1mzyQ9KHAF3A06UeMTeDUqW3gBQF9FgcGrG44jLaWyA7RSAqHw6oXsJaEEl4KvYZIZg2bhzfiqdXnkSpwsJus6LO4cOrkVjhNDvGcSNWxkELeSj0bQoCUPNEKgoBkMolgMNjVxbq1UNkdFI/HEQwGxUjZkSNHytrhSR3QsXAMiIxAq+e1UTcYiQZpnQaTQxqkrDGIJlA2m8Xa2ho4jitLgzWjCaR03npTWj4e0CM8TaDRIkG6gHw+H0ZGRhSJ4MlFq4SHiMVls9kqNWc5YzNgsDgwj1HnEELpKExGI4Ydg+g3S7thcwKP16Jv4Lev/gFudz+GnYOwmixIFbJwmO2ghXJW435MmUdRSOdlda8xDANXnwsvBV/Hy4HXwAsCGAAuixMnjW3GhGO0JbJDlK85jtPk6T+WT+KVwD7YzFZsGtmI8ze8HYfjK7AaLVgcnMOQw1P3fBiNRklBRKJo7PP5YDKZyghQrQ2cpF4LhcIxH+mgC5NnZ2fL0mD5fB7ZbBaZTPvTiFqA53mEQiGk02nVZAQq0Yk0WDN1jLQmUGXUj2VZSU0gta0xtFCIJnWdvQhPbfQIj8ogmjWlUgkjIyMYGRnR5O+0ouZMPLrcbjeWlpaqFhy5YxsYBgOWfniG3A0JRaqYxl8Cr4Hn10PeJZ6Dy2hGlI3BarLAyBgAAWBzOTgFO5w2ByaHJ2R7+zB2I173HQQvRosYpAsZvBJ8HbPuSTiNyrtsaLsEj8ejGXH1p0IQIGDGPYn/XX4Kh9aOYMQ5BIvRDH8yhOn+Cbxzbodku3+tMaXMP7YqawAAIABJREFUHhsV8mYyGfj9fvT392NycvKYflKsV5hcyyuKbocvFotltiJqt8OriUKhIHZcaikjUIl6aTBChOifm0mDqVUHU0sEM5fLidFOum3eZrO15AWnVXQHgC7qk/SK3pFRCZWaNdlsVtOFpRnCQ2oyyNN71VNesQg+n4fTbEZB4XzyfBEFrgCTwQSHqTp9lMyn11NfDAMBQCKXxMLgDFwWJziBAzgBmWwGBsaAU+e2wdPvkR2VEQQBPHjEkjGYTCaYTSYwhvWFKFPIro+vEHLsEtSAIAhwWtbJ2IC9H4fWjqDIl+BLBeG0ODDpGoM/GQDL5WA2SUfPGoHewAcHB6sKedfW1lAqrddQ9ff3d6TdvF1otjC5UTs8y7KwWCwtt8OrDeJjJUcyQWs0mwar/AwNLepgyN+rFMGkNYFoaww6DSZ3zdeiw5Y2le1BGj3C0wQqL9RMJgOv1wubzSZq1uTz+baqIdcD7R4+ODiImZmZqpui6Pcj/uJfUEgkwJvNcG/fBng8QIObR4AAbyqAvwRfQyKXgtPiwEkjmzDnnl6P2vwNdpMVBoYBwwAQBAgQcCS2iv+z8WysxP3wRnyYHpzEtsktWPDM1CQ7DMMgU8wiW2LBwACHyQYzTOh39sPt6Ec8kwDLsmCY9Sed4b5BMEJtTSSpY0We6AYGBjA9Pa3pJiEIAsZdI3Db+gEwYj0TAwYDtn4YGAa8oK5dKF3Ia7VawbKs2OpL6s7y+TxsNltZGkyvEQy5UFMxuV47fGVEgBy/drrD8zyPQCAgpuv0mpqUkwYDIIohVqbB2qW0XE8TiGVZURPIYrGU+YOZTKaaWmy9CE/70TsyLaCyu4l+WlRiL9EM5BIe4h7OcVxNj67S2hoCTzwJjl3X1cnn8sgEArD/34thHh+vO/5aLoHfLz+1LioIIFtk8YfsM7AtWjHpHBPf12/tw7R7ErFYXHT7jqTWsKFvFjtGToZp9nRYjGY4zPa6ZMefCeHp1RcQzcQgCMB43zDOmjkdA9Y+nD67Hc+uvIiSwIPnOYAXsHV0C1LhBPxpb8MaDFK/ks/n2yo212d24fyNf4dMMYu3jJ+A1yOH4LG74TTZAQEYcQ3BZlS3Q4pYmsRiMclIB72YVwr60XVA3QKtFZPpdvihoSHJwli6HZ5cg1o8jZO0ut1u7zodIaVpMNJZ1wnUIr25XK5KE4hEgMi6oyXh6fYHEy3RIzxNIpFIiPUOpLuJhsFgQLFYrPHp1tGI8BD9lEbu4QzDIHPwsEh21l8DSvkCkq+9juHx8ZrRBYZhcCS+KpIdA8Ng0O6B2WjC0bgPE85RUXHZarDgzOlTYeXNiBZiAA8sjMxizj2FfkefOLd6aaxMMYunjj6PaDYmkqZAKoy9vpdwztyZWPIsYMg+iCOJVTAA5j0zGLS6wQhMzVZusvkUi0UEg0G43e621a8wDINUKYN9wYMIpSOwWxw4c+Y0zA9M49XQAUAAhp2D2DF9iqpeWsSo1mg01ox01ItgEJkFEvYnx7FTxp71QCId2Wy2rSS2VmEsqQOio2h0FKiVp3NaH2psbKzK6bobUS8NRmrShoaGxLVWThpMK9Ckl2gCFYtF5HI50RqD1H4ZjUaRsKlFUMh+oLd7UE/oEZ4mkMvlEAqFMDMzIxaGVqLdBp+V8yOhe1q7pha4TKb8BYaBAAEcm4PAcUCdGzJXWpd299jdWPDMIpSJIJXPYLzPghyXLxMRdJkc2GCdxoJzGiajCcMDQ4oW+HQxi0g2hvWOrjfNK4KpMLKlHPpMTozYPBhzDAH42wLwN45QWYNBujLS6TRWVlbA87yYcshms21J4WS5HH5/+M+IZNbE17xJP85ZOAubhhfB8SU4THbYW2ypJ6A3xJGREQwMDMheHGtFMGhj1E4Ye9YDXZi8YcOGjkc6TCYT+vv7xWga6aYjgoh0Nx1dByTnGHIcJ3qLqakPpUcwDINkMolIJIKJiQk4nU4x+kNIBFA7DdbOeZI0WKUmECE/hw8fhslkKkuDNSuBQKJGPcJTGz3C0wTsdjsWFxfrXlgGg6HtKS2inkpUceUqADvmZpHcv1/8mcG6Q7djdgaMyVRzsxUEAbMD01iOr2JuYAaPv/G/iGbXlZSX46tI57PYMXUyLAYLeJ4XUyhDQ0MYGx2VfWOKoWwBfyM5le7ihrLX5BBNQrTS6TTcbjeGh4fFTiZShEpMCskGrmZunGEYhDKRMrIDAMVSEa+H38A582fBKLxZp9Aq6NZ6NTZEOoJRy9iTPM3SabB2bDrdophcr5uush2eHD8pEknMXF0u1zHfXUeIXbFYrBLD1LIbTC0QTSBCxsbHx0WH+MoHBzoNJmeuPQ2exugRnibRaLM2Go2aR3jolFmzzuuCIMA2MYH+LZuR3H8A4HkwBgZ9GxfgXFhouNmOOoZw+tR2POd7eZ3sMAz6LC70WZx4I3oYGwfnMGjsx+rqqlgD4nK5FJEdslC5zA6MuoYRqjAHnXSPr9e7yATRX8lms5icnBQ3HJPJJP4/aUvNZrOIx+Pw+/1lWjatduEwDCOqU1ciW2BR4kowGtSp2yH1K1q11hMoEUQU3eFVLrBUszC53ahshyc2CSSKFo/Hy0ik3W4XX++UmWs7kcvlsLq6CqfTWZfYadENpjboaIyUBAJJg4XDYRQKBfHhi0SCpO6bHuFpjB7haQJKrCW0AhmfCIqR4lMlaQoRNis8Z+5A36ZNKMTiMLicYE1GMM7G4mQWgxmTfeN4kXkNk/3jMBlMsBrNMDAG8AKP1TUfClwWQ0NDGBgYwNGjR2UdF7pTg3xfG2PF22ZPxzPeFxFKRWBkDJgaGMdpE9tggLzvnE6n4ff70dfXVzfNIdWWWkvLhvxTUsMiCALGnMOSvxtyeGAxWtBqaxYpqs/lch1x/JYSRKxFIpWmcKSgdWFyu0FsEqTa4dPpNLxer5iKzWQyEASh64rJ5YBWOW+2NqleN1gn0mD1yAmxxqi8b1iWFTsAjUajSH4AdZSbeZ7Ht7/9bTz88MNIJpM47bTTcNNNN2Fubq7pMfWGHuFpEsSxvBbaQXgKhQLeeOONsnb4ZsFYLDCPj8EyMb7embOygkGZn7UZrfDY+5EupMXXSiUOmUwafZNOTA9Ni3OzWCwNo0Z0VIdWIxUEAR5LP86bOwsslwfAwGG2iamfeuA4TlSapaM6ctFIy0ZpDYsgCPDY3FgaXsCByGHxdbetDyeObgLTItkhLth9fX266dSpRyJJCgdA2TGU08rdqcLkToCkY1KpFDweD4aGhsSUCF1M3ql2eLVBzm0ul1O1Nqkdooj1oEQwUeq+IR2AuVwOd999N37xi19gYmICJ554Is4991yceuqpiu/773znO3jwwQfxla98BWNjY9i9ezeuuOIK/OIXvzhmasIYQa3Wj+MMhUKh7sadz+exvLyMzZs3q/63OY7DysoKMpkMpqenVe/GyOfzOHLkCDZt2iT7M950EE8e/hMKpQJybA65fB4bRmdx4ZbzkCvl8P+z9+axkaXnee/vLLVvrJX7zt571tZMzxppJI9t2bLv+AqWfe3EcSzDSGwnhhzBQWQgnkSOHCQQjCCxjRjSwEai2LBydS1ZG5zIWsYjjaZ79t735l7FpYq1L2e5f1R/p0+RRbJIVrHJHj5AA83D4ld1vjrnfO/3vs/7PLczM5S1CnFXhKg7TJd/La9idVanXdLrQkHY5/ORSCQ6Rka2B0ClUqmhE8zuym1H1ayxWFpmLpci4PTTH+zBr3q3zdsRGb9cLtfUBXsvwx5ECk7DZq3cdmJyT0/PngjsOgW7w/l6QbudTC7mUNf1NZIM+2GeKpUK09PTeDyee/Ld2rM/wrZBYKdlsIWFBRRFIRJpdVu5PjRN48KFC/z93/89ly5d4urVqywsLNDV1cVzzz3HH/zBH2z6HK1Wq5w+fZpPfvKT/OIv/iJQz4Y/88wzvPjii7zwwgs7/px7AQcZng6hUxwekcoWPi+daD3dTnaq15fguZGneGvyXUpKhZGRIY4nJkiXM3zn5g8o32ldLxSKPDrwAI97H8Yh37381svq7AS7vfiv1wlmd+VeTeJ1Kg76fT0M+Huth+p2g51SqcTs7Cxut5uxsbF9p8exnrP5eq3cmqaRz+fp7e3ds8TkdqFVh/NmZPL15tAeiO81sTqhEJ1IJAiFQvckQ9UsC7ReGUy8vtUskGEYbSs9qqrKgw8+SCwWQ1VVxsbGmJmZ4c033ySbzbY0dxcvXqRYLPLEE09Yx/x+P8ePH+fs2bMHAc97Hbtd0hLdCYVCgb6+PhRFYW5urm3j27HVzy4Ci3I+zwdGnsTpduJUnVT1Gu/MX7SCHahr/Jybv8TR7gmi7q6OZXVE54poSb4Xi38zf5717Ah20glm70q6X/RXBJq1cudyOVKplLXwLC4uUigU2kIm34sQDufRaLTlzks7OtkO324YhmE1FOw1heh2lsE6QTC2v+fAwAADAwMt/20ymQSgu7u74XgikejYOnMvcBDwdAgiIGqH9Hk2m2V2dpZgMMjExASKolAulzvGERJKoHYUtCKaoaPKakNHVLFYZGZmBpfLxcjIyN0F24SKXmWlnG0cGwnN0FipZIm6uzqS1VlYWCCbzdLT07OnOleatSGLhUcIWW514alWq8zOzlo7//ttsV+NfD5vWaREo3W9JUGEFmRyO4dlrwoitgK7w3k7uUkbtcOvVgjeqB2+3bCbnI6MjOz5DOV63WD2bPV63WDtMj21YydBVOmO8Ozq7l6n00m1ulVnxb2Lg4CnQ5AkybKX2G66WNM0ZmdnrS4be82+1SyMJEmYtVq93dzVmlqvSIsLU87r6ducS10hXy0QcPk4mTjKaGiAxdSi1RIbDAbX3MCqrOJ1einUSrbBQULG5/A2qKO24+a3l3Q2SvtvBEmS0EwdwzRwyo62qRuv917rkXjz+TypVGrdTjB758peMIbsNDYiJouyDKzlsOxFQcRWUKlUrNJ1px3O1yPki2B8dTt8JzSVRIfdfr+WWymDifkVwVC7yNCGYWz7OhGZtGq12hD0VKvVtSbT+xgHAc820cnWdNM0rR1/OBxmYGBgzQ3R0tjVKsXJKbKXr4Ch4xsdxTc+hrTBTlE3DVaqOfLuMtP5eVyqi3PJS2TuaMakSyu8fPNV5MEncesqExMTTQMLExOfw8Px+GH+vvhDyxQTJIZDffhkt3UeO4UQNcxkMlsWmjMkk0KtSEkr43V6WC5muLx4A03XGAj1cigyglvenQ6FVjvB3G63FSzutbR/J7AVxeT1BBHF4r1aEFEES3spmyD4K1tVw24X1jPKbKapZM8CbSe7aJqmxbO7HzvsVgdAgoulKAper7et3WA7CXh6e3sBSKVSDVzHVCrFxMTEtsbcizgIeDqI7QQ8okShadq6Zp+tjC0B2YuXWHrtDNzJUpTm5qkuLxN58glo8nDSTJ1LS9d4c/YcC8uLhIIhJEXikd6TXExdpVSra0FUq1Wm8nM8N/ZkQwZEN3WSxUWuLN1E02uMhocZCPbwofFnubx4napWpa+nmx53nLmpuYbuEZ/Pt63SQ6lUIpVK4Qv56R7tBUmmoBcxkQATn+pdt8Vbw+DSwjXOJS/R5QnicXo4M/02YU8IBZlkfoGl4jJPDz6GQ9r9W6UZiVe43osM3O3bt61FezfVjHcD7VJMFromdnl/EQAtLi6uWbw7IYjYCvayw/l6mkqr2+G34q1mX/w7ncXaCyiVSkxPTxMKhYjH49bctFoG2+y+3klJ6+jRo/j9fl577TXGxsaAevn4woUL/MIv/MK2xtyLOAh4OoitBDzC5yiZTBKNRonFYpvuZDfiCOm5PNmLl6xgRyB37TqhkydQmrRDZqs53pw9R83QgLqf1mw2hUtx0e2N8u7kRRRVJRgMYcqN40qSxPX0JD+YPGtlcyZXZjkeP8xjfQ/S44mhmzqOO2WiqD/ckL1YXl7eso7N4uIiuVyOrt4IbybPMzOVZKWSw+fw8GjfSeaySXoCCU7GjzQNWJbLad6YfQfDNIl6I7w6/QZLpTQOxUHI5QcTJjMzPNBzjKiza93vYjdgV4ceHh62AmGx896oE2w/LiSdVExebYxqF0S0c6ns12KnSbz7zeHcrg2z2lttdSmxWTt8Pp9ndnZ220Ts/QR7+bmZIna7usFEhmg7cDqd/MN/+A/5wz/8Q2KxGAMDA3z2s5+lu7ubH/3RH93WmHsRBwHPNtHKDSo4PJtB1OtN02R0dLSlnZ244NdLY5q6hlYsrj2uaejlCqv/QpIkFgvpO8FOvZsKE4IuP1PpGQZ9PXi8XpzO+sIzERlpyO4UakXOpy7bSlf1v7+8cJ1D0RGCih9FUqy/aZa9sAdAmUxmjQaLx+NBkiTLHNXhcDAwPMj3Z84ytTJL2agyn0thYFDWKzze9xBvz10g5A4wHhpu+LyyLHM7M41x55iqqOQqdeHEXCVP0OlHAgzTpFgtEXOFO8rn2Qh2HaHVJZ1mO297J1i5XLYsPe5l9mIr2G3F5FYEEVvxtNoO7heH881KifZ2eBEc9fX17ammgk7ALpy42vtrPawOgMQ4G3WDidfsZHPzL/7Fv0DXdf7Nv/k3lEolTp06xec+97mWbYr2A/b2k2+fY7MMj0jZLywsEI/Ht/xw3yjgkZ1OnF1dVBYbfacUjweliWWEaZq4HXe5KpIkUa3VkE2JhD+O0+nC6XTgVJwcjU+ssUXQDI1C1RZgmfVgwTBrlGpluhyblyPsOjbCfFXIqc/Pz1Or1VAUBU3TLF+oollhNlu3echXCxjU53t6ZY5nhh9DlVWuL91mODSAQuOOSJHvzptpGIyEB5jMzKIqKncqYjgVB0GXf9eCHUHgFq36ouOsVR2h9TrBROnhXmQvWsVeUUxu5mllX7zT6XRDML6eqORm2MgI837A6lKi6MISdhhi07KaB7QXrsV2QJyv6GDdSdZuPWsMEfQUCgXrdduFoih88pOf5JOf/OS2x9jrOAh4OoiNAh6RwpZleUtmn62OL/t8hB99mNR3X8ao1HVwJEUh/PBDKHd2VZIkUakZaJqB26UQ80SIeLpYKqQJugLIkowmGTw5dIqhUC/DXf14HG4CDj+K1HhjqYqDgMvPUiFdL7XdMYJyKk58jtZZ/pIkkc5rXJ1aIp0r0x/3M9rbTSRSY3p6GoBQKESxWOTq1auE+6KUK2VMuYmruNn8/7IsU9RKDHcN1B8cgFt1cTR+iBvLk0iShG5oOBUHJ3uOEnBszYZiO5AkiZxWYDabpFgr0e2LIZfrnXrb7TgT464uPTRrQd6uJ1i7IDrsRElnL5XhmpF41xOVtAeSG52DkHMIBAL3vcM53LU66erqsgxsTdO0eEB2f7rdbofvBETJrlNdZ3Zej9g0h0KhfaWsfi9wYC2xTRiG0eBW3gyzs7M4nU5isbvZELFrFyJxO7kZrl27Rn9//7o7Yck0qS2nKU5OYWo1vMNDOCIRUFU03WQymef8jSXKVY3uiI8HJmIojgrTS2kW0kUMU2KsN05fVxDHJgtQncNzm5dvvUbNqFFX3IEHe4/zcOIYMq090JdzGt989Sb54t25jQWdPHkyiluloWvFMAzKtQrfm3yNaws3QJW4nZkBCYbC/Tw58Cjnkpf5B6OnrZKWjsGtlSkupK6SrxZxqU6OJiZ4a/Y8Za3Mw70nWCguIyHxYM8x4u4wypoCYHshSRKLlTTfufF9spUClXKZSrXK46OP8Fj/Q6hS595/dSdYsVjc1TbudhGT7zXspUSRUbNnL0QmzX6+7wWHc/v59vX1bbggb2Qtsl9I+eJ80+k0/f39HW3p1nWd2dlZCoUCQ0NDDevMAZrjIODZJloJeMSOJZFIAHd3dU6nk76+vh0TMW/cuEF3d/emRph2803x8/WZLH93ZhLdqB8zTJPeqJtHjvTw8lvTLGTydX6I18sTJ3s4OtTFemuepZZsGixXV7i+fJuqXmM8PETCF0VtOZEo8fI781y8uVT/TLpOoVhEkiR++gPHODQQbFpaymkFfjj9FrPZedKVFTyqm0d7T3IjeZv+UA/HY4dQjHqpJ1lZ4v9cexnjDtfIlEwWi2men3iWd+YuIgERTxhFknmo9zh9vu6Ol7MMTL439UOuL96mWCwAEj6vF1VR+cmjH9p1wrS9fFMsFi1PsHYvOnZicjvuh70Ee/ZCzKPIasiyTE9PDz6fb19mL1qF0BEzDIP+/v5tfb8ikybmUXTU2bvB9gonTQQguq5v+3xbhfAZM02T8fHxLZshv1exN66U+xSChyLk0ldWVujp6WmbN0yrXWCrF+xy1eD8jSUr2KnVNArFAhP9AX54PkWhpGNoBgoK1arGmfPz9EZ9dPnX3sCrPbDi7gjdA/WUtTjeKqqaSTpbBqBSrlCulHG73bhcLpZWKhwebG7nEVB9vH/4NCW9/rfynXLb0dgEDlMhn82TL+bIF/K8tXKJfCGPqqqoqoopmxSqRaZWZun2xZnLJVko1AOuoXKWPl/3mvdrN8p6heTKAsViAZfLhcPhvEOYNlgqpnedMN2sjVss2kKDxa5js1n5phl2m5i82xClGY/HQyQSsewh3G43qqoyPz9vyTLYeUD3yzyIzV0wGCSRSGz7vJrZs6zuqBOaNq22w3cC5XKZ6elp/H4/AwMDHX3/XC5nlX/Hx8fvq41Cp3EQ8GwTrQoPFotFrl27hsfjWVekb7vYrrChphuUKxqGaVIqlqhpNbxeL+Ggj/M3Z/G41fpu9A7xpVTRyBaqDQGP3ehSBDuaITG/UODq9AoORebwUJh4lwt5PSGcVXA6JEI+J1dvzQMQCASsTEKsy73hou+QVBxq83S5K+oiGo1iYKAUryIrCtVajWKphNPlQJUUStUyLpcTkzpfWUIi7o10NNAwMMhW82iGjsflxnRJFM0ybgNcsgMJiYDz3u/cNuoEW15etsinrXSC7RVi8m7B7nDe39/fsBO3Z9JWVlaoVqtrMml7icvUCuxdZ50o2TXrqBPt8HZxztU8oE6WwYRQZKe77IQUx+LiIvF4nMHBwfsmQN4tHAQ8HYKu66ysrFAqlRgcHOwIN2G7AY/HpRAOOrkxlcSh1nfysiRhmCahgItqVWswR5UkcDltHU2mSaGss7RSJlesEu/yEPK7ePvaAm9cSlmvuzqV5tmH+zkyFNr0c5qmyUomw2iPh5vTfqrG3Ru5L+ajL+bbcfChSApj0WHm8ilwuTCpL8BRVWE8Mszbk+fJlurZnwf7j+FXvG3xQmuGmqnxxuw5Ls5c4YmJU4S9Xbw8eQZFlpGRiXnDHIsfIuzpumft8OuhWSdYKzo2Qk5gLxKTO4HNHM43EkRsp7nsbuFedJ1t1g4vAslOaFPZVaI7LRSpaZplHj0yMmL5yB1gazjg8GwTYmfRDCLlKFKrIyMjHfkMs7OzuFyuLV38uq6TSqWomS5eOZciW6jr7siyxLMPD5Ar1Xjj4jzFUgkJCY/HzUB3gA+9bxCXo94unS/pfPuNGeYW67o1DkXiA+8b5juvT1HTGnWHIkE3P/HUKF7XxoqrotY/MDBAoSpxdTJDOlepd2n1BTf8+62gYtY4O/sON5ZuoZsGDsXB8cRhjsRGuZ2ZJVfO0+OL4Ze8pBeW0XW97QRewzC4kZ7kaxe+xYODx5gvLiJLMgl/lBvpSapajcFQL08PPUbEGdpzAc9msHeCiYVHtNEGg0EikQgul+u+3p2Kkt1OhPXsasYigyHKN2Lh3ivGqPaSTiKR2FPEYhFI2nlAqwnlqqpuaR41TWN6ehpZlunv7+9o8G7n60xMTNxX3la7jb27XdjjaHZziCi8WCzS398PYLX9dgJbzfCIQMzv9zPQ28VPhUPMLxcolDR6ol7CgbpdgUuVOXc9SU3TGR+I8uChOC6HdEdEUeLS7bQV7IjPsZAukS1UcTtVNN1AkkBVZHLF6p2fVco1g2rNQJElfG4FwzAsBVI7l6PLAadPJDDNenapna7wLsnB6f6HOR6foFgr43d68Tt8KMgcj040KFhHguGWxRBbhXD1vp6dJBgI4Hf7mU2eZ7G4TNwb4XBsjLAnRLFSYrmUJuLcf0J0dh0bkeUwDINQKGQ9vPejoWcraKfDeTM140qlQqlU6rggYquwqwjvVeHE1cra67XD23lAGwXkgp9kb7HvFMQzW/jIHfB1doaDgGcHsC+O2WyWubk5QqEQhw4dsvg77VysV6PVgEekmguFAv39/daN7/fAoYH6A+puFkHioUNR+qNOappGItaFaehoWj0TVDNgdrHQMH5NM3A6FKo1g3ypRCZXQZYlQj4XEwMhVEVmaqHIG5dSZHJlvG4Hx0YjdAdMctlM03SwOK9OJDdUFMLOEOFVwYT4Lu2wiyFCPagVAdD8/LzFu9jMRVpwOTKZ+vn6NS9SUUI3dLyO+rkvFJeppmr0+hKYpsnxxOH2n/wuYiNi8kaGnvuh/bgZOu1wbg8k7eUbe0DeqY66ZjAMg7m5OSqVCsPDw7hcu2Oyu1OsJpSvboe3C0va51GSJIuftFmL/U4hnhdLS0skEomOE6HfKzgIeHYIUY6pVqsMDQ01pBtbtZbYLmRZ3rQ1PpfLWd0SExMTax7CzcolpmkimVW0ShFD91s8FkmScEjgdjZeNoZpspgp8sBEjJffnEbTDdAhnSszPjiMicnfnZmiXK0HTdl8idszKX7q/Uc5PDq6w1lYDYlcSSNf1HA4ZIJeFafavgeFqqqEQiECwVA9+7RBB5N4UIprxOFwMDY2hqqqHIqOcXN5kqVimhOJI0xn58E0CbkCmKZJ3FcXgdxv5SxojZi8XieY3Y17p51gu4lMJkMqldp1h/PVAfl6rub267EdPCCRqfN4PDtWEb7XaGZzYxeWFPMonoGxWGxX+DrFYpHR0VEiTXwPD7A9HHB4tgnTNEkmk9YONh6Pr7npa7VMfmFDAAAgAElEQVQa169f5+jRox35DMvLy5RKJat8Zsd6WZ1WINLUyWSSYDDYoBwrSRJTqSJ/+8Pb9cCGekeTw6nwwUcHmUzmuDqVRlFkjgyHCXgdVCo6Zy7OYxqmlfXy+ryM9nfx/ONDKC12cW36uZG4MZvjh+fmyJdqyJJEf8LPMw/3E/Kq6IaJLEmY5vazblUNphcKXJtOo0gyR0bC9EY8KHe+ensHU6FQoFwuY5omPp+Prq4uiy9gYHIrO8W5+ct0B+KYmExn51AllW5/jBOJI4Qcu2dp0S7YFZO7u7u3HaTY59HOu7Av3HshvS9MXcV9uJcczuHuPNpFEVVVXTOPWwnQRFdSIpGwAoT7GdVqlampKWvexHzaifnt4lMJLpQkSYyPjx/wddqMg4BnB5iamiIQCKxbp9d1ncuXL3P8+PGOvH8mkyGXyzE4ONhwPJvNMjs7SzAY3PKiI3R1mj0oxU7RHwyxsKJz7voSxXKNaJeHQ4Nhzl6cR9MNIsF6C3kqXaIv5sPtcvL6hen6GE4n7jvz1Rvz8+NPDOFo08Y9W9T58veuU6po1jGPS+WZhwYolmtMpXL43Q6OjkaIBZ00ek9sDsOUuDy5QiZfwanKVGo6C+kih4fCHB9pzMbYidiRSKQhZS4W7lAohOkAHQOX6kICNF3DrbppUwy4a+i0YrK9E6wZgfdeeILZHc57enr2RZZjPUHEVvgrQk9McBT3WnDXCaxnEWHnU4n7eqe8NPHc9vv9Vib4AO3FwYzuAD09PRtyaATHplOtzas5PCIVKlrht6K+aakl3xlPluUGop/oGCkUCiymkkjAY4f8yKoTr8eJgUqhVKNU0cjkKta4bqfKYNzDD6oV/D4/ino3uhnpDeJybK+1fjUkSWIxU2oIdgCODEf49htTlCsaqlL/Dq7PZvix0yP0RjfW9lmNUsVgaaXMaxfmWc6WiYU8nDrWzdJKmXxJx+eWLSO/bK6ILxSjK+DDod5t8bcv3IuLixSLxYYdt9fr3bVgR5ZlKnoVExO34tr29yCCO9M0GR0d7Ujmxc67WO0JJgi8QMvE051gPzucb8RfaWaMKhZu0ZXkdDoZGRnZ0+XFdkBo3mQyGQYGBtZkWtbjU9l5aa3qKtn5Ot3d3fT39x/wdTqEg4Cng5AkaUNH853CHvCI3UEoFGJiYmJLu83VasnNbrZmgl+WZshCGllReHiii9fOJylXDRwOlUTEw1DCid9j8tzjh3jn6gKVmo4iS4z1d3FoYHN9nq1AVRvP2e91oBsmVyczDPXcLelVqjrvXFsg3jVolaJawVK2wldevk5Nq3/mQqnG/HKBj/3IEQyzHnCmMxnyFYVLU2XyxRxBn4sHD8XpjbgBc92Fu1arUaqa5NNFJNPA7awLIPr9fpxOZ9tLWzU0bi1Pc33pFgYmw139jIdHcMtb0065V4rJ6zmai4V7eXm5Qcm4XZ1gwj5A07T7wuF8Pf6KyFokk0kqlYrVDbafgrvtQtd1q7NwK8a96+kqlUollpaWGsqyr732Gt3d3Zw4ccKiJoyNjVnB0wE6g4OAZwdoVW25kwGPrutMTU21JauzXrDTDHYXbjFWvKbRG/OTyZVRZHAqBrJeQjJ9nBzrYqw/RK5Qxet2EPCqbStlifePBN10BVxWhsnjVFnOlgn6nKiybClHA2QLVSqajtfZWsQjSTJXp9JrusYqVZ2bMyscGw5y+/Zt/OFeJlMreFz1ev7cYp755QIfeXqMaHBt5qMunOZhZlnjzcsLrOQreF0Kjx7rpifs5epUhmyhQl88SMin4nLIO89cSHB54QZnZ962DqXyi6xUcpzuewSlBaNXUd4oFAp7QjG52cLd7k4wu11Cf3//vihhbQeqqhIMBgkEAiSTSTRNIxqNout6U2XtvcKnagcEhyYQCOzIEgPWb4cvFAp86Utf4p133kGWZcbGxnjyySc5ffo0jz766AFJuYM4CHg6jO2qIbcCQeb0+XwdyepsBZIk4XI6qFbKKNoKPp8Pv99PqVSy1HcdDgd+vx/VdIPhAaW9D8mAR+G5U4OcuTDPQrqEfCeTNDWfawh2AMIBF251KwtWPTiMBN0sZIpW4CNLEg5FopRfYWh4mNupCtMLBdLZMt0RHw8d7mZyfoVUukQs5EaSzDXXw1KuyvfenKamGXhcKuODEWRZ5dzNDD6PSq5c4Wuv3OT4aJQj/W6yK5Pb5ApI5Ms6mUIZo+LjRPQkk/lb5Cp1TaUbS5N1svQ6Fh0CdmLyXlZMbrUTzD6X65Uc3ksO51APFqenp1FVlbGxsYZ5Wa2sPT8/b2WA76Wf1U7RaYsIe3b3P//n/8y1a9eYnZ1lbm6ON998ky9+8YuUSiX+6T/9p3ziE59o+/sf4CDg6ThEFqadEC7EpVIJRVHo7e1t+W93ktXZCKJbpVgs0tfXZ2V+VoumNXtI+nw+i3S6E5imSaLLxfOPD1Gq6MiyhFOVmVvo4tp0xnqd1+3gwUNxFKW5Gel6GB8IcWNmBY9LZaVQRcLE71E5Phqmq8tPpmDwd2duM53KkyvWuD6zQlXTee7UIPNLBd66tojf46A36rOUo2VZ5sbMCjXNQFVkTo7HOHd9iQs3r7OULRMNunnmoT6OjiW4OpXh6FiCsbEuCoWCNZdCe2VjMcR6d90r78ywlCtyOzPFUCLMsw8f57LxLsVaCc3QqOm1dZ8KnSYmdxobeYKl0+mmmQvAcnTvFD9pr0EQdddTiW5Wlt3Iz2qvC0uKjtt8Pt9xiwhhR7G8vMzY2BjPPPOMNS+1Wo2LFy8eZHg6iIOAZwdo5QZWFKVtGR67wGFXVxc9PT1cv359S3/fzqyOQD6fZ25ujkAgwNjYWEOmSdOhWKkHfG6Xm2j0LllSBEBC7dTOE9pu141pmjgUcHjv7kiffrCXkb4gt+dyBH1OxvtDhAOOLQU7pmnSH/fzwHiMS7eXcaoSkgxHh6MkIn6QVDL5HIeGIkwMRqjWdK5PZ3jseA/nby5zdSqNKtcNVctVnfH+IG5H/dx03aDL7+L4WJS5pSLzy0Wr5b9c1fnOG9P8Pz96FIcqM79YoDcStfSAoDnnYnUAVKzCK+/MkC1U72SlnNxKLuO56ODokX6uLl/D7/TiVpuLx+0GMXm30YonmGEYlk+T3Sj3foTd6LQZUXc9rOdnJa7JvSwsKZTAFUXpeLZS0zRmZmYol8tN+ToOh4MHH3ywY+9/gIOAp+NoV0lLZHUqlYolcGjP1GyETmZ1UqkUhUKBvr6+NfyhbEnnB+/OMZPKY5omPVEfTz3YRyRQXywF6VQEQGKXWCgUSKVSDTwhn8+37bZjl0NivC/ARH89IyE657Y8jgqPHo4w2hdgdrGAoiis5CqcuTBHPOLn69+/yfxSgVJZY6A7wC//5HFev5ji69+/ic/jAEwu3VriA6cGiYc9eLrqnVFHRiLIisxytsLZiymWMiVCficOVaFc0dB0g5mFHKGAC793baAmOBfNTCgXFhao1Woo3hipxRVUVUV1OIj5wsznFphMpTl1/BCqrPJQ7wn8qndNt/69IibvNkTmwu12W3PY09MDYHXVAQ3B5Fb5VHZT3r0EEdBKkrQlou562IogopjP3W7DFpyscDjc8eu6VCpZRrJHjx6955y39yoOAp4dYCuk5e3CNE1rtxkOhxkYGLB2RnZri/U+S6eyOoVCgbm5OXw+X9OdkabDq+/Oc3suax2bWcjz92/P8KOPD+FyrE2T23eJq9PkYrGxZ4C2whNoZhuxFYhW5GKxSLri4dLtNJWqjm4YvO9YL9/4/g1qNQMZCa9bZSFdpKYZvH55Hoej3q5uUuf8nL2Y5OHD8Ybv79V35+jvDiBLsFKoUKnp9ER91HQdTTNxO1UkyaQn6t30PFaTJQ3DIJmpgCRRrlTQCgVcLhf9/m5Up0nEE+Inj36ILmegIdjZa8Tk3UC1WrV2/HYtFHFNNmvhbqUTzATyWoFsJYdLdRFw+nFJeyNLVigUmJ2d7ag3VLNyosimZTIZ5ubmLHkGEQB1SlfJLivQaYsIwHp+B4NBRkZGDvR17iEOZr7D2Im9hN22Ynh4uKkWxHpdYJ3K6ghzxFwuR29v77oPi2JFZ2Yht+b4/FKBYkXH5dj40lsdANV0k3xJQ9N0NLNGKpWiVCo1LDad0l2p1WrMzc2h6zp9/YO88/ochVLd0mOsr24xoekQ9LvQdINsoYpDkanWdAwDMLnD+QGHqmBi4nQomKaJLMtcn67zggzN4LHjPcwvFQGoVDXCATeqLDHWHyIacuN3K5sGPIYpkS9pGCY4HTJ+t0rYLzHcFyW5XADTRNN1dE3neH8CrwErM8tU3EVrwREZxb1OTG4nNnM436iFe6NOMEmRuJq+xRsz71LWKkhI9Ia6eXrwffiUexdE2jlZu7Hw29FM5kKUuPP5vKWrtJNsWjPY/b86LSsguEHpdJre3l56e3vv2+zofsFBwNNhbCfDY8/qRCIRBgcH1613Nwt4OpXVKRaLDc69my2C663LW0205MsG339njqlUFsMwiXV5eeqBPvr6FHK5nKW7YhhG2wMgsQiGw2FisRhIMg5VRkJiYrALkCiUNOaWCvgKDnqiXkJ+F4VSjapmMNYf4uzFZP28gaqmMxEP4XWr1vdTLNUAk0y+TG/Mx088PcLbV+oP/OOjUR45Eqc36oYWslTlmsnZi0muT2eo1HQiQTePn+hhuNvP+x/p57ULSeaXCqiKk5G+EI8cjuN2gM/rsbgrgqsj5rFTsgp7BTvJZG1WTqxUKgR7uvje1VcxJbO+u5dgdmWe867LPN738FYFv9sCEdAKrZl7zcmy6yptZuhpz6ZthQdUrVaZnp7G7XZ33P9LbFYrlQrj4+PvCQuO/YCDgGcH6ERJy57VGRkZ2fThax+/k1mdhYUFstksPT09LbXlelwKfXEfk/ONWZ5E2IfX3friaZgSZy4kuTW3Yh1bSBd5+a1pPvzUCF1dXQ26K9spNzSDruvMz89TLpcbFkFJgmMjkbojvCIxu1jk4UNxfv5HDlOqaly+nUYCgl4HDkXixFiU5WyZdLZCVdMJB9w8//gwiiKRK+n4PDIj/SGuTKUBuHhrid6Yn6cf6mMwEaAn6sbRYjeZJElcuLnEhZtL1rHlbJnvvTnNTz87Tjjg4IPv66dY1pFl8LlVpDurrTCVzOfzuFwu4vE41WqVfD6/hk+1X9uOm6HdDufNtFcuL1+nZmjomkahWECWZVRV5fbyDCdih/E5dtcvSXBXQqEQ8Xh8T36PrQoitmowm8vlmJub2xVz11KpZLX0Hz169D1hwbFfcBDwdBitOJrDXcNOQQ7dKKuzenxBwu1EVkdorrjd7i2RGZ2qxBMne9F1k7mlAqZp0h2pk5Y9TrllPk2hrDUtjS1ny+SKNXyuu11Fq4mSQniuUCiQyWSsHaJog18vACqXqyxnizg8YWIJN27n3TKSaZr0xbw88UAfqXSRmWSKV96aoTvqA0weP96Lbhj0Rr2Eg27mlwp86LEhZhfyKLLEsZEIyeUi/+/fXUUC+mJ+Hj/Rw/HRKJdvL6MbJvNLeXxuB7EuF6rc3NG+GYoVnVs2zpR1vKyRXC4S8gVRZQhaHWw2IcZ1iMmrCeWCT2XPAHWynNgpiCxqpx3OZVnG4/TgEYrQgK5raJoGhslKZoW5ldld8QQzTZPl5WWWlpb2pZ7Q6myaXVZgtSCinQgtOs92g4eWTqct0+X3Sil4P+Eg4OkwFEWhXC5v+BrRGink6rdyU8qyjKbdeYDS3qyO8JLZjuaKaZqE/Q6ef3yIYqUeiHldCu4tBDt1rH8um53lauG5ZnyL1RyBTK7M6xeTzC6VMFBIhL2cPtlLouuuvYMi1zM4P3h3heszK5imyexCnoDXyTvXFvnFHzvCQNxLplBldjHPVDJLrMtLX9zHmYtJrk5liIfcmJhcn8mgGQbvf6Sfo8NhMvkKQZ+LLr8T5zbuzq2WC1sp5zRrO7YHQKKcuF90V+zZu93QXYl5IwTdAbLlHBKgKiqqovLgwHH6I32Ug+UGT7BOBJO6rlvX/P1giQEbywpks1mSySSGYSDLMtFoFFmWOyYrIO6jTCZDX18fPT09e/b6fy/jIODZAXZa0rJndaLRKLFYbMtqyYqikEwm8fv9G4jObQ0iq+NyuXbk2muaJk4VnGojv2gr8LkVBhJBrkwuNxyPhtwEvFvjHazeIWqa1mA94HK5uTxb49LtNKrDgarA7GKe774xxU88NWqJBZqmiWHWu87K1btmpeVqiWyhSiZfoT/mJuhRef8jA7x2YZ7pVI6BRIDbc1nCARdIJr1RP5GgC1mS0XSIhZzEu1zb7ibzuhRG+4IsrZQaj7tVemI+dANqmonLIQPmjhSTW7Fx2AnfolMQ7cGiu3A3PpNf9fL+kSd4Y+4cy4U0LtXJRHSUsa4hTNNc0524XmnWbkK5lXtcnLPf76evr29PfA+dgF0Q0ev1Mj09TTAYxOPxWDygTgTmYsNarVaZmJh4T/iN7VccBDwdxnoBT7VabTAh3EpWx87V6enpsQzqUqlUU9G5Vh9wQngsk8nQ3d1NMBi857sUWTJ537EEmq4zOZ9DNwy6Iz6eONmLr4WOpY2gqip+v59KpVI3mgxEmF26jUl9kdB1HVVVKZfLLK0U8cS91nw4VIlIyE34jneXCThUmYGEH49L3FYmgwkv0dAwVc1E0+t+X4ZpcHI8TiZX4Yfnk+i6gW6YPHIkhs+1/cXINE2Oj0YolDSuz2So3iEtP/tIP8VSjR+em6NQ0oh2uTk8EKSUTRGLxdrygG6WTRMB0HpiiLu58NpbkXdbJdo0TWLuMM+NPElZryAj41M9Ta/dzTrB5ufnWxbxExuqhYWFfamMvV1kMhlSqVTDOa/naG6fS3F9bsdfTVVVjh07hsvVXLjzAHsDkrkXVbD2EarV6oaLrrixxsfHgbsP3mQySTQa3TJpUHB1RKp29d/au0QKhULLC025XLZq4D09Pfe8a2M1NKPOUTENcDvlbZTG1kIEnbIs09vbS0VT+MrL18nfaTk3TRP9Trnwx58aRamlURQFr9dLMBTi/K0C3393jnJVw9BNPG4H3RE3P/n0KH732jmu6vC/X5tEVWQqVYNvvz4FgNupMtjt59BgFx94tH/HpqqGCfmSjmGauB0K2WKNr79yi5VChWyhimHoDPcE+b8/eIigZ3c4BvbrUnjArV60O8V30DSNubk5NE2jv79/35dzms2lEPET/yRJstqv+/v73xMLsSgrFYtFBgYGWjrnjeZyM0FE8RwPhUKMjIxY1+9f//Vf86d/+qdMTU0xNDTEb/7mb/LhD38YgOnpaT796U9z5swZ3G43P/MzP8MnPvGJhvf4whe+wEsvvcTCwgLHjh3jd3/3dxsUmHdrjPsR9/fZ7QHYMzxC1EzXdUZHR7fEHVjdgdUs2IG1XSL2G7pZBsjtdrO8vEw6nSaRSBAKhe55VqcZVJmGxXmnIoJi5xuLxQiHw0iShKJK9MUDVvlMkiRUh4NwyEdfoguvs8t6MCbn5+nrCvDAWJBbszl0wyQe8XP6ZC8BT/PMk0uVeOJEL3NLBf76u3VLEJdTIRHx4lBkJuez5IrdRAI7uy1l6S4x2TQlzt9YIrlcZG4pX1ddVhSyRY2T4ys8fjzWMXNbO5qJIYrrcmlpiVKp1LBoi66xncLucD4wMLAnr+2tYr25FGWbmZkZoJ51E9yV+x3C7NThcDQEH5uh2Vw2E0T0eDxcuHABl8vFww8/TD6fJ5PJ0N/fT3d3t3VdffnLX+ZTn/oU/+pf/Ss+8IEP8PWvf53f/u3fJpFI8MADD/Dxj3+c0dFR/vIv/5KpqSk+9alPoaqqZRb6pS99if/0n/4Tn/70pzl27Bif+9zn+NVf/VW+8Y1vEI1GqVaruzLG/YqDDM8OUavVNlwwqtUqN27cIJFIkEwmicViW1Yz3SyrsxXYA6B8Pk+lUkGWZbq6uvD7/XvG46ZTEPojuq7T19e3ZheYKxm88s4M06k8hmHSFXDx5AN9DMa9rBZMMQyDak0jX9TuCPmVQa9airHN2mQlSWI5p/GX//sy1ZqOQ1UAE9Osl8R++tkxeiLt243XdPjqK7d5+c0pqjUNVVWt6+fZh/v5yDMjOPdAI4l9oRGtxw6Hw+qq83g8W8o6mqbJ4uKiJfq23zqStgthzBsOh1EUxZrLdvjU7VUUCgVmZmY6Yn1iF0T89//+3/Pd734X0zQZGRnh9OnTPPXUU5w6dcoyUf3Qhz7Ej/zIj/CpT33KGuPjH/84jz76KENDQ/zrf/2veeWVV6wy8he/+EU+85nP8IMf/AC3282P/diP8aEPfYjf+Z3fAerP6+eff56PfvSj/MZv/AZ/8zd/sytj3K84yPB0GKKDKp1OdyyrsxUoioLP56NcLqNpGolEAqfT2RYO0F7HahHBZnMZ8Mh88NQg+ZKGbpj43Mq6XCFZlnG7nLhd9RKJYfjWtMk6nc6GhUZRFLxumYGEj8n5PMnlIsVyvYTWF/Pj2k5r1gYwjRrRoJOapq8JGHqiPqo1A6dy77/fZsq7q408RTlxs0VbaFnBe8fh3F7OGR4etp4zq93MO9kJttuwt9n39/ev8fJrB+yCiJ/+9Ke5evUqt2/fJplM8vbbb/PlL3+ZYrHIT/zET/Cbv/mbzMzM8JGPfKRhjM9//vMA/N7v/R7Hjh1r4MydPn2aYrHI+fPnGR4e5tatW5w+fdr6vaIonDp1irNnzwJw9uzZjo9x6tSpNs7g3sJBwNMhiJsxlUoBbLkjpJ1ZHTsqlQpzc3NIktTQnip2wJuVwPZjAKTrurUYtKLF4VRpKCu1mgRt1iYr5jKdTlscqVAoxCOHE1yZzJAtVADwuFQeOhzn9twKwfEoiryzxKu4/tLpNMfHhhi/HOH2fF2jRwIeOhTH69oDqZ11YO+4EYu23XpgPTFE4Qu1UVB7v0EoCLtcrqblnGY+dXbybjMV43Z0e3YS9jb73Qhql5eXSSaTdHd38/jjj1tzrGkaly5dwjAMbt26BdS/j1/7tV/j3XffZWBggH/2z/4ZH/zgB0kmk5YZrUAikQBgfn7eClKbvebdd98F2JUx7mccBDw7RLOHguDqCNn269evtzxeJ7I6YlyxG7LzVlZjqxygvR4A2U1Ox8bGdvWz2hdkuJu1KBQKeJzwDx7qJZ2voOl1J/mFdIm3rmQY6w/ZxAG3DpHhME2T4eFhVIeLn/nABDdmVsjkyvTEfMgSOFWZgEfFNDvP4dkpmlkPrBZDFJ51XV1d75kSlshabnRPr4a9E0zs8rfbCXYvUKlUmJ6exuv1drzNXmyWVlZWGBgYIJFINMyxqqqcPHkSqPN3AH7nd36H3/iN3+ATn/gEf/u3f8uv//qv8/nPf55yubwmCyU2nJVKhVKp1HDM/ppqtQqwK2PczzgIeNoIe1YnHo9b9WRBXN7sxuxUVkd0IwFbFh3bLACqVqvWg9Hn8+263oqYo9VZmO3YYXQa9qzFck7jrWuLBDwOTEPn+u0FypUqqipTrlQwazWrbLMVNC/bGfRGPcgyzC6qVGsaQ91B+uPefRHsNIM9a+Hz+Sx7iEAgQLlctoQ8t2ststchjCnz+XxbFIQ38wRb3QnWLlL5ViGu73g8brWadwpi41qr1Th06NCmbf3iXv0n/+Sf8NGPfhSAY8eOce7cOV566SXcbrcVdNjfA7CuT/sx+2vEpmk3xrifcRDwtAnCkwdgbGysgQy7mZ9WJ7M6QntkPQforWKjACiZTO5aACRJElI5h55dxDQN1GAc0xPANO+22At/pL3YaulxycRCHkwgGgrQ3xMmmy/XhehUiWymrhQruC3CDmO9AMiumDwwMLDmweVUYSjhpT/mRZLqnVz3Q7/CRg7n+0UMcasQHUmqqnbMvmCj7iV7edY+n50sKwmNsN2yiBBEaKfT2bK+jigTHT58uOH4oUOH+Na3vsWTTz7JxYsXG34nKA89PT309fVZx44cOdLwmu7ubut1nR7jfsbeWwn2IRYXF1lYWGjI6tihKIqVbl+NTmZ15ubmrLJGp3Q47kUAJEkSZnqW7PmXMUp5AGSXF/+xp8g6QiSTyT3dYg91VeQPnBrk/M1l3r6ygKYbHB2J8PjxHroCKqHAoMVbKZVKaJpOaikHkozHCaoi43a7cTgcVoC3mWKyaZrIkvj/Lp5sB9CKJcZeF0PcDoQJZrs2MK1iI1K5sHGwv8bj8bTNYFZ0Vpqm2fENjD1LH4lEGBoaajmgPH78OD6fj3fffbeBNHzlyhWGhoZ47LHH+NKXvkQ2m7WuyR/+8If4fD6OHz9ubdBee+01nn32WaD+PH399df5uZ/7OYBdGeN+hvLiiy++eK8/xH6GUCYeHh5eV5k4k8ng8/kaSkkiqyMCoXZmdTKZDLOzs4RCIXp7e3c1wyHLMk6nE5/PRzgcJhwO43A4qNVqpNNpFhYWKBQK1Gq1uvaNomz5vGWtQv78y+j5jHVMr1bIpqbx9AwT7e7D5/Pt2WAH6kHb1eksZy/OU9Pq/j7ZQhVNN+hP+FHkOzpAqoqkurk0VeTt62muz2TJ5HViXV6WF5OkUikymQx+v59YLNbQdn6/olKpMDU1haIoDA4OtlyilWUZl8uF3+9fc22urKxYJSIhJqooyp4JgESGY2lpiYGBgXsezEuSZBl1hkIhIpGItekRnCqhr6Rp2rbv9VKpxOTkpGWL0UkzTkGETqfTDA4OMjAwsKXvX1VVdF3nv/23/0Zvby8ej4f/+T//J1/84hf5t//23/L000/zla98hVdffZUjR47wzjvv8Pu///v80i/9Ek899RRQ59L8l//yX+ju7sbhcPDZz36WK1eu8JnPfAaPx8Pw8PCujHG/4kCHZ4eo1Wrour7hjXzr1i2i0ajFI+mUs3mtVmNubm5djZm9gNVK0NvJAMnlLFn03rgAACAASURBVOlXvgSmAWY9m1Uql3C7XHR/4Ocx/ZFdOpvto1A2+NorN8nkG0mCiizxMx84dLdLTJJ49VyKd64tABDr8jCQCBDwOYkGnBi1Ak5VsuYUWNO5dL8EQHaH805k8OxiiEK/RgTvm6nudhKChC5JEn19fXuyRNsMdk+wYrG4phNss3tdbJB2wxZDdLrpus7Y2NiOOH9/9md/xv/4H/+D+fl5xsbG+Of//J/z/PPPA3D79m3+3b/7d5w5c4ZgMMhHP/pRfuu3fqthHl566SX+/M//nEwmw4kTJ/jd3/1dTpw4Yf1+t8a4H3EQ8OwQhmFQq9U2fM3k5CShUIhgMNjA1WlXsGNfCDohvtVJbCcAkis5Mj/8G4xykUKxiGEY+Hw+FKeL8BP/F4Zn75v3ZYs6/993rlGprS11fuSZMfqidfJhvmTwlb+/Tr5YozvipTvq57uvTzKVXOHwSJSjwxGeeqAPv1teYzxZvDM3+11vBe46nO+mVcJGYoi7wVsBrDb7rq6ufd9mb+8EK5VKDSVFu4/VdiwidoJ8Pm+ZJY+Pj+9765EDrI/9sVXY5xAcnk5ldUQb6dDQ0J5XyZQkCcnQQKuBwwXb4QC5/KixIRYvnrF24JIk4U4Mgcvf8meRJQkqRcAElxdji6G/+B5N09wyAdjlkAkH3cwvFRqOe1wqfs/dRdQwTTStPvZQd5Avf+8ai5kiqurA0CVuzWbxOFWeeagXSWruYl4oFCwxxP0YANkdzkdGRnZtF7pV3ko7FYyFUnQmk6Gvr68jonq7jY06wRYXFymXyzidTjRNw+FwbKlcuR3Y+TrRaJShoaH7PsPxXsdBhmeH2CzDY5qmZeLX1dXVtrS4aZrWQ3e/iKzJmBiZeUqT5zFKBdRAGPfwSfBH1w0YmpmhOhwO4kEfanoSY3kGMHHGBnEPn8BwtNZWqdRKlCfPU1mYBMAVH8I1dALDsXn3hyRJSJU8tcVp9MIKjkgPSqgbQ219JypJEtMLJf7u7CSlilb/TIrEEyf6OD7ahXTHxkIz4Ftnpkkt5xnqDfK/vnUVVVVxOhSGuwOoqkTA6+SnnhnF34IRaLMyw15t3b6XDuetwC6GKP41E0Pc6nwKkq5hGPT3978nlKKhnmmZmZmxCOilUqnBpqWdGTXB1xFt/fF4vC3jHmBv4yDg2SGEANp6v9N13SJFCjdekZUQN/FWdxWaplnp/b6+vo63aLYDkiTBSpKVN/4WU7s7X4onQPB9P47h2rxmLowgnU5nfWHGxGFqOFQV0+lBUR2tcYBMg9Kl71OeudJw3N1/CO+xpzGk5lYS1t+Xs+Te+hZaPm0dc/WM4D369JaCHpBYKWpMJ3PUdIOBRICI34Eqm1DOY1ZLSA4XWc3J/3l9kni0i//1d9dwqDI9Ud+dTJBJ0OfkI8+MNXVo3wyidVtkgQTPQlyf9yoA2o8O53YxRBGkbzWjJq7xUChEPB7fM8FnJ2GaJktLSywvLzdYRNgzamI+RUZNlMC2E1AKCRFd1xkfH7eyywe4/3EQ8OwQzQKe1bo69hKWnRhZKBQol8sNJZvNWmOF7oio6e+XFKyMSen89yjPrVWd9h89jWPogXW1ikSHSiaTWWMEKTJAYsG28wI25AD94MtW4CUBaBWMSpmup17AlGTUcA+6sjZ4kSUoX3mN0u1za34XevR5pOjQlstb4vowDAPZ1KlOX6R08120coFSpUZg5Djew6dZLit8/fu3yBerOFQZYWZ6cjzGUye7rZ93AjvPolgsrtGu2Q3LAbvD+Wpl2/2GVjNqdiX095LZqd0iYmBgYMMMzmp17WKxuGVPMHsWaWxszAqkb9y4wYc//OE1r//93/99fvZnf5aLFy/ymc98hnfffZeuri7+0T/6R3z84x+3XmcYBv/1v/5XvvjFL5LNZjl16hS/93u/x/DwsPWa3RjjABvjgMPTZmzWgWX3W4rH41YAJEz9KpVKU0l3TdNIJpOUy+WmwnJ7HoaOXik2/ZVezLHeY07sxhwOB2NjY2vKgevpABUKhTVaK1YAZBgNwY5ZLVJbmsU0dIxygdyVM7j7DuM58jiGvOqTaVW0lYWmn7WWTuKKDW854LFzgMzcEoUrZ6lV66USl9OJMX8dKdZHd/8RPvzkMD94d46FdBGHqjDSG+LhQzHaEezAWp6FXbtmteVAu4Ul7Q7nfX1998XOW3inNbNwEAu92+1G0+plzU5qZu01bNUiYrUnGDQGlM08wUzTxOfzWVmkhYUFYrEYQ0NDDc/my5cv4/f7+eY3v9nwnoFAgOXlZX75l3+Z559/nhdffJF33nmHF198kUAgwMc+9jEA/uiP/oi/+Iu/4D/8h/9Ad3c3n/3sZ/n4xz/O1772NVwu166NcYCNcRDw7BB2a4PtdGCtNpw0DMO6gYV1g9AKEaTNTmpRdAqm4sAZ6aW2PLfmd85Y/5ogwb7j3UoLcisB0EA8jO70YxYzOBQZPbuMaeg4wj0Y1TKYJuXZq7gHj4A/tuoNHCgeP7VMcu17+0I7Ui+WZZly8hbFQh7tzvctArzK/A3UnjG6wy5+/MkhylUDSQK/W7HEBDsBVVUJBAIN5rLi+myneN97xeF8dUApOoSEftLNmzf3nRjidiAy1YlEwiLYbwfrBZSlUolr167xK7/yK4TDYY4dO8bhw4d57rnn1gQ7UBcHHB8fb8rl+bM/+zMcDgcvvvgiqqoyPj7O7du3+dM//VM+9rGPUa1Weemll/jkJz/J+9//fgD+8A//kGeeeYZvfOMbvPDCC/zVX/1Vx8c4wOY4CHjaBBHs7LQDS5Zla8EWrbiFQgG/30+1WuXatWtrMhb7IeVvmiauvkNUl2appYUjr4S7fwIllEC3BQp288uten+tRrMAqFar4Tj8GOm3v41RLqCXCyhuH96jT1BZmr3zgQ30fAYlEG8IYkxJxj10nMriDGatbB1XA1HUSD/GDgKearVKqVIF0ySwWsRSVuBO6cOlSrjUexP0KoqyJgBa7blkz1C2wlETi/5+Id+vRlUzyZU0NM3A53HgdytslnETAqGrdWY2mk9x3+/HDY+AaZqkUilyuVxHLCLsAWV3dzdf+MIX+Na3vsWFCxf46le/yuc//3nC4TCnTp3iH//jf8zjjz8O1DM84+PjTcc8e/Ys73vf+xqyy6dPn+ZP/uRPSCaTzM/PUywWeeKJJ6zf+/1+jh8/ztmzZ3nhhRfaMsaZM2c2HENYRxxgfRwEPDvEysoKk5OTuN1uK1PTjt1pPp9nbm6OQCDAxMSEtWiIHbZwARcci3tNMm0FhsuP/8Hn0LML6IUV1FAc2R9BV+oBjb3zrFN6Qoqi1BVfe0fpC8fRV1JoKwuYioPszA0KyUlURUF1OAi4fOi63rBgm6aJFIwTevR5ypMX0MsFHKE4roGjmC7ftjwbRDYrnU4zOnAYafEmpmbr/JMk3P2HMSVlz3lCNPNcsjuYrw6APB6PtWCLxS+bzW6rTCvLMppuosj3zhcsXzJ4+e0ZZlJ5DLNOID99spfRHj/rBT12TaHVJaxm8ykCoOXlZcvDyh5Q7hchQk3TmJmZQZIkRkZGOv65c7kcAB/5yEf4rd/6LVRVZWZmhrNnz3L27FkWFxet1165coXh4WF+/ud/nsnJSUZGRvj1X/91nnnmGZLJJBMTEw1jJxIJAObm5iwfqtUBRyKRYG6untHe6Ri6rpNKpTh06BBf//rXKRQKlpeaGGO/8912A/vjTtnDEJL+uVyOZDKJYRg4nc6GIGQrAZC4sAuFQlP9jdU7bDsnYHZ2toEU6fP59pTOimmamA4PcmwY+U6lSGREROdZtVrdFT0h0wTT5UfpCSK7/Ky89bc4alVCoRCapuFMjJDTZeavXm3OWQnE8Zz4B2BooDjr57GNRVeoYxuGwdDQEA6nk8DJZyneeButsILi8uIZOo4S7dtR9mi3YM9QQuOCLawGBA+jWCxu2+C1WDG4PpNhKpXD53ZwbCRCIuyuq2/fgb3c3AmYpsSbV1JMJXPWsWyhyitvzxINjRFsIhMgnNw9Hk9LmkKrS96maVrzmclkmJub61jrdjtRKpWYnp7ele4zwQVbXFwkHo8zODhovd/AwAADAwO88MIL1uuLxSLT09NEIhH+5b/8l/h8Pr7yla/wq7/6q7z00ktWZ60d4mfhdWc/Zn+NaGjZ6RiKolAsFvnGN77BN7/5TarVKuVy2SrlFQqFPfOc38s4CHh2CEVR6O3tpbe3F9M0KRaLrKyskMvlmJ+fxzAMXC5XSzsykbXx+XwtuyA3I5mKDFAmk0HX9T0nNLd6ARLZrGAw2BJ5sZ0wDAM5lCD06I9Rnr6MUS0TSAzhjA+hq2784VgDB6ipEOI2F1RhBGkv5RiAkhgjEOoGvQayiuny7YtgpxmacdSWlpZYWlrC4XBYXkl2mYbNrvtKzeR7b84ymcximlCpGbx1ZYGffGaMgbgPr0smV9JIpUuUyhrdES9dfieK3N45LFR0Zhbya44XyzWWVyqEvL6Gaz2TyVi2GNvlrdh1fmB3xRC3A3vpbje6z4ScQaFQYHh4mFgstunfeL1eXn/9dRwOhxVwnDx5kuvXr/O5z30Ot9u9phNX/Cyy6uKYPWCpVqvW97TTMUSA63K5+OM//mPC4TBdXV38wi/8AgDnzp3j6aefbn2i3qM4CHjaCEmS1uzG8vk82WyWXC5HJpOpczDuBECiDT2ZTPJHf/RH/NRP/RQPPPDAjrpTVgdA9i4Gu9KuWGDupdeSkJDP5/P3VE3WgDsZmwQYBqYkW5yizUjQ2/ECszt9NyvlGIYBDk/9H+y5MtZ2YT/v4eFhPB5PQ8YinU5vWrKRJInFlZIV7CyulFleKWOYJt86c5snTvbSF/Pz3TemWM7WOVaKLHHqaDcPTERQ2nipS1J97GZQbG9kGAbz8/OUSqW2Zy8lScLj8eDxeIhGow1iiIVCgVQq1RYxxO1AnHe5XN4xF68ViK4v0zQ5evTolkqkzZ49hw8f5tvf/jaDg4NWyUlA/NzT02MFtalUquHZnUqlrDJWT0/PlsYQXb5ijEwmQ7VaZXBwkIcffhioB9DpdF0L7NSpU2QymR0RwN8LOAh4OghJkhrKT4ZhNARA6XSa73znO/z3//7fGRoashaBdmJ1F4PdamBpacnyoRIL9m7tBkulErOzs5Yexl4gYhqGCUgbBhg7DYBEScPtdrecxbsfIOQFXC5Xw3mvl7EQGUpRshHXqN/vJ5Ov74wrNcMKdgDS2QpOVeHlt2ZYzt41ZdUNkzcupxjqCRAJtK/c43erjPV38fqlxo69cMBNJOCygg9x3rvRYSlJEm632wqqVvuriXu+01nfarVqiYTuhh1ILpdjdnYWr9fL2NjYlsp6b775Jr/yK7/Cn//5n/Pggw9ax8+dO8ehQ4d44IEH+MIXvoCmaVbw/eqrrzIyMkI8HicUCuH3+3nttdcYGxsD6lnrCxcuWBmYxx57rOUxxHzZxxDK20KM8zvf+Q6//du/TU9PDx6Ph6997Ws88sgj/PRP/3S7pvS+hPLiiy++eK8/xHsFQkdCmIj+x//4H/nqV7/KL/3SL/Frv/ZrVhYmn89b6U7RrtouKIqC2+0mEAgQiUQIBAJIkmSRTJeXlymXy/VSjyy3/QEt6uuiq2A/iSc2gyzLOJ1O/H4/4XCYcDiMw+GgWq1aqfxCoUCtViObzVoljXg8vq/Pu1WIksbs7CyxWGzT85YkycrwhEIhotEoHo/H2iwsLS3hcLm5dHOJfFknX7pL7h7tCzHaF+K18/MoitzQrm+YJvGwl3hXO7lhJpGQB5Ao3PkcPTEfTz/UTzjgIJvNMjMzQzQavWff9//P3nlGR1Wubfiamt4r6SGFBKRJiXJQEQLHD1EUpAgoCHZQWkBFBaSKiIiAiIIHVAQberBwUKzHQiCgghRDTZ9JQspkMr18P3L2doaEPhMSnGstl4vJ5J2Smb3v/bzPc98SiaTRdz4wMLDBAsFgoKqqioqKCvR6PRaLBYlEctnHHK1WS1FREaGhoURGRrr1dQumpGq1moiIiEu6iAgPD2f79u189913pKeno9frefXVV9m5cycvvPACPXr0YMOGDRw/fpzk5GR++OEHXn75ZaZPn05mZqbYX7Nu3TqSk5MxmUzMmTMHq9XKs88+i0wmIykp6YLXaNu2LRqNhnvuuQelUsm8efOQSqW8/fbbWK1Wtm/fzttvv02fPn04fPgwM2bM4JNPPsFoNDJgwAA3vdNXBx6n5SuAzWajb9++hIeHs2TJEnEc0mKxOFWADAaDWLJ23AJzVwVGuBoUKkD19fVO/QCXO4FmNBopLS0V+55aYnOlq7FardTV1VFZWSmay53PCfpqwR0J53a7Hb3Ryi9/lJF3uIJCVS0SiYTgAG9uvzEFJBJ++7MCm93eyJ/o/3olkxDp64YmZgl1+gazUR8vGQoZqFQqtFotsbGxLT765Xzu2hf6GRXM/aqrq4mNjXW7OarQr6PT6UhMTCQ0NPSS11Kr1SxbtoyffvoJjUZDhw4dmDZtmji2fuDAARYuXMjBgweJiIhg3Lhx3HvvveLvW61Wli9fztatW9Hr9aJLcnx8vHifi1lDq9ViNpvp3LkzixcvJjk5meeee453330XqVSKVColKiqKcePG0aVLFx577DGmT5/uqfCcB4/guUIcPXr0vNMpZrNZFEBarVYUQMJ4b3MIIMHKXRBBQhPqxUyEOIZARkREEBwcfMUbp5uLMxuTHZ21dTrdJfUAtQYcE86joqJc/pqMFjvF5Tr2HilHIZeQGBVAfkE5dpuNjLbR/PRbIfyvWiGVSokM8eWf1yXio3Tt507w3RLcss1mM8XFxcjlcmJiYlrllqWjueSZcS1nM0O0Wq2UlpZitVqbJfDUYDBQXFyMRCIhJSWl9TnPO+C4zeXIW2+9xRtvvMF1113HwoULKSkpYf78+Rw5coSXXnqJqKgoTp8+zapVq6iqqmLNmjXExsZegVfQevAInlaE2Wymrq6O2tpaMTlcKpU2mxFhU+nQMpnMqQn6zC+uMHZttVqJiYn529jmOzboxsTEnPWAfGYWWGsXQI4O2e5OOJdKpdQbrBwtquFYUQ0KhZQObcOICfflVGkt+4+WU68zEBMRQOf0SHwVNry9vV3Wp2ay2KmsNaI6rSPIX0l4oILqyjJx6+hqEfWOZohCALKXl5dT9p9KpRLFrbtft0ajEadZm4qbaQ0UFhYSFxfn9N0+fvw4KSkpWK1WUSg/99xzfPnll4wcOZLHHnuM48eP88QTT1BcXIzFYiEkJISwsDDWrl1LUFCQ2OzsoWk8gqcVYzKZxO0voe9HJpM5CSB3jqELAkg4Wet0OqcGU6vVSkVFRat10L1UHBuTo6OjL+oq/8yr69YkgCwWi9NVfnMlnNvtEvQmK3KZBC+FVDzoG802LFY7CpmNeq1W/JwClz21ZLLAL3+o+LOgCgCDXk9wgJLbbmxHSEDLT3a/HBy9lTQaDSaTSYwgcacZotCvI8TNxMXFtcpjyvr161mxYgXvvPOO2CS9cOFC3n77bb788ksSEhLEqo9Wq2XatGkcOXKE6dOnM3jwYMxmM7/88os4wn7jjTcCOAklD03jETxXEUaj0UkAmc1msQIjnDDdOZIqTNhotVpqamqwWCwoFAr8/f1FEXQ1fyEdqxtRUVHiZNzl0FoEUH19PaWlpc1iLHc5nDm1pNPpLnpqSSKRUFyh4/OfTmKz2UTTNz9fP7q0iySrfSSuCnNtqThGRMTGxooeZEKOlavNEAUxrdfrSUpKEsNDWyMHDhwgJyeHyMhIXnjhBdq0acNPP/3Ec889R0REBOvWrcPHx0cUMEeOHOGee+4hNDSUefPmkZWV1WhNj9i5MDxTWlcRQnUlJCSEqKgoQkNDUSqVmM1mamtrOX36NDU1Nej1ejEywZVXYhKJBJPJREVFBQEBASQkJODn5yc27goHSJPJhN1uF/srrgbMZjMlJSUYjUbi4+Nd5ikklUrx8vJymgKTy+VNToEJEznNKTaEqbvy8nJiYmIICQlpsWIHzj61JJFI0Ov1oimi49TSme+pVCrlyKkaitW1aLValEolvr6+/7uPhLaxgWf157kasFgsFBcXY7VaSUhIQKlUNjlZJ/iQVVRUUF1djcFgEI87Uqn0gj8nBoOBwsJCbDYb6enpjcwLT548Sd++fQkLC6NDhw4AFBcXM2PGDObMmcPGjRupqqoiKyvL6XizadMmpk6dyrJly/juu+9o166dU7SDu9YYNGgQSUlJvPXWW5w+fZrevXuTmJhIWFgY7733Hmq1mr59+4rvT0REBCdPnuTw4cP89ttv9O7du9FW8dVyHHU3rW/z08MFI/hxREZGitUXoQJUUVGB1WpFoVA4bYFd6jaEzWYTBU2bNm1EnxrBFA2cbfGFXCClUunUBN0av7hNOSa7i6bCOx3Ty5uzAnS1JJwrFAqCg4NF0zbHClB1dbVTXIvQtOutbKhqOSbaA/j7yJHLpFytFR6dTkdJSQnBwcFn/aw7miGGhoY2Gn6oqKgALmxbsba2VswUbGrIw2w2k5OTI25VQsNW/4QJE0hOTmbLli0UFRUxa9Ys5HI5U6dOBWDr1q0sXbqU+fPnk5mZybp167j//vvZvn07YWFhbl9j4sSJTJs2jUWLFtG2bVseeOABbr75Zh5++GFefvll0tLSGDt2LABVVVUUFxdz3XXXERcX12Siu4cLw1Ph+Zsg+Jv4+/sTGhpKVFQUwcHBYrWgurqaqqoqamtrRR8eIWjzfOj1eoqKipDJZMTHx5/VSdbRYyU4OJjQ0FC8vLywWCyiLb47PYhcjeAkK4zhXonpsytVAaqrq6OoqIigoCDatGnjknK6VCpFYtIhMRuQYAfZlRFQZ1aAgoKCkMlkom+NWq0mNDiIilozJrNdFJRKuZRenWIJ9Lv6riOFSUuVSkWbNm0uqpInePv4+PgQGBjo5P91ZlVt165dHD16lJCQELEqHB0dTWJiYpOfsRUrVlBVVUVJSQl9+/alQ4cOfPHFF3z88cds3ryZ2NhY2rZtS3BwMKtWrWLcuHHI5XKmTJnC4MGDGT9+PKGhodx8882899572Gw2evbs6dI13n33XWJiYpzWmDBhAmlpaWg0GjZu3EhaWhoZGRkkJCRQU1PDhg0biI6OxmAw8Msvv7B//36eeeYZBg0ahFwu9zQnXyJX3zfTwwXh6HArWJsLTYiOOWBCub6pKSyj0UhlZSX19fWX1LPSVM6SMLFUUVEhJm0Lj93UOOyVoqU6Jru7AnS5CednQ2q3YSk9ir7gIFa9FplvIL5tOyMLT8CG+w7swkj5uVAoFCgUCrGJNCQkBH9fOdk9EzhwrBJVpZagAG86p0cR7CfBZrNdVScjweHXaDS6JCJCIpGgVCpRKpWNqmq5ubl8/vnnGI1G4uLi6NmzJ71790YulzdKEt+zZw/vvfcen3zyCX369BFvz8vLIzMz0+l4lJWVhU6n4+DBgyQmJnLq1CmnXhiZTEa3bt3Iy8tz+RrCazxw4AAFBQXodDr27dtHr169mDhxIidPnmTBggXExsaSmZnJ/fffj0ajYdasWQQFBWEwGJg1a5bo1yaYwnq4eDyCxwPgnAMmBKHW19eLAqi0tNQpB+zgwYM8//zz9O3bl5ycHJdsZzQlgISTdXl5eaOTtTs9iM6GOxqT3cmFCiBBVJ5LAAlxAXK53KVxIBKJBFuNmrpDP4GtIe3coqmk7sAPBPcYCP5hLnkcR6RWM3a9BpuxHplvIPgEYqPp1+34N3cMwPT3h+iwAHRGM1it6PVaSksqMJvNLbKx/FIwmUwUFxeL0Rjueh1CBE5OTg7Dhg3j5MmTqNVqfv/9d+bMmUNdXR2dOnXi/fffRyKRoNFomDlzJs888wxt2rRxWkutVhMdHe10W2RkJNBgCClUoJu6z4EDB1yyhtVqdVrjjTfeYNWqVeLvLV68mBdffJF27drx1FNPMXHiRBYuXMhLL71ESkoKq1ev5osvvsBkMtGxY0dR7Njt9lb7WWoJeASPhyaRSCROmVHCNEpFRQWvvvoqn332GdnZ2QwePJiqqqoLTrq+GKRS6Xlzq840RHOnABI8hWw2W7OEIbqDSxVAtbW1qNVqwsPDXd6YLMGOvvhPUewI2K1mDKrjeKdHNASqugipWY/uz18wqgvBbkMiV+Cb2BFlYkdsUufPr2CoZ7FYztKnZMfXSw7I8fX1IiwsrNF7eiHGfS0RrVYrRoI0RzO60K8TGBjI4MGDxWqy1WolPz+fyspK8TnMnTuXLl26cNtttzVax2AwNBoaEL6rRqMRvV7vdJvjfYTt9EtdIz8/H71eL26B+vn58corr7B582ZmzZrFtddey6BBgzh27BibN2/mkUceITMzk6effprp06ezevVqpk2bRlBQEAMHDhTXFao6V1Pl8ErgETweLgipVIparWby5MnU1tby+uuv07lzZ+rq6qirq6OqqsGPxPFk6eom5KaCO4UTi0qlEi3xHU/WrjpANGdjcnNyPgEkmFsCYsihy1+73Y7damn6R2aTSx9KIpFgKjuGUXXqr8ewmKk/8RuKsBgIjBRvF9yi/f39z+n5IpUAxnqQSLErfZt8T4VmfcetWsem3ZYkgITJu5qaGpduW57r8dRqNdXV1bRp04Y2bdo4vdcymYzMzEzx35988gl5eXl8+umnTa7n7e0tChcB4d/CccHxNsf7CK/1Utb49ttv+eGHH8StN29vb+rq6vjpp5+YNGkSd911F0ZjQ6htZGQkH3zwAZGRkYwdO5Z+/foxZcoUnn/+eWJjY5kwYYLTxWNL+ny0ZjyCx8MF89NPP9GhQweeeuop0QdD2NIRcsAEAXT69GkxudmxB8jVQaiOJxbHTKCysjLMZrOTB9GlmDA6OiY3x8H/SuP4ngr2StzmjAAAIABJREFU/XK5HG9vb2pqaigvL7/gLbALxS6R4h2djKmi8IyfSPCKbuvS7CuJ1YSpoqiJJ2HDVFmEV3A0VqtVjEI5n1u01KjFcOoApqpSJFIZXtHJeMVlYpX9deV/plB33KqtrKxsJIB8fHyuWE+YY0TE+aJvXIEw6WcwGEhJSRH7Xc7FRx99xOnTp536dgDmzZvHhg0b6NGjB4cPH3b6WXl5OdCwBRUTEyPe1q5dO6f7CGIlOjr6otfo1q2b2O8l3G/v3r0UFhYyatQoZDIZ3377LQB9+vTBZDLxxhtvEBcXx//93/8xatQojh07dtYGbQ+Xj0fwuBmbzcaqVav44IMP0Gg0YqhcYmLilX5qF40wJtkUcrncabTXYrFQV1cn5oAJAsidQahyuZzAwEDxBCUIoPr6empqarBarU4VoPMJoJbamOxuhITziooKIiMjnao6l9MDdK7Hk4fH45PQHkNJPnarBYlciW9Ce6RBEdhc6Y0qkSI5y/SXVOElnvAvpEFXajVRf+QXJwFlqavCbrXg1bYbZ9uEO3Or1tG5WJhWEnrl3LFVfDYEgXu+ipar0Ov1FBcXixWcs013nsmLL76IwWBwum3AgAFMmjSJQYMG8dtvv7F161Y0Go14LMjNzcXPz4/27dujVCpJTk5m9+7d3HDDDUDD53rv3r2MGDECgB49elzUGjabTQxX1mg0VFZW0qNHDz766CP69u1L9+7dOXLkCE8++SRyuZynn36awsJCPv30UzZu3EhRURGPPvqomIzuwT14BI+bWb16NZs3b+b5558nKiqKZcuWMWHCBD7//POrOldKLpeLY9LwVw6Y0ARdWVnp9hywMwWQo79KVVWVk8Ouowt1a2tMdiVWq5WysjJMJhOJiYmNPqOubIJ2elyZEu/U7njHZWAz6ZB6+YFPwFkbiS8Vu1SOd3wGptMlYP9Lkki9fJCGxnLy5El8fHwurEHXoMVUWdL45rLjeMVmgNeFmU+ebVrR0a9KoVA4bRW7WgAJPVrN9Xmvrq5GrVYTGBhIUlLSRVWSzpzWEggNDSU2Npbw8HBefvllpk6dyowZMygtLWXZsmXcd999ooAdP348CxYsIDk5mU6dOrF+/Xrq6+sZNmwYANnZ2Ze0ht1ux2azceLECbKzs4mOjqa6upq6ujpmzJgBwMMPP4xSqSQ/Px+5XM7p06cpKiryNCQ3A55oCTdiMpnIysoiJyeH0aNHAw2NgL1792bu3LnccccdV/gZXjlMJpNTBchkMiGVSp2uat0ZhAp/CSAhY8lms+Hj44PJZEIikRAbG3tVi9IzcexZiYyMvKSD79miMFy5BXa5SO1WLOoT6AsPYTPqkPmFoEjoQLlBgp+//wVtqwBIalXU7Pmi8e0KL0J6DcUuVzT09Uikl7Ut52jYKUQ3CH5Wl5tdJfTPaLVa4uLiLrjKcqnY7XZUKhU1NTXExMQQHR3tku94u3btWLBggShYCgoKmDdvHnv27CEwMJChQ4cyefJkp8/em2++ycaNG6mpqaFDhw48/fTTolPzpa4xffp0HnvsMYYNG8bUqVMpKCjgueeeY8+ePVgsFnr27Mm//vUvNBoNK1aswG638/jjjxMaGnrZ74GH8+MRPG7k999/Z/jw4XzxxRfiWCHAqFGjaNu2LQsWLLiCz65lYTQanQSQkAPWXEGo8NdVp1KpxGq1Ajj1H7kqZbul4c6E85YqgKRSKRjrsVst1BlMVNfpiIyMvCiBKzVqqd3zOTZDvXibRK4koF1P7FYzxopiJHIl3vEZSIOjmqxWSSQS0WXIDhckjATXdEGoO2ZXNeWXdTaEOBSZTEZMTIzbt82ExzOZTCQnJ7fKymlxcTGPPPIIQ4cO5ZZbbnEaS7fZbFgsFubMmcOff/7JG2+8QVhYg6WCyWSib9++JCQkcP3113Pq1Cl+/vlnFi9eLPYiefKw3I9nS8uNqNVqoHEJNjIykrKysivxlFosXl5eeHl5ER4eDvwVhKrRaKiqqqK8vNxtQaiOjckJCQn4+vqKIZP19fXU19dTXl4uehUJj99aoxQcEUIZ3TVq764tsMvFZrNhtEkpKa3Ay8uL2NjYi+898vbHP70H2iO7sJkaekp8k67BVH4KY0WxGC5hrCgksOONSCOdG7ClNgu2mjL0JUdBIsE7Nh1pcDQ2yblPeo7RDfCXANLpdOJo9/nCO4WIiJCQEMLCwtwu5IXHk8vlZGRkuL2S5C7y8/Px8vJi6dKl7Ny5k/vuu49+/foBDSJaqVRyzTXX8NVXX4mj62azGaVSyZIlS5g5cyZqtZqgoCA2btxIenq6uLZH7LgfT4XHjfz73/9m5syZHDhwwOlEMnPmTMrKynj77bev4LNrPdjtdicBpNVqsVqtTgf1S80Bc2xMjo6OPutB58w8IJ1OJ27BOVaAWhMtIeH8SlWANBoNKpWKiIiIy4oEkWIHQx0WTSVSuRdSuZyaPdux251blhVBkfh37Y9N7iX+nqlgP/VH9yHmbkkk+Kf3RJHQAdtlHJWF74vwWdXr9U7bxWazmerqamJiYsTGaXciVE6DgoJISkq6Kk7smzdv5s0336SoqIj77ruPgQMH0rFjR6Dhe5Wdnc3o0aOZNGkSgBgFUVVVhdlsJjw8HJlM5qnqNDOeCo8bcfRqcDwZO/o9eDg/wnh7U0GoGo2GiooKVCqVGIQqNICeS4BcbGOyRCIRq1AhISGiAKqvr6eurg61Wt2oAuXukd5LxdFnxTHo9UpwsRWgy53sE0JutVot8fHxYpXkktdDAt6ByHwaJtms6hPY7f8Lw7DbwGpuuJ9eg8RmAf63ZWasR194CKeQUbsdXeFBgiKTLrjhuSkcvy9CeKfRaBSNQy0WC3K5HI1GIwajumO7Vqic1tTUEBsbS1RUVKveEnYULUOHDqVfv35s2bKFt99+mx07djB58mT69u1LQEAAPXv25NChQ2i1Wvz8/MTX7dir4xE7zU/LPCJfJQiW5+Xl5U4nlfLyclJTU6/U02r1OJb0o6KixKZOQQCp1WoxB8xxDF0QQEeOHKG8vJyEhIRL3sZxFECOJxUhj0ylUiGXy50ma1qCABL6KKRSabP4rFws7hRAjtEYrrYZsNvt2O12ZH5BSGRyMOow16ixGRtSvL3jM7FbTEi8/Bu2taxmbEZ9o3Vsei3YrS57XtDwWRXcsv38/IiKinKybBC2ax3F+uUKILPZTHFxMWazmdTU1FbZr9MUJSUljB8/nnvvvZfRo0fz+OOP06VLF7Zt28YTTzzBzTffzOTJk/nnP//Js88+K4qkpsI+PWKn+fHMwLmRjIwM/P392b17t3ibVqvl0KFD9OzZ8wo+s6sL4WAdHR1Neno6Xbp0ISMjQ7T5V6lUHDt2jKNHj/LKK68wZswY9u3bR2Jiost6VoSr6tDQUOLj40lPTycmJgaFQkFNTQ3Hjx/nxIkTqFQqNBqN2BTdnNTV1XHy5En8/f2Jj49vcWKnKQQBFBUVRXJyMqmpqYSGhopZRfn5+RQUFFBZWYlOpztr029dXR2nTp0iMDCQuLg4951svAPwic/AUlshih2ptx8+iR3QnfgNifl/IkemaBi9P/P1+gaA1LXPTXjtISEhYrK9UKmMjY0lLS1N7F3T6XQUFBRw7NgxSkpKqK6uxmg0XtSUmU6n4+TJk0gkEqcATrVazbRp08jKyqJr1648+OCDHD16VPy9w4cPc88999ClSxf69OnD+vXrnda12Wy88sor3HDDDXTu3Jnx48dTUFDgdB93riFMbtpsNv7880/x/jfeeCMvvvgiCxcupK6ujhEjRlBUVITVauWTTz4Rf9fDlUc2d+7cuVf6SVytyGQydDod69atIzk5GZPJxJw5c7BarTz77LMehe8mhDTmgIAAwsPDiY6Oxmg0Mm/ePL755hsmTJhA37590Wq14sFcJpO5tFdEIpGIo8NBQUGEhYXh4+OD1Wqlrq5OTBw3mUwNxntyuduadYWthaqqKuLi4twTD9FMSKVSvLy88Pf3F32e5HI5JpNJdIKur6/HYmmIqpDJZFRUVHD69Olmee12iRSlXzBy/2CkXr54tUnBr10Wpmo1pooifNqkYFf6IJErkUklmKrKELe1pFL8Uq5FGhrrmudit4uvPT4+noCAgCZfu0QiQS6X4+PjQ2BgIKGhoeJ99Xo9p0+fFg0RLRYLEokEmUzW5FpVVVWUlJQQHBxMSkqKWFW12+2MGjUKk8nEiy++yOjRo9m7dy9r1qxh5MiRaLVahg0bxrXXXsvixYtJTk5myZIlhIWFiaPiq1atYvPmzSxatIhx48axa9cuNm7cyIgRI5DL5VRVVbl0jQULFtC2bVtxjbS0NGQyGRaLhU8++YTs7GyCgoLE6k379u256aab8PX1ZcuWLdTU1HD99dfTtWtXl/w9PVw+Lf8Sr5Xz+OOPY7VamT17Nnq9nm7durFu3bpWGTzZWjlw4AAPPPAACQkJbNu2jfj4eLRarWiCWFNTg91ub5TE7kpB2tRkzZnmckql0uWP766E85bCubbAVCqVmAUmGGA2tbXgaux2G4aKImS+QdhtFupP/NaQFSaRNPwH2ABFbAZBAWEYy44ikUjxapOKJDDcJc7SVquVkpIS7Hb7RW9dChcMSqVS9CQSJhbPNO3cvXs3vr6+dO/eHZPJRG1tLXFxcURGRjq9z5WVlaSkpPD444+TnJwMwKOPPsrgwYPJz88nNzcXhULB3LlzkcvlpKSkUFBQwOuvv87w4cMxmUy8+eab5OTkcNNNNwGwfPlyevfuzfbt27njjjt4//33Xb5GUFAQ/fr1E9cAiI+Px2QyUVpaSnx8PPDX5yosLIyHH36YrKwsTp8+TXZ29mX/LT24Do/gcTMymYycnBxycnKu9FP522IwGHjggQcYO3aseMXp6MBss9mcXKCFIFTHhGtXBzw69kyEh4djs9lEbxXhavpyAybdmXDeUhEEkEQioba2VqysOfYAnZlc7vL3xcsXuV8wRvVJp5sVwVGg/GtYwSaVIQlpg09oQy6TzWbDFSOzQkREQEBAI+FxqSgUCqfoGMG0Mzc3l++//x6j0UhCQgLXXXcd//jHP+jRo4foQQMNwbPLly8X/11ZWcn69euJjIwkPT2d1atX0717dydhlpWVxZo1a1Cr1ahUKnQ6Hdddd534c39/f9q3b09eXh533HEHeXl5Ll1j37595OTkUF5ejtlsZu/evXTr1o3+/fuzdOlSPv30U7KyspzeXyHV3LGq42lObjl4BM/fmLVr1/Ldd9+xefNm8bbi4mLmz5/Pnj178Pb25s4772Tq1KlOB5FNmzbx5ptvUlFRQWZmJk8//TSdOnW6Ei/hgujZs+c5e6akUilBQUGNglDPzAE7MwnelSdKx7Fh+CteoL6+vlHApFABOpsAstlsqFQq9Ho9CQkJrdbz5FJwnECLjY0V4xrO1QTtagFkk8jwTe8B2DGdbqiyKIOj8WvXE7vCGxwqOEKzs6sQtvVcbSB5JgqFgqCgIGbPnk1BQQGFhYWoVCr27dvHU089hU6n4/rrr2fDhg2NfvfJJ5/k448/RqlUsmbNGvz8/FCr1Y0GOSIjG5Lry8rKxODOc3maXe4aQl9Yamoqy5cvZ9euXYwcOZLTp0+zYcMGZs2axd133824ceO4/fbb+f777ykoKHDKRWzqO+kROy0Hj+D5m7Jp0yaWL1/udCViMpmYMGECycnJbNmyhaKiImbNmoVcLmfq1KkAbN26laVLlzJ//nwyMzNZt24d999/P9u3b3e6omvNnC8ItbKy0u1BqE3lKwkn6oqKCgwGQ6McMqlUKvoK+fj4kJycfMVjHJoTwUTRbreTlJTUpC3BhUyBuUIA2bz88e1wI75mfUOLjtIbu0zpUnHj9Hj/69PS6XRNZqC5GsHWoby8nLCwMLp27Sqe2C0WC4cOHUKj0TT5uxMmTGD06NG8++67TJw4kU2bNmEwGBpt8wv/NhqNoolfU/cxmUwAl72GTCYTvz+nTp1i/PjxDBgwgKqqKjZs2EBERATPP/88J06cQC6Xi1t88Fdlx0PLxiN4/mao1WrmzJlDbm6uuJcusGPHDkpKSnj//fcJCgoiPT2dnJwcFi1axCOPPIK3tzdr165l1KhR3HbbbQAsXLiQ/v37s2XLFiZOnHglXpLbaQlBqGcmbFutVrECJJyo5XI5ZrOZ0NDQK2YkeKUQnHwv1kTRnQLIJpWDV8BfN7hJ7Agj4AqFolmM/YTJx7q6OuLj48UqioBcLj9nxTctLQ1oOHb8/vvvvP3223h7e4vCRUD4t2BAKdx2Nk+zy13DbDZTWVnJzz//THp6Ov/3f/+Ht7c3KpUKgMmTJ1NbW8srr7yCl5cXJ06cYNu2bWRmZnrETivBI3j+Zhw8eBA/Pz+2bdvG6tWrnUYy8/LynMZIoWEPXKfTcfDgQRITEzl16hRZWVniz2UyGd26dSMvL69ZX8eVRKFQEBoaKpqImUwmUfzU1tZSUVHhtEXljhwwmUwmCiCr1UppaSkGg4HAwEC0Wi3V1dVuFWAtBUcDSVc4BzdXBchVCG7ZzRURYTKZKC4uxmq1kpaWJr5P56O8vJzc3FwGDRokPkepVEpqaipqtZro6Ghxy8nxdwCio6PFyti5PM0udg2hKiOsoVAoSExMJD8/n8OHD6NSqUhKShK3zBITE4mMjCQpKYnff/+dgoIC9u/fT3V1tXgx5KFl4xE8fzP69u1L3759m/yZcOBxRLh6U6lU4hVSU/c5cOCAG55t60CpVBIeHt4oB6yuro7q6mq35oA5Jpw75kFZrVax5F5WVobZbHYyYbwaBJAg9CwWC8nJyW6J9mipAshR6Dn2KrkTrVZLaWkpXl5epKenX9SkaVlZGTk5ObRp04bu3bsDDZWpQ4cOcdNNNxEVFcWmTZtEF2iAXbt2kZSUREREBEFBQaKnWdu2bcXnc+jQIUaNGgVAjx49LniNhIQE5HI5Wq2WgwcPMnz4cOx2OwMHDqS4uJi6ujpeffVV5syZw549e0hKShKPhampqaSmphIcHMyUKVOora31CJ5WgkfweBAxGAyNDpwXu4/uoSEINSIigoiICAAxBqOuro7Tp0+jVqtF3xNBgFxsz8X5Es5lMpnTJJqjs25NTY0YKeD4+K1JAAlCLyAggLi4uGZ77i1BAFmtVlHEukvoOXJmv058fPxFb5t17NiRrKwsZs+ezbx58wgMDOS1116jpqaGcePG4evry7p165g1axYPPvggf/zxBxs2bGDOnDlAwzFmzJgxLF++nPDwcOLi4li2bBlRUVEMGDAAgKFDh17QGi+99BLh4eFERUXx6KOPYrVa+fDDD9HpdOIacXFxfPfdd1gsFr7++mvmzJnTyDW5V69eBAcHk5ubS1JSkkvfcw/uwSN4PIhczB74mffxZIOdnbPlgNXV1VFRUXHRQaiOCecXesKTy+VOAkgYK9bpdFRXV2O1Wt1SgXI1drud6upqKisr3T6JdCE0twAyGo0UFxfj6+tLTEyM23tHBHElZI+d2a9zoUilUlauXMmLL77IlClTqKuro3v37mzatEn0slm/fj0LFy7kzjvvJCIigunTpzNkyBBxjfN5moWFhV3wGk8//TS1tbV4e3vTv39/Dh8+zGuvvYavry/r169n3rx5lJSU8J///Ifbb7/daQ2JRILJZOLzzz+nqqpKvLDx0PLxCB4PItHR0Rw+fNjpNsc98JiYGPG2du3aOd3nzFFPD01zrhwwQQAJQahnJrHb7Xa2b99OUFAQaWlpl9WYLIwVC/1aggCqr693MpYTHr8lCCChWdZoNF5yBpq7cacAckx4b44tFKPRSElJCVarlfT09MvujwoKCmL+/Pln/XnHjh3ZsmXLWX9+IZ5mZ64heGoJVRmZTMYDDzzAH3/8QUhICE8++SRRUVF88cUXPPfcc6xfv57Zs2fzwQcfcOTIESZOnEhRURF79uyhR48e4rrFxcXs2bOH6Ohorrnmmot5GzxcQTyt5R5EevToweHDh53GSXNzc/Hz86N9+/aEhoaSnJzslA1mtVrZu3evJxvsEnHMAUtLS6Nz585kZGQQEREh+ukcO3aM/fv3M3HiRObOnYvNZnOZoZyAIIBiYmJITU0lOTkZf39/9Ho9hYWFTtlKQhxGc2IwGDh16hRSqbTFip2maCoLLCQkxCkLrLCw8JxZYHa7nfLyctRqNfHx8c0idrRaLadOnUIul5OZmXnZYudKUFNTw7Rp0/jqq6/ErSho6Ec8evQot956K5GRkVRWVnLgwAHatm1LXV0dq1atQqVSkZGRwfTp09m7dy/ff/99gzHk/9ZITk7moYce4quvvrrkqpeH5sdT4fEgkp2dzcsvv8zUqVOZMWMGpaWlLFu2jPvuu088wYwfP54FCxaQnJxMp06dWL9+PfX19QwbNuwKP/urA4lEIvrvtGnTBrvdzq5du3jiiSeQy+XMmzeP8PBwTpw44eTC7OogUEdnXbvd7hQtUFlZCeBkwuhOASKY6UVFRbX61O2LrQApFArKysouKSLiUrDb7Zw+fZqKigrCw8NJSEi44pW9S0UQKC+88AIJCQnk5eUxePBgsScpNTUViUTCyy+/zJ49e1ixYgVff/01K1euZNWqVcyZM4eBAwfi5eVFv379nNaWSCRi87SH1oPE3tyXah5aDE8++SQFBQVOTssFBQXMmzePPXv2EBgYyNChQ5k8ebJTr8Cbb77Jxo0bqampoUOHDjz99NNiOJ8H1/Lvf/+bWbNmcfvtt/Pss8/i4+PjlAMmVAW8vLycBIg7vVjsdjsmk0k8UdfX14tj+I5bcJeLo2N0XFyc2830WgKOAkir1WIymcTqmzsMLs98bMd+nauhN+Wbb75h3rx5qFQqOnXqxFtvvYXBYODHH39k0KBBLFiwgI8//ph33nmHzMxMjh8/zq233gpAt27d2LRpk7iWJyKi9eMRPB48tGB27NiB1Wpl4MCBTf5cyAET/hOcX90ZhHomggASKkA6nU50inasVFwMQv+Il5cXbdq0+dsZuwlVrcjISGQymfi+umsKTGiGttvtpKSkNMuYuzsRPHa+/fZbJk2ahEKh4Oabb3bK89JoNIwYMYJHHnmE22+/HZ1Ox7/+9S+++uor7r77blJTU+nWrdsVfBUeXI1nS8tDq0Cr1fLKK6+wc+dOqqurSU5OZuLEiWKp+WrMAAP45z//ec6fn5kDZrVanQRQVVWVmAN2ITlcl4JEIsHLywsvLy9CQ0Ox2+0YjUZ0Op3YaCtMoQki6FxbM47NucHBwa12S+VSOFtEhDunwOrq6igtLcXHx4eUlBRRnDbHd86dawju5M888wyHDx/ms88+41//+hf33XcfABUVFVRWVnLy5EksFgu///47O3bsYMCAAQwdOhS5XO6JjLjKkM2dO3fulX4SHjycD6F5cN68eTz44IPY7XbmzJlDt27diIqKYuTIkYSEhLBs2TK6devGypUrxQBDaMgAW7x4MbNmzWLSpEmcPHmSFStWMGTIkKtqpF4qleLt7S1GLERGRoqvTwhCraqqor6+HrPZjEQiQS6Xu1RUCGv6+PgQFBQkJpbbbDa0Wi3l5eVoNBqMRiM2mw2ZTIZUKhVP9tXV1cTFxYmp538XzGYzRUVFSCQS4uPjm6yKSaVSvLy88Pf3F+NO5HI5JpNJrArpdDrMZjPAOf+2QtCqSqUiPDyctm3bOlUC3f2dM5lMLl1j6dKldO/enZUrV1JVVUXXrl1JSkqiY8eOJCQkUFRUxLZt28jIyCAxMVE0Pvzss8/4z3/+w3vvvUePHj145plnkEql2O12j9i5yvBsaXlo8VRUVNC7d2/Wrl1Lnz59xNvHjh1LeHg4ffr04amnnuKnn34SKx0ffPABixYt4pdffsHb25t//vOf9OvXj5kzZwINV8r9+/dn6NChV20GWFOYzWanHiCj0ShOijVXDIXgRST0/+j1euRyOVarFaVSSUxMTKuZwnIV9fX1lJSUEBYWRmho6CW//44VoDO3wLy8vDCZTISFhWGxWCgrK6O+vp6EhATRJVygOb5zn376qcvX2L9/P9OnT6eoqIj4+Hi6d+/O9OnTCQ8PZ/fu3SxZsoTa2lo2btxIbGwsf/zxB/v27ePUqVN07tyZwYMHi4/j6de5+vBsaXlo8fj4+PDGG29w7bXXOt0ukUiora31ZIBdBAqFwikI1WQyiUGoTeWAuSOGwtGLKCwsDI1GQ1lZGb6+vthsNk6ePIlSqWy2HqQriTAVVVVV5ZKIiHNNgX3yySe89NJLREdH0759ezp06MCtt97aSOxA83znXL3GN998w/Tp0+nZsyeFhYWkpqayb98+Ro0axauvvkrPnj259957WbVqFdOnT+epp56ipKSEu+66y6nK6xE7Vy8eweOhxePv78+NN97odNtvv/3Grl27eOaZZ/jxxx89GWCXiFKpJCwsjLCwMOCvIFSNRiNukchkMqcgUlfFUAj+MhqNhvj4ePGkY7PZMBgM1NfXc/r0afR6faMptKthq6E5IiIcBdCYMWPo0KEDP/zwA3/++SdvvfUWK1euJDExkZ49e3L//feLEQnN8Z1zRXaf4xofffQRN9xwA08++STfffcdgwYNIj8/n7Vr13Lw4EFSU1MZPHgwKpWKt956ixEjRtC9e/dG2YIesXP14hE8Hlodx48fZ9KkSXTu3JkRI0awc+dOTwaYizhXEKqQp+SKIFSz2UxJSQlSqbSRv4xjhQkaBJBerxc9gAwGg1ubsJuD5o6IsNvtVFRU4OPjwz333ENcXBx2u538/Hx2795NXl4epaWlZ82Ecsd3zhXZfQaDAV9fX8rLy9m9ezf33nuvKIA+//xzvv/+e2bOnElkZCTffPMNffv2ZfTo0dxwww0cPXpU3MLy8PfAI3g8tCr27NnDpEmTiIkxrqS6AAAfw0lEQVSJYe3atSgUCk8GmBtxDEIVpq+ECpBjEOqF5oDBX6nboaGhhIWFnVcsCSPufn5+ogO1sE1TUVGBwWBoNKnUkgWQMIUWGRlJcHCw2x/PsV8nOTmZ0NBQoGF7KiMjg4yMDO69996z/r67vnOXsobNZmu0htlsJiQkBB8fHywWi3j/n376iSVLlnDLLbcwe/ZsysrK6NmzJ/7+/rRv35727duL74+7DR09tAw8f2UPrYZt27Yxa9YsevbsySuvvCLa3XsywJoHYbz9XEGoQg6YYxK8IIC0Wi25ubkkJCRcVr+KMG4s/P2tVqtYASovL8dkMuHt7S0+vjvN+i4GYQuvrq6O+Ph4fHx83P6YBoOB4uJiUdxcrMB353fuYtcQRsTLy8vx8vJi69atKBQK1Go1UqmU6667jo0bN/L+++8DMHv2bAYNGoTZbCY/P5/Y2NgmIzI8Yufvg+cv7aFVsGrVKlatWsVtt93GokWLnPodevTowdatW9FoNGJ6tmMGmFKpFDPAbrjhBuCvDLARI0ZckddzNXCuIFSNRoNarcZms6FUKikrK2Px4sV4e3vz3nvvufQkI5PJGgkgoQKkUqkwm81OPUjunkJrCovFQklJCRKJhKSkpGY5yQrN4H5+fpfUI/Tpp58yc+ZMt33nLmaN3r17I5VKsVqt7Ny5E4PBwNy5czEajchkMqqqqhgyZAh79+6lpqYGHx8fbrvtNgwGAz///DNVVVXcddddwF9Boh7+fnh8eDy0eE6cOMHkyZOx2+2UlZWRl5dHdXW1GKHQrl07tm3bxq5du2jXrh379+9nwYIF3HvvvfTq1Qto2PdfuXIlUVFRKBQKli1bRn5+PosWLWqWK+2/AxKJBIVCgb+/P2FhYURHRxMUFMSOHTuYN28e6enpTJkyBaPRiNFoxG63ix48ruRsXjVGo5Hq6moqKirQ6XRYLBa3+BCdiRDA6u/vT5s2bdzeFCv066jVaiIiIkhOTr7ox1SpVEyYMIFu3bqJwkIQkWaz2SXfucTExAteIzo6GolEQk5ODoWFhbzwwguMHz8emUzGgQMH+O6777jjjjvQarX8+uuv+Pr6olKp+Oqrr3jttdfIzs7mscceA/CInb8xHh8eDy2eZcuW8frrrzf5M29vb6ZPn0737t1ZtmyZJwOsBWG323nmmWf497//zcyZMxkzZgw6nY7a2lq0Wi319fViDphjA7K7qx8Wi8UpB8xisTj1ILlyCq2mpoaKigratGkjjoq7E4vFQmlpKXq9nqSkpEtOVn/rrbdYuHBhkz+79tpr2bx5s0ty9y50jfXr11NVVYWvry/9+/dn0aJF4n1GjBjBH3/8gUQiITg4mH/84x+YzWbUajWhoaH069ePO+64A/CMnP/d8QgeDy2eDz74gMWLF/PWW29xzTXXiMGVp06dYvPmzXz//ffk5ORw1113ecrVLQij0ciMGTOYMGECnTt3bvRzwXnZMQfMbrc3aw4YNIgExxwwq9Xq5EN0KQJICD41GAzExsY2S/CpY79OamrqVVW5/PTTT3n99dcpKipi5cqV3HDDDeK0nlarZeDAgQQFBbFgwQKnz5rJZBJ7yDwxER48gsdDi8ZutzNt2jT279/P119/Dfx14LLb7dTX1zNixAiUSiXr1q0T/WRaEmq1miVLlvDTTz9hMpno0aMHM2bMIC0tDYDDhw+zaNEiDhw4QHBwMPfccw8TJkwQf99ms7Fq1So++OADNBoN3bp1Y86cOSQmJl6pl+QWzhaE6jiB1RwePGazWaz+6HQ6bDbbRY3hm0wmSkpKUCqVzRZ8WltbS1lZGQEBAY3G/FsbZ6vCvP3227z44ov06NGDdevWAX8JmoMHDzJy5EhuuOEGpkyZQnp6OuDp1/HgjEfuemjR1NfXU1BQ4DRNJZxAJBKJ2BdRVFTUpNixWq3YbLZme75nYrfbeeCBB1CpVKxfv54PP/wQb29vxo0bR319PVVVVYwbN47ExEQ++ugjJk+ezCuvvCJOmgCsXr2azZs3s2DBAt577z1kMhkTJkzAaDResdflDoQg1Li4ODIzM+ncuTMpKSn4+/uj0+koKioiPz+fU6dOUV5eLm6JuRqFQkFQUBAxMTGkpqaSnJyMv78/BoOBwsJCjh49SklJCdXV1ZhMJqfnoNVqOXXqlPj7zeGvo1arKS0tJSoqitTU1FYrdux2u5itVlNTwyeffMLWrVv59ttvARg5ciRjxoxhz549rFy5Emjo8bFYLOJ22TfffENxcbG4pkfseHCkdX4zPPxtKCoqorKykvLycjp16kRqair9+/enT58+KJVKPv74Y3Jzc8nOzhZ/p7a2lurqapKSkq74fn1lZSUpKSk8/vjjJCcnA/Doo48yePBg8vPzyc3NRaFQMHfuXORyOSkpKRQUFPD6668zfPhwTCYTb775Jjk5Odx0000ALF++nN69e7N9+3axN+FqRC6XExwcLHrVWCwWMQZDCEIVJsUct8BcfZJTKBTi87Db7U4VoMrKSgAxFkOv1xMXF9cs/k7C5JfBYKBt27aX3K/TEhC8cCQSCbt27WLGjBl4eXmhVquxWq307duXKVOm8NBDD1FaWsqGDRtIS0vjlltuEdcYOXIk7du3d0pS9+DBEY/g8dCi2bdvH/X19axfvx4vLy++/vprXn31VVasWEFAQACRkZEMHTpU3ALatGkTX3zxBSdOnMBgMNCnTx8nsXEmVqsVcJ+dfEREBMuXLxf/XVlZyfr164mMjCQ9PZ3Vq1fTvXt3p6vyrKws1qxZg1qtRqVSodPpuO6668SfC8ZpeXl5V7XgORO5XO6UA2Y2m0UBVFdXR2VlJVKp1K0j6BKJBKVSiVKpFAWQwWCgrKwMi8UCQGlpqdMWmDsiI/R6vehUnZGR0Sr7df773/+Sl5fH1KlTxc//gQMHmDVrFgMGDGDYsGEolUry8vKYPXs2BoOBxYsX8+ijj6JSqViyZAkpKSmkpaWJW1uC2PH063hoCo/g8dBisdvt5ObmEhISwj/+8Q8AunbtSrt27Vi8eDHTpk3j+uuvJz4+HolEwvLly1m/fj29evVizpw52O12Vq1axcqVK1mwYAG+vr7inr7w/+asAD355JN8/PHHKJVK1qxZg5+fH2q1mtTUVKf7CXlCZWVlohHbmQaJkZGRlJWVNc8Tb6EoFApCQ0NF52AhB6yurq7JIFRXTmAJGI1GSktL8fPzE/9GJpOJ+vp6tFot5eXl4nMQjBAvVwBdDf06u3fv5oEHHkAulxMVFcWoUaOABh8eHx8f7rvvPuLi4gBo27YtJpOJhQsX8uGHH/LII49w3333sXTpUh566CH+85//NHL39ogdD03h+VR4aLHU19dz8uRJ0QJeaGa84447iI2NZfPmzWJujk6nY8OGDdx22228/vrr3HLLLWRnZzN16lS++OIL9u3bBzRcoR84cICXX36ZkSNHMmHCBD766COxQfZMrFarWAW6XCZMmMCHH37IoEGDmDhxIn/88QcGg6HJrCDw5IBdLEIOWHJyMh07duSaa64hLi4OhUJBdXU1J0+e5OjRoxQVFVFVVYXBYLisHiCNRkNhYSHh4eGiT4xEIsHLy4vQ0FDi4uJIS0sjLi4Ob29vNBoNJ06c4Pjx45SVlaHRaMSq0IVgt9tRqVSUlpYSHR19zn6dtWvXcvfddzvdVlxczEMPPcS1115Lr169WLp0aaPH37RpE/369aNTp06MGDGC/fv3u2WNZcuWYbfbsVgsrFq1ih07dgANMRa+vr7ExsYCf1Vgx4wZQ69evfj8888BGDBgAHfddRd33XXXeaNMPHgQ8AgeDy2WgoICqqurnbZzhIPr8OHDOXLkCJ999hkSiYSff/4Zo9FIfn4+H330kRhxkJ2dTVRUFL/++isA+/fvZ+zYsXz00Ud06tSJ2NhYVqxYwTvvvNPkyU8mk7msCpSWlkbHjh1ZuHAhsbGxvP32254cMDci5IC1bduWTp060aFDB2JiYpDL5Zw+fVoUQMXFxVRVVV1wE7jQKFxeXk5CQgJBQUFnva8QxxEaGkp8fDzp6enExsaiVCqpra3l+PHjHD9+HJVKdU4BZDabKSwspLa2lpSUFGJiYs5aqdq0aZPTNio0fF4mTJiARCJhy5YtzJ8/nw8//FBs/gXYunUrS5cuZcqUKWzdupXk5GTuv/9+Tp8+7fI1QkJCyMrKws/Pj5qaGhYuXIharaZDhw6cOnWK6upq8f0T3pNevXpRVFTEiRMnALj//vt59NFHxb+JBw/no/XVQj38bfjhhx+orKykW7duQEOZWjjIp6Wl4ePjQ15eHkOGDCE3NxdfX1+io6NZuXIlTz/9tHi1X1lZiU6nw2QysWLFCoKDg/nwww8JDAzEarUSExPDSy+9RHZ2Nm3btqWqqoqvv/6a7777DplMRnZ2Nv369WuU/eR4kD3byae8vJzc3FwGDRok3kcqlZKamoparSY6OlrctnL8HWjIExIeo7y83CkHqLy8vNFWmIdz01QOmNADVFlZeUFBqEajEZVKhUQiuSQHY8c8srCwMPF56HQ6ampqKCsrQ6FQoNVq+f333+nevTsJCQmUlpYik8nIzMwURfCZqNVq5syZQ25ubqOetR07dlBSUsL7779PUFAQ6enp5OTksGjRIh555BG8vb1Zu3Yto0aN4rbbbgNg4cKF9O/fny1btjBx4sRLXiM7O7vJNXbu3ElwcDB1dXX88ssvzJ8/n2HDhvHhhx+ydOlSFi9ejFQqFSMljh07RmZmJm3btgX+6rvzjJ57uFA8gsdDi+WWW27B399fDB8UDmp2u51u3boRFxfHjz/+iMlkwmq1EhcXx+LFiwkMDOTIkSPs3buXP//8E5PJRNeuXTl48CAHDx5k9OjRhIaGYrfbkcvlDBkyhC+//JKSkhICAgKYMmUK+/bt45ZbbsFqtbJ8+XJ+/fVX5syZIz43R0MzAWH83bF/oKysjJycHNq0aUP37t2Bhqv1Q4cOcdNNNxEVFcWmTZucEpt37dpFUlISERERBAUF4e/vz+7du8UDvVar5dChQ2Lfg4eLxzEHTBBAQg7YmUGogvj58ssvWbZsGa+99hqdO3d2yUnW8XkIAkiv17N37142b97MSy+9hL+/P506daJPnz54eXmRmpra5GMfPHgQPz8/tm3bxurVqykoKBB/lpeXR2ZmplM1KisrC51Ox8GDB0lMTOTUqVNkZWWJP5fJZHTr1o28vLzLWqNjx47k5eVht9ud1khOTmb16tXMmjWLn3/+mT179hAQEMDw4cN57bXXCAwM5O6770apVPLbb7/x888/M2zYMMBZ5HjEjocLxSN4PLRYkpOTSU5ObnQFJzQd9+3bl7Vr17Jnzx769+/Pu+++yw8//MDAgQNp164dGRkZ/PzzzwwZMoQuXbqI8RTCFpng+xEZGcnWrVupq6tj+fLlHDt2jNWrV9OnTx/0er14xZmZmcnw4cPRarUsWbKEyspKFi5cyOHDh7nmmmua3Nro2LEjWVlZzJ49m3nz5hEYGMhrr71GTU0N48aNw9fXl3Xr1jFr1iwefPBB/vjjDzZs2CCKK6VSyZgxY1i+fDnh4eHExcWxbNkyoqKiGDBgQDP8Ff4eSCQSsbIjVNZ0Oh0ajYaqqipWrFjBzp07GT16NBEREWg0Gpc0IJ/tefTu3Zs333yTY8eOUVJSQn5+Phs3bmTRokWEhoby7LPPMnDgQKff7du3L3379m1yXaGa6IjQHK9SqcSqUVP3OXDgwCWtYbPZmDp1Kj/++COBgYHU19ejVqvF5u5rr72WlJQUsS8nISGBffv2ERoaSk5ODkuWLOH9998nMjKSyspKhgwZwqRJk8T3yYOHi8UjeDy0WGw2m9gIeiYSiYS7776buro6Tp48yZgxY7j77rtZs2YNpaWlJCcn88MPP7Bt2zYmT55Mly5dsFqt6PV60YUVGq5AhWpNSUkJP/74I127dqVHjx5IpVL8/PwYO3YsO3bsIDc3l+HDh4v9Gz/++CPz58/njz/+oKysjP79+zNt2jTi4+OdnufKlSt58cUXmTx5Mlqtlu7du7Np0ybxfuvXr2fhwoXceeedREREMH36dIYMGSKu8fjjj2O1Wpk9ezZ6vZ5u3bqxbt06T7OmG5FIJPj5+aHVapk/fz6FhYWsWbOG9PR06urqUKlU2Gw2MQdM+M8VE1Nms5mSkhJMJhNZWVlOQrq4uJg9e/Y4fYYvBIPB0GhL9mKb4y92jddff50dO3aIVasXXniB6upqcfoK4NZbb+WDDz4AIDs7m2+//Zbvv/9ezGD773//i0KhoF27dmLlyJOH5eFS8QgeDy2W842WRkdHM3v2bPHfjzzyCH5+fmzZsgWDwUBcXBxjx47lzjvvBBr6foTGZkHQAOzcuZO0tDRKS0tRqVRUVFSQlZVFfHw8119/vVjVEZpaKysr+e233+jRowf9+/fnySef5LvvvuP5558nLS2NRx99VPQBsdlsVFZW8uyzzzJ//nyn52+327FarXTs2JEtW7ac9XXKZDJycnLIycm5pPfRw6WzZMkS7HY7W7dudapu2O12tFqtuAVWU1PjkiBUnU5HSUkJcrmcjIyMRv06cXFxToLhQnFFc/zFrjF+/Hj27NnDoUOHRMf0gwcPAlBVVUVoaCgZGRkcO3YMgMTERB566CFeeOEFVq5cyaJFi5wiVgDRidmDh0vBM6XlodVis9mcRsYjIyPJycnhm2++4csvv2T16tVMmzZNNKrr3Lkz1157LS+88AK//fYbR44cYcmSJUyfPp3du3ejUCiwWCy88847vPXWW9x2220cPXqUMWPGkJ+fL568jh07Rm1tLdOnT2fgwIFERUUxYsQIoqOj2b9/PxqNBqlUytdff83EiRN5+OGHycrKYvjw4Xz//ffi85VIJK3KQ+XkyZN07dpVvCIH14wpt2QWLlzIu+++22grRyKREBAQQGxsLBkZGXTp0oW0tDSCg4MxGo2UlJRw9OhRTp48iVqtRqvVntfeoLq6msLCQvz9/ZsUO5fD+ZrjY2JinG5zvI+wBXWxayiVShYtWoTVakWpVHLXXXeRkJDA/v37eeaZZygqKiIjI0P02IqOjubmm29m5MiR/Prrr+zcuRPAKRrG46/j4XLwfHo8tFqkUqnT1Z7g62Gz2fD39yciIsLp/hERETzxxBPIZDLGjh3Lww8/zH//+18ee+wxRo8ejVwuJyAggKKiIrp3786kSZPYtGkTn3/+Oe+//z4TJ07EYrFw4MABgoKC6Nixo9NJTK/Xo1Qq8fHxIT8/n9mzZ1NSUsLYsWNZuXIl0dHRLF++nKKiIo4dO8YDDzzAL7/80mzv1+VgNpvJyclx8ityxZhyS+dC09qlUimBgYFOOWCpqakEBgai1+spKiri6NGjTjlgwoncZrNRVlaGSqUiJiaGtm3buryK0aNHDw4fPoxGoxFvy83Nxc/Pj/bt2xMaGkpycjK7d+8Wf261Wtm7dy89e/a85DXCw8NRKBRIJBK++uorJk6ciFQqpbi4mBEjRvDhhx+KBqCCmBk7dixr164VqzsekePBVbSey0sPHs7DhVRMOnfuzJYtW1Cr1Rw5coS0tDTxyrRTp0507dqVN954g5CQENLS0iguLubdd9/Fz8+PnJwcVCoVhw4domvXrsBf/QRHjhxBo9HQoUMHFAoFGzZsQKvVsm3bNjHUNCUlhX79+rF9+3buv/9+fv75Zzp16sT1118vPj+TyURRUREpKSluepcujZUrVzbq33DFqPPVikwmIygoSOy/sVgsTltgQg6Yt7e3KNJTU1PP6elzOWRnZ/P/7d1rSJP9Gwfw76M7aB4oDw3TYU/LDiqdNAuzECIjCArCJZWiFkkgkaWhkTnygFDTkoQOUyQCO1BYJvmiiKDA0mFpmWlmomkKRS2blm7+X8TuZ0t7+vvkcft+3vmbu70H03297+t3XadOnUJSUhJSUlLQ2dkJtVqNuLg4oeYmPj4eWVlZ+Pvvv7FkyRIUFRXh69evws6o/3qMwcFBJCQkQKPRwNvbG3PmzMGsWbPg7u6OwsJCvH//HiKRyOLq0dq1awGwXofGFgMP2RRTIbRMJhs2rsHJyQkpKSnIycnB3r17IZfL8eXLF+j1emRmZsLR0RHt7e14+/YtoqKiLJ5bVVUFsViMwMBAdHZ2QqvVIjw8HO7u7hgYGIBYLIaXlxdSU1OFnT0LFy5Ec3OzxXFu3ryJ9PR0nDt3ThgWOtmqq6tx5coVlJWVITw8XFgfi63OtuLnQagDAwPo7e3Fp0+f0Nvbi8WLF0MqlY7bz5dKpdBoNDh+/DiUSiVcXV2xfft2i9BpqlU7ffo0Pn36hICAABQXFwujO/7kGP7+/mhvb0d5eTmUSiWePXuG6upqSKVSyOVytLW1obKy0mIIMDB+M+7INjHwkE0xvzw+UsMyhUKBoqIitLa2Ch/K69evF/7o19XV4fv371i3bp3F8R4/fgyZTIZ58+ahq6sL7969Q3x8PADLLbQxMTEAfnzg+fr6WszDampqwtmzZ7Fly5YpE3Z0Oh0OHz6Mo0ePwsvLy+KxsdjqbKvEYrHFINSxlpubO2zN19cXRUVF//q8+Ph44X07kj85xpEjR9Da2oqKigqkpaWhqKgIRqMRRqMRpaWl7CtF446Bh2zWSNvdTburTD2AzA0ODkIikcDb2xs+Pj5C40IAaG5uxrJly+Du7o7Pnz/DYDAIAcH0X6per4ednR1EIhHEYjEcHBxgNBqFoZd5eXlwdXUVeo1MhYnPKpUKy5YtE25JmRuLrc5kO6RSKU6cOIHY2FhcuHABcrkcCoUCdnZ2iI6OBgCLBpxEY43VYERmTAHD1JTQnEgkQlxcHO7duyd8D/Cjw21HRwdkMhkkEglkMhlmzpyJq1evQqfTCcHq0qVLKC0tFY5raqhmMBhw8eJFNDQ04MCBA0J/nskOO2VlZaipqYFKpRrxcc4Bo9GSyWTIzMxEfX09CgsLR/wdIxovfHcRjeBXDQ/Niyjt7OwwNDSEgIAAlJeXCzUYs2bNQlpaGvLy8lBQUIClS5fiwYMHuH37NmJiYoQrHitWrIBGo0FraytKSkqQmJiI0NDQiXuRv3H9+nV8+PDBom4HAI4fP46SkhJh1465X21TNo0HMX39c/0U2Y7g4GBkZ2cjKCho0kM92RYGHqJR+LmI0nyYqbmNGzfi27dvKC0tRWVlpTAl3byDsoODA4aGhpCeng4/Pz/s3bt3/F/AKJw8eRL9/f0WaxEREUhMTMTmzZvx9OlT3LhxAzqdDq6urgAstylLJBJhm7L5rhutVovt27dP+OuhqcN0i5S7sGgiMfAQjYGfC6ClUikiIyMRGRkJo9EIvV4vTDs31eY4OjrCxcUFg4ODQhflqfQB8KurMG5ubvD29oaHh8cfb3WmsWM0GnHmzBlcu3YNOp0OQUFByMjIgK+v72Sf2i9Nlfc62QYGHqIx8PPtL/MaIHt7eyHsAP/U5nR1dWFgYABxcXHCFZDp9AEwFludaewUFhaitLQUubm5kMlkUKvV2L17NyoqKsZ1yzvRdPHXkKnykojGjekKkE6nw+XLl4WwsGbNmhG3EBONhmnQaHJyMnbu3AkA6O3tRVhYGFQqFbZu3TrJZ0g0+XiFh2gCmK4ANTY2oqCgABKJBOHh4UhNTQUwck8gov/Xy5cvodfrsXr1amHN2dkZ/v7+qKmpYeAhAgMP0YQZGhpCSEgI7t27h3fv3mHBggVwdnZm2Blnb968waZNm4atZ2VlITIyEi9fvkROTg7q6+sxc+ZMREdHW0zpng61Md3d3QCG113Nnj3borklkS1j4CGaIKZQ8/NYC4ad8fXq1Ss4OzujsrLSYt3FxQUfP35EbGwsNmzYAJVKhbq6OqhUKri4uECpVAKYHrUxbPJI9HsMPERk1ZqamqBQKODp6TnssZKSEojFYqhUKohEIigUCrS1teH8+fNQKpX4/v07iouLkZycLIz7yM/PR1hYGO7cuTNlbhWZN3k0Dz1s8kj0D3Z9IiKr9urVq19On6+pqUFwcLBFh99Vq1ahvb0d3d3dv62NmSpMY0zMJ46bvmaTR6IfGHiIyKo1NTWhp6cHUVFRCA0NxY4dO/Dw4UMA/z4Ataura9rUxixatAjOzs548uSJsNbb24uGhgaEhIRM4pkRTR28pUVEVkuv16OjowNubm44dOgQnJyccOvWLezZswfFxcXo7+8fse4FmF4DUCUSCXbt2oX8/Hx4eHjAx8cHarUaMpkMERERk316RFMCAw8RWa0ZM2ZAq9VCLBYLoSUwMBAtLS3QaDSjGoA61Wtj9u/fD4PBgGPHjqGvrw9BQUHQaDTDwhqRrWLgISKr5uTkNGxtwYIFuH//PuRy+Yh1L8CPAaimvqw9PT0W3bJ7enowf/78cTzr0bO3t0dycrIwpoSILLGGh4isVm1tLZYvX466ujqL9efPn8PPzw8rV66EVqvF4OCg8FhVVRXmzp0LT09P1sYQWREGHiKyWoGBgfDx8UF6ejq0Wi1aWlqQlZWF2tpa7Nu3D9u2bUNfXx+OHDmC169fo6ysDCUlJUhISABgWRtz9+5dNDY2IikpibUxRNMQZ2kRkVXr7u6GWq3Go0ePoNPpEBAQgIMHDwpXaOrr65GdnY0XL17A09MTsbGxiImJEZ5vMBiQn5+PGzduCLUxGRkZkMvlk/WSiOg/YOAhIiIiq8dbWkRERGT1GHiIiIjI6jHwEBERkdVj4CEiIiKrx8BDREREVo+Bh4iIiKweAw8RERFZPQYeIiIisnoMPERERGT1GHiIiIjI6jHwEBERkdVj4CEiIiKr9z87KOUHhEHswwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plots_model()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Adding column groups to DataFrame:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "      <th>groups</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>195</td>\n",
       "      <td>51</td>\n",
       "      <td>157807.81</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>64</td>\n",
       "      <td>20</td>\n",
       "      <td>70488.44</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>264</td>\n",
       "      <td>26</td>\n",
       "      <td>94117.26</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>83</td>\n",
       "      <td>46</td>\n",
       "      <td>153996.13</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>187</td>\n",
       "      <td>7</td>\n",
       "      <td>24179.96</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Recency  Frequency  MonetaryValue  groups\n",
       "0      195         51      157807.81       2\n",
       "1       64         20       70488.44       0\n",
       "2      264         26       94117.26       0\n",
       "3       83         46      153996.13       2\n",
       "4      187          7       24179.96       1"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Distribution of the size of individual groups:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    47.777778\n",
      "2    34.444444\n",
      "1    13.333333\n",
      "3     4.444444\n"
     ]
    }
   ],
   "source": [
    "print((df_norm.groups.value_counts(normalize = True, sort = True) * 100).to_string())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Statistics for the whole set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "      <th>groups</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>185.866667</td>\n",
       "      <td>27.9</td>\n",
       "      <td>99335.664889</td>\n",
       "      <td>0.955556</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Recency  Frequency  MonetaryValue    groups\n",
       "mean  185.866667       27.9   99335.664889  0.955556"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.agg(['mean'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Distribution of the size of individual groups:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>mean</th>\n",
       "      <th>mean</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>groups</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>208.162791</td>\n",
       "      <td>22.093023</td>\n",
       "      <td>78633.205814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>323.250000</td>\n",
       "      <td>10.666667</td>\n",
       "      <td>35628.653333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>125.548387</td>\n",
       "      <td>37.129032</td>\n",
       "      <td>133158.014516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.500000</td>\n",
       "      <td>70.500000</td>\n",
       "      <td>250884.925000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Recency  Frequency  MonetaryValue\n",
       "              mean       mean           mean\n",
       "groups                                      \n",
       "0       208.162791  22.093023   78633.205814\n",
       "1       323.250000  10.666667   35628.653333\n",
       "2       125.548387  37.129032  133158.014516\n",
       "3         1.500000  70.500000  250884.925000"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.groupby('groups').agg(['mean'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The sum of the values for each group:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rfm_df.groups.value_counts().plot(kind='bar', figsize=(6,4), title='The sum of the values of individual groups');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "      <th>groups</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>195</td>\n",
       "      <td>51</td>\n",
       "      <td>157807.81</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>64</td>\n",
       "      <td>20</td>\n",
       "      <td>70488.44</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>264</td>\n",
       "      <td>26</td>\n",
       "      <td>94117.26</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>83</td>\n",
       "      <td>46</td>\n",
       "      <td>153996.13</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>187</td>\n",
       "      <td>7</td>\n",
       "      <td>24179.96</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Recency  Frequency  MonetaryValue  groups\n",
       "0      195         51      157807.81       2\n",
       "1       64         20       70488.44       0\n",
       "2      264         26       94117.26       0\n",
       "3       83         46      153996.13       2\n",
       "4      187          7       24179.96       1"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Function to assigned segment name to each value:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "def name(row):\n",
    "    segment_mapping = {0: 'departing', 1: 'new', 2: 'inactive'}\n",
    "    return segment_mapping.get(row, 'active')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Adding column SegmentName to DataFrame:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\PC\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "rfm_df['SegmentName'] = rfm_df['groups'].apply(name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>Recency</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>MonetaryValue</th>\n",
       "      <th>groups</th>\n",
       "      <th>SegmentName</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>195</td>\n",
       "      <td>51</td>\n",
       "      <td>157807.81</td>\n",
       "      <td>2</td>\n",
       "      <td>inactive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>64</td>\n",
       "      <td>20</td>\n",
       "      <td>70488.44</td>\n",
       "      <td>0</td>\n",
       "      <td>departing</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>264</td>\n",
       "      <td>26</td>\n",
       "      <td>94117.26</td>\n",
       "      <td>0</td>\n",
       "      <td>departing</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>83</td>\n",
       "      <td>46</td>\n",
       "      <td>153996.13</td>\n",
       "      <td>2</td>\n",
       "      <td>inactive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>187</td>\n",
       "      <td>7</td>\n",
       "      <td>24179.96</td>\n",
       "      <td>1</td>\n",
       "      <td>new</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Recency  Frequency  MonetaryValue  groups SegmentName\n",
       "0      195         51      157807.81       2    inactive\n",
       "1       64         20       70488.44       0   departing\n",
       "2      264         26       94117.26       0   departing\n",
       "3       83         46      153996.13       2    inactive\n",
       "4      187          7       24179.96       1         new"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save recommendations to a CSV file\n",
    "rfm_df.to_csv('C:\\Python Scripts\\Projects_done\\Customer_segmentation\\customers_segments2.csv', index=False)  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Description of the results\n",
    "\n",
    "\n",
    "**Interpretation of the resulting groups:**\n",
    "\n",
    "From above analysis I have obtained segmentation for four groups. They can be interpret as the following customer segments:\n",
    "\n",
    "- **Group 0**: Customers who were shopping for a long time, buy moderately often and they shop for high amounts.\n",
    "- **Group 1**: Customers who have been shopping recently, they rarely buy and make purchases for relatively small amounts.\n",
    "- **Group 2**: Customers who have not been buying anything for a long time, they rarely buy and make purchases for relatively small amounts.\n",
    "- **Group 3**: Customers who have been shopping recently, they buy often and make purchases for large amounts.\n",
    "\n",
    "Based on the received groups they can be name as follows:\n",
    "\n",
    "- **Group 0** - former best customers - **departing**,\n",
    "- **Group 1** - new and random customers - **new**,\n",
    "- **Group 2** - old and random customers - **inactive**,\n",
    "- **Group 3** - best customers - **active**.\n",
    "\n",
    "## Summary\n",
    "\n",
    "In this approach I tested a second method of customer segmentation. I have used raw calculated RFM variables and K-Means Clustering to determine segments of clients. From above analysis I have obtained the similar results to previous method, i.e. RFM scoring + K-means clustering."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}