hist3368-week9-grouping-data.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hist 3368: Grouping Data\n",
    "\n",
    "#### By Jo Guldi"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One of the reasons why we work with tabular data is that it makes it possible to aggregate information around features o the data.  For example, rather than counting the top words overall, we can compare the top words of each speaker.  Or rather than studying the words for crime overall, we can aggregate the top words for crime of each year and study how they change.\n",
    "\n",
    "When we want to study how the facets of data affect count, we need to \"group\" the data.  Typically we \"group\" data by a column -- for example, year, or speaker -- and then we take a count of the words that result.\n",
    "\n",
    "This Notebook offers an introduction to the **.groupby()** function in pandas. We will learn a few major commands applied to grouped data:\n",
    "    \n",
    "   * **.sum()** -- to add up grouped data\n",
    "   * **.agg()** -- short for \"aggregate\" -- which has similar functions to .sum()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will also learn some important features of the .groupby() grammar, including:\n",
    "    \n",
    "   * the use of **.copy()** to preserve data integrity when executing grouped transformations\n",
    "   * the use of **.reset_index()** to renumber rows when working with grouped data\n",
    "  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Watch the following video:\n",
    "    https://www.youtube.com/watch?v=ipoSjrN0oh0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Load some packages and data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import csv\n",
    "import glob"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/users/jguldi/digital-history\n"
     ]
    }
   ],
   "source": [
    "cd /scratch/group/history/hist_3368-jguldi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "congress = pd.read_csv(\"congress1967-2010.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TROUBLESHOOTING: if the line above doesn't work, you might have missed something earlier this week."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>speech</th>\n",
       "      <th>date</th>\n",
       "      <th>speaker</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Those who do not enjoy the privilege of the fl...</td>\n",
       "      <td>1967-01-10</td>\n",
       "      <td>The VICE PRESIDENT</td>\n",
       "      <td>16</td>\n",
       "      <td>1967</td>\n",
       "      <td>1</td>\n",
       "      <td>1967-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Mr. President. on the basis of an agreement re...</td>\n",
       "      <td>1967-01-10</td>\n",
       "      <td>Mr. MANSFIELD</td>\n",
       "      <td>35</td>\n",
       "      <td>1967</td>\n",
       "      <td>1</td>\n",
       "      <td>1967-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>The Members of the Senate have heard the remar...</td>\n",
       "      <td>1967-01-10</td>\n",
       "      <td>The VICE PRESIDENT</td>\n",
       "      <td>40</td>\n",
       "      <td>1967</td>\n",
       "      <td>1</td>\n",
       "      <td>1967-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>The Chair lays before the Senate the following...</td>\n",
       "      <td>1967-01-10</td>\n",
       "      <td>The VICE PRESIDENT</td>\n",
       "      <td>151</td>\n",
       "      <td>1967</td>\n",
       "      <td>1</td>\n",
       "      <td>1967-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Secretary of State.</td>\n",
       "      <td>1967-01-10</td>\n",
       "      <td>Mrs. AGNES BAGGETT</td>\n",
       "      <td>3</td>\n",
       "      <td>1967</td>\n",
       "      <td>1</td>\n",
       "      <td>1967-01-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              speech        date  \\\n",
       "0  Those who do not enjoy the privilege of the fl...  1967-01-10   \n",
       "1  Mr. President. on the basis of an agreement re...  1967-01-10   \n",
       "2  The Members of the Senate have heard the remar...  1967-01-10   \n",
       "3  The Chair lays before the Senate the following...  1967-01-10   \n",
       "4                                Secretary of State.  1967-01-10   \n",
       "\n",
       "              speaker  word_count  year  month  month_year  \n",
       "0  The VICE PRESIDENT          16  1967      1  1967-01-01  \n",
       "1       Mr. MANSFIELD          35  1967      1  1967-01-01  \n",
       "2  The VICE PRESIDENT          40  1967      1  1967-01-01  \n",
       "3  The VICE PRESIDENT         151  1967      1  1967-01-01  \n",
       "4  Mrs. AGNES BAGGETT           3  1967      1  1967-01-01  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress = congress.drop(['Unnamed: 0'], 1)\n",
    "    \n",
    "congress.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>speech</th>\n",
       "      <th>date</th>\n",
       "      <th>speaker</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2329890</th>\n",
       "      <td>Mr. Speaker. we in Delaware are proud of the o...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. EVANS of Delaware</td>\n",
       "      <td>122</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329891</th>\n",
       "      <td>Mr. Speaker. it is logical for Americans to be...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. DERWINSKI</td>\n",
       "      <td>82</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329892</th>\n",
       "      <td>The Chair has examined the Journal of the last...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>The SPEAKER pro tempore</td>\n",
       "      <td>32</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329893</th>\n",
       "      <td>Mr. Speaker. I ask unanimous consent that the ...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. WHITE</td>\n",
       "      <td>36</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329894</th>\n",
       "      <td>Is there objection to the request of the gentl...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>The SPEAKER pro tempore</td>\n",
       "      <td>11</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    speech        date  \\\n",
       "2329890  Mr. Speaker. we in Delaware are proud of the o...  1980-01-03   \n",
       "2329891  Mr. Speaker. it is logical for Americans to be...  1980-01-03   \n",
       "2329892  The Chair has examined the Journal of the last...  1980-01-03   \n",
       "2329893  Mr. Speaker. I ask unanimous consent that the ...  1980-01-03   \n",
       "2329894  Is there objection to the request of the gentl...  1980-01-03   \n",
       "\n",
       "                         speaker  word_count  year  month  month_year  \n",
       "2329890    Mr. EVANS of Delaware         122  1980      1  1980-01-01  \n",
       "2329891            Mr. DERWINSKI          82  1980      1  1980-01-01  \n",
       "2329892  The SPEAKER pro tempore          32  1980      1  1980-01-01  \n",
       "2329893                Mr. WHITE          36  1980      1  1980-01-01  \n",
       "2329894  The SPEAKER pro tempore          11  1980      1  1980-01-01  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress = congress[congress['year'].isin(range(1980, 1989))]\n",
    "congress.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Groupby with one column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "The PRESIDING OFFICER      155043\n",
       "The SPEAKER pro tempore     48203\n",
       "Mr. BYRD                    35055\n",
       "Mr. BAKER                   32142\n",
       "The CHAIRMAN                29128\n",
       "                            ...  \n",
       "Mr. RON SMITH                   1\n",
       "Mis. BENTLEY                    1\n",
       "Mr. RIEGLES                     1\n",
       "Mr. LuNORN                      1\n",
       "Mr. SYKES                       1\n",
       "Name: speaker, Length: 12145, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress['speaker'].value_counts() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is another way of writing the same command:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker\n",
       "The PRESIDING OFFICER      155043\n",
       "The SPEAKER pro tempore     48203\n",
       "Mr. BYRD                    35055\n",
       "Mr. BAKER                   32142\n",
       "The CHAIRMAN                29128\n",
       "Name: speaker, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby('speaker')['speaker'].count().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>speech</th>\n",
       "      <th>date</th>\n",
       "      <th>speaker</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2329890</th>\n",
       "      <td>Mr. Speaker. we in Delaware are proud of the o...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. EVANS of Delaware</td>\n",
       "      <td>122</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329891</th>\n",
       "      <td>Mr. Speaker. it is logical for Americans to be...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. DERWINSKI</td>\n",
       "      <td>82</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329892</th>\n",
       "      <td>The Chair has examined the Journal of the last...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>The SPEAKER pro tempore</td>\n",
       "      <td>32</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329893</th>\n",
       "      <td>Mr. Speaker. I ask unanimous consent that the ...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. WHITE</td>\n",
       "      <td>36</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329894</th>\n",
       "      <td>Is there objection to the request of the gentl...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>The SPEAKER pro tempore</td>\n",
       "      <td>11</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    speech        date  \\\n",
       "2329890  Mr. Speaker. we in Delaware are proud of the o...  1980-01-03   \n",
       "2329891  Mr. Speaker. it is logical for Americans to be...  1980-01-03   \n",
       "2329892  The Chair has examined the Journal of the last...  1980-01-03   \n",
       "2329893  Mr. Speaker. I ask unanimous consent that the ...  1980-01-03   \n",
       "2329894  Is there objection to the request of the gentl...  1980-01-03   \n",
       "\n",
       "                         speaker  word_count  year  month  month_year  \n",
       "2329890    Mr. EVANS of Delaware         122  1980      1  1980-01-01  \n",
       "2329891            Mr. DERWINSKI          82  1980      1  1980-01-01  \n",
       "2329892  The SPEAKER pro tempore          32  1980      1  1980-01-01  \n",
       "2329893                Mr. WHITE          36  1980      1  1980-01-01  \n",
       "2329894  The SPEAKER pro tempore          11  1980      1  1980-01-01  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress = congress[congress['year'].isin(range(1980, 1989))]\n",
    "congress.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Groupby with one column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "The PRESIDING OFFICER       155043\n",
       "The SPEAKER pro tempore      48203\n",
       "Mr. BYRD                     35055\n",
       "Mr. BAKER                    32142\n",
       "The CHAIRMAN                 29128\n",
       "                             ...  \n",
       "Mr. QUILLEN of Tennessee         1\n",
       "Mr. MILTON D. STEWART            1\n",
       "Mr. MeCAIN                       1\n",
       "The PRESIDENT protempore         1\n",
       "Mr. ZABLOCKi                     1\n",
       "Name: speaker, Length: 12145, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress['speaker'].value_counts() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is another way of writing the same command:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker\n",
       "The PRESIDING OFFICER      155043\n",
       "The SPEAKER pro tempore     48203\n",
       "Mr. BYRD                    35055\n",
       "Mr. BAKER                   32142\n",
       "The CHAIRMAN                29128\n",
       "Name: speaker, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby('speaker')['speaker'].count().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The only difference: value_counts will return an object in descending order.  Count will not; it will organize things according to the order of the objects in ‘colA’\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Groupby with multiple columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What if there are many columns?  \n",
    "\n",
    "    value_counts() \n",
    "\n",
    "only works on one column at a time.  Its only purpose s to return the frequencies of unique values.  \n",
    "\n",
    "\n",
    "    Groupby()\n",
    "\n",
    "is much more flexible. Its purpose is to group data by unique values. It can be used along with sum, count, mean to return information about any category.  \n",
    "\n",
    "If you have blocks with aspects ’shape’ and ‘color’, you can specify which columns to count\n",
    "\n",
    "    df.groupby(‘color’)[‘color’].count()  -- organize by color and count how many there are of each color -- equivalent to value_counts()\n",
    "    df.groupby([‘color’, ‘shape’]).count() -- organize by color and count how many there are of each shape. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Let's apply it to Congress*\n",
    "\n",
    "This is how you ask: for each speaker, in how any years did they speak?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker                year\n",
       "The PRESIDING OFFICER  1987    19439\n",
       "                       1986    18423\n",
       "                       1980    18275\n",
       "                       1988    17950\n",
       "                       1982    17829\n",
       "Name: year, dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['speaker', 'year'])['year'].count().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The order of the columns you call matters.  So this produces the same results organized in a different way:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year  speaker              \n",
       "1987  The PRESIDING OFFICER    19439\n",
       "1986  The PRESIDING OFFICER    18423\n",
       "1980  The PRESIDING OFFICER    18275\n",
       "1988  The PRESIDING OFFICER    17950\n",
       "1982  The PRESIDING OFFICER    17829\n",
       "Name: year, dtype: int64"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['year', 'speaker'])['year'].count().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The line above asks: for each year, how many times did each unique speaker speak?\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use .groupby() with\n",
    "\n",
    "     .sum()\n",
    "\n",
    "to add up values, for example the numbers in the column *word_count*, which tell us the length of each individual speech.\n",
    "\n",
    "This line asks: for each year and each speaker in each year, how many words did that person speak per year? Or in plainer English: *how many words did each speaker say per year?*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year  speaker           \n",
       "1987  Mr. BYRD              867930\n",
       "1988  Mr. BYRD              584008\n",
       "1980  Mr. ROBERT C. BYRD    563799\n",
       "1983  Mr. BAKER             466734\n",
       "1985  Mr. DOLE              447936\n",
       "Name: word_count, dtype: int64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['year', 'speaker'])['word_count'].sum().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We could also use .groupby() with \n",
    "\n",
    "    .mean()\n",
    "    \n",
    "The line below asks: what is the average number of words spoken by each speaker in each year?  Note that the results suggest that there is a lot of bad data in the speaker column that needs to be cleaned up."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year  speaker             \n",
       "1980  sors. P.O               19216.000000\n",
       "1986  The REIT                16511.000000\n",
       "      ass.........eA          15159.000000\n",
       "1981  ps.. P.O                15105.000000\n",
       "1988  The MAAC                15047.000000\n",
       "1981  sors. P.O               14924.666667\n",
       "1986  The LIFO                13868.000000\n",
       "1982  CBS. WDM                13520.000000\n",
       "      sors. P.O               13056.000000\n",
       "1980  es. AFL-CIO             12735.000000\n",
       "1985  The FUTA                12376.000000\n",
       "      MR. JUSTICE BLACKMUN    11645.000000\n",
       "1980  car. S.A                10895.333333\n",
       "      The CTTF                10084.000000\n",
       "1987  Mon. AFL-CIO             9034.000000\n",
       "Name: word_count, dtype: float64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['year', 'speaker'])['word_count'].mean().sort_values(ascending = False)[:15]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The only difference: value_counts will return an object in descending order.  Count will not; it will organize things according to the order of the objects in ‘colA’\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Groupby with multiple columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What if there are many columns?  \n",
    "\n",
    "    value_counts() \n",
    "\n",
    "only works on one column at a time.  Its only purpose s to return the frequencies of unique values.  \n",
    "\n",
    "\n",
    "    Groupby()\n",
    "\n",
    "is much more flexible. Its purpose is to group data by unique values. It can be used along with sum, count, mean to return information about any category.  \n",
    "\n",
    "If you have blocks with aspects ’shape’ and ‘color’, you can specify which columns to count\n",
    "\n",
    "    df.groupby(‘color’)[‘color’].count()  -- organize by color and count how many there are of each color -- equivalent to value_counts()\n",
    "    df.groupby([‘color’, ‘shape’]).count() -- organize by color and count how many there are of each shape. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Let's apply it to Congress*\n",
    "\n",
    "This is how you ask: for each speaker, in how any years did they speak?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker                year\n",
       "The PRESIDING OFFICER  1987    19439\n",
       "                       1986    18423\n",
       "                       1980    18275\n",
       "                       1988    17950\n",
       "                       1982    17829\n",
       "Name: year, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['speaker', 'year'])['year'].count().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The order of the columns you call matters.  So this produces the same results organized in a different way:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year  speaker              \n",
       "1987  The PRESIDING OFFICER    19439\n",
       "1986  The PRESIDING OFFICER    18423\n",
       "1980  The PRESIDING OFFICER    18275\n",
       "1988  The PRESIDING OFFICER    17950\n",
       "1982  The PRESIDING OFFICER    17829\n",
       "Name: year, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['year', 'speaker'])['year'].count().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The line above asks: for each year, how many times did each unique speaker speak?\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use .groupby() with\n",
    "\n",
    "     .sum()\n",
    "\n",
    "to add up values, for example the numbers in the column *word_count*, which tell us the length of each individual speech.\n",
    "\n",
    "This line asks: for each year and each speaker in each year, how many words did that person speak per year? Or in plainer English: *how many words did each speaker say per year?*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year  speaker           \n",
       "1987  Mr. BYRD              867930\n",
       "1988  Mr. BYRD              584008\n",
       "1980  Mr. ROBERT C. BYRD    563799\n",
       "1983  Mr. BAKER             466734\n",
       "1985  Mr. DOLE              447936\n",
       "Name: word_count, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['year', 'speaker'])['word_count'].sum().sort_values(ascending = False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We could also use .groupby() with \n",
    "\n",
    "    .mean()\n",
    "    \n",
    "The line below asks: what is the average number of words spoken by each speaker in each year?  Note that the results suggest that there is a lot of bad data in the speaker column that needs to be cleaned up."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year  speaker             \n",
       "1980  sors. P.O               19216.000000\n",
       "1986  The REIT                16511.000000\n",
       "      ass.........eA          15159.000000\n",
       "1981  ps.. P.O                15105.000000\n",
       "1988  The MAAC                15047.000000\n",
       "1981  sors. P.O               14924.666667\n",
       "1986  The LIFO                13868.000000\n",
       "1982  CBS. WDM                13520.000000\n",
       "      sors. P.O               13056.000000\n",
       "1980  es. AFL-CIO             12735.000000\n",
       "1985  The FUTA                12376.000000\n",
       "      MR. JUSTICE BLACKMUN    11645.000000\n",
       "1980  car. S.A                10895.333333\n",
       "      The CTTF                10084.000000\n",
       "1987  Mon. AFL-CIO             9034.000000\n",
       "Name: word_count, dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress.groupby(['year', 'speaker'])['word_count'].mean().sort_values(ascending = False)[:15]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WQFxtHi-tODs"
   },
   "source": [
    "# The `groupby()` Method"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "psjrk-eBtODt"
   },
   "source": [
    "Understanding how to group data is essential for performing a number of operations on pandas DataFrames. \n",
    "\n",
    "Groupby refers to a process involving one or more of the following steps: \n",
    " - splitting the data into groups based on some criteria.\n",
    " - applying a function to each group independently.\n",
    " - combining the results into a data structure. \n",
    "\n",
    "The main takeaway is that with `groupby()` we can group categories and then apply functions to the categories instead of to the entire data set.\n",
    "\n",
    "Before showing code examples of `groupby()` let's get a quick conceptual overview of what it means to perform a function to a category as opposed to the entire data set.\n",
    "\n",
    "Consider the following image, a sample of the Hansard data. If we were to group this data by decade it would be as though we stuck each decade into a bucket to be operated upon independantly from the rest of the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mTuBGvx2tODv"
   },
   "source": [
    "<img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAABT8AAALoCAYAAAC+rIEjAAAABHNCSVQICAgIfAhkiAAAAAFzUkdCAK7OHOkAAAAEZ0FNQQAAsY8L/GEFAAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAi3pUWHRSYXcgcHJvZmlsZSB0eXBlIGV4aWYAAHjaVY7dDcMwCITfmaIjHD8GM04V2VI36PjFcaQm3wPcIXRA4/uZ9FowjKxF93RHYWkp7xIdGwVYwKtX3VxduZSUvTypbOHZA/ZfNMODpt59hkV488MPqXQZqmdlBZ1bK2y9kreLE22L53x0yP0A/QABNy1QmOR7hQAACghpVFh0WE1MOmNvbS5hZG9iZS54bXAAAAAAADw/eHBhY2tldCBiZWdpbj0i77u/IiBpZD0iVzVNME1wQ2VoaUh6cmVTek5UY3prYzlkIj8+Cjx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iIHg6eG1wdGs9IlhNUCBDb3JlIDQuNC4wLUV4aXYyIj4KIDxyZGY6UkRGIHhtbG5zOnJkZj0iaHR0cDovL3d3dy53My5vcmcvMTk5OS8wMi8yMi1yZGYtc3ludGF4LW5zIyI+CiAgPHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9IiIKICAgIHhtbG5zOmV4aWY9Imh0dHA6Ly9ucy5hZG9iZS5jb20vZXhpZi8xLjAvIgogICAgeG1sbnM6dGlmZj0iaHR0cDovL25zLmFkb2JlLmNvbS90aWZmLzEuMC8iCiAgIGV4aWY6UGl4ZWxYRGltZW5zaW9uPSIxMzQzIgogICBleGlmOlBpeGVsWURpbWVuc2lvbj0iNzQ0IgogICB0aWZmOkltYWdlV2lkdGg9IjEzNDMiCiAgIHRpZmY6SW1hZ2VIZWlnaHQ9Ijc0NCIKICAgdGlmZjpPcmllbnRhdGlvbj0iMSIvPgogPC9yZGY6UkRGPgo8L3g6eG1wbWV0YT4KICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgCiAgICAgICAgICAgICAgICAgICAgICAgICAgIAo8P3hwYWNrZXQgZW5kPSJ3Ij8+3bt7oAAA9OpJREFUeF7s3Qm8FeMfx/FfiyTZsi+hqCwpu6wlKUQIZQvZIkJCliyRJUsiW7bsf0KERCr7vpbQQiH7EpIkLf8+vzvPbZpmzj23e9uO7/v1uq/OmXPOLM/MHM53fs8zFWbPYSIiIiIiIiIiIiIFpmL0r4iIiIiIiIiIiEhBUfgpIiIiIiIiIiIiBUnhp4iIiIiIiIiIiBQkhZ8iIiIiIiIiIiJSkBR+ioiIiIiIiIiISEFS+CkiIiIiIiIiIiIFSeGniIiIiIiIiIiIFCSFnyIiIiIiIiIiIlKQFH6KiIiIiIiIiIhIQVL4WY5uu+02e+qpp6Jni8f06dOtdu3avi5JDz30kL/2ww8/RFMWzA477GDnnXde9Kz8ldd6isiS5c8//7Qrr7zSRowYEU0pnSXhO1ZERERERESWLgo/y1H//v3tpZdeip4tHpUrV7brr7/emjZtGk0REVky/PXXX3bHHXfYmDFjoimlsyR8x4qIiIiIiMjS5T8ffn711VfWtm1b22yzzaxx48b22GOP+fTvv//ejjvuONtyyy2tTZs29vrrr/v0Sy65xLbeemv/l9datWplv/zyi1dCfvnll/bkk0/azTffXOrPY/jw4da8eXOrX7++HX300fbjjz/69Pfee8/2228/f/8JJ5xg3333nU9PM2PGDOvcubPPC88//7ztuuuu/vfmm2/6tNKaMmWKnXLKKb585s0yAqo0d9ttN2vUqJFdeOGFXnmKrG3p1auXbbfddrb55pvbkUceab/++qtPz1rPrHYUkaXP4Ycf7v+ee+65/r2W9t3GtI022siuueYa+/rrr61u3bp21llnzfcdKyIiIiIiIpKP/3z42bt3b5swYYLdfffdHrxddtllNnnyZDv//PM9mLvrrrs8qOvQoUNx6Pj7779bxYoVPRAcNWqUDRo0yC6//HLbYIMNbP/997eOHTuW+vMEoJ06dbK1117brrvuOhs9erR3D6VS6qSTTvJwlnkRRB5//PE+n5L8/PPPHlauu+661qVLF3v77bejV0qHNiLMZH2rVq1qv/32m0//8MMPrVu3bh5OEmoSYF511VWZ2/Lxxx97t1UCEN5PyPnoo4/mXM9c7SgiS5cHHnjA/+X7ctNNN039btt2223tsMMOs3vuuccuvfRSW2GFFfx7IPkdKyIiIiIiIpKP/3z4SeBGWMcP65VXXtkef/xxW3HFFe2tt96ykSNH2iGHHGL33nuvTZ061T766CP/TPXq1T30o1JptdVW83CPMLNChQrFf6X9/AcffGB///23nXbaadaiRQt7+OGH/fWxY8fapEmTPCRkXgSDvD9UTObyxhtv2LRp0+yCCy6wAw44wPbZZ5/oldIh+Nx55519fc4++2zfPoSQsnv37l7FSTtSmZm1LVtssYUNGTLE2+p///ufzZ4928cAzLWeudpRRJYunPuoVKlSzu+2c845x0NPvnv4XqhRo8Z837EiIiIiIiIi+fjPh59nnnmmVyNts802HnwSvFEJSkBJleG4ceN8fDoqNOnGjSpVqhT/iF9mmWX836TSfn7WrFn+b3w6f8wHVEYxL8IB5kUYUBICSISgIMyrtJhPmAeVn2Gdw/yo6GTdPv30U++SmrUttEOzZs28azwVnssuu6y/nms9c7WjiCy9cn238R0Svkf+/fdf/1dERERERERkQfznw88jjjjCb8Cx7777evUi41lSebTLLrvY4MGDvcqQrpdNmjTxbpm5cLMh7lBOt/bSfp4x7wgD+/TpY8OGDbODDz7Yu4hvuOGGXp3K2JqEjKzvySefnFflU8OGDf1fupw/8sgjXmG1IFg3Kjr79es3z7ieO+20kwecd955p1dtUR1KV9Wsbfn888890FhvvfXsueees3/++cerP3Ot54LsBxFZMoULJ4zdudZaa2V+t11xxRX+/cDFKL4XfvrpJ/9c/DtWREREREREJB//+fDz9NNP9+7a7dq1827VjFXJmHOMMVe7dm2/WQ9drxn3km7xuXDjJKofqSAt7ecJAngPN2BinRgHj3kQFhAiMs4l4QDdRQkD8sGYelS2jh8/3m644Qavbl0Q3GyEgJK7yNM9PWwH20dIwU2iGLuPYJLu7Vnb0rRpU/+7+uqrvbv7Gmus4V3ec63nguwHEVkyrbrqqtayZUu77777/HxP+27jO4HvFL5LGB6EEJQbxCH+HSsiIiIiIiKSjwqzKb0TERERERERERERKTAKP5dSVFx+9tln0bO56HLOeJoiIiIiIiIiIiL/dQo/RUREREREREREpCD958f8FBERERERERERkcKk8FNEREREREREREQKksJPERERERERERERKUgKP0VERERERERERKQgKfwUERERERERERGRgqTwU0RERERERERERApShdlzRI+LDR8+PHokIiKydGjatGn0SERERERERKRIavgpIiIiIiIiIiIisrRTt3cREREREREREREpSAo/RUREREREREREpCCp27sUvEk160SPzGpMHBc9EhERERERERGRQqfKTylo8eATyeciIiIiIiIiIlK4FH6KiIiIiIiIiIhIQVL4KSIiIiIiIiIiIgWpXMLP2267zRo1amRNmza1p556KppaPvr06WO77bab7b777tauXTsbO3Zs9EqRqVOnWs+ePaNnZr169bIePXpEz4oMGzbM2rZt648fe+wxa9asme255572wgsv+DRkTUdyGdh5551tyy23LP778ccfbfr06Va7dm1/jb+tt97aOnXqZH///bd/Jtcykm6++eZ55r/ZZpvZxRdfnDkdb7/9th1wwAG+7L59+/o09O7d23bddVf/u++++6Kpi8YHH3xgAwcOjJ6Z/fXXX9axY0fbZZddrEWLFr7Occn3z5w507p37+7HVuvWrW38+PHRK4XrxRdftPPOOy96tmi88sords4550TPcrvhhhusf//+0bP5ldf6//bbb7bvvvtGz4pwLnIs5OvTTz+14cOH++OS1ntBlbROHONnnnlm9CzdlClTrF+/fv44n/fHxbcxH9dcc409/vjjee3z+D7gu+WHH37wx2n4zuE7sLT7KIi3gYiIiIiIiEh5KXP4+cYbb9jVV19tHTp08FCPH+1fffVV9GrZDBkyxJ577jl7+umn/cf96aefbscdd1xxmPjaa69Z586dbfTo0f4c+++/vw0ePNji93FiPvyA/+mnn+yWW26xAQMG2P3332+XXXaZzytrOtKWwQ/8KlWq2EcffVT8t+aaa/pryy67rL3++uv+R+j63Xff2cMPP5xzGWlOOeWUeea/44472pFHHpk5/c8//7SuXbt6ED106FBfDkEFYeKgQYO8DZ555hm76aabfJ0WtmnTpnkQfv7559v3338fTTUPX6tXr+7By+WXX27nnnuuT896PyEx60tbnnrqqR6EFjouJJx11lnRs/LDOTFjxozo2YI75phjbO+9946ezW9hrT+WW245e+CBB6JnJSMsTwbs5a2065SG4z8e+pfGothG3Hrrrbb66qtHz+bHxR2OrwVtj7K0gYiIiIiIiEiWMoefhFKrrbaatW/f3sOpWbNm2auvvhq9Wja//PKL1axZ01ZYYQWrUKGCbbvttnbBBRcUBzj86A+hY7DRRhtZjRo17P333/fnvJdKNMKal156yStIV1xxRVtrrbWsQYMG9tZbb2VOR9oyCON4X0lWWWUVr1qdMGFCzmWUhECA99epM+/NeuLT2Q8si3kTPhAYE1RUqlTJlllmGQ9rq1at6o8XBYIM2mnttdeOphSpWLGirwf/EoKyX5H1fkLbgw46yN9HNRkhNJVlhYLAt02bNn58chEB7777rl177bX+eI899rDTTjvNq2SpfP7jjz98Op9jWpMmTezGG2/0aVTVdunSxY+DAw880C9MgGpoQm8uDHA8E8I3btzYl3nPPff4ewICrFBJjEMPPdQ+/vjj6FkRPsMFBlBFyDpS0RwCr/j6H3744R5yt2zZ0i9AfPjhhz79yy+/9HmzroSpoaLwnXfesb322su3jSrnpH///dfDflDJeeGFF9p+++3n6/DII4/49GDSpEk+j2effdbPB4waNcrbkQs14f1Z6xKE4B20+6abblpc4bjVVlv5v2GdaCu+C1l/PjN58mSfTgUlFc9s20UXXeTT4rgYwr6hPZH2fqqgu3Xr5vt8p5128vembWNANfrRRx/t72fbwj4rSdY+YHt+/vnn1G286667/Pw9+eST59lHWfv/s88+82OU45CLWrRhsg1EREREREREykOZw09+DBM2gpCScO3XX3/152XFD3C6QhIa0pWdEJMf3CwHRx11lHeFTiLkodoR/JCvV6+eB7SEGvHQcp111vGKzKzpSFvGt99+a998843ts88+XuV2/fXXR6/M64svvrDnn3/ew8lcy8iFkIVg4fjjj4+mFElOZ53YF6wT3d6vu+46Dz4bNmxoW2yxhe2www7+d8ghh/iyF7aVV17ZTjrpJF9mHKEJoS9d8OlGSzCCrPcT4MQDUdown3ZbWlx66aUeTFKV+/XXX/s+jCM4JzTiOKpVq5a/j8pYhmGgopjKaAI5AjxCzVVXXdWrpBnq4IwzzrDff//d50NlIMEmx+HLL7/s5wddwJ988kmvUg4IBfk8FzEIs/g8x08aqnKZL6EalcZ0ex4zZkz06lzMi+URjN19990+jZD22GOP9Qpgzi8CXo5pgrArr7zSt5fnJSGIe+KJJ+yhhx7ybY7je4lKac4JAlKMHDnSu1bfe++9xaFx2rrEcUxSQQ3+pY1ZLpXX22233TwXFKiOJKxn/evWrevvAcc8+5q2or3Zr3GEuAyZcfbZZ/vztPcTGBJ+s3/4e/TRR61atWrzbWNAaEtlOBdeCDHvuOOO6JVs+eyDtG2kIp/vFV5LStv/HJtUefOdzoUZjutkG4iIiIiIiIiUhzKHn/ywjaNCLzltQRFeUE1GpRnBD126CcxKqvwLYRHrwb9UHYHnodIQoRtw1vQshB9UPPGDnh/tVF3xIx7//PNP8ZifBAJUq1JVVtplBAQ7VG4tv/zy0ZQiyelUTjIeKl3qQ2BC+PHmm296aPLggw964MP6pgVUiwoB2eabb+7BGxWJjA1JpVgWqt3iaLfktKUZxwnhG0EmQXCyWzFVvIT/2GSTTTyMfO+99/y4orKYKlqOPyqkCcyorDvhhBPskksu8Xb9/PPP/bNUzRIyEVZSLchYtIR8hGmhGg8rrbSSB1qEfISqnG9ZmBfVo8yXimaqC1m3JII5UDHJ+nOOUIFJYEkYzviTBGgEuBzPVFOC87gkzZs3t8qVK3t1NtXEJQnvDxcAstYljgpl2pe2JDylOpTtpMKVbY5jLFsqUvmuatWqVfHrhKRcgOE7gP0YKnizpL2/fv36PpwA60hlLW2Z69yh6pLglu8EQsl8wuR89kHWNmZJ7n8Cfv5lG9lnBK25jjMRERERERGRsihz+EkQGCo9GXeSH9j8aC8P/LgmiCD4POKII7ybKoEoFVq5UB3IZwiDqJCiWhRrrLHGPF1aqSDkvVnTsxAO0VWZYIJtJVgKgWJ8zE8qrqjeojKstMsICCvpUpyUnM5+oJqSfwmiCCjGjRvnoSwVYVS/UgVKkFZewxIsCMYjpcqOak5CEdY1Pp5qEqEW1Z8B7UZbFgqqFam0YxsJkpIVgQRvyaEKCH8J8ALOOy4IUAV41VVXeWUjfxz73HQLISQnTCW8C++hqzRhaRz7JYy1SxV1LvFAn0pjwukkQto4tofzhosarAP/cq4ShMXnx7aXhIrh0kh+N2WtSxLnE+cz7U7YR/BJAMo5F0fgyPcWgTDd1qmsRWm/E9Pez4UMqie5kHLwwQf7d1wuVAITUjIvPpePfPZB1jZmSe5/LvrEj2kC6JLCYBEREREREZEFVebwkzCN8JOuu3TF5YdzSZVApcGPd8YyBF3NJ06caBtssIE/z4UgifWhy3n48c26EgZyoyHWmYo3uq1nTc9C91Fu7EQIwdh8hLHbbLNN9Gq60i4DjKVH1/lkt+O06VSBEmoyTipVVQQ1hJ0bb7yxrx/hGAEDoSzTFhfGZA1VuVTS0dV7vfXWi16dH2M5EvSC8IdusWHYg6UdQzrQzZz9wfHE/hwxYkT0ajYCTQI6jgOCIy4MULHHOJBU+YHgmxCTdo7jPXR/p1KYY4JxF0N1aMA6EWhxoSE5Bmvc9ttv7yEpYRYBLMcc1XwlIWBjm/ks6L7Nja84r2mT0AZUN5eHXJXCWeuSRPhJOLrlllv6dwpjU3IeJ7+LGCaD6QT8hIRpQWqWkiqa6WrPhZy2bdt6yMy5Eyo/0z7LuMeEngwfQnVrPpXm+eyDXNuYT1U2F30IP/kOZDsY0iQEqOHzHJ8sI9djERERERERkXyUOfwkdKMrJt3TGUuQwDGfcDIf3EGe7r2Mw0f3YLqlUmlE98mS0B2XbuChyzsIcpgn3dCZFzd2oSIua3oW3kMAR0hEd02qQEsKfUq7DBCWst3xSiykTd9www29mz1jelKtx3ox3h8VYoQ1BCB0YeVf9tniwjh/VHpyoxPG+KT7brIyLI51pis220OV5BVXXBG9svSjqo5jm0CLsJ52YP+UZP3117fOnTv7vqX6l88QnHJcEVxRiUzoxbkYrxAFwSnv57xguRwvDEMQRzUuXa2TY90m0YWc85NzjTEgOb6pMM4HgRdVllwUIIxlvEcCMfYx4SPBLecY1aRlwbFPtXGuu4inrUsSNxbjIgwXOTjvqP5m25PYB4xZSdtywymO8XxwjIcgMAvnANWmHCu33367tz9DJWRtIxW9HAO8n4ppgvD+/ftHr6bLZx9kbSPtQTBaEtqPi1rdu3f3NudiDTdIircBQTrfZ8h6LCIiIiIiIpKPCnN+bM7fT1WkQEyqOe8d8lFj4rjokaQhQCXQpEKXrvRJBGpU4BKcioiIiIiIiIgsyRR+LkZ04edGN2natWuX15igkpvCz9Kh+zzVe9zQi4rmJG6CxB27GfNRx6eIiIiIiIiILOkUfkpBU/gpIiIiIiIiIvLfVeYxP0VERERERERERESWRAo/RUREREREREREpCAp/BQREREREREREZGCpPBTRERERERERERECpLCTxERERERERERESlICj9FRERERERERESkIFWYPUf0uNjUqVOjRyJLt2n1GkaP5qo6ZkT0SEQKSbVq1aJHIiIiIiIiIkVU+SkiIiIiIiIiIiIFSeGniIiIiIiIiIiIFCSFnyIiIiIiIiIiIlKQFH6KiIiIiIiIiIhIQVL4KSIiIiIiIiIiIgVJ4aeIiIiIiIiIiIgUpHILP3/44Qe74IILbNq0adGU8nHrrbda8+bNbe+997bjjz/exo0bF71SZOrUqdarV6/omdmNN95oPXv2jJ4VefHFF+2oo47yx0888YTtu+++tt9++9mwYcN8GrKmI7kM7LHHHtaoUaPivx9//NGmT59um2++ub/G30477WRdunSxv//+2z+TaxlpevfubXvuuae1bNnShgwZ4tNmz55tl112me21117WrFkze/jhh3NOR2mXW54++ugje+aZZ6JnZn/99ZedccYZvo6tWrWyd999N3qlSPL9M2fOtCuuuML22WcfO+yww2zChAnRK0uGtm3b+r7Px2effWYvvfSSP7755pvt8ccf98fl4eWXX7aLLrooelY27JNzzz03elY6ZfksLr/88vmO0Xi7ZWGZb731VvRs6bIwj4ssDz30kN15553Rs7kWdP/ls49eeOEF++KLL/xx+/bt7euvv/bHcaU5n0RERERERETyUebwc9KkSf5D+pRTTrEnn3zS/v333+iVsiME4Qfzo48+as8++6x17NjR/0KY+MYbb1jXrl1tzJgx/hyEfASFhIEB8yEU/Omnn+yOO+4o/uFPSMq8sqYjbRmEnFWqVPGwJfytueaa/tqyyy7ry+Nv0KBB9v3339tjjz2Wcxlpnn/+eXv//fdtwIABdvvtt9tVV13lbU3AMHHiRG8P5nX99dfbr7/+mjm9tMstL4TgbP/FF1/swXjAeiy//PK+jy655JLiwC7r/cOHD/c25LUOHTp4ELokueGGG2y11VaLnuX25Zdf2nvvvRc9K1/bb7+9nX766dGzwrIw221JsCRtX4MGDeyss86KnuUvn21488037ZtvvomepSvN+SQiIiIiIiKSjzKHn99++60HlFRHlrdffvnF1ltvPVthhRWsQoUKtvXWW9s555xjM2bM8Nf5wb3GGmv446B27dq2yiqr2IcffujPee8rr7zi1aOvvvqq7bbbbrbiiit6WFm/fn175513MqcjbRmEcSHszIX12HnnnX0euZaRZuTIkbb77rv7tq+77rq+7YSsK620kp1wwglWsWJFW3311T0oYBuzppd2ueWFMJN2WmuttaIpRVi/qlWr+r+EoOxXZL2fEPmAAw7w9zVp0sTGjh27UI61BXXmmWf6cUrF3qWXXmoHH3ywV+oSeMcRXPft29dDbQJqfPrpp17NTEVueD+Vseedd54fr4ceeuh81Yy00y677FLcBh9//LFXxBKUExzhq6++sqOPPtrnceKJJ3ol3dVXX23/+9///HVC5+OOO84fv/766x4qJ/32229eobv//vt7RXFAEN+iRQvfJw888EA0dV5pn6WCt3v37n4RomnTpl7VDC4kXHjhhT6d6uxkOJZsN7aF45z3s32hIjqgXY488kh77rnn/Hna+lLd2KlTJ2vXrp3vq/j2BVwk4HstoJIbnTt3tm7duvk8qUYOy5k8ebJvM/M76KCDfH+gvI6LtH2axAUCtolt5eIHOH+opg3Y16F6/vPPP7cjjjjC58n2csGI751rr73WX6eamHVgvbkoES5ssRyqttkHt9xyS+o2JHEcc4GGYzRUf1KRzvrSNiNGjPBp4XwSERERERERKS9lDj+32GIL69evn3dXLG90+SYMoss7P875MU53aQJBHH744R6wJFH9GUIJqpHq1q3rYSCBQTy0XHvtte3nn3/OnI60ZXz33Xce+h544IEeyPXp0yd6ZV7jx4+3oUOH2kYbbZRzGWkIcak6pUqT97EdVHISgm633XbeNZaQhyCJ+WZNL+1yy8vKK6/sAQ7rFEfYQfjKviXcI1RB1vupXI2vP48XxfoviE8++cSHG7j77rs99IqrUaOGB08hNMOoUaPstttu84CO4R1AlS7vJUQisCTs//333/01EBzvsMMOfmyAgI6AKo7wlCCRYJBjl0q+HXfcsXiIAcI1qoQJswjpCOiTeC8hH9XcBFcE+K+99pr/cYzdf//99vTTT3s1X1LaZ0ePHu0BIdvFHxXNBLmEe4RdBGqc42+//XY0lyLJdiNAY/s5vxmKgu+e4J9//rHTTjvNwz+CuVzrS+jL55966ilvgxDI5YP1HTx4sFdSE2xyXrL/uEjBdjDt7LPPLg4Ly+O4SNunSVdeeaVdd911vr3s35LOE/YT86cN+H6Jh70EmmwHoebAgQOtcuXKHnRSlU273Xvvvb5/2Zd//vnnfNuQxNAgfFdSncz3Idj/7Be6wD/44IM+TURERERERKS8LdE3PCIYuOuuu/wH+AYbbOCPCVlLqvwjLCV0nDVrlv/Lj3LwPFQagkonqiOzpmcJgQUBDuENgQTBLAhgwpifdNEnlDzkkENKvQwCjnXWWceDLYYUIMAl+AoYT5T50i2cMDZITi/tchc29uFmm23moQdd+eMVZWmoGIxj/ZPTlhTsc0IiAloqW0sS3k8gHRAMUwXHPqdij7YhRI/jeGYcW/BvOL7B8Ue4+cgjj/gxSkBFNd+2227rVaIEkMstt5xftCCQpEKaYyaJ45YLBhw7HHt8joCMqk7GsSV8Y1qo2ItL+yxj4RJ8sT5U//3xxx++bQRwHOOVKlXydthyyy2juaTj/CcgD8M4UDkaEBaz7WF7cq0v60iFNMutU6eOr0+++H5h/3J+0o6ElWxHCP7YVtqYABJlPS6y9mkSAff555/vw4RwXrF9uVBBTEU43yscQx988EH0SlFAzv6hWpdjkW3k+OE9tB1V7XyO70Cq8xcEFafgGClN+4uIiIiIiIiUxhIdfhJu8CN/ww039CrB++67z390U7WVCyEDnyFICtWiIAyIdxelMor3Zk3PQlhC11bCHUKexo0bF3cljY/5SXUaXXqXWWaZUi+DUIbuyYS3/fv3t2rVqnkYwvbQ9ZvHVLg1bNjQq+Wyppd2uQsbYR3Vn3RvJ3Chijc+nmoSQw5Q/Rmw/iWFOosLQw+Uxqqrrho9mov9HCru+KOyLhkI7rrrrn5sE0jVrFlznvlwrPGcqkE+z7+0OWHc+uuv71V8zG+bbbbxqkzaduONN44+PRfHdRJhF92kw7oRslG1l5T2WY5Fwk/CeI5PLmaAMDAezjMUQi5sD9WTLINK4Tjme/LJJ3sIilzry/dIvqi+jgf08fXlPOViAqF8cjrbirIeF1n7NInt5oIL+7RNmzZecZs0ZcqU6NG820HYyl/Accg4sqHtuFhBV3ouPMTfR9Xngg5DkXb8i4iIiIiIiJS3JTr8BEEHXd9BV3P+CHxKQhUW3TPpYhmCDkJKxv8kzKCrKlVg/MDPmp6FrrbcEZlwg8oyutZutdVW0avpSrsM7p5MSEjownh/PKe7L9WcdFWl4o3u0NwdnaA3a3ppl7uw0Z0/VOXS1ZjqOLoLZ6GrbBjCgACtVq1axcMeLI1KqlplH4exORmTsXXr1sUhWkCQyditdHFmiIc4wkTaOHRhpjIv3FSKysB77rnHj1UqQemKXdJxG0fXZaoO6a5M4MVYo/l2F6frNxchGN+RoJCxPTm2CWHpjs42Etolu70Hod2oVGXMT4ZNoDoxXsVMFSbrNGHCBA92y7K+1atXL76gwYWM+D7geGQb6ALOecZyqUYNxykVtYSChM35ynVc5NqnAaEmFbS8j67/VJ9SqUllZ6gc5juCaQHfW6wnlaV0449/L9SrV8/3GXdkZ9u5EMMYnQTnBO9U0fK5Y4891r+fELaB15hvmpKO/4B9xvdVrsciIiIiIiIi+Viiw09uykIXS0ILuoKedNJJHnxssskm0Tuy0aWS8Iix/wKqDfmxzo1DmBfj6FFpljU9C6Ek4QhhAxWpVIES4uRS2mXQNZzqTUJcglZuQkIFGOER4R/d4qnu4jkBVtb00i53YaP7MRWqVH1SCcjd23NV4dHFmACH7s4E4YQwSyuCeCr2nnnmmWjK/Di2CPvZ79xAh5vxxCvtAtqPcIqbYiXR5Zmb+9B2hH8cPyD8pAqYAIvqZQJxuj7ni88ROnK8c5wx5i3HaT4IaQkuGZKBiwecz9dcc40fpwwjwbnEOJmhK3RcvN2o3OSiBvOhCphgjjEuA9qKY5zxLwmIF3R9CZ0ZH5T1og3jVYo8Zp58P11wwQVenUzFKSEgNwKi2pvztUqVKtEncsvnuMjapwHfRyyb7aRtGEOXbSfk5vyi/ekyz93cA9rimGOO8W1lqABujBbwHcG4rdwYiu9QQku+67jwdOqpp3pFLfNkGbRzfBsYA5nxQZMIidl3fC+XhBu1UcWa67GIiIiIiIhIPirMpoQpYUG7MYosaabVaxg9mqvqmPnHqRTJB3d7J6DmokwhokqcKlDC4wVFUMqYrLTTokZ3fREREREREZE4hZ+LEV1/uWFSGqpdF+fYnIVC4aeUp0IOPxkugMpsus0zXMaC4qZIjOm6OMb0VPgpIiIiIiIiSQo/paAp/JTyxHi6BGz5dmmXRUvhp4iIiIiIiCQt8Tc8EhFZUjCWpoJPERERERERkaWHwk8REREREREREREpSAo/RUREREREREREpCAp/BQREREREREREZGCpPBTRERERERERERECpLCTxERERERERERESlIFWbPET0WKTiTataJHs1VY+K46JGIiIiIiIiIiBQyVX6KiIiIiIiIiIhIQVL4KSIiIiIiIiIiIgVJ4aeIiIiIiIiIiIgUJIWfIiIiIiIiIiIiUpAUfoqIiIiIiIiIiEhBKnP4OXPmTLvooots6623tkaNGlmvXr2iVxauPn362G677Wa77767tWvXzsaOHRu9UmTq1KnWs2fP6JlZ27ZtbejQodGzIpdffnnx+j722GPWrFkz23PPPe2FF17waciajuQysPPOO9uWW25Z/Pfjjz/a9OnTrXbt2v4af7RVp06d7O+///bP5FpGmt69e9uuu+7qf/fdd180da4bbrjBHnrooejZXFnTF6YPPvjABg4cGD0z++uvv6xjx462yy67WIsWLeztt9+OXimSfD/HV/fu3a1p06bWunVrGz9+fPTK4nHCCSf4Ptx+++2tXr16xfv02WeftcMPP9y++uqr6J2LBsvj2E668sor52nHxYVjrn///tGzReOVV16xc845J3qWv08//dSGDx/ujxdH+8WXXxZ83/DdiIMOOsi+/fZbf1ze+vbt68tKik/n3MjXlClTrF+/ftEzERERERERkfJR5vCTgOCBBx7wUOiAAw6wm266ab6QsbwNGTLEnnvuOXv66ac9LDj99NPtuOOOKw4TX3vtNevcubONHj3an6NVq1Y2aNCg6FkR5rPvvvvaTz/9ZLfccosNGDDA7r//frvssst8XlnTkbYMfvBXqVLFPvroo+K/Nddc019bdtll7fXXX/e/YcOG2XfffWcPP/xwzmWkIRxkO1j3Z555xtubeQXvvPOOhw8Es3FZ0xeWadOm2VNPPWXnn3++ff/999FU87C2evXqHlARPp977rk+Pev9hMFsH2126qmnehC6ON1xxx2+D++66y7bdNNNi/fpPvvsE71jyXDSSSd5oC75I1gPYfziaL/48ksye/ZsmzFjRvRs8eCiTdo6ZE0vCd8BS0JgLyIiIiIiIoWlzOHnxIkTbZNNNrGTTz7ZwymMGzfO/11YfvnlF6tZs6atsMIKVqFCBdt2223tggsuKP7BTYgQQseAcOrll1/2H9igyqpq1apWt25de+mll7yCdMUVV7S11lrLGjRoYG+99VbmdKQtg5CO95VklVVW8cqsCRMm5FxGmkqVKtkyyyzjISvrz+Pgzz//tKuvvtorEOOypi9MtDPtsfbaa0dTilSsWNHXm38JQdl/yHo/IS/Va7yP6k/C5kUV4C4Igp82bdrYfvvtZx9++KFPe/HFF6158+a2xx57eLj777//+vQ4qnjZP/x72mmn2Y033uiVsVR1/vHHH/4eqnZpA6pmjz/++PlCckJxlk373HnnnX4RgjCtQ4cOPp3lU6UdXHXVVT6NymmWk6yeZrkE/HvttZcdeeSR9vHHH/uxtPfee0fvMLvttts8tAfhNevcpEkTX/9g1KhRvgwqmx955JFo6lxURx999NH+Oc6LwYMHR6/MxbzZBpbNsRxw4aBx48Y+/Z577ommzpXV9sl1nTRpkt18881ewctFldB+uOaaa/zzhKFc6EGudg2y3kM1c7du3XzZO+20k88/ufwePXr4BZ4gVFBSqc4Fj/3339+/g0o6JoKzzjqruF2pvmZ+8apN9j3HABewaE/We/LkyZn7m/Cf85Xv/bjkdLaVixxcfOIYYp5p2w8u/LBN4bmIiIiIiIhIeShz+EnVJT/YQZAHunsvTIQxdJEkNCQkIOAgyCAMxVFHHeVdpOMIHOluTgCK559/3gMq/PDDD/OEluuss45XZGZNR9oy6F76zTffeNDKEADXX3999Mq8vvjiC19+nTp1ci4jTcOGDW2LLbawHXbYwf8OOeQQ/wwIVwjO2Na4rOkL08orr+zVc6xjXPv27T3cJeQjaCEYQdb7CcbigShtlat9FjfCJ7p5Uwl97733eqhF0EN3XvZ55cqV5wkGA46dM888088hQlNCOt6//vrre9XrrFmzPFh9/PHHveoYVPMGVNJSlcpftWrVoqlFeI1gjSD53Xfftc8//9zPmc8++8wray+88EKvKE568sknPaAmhCOcfOONN6JX5ke1LkNAUM3M+wkXv/zyS39t5MiRvv20R9q2s3077rijbzvryTYkXXrppR76Ue389ddf288//+zbxflM6Eubs75U4QZZbZ+2roRyp5xyip+74XsBtA8hJsEh1dlUT48ZM8ZfS2vXpLT30O4Ey1St8/foo4/6PktbfhrWh2Nh4403znlMxLVs2bI4/GSbCGO5gBLH+l188cXeplzYob2zUGnP986tt94aTSmSnM65yhARtDHHMvsqbfu5+MFxyPAgZ599tn9WREREREREpDyU2w2PnnjiCR9n75hjjvEgY2GqUaOGV2BR6VWrVi2vRiJIK6kikOojwhMQhtDlHQRLoQIRoUtp1vQsq666qle/8gOf5RAKEzLhn3/+KR4fkoCAatVDDz201Mt48803PTx48MEHPUxiOYQxhAsrrbRS8Vh/Qdb0xYXwaPPNN/ewiiq28847L7USMqBKLI72SU5bkoQqOaqhCXioemT7COEIRAkCGQ4hiWOHY4LK3g033LB4PhtttJHPh0rZu+++28M9xtH85JNPiiv3GPeTyr769ev7vk7abrvtbPXVV/d5U+n8+++/+3FEJSbzZRrjlybxOYJJxr8knKYaMMt7773n60/ATmUvxz7V2aDykuAxhPRJVLwSelNRSGiWNo4k502XLl08sCQwZ3sIAQk4GT+XcJ92CtW2yGr7XOuaxDLYFwSFVGdzHvF5pLVrUtp72E/sL0LLa6+91qflOgeSqPRkfXIdE0m03/vvv+/fQ3QtT164AW2w1VZb+WNezwpSS4NQlwpPMEwE+6is2y8iIiIiIiJSGuUSfoaxG+l6ntb9s7wRdhJiEHweccQR3pWWQJQqq1wIewgzCEUIDwiZsMYaa3gFZkC1EhWGWdOzEG7QxZUwc7XVVvOQIlSJxcf8pMKNSja6rJd2GYSpVIYRVlEFSvXrq6++6iEnVVRsI5VuhIwEIlnTFxe6Eh977LFezUmVG4FSfNzUJCrQqP4MaB/abEnFfo8j/KEKmLbnj9CXrtpJBFlxhIVxhHx0/6eymC7OIVACn6UqkYo9hnNISqv4JUCOh+50b07abLPNvAs2FcqsM93Gk8LnmF98nZkeLkYk2ySJeXNM8r5kN+qAm3xRZc6xwEUMhoxYbrnl/GJLaFvWlZAzyGr7XOuaJt5OhJgE8MinkjrtPQTPbCcXPg4++GD/HsuFauJ4OLj88sv7v7mOiSS+7whAOU7olp5WnZ/czrSLMGnHSS6E8fGhOVDa7RcREREREREpizKHnwRrjPNGN0p+6FKJmBw7cGEgLGHsOvDjn7FHN9hgA3+ehbCEAICQlm6gAQEioSIhw6+//urVY4QmWdOz0F2Xrsv8qP/tt988jN1mm22iV9OVdhl0dWW+hDVUURGmMo0xCplOl9b27dt7lR5hUdb0xYVKRqpuaSO6ANOFeb311otenR/HVajWJTShW2wY3mBpQLUbY2VSnck2EyDSzbe0OMYJghkagNB+xIgRxeEUFXu0Id2F870hFF2RCaIJ8gjoOX+SOMdYVwIq1pv2J0TjOOXco8owjIvJkBIMZ0D3caoLuSgRur2XhIpEwjCGs+DCRDJ0Y4gLwnuOc84vhn1g+xkvkm7fdJnmfDjwwAPn6Xqe1fa51jVZVUw70TWedSL448IF1ZxlwTqF8Vxpf86BEG6G5XOMh+9Rusyz/km5jok0XGxgnFfC4zSsB9WxoJKfbc/a30GyvYKs6chn+9mnLDfXYxEREREREZF8lDn8JHwDAQFdT/lLu2FJeSLAo9sqXTMJMwn1Onbs6GFHSbhRCD++4+EnVYjMk27ozItx76iuypqehfcQWhDU0A2fKtCSgpLSLoMgiko8giK67fNvroqvJQ0BFJWe3FSF0IZur7kq6NhGQnXak+q/K664Inpl6cC+pMqXfcx+Itzhxi+lxY2wCDgJrjjHCM0ZtzIuhFpUQJaE8ImKO7qkU0nNUARU4cYRJnJes6/YBo5NKpgZtoHziOM8VE8zniM3R+L45MZAHJeElPmgWpNtYf2p7CXo69+/f/Sq+bijnOusM+/heGH+hJj8y7nMa4x/y3YEWW2fta6cV4R78TuO0z5819D1nSpL5pU2REBpcCzTdZ5tuf32230ZVM/Hl8+2MI4nrxHwplXP5nNMxPFdSTjK8tNQ6cuYqOxXwnDmmbW/Qbsw9nFS1vQga/s5zwlDGceZkJnlIuuxiIiIiIiISD4qzPmxWdSHU6QATapZJ3o0V42J46JH/13c4IiKz8MOO8y7TxO2hZsCSWGiwva6667zO8QnUWXK3dYZd1VERERERESkkCj8XAJRdcWNXdK0a9cu55igMi+Fn+noxkwXebpOM0QDXeYZo1YKE2P/Ujndp0+feapjA4WfIiIiIiIiUqgUfkpBU/gpUjK6wzMGKjeOExERERERESkk5XK3dxERWXox3IGCTxERERERESlECj9FRERERERERESkICn8FBERERERERERkYKk8FNEREREREREREQKksJPERERERERERERKUgKP0VERERERERERKQgKfwUERERERERERGRglRh9hzR42JTp06NHoks3abVaxg9mqvqmBHRIxEpJNWqVYseiYiIiIiIiBRR5aeIiIiIiIiIiIgUJIWfIiIiIiIiIiIiUpAUfoqIiIiIiIiIiEhBUvgpIiIiIiIiIiIiBUnhp4iIiIiIiIiIiBSkMoefs2bNsh49eljjxo39r1+/ftEr5ePWW2+15s2b2957723HH3+8jRs3LnqlCHem79WrV/TM7MYbb7SePXtGz4q8+OKLdtRRR/njJ554wvbdd1/bb7/9bNiwYT4NWdORXAb22GMPa9SoUfHfjz/+aNOnT7fNN9/cX+Nvp512si5dutjff//tn8m1jDQ33XST7bnnnv730EMPRVOz5/Puu+9a27Ztfdl33XWXT+vbt+8867nNNtv4/lpUPvroI3vmmWeiZ2Z//fWXnXHGGdasWTNr1aqVr3Nc8v0zZ860K664wvbZZx877LDDbMKECdErUhocE/9Fl19+ec5z7dprr53neFvYbr75Znv88cejZ/njO4hzYHH57LPP7KWXXoqeLXp8p/OdwXH84IMPRlNFRERERERESlbm8JMfovyYP+ecc2zXXXf1MOHzzz+PXi0bQosXXnjBHn30UXv22WetY8eO/hfCxDfeeMO6du1qY8aM8ecgFBwyZIjNnj07mlI0n7322st++uknu+OOOzxIvPPOO/0HNfPKmo60ZRByVqlSxd56663ivzXXXNNfW3bZZX15/A0aNMi+//57e+yxx3IuIw0h4PPPP29PPfWUf/62227zeWXN588//7QLL7zQw1/CnIEDB3og26FDh3nWc/vtt7dDDz00WsrCM23aNN/+iy++2H744Ydoqvl6L7/88r6PLrnkErvooot8etb7hw8f7tvNa2wLQagUnn///Td6tGgdd9xxtvvuu0fPllzLLbdc8QWNYFG22Zdffmnvvfde9GzR4gIJ34d8Fz788MPWp08f/24TERERERERyUeZw8+11lrLTjnlFGvZsqWHn5gxY4b/W1a//PKLrbfeerbCCitYhQoVbOutt/aQNcyfH+RrrLGGPw5q165tq6yyin344Yf+nPe+8sorXj366quv2m677WYrrriih5X169e3d955J3M60pZBGBfCzlxYj5133tnnkWsZaSpVqmSVK1f2kLVq1aq2zDLL+PSs+VCVxbKYRlBCYLraaqv5ZwJCUd6/8cYbR1MWHsJM2onjI65ixYq+PfxLCMp+Rdb7CZEPOOAAf1+TJk1s7NixXgUnpUMF7dVXX22HHHKIB36TJ0/26Y888ohXFFJVFy4sHHvssR6Ug4CN443333777daiRQvfHw888IC/HkewfdZZZ/n8eF+olmYfUoEZEGJTwc0yCOxPPvlkD7ZGjRrlr1ERfOaZZ/oysz6LtPXhwgTz5GIH1d7ffPONT89y7733emU4+A5j/Vk+VeZ//PGHTw94zoWQ/fff39uQ9cVBBx1kv/32mz8eOXKkde7c2R+PHj3aLzSwjmeffXbxcfvpp5/6/Knc5jxN+v333+20007zdjziiCPsq6++8v3AMsF20dZHHnmkX+S57rrrfN258BOqM7PWqX379v4dyjpxLFBF3qZNG9/m8J3JhYZ27dp5u15//fU2adIkryDnYgwXoQgeTzjhBG9jvle5kAGqWi+99FI7+OCDfX3CtqVtD9L2X1ob83120kknWbVq1WzVVVe1lVZayXsciIiIiIiIiOSjzOEnXbL5IU8VIlV7rVu3tk022SR6tWyYN92k6fJOhePLL7/sIQ1hKA4//HD/kZxECPDcc8/5Y6qV6tat60EgP9rjoeXaa69tP//8c+Z0pC3ju+++s2+//dYOPPBAD+SoREozfvx4Gzp0qG200UY5l5Fmiy228C70YTgBlsVnsubDOhEW8z66hlIBSoAaEAoR9BxzzDHRlIVr5ZVX9uNiu+22i6YUIVQhrGXfEgxxzCDr/VS6xreXx7naTdLRZgx5QBU1FxQ4PwiQnnzySa/e5jgF5wuBVggE33zzTb/oQID22muveZX3/fffb08//bS/lsTn//e//3nlMa+X1FWa16kAJqzjO4QgjCCUgJ5lZmFd0taHwI3zgACP74y33347+kTJuEjBdw3L32CDDWzw4MHRK0W4eEBgz7bxvVDSvAkZCT2ZDxcvwvwI9KjkJvxjWI8kvk823XRTDxppl7T3sD8Jrgke6ZLOOjNMxpVXXhm9IxttS5DJhSLajeCRUDR0x2ceBKo8nzhxogfnhM4ElQSY7LMddtjBjyEC7vhQJ5988olXZ959990ehiJte7L2X1obc8GG70BcddVVtssuu/j3noiIiIiIiEg+yu2GRw0bNvTKQ37gU9lUHmrUqOFdPak6IozgMWNallT5R4BBmEO4w7/8aAfPQ6UhqJqiMjRrehbWizBgwIABHrYQQBDM4p9//vHwkT8q6QiOqLAq7TL40U9Xe0IEurkTNFD1mDUfKicZboAAiWougoQQYIFwhMpcgoXFiX242WabeeBBkEH4mav7LsFLHNubnCYlo2qOCk7Uq1fPqyqpviWIogKToIoQjZCcCwzheGbYCSoUCTWpJGQMW6oj+fyIESP8PXEEU1TmUd3LeVdSV+ktt9yyONzecccd7ZZbbvFjmMpB5pUla33oIs36EvwTkDH/fFFhGNqICybJyk/OZYI/hvYgrA+VmGkIJ/k8gTPtTEUk6wW+G6jqzgrwOPepqgQXMwiHkwiowb7jQgLbu+GGG3qYXZKmTZv6v4SfdPmnupwLNLQh2A/nn3++B+Wcn6uvvrpPD/gO5iJFGHqDYyYI28Y+ZbuRtj1Z+y9XGxP2cpGHdRMRERERERHJV5nDT7ql8iOZH8P8qKXbLF2zywM/rKn+4kc9VYL33XefdyV//fXXo3ek44c3n6HCMFSLgh/xVE4GBBS8N2t6ljp16ngFEyEkFaVUJYWuuPExPwks6YJL1Vdpl0FXfaqsCGGoAiWUYfzRrPnQHZQbLPEvXeJ5/MUXX0TvMl+XJWFsQwJZqj/p3k44RhVvfDzVJIYcoPozYHuTYYyUjGMiDJ0Q0J2ZbshUMXO8hGErOKbZP1TxffDBBz6dMJP3ciGCP855qgVzIQRLC6qnTJkSPbJ5wnhCtRtuuMHDOG6KRTVgUvhs1voQuMUvDpQm7Oe9yTaKo3qR5RAUUrUZqpbjGHsXbHd8XlwUCWEq52gufJa2A2E/x3xS2C4ufMSXwz5NXlQJ6xTEK8LjjwOGR+DCDecdXeKpiI1j2wnLOU6o1o4j+E5K256s/Zerjamkpxo0bZ1FREREREREspQ5/KTasHv37h5MElyAiqLywo9sur6DkIa/mjVr+vNcCA7pksmPaAJTEFISKhLQ/vrrr15pxA2AsqZnoZvnueee61WYVC/RfXOrrbaKXk1X2mXQhoS8VLkSmlDJybSs+RBQEY7S5Zcx9ngPoSmoqiI4oPvo4sY2hKpcwlm61a677rrRq/NjWAGCW1BBVqtWreJhD6RsOJcIlwmw1l9/ffv444+Lw0oqC6m+49gijGzUqJFXFVJhzDHJnffj4XrAMcvxRjUgVeBUPhK8cvyByj2Wk4b1ILxjTEtCNy5eZH02a31YHqEpxxfhHccMeB/nS1nwXURVNdWLdGcPY/ayjqEt6E4OLkgQSnJ+EvjRBT/fu8qzDWE+dN/P1ZWdi05UnrO9nEvhZmZp65QPwmUqVDlPGaeToTdCm4djg7FBGfOTilOq/HNVsCNte7L2X1Ybg7FhmZeIiIiIiIhIaZQ5/Dz66KP9Zhl0V6WLLN3B+VFcHujySDdIfhjTnZKbXvCjO58xRQlvCGZDd0tQzcbNXFhn5nXeeed5BVXW9CxULlavXt1DAsIGqkBL+lFe2mXw45/gljEQuYEIbUr1Z9Z8GBaA0Ih1o4KOdWNcPhCEEjbEK+IWF7q40n2fqs/TTz/d794ewuk0DGFAkMM2EYykdQGWBUM4vs466/iwDARNBOvhJkUcb3RNZvxc0H2caRzrjIHL8cnwBUnMk+8Ajtttt93Wz1v+ZR8zL15r0KBB9O55cexecMEF/n1CaEkYmvXZrPXhXGFYCo5/til0D6ca/cQTT/THC4r14juOY5chGzj3wPnITZnoqh8qUznXqKDk+OYY5gJGCCZLws2J3n//fZ8f43FyzmShXTj3aQO2j/ODKsu0dcoH32tsJ/PjuKDrOe3MdxFV2wS4VGhynPA6VZwEr2G80DRp25O1/7LaGFxkCzfiEhEREREREclXhdmUJSWUNKamyNJiWr2G0aO5qo6Zf6xKKTvGgKSLNSHlkogbfnHRQAoXY9uKiIiIiIiIxCn8XIy++eYbv2FSGqpdc40JKvlR+LnoLMnhJ13C+V7jZk9SuBR+ioiIiIiISJLCTyloCj8XHb436AJN12mRxUHhp4iIiIiIiCSVecxPEREQPCn4FBEREREREZElicJPERERERERERERKUgKP0VERERERERERKQgKfwUERERKSgTbFrLOjap5py/lnfazGiqiIiIiMh/kcJPERERkdIYf6dNJliM/U0ZGr22JBh6m00d2dqqTxxnNQYdb5WiyWU1s++Bvq2T+06IpkSGdp2nLZKvZ34u0Y5LVBuKiIiISMFQ+CkiIiKSL4K+xj2tYr9xVoNw0f/6mrXvatOjtywRGtQpt9DTbLhNqVnH/rJOVq1BNKnYnNfaj7VqL0dt8XJXsx5n2rTx0Wu5PjenHa3bkKLP9Wtt09sfGH1ORERERKT8VJg9R/RYpOBQSZLEjywREZHSI8zrYNZvnFVvFk2aD+/pY5W71bWpPQbMeU4FZk+r4l3Rm9vUkUXvmjsdRa/N6BybLyHr9XVsJa/c5PUzzfp0shmNO0Qha32r9vITVrW2PylGleUfPUZFz+Zo09dqXNd0zoNcy89a56SieUxvNcRW7FCraBLVm43HWdXkvOZZt5TPsX3tLXcbiIiIiIiUA1V+ioiIiORj6PM23VpblRLDuVE2tQfBHtWQhHtR8LhJ36Iqxzl/K3Uba1NqlqZadM48GxMqhs/bnOfzf75Shye8itIadLWVWFY8+My5/OQ656n28Va1zYBoXiyng83q1mu+UDZp5hdjU6tTZ32R6BovIiIiIlJGCj9FREREFkAYy7Lob94gskq/WIA4fphNZwxODyKLVOrQac7rY21WKbp5V+k3t5qyVJ/Pc/nzrHMpVLlunFX3ALS5TbWutnyo7iyVWlZxk+ihiIiIiEg5UvgpIiIisgC8ypJKSSotcxk/zmZED+eqZZUbjLIZZRrjMs/PL7Tlg27udWz63lFFaatB9kfNBRm7c4LNGh09FBEREREpRwo/RURERPLRrIVVsQE2vbR3Ja9dxypHD+eaYDNG1rfKJXQPzy3Pzy+05VP92semN+hqy0VDAVTq0MuqNRhl01/I3X290kZ1zUaOs5nR86DiRgtSNSoiIiIikk3hp4iIiEhemtpy3erb9PZ1bEppAtDae1iVBgNsSpfh0YQQGra0ZTx8LOryPX1w9Do3EWrPjYfmFb8b+ryfL0GJy19w84WY3sU+jxAzGSQPvc2mjsxnPFURERERkdLR3d6loOlu7yIiUu78Duc95+lKXrlbuJN51h3hi6YXjwvKDYn8Tu5B/PXWVr2f2ZR57vbOndD7mrWPvSfrxkTz3Ck+yLX8rHUOEp+NhG2e3qWOTekfTZwj2RZZn5u3HdPvXi8iIiIiUlblFn5+9913dumll9p6661n3bp1i6aWXZ8+fezRRx+1SpUq+bwvvPBCq1u3bvSq2dSpU/09Xbt29ee9evXyafF1GDZsmN1+++32yCOP2GOPPWa33XabVahQwc455xzbc889/T1Z05FcBnbeeWf766+/omdmzz//vK2yyiq2ySab2Nprr+3T/v77b3/f1Vdfbcstt1zOZWRJLvvmm2+2O+64wx9j+vTpdsghh1j37t19G2kr1mv//fcv/kzauq655prRs4Xrgw8+sIkTJ/r6gPU4++yzbeTIkbb88sv7MbPDDjv4a0i+f+bMmdajRw97+eWXbeWVV7Zrr73WatfO/5fR4gw/+/Xr5/tkyJAhtsIKK/i0t99+26dzHJTW2LFjfT9//fXXVq1aNdtrr72sc+fOfozsu+++Nnz43Iqe/7Irr7zSNttss+JjqLT23ntv69+/f/E+w6effmo//PCDNW0692YhuXDOvf7669Gz3DiHmzVrZq+88kqZ1z3Liy++6Mch88/Hu+++axdffLF/d9WvXz+aKiKLTwg/s8JJERERERHJUm7d3gmo+IH91ltvRVPKjh/rzz33nD399NMe7Jx++ul23HHHeaiI1157zcOf0aPnjpBPaDB48GCLZ7rMh3Dop59+sltuucUGDBhg999/v1122WU+r6zpSFsGYUWVKlXso48+Kv4LYeKyyy7roQd/hK6Ewg8//HDOZWRJW/Ypp5wyz3J33HFHO/LII+2dd96xJ554wp588kl79tlnvb1eeumlnOu6ME2bNs2eeuopO//88+3777+Ppprdd999Vr16dQ96Lr/8cjv33HN9etb7X3jhBW9D2vLUU0/18G9p8s8//3hgWx44Fg488EBvO0JuQtA777zTg/UHHnggelcRjv8ZM+a/vcWi8u+//0aP8lPa9+dy0kkneZgYV9b5jx8/3oPrhS1t3ctDo0aN7KyzzoqelYzvzI4dOy6y4LM897+IiIiIiIhIXLmEn4R0hAO77LJLNKV8/PLLL1azZk2vwKJactttt7ULLrigONRhmckgb6ONNrIaNWrY+++/7895L6Es1VyEgbvvvrutuOKKttZaa1mDBg08rM2ajrRlEMbxvpJQCbrbbrvZhAkTci4jS9qy4wYOHOjzqVOnjk2aNMkOPfRQr6akQnKnnXby9cx3XcsbYSbLDlWwQcWKFa1q1ar+LyEo+xVZ7yeEOeigg/x9VN0RBFPpuLRgnxCEEzonjRkzxrdtjz32sKOPPnqe0DcN5wOVxbQFx9GZZ57p4RTBEQE4qHy+6aab/CIAx09SCJFB2Lzpppt6QE6bbrXVVj4vQvrGjRv7OXPPPff4e+OoxqWyukmTJn6cXXPNNT79jTfe8DCbCxQE8V9++aVvP+fAMccc45WTSW3btvWLAVQvI23ZnMMXXXSRnz+tW7e2E044wUNx/i655BJ/D9q3b+/VsQTCQ4cO9fVs2bKlh+yE5lQdd+nSxdeHEJn1Be1+xBFH+H7gAgvHYhznFhXXXFTgQkw++43wuWfPnta8eXPbb7/9bNSoUT69pDYJ6w6+t/g8y+GiAPuG9QzrzfPtt9/eJk+eXOI+o5IzhPCHH364twltw0WhDz/80KcHfHdyAYl5fvLJJ/bQQw/5ucf3+/HHH+8XbbLWgzZq0aKFrzPHCMfWn3/+6esVUPXMPk/uHxEREREREZGFoczhJz986bpMJeMyyywTTS0fdOudMmWKhx6hspQf1qE76lFHHeVhSBLBz6BBg/wxFZH16tWz1VZbzYOGeBC4zjrreEVm1nSkLePbb7+1b775xvbZZx+vqLr++uujV+b1xRdfeBdzwslcy8iStX0gVLjrrrs8jABtRRCDr776ysNWApZ817W8EcBSxRbv0g4CKkLfXXfd1Q444AAPPpD1/h9//HGeQJQ2LKndliRUAnN+xEP7gCCOSl4CSfZVaIssDBdAYEb4d++99/pQEOzTJCoUGWIhPjxEQPsytAD4d9VVV7WPP/7Yw9ntttvO3nzzTR9igPOHrt9UEie7b3/22Wf2xx9/eHUxf1ShhsCQbbniiiusTZs2vn3HHnusV6pyHJ922mn+niT2J8vhfWnLZls4jgkFCeRC6JYPwlCONb4/+Czbyzr37t3bzjjjDPv999/9u4v3sO6EjZw/cVxMYT9xDhFk5rPfOD/5HOE9x3Wousy3TQhcCQ8ZIoHvkMqVK9uNN97o6xDCUdqGC0Lsu5L2WdKsWbP8/SeffLLdfffd0dQi22yzjX+fsG6E47T/448/7he5wHdq2npQ5UwYTbBJMP3zzz/7+uQS3z8ikkstqzpIXd5FRERERBZEmcNPfjhTMUZwUt4ID+jOSyVTrVq1vGKIwKykyj+qmQgM+IHPv1QXgeeh0hCha3DW9CwEKFTPER4888wzXu1EMAsCAMb7448KOEIBKr1Ku4ySUFlH5R2VngHzpFs5ARlj9TFGaq51XRz69u1rm2++uYciVCmed955HqBnoTosjm1MTlvSUR1JiERYHVANRyhNtR4I1QiVcjn44IM9NGvXrp2Pi0p4RjCVRJUeQx2kodqWaurPP//cx13lWHnvvfe8MpAgj+CU4K1Tp04efhFyJisDqTYlzCMQo5qQADHsw6233tqrlTkPqHakapDAm/emVb8iVAVmLZuwk/OesJfwuzTfNZwfbBcI3Zkf4TEhHetMO7Dt4SID4/BSmZylNPuN10CgSnUoYWC+bcL7WD8CUNaXfcV7ufhDeAuGBKFd8tlnSYSX4Lhk/2WhQpvveALOG264wStBCXbT1mPEiBHWsGFDW2ONNfxzfO+WdEzH94+IiIiIiIjIwlDm8JOKIyq/uAkNVU7cGIRAqzwQdvKDn+CTbpbcsIhAlAAoFwISPkPYwQ90fqiDH+XxbqZUnPHerOlZqKijso0wk4pSwia6wiI+5ifVl1T9URFb2mWUhCAzPjYggSDVaAQQdIcPAVGudV0cqBaj8o1qTgIYum/HxzRNIkij+jOg3WjLpQ3dlqmII7QM4mE4YVEufI4KQ250RFhEKMb5EQ9Ug3ggnobuyxybVBMyL8I/AlAq8Bg/lECUikP+6OZN+BZHdSgVgwT6BLKca0FYNsc8xxsXLpgP/2YN8xA+k7VsLhLE24oQNA3BZFK8LWi7q666qnj+fDcQ1sbnTZvkCj+R734L7+Nf3sd8820T1pWq3rCuXCygcpXPc+4QjrLfuACSzz5LYkiOfBCq0sWfwJfjhuUhbT24MJFsS46RpPh+KulYFRERERERESmrMoef3H39f//7n/8RuG244YZ24oknRq+WHdVG4U7l/AAnBNpggw38eS6tWrXyMffoch5+6NN9nqpHxqz79ddfvTqKgCFrehbuts54i/yw/+233zyMpatoLqVdRi6MrUeX+i222CKaYl7Ryfpcd911HigGC7KuCxNjsoaqXKruuGkPFapZGDuQoBeEboTs8btwLy0I7elmHcbHZBsY+uDVV1/15wRWIbBmbE8qJ+Oo4KU7c3g/1XcElvHgMYnjJC0QJMQiJNtyyy39/GBsUI5JziuqVKkmpRs7FdaMjcl+iqObPBcUGK+TwIt9mKzeJezbeOONvSoQXCAJN7fKkrVsjle2nWOGIDwEhhznYd3oFk/wnwvzJ4DGuHHjPHxnnrQ7ldTgHM2qhOQCQ679lkS1NegCzvcin823TajIpJ3pgs86Ep7zfrDeDC3Azc6o8M1nny0ovnO5AEHXfbaBNg4V68n1oOqTi1UcvxwXtA3DLPAaxxff4xy3obt8LmwLn8n1WERERERERCQfFeb8SJ17W/Qy4gcyP5ZDWFVWrBrVTtyYhbE/qYai4iw+DiY/tglICXPiCH744c1YclQuBf3797cHH3zQ583NTQjXck1Hchn8gGe+VHcSBFB1RTdWphMAMCZimlzLyJK2fVR2Mv5eCNJw8cUXezfaeNUaoSeVeWnruqhQnQiODRCMEPhQfUqVLI/jFazJ9xP8cDMUQhfeT6BNCJOvSTXrRI9ERETS1Zg4LnokIiIiIiKFplzDT5EljcJPERHJhwJQEREREZHCpPBzMaIL/8MPPxw9mxc3tSnLmKBSROGniIjkQ+GniIiIiEhhUvgpBU3hp4iI5EPhp4iIiIhIYSrzDY9ERERElmYKPkVERERECpcqP6WgpVV+6keuiIiIiIiIiMh/gyo/RUREREREREREpCAp/BQREREREREREZGCpPBTRERERERERERECpLCTxERERERERERESlIqTc8mjp1avRIZOk2rV7D6NFcVceMiB6JSCGpVq1a9EhERERERESkiCo/RUREREREREREpCAp/BQREREREREREZGCpPBTRERERERERERECpLCTxERERERERERESlICj9FRERERERERESkICn8FBERERERERERkYJULuFn48aNbfPNNy/+e/fdd6NXyu7WW2+15s2b2957723HH3+8jRs3LnqlyNSpU61Xr17RM7Mbb7zRevbsGT0r8uKLL9pRRx3lj5944gnbd999bb/99rNhw4b5NGRNR3IZ2GOPPaxRo0bFfz/++KNNnz7dt5/X+Ntpp52sS5cu9vfff/tnci0jDe3Ytm1bn9ddd90VTc2ez9133+3TmjZtWry+s2fPtssuu8z22msva9asmT388MM+fVH56KOP7Jlnnomemf311192xhln+Lq0atVqvmMl+f6ZM2faFVdcYfvss48ddthhNmHChOiV/4YpU6bY/fffHz1bOvXu3duefPJJe/nll+2iiy7yaa+88oodeOCB9uWXX1rHjh1t9OjRPj0fBxxwgP3555/RsyKfffaZvfTSS9GzkrVv396+/vrr6FnJLr/88rzO2QXxxx9/2KGHHmr9+/ePphSJb9PNN99sjz/+uD9e2Pi+SXP44Yfbd999Z9dee+0856iIiIiIiIjIkqzM4SfhzC+//GLdu3f30JG/2rVrR6+WDWHDCy+8YI8++qg9++yzHpLwF8LEN954w7p27Wpjxozx5yAUHDJkiId+AfMh/Pvpp5/sjjvusIceesjuvPNOX1fmlTUdacsg5KxSpYq99dZbxX9rrrmmv7bsssv68vgbNGiQff/99/bYY4/lXEYawp0LL7zQw1yChoEDB3rAmjWf9957z5566ikPNwcMGOChCQET/06cONHbj89cf/319uuvv0ZLWXimTZvm23/xxRfbDz/8EE01X4fll1/e99Ell1xSHIZlvX/48OHehrzWoUMHD0L/S2iXQgmatt9+ezv99NP9MfufCwMbbrihHwdl/c4gROUcyNdVV11la6+9dvRs8frggw9ss802szZt2kRTipR2mxaV4447znbffffo2cLx77//Ro9EREREREREyqbM4edXX33lYdaee+5p22yzjYePq666avRq2RCqrrfeerbCCitYhQoVbOutt7ZzzjnHZsyY4a8TDqyxxhr+OCBEWWWVVezDDz/057yXEJDq0VdffdV22203W3HFFT2srF+/vr3zzjuZ05G2DMK4EHbmwnrsvPPOPo9cy0hDaMlnee9yyy3nAepqq62WOZ/ffvvNDjnkEN8XK6+8slejsp4rrbSSnXDCCVaxYkVbffXVfR6h/RYmQjuWv9Zaa0VTirAeVatW9X9ZV/Yrst5PiEylH+9r0qSJjR071itxlwbXXXedtWzZ0iuWqTwOVcu33367tWjRwrfrgQce8GlUwHbq1MnatWvnn6FaF4TbHD9UT6ahapB2AcE4lcfvv/++Pz/iiCNs5MiRHq61bt3al8m/X3zxhV8cYJ4HHXSQVxc//fTT/pm4n3/+2Y8d1oeLDlRqEsqzTlyUCEKlIOE87+dCA+cbAWcc63XDDTfY66+/bq+99povn20j/Bw/frxXBZ933nn+WSohuagAwvBjjz3W1+Pss8+2f/75x6cHkyZNsr59+9rzzz/vIT/HCFWKvJ/1iYfpwbnnnuvHG8cXVZ0BATv7iQscXHxgW9h333zzTfSO9P2ate5xkydP9qpnPku70x4cy1zgoCqWiu4guU349NNPfZlUd/N9AL5/jz76aF/uiSee6Psgie/kM88807eFgJX2AevA9wY4Tjp37uyPOTaoHGc5Bx98sH3yySc+Pbj33nu9mh5clKCCm3nfcsstPi2O1zmmOda58BKknQO0GW1+8skn+4UckSXZzL4H2uS+C7snwgSb1rKOTRkaPS13RfOfVHPOX8s7bWY0dekx3KbUPNCmjY+eioiIiIhkKJfwkzCNrtZ0ZebHOcFBeSBQJVSgyzuBCwEByyAMBQHH/vvv74/j+LH/3HPP+WMqp+rWreuhH8FAPLSk8ouAJ2s60pZB189vv/3WwyCCpz59+kSvzItAZ+jQobbRRhvlXEYalkH4yzIIlwhIKlWqlDkf2oqwC3TnJSTdZZddPDDebrvtvMvskUce6fsp/vmFhQCWY4FlxxGEENayvoREVHoi6/0EevH15XGudltScKzSlZtQkcphuvOD0I8/9gfd2Xn9zTff9NcIBQmdCH4Ixggp+SzVkYRmaQi3OQY+//xzGzFihD/ms4TJVPwShhKqcdGAII1Ai/kT2BFgUVVNmMc6JdG9mWpNAixC0xCaZSGw32GHHfzcYzv69esXvTIvQn3+zjrrLN+2gIrmGjVqeNh39dVX+zr//vvvXqXJEBKsB+dCsrs6nyG0JExjeARCSIJA3s85wLxKi3CR84958N3z9ttv+/Ss/Zq17nEM4bHuuuv6PC+99FIPcpdZZhk79dRT/bzkXA+S24RRo0bZbbfd5sEh8wLbSgBL0Mz3FG2axHpQWcp+IUTu1q1b9Eo6vr+5cMM2ck6ef/750SvzIlRmPxOGMqwB285/D+KuvPJKP7443jkeOXdznQMcQ4ThBLMiS67h9ncPsyp71oqex0LE4r+uVj7/J7TgCGgndRkePUsx9DabOrK1VZ84zmoMOt4qRZPLypc7pw3mC4eHdp2njZKvZ35u/J02Ofa5uWFwU1tuztfZ9Bf+W8PhiIiIiEjplTn8JFQkVOMHfY8ePfxHLI/LAwEAY11SAbXBBhv4Y6rUSqr8IywldJw1a5b/S4AAnodKQ1DhRHCbNT1LCCboXk5IQvBAKAKq0gho+KNajvCRiszSLoPwikCLru2EV7QrYVWu+fCYbuWsG/uCoCUgPGI96EZOsLq4sA8JYgg9CLUIP3N1cWXMzzi2MTltSUTAS6BFhWudOnX8D4TxVNvR5ZugimpAQktwrFCdS8jN+6nqzMeOO+7o86XambEsqfT8+OOPbdttt/V5MbQAxyVjwhKAEW4ROvIvQRgVhRwvSVTiUdUHxvWlUjcXzk3C6zAkQ2kvgtBmtMUpp5zi1ZgcF1xAIMwN60GbUjmchcpULkzsuuuu/pzvggXpOk4lLkEx7UegvOWWW/r0rP2ate5xzDMEmYTSVHQTCOaL75TKlSv7+oB9yr575JFH/JwngKSCMw3VpmAeBJRcVMolrCfvJ+TkOE3iOOOYJShln/B9SKV+HMcm4SkhO+c6x3euc4B2XhQXZ0TKZOjzNr1BS1smMVpH5W5DrAZB4sQhVq3BAJuSK3hcUjSoU26hZ1ElZh37yzrN2f5oUrE5r7Ufa9Vepn3m/L3c1azHmVHVZgmfa9zTLLRtv9Y2vf3cas9KG9W1GU8NWwqrVkVERERkUSpz+Ln++ut76EF3ZbpeEszRnbQ8EKDwY56ghirB++67z39oUyGXCz+e+QyBRKgWBT+8491CqULivVnTsxB4UJnEthL+EgyFLs3xMT8JmujGSXVXaZfB0AEElvxLF3ceUwmYNR8CQbqtEnoRhjAEAdh+KvYITOhi2rBhw+IqtsWBAJfqT44XQmmqeOPjqSYx5ADVnwHbSxss6dgf8ZCasXFBSESFLoE+f4RCBJbg2F4QVFESKLHvqf6jMpJjn+mgGzGVwPXq1SuuqGPMWkJ1wi3GtQ0BfRyhetgG/o1vT8B4syG8piKRG/NwTlAxWFrVqlXzisjQNlQSEobFl0v4xzmWS/z9hJSlEfYTn4vPJwS/Wfs1a93jCO7jnyVYTbZ5LsnhRPheYRrtzjL5N3RHz8Ly2bbkBYTkDaSC8H7WNYl5sD8C5pG8MEUVLBeBOIfpcs8wB7nOgZICdpElwfTBA8w2qZ0jNKxly7SqbzZ6fBTKFVWGztN9nSrIeFfzzOrGGK+cjHfzTlachmrToul/9Bhl1r+Dv5ZaZdl+znaM7Gl/8NnioDZrnoi6mfcNFZzJ6tamXkW6YodQERszfrzNsrpWMQTGtWvH/gc0x+cImq21VQ2vNTvJqjUYZTNCGzRrYVVGDrJ/573WJCIiIiIyjzKHn3R5DGPQ0cWRH/jJrstlQZgSqpSo6OKvZs2a/jwXKpfokkmX8xAqEVIy/ieBDTf9odqIbr1Z07PQnZcxAwkuqGCiC+dWW20VvZqutMugco1Qiq63dFvls1tssUXmfOjSzPqwDwhLA6o86SJLFR7zoZtuvKvxosaYrKEqlzCXyrd4hWoSwwqEIQwIbWvVqlU87MGSjKpLgijOB8LnMGYkY7FSoUdlL0ERd7CnHXIJQRWfSbtZVYMGDXxcRtq0evXqPswDVXgE5kxj+XSxZp04XggruYBAl2mOMyrzCK7iITMI7+iSDEL0EPSxjBD2E/KzDFB5yhibDGlARWKuyuY0dJn/3//+54+peqarPfMmyI+vR1ZFLO3EsUHQHy6QDB48uPhCQBrOlVChyblCgIywTJZPu4QLBln7NWvd4/heDMcyXedpcy4e5ZIMKeMIJTmfwvir7PNwA7Ek2gF0K+fCA9vNXzj2+P6IC8+pFKfqPi2U5PggZKdykypUutTHu71zvPDfBtbxtNNO82pX2jefcyB+rGc9Flk8Jtis0WaV66YEdcUm2L9PjbLKrfbIs6oyUd04sa/Z9YkxOAk+vXLyCasaBYjTuzS3qZv0jT4zzlbqNtameKBay6oO4nl9szZFryeDxUodnvAqSmvQ1Vbi89c1nTOV4DNlnvOEnKNsao85/x3w13talWhqiWofb1XbDIjmxXI62KxuvYq3JcvML8amVqfO+iKEubWscjwMFRERERFJUebwk6oybsBzzTXXeEBAN9ryCj+5qzDdKvlxTIXaSSed5OHKJptsEr0jG1WohBDciCPgRz8/0LlBCPMi/OFHfdb0LFQuEgDxw56KVKrpcgUsKO0yCBwYy49lUVnLsghYsuZDF1S6xjPGYfijApQblhAIURFI5RXPSwpqFya6uRIaUfXJnb+5e3uuike6LRPS0AYE4YwHuDSg2piQme7adPXfdNNNfTsIjAgHOWbYJ4zzyDAAWRjTk6CNcSep3qSKLokKQAImbn4FjkUqEQkBCcjoEk2lHcfTxhtv7MM0hDbnPOF44jWOrThCUcJGtoEwLAR1BHscaxyThFihIpHqPS44UEVKhS7hH+M65otjnQsdXLhgjFNu+kRlIWNrEmayHlRUph2/XOQglOTO+FwAYGxM3k+bcaEiC2EmbcE4wbQTQTI4Txjegm0kOKadkLVfs9Y9ju9KwkE+S0U4Y6pSgZslvk1Z6ErODYM4T9gXadvKMUA1PpXfVIfSPmC/00WfLvEh2AbtQZDN8clQCWlDIoCLUIxXyrFF+3Fch2MQfEeyrcyHY4JxfXlPPudA/FjPeiyyeEywGSPnnFcbzR9+zujRPKqIbG5TR9aPjQlagmR1I5WQ8TE4v7jTJieCTypFp/Wvb9VOIbQsUqlDp7JVQY4fZtMZA9SD0CI+Txtrs2LzrNKvFKFnTJXrxll1D0DntI91teXTKj1LVMsqzvO/gEXP54ahIiIiIiLzqzCbZCWhpDE1RZYW0+o1jB7NVXVM0fiCCxMVtoS8BM7cuZtwbODAgfMFYksTwjPGal0aKm9LwkULLtjkU0UetzTuV6rFCbH/Cwj9RRYuun53MOs3zqoXjagzR1HF5PRWQ+ZWWM5TqVn0+ozOsc/w+vV1bCVCzvjj6OUiUSWmD+Vbf77wc3LjnjZ/ff3c99G1/Y+xnaKqzhTJ5fo6U9UZDzfj65627WlS2iPRbr5uPeacs/FtSvmcv++plrG2mf8907vUsWl148sSEREREZlXmSs/ZcHRZbZ3796pf/FxPWXpw7iwdDGnGpIKO6oHl+bgs5BQRUtVZ/LmPPnQfhWRvDAWpZWiO/bIcZk37anSj+7nZlM7JbrCW3Sn9nn+4mFiKdWuY/N/m1HpWt8qL+g8IzP79rHpDbraclFoWqlDLx+7s6Q7tXNDo7S2Sau8FRERERHJovBzMSJ8oYts2l+umyHJko+hCOhSzPiwdEdm7NKlHWNaFkLVJ93ob7nllnluPpSvpXG/ckMhESkvjDGZRzdr78oeQsOirtnTB0c3FaJqk5sNBR6UDrBpxTclGh6N3TmXh4XW0/4INyaqvYdVaRD/TIbimy7lIZpn/C71RaHl/He2L635QkzvYp9HiBm1zfRwA6iht9nUka2tSqzqljFYFYaKiIiISC7q9i4FbXF1exeRRU/d3mVRoJv1FOsb604e754+FxWbc7uHF3X7LrpxUGur3s9sSrzL+Tzd2EPX9XiXc6YXzWNWt9DFO2W53MAoZZ6Viz8Tk9rdPr6ec8Tnl+i6Pr/EZyNh2d5u/aOJc8xdp9yfS28bfzIHn+1jleeZJiIiIiIyL4WfUtAUfor8dyj8lEUic4xOWeRSxykVEREREZmXur2LiIiI5Iuu2GW5q7qUm+mDB5i1oWu8iIiIiEg2hZ8iIiIieWtqy3WzEm/WIwvbcJvev75VOyXjbvYiIiIiIhF1e5eCpm7vIv8d6vYuIiIiIiIiSar8FBERERERERERkYKUWvkpUigm1awTPZqrxsRx0SMRERERERERESlkqvwUERERERERERGRgqTwU0RERERERERERAqSwk8REREREREREREpSAo/RUREREREREREpCAp/BQREREREREREZGCVC7h57PPPmvNmze3bbfd1i677DKbOXNm9ErZ9enTx3bbbTfbfffdrV27djZ27NjolSJTp061nj17Rs/MevXqZT169IieFRk2bJi1bdvWHz/22GPWrFkz23PPPe2FF17waciajuQysPPOO9uWW25Z/Pfjjz/a9OnTrXbt2v4af1tvvbV16tTJ/v77b/9MrmWkSVsGN+e/6KKLrEmTJrbLLrvYgw8+GL3b7JprrrFdd93V9thjDxs8eHA01ez222/3Ze60007zbcfC9sEHH9jAgQOjZ2Z//fWXdezY0de9RYsW9vbbb0evFEm+n2Ope/fu1rRpU2vdurWNHz8+emXpdsopp9hHH30UPSt/zz33nH3++ef++IADDrAffvjBXnnlFTvnnHP8OOWcKsnee+9tf/75Z/RsfvFllJcXX3zRzjvvvOjZgmPdkt8D6Nu3r2//fxXn25lnnhk9K52yfBaXXHJJXt97IiIiIiIiIuWpzOHnhAkT7PTTT/dw7eyzz/bgKhloLaghQ4Z4iPH000/b8OHDfTnHHXdccZj42muvWefOnW306NH+HPvvv78Hf4SEAfPZd9997aeffrJbbrnFBgwYYPfff78HtcwrazrSlkF4UqVKFQ+vwt+aa67pry277LL2+uuv+x+h63fffWcPP/xwzmWkyVoG7fD111/7v8yLMPPXX3/1APrdd9/1f++9916f/6RJk+ydd96xJ554wp588kl/jc+99NJL0VIWnmnTptlTTz1l559/vn3//ffRVLP77rvPqlev7kHc5Zdfbueee65Pz3o/YQltSFueeuqpHoT+l3Fcz5gxI3qWjeNv4sSJ/vjWW2+11Vdf3R+Xp/gySvLvv/9Gj3Jr1KiRnXXWWdGz8scFiHzar6zy3d6yWlTLEREREREREVlalTn8HDRokIdZVDhSWUgAx7/l4ZdffrGaNWvaCiusYBUqVPDK0gsuuKA4vKAKMISOwUYbbWQ1atSw999/35/zXqrJqGIj9KOCdMUVV7S11lrLGjRoYG+99VbmdKQtgzCO95VklVVW8Qo7AuJcy0iTtYyVVlrJTj75ZKtYsaIHWvyxjYSjVJXSVuutt563FeEUAeihhx5qyy+/vK288sq+b5j3wkaYyXLWXnvtaEoR1rtq1ar+L8cN+xVZ7ye4Puigg/x9VH8SQlOJu7DdcMMNXqHKMqlSJTQ7/PDDvar2oYce8vdQxdqlSxffxwceeKC98cYbPj2fz+KZZ56xli1b+nFBQI2seVLRfNNNN3m4H69+JQy99tprPdynwpOQm88QFjOdykxC459//jn6RDZC5yOOOMLXkwsN7BOkrVNyGZxjVH/zWQJsQjmqdtk+Qm5Cay6K0A6cixyHbM8ZZ5zhx22o9uT7g3mCeZ122mleIUzV9x9//OHT05YFQn8qomkL3pN01113+THG+QMqpZkHy3/ggQd8WhwXKVg++7Fx48bzVE1zIYNpbMs999zj02gTwnwu0LA/CfPbtGnj77n66qv9PUh+lnbmfA3H9YgRI7zKGfksJ6s9svz2229+fO61115eRR6kLSsp7bPs527dunnbs19pV3ABh/VkOpX3aUF51mdFREREREREykuZw08Ck1mzZnkwwg9nwplcXWVLgx/YU6ZM8XCILqz8yCcIIeDDUUcdVRwSxLEOhLKg8rFevXq22mqredffeKC4zjrreEVm1nSkLePbb7+1b775xvbZZx+vVLv++uujV+b1xRdf2PPPP2916tTJuYw0WcsgJNlhhx2sf//+dsghh3jFLeHsxhtvbK+++qpXkxJ0sd2Ex7Th0Ucf7Z/96quvPIRlXy1sBK0nnXSSr2tc+/btPfSlez5hHcEYst5PV/94IEob5mq38kQoRfhKsM/xR1hHKETXaRAWrbrqql5N27t3bw/yfv/9d3+tpM+C/UEASvBGOMgxkmuehIcEqXXr1vXnYBiIoUOHetDGEBEvv/yyh0gEYVRQclzki2ph9guhJmEa64e0ddpss82Kl8HFBgKsfv36+fFeuXJlu/HGG/2zrB/zDF3QCUqphCYEJtDlGOYzb775ph/zcVw0IMjk9Vq1anlbEeanLWvMmDE+vANh4OOPP26ffvppNJe5CAs576iEpaKY9qRKnApq9gvzSAqV0yyLiwm0AVXLtDPfMZyHBM68Btruiiuu8NDz0ksv9f3OelOtzXmZ9lku1Oy4445+/oJqd86NfJZDcJvV9lk4/1g3tp350c65lhWX9tnPPvvMg2nahr9HH33Uj3/2M9vMuvL9RWiblPVZERERERERkfJS5vDzn3/+8bCTaje6NI8aNcoDmvJAqEIwRGhE+HHbbbd5KFBS5V8ITAhl+ZfqM/A8VBoidCHOmp6FIIhqOoICgg26k4dKM9ojjPlJ2EJYSeVleS4DVBQedthhHuIQGhHQUvFJGHr88cfbJptsYsstt5y/l2Wxb4455hivQON9iwsh0+abb+4BC+EXFX+5KtWS48eyLclpCwsVf4RJVBMTynI8UokcwhmCoA8//NBOOOEEH8+Q7QhjYJb0Wey3335+TKy77rq21VZb2ciRI0ucJ0MhxHFecMx17drVz714hWFpUXUZgn6OIyp0kWudwHKZRgjHe9gOKpFBxXE8bN9+++29ennDDTf0ymjOk0qVKvl2TJ48OXpXEY5fLnyA45kQOGtZhJS8l3ky9ATrnwvBJxWOtCfV2Kzje++9F706Fxd0WF/aggsRfI4/QliCbSpDCe9oHzDOb6gUZ9uomCUEJOSnSjvrs8ybkBCE2Xxn5bOcXG2fZbvttvOLQRx7tCvzzbWsuLTP1q9f30NwQmeqdtlPrBPz5Pua/UvozDonZX1WREREREREpLyUOfwk1OHHMJVLVJwRJFBxWB4IO/khTzBCd9xHHnnEl0eVUi5UB/IZQhuqiagWxRprrOHVdQEVhLw3a3oWKu+o7CIAYNsJpULVWHzMT6osqZJaZpllym0ZBKD8S5hAd3BCM6rmCBgIWBijlHFXq1Wr5u8hKOTmOnSlZTrhxeJEsHPsscd6NSeBD8dLfDzVJAIeqj8D2o22XBTomh8QZCbRxldddZVX3fHHsRYCnpI+m8T+4zjJNU+CxCTCO6rwqNRk37dq1cqD9gXBsRawziH8zLVO4HWqk8PrBNtUiyK5zmxnUFK7MCwCbRKXtSzaO77+oTo8l/j7WS+C9VxYX84nQlkuJIR1YExigkfEt5cqWYYP4Phlv1AlmfVZun1zHn/88ce2wQYb+MWPfJaTq+2z8H2SlGtZcWmfZb0ZSoDj7uCDD/bvXiT3CfszKeuzIiIiIiIiIuWlzOEnoRxhJ1U7dMOkeivZdbksqChlzEHQDZxx4wgHSkLYwBh9dDmnGgxUhhEe0jWcmwRR2URwkDU9yx133OF3PeYHO2PgEcZus8020avpymsZVHnSrZXx9KiS4u7oBAZ08yUspWrqyy+/tE8++cQDaSpGmcd1113nQePiRiVkqMqlepDuwLkqUelaTeUrCEq4m34+wdaiQNh///33++Nx48Z5mMt25YuqXgI3xqEk5Gcc2NLOk67SVBdSoXjxxRf7+RfC4tJWyBKMh7FHOVZDd/tc68QyNt10Uw/t6CbPdMadpPvywpC1LCqsCWWpROQcCMdMGtaZClS6l1N9TeU6FyrSLgxw3tGmnG/Mk/fQHlS3U8VLFTrjoMYrYcFwHVSfMuwA5/EWW2zhFyCyPkv4yP6/8sorfdgO5LOcrPbgMRdbSgp0g3yWlYXlc4GJcT1ZHuc0+4A2pvs868JFi7Ru71mfZT34nkTWYxEREREREZF8lDn8JMAjdKGqkHEH+aFfXmNKdujQwUMNuuLShZTxIrnZBj/4S0KXVsYbDF3eQbUh86QbOvNivamgypqehfcQwBFu0K2T0LGkisryWgYhASEmgQEBL/MjFKUrOVWghNHsAwJSquboyktFIO0Y/uI33VnUCGeo9CSsY4xPQvMQTqdhCAO6HdMGVNIRsC8p2EeEXLQ51WuE7flUeQbsX45PxmSlapdqv9LOM1xo4JyjOjpU1TZs2NCrNQkr80XXacI+ji3O5RDoZ61TWAbhLRXOHN9UMBIuHnnkkf7Z8sY5k7YsxvWlUpHQjororAsLHP+M40ulLI/5nuD9zI95JLGNVEQydi5hHp+j6pXn7DvainFLOf/iqHLkeyucpxzjfCbXZwmVCQMZxxP5LCerPQiBCTTZb/nIZ1lZODf5nmE7GXeVNuJ45ruKY5rvMLrSs31JWZ8ljGbYEGQ9FhEREREREclHhdn5lgaJLIUm1awTPZqrxsT8A0H577r//vu96pNhI5ZGjJVMEFqaQF5ERERERESk0Cj8XIzows/NUNK0a9cu55igkh+Fn7Kglubwk5tgMUwE1aAiIiIiIiIi/2UKP6WgKfyUBcXYl4xXmXajHhERERERERFZOpR5zE8RkULEndQVfIqIiIiIiIgs3RR+ioiIiIiIiIiISEFS+CkiIiIiIiIiIiIFSeGniIiIiIiIiIiIFCSFnyIiIiIiIiIiIlKQFH6KiIiIiIiIiIhIQVL4KSIiIiIiIiIiIgWpwuw5osfFpk6dGj0SWbpNq9cwejRX1TEjokciUkiqVasWPRIREREREREpospPERERERERERERKUgKP0VERERERERERKQgKfwUERERERERERGRgqTwU0RERERERERERAqSwk8REREREREREREpSGUOPzt27Gibb775PH+dO3eOXi27W2+91Zo3b2577723HX/88TZu3LjolSLcmb5Xr17RM7Mbb7zRevbsGT0r8uKLL9pRRx3lj5944gnbd999bb/99rNhw4b5NGRNR3IZ2GOPPaxRo0bFfz/++KNNnz7dt5/X+Ntpp52sS5cu9vfff/tnci0jiZvwX3bZZbbXXntZs2bN7OGHH45eKXk+N998sz3yyCPRM7PevXvbnnvuaS1btrQhQ4ZEUxeNjz76yJ555pnomdlff/1lZ5xxhm9Tq1at7N13341eKZJ8/8yZM+2KK66wffbZxw477DCbMGFC9Mp/1ymnnOLH12677WZbbrll8fH2/PPPR+/I9sILL9gXX3zhj/nMguA455zMJb6czz77zF566SV/vCA4v3777bfoWX6uvfZaP47i68p58OSTT/rjONqT74xRo0bZaaedFk0t2V133eXzz4V2YN58h3AMl1W8XReV8tqX8fm0bdvWvzNFREREREREFrYyh5/HHHOM/7jnj6CvQoUKttVWW0Wvlg3BHj+YH330UXv22Wc9aOUvhIlvvPGGde3a1caMGePPQShIwEd4GDAfQsSffvrJ7rjjDnvooYfszjvv9HVmXlnTkbYMAo8qVarYW2+9Vfy35ppr+mvLLrusL4+/QYMG2ffff2+PPfZYzmWkIWCYOHGibzefuf766+3XX38tcT7vvfee3X333cXTCMTef/99GzBggN1+++121VVX2aRJk/y1hWnatGm+/RdffLH98MMP0VTz9V5++eV9H11yySV20UUX+fSs9w8fPtzbkNc6dOjgQeh/HeE2x9ctt9xi9erVKz7eWrRoEb0j25tvvmnffPNN9GzhiS/nyy+/9OMyH//++2/0qGyOO+4423333aNn2Ti2CFY5x+vWrWsXXnhh9EqRXOtDkDpjxozoWW7LLbech6VxfEfl+/lgUe2/uAXdl0nx+dxwww222mqr+WMRERERERGRhanM4ef222/vgSN//Cjedttt7cgjj4xeLZtffvnF1ltvPVthhRU8VN16663tnHPOKQ4M+CG+xhpr+OOgdu3atsoqq9iHH37oz3nvK6+84tVfr776qlfLrbjiih5W1q9f3955553M6UhbBoFJCDtzYT123nlnn0euZaRZaaWV7IQTTrCKFSva6quv7kEB25JrPn/++aeHpG3atPHnGDlypIdAtOG6667rbUhYu7ARZtJOa621VjSlCNtTtWpV/5cQlP2KrPcT6h1wwAH+viZNmtjYsWO9ik7mR9scfvjhXuHLsRMPkcF+J1QneKICj6raq6++2g455BAPCydPnuzve/nll72qmPkQRucKAKneY1lcXOAcI9SOL+eDDz6wvn37eghPkP/VV1/Z0Ucf7e898cQTi6v/qMwmGM/3u4N1/vzzz/0x60mFOKju5KLAvffe6xXfJeHiwfjx431daT+qrWmX1q1be9sQtv/xxx8eju6///7eTlSIMn+O19NPPz2a01wPPvigt8dBBx3kbQnakM+C6vTbbrvNz1MqmdPam2D0uuuu82l8t9Keyf0X1759e/9uJACnbagOZ/5UV/NdyPMePXpE7zbfB2xHXHntS9aFtqMNaYMRI0bMt+5nnnmmf79zzFEJznbyXi7UgID/0ksvtYMPPthf4wKSiIiIiIiIyIIotzE/qZAkZDz33HM92CoPdNWmmzRd3gkpCAnoLk2QB4IeAokkwoLnnnvOHxPIUtFFeMiP83houfbaa9vPP/+cOR1py/juu+/s22+/tQMPPNADuT59+kSvzItQZejQobbRRhvlXEYaQsrtttvOHn/8cQ+EmjZt6p/PNR/CjZNPPtlWXnllfw7CYPYNlaC8j/aggnRhYx0YpoBtiGvXrp2HtezbQw891MMeZL2fStf49vI4V7v9l5133nkeQlElu8suu3gAFUf3cY5XAjuOSdpxm2228cpqLjJwzlAVTOhEyDVw4ECrXLmyV5hmIdDaYYcd/LMMDdGvX795lsNxTMUuoRzdvllHgk6CNc6rs846K5qT+frEh2vIhSElGDKB9aUSO4RmBHS8li8CQ74fkiEmwSrz4fik+zxBPe3B98Hbb7/toR/nHmFeHAEqldf/+9//PMwdPXp09Mq8OA8JSVddddXU9qZ96GL+1FNP2U033WRXXnnlfPsvaeONN/ZgknP+6aeftgceeMCDSL5D+P7g+3PWrFke2v7+++9+4SSuPPcly6HanDD1vvvuy1x3QmsuynDM0g5nn312cdj+ySef+HAftCdhqIiIiIiIiMiCKLfwkx+oVCRusskm0ZSyq1GjhncVJRjYYIMN/DFjxZVU+UdYSujID3D+Dd2BeR4qDRG6nWZNz8J6EQLw456KJH78hwqvf/75p3gMRrroExhQiVXaZQQEMHye7t+ErlnzITygGpTQK45QYp111vHKMsY2JOih8nJxYR9uttlmdv/993sX/JIqC6nCi2N7k9OkqOqXQH7XXXf155wDJXVPrlatmp+zoPs8VXiffvqp74/u3bv78UJ14Mcff+zvScP5SGAdhmHINQYm5wbzJ+Dk/KHbOJXJARWE+dpxxx19+6gq5GIH2873AsHepptuGr1rwdE24VziHCYYZBxRQvpQwZmGdaJNqfpmCAxCxzSNGzf2oTOy2ptKSS4QVKpUyTbccMPUsUqTwrIIP6n2Zv4EjexXKskJRxlTl+8Kvg+SynNfhu9cjisqZ7MQYIexUBkvmeEBGO4DfIcSBnPBo7wuqImIiIiIiMh/T7n8oqRii2oofqyXJ36A84OaH/9UCVJBRKjw+uuvR+9Ix49lPkOFYagWBd3HQ9dMsN68N2t6ljp16ngXTUJIKkoJMsKNmOJjflJBxRiCyyyzTKmXwXpTRUZ1Gd2+GzZs6G2cNR8CDT5DqEGwSMhItRThCWNrEgL379/fQx3mubjQFZnqT7q3E5BQxRsfTzWJIQeo/gzYXtpA5hcPxfMJiwjLOTbjOD4YyoILDvxxLDGUQhbGkeU44zygcjcXlkWlI59h3vwb75pOdWW+CCQJCgnzeMxFFyocqWSNt8OCiq8LYSrVsQSJrHOoVk5Du8eXH6rUk2jn8G9ae3NBI75vqHAt6WIJ53oQfxxwvjGGMl3WCYyTynNfxqvPc+FiRry9WG8u8IDAVkRERERERKSsyiX8pPqKH6zJLsvlgR/jdH0H1V381axZ05/nQjURXTcJLAhMQUhJ13y6gNP1m/UmeMianoXuoHTvZ5u5Wcprr71W4k2eSrsMqjzpEkr1FV1UCXkIdLPmE7rK0t2VcJHqNCrJ6DrLc6rLGKOP53RtXVyoSgtVuYz9R5UX3V6z0FU2DGFA+FurVq3MQOm/jDYh1A4XBgYPHuxBYJpclbNU6tHd+Ouvv/Z9RHDOeJFZGE+Srs1c+KASMB7QxZfDY4JB9j8BHKicDje8Ki2qGqkGJ8jjjveMNXzPPfeUqst7vvgOog0Y5oJu2fGxepNtSRDLRQiqHTnn2A+5ZLU336Wcz0zjHOHiT7Cglc9UhvL9QOV6cmxdLIp9mVx3tjOc3wwRQAXz+uuv78/TMDbwohi2Q0RERERERApHuYSfVD1SsZP2g7osCPAIEw477DDvAnnSSSf5j/N8utbThZZx+7h5R8D6HXvssT5eH/NizDoqvLKmZyFMrF69uldZEkpQBZoVNAWlXQY3+iDQots6Ny7hOQFraedDF3OqRgmDCWzpupus9luUGBeQilaq0Bj/jxvKhHA6Dd23qVCkSy4hFOGQpGNcSG6kw01uuDEW+ztpiy228IsC4WZBSRxL3bp1s06dOvm5Q1gVD96SGFOS+TE0A1W5hHVUYMaXwwUIqgIZO5OqScaiZL/SVTptHfPFzcQ4DzlPOP+oiKY7fHmjPQn5OGYZqoFzDpyPfB/FMawEbcJ3FuOD5rrAgaz2pjKTcJfzn3FcOe7pAl7S/suF84iwle1Js7D3Zdq6M0YxF2VYJ6rk+X4i2M7Ccc1wIiIiIiIiIiL5qjCbfocJupu2FIpp9RpGj+aqOmZE9Ejkv2PKlClevcqNlUK3+0JTqNslIiIiIiIiC07h52L0zTff+A2T0lA5lmtMUMmPwk8R8+pKqlS5ORGV6oVK4aeIiIiIiIgkKfyUgqbwU+S/Q+GniIiIiIiIJJXLmJ8iIiIiIiIiIiIiSxqFnyIiIiIiIiIiIlKQFH6KiIiIiIiIiIhIQVL4KSIiIiIiIiIiIgVJ4aeIiIiIiIiIiIgUJIWfIiIiIiIiIiIiUpAqzJ4jeixScCbVrBM9mqvGxHHRIxERERERERERKWSq/BQREREREREREZGCpPBTRERERERERERECpLCTxERERERERERESlICj9FRERERERERESkICn8FBERERERERERkYKk8FNEREREREREREQKUpnDzxkzZthFF11kW2+9tW2//fZ20003Ra+Ujz59+thuu+1mu+++u7Vr187Gjh0bvVJk6tSp1rNnz+iZWa9evaxHjx7RsyLDhg2ztm3b+uPHHnvMmjVrZnvuuae98MILPg1Z05FcBnbeeWfbcssti/9+/PFHmz59utWuXdtf44826dSpk/3999/+mVzLyOWGG26whx56KHpWJN91wttvv20HHHCAv963b1+ftqh88MEHNnDgwOiZ2V9//WUdO3a0XXbZxVq0aOHrFpd8/8yZM6179+7WtGlTa926tY0fPz56ZcnCcXnEEUfYrrvu6tt1/fXXR6/kp1+/frbvvvvarFmzoikLZsiQIbb33ntb48aNfZ8PGjTIp7/44ot23nnn+eN8lNf6ZPn0009t+PDh/vjwww+3r776yh8vDgcddJB9++230bP8cP5xTJZVeWz7lVdeOc85syDuv/9+u+2226JnIiIiIiIiIoWjzOHn448/bg8//LBdcMEFdvLJJ3v4mAy0FhRBznPPPWdPP/20ByWnn366HXfcccVh4muvvWadO3e20aNH+3Psv//+NnjwYJs9e3Y0pWg+BDk//fST3XLLLTZgwAD/sX/ZZZf5vLKmI20ZhJxVqlSxjz76qPhvzTXX9NeWXXZZe/311/2P0PW7777z9sm1jFzeeecdDywJW4LSrNOff/5pXbt29WBj6NChvvwffvgh+tTCM23aNHvqqafs/PPPt++//z6aanbfffdZ9erV7ZVXXrHLL7/czj33XJ+e9X5CYtqQtjz11FM9CF0SsT8OPPBA365HH33Uvv76a7vzzjujV0v2xBNP2D333GMVK5Z8Sv7777/Ro3n9+uuvds4559iNN95oL7/8st18881+MWLEiBHWqFEjO+uss6J3FsmaD9LWJ9f7S4sQu7y+JxaH5ZZbzh544IHo2eJ10kkn+UUVEREREREREZlfmcNPQoCqVat6lSN/lStXttVWWy16tWx++eUXq1mzpq2wwgpWoUIF23bbbT1kpdoUBCghdAw22mgjq1Gjhr3//vv+nPdS9UY13EsvveQVpCuuuKKttdZa1qBBA3vrrbcypyNtGYRxvK8kq6yyiletTpgwIecyshBcXn311V4dFleadSI0ZB14jX1FkLz66qtHry48hJms09prrx1NKUKYxvHCv4Sg7FdkvZ/gmso83kelHYFvPAheUnCsbrLJJr6e7OMzzzzT6tev7689++yzXg26xx57WLdu3TyojnvwwQf9GCHcpdKSKl+2lerY448/3kNyKmBbtmzpgXFWAPzHH3/4Pl5nnXX8+brrrutV0Jw/7777rl177bU+nSpoAvhDDjnEnyfF14fAM77cyZMne+Uu28IFBeYLqpOZHtabKmeOW96XrFqeNGmSB7O0C8cjeH+bNm1sv/32sw8//NCnffnll3booYf68XvMMcekhvZp7xkzZoyvG+tO9XPz5s39XCIgZJuaNGniF0l4X9I111zj68x7Q7j58ccfW/v27X0f8nnagHkfeeSR/jrz4Rjlc0cffXRxeM/20260H+sTtispbdvfe+89/xwVvPz7+eefe6gdryg+9thjPeQmZOfCBmFyhw4dfF6sCxX5wVVXXeXTqJ5n/ycr6MG0gw8+2NuS4yZcQOKiDevBdyiBeJA2nePgwgsv9G1heY888ohPj2Of0E7sB5bFxSoRERERERGRhaXM4Sc/zDfccEP/oUvlG8EHAWR52GuvvWzKlCkeGvJjnBCTAIIwB0cddZR3hU4i2AjdfamcrFevngeyBCPxgJCQiIrMrOlIWwZdZL/55hvbZ599vKIuq4vzF198Yc8//7zVqVMn5zKyEF6cdtppHqLGlWadmP7zzz/7dLq9X3fddVapUiV/bWFaeeWVvSJthx12iKYUIUQi9KV7ON2yCYeQ9X6CknggShuW1G6Lw9lnn+3h2wknnGD33nuvtzH7gba/5JJLPGykipXn/fv3jz5VhO7yG2ywgd16663+nDCMimoqfMExDMIp2i05rEPAkAsEfQRKrM+TTz7pgSzTk2hDXk8TXx9C6vhyCeDWW289D9UJ1KjGDhWhBNiE1Qz1wHupOiUUSw61wMWJU045xY9JQjIQ8NIuof3QpUsXD/iopuV451xISnsP5zvfR3fccYddeumlXg3Ld8bvv//ugTQXIk488USfHsf+IUAkjKNCmvUm2KQdCDc5l+vWretV1XGsA9sTLjSEYxqE2XwXURV/9913R1PnlbbtHANc6CHc5DxhfWivMFwGYS7rxjkdRzsQLLMfCKYJTfne/Oyzz/yzBJMMLZGGbWcd+SzHHFX3zI91YBtYR44ZqtqzpoOwmMphQu/evXv7tDjaaccdd/T9wLqyn0REREREREQWljKHnwQc48aN87CGkIAftPxQLw+EJFRfEaDUqlXLu24TBJRU+UeVFUEFwQP/EtCC56HSEFQ2URmaNT3Lqquu6hVg/PB/5plnvIKNgAH//POPBxL80UWfalUq00q7DLqAr7TSSh6m5CNrnQikCK8I3wh1Qnf8xYVAafPNN/fAhCESGIcyV3dqKh7jaLfktCUBFXMEQlTWTZw40YM4Aiy6nDds2NDWWGMNDxI5FkOYmYb3EECxn6ii++STT4orRZdffvkSjweqMznmeR/7nwsIzCOJar18xZdLQNaqVSt/TJBIpSld/EHVJ5XfXPwgxOb8pXKbY7AkYX0Ia6lg5TwaNWqUB2gE5ny/JEPHXO8hjOQ8IISOdwkP605ITNUoY9AGbBvrwfARVO+yzVRgckGHfcH3D5+P7wNCSC46UAEJwtz4/iWwxKabburha5rktoNhMdi+22+/3c9pjoGNN97Yv0dYHucw+5b2jttuu+28spvtJqhlmW+++aaPM8yxxTTC4TRsA9tNZTbrzbbTJlTqEmgTLLN+VKdmTQdty3pRnc4yk6iI5fjgO4n/ZiQroUVERERERETKU5nDT8IdgsmtttrKf/RSYZUMKRYUYQPzYv5Uo9GFkkCFkCkXqgP5DBWGjBVKtSgIoOJdZ6l+471Z07MQINC1lDCTilJCn9CFNj7mJ0EwlWfLLLNMqZdB+Mm6E1pw8xlCQwKYLFnrRChK1R7/EmwQ5BBWLy50z6VSj2pOAhbWKT52aRIBSrhxE2g32nJJQthJtV+1atU8GKNrO8fuXXfd5WFtPPQmFCLAykKgRJUhARf7iq7BASFkLhwvjDfK/ieE43jhvGFaUknziou/N9f2xIOuZChXkuRQGZwzTOPCB8c//yaHicj1HgI11ovjhXVOYhtY32SQHt82AkQ+S1jH/iQUpWt/smI2/plk2Jes2k6T3HZwjvD9QWgabtYGglLOIS5kcCEoKW15bGN8HQls0yT3K3+E21Q00778MUwBFapZ00EVdy50l+fYZLupiBURERERERFZmMocflLFR8hGOHfFFVf4D+stttgierXs+JEcqrMIhAia6JJbEiq0uBs6Xc5DIED3earh6GbKzWGoVKJrctb0LHTTZExHwpXffvvNw9htttkmejVdaZfBOH7Ml66qVLUxlh9djLNkrRPh2auvvupjUlIFRqBCJeLiQlVgqMqlSy5Vg3SjzkL3ZSr4QAUbXbjDsAdLCoJlAjHaGQRvVM0RwNPWBPi0P0EaIVGya38cxziBL0MAMJwEFxeSFcK0HUF6MtTjOKdbeqjEpBqPqk/WIwvrRYVhvlj3sD/oSs34l/mcj2lyVfASIlLpSNdrEOCGm2MFud7Duc94qQwvQQVuENadykkCeML3YPvtt/d50d58j3GuUEnJMBOcswSSBKHxEJZjkWWEfc/+5TNlwf7lO5Wbf7FOfFeE6mjCTy6McJzk+z3LPAhMOV6YL9+haejGznZzPNBOdE3faaedvP2o3qXinqFNOG+zpueD8ZgJPalcpXI3rQKe+dLmuR6LiIiIiIiI5KPM4SfdS7lxCuHnwIEDPaSjC3B5YF50G6cLMd3ICQGpvKIaqiSEBHT3Dl3eQdjBPOmGzrwuvvhir2rLmp6F9xB6UJVJ9RUVlyUFHqVdRmllrRMBGt3v2UeMhcprhBqLC4EOlZ50sSXg4yY8uarjGMKA7v+sN+MHErAvaaj45I7+dFGm8pM/AiuqfqlSZV9zcxwqkAlK45V8SdwIizCY8J7uxITmBHlxhJqET4yHG0f1Ncti+AOqfams5aZH4cY8aeimX1IldRzrRHdxtqVr164etlIRWVpclCCQ4zsjC+OGUs1JGxC0MV5lUtp7CMkJ4qig5XhjXEkCM8JSbqrFdwPVh4x/G0flOt83vM5nOV/pIs65xRiqbPMbb7zhx20c8+nTp4+/ToCYtp6lwXryvcr3KMcKbUWlJ99nVHiz3wkO88V6EYCzfVSwcsEqHvoGDGNw2GGH+XHD8cOFE25ix7L4HmU+fI/w+azp+aBClGOaY5zKXMLe/v3nHQeX4JnvrVyPRURERERERPJRYXayfEykgEyqWSd6NFeNiYuv2395IfAj1Cxt9/IkKlSZx5ZbbhlNKVxUrTJW5X8NNzii4pNgk2EVCCqpvi7rsSMiIiIiIiKyNFD4uRjR/ZSKwTRU5OUaE1TyU4jhJ92SqW6Mjwe6oKi+ZHzY5FiVhei/Gn4ybAg3wqKrPENiUMXKPhcRERERERH5L1D4KQWtUCs/pfSWxJtliYiIiIiIiMjCVfjlXiIicyj4FBEREREREfnvUfgpIiIiIiIiIiIiBUnhp4iIiIiIiIiIiBQkhZ8iIiIiIiIiIiJSkBR+ioiIiIiIiIiISEFS+CkiIiIiIiIiIiIFqcLsOaLHxaZOnRo9Elm6TavXMHo0V9UxI6JHIlJIqlWrFj0SERERERERKaLKTxERERERERERESlICj9FRERERERERESkICn8FBERERERERERkYKk8FNEREREREREREQKksJPERERERERERERKUgKP0VERERERERERKQglTn8nDFjhl1++eW288472/77728vvfRS9Er5uPXWW6158+a299572/HHH2/jxo2LXikydepU69WrV/TM7MYbb7SePXtGz4q8+OKLdtRRR/njJ554wvbdd1/bb7/9bNiwYT4NWdORXAb22GMPa9SoUfHfjz/+aNOnT7fNN9/cX+Nvp512si5dutjff//tn8m1jFxuvvlme+SRR6JnZjfddJPtueee/vfQQw9FU83effdda9u2rS/7rrvuiqYu+HLLw0cffWTPPPNM9Mzsr7/+sjPOOMOaNWtmrVq18nWOS75/5syZdsUVV9g+++xjhx12mE2YMCF6Zelz8skn2y677OLnTPD+++/bgQceaJ988ol99tlnZTp/2Occg7TpueeeG02d6+WXX7aLLrooelZ6nAfsh/J0+OGH23fffRc9W3g4LzhHFxTtyvfQgnrhhRfm+17q2rWr9evXL3pmfpwfccQR0TOzd955J2d79+7d25588sl51u3aa6+d5/zJ15QpU+z++++Pni1ci3JZIiIiIiIiImUOPwnl+vfvb6effro1aNDAOnfubN9//330atkQ1BEaPProo/bss89ax44d/S+EiW+88YYHCGPGjPHnIOQbMmSIzZ49O5pSNJ+99trLfvrpJ7vjjjs8MLzzzjs9jGBeWdORtgzChipVqthbb71V/Lfmmmv6a8suu6wvj79BgwZ5Wzz22GM5l5HLe++9Z3fffXfxewkHn3/+eXvqqad8vrfddpsv488//7QLL7zQw1/Cj4EDB3rYs6DLLatp06b59l988cX2ww8/RFPN12P55Zf3fXTJJZcUh3FZ7x8+fLhvH6916NDBA6Kl0W+//WYffvihrbHGGvb2229HU4uOzRNPPNFD8y+//NL3dz7+/fff6NFcBGHxYDVp++239/N0QS233HLzhOpLmrQ2CW644QZbbbXVomcLV671iNtuu+38fA74Hpk4caJNnjzZn48YMcK23XZbf5yv4447znbffffoWf44/xYkNM3C92/WsVjeyxIRERERERHJpczhJ4HOJptsYm3atPGKPoJBKi3Lwy+//GLrrbeerbDCClahQgXbeuut7Zxzzin+UU1YRJgUV7t2bVtllVV8vcB7X3nlFa+MevXVV2233XazFVdc0cPK+vXre3VV1nSkLYMwLoSdubAeVMQyj1zLyEKgef3113vbBpUqVbLKlSt7+Fq1alVbZpllfDoVgyyLeRNSEYwS9izIcssDAQfttNZaa0VTilSsWNHXm38JQdmvyHo/4eABBxzg72vSpImNHTvWKxCXNgTWu+66q1fyDR482KdxjBIC33777fbmm29a3759/X0E/VTInnfeeX7cHnrooR6MgQpmAuQjjzzSnwf33nuvt18INwlbOR+pxr7ssst8GlWmhIAgTG7Xrp23LcdY0qhRozxspjr3zDPP9ECOUI9wDewHjkvmf+mll/pFD7Rv396uvvpqa926tR100EEe4OGDDz7waS1atPB/v/jiC5+eJm3ZoHKV6uWWLVt6SM76UBnM/Fgmwfixxx5b3Fa8zrHP55kP3ye///67nXbaab4fqLL86quv/L1p887CtlOxyntPOOGE4rA+uW8efPBBv+hCOzD/JMLP8D3FhQrOadY3nJ+0He8Bxwhtx/564IEHfFoajgO+f6n+7dSpk+9j1jMcA7juuut8GpX0rDPV9FwU4XuKalLai2OH97DuHDegAp19ffDBB/trfMckcfGFCzIcG1Rpp7VrfFmc3/QcCNjvyep+ERERERERkbIoc/i5+uqr+49cwgx+yOLnn3/2f8uKbt2EQHR55wczP6TpLk0YCgIIwpckqj+fe+45f0wlXd26dT0IJGCIh5Zrr722r2vWdKQtg2663377rXdXJpDr06dP9Mq8xo8fb0OHDrWNNtoo5zKy9OjRw7tKr7zyytEUsy222MKrBBs3bux/rAPzYp0Id3hOt3dCCILSBVlueWCdCVdCeBMQxhDusG8J9QhEkPV+Klfj68/jRbH+5Y3AkyCsadOmfkxwkWCrrbbydqCaeccdd/Tgh4CLYI5q3Ro1angQSrBH6E9wB7Y/PgwCjj76aN+3Idwk/OrWrZtXgxIGEjbFXXnllR6CPf74415tmGxTqoQJ2qgw3njjjW3kyJHRK0WoMqZrPRXGhPRxs2bNsgEDBngweN999/k0hl5gGwh3CcOYb5a0ZU+aNMmDNwJilskFgFtuucXf//nnn/sQExxLhMXh4guBMhdMCP4DztVNN93U25Vgj2E1cs07DaE01boEyAxjwP4Jwr4hIKVi+3//+58HoqNHj47eMdcGG2zgy/r66699XTkGmN/rr7/ur7PdVH6+9tpr/se+orv4008/7e8vCfNhuA7akQCT72i+Q1kX5kFFe6g85fGGG27ooSdtsu666/r20S5nn312cRjM8AwPP/ywbxthaBq+cwl+V1111dR2jS9LREREREREZGErc/hJ5RDhAlVa/MilQo+KvvJA+EM3W+ZLUMBjxu4rqfKPsJSAiRCGfwmUwPNQaYjQNTNrehbWi6CKgIfqJ6r3QmXXP//8UzzmJ6EW4cshhxxS6mUQPNCuhCFxdJmmCz7hAwEZIS9BC5WThEAERwRNhCOEQKVd7sLGPtxss808xLnqqqtKrLKjsi+O9U9OW9JRGcg+Y19SmUwQT5iVCwExlX+nnHKKV8bRRoTpyGfsSY47lsO+J/wP1ZMBQdv555/vQ0qwD7iIEcfrBFUcT1TtxY9DLkgQtrMMEOrGhfOtXr169scff/hjhjfg3OC45Zgl/M2StuxPP/3U26B79+7eJlSHfvzxx/7+atWqFa8fF0fCuciQGQStcZw/YX25UMDQC7nmnUTQy4UPqnjBd018qIKwb5hGFSfV3wyFQeidJlR/ElCH8YN5TCDKdhFoMy8qeRk/+KyzzvJ9GSpqc2H/sF+5CFKnTh3fFxxXrAuV10zjL4ngPIw1yoUWKskJyMH3GiEmFyGYRxouylDFWpp2FREREREREVlYyhx+Ut3DmJ+EKHTHJJyiaqg8EH5Q/USVEFWCVJERJoTKqCz8MOcz/NAP1aIgCKASMqBKi/dmTc9CYEDVGMESARM/9kNXzfiYn4Q8VMjRNb20yyD8ZN0JbwgKCQ2ptKILP8EEgRZVoAQsjEvKfqD6jX8JTXlMpVdpl7uwEchS/Un3dkIyqnjj46kmMeQA1Z8B6882LU2o+iQoorsw+5MALVQmZyH4ClVz/FGpuOWWW/pr+Vxc4LjMhWpFwnnali7KycpQLjJQRUqIRYUelYIB4XM8UOcGNnHxSuWACmaGYCAQ5dzJJW3ZtAdjlob24JwI3fXj7cF2c2xRoUhX+xBSBqw74R34ruJ4yjXvNPFtTwaAYV2YHn9fqFZPCuEn67rNNtt426200kq+zWG8T4aJoIt+WD++axleoCR8VyaVtO9Au8TfQ3jKRRSwbiWhPcO/pWlXpK2PiIiIiIiISFmUOfykeyNdvwkbCCz4wZusViwLAj8qzUDFFX81a9b057kQENLlky7nIQQgpCQ85IY/v/76q1dP8eM8a3oW7tBMl18CASqyqOKjC3MupV0GlW9UlBKCEBYy1iLVU1QOEv5S/UolFxWeTCPkIQQN4xqyLMLR0i53YWNdQ1Uu4SwVZbnCco6tEBRStVerVq3MIGlJRZDNEAbsS/4IsgmB0248Fapad9hhB+8yDSp6GdcyBFC55FMVS8BECMu+YPxLqvuSFXkMQUB3cCq7CUfDOJQgXCdkD12mQzfzLKw31cl0nybQ4xjMVe2btmxCU75jqIhkflRsUuGchupL7nrOcU6AGkfASNd7sF/o/l+aeXPsUY0ZLsAQbDPPJKouuXjBOcq2hnFekwg/Ob75TiPkBGP3EhSGISCoBmX4Aqq7Oe8PO+ywnGOm5kL7s78IONkn33zzTfTK3GMnrBPoIk9Yv/766/vz0sjVrmFZHEuhoplq4nAcsq18X+V6LCIiIiIiIpKPMoef/BCnGylhAz+UGVOvvMIpAj9CBJZBd8uTTjrJxxHkBkslIQAhNIp3yaUijBuiMD4i82LsPiq1sqZnIYysXr26B0hUpFLJlhaAxJV2GVnoqkugy5iIVBIyZiTVnwwLQFjEulE5x7oRoJXXcssL3XYJXaj65OY83KQmrUItoFsxAQnbRBBOgLI0YTxcqivZRwHjWBI8cpOqOPYrwRR3wmY/EvoT4lP9yA1rQsViFgJ4zo+ScOwyTAVj2TIkA9WGHEdxLP+CCy7w9xE6E0jG0RWfdWK/cIzlOuepgmSYCKoXOUbZfoL9rBvbpC2bY5YxTLmJD+c04RnnXhq2ha7ijP2bxI2ZGP+S7vRUqnM8lmbeIDDlpj6sH9WsXAhJojKb6ky+uxg3OOuCA+ctyyfgDAg/CRxD+EnFL9vE9wz7jO8Aho5YEFTBUxXPuhPCM/4p5xcVnQSijK1MlS43guI9VK7z3Z4MkfOR1a7xZRHGcv6zrzhGGjRo4J+lXalMzvVYREREREREJB8V5vwInR09LrY03k1bJM20eg2jR3NVHVPyeImSG1XVBOpUgHLBgwCV0FKWbFTrcvGBilqqa9ln4WZEhSB0uRcREREREREJFH4uRnQ55YZJaagYW5xjcxYKhZ8LB0MX9O/f3wM0hiLgpjYKnpZ8VBNTbU0Xc4bsoKqYoSUKhY5BERERERERSVL4KQVN4afIf4fCTxEREREREUkq85ifIiIiIiIiIiIiIksihZ8iIiIiIiIiIiJSkBR+ioiIiIiIiIiISEFS+CkiIiIiIiIiIiIFSeGniIiIiIiIiIiIFCSFnyIiIiIiIiIiIlKQKsyeI3osUpAm1awTPTKrMXFc9EhERERERERERAqdwk8REREREREREREpSAo/peA12fv96JHZS4O3iR6JiIiIiIiIiEih05ifUtDiwSeSz0VEREREREREpHAp/BQREREREREREZGCpPBTREREREREREREClKpw8/vv//ezjnnHJs2bVo0xey2226zRo0aWdOmTe2pp56KppZOnz59bLfddrPdd9/d2rVrZ2PHjo1eMbv99tttzz33tJ122sl69uzp03r16mU9evTwx8GwYcOsbdu2/vixxx6zZs2a+edeeOEFn4bSTsfUqVOLl4vp06db7dq1beedd/a/rbfe2jp16mR///23v55rXlmSywjSpmfNP62dkDV9Ufjggw9s4MCB0TOzv/76yzp27Gi77LKLtWjRwt5++22f/s8//9jZZ5/tx1Djxo3tmWee8ekzZ8607t27+/TWrVvb+PHjfXqhOuWUU+yjjz6Kni08L774op133nnRs/ndcMMN1r9//+hZ6e299972559/Rs+KfPrppzZ8+PDoWfmYMmWK9evXL3q24LLWLasdrrzyynmO63zxfVEazz33nH3++efRs7lKO5+4vn37+ncYyjKfknBun3nmmdGzubL2WXntSxEREREREZG4vMPPSZMm2f33328nnHCCh2///vuvT3/jjTfs6quvtg4dOnjAxo/dr776yl/L15AhQ/xH/tNPP+0BxOmnn27HHXech4nvvPOOPfHEE/bkk0/as88+66+/9NJLtv/++9vgwYMtfr8m5rPvvvvaTz/9ZLfccosNGDDA1/myyy7zeZV2Ol577TXr3LmzjR492p8Hyy67rL3++uv+R+j63Xff2cMPP5xzXlmylpE2PWv+We2UNX1hIxwnCD///PM9MA/uu+8+q169ur3yyit2+eWX27nnnuvTCWQqVqzobUlYy+f++OMPD3dpW6afeuqpHoTK/DgPZsyYET0rGRcrzjrrrOjZokFwHcLu0gjfNWk4zpIhZGnbAqVdt5NOOskvQCxsfL9MnDgxelY++P4ubfuUp7R9hqzpIiIiIiIiImWRd/j5zTffeNhI5V4codRqq61m7du393Bq1qxZ9uqrr0av5ueXX36xmjVr2gorrGAVKlSwbbfd1i644AL/gU7oeuihh9ryyy9vK6+8slcvEoZttNFGVqNGDXv//aIb2PBeqtmoOCPco4J0xRVXtLXWWssaNGhgb731Vqmng1BkzTXX9MdZVlllFa9anTBhQs55ZclaRtr0rPlntVPW9IWNIIPlrL322tGUIgScVatW9X8JQdnfIOA65JBD/Hm9evVsvfXWs19//dUD7YMOOsinU/1JEEw17OJCQHb44Yf7cUZb3nTTTXbGGWd4EBavpCSgpoKV991zzz0+jSpCql7ZDipfCaGY1x577GEPPfSQvwdUvbZs2dL3M8E1OO+6dOnix9mBBx7oFx1ABTTrwMUAjpeAYPmBBx7wx5xLRx55pD/m3ORcfffdd+3aa6/1aYTUbdq08XXlQkYwatQor8LmosYjjzzi07LWg4D7iCOO8G3h4gX7P47j8Oabb/YAnoscY8aM8f3K+48++uh5AvKAKm4Cfo4LpLUp4T/bfc0118zXFu+99563I5/h31BByfZSddykSRO78cYb51u3pLR2uPPOO23o0KEetNKOXHQ54IAD/CIDCO2ZlsT7qb5u3ry57bfffj5v8N7ffvvNH1P5SwUwbcv3K/NPVn9SEc0+btWqle/byZMn+3S+A5k37coFhGRwfNddd/l5efLJJ/vzrPnQFrQR8+nWrZtXilLJS9sHVPyzf3DVVVf5e2kn9luo3GebOOb32msvu+iii3xafJ/FZU0XERERERERKYu8w09CNgIaAo64n3/+2UNIEF4us8wyHlqVBj+M6fJI2ENXdn7A88Ob+fEa4QioKCX8I3gBIcegQYP8MRWOhGYEsT/88IMHg8E666zjFZOlnY6jjjrKu1vn8sUXX9jzzz9vderUyTmvLFnLSJueNf+sdsrVfgsTQSvVcTvssEM0pQjBG2Htrrvu6mERwQsI1Ai9wf4nlGFogR9//HGeAJVtL6k9FzaCKKp8OR8I3Ajn2P9vvvmmffvtt17V+vLLL/uxSZdpAjEq+EAoSKDLMAkc64R1BHlUvgbsJwJQwksqXdnnBH+rrrqqV+727t3bA9fff//d308gS5Bat25dfw7C1RC6E7B9/fXXHoS9++673vZxl156qa8Hy+R9nNMYOXKkd0O+9957PSRE1noQXDFfwjrCt2T1N98RBHr77LOPh37sb57zfo7HcBwksa9pv6w2vfDCC/04YcgExNuCfwl++RzHGtXShKyEj+w/qs0JQgn84uuWlNYOASEfISjzYegOlgXa4u677/bHcYSItAXHAOdHrupbwnUCRd6z8cYbR1OL0C7bb7+9L3f99df3diHEJahkXTkeK1euPN/6UlHPd8att97qz9Pmw/6/5JJLPNgkxOV5Wtf/gPP1s88+8/eyPxjqIuAY5Pjiwhn7iwtEyX0WZE0XERERERERKYsy3/CISs84KvSS00pCGEDQQwhUq1YtrygirAgVflRL0V36mGOO8co0qgJBtRQ/8lke/1LdBZ6HikKEbrClnZ4LY1SGMT8JFAjuqLBckHmVRq758zitnbKmLw6EfJtvvrmHKQSHVEuG6jSqCqkOo0KS9QUhaBzbkpy2qBEWrbTSSrbhhht61S/HQKVKlfzYJUgjgCOIIuA87bTTvPv+hx9+6J+l6pNQisplgmGOfaqe45WSBHDs43XXXde22morD98IkZgHw04QTNFmoRqQeVapUsUfB6zjiBEjfH2WW245a9iwoY9rSTVkMvxk/QkjCQQJIVdffXWfTojJuhKWBVnrQagagnoqJKnuzUIFIZXkVGSC7eXiRZpQaZirTePibUEgy3nKMAqEegSPbD/nKvuNdaTCkfbPJa0dAvY5y+jatauHzKFylnmvscYa/jgpBKzMlzA2VFuWRrVq1bxyFZtuuqm3B8tnfxCAsn84bkoaPzZtPhw3HC+sPxXafK9m7R8Q+rPPeS+hMxehgu22284vSHE8b7LJJj5/ERERERERkUWpzOEnVWCh0pNQg4CBH7ulQdjJj3SCBCpL6VpKKES1F0EXVVn8IGc8OH5MB1QB8hkCGSrRqBYFP9qplguobuK9pZ2eS3zMT6opqW6i6nVB5lUaWfPPaqdc7bc4UCV37LHHejUnlXZ036crO8cQ3aCpPnv88ceLu/vzL9WfAdtLGyxOBJ0BoVgSYSNBMxV4/NGVmjAKBERB2meTWBbHFSEVXYvDPDneudEWGNIgiXXYYIMNvC15H2Eo3ahpSyqU46jgpKs6r9H9meo8pJ3HWesRD+TZrlzhJ+Lvj7dJUti2XG0aF28LjjPOTUK9cCM0zod4u/OdVdIwCrm+zwhaqWokyGSMXtqPCxS5hG3nX7Y9fjwheaOoNITvHBdx7BvGcg1txAUGKnVzSZsPFxiS+zNtm8J60qbx98fXv7T/LRAREREREREpb2UOP+mqTnBFt1m6k/IjeEG6VVPtF8YTpSqMm3wQ3lCZxQ/v6667zoOyJMIGlkugQzUXWCe6YnIjINaNCjFCgdJOXxDlOa80WfPPaqeS2m9Ro+IxVOtSMUg3aypRr7/+eg+ojj/++HnCILr9hju/U2FGt1iGQ1iS0V2ZLtdUcxKsMTZmcszGXKhSJIBibEYuCjDkBPMM4yuOGzfOg+O0QCqOru+MTUmlI+En1dVhaIGA4Sao2qNbNTcr22KLLTwoz5K1HoTqYXxSjs/QJT+JoIz9RxVlGBuYILOkUD5XmzLPJNaJcUUZ95Jt5zyhKpKgloslVFtSscnFli+//NI/kzafkrANVM1SxXrxxRf7fAmRWU/OzzTsX9BNnOphAlvOTdoTnLNx+a4XIe/HH3/sQw6w/Wz7o48+Gr06r1zzpOqT446xmDkO2T9UKRP0sk18T3ORiwsZoH3D+Ke0eT43aMpafpgeb7+sxyIiIiIiIiL5KHP4SZdJxqQjWGG8O4JIQsvS4E7xhDJ0m6ULbvv27f0mGfyYp5sqFVW8Hv7iN4ehWyzj7oUu76CqkHnSDZ15EUoQMJR2+oIoz3mlyZp/VjuV1H6LGoEMlZ6ERYx5yM1cCK1ZTwLw+HoSpDC0AdVpDINAheIVV1wRzWnJRcDGWKsck1QjMyYoXf3zRTjIZxmrlW7oVFezrwkq6dbNzWo4z0qqHCX8pEqY9aE7MoF58sIEN53ivGM9uZDAvmDds2StB12tqdRmPowfuc0220SfmIsLFIRkVCATxjNGJu/ne4PxHnPJalOODUI3xk+No6KSiueDDz7YQ3WWTYUm4Vnnzp19OjepYp4EvvF1K40wpi3tSpAaqpqpOA1j7cZR0c5N2lgu3fFDN3nOaYYRIOyPV04SRFJpG4LRXPgeoAKdefG9TJAYbnQVx7nFeMJZqKzme4XP0tYcf7Qh1e4M8cFYywTmBLfgPVTgU/1KFT/7JdeFlqx9Fp9O+7EsZD0WERERERERyUeFOT82Z0ePRQpOk73fjx7N9dLg+YM5kfLGndX/C0EdNzjiQsVhhx3m47KGG4CVFM6LiIiIiIiILAoKPxcBuoFyM5k07dq1K9cxQWVeCj9lcWA4B7rnc5OfQkc3+O7du/twJb/99pvfrZ3KYBEREREREZElgcJPKWgKP0VERERERERE/rvKPOaniIiIiIiIiIiIyJJI4aeIiIiIiIiIiIgUJIWfIiIiIiIiIiIiUpAUfoqIiIiIiIiIiEhBUvgpIiIiIiIiIiIiBUnhp4iIiIiIiIiIiBQkhZ8iIiIiIiIiIiJSkCrMniN6XGzq1KnRI5Gl2z4HfRY9muvZxzeNHolIIalWrVr0SERERERERKSIKj9FRERERERERESkICn8FBERERERERERkYKk8FNEREREREREREQKksJPERERERERERERKUgKP0VERERERERERKQglTr8/OGHH+yCCy6wadOmRVOKZE0vq1tvvdWaN29ue++9tx1//PE2bty46BWzu+++2/bbbz9r2rSp9erVy6fdeOON1rNnT38cvPjii3bUUUf54yeeeML23Xdf/9ywYcN8GkoznRvkX3bZZbbXXntZs2bN7OGHH/bp06dPt80339z22GMP/9tpp52sS5cu9vfff/vrWctIk7UM3HTTTbbnnnv630MPPRRNTZ9/3759rVGjRsV/22yzjfXo0cNfWxQ++ugje+aZZ6JnZn/99ZedccYZvk2tWrWyd99916f/888/dv7559s+++xjLVq0sMGDB/v0mTNn2hVXXOHTDzvsMJswYYJPX9Q41ti3Y8aMiabk7/LLL59vf0+ZMsXuv/9+f3zzzTfb448/7o/jrr322nnaLpfDDz/cvvvuu+hZydLWqaw45hfE1KlTff8Wot9++80OOuig6NlcL7zwgn3xxRfRs5ItaNuWBt+tHTp0iJ4tHm3btrUff/wxeiYiIiIiIiJSdnmHn5MmTfKg7ZRTTrEnn3zS/v3335zTywPhDCHBo48+as8++6x17NjR/wgT33vvPXvqqac8FBwwYIC99NJL9sorr3j4N2TIEA8PA+ZDiPjTTz/ZHXfc4et75513ekjKvEo7nWVNnDjR14nXrr/+evv11199Wcsuu6wvj79BgwbZ999/b4899ljmvLJkLYMw8fnnn/dtZ7633XabLyNr/oQZb731VvHf9ttvb4ceemi0lIWHEJztv/jiiz0YD1i/5Zdf3vfRJZdcYhdddJFPv+uuu6xixYr+mT59+vhrf/zxhw0fPty3j+lsC0Ho4sB+2HrrrYtD2bKifUoKNo877jjbfffdo2eFa7nllvP9/1/y5ptv2jfffBM9k+CGG26w1VZbLXomIiIiIiIiUnZ5h5/ffvutB5FUacVlTS8Pv/zyi6233nq2wgorWIUKFTx8Ouecc2zGjBleUXXIIYd4kLbyyit7VSMhWe3atW2VVVaxDz/80OfBewlFqR599dVXbbfddrMVV1zR1lxzTatfv7698847pZ6+0kor2QknnOBh3eqrr+4/1llOEuux884725dffpk5ryxZy6hUqZJVrlzZqlSpYlWrVrVlllnG35/P/AnbmL7xxhtHUxYewj32x1prrRVNKcL2sN78y75jv4LQvHXr1v68bt26ts4663iwToh8wAEH+PQmTZrY2LFjF8qxlsv48eM9eO7cufM84SdVq506dbJ27dpZy5YtvVIXVABfeOGFHrhTcZwWchFOc1z07t3bn3/66ade2UzVLqE27r33Xq9aJsjnfVQQUhn39NNP++tpqJTt3r27L5uK6DD/rHWi2pYgGlTlUmHIe5NYjzPPPNMfE0RzLvI+9sWOO+7o+49lX3311X5eEtxOnjzZ309YT5UvbcS6JefPZ3k/mDftyT4n8E8iEO/atavtv//+/plRo0bZn3/+6e8PCP8J2alkZFtpM6qJ2VesU9Z0pK1rVvt/9dVXdvTRR/t3y4knnlhcsciFGdaP+dx+++0+LY6LEFzcIOij+rOk9kFW2z7yyCNeNUsldbgwdOyxx/oyQNvyvcD7X375ZT++WA4XJdIuVFGRfNppp/k8mR/nINgO2op2fuCBB3wa0qZnnRdxafsRHGN874uIiIiIiIiUl7zDzy222ML69evnP/7jsqaXB7p1E8jQ5Z2wiB/v/MgnDOW1I444wt/39ddfe/i3yy67+HOqP5977jl/TBBBmEZ4SDhBMBisvfba9vPPP5d6OsHPdttt512VjzzySA+Z4u8LCM2GDh1qG220Uea8smQtg/am+3Xjxo3978ADD/R5lTR/AhXCtGOOOSaasnARSBPmsQ1xBCKEsuw/KlAJYXD66af7NoP9TNhTq1Ytr2iNbxePc7XbwkDVJwHPVltt5d3zP/744+gVs9dff92HXKAS9/333/cwi/CSAIcgj+P27bffjt49F8HPhhtu6EMAgPCHKl7CJIZ6iCOsI3ykAvq6666z1157LXplfqNHj/agi3Xmj6poguisdWK7QvhJlS0BM8F6EvuRqmPwb40aNeyTTz6xkSNH+lAKhPDsFx6znly04BxkGl3sCSQJDVmHtC7+wZVXXunbyHuofE7uawJ8QvOBAwd6V/+0to1jnxDwEjCuscYaPgxE1vSsdc1q//POO89DVNqPEO+ss87y8+zss8+2Sy+91I+JtCCTCzW0M8c8FyvyaZ+0tp01a5ZX2z/44IP+PQO+7whjWV9QYcp5xYUT1ontpO24gHLLLbf4e+LYtwS5HDt8h/Aetpc/1ouhGlhP5ps1HWnnRVxp96OIiIiIiIjIglqib3hEwEJ3WH6wb7DBBv6YkDVU/lGRRYUX3aEZx3Ldddf16YSlhAGEA/xLwAOeh0pD8HlCgdJODxjTk0osQqMw3iLhGNVz/FE5RfDAe0qaV5bkMggJGHeSMSjp5k4IQjVkSfMnhNh11109cFic2IebbbaZhyVXXXXVPBVoBN1UiRG4sG0gBI1ju5LTFjaqPTmGaF+6oRMMBexfKnOpyK1Tp45XtFH5RoUd0wiQttxyy+jd2TheCKR4fxIhKSFat27dvEI015itBOOEaoRiVBayPrRv1jpRtUmVNMctgRQhXprq1at76EaIRVBLIP/BBx94sEV1M6pVq+ZVhqhXr56HsATFBPYEjFT7UnnKZ7KwPoSShHwcG7RtHO1N1STjoRKwh4rRLKxzw4YN/THbRjiItOlZ65rW/rQXj6m85PuH9iYIphqUcyzMm++iXPJtn7S25f0E5VR5Mm7sZ5995uvJBSIuIICqfPY768pxQGUpQ5SwD+MhfsC2Uh0O2oV9TNtQac/4xQS8LHvEiBGZ05F2XsSVdj+KiIiIiIiILKglOvykGopAgR/kVAned9993pWcqiICMLoh8wOeAIKqqIDqQD5DhWGoFgU/xqmQDKim4r2lnc48CRwJkejuSdARKpfiY34STNLVmKq4rHllyVoGXfjpkko1K6EJgcgbb7xR4vxZlyVh/Egq0qj+pDs8gSJVvIS5dCunAqxmzZoeaId1JxSi+jNgu9jWRYWQiO77jE1KiERlW6i6A8djEqFUPIjOJ3BeddVVo0fzoxKTm1kRkLKvQ5iehmOE8JPXOW64aICsdWLeVCJSgcd2NmjQwKenIZikwpqQlsCTkI7gNISfVDGGYRgCwur4cvls1rqDrt1cNGCft2nTxocGiNt00009GKWamirZUDkcR9ftIL5s2iBcEEibnrWuae1PqMc+Yx24OMO/HNul3ff5tk9a29Ilnep3hh7hIgkXN0CVO+cXlbmEl0wnPGW8X9aVPy4+pA0rEMe2sEyGqWA54bO0f/v27TOnI+28iMtnP4qIiIiIiIiUhyU6/AQVTVQEgh/5/BGQ0V2VkIBusgQDSQSEdLvkx3X4IU43ccJDxsUjbKNKiUCgtNOpwKTiiiqr33//3buKErbmkjWvLFnLYExTwl+qX6mmIoxjWq75U5FFF/xQ0bU4sa6hKpcqQro2U7HLHewPPvhg75ZPsBTQPZiwEQR7dIUnMF1U6CYexnnkj+cEVoRKWQjieS/bSIgXgvGkfCtY2d90sSbEoiqSMS7jgXAcgRdhP23JejK2JxV/udaJEJru3IzPCI4XlpFEwEZoRkDKuLEEkxxr66+/fvSO+RHQcwGD7tysD1Wz2267bfTqvAgtCZg5Rhh3kirWZHUi3wcEkQz3QPdyLnAQThIE8j3B+UKVdMDxFebB9odlp03PWte09uc9rCeVlWB4AQJy2oLtiM87C/u/NO2TxHchFwcYXoLlssxwTNH1napKvgNoH6pFOTYYIoRjgBuK0Y5J7FOqREEgvsMOO3g4TmUrwyfwvXPYYYf5uZs1PR9p+zGOeXJs5XosIiIiIiIiko8lOvykKyTdI/lRTdXVSSed5DcB2mSTTTx8IvhjnM/wRwVowI//zz//3LuRBlRDcTMQblLCvAg0qMwq7XSCJQI4uoVSncZzxoPMJWteWbKWQVhAoEtVH9MYO5Pqz1zzJxQlqIhXmC0udI+lopXAjQpF7t5OOM3+JBCJ70/eR7dhwm0CSF4ntFlUCIkIXmnzgGCWMWXjXd+T2C8M2UCQR7DDsZjEDa0Iuxh/syRh3FTmwz5mnEn2dxrWjWpMqhMZi5fz5pprrsm5TlR0EprxGrjbftp4kITnBIwchxxLBKCcn7kQzhEYci5zUx/WgXVJQ9d63sMxz/rTHZrjO47XCRw5fhg2geOcamuOe44RtiFUu4LvCi4iMJ0g+OSTT86cnrWuWe1PtSI3+eEYJQQ899xzvVKS6lUqvjlu2KZ4mB8QenJxhqA53/ZJYh7cGIy2Yp9yAYR5gnajWzrHA/guoNs+NyLiO5H9TTV9EhXmtCufo8qa71uGSGB+3PCJfcN2MXRF1vR8pO3HOCqMqQDO9VhEREREREQkHxVmk8AkhDE1RZZ2+xz0WfRormcf3zR6JCAsvfHGGz0sBcEYQw8wPMHSjBsVES4zfEZc1nRZ+tG9X0RERERERCRuie/2XsioOOvdu3fqX3z8TpGFhW78F1xwgVcOBgyZwLARIiIiIiIiIiJLO1V+SkFT5ed/Fzcxols5XcnjsqbL0k+VnyIiIiIiIpKkyk8RKUjcOT0t4MyaLiIiIiIiIiKFR+GniIiIiIiIiIiIFCSFnyIiIiIiIiIiIlKQFH6KiIiIiIiIiIhIQVL4KSIiIiIiIiIiIgVJ4aeIiIiIiIiIiIgUpAqz54geixScJnu/Hz2a66XB20SPRERERERERESkkKnyU0RERERERERERAqSwk8REREREREREREpSAo/RUREREREREREpCAp/BQREREREREREZGCpPBTRERERERERERECpLCTxERERERERERESlIpQ4/v//+ezvnnHNs2rRp/nzmzJl20UUX2dZbb22NGjWyXr16+fTS6tOnj+222262++67W7t27Wzs2LHRK2a333677bnnnrbTTjtZz549fRrL6dGjhz8Ohg0bZm3btvXHjz32mDVr1sw/98ILL/g0lHY6pk6dWrxcTJ8+3WrXrm0777yz/7HtnTp1sr///ttfzzWvLMll4O2337YDDjjAl9G3b99oavr02bNn+35o0qSJ7bLLLvbggw/6dPTu3dt23XVX/7vvvvuiqYvGBx98YAMHDoyemf3111/WsWNHX8cWLVr4tuCff/6xs88+25o2bWqNGze2Z555xqdzfHXv3t2nt27d2saPH+/TlwYctxwno0ePjqbkh336f/buBF6q8Y/j+C9rRcheEpXI9rckUtIiUUqICJWQ7Hu2kFD2fUlCZE9ZiiJJSUSy72mRlC2RVNb77/O756nTNGfuzL23bXzfr9d9NXNm5izPOWeY7/k9z+EYy8bcuXOtX79+/pi2PPfcc/3xitS8eXP77bffomfF8+uvv9qhhx5qjz/+eDRl+eFc5HhL8uqrr9rFF18cPSu+F1980b766it/zLmcqqj1yOSKK67w757SWtd04uvP99F3333nj0VERERERERWKgVZmjVrVkH//v0LDjrooIJq1aoVzJkzx6cPGjTIn999990F11xzjT9e+KPbX8vWSy+9VNCiRYuCX375peDff/8tGD9+fME+++xTsPDHf8Fbb71VcOCBBxbMnTu3YPbs2QXNmjUrWPiDvmDhj+6CevXq+fuDCy64wNfx+++/L2jcuHHBr7/+WjBz5syCBg0a+LxynY4xY8YUnHTSSQXHHXecP8cff/xRUKtWrehZQcHPP/9ccNhhhxU88MADGeeVJN0yaN+GDRv6PPg8283jpOkjRowo6NixY8E///zj67DzzjsX/PTTTwUTJkwoaNq0qb+X9q1Tp07Bt99+Gy1l2Zk/f37Bc889V3DAAQcU9O7dO5pa4MdJ165dfT3Zz40aNfLpt912m+8/9ufnn3/u68/6Dhs2zNuG6a+88kpBhw4d/P3ZanjgO0v9LS8tW7YsOOKIIwquv/76aEp22F+///579CyzH3/8saB169b+eNy4cQXnnHOOP16ROF/D90NxcTxfeuml0bNl488//4weLYljLdM5wrnEuVVSbN/IkSP9Md9lqYpaj0y6d+9eMHz48BKva1IbIb7+M2bMKPj777/9sYiIiIiIiMjKJOvKz+nTp9uwYcO8ci/um2++sVq1atkpp5xip59+uk+bOHGi/5uthT/Obcstt7QKFSpYmTJlbI899rBu3brZwh/T9vPPP9tRRx1l66yzjm2wwQZe/bnwh7bVqFHDNtxwQ5swYYLPg/dS5UTV2ahRo7yCdL311rPNN9/c/ve//9m4ceNyng4qDTfbbDN/nKRixYpetTplypSM80qSbhlUsTJP5lGuXDkbMmSIbbLJJonT119/fd8Hq622mj/njzZZffXVbc0117S11lrLypYt64+XByqD2U+VKlWKphRi/VgP/l133XV9f+Ovv/6yI444wp9vt912VqVKFZs1a5YNHz7c2rRp49OpgqOKct68ef6ZldmkSZP8uKZKOlSxgorGc845xw488EA79thj7aOPPopeKXT//fd7u7EvMXToUK+Q3W+//ezSSy9dqiL0qquu8uPnhhtu8OezZ8/2ylrmTyVwcPfdd3tFLefHgw8+GE1djKrRo48+2l/nHLvzzjvt7LPP9grmeOUgy2FdmP7II4/4tJkzZ9oxxxzj088666xFVeF8V5x33nl+vFLF+cYbb/j0uDlz5vj68tmWLVva+PHjff/edNNNNnLkSK+ijks3T5bH9n799df277//2uGHH27vv/++TZ061b87eO9xxx23qDKR6vCHH37Yj7eF34F27bXXLtomlsmxyL4B+6dTp06+D/h+Y31ZxxtvvNFfp8169uxpBx10kK//e++959O/+OILa926ta/XZZddZqeddppPD1hvzmXmQ/UkFc7M5+CDD/Zls5yi1iOO4+Kiiy7yym+2j+9lxNd18ODB1rZtW9/H119/vU/75Zdf7OSTT/Zzi3ajzVgXtof1oeo6XZunrj/r9OOPP6bdn7jtttu8HVq1auWvPfnkkz5dREREREREZFnLOvwkxHvsscc85Igj7CCgAcEfdt11V/83WwQEdN8lNKQrOyEmP/IJQ3mtY8eO/j7CDZbBj3AQLrzwwgv++O233/bQbOONN/aQg3AwqFy5sv3www85T0eHDh28u3UmBF0vvfSS1axZM+O8kqRbxrfffuthQosWLbxLLGEQQWbSdALjvfbaywYMGOChDl3uCVR32WUX23nnnf01/niNdVrWCKoJVVhmHAEOYTBd8OkqS8ACwhW2Aex/Ahi6jH///fdLBKi0bVHtuTIglGYf1a5d27v0f/DBBz792Wef9dCXLsMc16mB4AknnOD7p3fv3r6f6b5MUEcXZp6zf+MIlGgnhgwAbXvllVf6hYqxY8d6IP/aa6/Z6NGj/Vzh86wDr6UixHriiSf8PGdYCY4Vjus333zTjzvWgZCUeT/99NPePZ+QjwCW/UkY1qxZMz9PQeC60UYbeaDI0AuEqYRtcbfffrsH3XyWEJLvEwJ6AmKOYQK5uHTzJPwk6OWP4R522203/w7imDr++ON9+zm/zjzzzGgu5scQ7cC2fPrppx6y9+3b18O+OPYD4TvtsO2223qomorAlbYlsH7ggQd82oUXXuj7hv2cGlSCgJkQ8Pzzz7dtttnG12fPPff0gLJq1aqLvteCotaD/cbxQTvecsstSx1X4Lgg1CaMnzZtmr+f/bzjjjt6exKasj/AsCPsU76P07X5DjvssMT6B+n2JyEuCHCfeeYZP76Yj4iIiIiIiMjyUGo3POJHLVVuVFjtvffe0dTsUMFJFRkVadWqVbN77rnHg7FQ4Ud1Vv/+/X3eVCzx4xpUFhEGED7wL9VK4HmoKASfpwoy1+mZEGgRPvJHYEVwR5VZceaVDoEOAQTBVwiyCBSSpgeMpdmuXTsPqgisCK4+++wzD4UeeughDz4IrFYUAjPCFkI4gheqCkM4QoUZARZVYuxvEILG0Z6p01ZGIfzkWKCikOeoU6eO769rrrnGA+IuXbr49HQITAmvN910U6+U5fgm5M+E+XMBgOVSkU2lKYElFdSMS0sAyLRQoRhH+EYF8dZbb+3VzBzbBOuckwR4zIeqQaqIqWzmIsQ777zj1X0hvCewpLIXBLEsp3Pnzh7isp/DGJEB86TaETvttJNXMxPMJUmaJ2EiYT/fIYx7yvn58ccfe9BG4D5o0KAlAkO2A4SEXGAJ20lYGcf5xPHIdxLrGS68xLGfsf3223u4y3HMuRfCfL6nilK+fHmv2gTzYR/FFbUetCPfmWwH4TnjEKdifxIIE5Ry0YHqcLY/fG8SNPfq1csfU2kflpHNfgwy7U+C8TXWWMP3E8eziIiIiIiIyPJQKr9A+SFNl0u6qse72maLH/QEE4QPVJbSJZJAlIotgi66jBIEceMcwp2AKkA+w49zqpKoFgVhUejiCqqqeG+u0zNZe+21PXjkj2pUqqqoWCvOvNKh0orKK/4laCL8YDiBpOlUSxJqEnxQIUb1G8En0+lqSlUsQRrVtWPGjImWsvyNGDHCq/Go5iQ0Yhvoyk4Xd9abqjeCqjAMAP9S/RnQnrTxyowKN7quc04QBr7++uteyUcwTsUcQShVwlTUXXLJJdGnlkbQGw/SCY6YRyYEn6kIoLhwwI2R+GP5BFmpCM4ClpVOfH14f7p1DOEngR7Vf2G5nKOpoVyu25hpntxkie7fXDThXKQtCEN5H//yPREQ7oELE/GhIDgO4xcr6NbO9xOBL925qRZNRVAcx3dWfJuyufkToXOmISmKWg/CxPgyqS5ORbUllZicTwSUVAWzrmFfsy84vxDaB9nsxyDT/iTsFxEREREREVneShx+8kOYcQrpAskPeCoL43dqzxZVTVRMgfFFGbNuq6228i71/HimezdBWSp+xHOXdMKkEEIQ8BH6cfd1wgyqlrgTfa7Ti6O05kUVGCEl40ZSTUbASniZNJ1KM7qcEv4wnbusEwzTJZUQmUCIajLC2ng31eWNsVpDtS7VY1SFUclLV1263Z544olLhHAcV2HMTMJcungzHMLKjJCeEJ/qW/6o9CQUokqS4/ypp57yKjuCT7YpHUIp9isXBdjXfJ7QMnUYARRVCUtVJGNnUjXMccC4jUmVe5lQGUplJOEggR7HHhcj+KPyGxz7oWs7y6VCGQT0hN2pwSbbE/YvFcpUmHLeJ0maJ/Mg8KO6lepFwkCO81DJSZsTRqdi+VRQMw+ORdomjiEpOI8J7Akg4wFqEr6nCF45B0Hgn6SofRcUtR7sGwJRtoMAM7XbO8OKEMTTJlTGMhQGF5T4XBi2hC73XMRJlWk/pq5/rvuTY5LtShWfnvRYREREREREJBslDj8J00DIQJda/ggTckHXX7qI0nWWrpl0U6W6ie6fBEZUzvF6+KMra0D3VcLW0HUTVBUyT7qhM6/u3bt7JVOu04ujtOZF12O60zPuImOb0qWV4QSSphMcErpQ/UogzPIZb5KQjWCYrr10v+Xf0L12RSDwo9KTm+8wJig3TCG0DsFgfD9Tyco6E6qznVSuhW65KytCIao82T8BYS77ivCScI1zhe0naOL4SMW2E3ZR4crr3PSG/Uq1L/s5jrYhGGVsxiRU6bHfOUeYD+vG0AO5otsy68Y5R5UuxzkVxdyIiYCdeROgcdyB45/QjZvpMB4mFylSK0r5vuAmO3yWcTIJ8KluTJJungRsHBuEm2wbF074XqJNqPjkggThL2NwpmK/cE6xfMZg5fiKryPLYzxVXidQ5JjNBjeGogqe45bvhHQXbgi3qagkUCxKUevBccHxQcBJm4au+AHBMN+vfJ7vB845jgmGKmHYAi4y0Fbxm1sFSfsx3frnuj8J0MOYznFM53su02MRERERERGRbJQpIDkRyVONmk+IHi02alhhOCeyrBAQUsVMIElVM12+CRFlafQcUKApIiIiIiIiy4rCz+WASjRuMpJO+/btizUmqGRH4aesCAzt8Pjjj3sXbYZq4AZXjJ0pS2KoAYZi4OZcIiIiIiIiIsuCwk/Jawo/RURERERERET+u0rlbu8iIiIiIiIiIiIiKxuFnyIiIiIiIiIiIpKXFH6KiIiIiIiIiIhIXlL4KSIiIiIiIiIiInlJ4aeIiIiIiIiIiIjkJYWfIiIiIiIiIiIikpfKFCwUPV5k3rx50SORVVuLNp9FjxYbOmj76JGI5JPy5ctHj0REREREREQKqfJTRERERERERERE8pLCTxEREREREREREclLCj9FREREREREREQkLyn8FBERERERERERkbyk8FNERERERERERETyksJPERERERERERERyUs5h5/fffeddevWzRYsWODP//33X7v66qutYcOG/tevXz+fnqvevXtbs2bNrHnz5nbiiSfaxIkTo1fMHnjgAWvVqpU1adLEbr75Zp92++2323XXXeePg1dffdU6dOjgj5955hlr2bKlf+6VV17xach1erpl//nnn7bjjjvafvvt53/16tWz8847z+bPn++vJ80rSbplBPPmzVtiWkFBgV111VV24IEHWtOmTe2JJ56IXimU+v4gafqy9v7779vzzz8fPTP7/fff7eyzz/Z1P/jgg238+PE+/Y8//rBLLrnEWrRoYQcccIANGzbMp//zzz/Wq1cvn96uXTubMmWKT19Z9e3bd9FxsdNOO1njxo39McfqXXfdZYMGDYremZtzzjnHPvzww+hZ8YwePdouv/zy6Fnx/Prrr3bUUUfZgAEDoimFPvvsMxs1apQ/znU7X375ZZs0aVL0rGhHH320zZgxw2699VZ79tln/djm+MgHc+fOtYcfftgfl+R4yUV83914441LnK/Ziu/DI4880r7//nt/LCIiIiIiIrKiZR1+/vzzz/bYY4/Zaaed5oHDX3/95dMfffRR/4F+wQUXWIMGDfzH81dffeWvZYuAkB/PTz31lA0dOtROPfVU/yNMfOedd2zw4MEe8j399NP+I/21117zcHH48OEeBgbMh1Dwhx9+8BCK9b3vvvs8eGJeuU5PWjbWXnttXx5/L7zwgs2cOdMGDhyYOK8kmZbxxhtv2IUXXmhffPGFPwevf/PNN95OLOOWW26xWbNm+Wvp3o+k6csS4Tjt0r17dw/MA9Z5nXXW8X13xRVXLArj7r//fltttdX8M3fccYe/RtA2cuRIb1umd+nSxYPQlVnnzp0XHRfly5f39eYx7b+i7bnnnnbWWWdFz4rn3XfftR122MHatm0bTSk0depUP5aL480337Tp06dHz3JXrlw5P37yAedNccLHkojvuxNOOMED+1zF9+Ftt91mG2+8sT8WERERERERWdGyDj+//fZbDyipsorbfPPNPRA96KCDPPzE33//7f9m66effrIqVapYhQoVrEyZMrb77rt7mMp8Zs+ebUcccYQHZhtssIHVrVvXw7Dq1atbxYoV7b333vN58F5CQ6pHx4wZY/vuu6+tt956ttlmm3kF3ttvv53z9KRlp2I96tev7yFC0rySZFoG89t00039cbD++ut7wEZQuMkmm3jIENo73fuRNH1ZIsRhOzg+4ljvsmXL+r9sM/sbhOmHHXaYP992222tcuXKHrgTHB5yyCE+vVGjRvbll18udQyuSj799FOvbKbSl7AcVMNefPHFfuxSVTlu3DifnopqWNqI6mgCc1Bhx/FA6M/nCZXx8ccfe1hMde25555rc+bMsQkTJngwBULZ9u3be9sSoKfi/VTocl63adPGP0u7U3FNBSnVzQH7qU+fPvbSSy95KI9025luXdlWAn3WK7X6k6CV7aUSmH+TqkM5dgjtkPSZTp062fXXX+/T2J4PPvjAp3/++efe5ry/a9eui46te++916fRPo888ohPi+OcDRd5aCMq10E1Khc9ktajqHbnYgnnK/NBcY+XdO/hnGzdurVNmzbNK/aPOeYYe/3115fYdw899JBX0FORfcYZZ/i6sn1Um4OLTTfddJNP4wIU+y51H3K88Z2e7hgCFa1XXnmlHX744f5a2C4RERERERGRZSHr8HPnnXf2Lu10aYzbf//9/cc5P/ip8uOHfq1ataJXs8M8+LFOqMOPf8IVukUThvIaP9LBj3bCxX322cef8+P7xRdf9MdULhGaEQYSshA8BpUqVbIff/wx5+mZlh03efJkGzFihNWoUSNxXkkyLYPuvYQVcQTDderU8WrbY4891rvKh+Wlez+Spi9LBLkcF6xrHGEKYTDbTSjDMQMqEtk2sP/p7l6tWjWvpI23J48ztefKjlDynnvu8XAtBGZUCm+44YYePhHQEfz/8ssv/locxwf7nSpHKmA51gid9tprLz8PGNYgDDvB+UjARki6zTbbLNVl/pprrvEQi/lRSZzapqzbFlts4WEdQRXB4Jprrmmnn366H3OHHnpo9E7zdSdoJegL3c/TbWe6dSXsJ9Rm/3P+xBGw0hYEc4R/IfDNJNNnCPyoriaA7d+/v0/jvWwbwTLbx78EgvzRNnRBHzJkiFc2xjHUBQEhwS/DYIRgj9CT15LWo6h2p0J466239tAQxT1e0r2H8JNAlCCTSvNddtnFv2tS910wduxY30+sO9tHsElgTTd5pt15552+PUn7MN0xFHoMfPLJJ74ODPlBGCoiIiIiIiKyrJTaDY/4IU31I+EB1Uq54Ec6gQ4VSFtttZU/JmQNVVhUG9Fdmh/pjC/KD2oQlhI6EmrwLz/gwfNQUQg+T3VkrtPD43TLZozKMLYjXfQJ7qgGyzSvJEnLyISAheXRLZzxD1cV7Fu6TRMqXXvttR5+hkCEAJxg5u677/bwBoSgcbRV6rRVCcfLGmus4aF4QBhMJSIV1D179vT2IFBPRTjFsUVVLOcbwRjnCQEzxw+BJ0Ec9t57b29HplFdlxra8zpjrDLUBPuAKuI4gr0QhjG+LV3LCeuylW47k9Y1CUMicJ4RkBGYFvV+ZPpM+H7YbrvtfEgFgkf+rV27tlciE9ARVHIhhYpsxvE9//zzvYIxVIoGtB/vYzoXYaiM5/uKauftt98+cT2KavdUxT1ekt5DUEkVOMOVUNmZCd9prN/qq69uNWvW9LaiypMLF0wjpGUIlCSZjqGwXVzMoO1FRERERERElpUS/+qkWzI/5Ak1GKeR8S2pXswFQQiVafyYphqQqiy6klN5RNDFzV4++ugje/LJJz2oCPjhzGf4oR+qRcEPdqriAkIO3pvr9EzLjo/5Sbhx2WWXeeVY0rySZFpGOmwnXb8JQ6jsIwR76623oldXfnSppfqT7vCEUVT3MhYp45ZSobrlllt6OBbajKCG6s+A9qSNV1UbbbRR9GgxxgYleCP8549qvV133TV6NT3CJ8IjKgKpnKPimUrbgKCRbshrrbWWVxFSvRhHNSChPW3L+J10tY4jZI6H+CyPYD9b6bYzaV2TnHLKKf5dQlhJt+lsZPoM1chxnHucswFhJQEfwzJQjR32B99vdJuPIxgkfOaGXjym2p1qTs5f2i1pPYpq91TFPV4yvYebKhHGFjV8BN/BqbiQE28zKl+TLu5kOoYYvkNERERERERkeShx+Mm4dz169PDAMownyHicuSIUofIPVFHxRxBGt1F+MNO9knE0U1FZRNdMuluGH+vcdZ7xPwliCdWogOJmL7lOL2rZ6STNK0muy6DKk+6khBd0dSV8IQBeVXBshGpdutFSCUalK11oGQPwuOOO85AkoDst4TIIeekKT2CaT+gK/vjjj/tjzieGjgghURztQKBEdSEXCxhPljFv6cZNNR4V1yGIIlwkmOrQoYOHbFwgCAi/qHBkX5x55plelUf4HsfFjNDujIv522+/WdWqVf15kqIqcpPWFamfZfsJ+ekqvccee/h5FCqEk+T6GQJ2gjzeR7sy5AY3G6I6kopGuokTELZr127RmJ0BoTJV6iFUZHkPPvigV2QnrUc27Y6i2jGb4yXpPVTmM9buySef7DenC4paZsBxQdd35sW5y8WqIHUexTmGAtqe789Mj0VERERERESyUeLws2PHjn5TFbrYckMkum4TbuSCm5VQPUXIQHdIfpgTklBNxRh6jLdHt93wR4VkwA09+HHPTVQCqgqPP/54XzfmxTh3/ODPdXpRy04naV5Jcl0GASHhH2N4EmrxfLfddoteXfnRjZhgiKpPxghk7EpCa9qBADzeDryPoQ0Ihalk5HWqi/MNlbAE/wT5VGnS9Z+qzlTsd0Ksk046yduBqkAqEgn/GQKBqlhCKSoQmWe3bt383CQ0jldarrvuuj6dY4jPURGZes5Sufj111/7+6hqJigj8EvCxQeqejPdqTxpXRlPmOmcxwFdofkuoQKTAJdxSwndJk6cGL1jabl+hqpEKjE5BjnOqPokzCPMpD2o2KSNGN+UoRpSMcwHbcl+oeKTim+6tSetB6F1Ue1ORWQIYpNkc7ykew/d9zmH6M7PccTd2fnuyWbfBXTxJ/RlG8JxyLLT7cNcj6E4qmapkM30WERERERERCQbZRb+0C6IHi9SVHdIkVVFizafRY8WGzpo++iRiOQTuvuLiIiIiIiIxCn8XA6osBo4cGD0bElUu2YaE1RKRuGnyH+Hwk8RERERERFJpfBT8prCT5H/DoWfIiIiIiIikqrEY36KiIiIiIiIiIiIrIwUfoqIiIiIiIiIiEheUvgpIiIiIiIiIiIieUnhp4iIiIiIiIiIiOQlhZ8iIiIiIiIiIiKSlxR+ioiIiIiIiIiISF4qU7BQ9Fgk7zRqPiF6tNioYbWjRyIiIiIiIiIiks9U+SkiIiIiIiIiIiJ5SeGniIiIyH/OAnv8tAneQ6LRaTNtWjRVRERERCTfqNu75DV1excRkRVi+kzr3HmGTYyeoln32nZJ3ejJijZusjXqYdZzWHWrH00qDdMGfGod+s23mp12tL5ty0ZTF/LlzY6e2BKvh88ES7RTSjuuVG0oIiIiIqsEVX6KiIiIlCaCvs4zrFr32n7BrfCvulmPyTY2estKoXo52zJ6WHKzrVfzCXaVVbIuCzd1SQtf67HAuvSN2qJvZbN+k+3x6Qtfmj7Trhpd0fpH7dS/UzkbvqidFn5uYTtapx0LP9e94sLXPi38nIiIiIhIlhR+ioiIiJQagr7ZaSoUK9oli6osCQo/tccHTC7sdt48hH2xruhLTEfha73GRU9ByLqoyzqvEwwWhpCFn08fFFJp6VWYk2dYB953Q6jIzLT8pHUO2L7a1rdtueh5zPQFNsXKWtUq0fMqZa1a9NCqVLK+d1WyqtHTqvUqWs2F6zGN9R4324YvnO9xoYK0bmXrUn2+TVH4KSIiIiI5yDn8nDlzpl1wwQW2YMGCaEqhGTNm2Mknn2xXX311NKV03HHHHbbvvvta48aNrX379vbll19Gr5jde++9tv/++1u9evXsuuuu82k333zzUuvwyiuv2JFHHumPBw4caE2bNvXPvfzyyz4NuU7HvHnzFi0Xf/75p1WvvvCHTf36/rf77rvbGWecYfPnF3blyjSvdG699VZr0KCB//Xv3z+amn67kW76XXfdZbvuuuuivx122MG6d+/ury0P7777rj333HPRM7Pff//dTj31VNtnn33sgAMOsLfeesun//HHH9a1a1dr0qSJNWzY0J5//nmf/s8//1iPHj18+mGHHWaTJ0/26auKvfbay+rWrevHA/927tzZfvvtt+jV9F588UX76quvomeLxaczv1Qcj7TTinbOOefYxRdfHD0rntLalk8//dRGjhwZPStaUW38X9GvXz9r2bKl/fvvv9GUkjnkkEPsu+++i54tKb6v27RpY99++60/FlllRYFdwyK7Zs+3Pv3odl5YFVrfg8dPrM/W1QurHL0KcoF1WypkzGThPDvPtGpRhWX/Trbw+dKfr9p2B6+itOqVCysuuy58nNXyU9c5S1Uq2XFNZkfzYjmTbUqn6tYuhKFx0+fbxCgonTZt4f9rpqlOncJ0EREREZEsZR1+/vzzz/bwww97eEOI99dff0WvFCJwfPXVV23cuHhJQskMHz7cw4ghQ4Z4gHHWWWfZCSec4GHi22+/bc8884w9++yzNnToUH991KhR1rp1axs2bJjFhzJlPvyQ/+GHH+zuu++2p59+2rflqquu8nnlOh2vv/66hzyff/65Pw/WXnttGzt2rP8RuhIKP/HEExnnlQ6h4QsvvODrThB45513+ryStjtp+mmnnWbvv//+or+9997bjj322Ggpyw7h+ODBg+2SSy7xwDwgxF133XXttddes549e9pFF13k0/v06WOrrbaatxkhLp/79ddfPSRmu5l++umnexC6qmH/cTy8+eabfnw88MAD0Svp8d5vvvkmerZY0vSgXLly9sgjj0TPVoy5c+faxx9/bNdcc000JbPU75GgtLaFsDwE7Nkoqo2XtaT2WN74LnnwwQf9nCwNvXv3tk022SR6tqSV4bgVWZa8yjJtJSXjV8YCxOmzbeTkitbTg8hCVdtWsmahCjJLzbrvsChUzOnzWS5/iXXOQf2uta2nB6CfWB+rbJfFxwNdZIE9/vBsq9mpcsIyylrVraOHIiIiIiJZyvqX7fTp0z1UpHIvFUEgIQPVfKXpp59+si233NIqVKhgZcqUsT322MO6detmf//9t4exRx11lK2zzjq2wQYbeLUjIVmNGjVsww03tAkTCm90w3sJZZs3b+5hIBWk6623nm2++eb2v//9z8PaXKeD7d1ss838cZKKFSt61eqUKVMyziud1Vdf3dZcc01ba621rGzZsv4YSdudND2OCkyWW7NmzWjKskP4yfIrVaoUTSlEmML28C8hKPsVhD5HHHGEP99uu+2sSpUqNmvWLA9/qQZjOtVhhM1Uiq2K2Ab+wnFDID97dmFXQ4Jpguo33njDg94bb7xxierP1OlUxBIeH3zwwR5mz5kzx9swBNsfffSRderUyatrCY15PY7PTZxYePuI/fbbzyusccMNN9g999xj33//vXXs2NEaNWrkxzDnPoqaLxdBCLtD9TXzY/5UPIeAi20h9OZCBiEb52ezZs38fYTebEd8W6iUPemkk7yq+cILL/T34bbbbrPLLrvMWrVq5dOefPJJnx5wTlD5zMUALqB88cUXfizxXrYtHsojmzZGuvWNozqdSnOqHali7tKli3+W7eB7KKCduRDCcg466CBfFuE+084880w/3vl8qOJmHlRNs1yOnfHjx/v0pH1S1HqCtuFzvOfSSy/16vVHH33Uv7OYV7zyk/XkPRwTfL+wb1Nx0enaa6/1+bHPQ9Ut8yJU5js8nL8ffPCBV3PH97VIPvIqSyolqbTMxCseU5WzaiXu5p3l55fZ8lHYFX90g6iitOFs65CmS/7YGzIFo1hg06ZGD0VEREREspR1+Elo9thjj9kxxxwTTSnED9crr7zSKxlDQFdaDjzwQK8kIzQMlaX8UCcM5TUCDHz99dceLhLSgOpPqiZBRSRh2sYbb+zdLgkeg8qVK3tFZq7T0aFDB//hnsmkSZPspZde8rAx07zS2WWXXWznnXf2btP8EQzymaTtztQeINS4//777cQTT4ymLFsEsAyDwLrHEdIQ+tKVn3CIwAfnnXeeByNgPxO0MIQAIVw8QKUNM7XbyojtZF/Url3bj4lMxw2hEsHR+eefb9tss000denptMGee+7p1bVVq1ZddLwHVNoR9HH8bbvtth6uxnGhgopIAkKOjRCkvfPOO75vCAGpEuY4IkDs27evv17UfLk4wQULQjKqdlkGwSkVz1T3EkCC+ffq1csDMt5LN2vmucYaa9jtt9/u7wlYPucD8+N7iIrggOCPAJXvJoaJiOMiCIFyixYtPCDlGOM5y2Z/hGMvyKaNaa+i1he0J8NLjB492sNuKrczITCl3UNoHCq5WQbVqISILIeLAqw/ASOV8Hz/ptsn2aznjz/+aFdccYWHrbQtzwcMGODf8VtttZXPN175+dlnn3nbsy78PfXUU0sNf8K+ZqgBLlpwzMQrtbnowTE1ZswYf05VP+eGSN6pW9Ga2WwbnWtHmCrlbOlLk/NtyuRyVi1d9/CsZfn5ZbZ8ql9n2vDqle3YaCiAqm2r+9idI99Y/B0y9oYJ1m1qZesfH/+zalmzyfMttR6/GtNFRERERLJU4j6NdOHdbbfdrE6dOtGU0kN4QbUY3S+rVavmlVL8WA6VQ1QZ9e/f34477ji7/vrrPRgAVVH84KdqiX+pqgLPQ6Uh+DyVoblOz4SxK8OYn1S2EehRkZnrvOgiTdhAFdZDDz3kXadDcJS03UnTQZBCxRaVoSsSAdiOO+7oIQvjszI2ZKhIo6r48ssv94o+tgOEoHFsY+q0lR1DEdDNn9CH46Kk42GifPnyvj+x/fbbLxEIgnCTduScoXIxHoQjhJ/vvfeen1NU5XFeUa3LuLBHH320B9cEY4RgBKThc5nmG8f8qXSkepmKZ95LuArGwyUUpIs8+5+gjiE1Pvzww6UCVc4FKhjBMuOYTrDHvDJ10abqkup1KilBGErAmEm6Ns5mfUEAzPciCLuLWhbnZbwtWc/111/fA0PCW9qSv7D9O+20k3cXnzZtWtp9ks16UnlJqLzpppt62/E9mWk9WSbB8KBBg7w69pdffll07gZUz3Ihhsp1vrMJOOPYFsJbjBgxYtF3s0h+qWjH+h3LU25OVJQqFa1J9dnWbdHNh0JoWNHq+3/OC7t8Dx8TvT59pnXmpkUp4ndDX/LzRShy+cW3VIjpXexDiFl4k6XU4NOlBsnjZlifydmMpyoiIiIisliJw08qm6gAokqPah+qfkoj3AE/5vnBzo9oqpHo1kogSpBEAEYVFz/g6c4dD1+pDuQzVBhSoUS1KPiRH7/pBpVdvDfX6ZnEx/ykao6qWCpic50X1Y8ENFStElBQ/Up4lrTdmdoDhKdU2a1oBB7HH3+8V3MShBCK0ZWdLu5Ur1FhR7gSuobzL9WfAe1GW66KCNM4jtMFTEXdBCkVwVimSmvCS84fgke6ShPAxhHKE4gxPASPCTw5vzhuCOkZn5ZAjYrpU045JfpU0fNNFQ/8CcQIrxFCeNqEG0FRocgfoTjLjuMiQZhPajtRYZyt+LpkCkqDdG2czfoidbvTXeiIb0umixKEu5zftF18vkznokq6fZLNeibNLwkhNMcC7zn88MP9OzYV2xlvM87r+LYTJjMfKnapLt1oo42iV0Tyi3d171vZpvQIY30u/OuRaSxLlLV2d1W3ZiPD3dQnWIfRFZcIBOt3jb3eeb4dl6YrfbPulWxK5+jz/cpaz9RAMVHRy88s3GX+E+sz2Wxiv098Hp0HLDCrWz0a77Nwvo06zzDrtGPh3fA90Fz4b7jzfPjzELaiXRJvxx4LrEvf4o05KiIiIiL/XSUOPxlz7/HHH/c/gpOtt97ax+crLQQwYZxRKreoUONHM2PV8SP8pptu8gAtFRVQjJVHl3PG3gQBIqEiNxriRzlVbwQEuU4vjlznRbdbQl6q8ag4I0xlWtJ2Z2oPxgCkuzXd6Fc0xmQNVbmMq0jlGhWqt9xyi4+TSLd8wqKAbsjhzu+EJoTsDHuwKiJsouptiy228OfspzDuJvsvLqm6NWl6KoZl4DgjaCYcSx1flqCM8IqxMKnCpHv3fffd512vQShK0EUVH1WEIcAqar5xzJPKPz5L0MfFgNRQnopKgjCGauCYYGxKLqbEMR+6ZYMLLLmizThmGDYidLlmu1PXJcjUxtmsLziuCZdB1TXbQJvTdnyfUUnLhYAknPuct7yP4591pRI3nAtUhfM634Xp9kk268lFFS4uMbYyxyZtkjpMRRzz40IS5ynvZxtTKz/5PF3fWSavH3roodErhahWZegCbojF8CRJ6E7PNmV6LLLSq1LJ+kZ3TQ9/fReNZVnRLln43MO/JRROX/SZpYLH+OvVrX7d6kW/J5q6lCI/u/BvideT1jlI+Wz0F7aZGx6lm+7rEZu+6C/ceGmJdlx8MycRERERkWyVOPzkRzY/ePkjZKTiKF1FUHFwoxCq0ug2SnfhTp06eWUTy6T7LDda4vXwx7h/Ad1tGUcv3q2SakPmSTd05sWYfFRc5Tq9OHKdF5VVBLeET3Tj51+qppK2O1N7ELrSfvEqrxWFEIZKT7r1MiYo3Wc5blh/gu74+tPNn22nAo+u2YzpyDiRq5ow5idhN1WtdEUGxwNjLhLwxqsACaUY0zEEo0HS9HQ4xrp27ephFV2RaetUBJ2EgvwRzlGZTBdq0FWaiwdcRKDaljBrwIABWc03oEs6+5FzkapetpdK5jjOAaqjeY3jm+Ax9eY3jG1JtSyVwtwgKt3FjiScQ4SMVENzYYAbO7HuVKxz4SZVUW2czfqCSlrG2OQmTVyw4QZGVIUzFAahH9O5UJSE9WD4Cs579g1tyTymTp3q68+Nn5g/gWq6fZLNelJBzfcQ0/ksVZgEm0k4jjlPOSbuvfdeX6fUcVPZNraL+TEGMecrFaVx7EeC1EyV6ATltFWmxyIiIiIiIiLZKFNACY9InqKbXCqqR2TVMXDgQB8OgRCQcI/xYOnmvbLiogs3gGO81OLgc1R1MoyFiEj2GDvzE5vSPlN1poiIiIjIf4/CzxWIirAnnngierak9u3bFzm+qBRN4eeqjypMKoTp7sxYnVSkllZ1+bKg8FNERERERERk5aHwU/Kawk9Z3hjjlPCSm7MVB+P8MszAuuuuG00RERERERERkeJS+Cl5LV34KSIiki1dMBMRERERWbWV+IZHIiIiIvlKF9FERERERFZtCj9FREREREREREQkLyn8FBERERERERERkbyk8FNEREQkgcb8FBERERFZtemGR5LXdLd3EREREREREZH/LlV+ioiIiIiIiIiISF5KW/k5b9686JHIqq1Fm8+iR4sNHbR99EhE8kn58uWjRyIiIiIiIiKFVPkpIiIiIiIiIiIieUnhp4iIiIiIiIiIiOQlhZ8iIiIiIiIiIiKSlxR+ioiIiIiIiIiISF5S+CkiIiIiIiIiIiJ5Kefw87vvvrNu3brZggULoilmDRs2tB133HHR3/jx46NXSq53797WrFkza968uZ144ok2ceLE6BWzBx54wFq1amVNmjSxm2++2afdfvvtdt111/nj4NVXX7UOHTr442eeecZatmzpn3vllVd8GnKdnm7Zf/75p2//fvvt53/16tWz8847z+bPn++vJ80rSbplBNyRP3VacNddd9mTTz4ZPTO78847bf/99/e/xx57LJq6fLz//vv2/PPPR8/Mfv/9dzv77LOtadOmdvDBBy86Vv744w+75JJLrEWLFnbAAQfYsGHDfPo///xjvXr18unt2rWzKVOm+PSV3Y033rjEdgc9e/bMuO9Hjx5tl19+uT++//77/Zhalm699VZ79tlno2eLvfzyy0udR5x7Xbp0iZ6tmjgvl7VDDjnEfvvtt+hZoc8++8xGjRrljzk/Bw0a5I+zwb6YNGlS9Cy92bNnW5s2baJny1Zxj8vXX3/d/9uBI4880r7//nt/LCIiIiIiIrIsZR1+/vzzzx6cnXbaaR6W/PXXXz597ty59tNPP1mPHj08LOGvevXq/lpJERLxw/+pp56yoUOH2qmnnup/hInvvPOODR482J544gl7+umnPVh47bXXPFwcPny4FRQURHMpnM+BBx5oP/zwg/Xt29e347777vN1ZV65Tk9aNtZee21fHn8vvPCCzZw50wYOHJg4rySZlvHGG2/YhRdeaF988YU/j+NzhKZh3oSPL730ks+L9bjnnnt8nZY1wnG2v3v37h6YB2z/Ouus4/voiiuuWCLoW2211fwzd9xxh7/266+/2siRI319mU7wRhC6KjjhhBOscePG0bPs7bnnnnbWWWf5Y86zv//+2x/nk/DdsSop6TpPnTrVz83iePPNN2369OnRs2Un220sjePytttus4033jh6JiIiIiIiIrLsZB1+fvvttx5EUnEY9/XXX3uYRVVh7dq1PXzcaKONoldLhlC1SpUqVqFCBStTpoztvvvudsEFF/gPbyqdjjjiCF/2BhtsYHXr1vWQjOC1YsWK9t577/k8eC+hIdWjY8aMsX333dfWW28922yzzWynnXayt99+O+fpSctOxXrUr1/fg4+keSXJtAzmt+mmm/rjOKrNbrnlFmvbtm00xWz11Ve3NdZYw9Zaay0rW7asrbnmmtEryxbhJ+u7+eabR1MKEXCyHvzLtrFfQfBy2GGH+fNtt93WKleu7IE7ITKVdExv1KiRffnll0sdg8sDYTpVklTXUbU2ZMgQn071WufOnT1c5xgj1MVDDz3kFcdUyF122WX+OtXHRYVYEyZM8GCIz9N+IQgNrr/+env88cf9MQExISvGjh27qCqT9TzooIP8XAzvJWzn/A3SVUA++uijvp5sIxWo6XCx48wzz/RKXC5EsI/A+6lSZrkE3qlB2rhx47wdTjnlFA/i+d7o2LGjt9lJJ520qAqwU6dOfo5T/cvxT7U0xzNVwuGcLsln2Y9UTLOuhx9+uH3yySc+Pd08qTrmmKTNCd2pWr744ov9PUcddZRvEwj3jz/+eN/2rl27ehVzHG3Up08fvwjBRRx8+umnXsnOenBRAumOJZbBhQ+OidTqT8LU1q1b+/bde++90dTk9kmaznHJRYljjz3WnwdcfOAiC8vgOPv444+XOi7T7XfajYthbAdV6xyPqc4991z/fqcK9sorr/R9wTxCW4iIiIiIiIiUlqzDz5133tn69evnwU8cP6gJGPmRS1dmftCXVlddAlUCB7q8E97wQ5tlEIby2jHHHOPvmzZtmoeL++yzjz8n9HnxxRf9MQEBYRpVRvzYJ3gMKlWqZD/++GPO0zMtO27y5Mk2YsQIq1GjRuK8kmRaxtFHH+2BRKqrr77awyXC0oD9Rld8hibg79BDD/VlL2usA8dCnTp1oimF2rdv76Ev20eARGACwhTCbbCfCVCqVavmFbPxduNxpnZbVujyTZhJFfJNN93kXXhBMLXXXnv58UaoxjkSR5hDyEPlKsfwW2+9Fb2SGSEV+4nQK27vvfdeNFQAAdo333zjgROhKUE7YTGvE/wRaFEFTGBcFN7DewlL+dznn38evbIkKokJzgjxWL+7777bwz0CLAK+5557zsN2pqeirQhsCVcJEQndCPg4ls8///zoXWbbbLONB4VcyCBkfuSRRzzYDF3FS/JZvpu4KMFrHJ8MtYCkeX711Vc+fAXHKZXbG264oW87gShB6y+//GLXXnutv4d9TKjM+RrHZwimCWUJjUGQSBU2oSVDeyDdscRFD0J/zg++RwK2g6CVdidMjn/nJm1LpnbjnIoPlQGGbeACBfuU7xyO3fhxmbTfOXbmzJnj7cQflevxYVJSEUBT4c7xRxgqIiIiIiIiUppKfMMjQkUqhvjhT/hGF00elwZCA7pD8+N6q6228seEr6HyjyoughqCBZa9xRZb+HTCUkLHf//91/8ldADPQ6Uh+DzBba7Tw+N0y6bqiwCEPyrjCPSoQss0ryRJy0iHNqeqNDWEJbCgezzBAuENwUo2Ydiywj7cYYcd7OGHH/bQKF4lSNB91VVXeYDCuoIQNI42SZ22PGy99dYeMF166aUeOrI/wPFIwMt+YjiD1OCfIJLKOCpwCY123XXX6JXi2WOPPeyjjz7ycKlcuXIebhM2UdlIAEfYT2Uflb4cDwSi7777bvTpZHyOymSCQYZu4GJGOrQDVcsgQGPetAf7kGo/hsUg2GMdU7HthNecI3yGsI1jm27UH374YfQuW7RsAkyGDmBbCP7Y5pJ8NggBJOcoVZsEf0nzLF++/KJzitD+gw8+8G1k7Fa2mQscBM9UX4LlU9lcFJZNWBi/EFHUsRTHRSeCyV122cWf852HpPYpqt04ZlLx3UUgy/i1XMwIVcYB80u337nYQljLMghJqSAN53g6oS04NqgIFxERERERESlNJf6lWbVqVf/RTvdmfkAT8KXrAl4cBAD8QCdwoUqwf//+Hs7QxZcA7JxzzvEf2/ygp8t9wI9oPkNYEapFsckmmyzq6glCD96b6/RMy46P+UnQSFdfuponzStJpmWkQ/gZuqASLBIyUkVFl3/CHqpfCcoIuBgzdEWhepLqT44XQmmqeAlnZ82a5dVlW265pYc/oW3o3k/1Z0C70ZbLGyEa1ZQENbRfCLSp3KOduQhAJWEqwpx46E1gVRIEnpxzVNoRJnJc0DWaNqLqEfHlEboSGMcxHmxqGJW6nuyXovAZjm0CQsYq5SIFfxx/DL+QKmw7n2FoDNqO9/Mvx0XAOgfxxyjJZ1OxvWwDYWXSPOP7i+0MlY78UdXIPoi3GyEe3wFFSTc0CMvNdCzFJR1XSe1TVLvxeb5zOAf5YwiE7bff3iudCY95f6jSDpL2OxdcCD85PxiyggtXmay//vrRIxEREREREZHSV+LwkzHgwrh111xzjQctqV2dS4IwgIpAMO4ofwRkdG3lxzXLpMItFYEfXUf54U5gCrp9EwYS/hC2UcXFj/dcpxe17HSS5pUk12VQLUl3VrrzEi5SpUU1FhVwhMVUy1KBRWUu01YUlh2qchnDkG7bVLRyR3rG/TvuuOOWCK3o8huGMCBUoSt8NsFcaaMN6TbcoEED7yrN+KoEjlRcMk4j3fiphEut5iWcZJ+wvbw/dHvnOWF4ajCZKl2VK13fH3zwQdttt928EpQuwzwGzxnbk/VgHRkugXVYd911ves+COZZfhxVfoTnoUov3G0/FePNsp0ggKWb9nbbbeddl+nuzXzp2k5QnITgjuMgjEFKt+hw46uilOSzAecWuJkWwRzHUzbzZFvDGKp0h2c8ULY37GOENkwn3b6My3QspX6WAJzxV0OFbVh+Uvtk026cd1x44I8hN/juZT8yVAZd7ONjFLM+SfudaVxw4nzm+Gac20yVn+nQTZ7vyUyPRURERERERLJR4vCTMSapJrzhhhs82KFbcGmFnwR4hDLt2rXziruTTz7Zw4FatWp5d1uCPLqkhj8qJAOqUAkouOlGQLUhNyZh3DrmRZhFxVOu04tadjpJ80pSnGWkQ3BBAEwFFmEEwQr7a0VhnEG63VP1SXUYN5IhnGZ7CVvi28v76M5L+Et1Ma8TsKwI4ZjmuGI/MnYi+5TxJAnZqQSlKpUQKIwvCdqc4Ru4QECAFLoXE5DRRZoAKwmBJsd7KsJPglOqDmvWrOmBOu0FzhPOGfY7FXysK1W/BHUcT6wH3ZFTKw95D9vCucbnkoJ5ulkzXAHj6rK9rB/HMef9GWec4ecbwRiV2plQRch4nOxf1ueiiy6KXilaST7LsUbISJd9hoIIwxdkM08uKnAhhgsrZ599tg/RQKUnNwUiHKfrO9WgIYiO4xyk0pJxNJMkHUtUbDOd77OASk7GHaWynH1NuB0uGiRtS67txvYQlnKuss/5zkI4LpP2O8cGbcx2MG4pxyT/fcgFoT3DhmR6LCIiIiIiIpKNMgVpSs9WxN20RZaFFm0+ix4tNnTQ9tGjFYsgirCIAE1ESo6u+CIri2kDPrWrrLr1bVv0OMDFt8AeP+0Tm9K+tl1SN5pUqgrn32fywofVK1v/uypZ1cIXVhGzrVfzmVat7w7Wrko0SURERET+cxR+rkB0B2W4gHSowMs0JqhkZ2UNP7kBDZXSK7IKVyTfKPyUlUdq6BYLERepaD2HVbf60bPiKVn4SUDb4etKNqpr4fBASxk32Rr1sFJYzyX5cvvNt5qddlwyHPblzY6e2BKvh88EzbrHtnn6TOvceYYVDu6y5GvLJ4QWERERkZWZbq27AlWpUsW7z6b7U/CZ37gpjoJPEZE8NW62Da9e0eqnVBsS5o0aVnvh347Wpfps63bD4qBvpVW9nG0ZPSw5QuEJdpVVWrj90aRFFr7WY4F16Uv7LPzrW9ms32R7fPrCl6bPtKtGV7T+3na1rX+ncja8x2QbGz7XeYZZaNvuFRe+9mnh5xaqWrWsTRw926YVPhURERGR/yCFnyIiIiKlaOyY2WZbl83QRbys1W9YzmzqgiiUo4JzgvUa508KUQV52szFoR3Vjc0nWKPob4n3Bnym+eLgL8w3fKZR8xAYFk73SsqRfGaCdR6wwF8JqJj0KszJM6wDn10U1CbNE4SbC5c/oHCeS76GinbJsNrWt+3CbU81fYFNWdguVUNgXKWsVYseWpVK1jfW5b5qvYpWc+F6TGM7CZoXzve4UNlZt7J1qT7fpoQ2qFvRmk2ebWMXtYmIiIiI/Nco/BQREREpNQts2lSzmlulCfgWWWBjR8+3mg0rZjmGZkp147DqZg/HglF4l3EqJxePbzn2hk+sz9bVo89QMbnAunmgWtba3VVYQWlNCl9P7RZetaGaFkIAANj3SURBVO0OXkXpY33yee8aX9jNfql5LhFyzrc+/egqX7ieWXeXr1LJjmsyO5oXy5lsUzpVTz9W5/T5NjEKSqdNW5C2OnUK0105qxYPQ0VERETkP0fhp4iIiEipmW9TJptVq7r0GJMT+30SVUQy/mc5a1Ivy3EoU6sbqaCM33xo2kzrnBJ8Uin64Mhy1qXd4vE8q7atVLIqyOmzbeTkitYzNkaozzNUYUaadS/eGKH1u9a2nh6ALmwfq2yXpR2nc4E9/vBsq9mpcsIyylrVraOHrvD54jBURERERP5rFH6KiIiILAeLx/xc+Ne9rPXpHO+iXoQMY28O77f4Zj9Lmr9wGUt2UR8evVIsXnGZqrQqKwvHAx3dIKoobTjbOizRhb+QV7MmBqMorLwVEREREQkUfoqIiIgsb4xFSZVotqHh5Pn2TfQwFXc379/JrM81KV3h/Y7yUdi66C9WHZqrKuWsZvRwMSpdy1m14s4zMm3ATBtevbIdG92lvWrb6j5258g3Fldsjr1hgnWbWtn6x8f/pMI2Tdukq7wVERERkf8mhZ8iIiIipYZKyCy6WXtX9hAaFnbNHs6NksDNjbjZUOBB6Wx7cNFNiWZbr/jNkBbysNBmWIdwY6IqFa1J9fhnEiy66VIWonnG71JfGFoufWf7XC0VYnoX+xBiFt5kKTX4dFHbjA43gBo3w/pMrmgNoxA1VIIqDBURERH57ypTsFD0WCTv0M0vFZUvIiIiy4pXKFr16CZBiG4UNDl6GqFi85JFIR3dvkO39IrWs7tZt4fLLQ77CEQ7h+7t5aLxPQvnO6V9mE/hPKZ02jG6gVGa5XIDozTzpEt+6k2P/CZK8XVw8fVcKD6/6DVbYrviUj4bCcv2dhsZTVxo0Tr5zZxiYXDAzZpo47Rt408WYpkzrdoS00RERETkv0Thp+Q1hZ8iIrLcpQ0NZYXw4JS7zxfvJkwiIiIisupTt3cRERGR0kRX7JLcVV1KzViGEmhSUcGniIiIyH+Ywk8RERGRUlXRju1kS9ysR1aE2TZ6ZDnr0i4MPyAiIiIi/0Xq9i55Td3eRURERERERET+u1T5KSIiIiIiIiIiInlJ4aeIiIiIiIiIiIjkpZzDz5kzZ9oFF1xgCxYsHsdq6NCh1qxZM9tjjz3sqquusn/++Sd6peTuuOMO23fffa1x48bWvn17+/LLL6NXzO69917bf//9rV69enbdddf5tJtvvtmuvvpqfxy88sorduSRR/rjgQMHWtOmTf1zL7/8sk9DrtPTLfvPP/+06tWrW/369f1v9913tzPOOMPmz5/vryfNK5N58+Ytmn9cuuksc9ddd1309/333/v04iy3tLz77rv23HPPRc/Mfv/9dzv11FNtn332sQMOOMDeeustn/7HH39Y165drUmTJtawYUN7/vnnfTrHUo8ePXz6YYcdZpMnT/bpy9N3333nbRs3e/ZsP95Lqk2bNvbtt9/aa6+95udVtl588cWljvNMkt7fp08fP26zMXfuXOvXr58/Zr+de+65/jju1VdftYsvvjh6ltmvv/5qhx56qD3++OPRlOzE1yPJaaedZu+//370bOV39NFH29dffx09yx/XXHPNEue/iIiIiIiIyPKWdfj5888/28MPP2ydO3f2MO2vv/7y6VOmTLGzzjrLgzXCK37ohkCrpIYPH+6hzZAhQ2zkyJG+nBNOOMHDxLffftueeeYZe/bZZz185fVRo0ZZ69atbdiwYRYfypT5tGzZ0n744Qe7++677emnn/ZtIahlXrlOT1o21l57bRs7dqz/EbrOmDHDnnjiicR5ZfL666/bOeecY59//nk0pVC66QRYa621lgc+4W+zzTYr1nJLA+H44MGD7ZJLLvHAPOjfv7+tu+66Hvb17NnTLrroIp9OCLfaaqt5mxEs8znCMcJa2pDpp59+ugehUno4l//+++/oWWbs06KCrLp169r5558fPcvsnXfesZ122snatWsXTclONushK4eTTz7ZL7yIiIiIiIiIrChZh5/Tp0/3UJHKvbgXXnjBwywqHOvVq2fjx4/3f0vDTz/9ZFtuuaVVqFDBypQp45V23bp187CGMPaoo46yddZZxzbYYANfJiFZjRo1bMMNN7QJEwpvdMN7qUZr3ry5B5RUkK633nq2+eab2//+9z8bN25cztOTlp2qYsWKXrVKQJw0r0yociTATJVuOstnvqmKs9zSQEDFOlWqVCmaUoiAs2zZsv4vxw37FYTpRxxxhD/fbrvtrEqVKjZr1iwPrqmOZDrVnwS+VL2uTDgnzjvvPN/XVDK+8cYb0SuLEfQddNBBXtXKv1999VX0ypJ+++03P1aDe+65x0NrPPTQQ9aoUSMP8jmmAx5Teb3ffvt5aBwuTCS9P7j//vt9H51yyin+/IYbbvB5EFY98sgjPi2O4Jxjj/eB6leqeA888EC7/PLLfRrn/4033ugXH/iXZR9yyCF+oSCOfXjTTTf5hQMC2KT2+eyzz7xNmc7FDz6Xuh5JqB5mXhz/XKxA0r765ZdfPKjjGDv88MNt6tSpPv2xxx7zaVQqn3jiiX7hIGkfJW0z8+L7gmUed9xxXkmcDu3Qtm1ba9Wqlb333ns+rajPhosC4GLD9ttv7xdCaKfddtvNz8NLL73UjwO+p0KbcWGBCyjsu2OPPdY++ugjnx6XbtvBMXLmmWd65TbV+Mzriy++8O3m2KPinOORdrrvvvtsxIgR/rl0xxcXN6644gp/jE6dOi1R3S8iIiIiIiJSUlmHn4Rm/Bg+5phjoimFqOr7999//cc5AQWVl/zoLQ38MKeLK+EFXXYJcPjBTRjKax07dvT30V2UkI91AOtAKAuqNAnTNt54Yw8O4gFh5cqVvTIy1+mZlh03adIke+mll6xmzZqJ88qkQ4cO3tU7VbrpdJ0moG7RooVX391yyy0+vTjLLQ2EwoRJe+21VzSlEOEG4WuDBg08IKL6EwRSoRs5+5nu7gwhQJASD1DZluWx/qlYJvs4/BFQBVTWbrTRRh7k3XrrrXb22Wd7mBZHsEVwP3r0aN9uKnFzQbhERSwh3qBBg+zTTz/16QTxhFt0A+dYW2ONNez2229PfH8cVdQcD7179/YQioptLnCwblTiMo+4yy67zPcJFd5gP1555ZX+GSqdCfkDAixCLwI5hq5gu+PKly/v4RsV44SNSe1DWxLockxQ2UygmboeSTg3eT9BGxXDnAtJ+4rhMnbccUefzhAZtCHfa6wX7Ue1Nfg+SZK0zRzbxx9/vFc7c94SHKZDuDhgwACvrie4RlGf5fxiaAnwL9tGkEnld506dTxEJpxku/h76qmnPBAlmOXiA5X1fJelBvaZtp39TNDJ8VatWjVvY75jCTb79u3rxwRDOPA9HWRzfImIiIiIiIgsCyW+4RFjNRJ23nbbbd6l+eOPP/YfzaWBCk6CiwcffNB/ZFNhRTASKv+otGKZVERdf/31Xi2I8MOcH/D8S/UXeB4qDcHnqQzNdXp4nG7ZtEcY85NwiUCPyq1M8yoNhB5UgBH6EkbQHZ/AaFkvN1eEHoRMhDwETowPGSoVqcqjgjAcS0gdP5b1L80xZbO16aabegAV/hiKISAEpFKP0IoqNrYntbKTakWODQJJ9lG242wGBE9cBKCamKEVCA3B+cbyCEBZ/ocffujBV9L7kxBMUc1IwEiVMAEv1ZiZEK5xUYHjq1atWh6yBZyvbO+FF17o68g5kkm69vnxxx89mGQ5VAozfiTnf7YIqFm3LbbYwqsgaZukfUX4F74nCGN79erly3zggQc82OWY/OSTTzLut3TbzHMec+GI4J8wkf2TTqgmDW2ZzWcJMKmOZxvYPr6P2G/jx4/3fciwAgxDwGepSqU92WbalKpR2pQLFV26dInmWCjTtpcrV86PLbCuIehnnFW+e1ZfffWluroX5/gSERERERERKQ0lDj8JKAlA9t57b+9WyQ9buquXBsJOfuwTKlBx+uSTT/ryCJ8IwPix/cEHH/j4f/yYD6gO5DMEHVQ7US0KAiyqvwKq+XhvrtMzLTs+5icVoVRBrbnmmonzKi3bbrutd5kl7GF/0F2VyqplvdxcURlHJRvVnFSpcrzQlZ0u7nRvr1q1qgc1oVs//4YbN4H1Z5tWJlQxXnvttV59yR/HHDe7imObOR7olhxuvpWNUEVNGBUPsUNVHcum0jcsm1CZ6sak92cSfz8BFkFzJhxnSQi5qPKj+zOVgwcffLAH8UnStQ8hPedOQBgYD1hzwfYwr6R9xTlN1SzYbo4zqmo5Jqmopus3XcfTCfso3TazXNqJCzgsj3/5XkontT1Z32w+y7rRdqw/oSLBJ8Ei1dVvvvmmD2tA2xPq8r2IHXbYwQN8qtI5Xqiujcu07QSu8f0SEI6yHNou3bGTzfFVWr0GRERERERERIISh5+EbISdVBVRLTVnzpylujqXBFVHYZxRfoh/8803ttVWW3llIz+0GTeQAC0VwQN3Q+fHPdVvoFqJaki6lxK2UQFGcJTr9KKWnU7SvEoL3U25+zbrxViMBMS1a9de5svNFWOyhqpcqtWmTZvmVbN00yf0YmxBgpGArrRUk4Egh+7O2QR5yxOhfxiXc+LEiR7qsn0BjwmiCZj23HNP3weh2jUVARr7iWOeMCmMl0gFMUEd4R+fDW1CWEg3Z7p4sxyWQdfmpPenQ/DHetEFmsCRAIowLR7qB9lW3Y4ZM8a7bDMURvfu3f17IR5ixyW1DyE9IRvPCcoY+iKMoxnWgy7ctFc6VJDyOcY15SIKQ3ck7SuWy3kNuq1z0YLvG8J3hm/Yeuut/WIH7ZO0j9JtM9Wr22yzjbct2DfhJl9FIcDO5rOEk4Sju+66q3/fMR4q68f3JMcGF384t2gLzjfalu9V5kcgSrtzbsUlbXsmfN9y/jKUQmr1f9LxxfdnqJJm6A6Wg/h+TXosIiIiIiIiko0Sh58EafzQpwKS0IAAjuqj0kBXTEIcxrqjG3mnTp38BisEPlQ2UV3F6+GP7qEBXSwZgy90ZQXVhsyTbujMi/XmpkW5Ti9q2ekkzau0ME9CQbo30zWYKlDChWW93FwRtFDpSUBEsEJoTjhNmxLIxNuUQIwhDNZff33fJsZoJGBf2dCujE3LhQCq7AiBQhUhCLGoFCZoIoQioKJCMN2NXagcZrgExq1lXxI8gTEV6arNTXqoyAsBNvuSoI59THUeoSA3sEl6fyramTFkqVbkMecN72d+zCOO/RBCyKKECyB8F1C1Hap900lqH8JJjgnG6yTEp3s12xZfD0K0MP5uKs4Hzn9eZ2xZhoZI2leMUUnFJGE7FZYMx0BYSjDPhRTG2mQdeH/SPkraZtaTefJ5QkHGLM1WNp9lPQliudhBdSVV4OxLcN5wbrENDCnAfqYtOC4IIzkPOX74XohL2vYkhKeEmBw7nON33XXXEiFl0vFFKEo1P93k2TcEuGC/0saZHouIiIiIiIhko0wBKYJInmrUvPCu/3GjhtWOHkk+4K71CsREREREREREJB2FnysQXfifeOKJ6NmS2rdvv0LH5swXCj/zG924uQEaN94REREREREREUml8FPymsJPEREREREREZH/rhKP+SkiIiIiIiIiIiKyMlL4KSIiIiIiIiIiInlJ4aeIiIiIiIiIiIjkJYWfIiIiIiIiIiIikpcUfoqIiIiIiIiIiEheUvgpIiIiIiIiIiIiealMwULR40XmzZsXPRJZtbVo81n0aLGhg7aPHolIPilfvnz0SERERERERKSQKj9FREREREREREQkLyn8FBERERERERERkbyk8FNERERERERERETyksJPERERERERERERyUsKP0VERERERERERCQvKfwUERERERERERGRvJRz+Pndd99Zt27dbMGCBf781FNPtR133HGJv3POOcdfW5Z69+5tzZo1s+bNm9uJJ55oEydOjF4xe+CBB6xVq1bWpEkTu/nmm31ahw4dbOTIkf44uP766+3222/3x88884y1bNnSP/fKK6/4NOQ6Pd2y//zzT2+X/fbbz//q1atn5513ns2fP99fT5pXJvPmzVs0fxQUFNhVV11lBx54oDVt2tSeeOKJ6BWzW2+91fbff3876KCDbPjw4dHU5eP999+3559/Pnpm9vvvv9vZZ5/t63jwwQfb+PHjffoff/xhl1xyibVo0cIOOOAAGzZsmE//559/rFevXj69Xbt2NmXKFJ++ojz88MPWuHFj++2336Ip5ttw5plnRs9WHb/++qsdddRRNmDAgGhKoc8++8xGjRrljzt16mTTpk3zxyvajTfeuMSxlGr06NF2+eWXR8+WFt+u0nLXXXfZoEGDome5mT17trVp0yZ6VjqKaoNsHXLIIUsc4yjJtqbiPGLb//3332iKiIiIiIiIyLKRdfj5888/22OPPWannXaaPfvss/bXX3/59OOOO86uu+46/yPQK1OmjO22227+2rJCQPjyyy/bU089ZUOHDvUAlj/CxHfeeccGDx7s4d/TTz/tYcdrr73m4dlLL70UzaEQ8yE8/eGHH6xv376+fffdd59vC/PKdXrSsrH22mv78vh74YUXbObMmTZw4MDEeWXyxhtv2IUXXmhffPFFNMV8Wd988423B/O65ZZbbNasWb7NEyZM8PW599577dprr/V9uawRjrOd3bt398A8YN3WWWcdD2GvuOKKRUHN/fffb6uttpp/5o477vDXCOcIrGkrpnfp0sWD0BWNoPa2226Lnq263n33Xdthhx2sbdu20ZRCU6dO9WN5ZXPCCSd48Jxkzz33tLPOOit6trRctouLCX///Xf0bOUSvnvTKaoNVoR0bcn3ZJ8+ffycFxEREREREVmWsv7l+e2333rgSMVhHD+2qVrkj2Bhjz32sGOPPTZ6ddn46aefrEqVKlahQgUPW3fffXe74IIL/Ac21VRHHHGEB2wbbLCB1a1b18MzqgnHjBmzqGKVKjACyZo1a/r0fffd19Zbbz3bbLPNbKeddrK333475+lJy05VsWJFq1+/vocxSfPKhM9tuumm0bNC66+/vnXu3NnDhE022cQ23nhjb48PP/zQAyPaaosttvC2GjduXPSpZYd2Zts333zzaEoh1q9s2bL+L+3E/gOBzmGHHebPt912W6tcubKHtITFVKExvVGjRvbll18udQwub+xj2pC2TUWgyHZwvPHvpEmTvLL36quvjt5h1rFjR/v444+jZ0uaM2eOV8ZSpUtlHME1qLq78sor7fDDD/fXCM5BJe3FF1/sVdBUcabbt+nmSRtS9UylIOsX0OaEUoTmBOng9fbt2/uyP/jgA5/29ddf+3aw3JNOOsm+//57n05IzXvZZwTwyGYd2T7WkYsUVAWzTKpOWecnn3zS3/PQQw/Zq6++6o+Zfv7553v1MJXfBOVsVwilU9cjdbuS1ok2ueeeezwQnjx5sldN02ZHHnmkDRkyxN+T6tNPP/V1oHI77Jd0xwH4jmzdurWvNxcj0uG79Nxzz/UqbtaDYx5Ur3PxgO9XKqJ79Ojh76HKnPVEpjZA0nZzgeL444/3du3atasH/Omk29Zs2jJesc3FIY4fLpZR+cm6s1y2+/HHH4/eJSIiIiIiIlI6sg4/d955Z+vXr5+HAOlQjUiV40UXXbTMq3nows0Pbqo2qZQkwCEwIeDjtWOOOcbfR3ddwsV99tnHA8ddd93VXn/9dX9txIgRHrSA4IbgMahUqZL9+OOPOU9PWnYqQhWWX6NGjcR5ZXL00Ud7gBJHqFmnTh3vlko4QiDCfKtXr+77hmpS5kv4QkXoskb4S0jCOsURxhDu0lYEJVSGgmo1tgHsT8KdatWqeWVsvH14XFT7LGtrrbWWXXrppV6dmlrRRmhHEE/IRkBEhRv7gm0i6CEQ/uWXXzzkTofhHAipCa4IOwmiQqXfJ5984sERQysQFoKq4Q033NADPYZxYNnMPy7dPNdcc007/fTTfd0OPfTQ6J3m86LCltAunB8E2XRTJox89NFHfRphF2EcFbwciwSRuOaaa+ymm27y45BKZPZVNusIlkN7nXzyyX5eMx8CS7Y3FRcAOP95/1ZbbbVomIQgdT04nuLblWmdOEfCdhK2UmHOvMJ3RyqCbEI+wkzaGumOA4a/oO3ZB+F5OqwHFbkvvviiB5IcawHtSRj8+eefe6jN+vNHZXe4sBPksi+oCGc4Do4RhuZIGuog3bZm05ZcZAo477fccksPPWlfho2gvQh22dch7BUREREREREpDaWWUvKjlQrGWrVqRVOWHX5o002aYITgg8eEsqEikG6W/JAm7KDijuAHVBeFkITwkfAEhFKhAhGhm2au08PjdMumkiqM+UkXfYI+Kggzzas4CDCYL93FZ8yY4cEUVZQEMAxZQFUllZcrCvuKYIcwjcCF8DOEewTajFt69913e6ACQqs42id12opAVS/HOtWIcXTjZ19zPhBeEXBRlbvNNtv4+KeES+yLJARBIXRknNhy5cp5cAWOnTXWWMMD4HCBgSCZakz2bc+ePb0tCdfjMs0zG1T0gWOHCku2jwpAQjiOc4bBCFWwe++9t4/dSmDIvqUKOZt1RMOGDX37COwJzTnPqfBODfXANvB9g7BecenWIy7TOrEeBNxbb7217z/CR7Y3Xr0bF/YLFy6CdMcB1Y5UO++yyy7+nvD9kw7fVWDefI5zA2FfsB+5YEDbU+nJ9ofzKMhlX1AxSjUqCMSTviPSbWs2bZmEcJRt4j1Uv1MRT9WsiIiIiIiISGkplfCTiqK33nrLq/mWB8bGJGwhnKCKqH///l7ZOXbsWA/GuOHSRx995OFM7dq1o08V/qgnCKKCjh/bBKcgFAjddsH2EDDlOj3TsuNjfhKGXHbZZV59lzSvXFFZSMUUoQTdXAlY2Cerr766VygS9nJjm/Llyy8RXCxvVHpR/Ul3eKrwqNZl7FKqUalopSKM8Di0Ad37qf4MaB/abGVAtSPddKdPnx5NMTvllFO84ne77bZb4mY2bCvDRlAdR/feJIS78TCc/UdADkLUVOxPKgm5EMAf86fCOS7TPLOx0UYbRY8Kcdwyjeo/lsm/oTs6lX+E++wzujtToZnNOiJeMU7AlgkhIuuRJN16xGVaJ14D3xFUJBL4UT0dLlakSm0fpDsO2L74fmAbisL7+VwI/MNnOLcJP1kfzvfwXRaXy76Irxdtz/dVOum2NZu2zCT12OR4FRERERERESktpRJ+UvXDj/DULs7LEt1+QzUU45HyR3BGN1PWhS6fVBLFUS1GNVQYKy+gOoku+3QNJ4RjexjLNNfpmZadJGleuaLKk26oVJjR5ZQqQ8JhxjYlbKQaiwoynu+1117Rp5Y/qvoIYmknxkGkApHq2DvvvNPHlOQGWgQgAeN8EhaDwIeu8ASmKwMqE6l2C+Mtsk0E0HRtZuxb9mWoxiN4Z8xIPkPwSzUj+zsV51DYXro2c8ftqlWr+vN02JdhnMSvvvrKx5dkPeJynScyVdcSxrEfCXNBl2sqHefOnetVrbzG3e+pTuRCQDbrWJqS1gNhu7JZJy6m0L2/QYMGXkFJu8WD+CRJxwFtzrqFdUkaQxShQp0bmXG8pH6fcAGHoT44ZwgLCeDDsYZc9wUXasL6cCEltZI2k5LsX9qH44hqd9qXwJh1iZ8fSY9FREREREREslEq4efEiRM9sEq9uc2ywl2f6Tberl07r8pijEBu9lOrVi3vMvnmm2/6WJvhL9wwBXQnJTiIh5+sN2PrcQMX5kXgQYVVrtOLWnY6SfPKFSEIoSDd3Kny4jl33aeLOVWgdHtmPNYbb7wxY8Xcska1JMEQlZBUrnH3dqp2aTsC7Xjb8T66BhP8MKwBr1PFujJhvEwCJhAK0g2ccV8ZD5Ou7oyJyfnBNlAFGLoWE/Jwo6BUVAwSUvM+qoPZX5m6DRNscxGA/csNgxg2ILVqMtd5MhYtlZzPP/98NGVpdKN+5JFHfP/Q9Zpja9111/VlcAxSJcm4r1SDZ7OOpSlpPeLblc06hYs5dMvmHGWfZvMdl3QcUHlJNSb7gOOG9YwH/QGfZ2xYKjqpquViSiqqh9977z3fPsZi5nvwhhtuiF5NboOk7b7wwgs97OUzVG7y3ZGtkuxf1pvvctqDym/amWEMOD+oWkXSYxEREREREZFslClI08dwRd9NW6S0tGjzWfRosaGDto8eLT9U4hHwPPfcc4u6AjNeKKG3SBzV4FRfSu6y6WYvIiIiIiIi/y0KP1dCdGEdOHBg9GxJVLsWZ0zQ/6qVIfyk6pLKZKoBw/iPdPfnPKMaVCRO4WfxKfwUERERERGRVAo/Ja+tLJWfItliXFFu9CW5U/gpIiIiIiIiqUplzE8RESkdCj5FRERERERESo/CTxEREREREREREclLCj9FREREREREREQkLyn8FBERERERERERkbyk8FNERERERERERETyksJPERERERERERERyUtlChaKHovknUbNJ0SPFhs1rHb0SERERERERERE8pkqP0VERERERERERCQvKfwUERERERERERGRvKTwU0RERERERERERPKSwk8RERERERERERHJSwo/RUREREREREREJC/lHH7OnDnTLrjgAluwYIE///vvv+3yyy+33Xff3fbcc0+78847fXppueOOO2zfffe1xo0bW/v27e3LL7+MXjG79957bf/997d69erZdddd59Nuvvlmu/rqq/1x8Morr9iRRx7pjwcOHGhNmzb1z7388ss+DblOT7fsP//806pXr27169f3P9rkjDPOsPnz5/vrSfPKZN68eYvmHyTN56233rJDDjnEl92nT59oavGWW1reffdde+6556JnZr///rudeuqpts8++9gBBxzg64w//vjDunbtak2aNLGGDRva888/79P/+ecf69Gjh08/7LDDbPLkyT59RaF9Oc7DPubvkUceiV4tPbTLueeeGz0r2qeffmojR46MnqV39NFH29dffx09W3Fee+01/w5JNXv2bGvZsmX0rHhoszfeeCN6tmLMnTvX+vXr549z3Y8lkc0xUJrOOeccu/jii6NnheLbftttt9mAAQP8sYiIiIiIiMiKknX4+fPPP9vDDz9snTt39jDtr7/+8umDBg2yJ554wrp162annHKKh48h0Cqp4cOH24svvmhDhgzxH/VnnXWWnXDCCR4mvv322/bMM8/Ys88+a0OHDvXXR40aZa1bt7Zhw4ZZQUFBNJfC+RCq/PDDD3b33Xfb008/7dty1VVX+bxynZ60bKy99to2duxY/yN0nTFjhrdP0rwyef311z1g+Pzzz6Mpljif3377zS688EK75557bMSIEf76d999V6zllgbC8cGDB9sll1zigXnQv39/W3fddT0A69mzp1100UU+nbB2tdVW8zYjWOZzv/76q4e1tCHTTz/9dA9CV7T7779/0T7m79hjj41eScbxyIWCZYVQuLTOOykZjv144L+85HIMhO/v4iLk/Pjjj+2aa66JphRaUdsuIiIiIiIikiTr8HP69OkeKlK5F1euXDkrW7asVznyt8Yaa9jGG28cvVoyP/30k2255ZZWoUIFK1OmjO2xxx4eshIiEcYeddRRts4669gGG2zgFZiEZDVq1LANN9zQJkyY4PPgva+++qo1b97cA0oqSNdbbz3bfPPN7X//+5+NGzcu5+lJy05VsWJFr1qdMmVK4rwyIczYbLPNomeFkuZDOMiymMY+ITDeZJNNirXc0kAIQptUqlQpmlKIgJPjhX8JQdmvIIw54ogj/Pl2221nVapUsVmzZnlw3aZNG59O9SdBMNWwK5s5c+Z4Ret+++3nQfv48eN9OhcDqIYmlI9XrRL8hopRjukQoI4ZM8Y6derkj6mEZJ4HHnigV1eDSthLL73UGjVq5MfdDTfc4MfjXXfd5UE8+z0TLly0bdvWWrVqZe+9955Pmzp1qh/PHD/HHXech+ZxIXgGgfb222/vVc7sh9122833HcumkpftZ/14neD6iiuu8M+B7YpXboMLCWwfn2Ub0nnnnXfsoIMO8opg/v3qq698OuvCtnBuX3/99T4tYN04nl544YVoytIOPvhgmzhxoj9mvakyB23KRYTvv//eOnbs6G1N2/D9h0zLBRcY2NfMB+n2I9+j5513ns/30EMPTVutSuXkZZdd5vuK9XvyySd9err1Sj0GOAZZLt5//3077bTT/DEV8FwEoW3AhRHalW158MEHfVoqtoPlUz0ejlmq67mokVpln7rtBKRU7FN1HtY/m20XERERERERKS1Zh5+EZo899pgdc8wx0ZRChBFbb721/zjmhyzdmQkgSwNhARVGhHf8yCbEJCQhDOU1AgDQlZeQjx/TIGgKoQfhCmEagSyhDgFgULlyZa+MzHV6pmXHTZo0yV566SWrWbNm4rwy6dChg3f1jkuaz7fffms//vijtWjRwrti33TTTbb66qsXa7mlgVD45JNPtr322iuaUogAjPC1QYMG3oWcEBCEIYTbYD8T8jGEAEFPPEBlW5bH+mdC9TP7mz+OR9x+++0e2BIUXnvttV6lHKrrqMYjdNx22239OThPQghNQDRt2jR///jx471twOtXXnmlh1tUmBKif/bZZ14RS7Uxf0899ZSVL1/ewy32PUFZJlT90hWZbXjooYd8Gm1//PHHezUux9uZZ57p0wP2IcMXgH832mgj++ijjzxUq1Onjv3yyy8echKqEXhyHGbT3ZmAlHaiepDzhOfp0HYExKNHj/Zjhipm0DYEywyRQPuxXDCEAsfe4Ycf7t9PSdgH7BuCQ5ZN24OwlX3Avtx77739/CZY7Nu3r7+etNyAwJJjl2EckG4/EjrSjuzDW2+91c4++2xvx1S0M1XmfPfyPqRbLy74ZHsMcP5Qtc7+pk35rmR/MY31i2N/0kasO+1OhfYXX3zh+4MLUwTdcanb/uGHH3o3eI41zhFku+0iIiIiIiIipaHENzwiBKB6iu7vvXv39h/kIZwoKX7QU2lERVK1atW8GovwI1T+0ZWYbtRUq1GBRfgEqp4IU/7991//NwQgPA+VhghdkXOdHh6nWzbBSxgLki76BHpU1WWaVy6S5kOlJVV1BFAhZCEkKa3llhbCkx133NHDFqoiGTMwhIRUhFEZR8UbbQtC0DjWP3Xa8kbYRHDEH8cXCIioJMROO+3k1bcEY6Bida211vLHAeOGfvDBB14xynt32WUXH7MxBG8gWCS0Z//VqlXLQ0/mff755/v5duONN3poFNovG1T4IcyP45XwlXCNYJr5EmrGUaFL0EXFJWEWxzzrSVhIAMx2sP6bbrqpV/RyvnHRoShUnFI9TfUoksb7pJqQ9WQ4BIK6EJJyjhHcMqwEITqVzuBCCdtEuJlJCD+pgOV75ZtvvvHvFiqWd9hhBx8jleCXc4rvtqKWmyTdfiQQZbmE0ATH7MNQ0RrXrFkzr6anApy2RdJ6ZSscAyH4ZVxiAm/WK1QDB7yH93P8Uj3O/mbfZyusPxddgmy3XURERERERKQ0lDj8JPggmCTA4IcuVZmp4UlxEXYyL+ZPxSndJglECZ0IwKh0YvmMMUfAEFAdyGf4kU11UajOI5yhEjKgAor35jo907LjY34SBFPxteaaaybOK1dJ86GSitCMfwkpCHYIpUtruaWF8UipMqSakyo11pWu7HRxp3t71apVPYAL3f35l+rPgPVnm1Y2hLLxkJnAh+AZBHypCDy32mor39ZwszC6/7KtVAoj3fARb775po+ty7ypbOQ4z0XqPDk2mcYFBir0+JfzJhXHE8cz20UARvAZgtpM2x7HuLRxhHnxzxGypsPxwrLpbh9uXAaqBqkcpc0InqmoBO1CoBeqipNwYYIwlyEyeEzgyXcM5zPrRYUiQTztQ5sHSctNkm4/Uq1LhTBtzh/fUxwHqaigTpW0XklS2z0cjxyDBNlhHeguTyCZKr6PqCZnf2erJNsuIiIiIiIiUhpKHH5SxUc3SCr6evXq5T+0d9555+jVkuNHfhhnlHFHqc4iNGJsOwIWuncToKUilOAu6QRJjL0Jus/TpZquv4RtVB/VrVs35+lFLTudpHnlKmk+jP/HeJGMk0o1IGER1XiltdzSwpAIoSqXai+qI6maveWWWzzYOvHEEz1gCRhOIdz5neCPLrUE7CsbKvHCetI1nYpOjtNMCBTvu+8+D94IP6lyDl3/k9ANmjCftiKECt3lESpiWXZq4JWEAHKbbbbxG4uBbvThJlRxrCtB1a677urnFOM6cjyxjRxnXKTg2GOdCNFoD86NUNHHsAxcLIjjswxrEabTvTsVxwnfL9wAizbi+GV7+RzjSLLu3E2d75wwH9aHcSZZRwJl5sEFgNTQjmpGwmPWNwTQ7I9QeUsoSrjIMBdUklIxnWm5cUVVJ9erV88rN8FFCi4EsJ7ZSLdeQVgubR/GM+X7Kh3WgSEFqBqn4pVhS1IrMGkTjg2WwTHF90r8Yk86pbHtrBPHV6bHIiIiIiIiItkocfhJBSQ3zyD8pAqyS5cuXnlVGpgXYRDjENLVtFOnTn7jECrAqDrjbui8Hv7ouhvQVZNu4PEx/6g2ZJ50Q2de3bt39yqoXKcXtex0kuaVq6T5MO4q3ezZF4x5SjdexgUsreWWFkIsKj25yQrjMtJ1m3CaNiXojrcpoRddoddff33fHiruCNhXRnQbphs3wSR33Wd8w9Su7qkIFAnlCN4YD5SAmqrKTGgH2opwn27gVFtT4UggSVUt5yDDCdCW2aKbOBWfBOWEYYzbmIoxf7kIUbt2ba8EZH3ZR6ASl+OKmzax/VQfE84SnFGpzY1yOPYITuOoOmWfErYSgBFqx4NvEM7yHcN3CvNkOxnWga7pfC+wPNqCY4gwMKD6lCEUqLwmLCNwI7hMRdDJcvljfdkf7BdQBckFFOZPxTEBHUFipuWC45WgNfVmQHG0B+vDkAgEmSyHdc5GuvUaMGDAEscA5zxdyrl4kBSEc9yx7nxHsj18d3AxK47ji/3M9ymV2cyXMZSTlNa2E7LyfZbpsYiIiIiIiEg2yiz8oZp9H0aRVUyj5oV3/Y8bNax29Cg/UXnHmKmETFKIcJdwNtuAUURERERERETyg8LPFYgu/Nw0JR267a7IsTnzxX8x/KQylO7cVGBK4U3IGDKBoSFERERERERE5L9F4afktf9i+CkiIiIiIiIiIoVKPOaniIiIiIiIiIiIyMpI4aeIiIiIiIiIiIjkJYWfIiIiIiIiIiIikpcUfoqIiIiIiIiIiEheUvgpIiIiIiIiIiIieUnhp4iIiIiIiIiIiOQlhZ8iIiIiIiIiIiKSl8oULBQ9XmTevHnRI5FVW4s2n0WPFhs6aPvokYjkk/Lly0ePRERERERERAqp8lNERERERERERETyksJPERERERERERERyUsKP0VERERERERERCQvKfwUERERERERERGRvKTwU0RERERERERERPJSzuHnd999Z926dbMFCxb487///tt69uxp9evXt9atW9uoUaN8eq569+5tzZo1s+bNm9uJJ55oEydOjF4xe+CBB6xVq1bWpEkTu/nmm33a7bffbtddd50/Dl599VXr0KGDP37mmWesZcuW/rlXXnnFpyHX6eDu92G5+PPPP23HHXe0/fbbz//q1atn5513ns2fP99fzzSvJKnL4Cb8V111lR144IHWtGlTe+KJJ6JXzG699Vbbf//97aCDDrLhw4dHU83Gjx9vRx55pK/T/fffH00t3vqUlvfff9+ef/756JnZ77//bmeffbZv08EHH+zrjD/++MMuueQSa9GihR1wwAE2bNgwn/7PP/9Yr169fHq7du1sypQpPr00XH755TZo0KDoWSGO5cceeyx6lpuXX37Zj8nRo0f7vFPNnj3b2rRpEz0rPvYtx+DKYO7cufbwww9Hz5a/eFtw3GdrRa93vuLcue+++6JnRcv1/SIiIiIiIiK5yjr8/Pnnn/2H6mmnnWbPPvus/fXXXz79ySeftAEDBthZZ51l//vf/+ycc86xmTNn+mvZIpAjOHrqqads6NChduqpp/ofYeI777xjgwcP9vDv6aef9nD1tdde8zCP4I+QMGA+hIU//PCD9e3bd9EPawIp5pXrdLzxxht24YUX2hdffOHPg7XXXtuXx98LL7zg2zxw4MCM80qSbhls5zfffOPtwbxuueUWmzVrlr300ks2YcIEb4t7773Xrr32Wt83v/32m1122WUeChM2Pvfcc/b9998Xa31KA+E47dK9e3cPzAPWY5111vF9d8UVVywKCQmxVlttNf/MHXfc4a/9+uuvNnLkSG9bpnfp0sWD0JXdnnvu6efDssL5x0WHlQH7OR5uL2/FbYtltd7hezGf/Re2UURERERERPJH1uHnt99+6wElFYpx7733ntWqVcvatm3rFX1UYVGBmYuffvrJqlSpYhUqVLAyZcrY7rvvbhdccIGHGlTLHXHEER6YbbDBBla3bl0Pw6pXr24VK1b05YP3EopSPTpmzBjbd999bb311rPNNtvMdtppJ3v77bdzno6pU6fapptu6o+TsB5UvvLeTPNKkm4Z66+/vnXu3NkDwU022cQ23nhj38YPP/zQGjdu7G21xRZbeFuNGzfOw1LWgWWWK1fOg1g+U5z1KQ2ES+ynzTffPJpSiO0pW7as/8s+ZX+DQOWwww7z59tuu61VrlzZQ13C5UMOOcSnN2rUyL788suljsFlhQpOqmWpsCXETRf6PProox64U9HJ+0E4fdttt/ljwnsqoqlyJaxOp1OnTnb99df79jOfDz74wKezrUcffbQvn2OBEPmhhx7ydk0XrlINzLz49/zzz7e7777bl0s1NEEy0m0T1bU9evTw7aC6mspiUJV7xhlnWPv27f39VCKnIkzn+A2f4V/ey8WJxx9/3KfFffzxxx5is17nnnuuzZkzx6dzEYXqXiqCw4UPKoFDZTMVw1R2xiteU9uC7aAd+b444YQTMs47db0DPvvVV1/5Y7aDinTwPi4evPvuu76fqE7m30mTJvnrtDHB/rHHHuvPA9YntAOBPuuFsWPHejtwgYJ9S9vz3RW2N6mdgkz7Jt0+KGq74riAdemll/o20m4vvviiT0/dxqR9zXKOOeYY3x7amQtUSccY0r1fREREREREpLRkHX7uvPPO1q9fP+9WHUcwR1dkQgCCKvz444/+b7YIawg36PLOj18CGoIKAj5e44cxpk2b5mHePvvs48/50R1+mBMyEZoR+BEoEPQFlSpV8nXKdToInwivMpk8ebKNGDHCatSokXFeSdItg1CzTp063i2bsIHAgPkS+lIpSoDDfNluKkJnzJjhIfKhhx7qIREVoKuvvnqx1qc0EFQzfAHbEEdYQ/jKfj3qqKM8gAMBFtsM9j9hSbVq1bxyNb7+PC7N9WeoAUKX8EeVMQher7zySuvTp49X0a6xxhoeJsYRTjIkA8EPodDnn38evVKIoK5r164+H+abqav6v//+69W8BGH9+/f3aRdffLGddNJJXvXKMU+Q1rFjR9+HIVyN4xggEOOcIEAl2GS5XFggHE/aJtabcI0qY/5YjzCsBSEdbcR8CHVD2BdQsbz11lv7hQ/Of0I5hlmgPWgb2iiOoI0wm/lts802Huaz7VRwEiRzHoHjmvAthIFUABN+r7XWWv4cqW3BcVG7dm2vIGebaYekecfXO44hLNgG2or9xTaD0JPX2DYuzFCBTYgcjhewfILWuL333nvR0A6ffvqpV3OzX5gvFyvYL3vttZevK+3MdyzStVOqdPsmaR8UtV2p+C5h6AnWg2OG7xiEbcy0r5lOuMp60dZcNMt0jKV7v4iIiIiIiEhpKfENj6gGoqqQCiVCFSr0qOjLxYYbbujdnvn8Vltt5Y8JWUOFH5VA/MCmEurqq6/2ikcQlhJoEHDwL2EJeB4qCsHnqZrMdXomjFFJyMgf1WQEd1RXFWdemRBMMF/CH8ItQlKqIgleGIKAwJdKSoIEKqgIKwgk3nzzTa/ALe31KSn27Q477ODjLdJlP15RSQBOBRuBHF31QQgax/qnTisJquoI2MIfxzEIqlgvqtVoZyrxPvroI38tIKihqpbKX4ZBIKCO+/rrr/1c2GWXXfw5x2uScOxut912XqXJMAZUWzdo0MCn81mWlwnnEcchoTfnEWEuCMwJnpK2ifFrCZ8JCQkSWX7YJ8yPCxzMs2bNmosqSNNh/VgmASXfCYR7hGtxhIHsX45TqgYJdakCJvyigvmuu+6yzz77zAM63ktlN+caXdSLughRvnx53x+gHdnmpHknYZlsB+ExF1fYB3wPUWG6/fbb+zANrA9hH4FlfF6hveP22GMPb2PWhYpsLiIRBLJdnNt8z3GBgO832iTML107pUq3b5L2QVHblYrjjbbju4Z15lhB2MZM+5p1ZRrfSxzXTM90jKV7v4iIiIiIiEhpKXH4udFGG/mYn1RbPfLIIx5OhXAyW/zAp7KJSiyqAfv37++BEpVNBF10wyRAoOKIyq6AKkA+QyVhqBYFgQAVjwHVSrw31+mZxMf8JARhvM0111yzWPNKh+2hkorKNirACNDeeustDzroPkvYS7sT+PAe9gNhCv8SJPCYSrDSWp/SQiBL9Sfd4Qk6qO5lrFMqy6iA3XLLLT0ICuvIcABUfwasP9u0rNGujN1JIM8fYS3jrsYRDsWDZbYlLvX1TBcFqJRNFf8s8ypK6nuo7IxL2iaOK4IpgnKONYLTgPMwF/F15ljl+yCOsI/wi9CMqsshQ4Z4NSLV3QRyHLch8OU9DHNBlSohHWMKZ8JxzzkYlzTvJASKBH3cqIvHDOlB9TXfO2zbKaec4tXnhKupN69Kt38JPKtWrerbsOuuu/p8CGI5pqnoZCgEQlkq1qmUDtK1U6qkfZNuHxS1XalS5xEumMS3MWlfx6dzDPKX6RhL934RERERERGR0lLi8JPupHRH/eSTT/zHOgFLuiqlohAAUPkHggr+CMLoXsoP5muuucbDjVSMSUfXT7qchzCgYcOGPv4nXcMJ1ah2IvTJdXpxlNa8qPKkYo1KsF9++cVDC4JeKtcID6maorKQ53SbJdShOzzdVXk/60DFVmluW2mgCjFU6xLO0g2YsPzOO++0ww8/3I477jgPUgKOLcJlEKDQFT41ZFwWCLc4phlqgXUlcKaiNo4QiZA6VLGFO9QHhF7cVTxUjKYLsJKwjYTaXAAA844H/8Wpfk3aJqZx4YD2J8CaPn36oqq8bIR1ocqRLssEZVSuEhLG1xkEfASSVIwzTjAXLjjXCbl5jTajvcI8Cchvuukmr35MkqktMs073ecIGwnm6JpNWMk2Pfjggx6c0mZckGAoA6ZzLmXTTlRdMo/ddtvNP8fN23gMKkAZ6oBhIKjMDSFjunbKRtI+yLRd6XDOcSwwzizfPXyXxCUtB6+//rpPo0KW45bvm0zHWLr3U8nO9xWSHouIiIiIiIhko8ThZ7t27fwmFjfeeKN35+RO3bmGU9wEhCCJedGN/OSTT/ZAgOokukDShZtANfzFx9Wj6yXdvVmHgKrC448/3scEZF6MnUjFUq7Ti6O05kVIQDvS1Zfwg+cEJnQZpwqU0Peiiy7ydqfajWCDoIRglKoxusUTipbmtpUGbsRDgESoRSUYd28ntGY/E4DH9zPvo/stoTfbxOsEdssDbcRNXxhDk2OLoIyq5DiGHOjUqZMft1StpobK7BfG6aQqmLFY11133SWC3aIQ+N9zzz3eFZ9wif0NjgPOj1wlbRPdoAnhGF6BMSc5B2+44YboU5lxYy7CLMbq5XOcx2wr7cFxRxvFcXx269bNt4kwm5CPYI3u1SyfYJHAngsaIDhkPTme0ymqLZLmHV/vVHThZl9x/hHoUTnNelBZy9AbVJJyrlG5yVAJEydOjD6ZHp9lHoSOdE/nQgTHNzh+WB/Wj6pmAlYqMtO1UzYy7YOk7UqHCnIqW/luZj1Sb8iWaTl8R3ERgxtCEa4yFEGmYyzd+zneGU4ESY9FREREREREslGmgAQgxfK6m7bIstaizWfRo8WGDlp6jENZORGYcfOucCMgWfYYZoTwlXBzVUPPAxEREREREZE4hZ/LAV08Bw4cGD1bElWDK3IMznyn8HPVxV3uqfalupnqQFk+FH6KiIiIiIhIPlH4KXlN4adIbhgzmBCRcUJXNQo/RUREREREJJXCT8l78QBUwadI/lL4KSIiIiIiIqkUfoqISF5Q+CkiIiIiIiKpSny3dxEREREREREREZGVkcJPERERERERERERyUsKP0VERERERERERCQvacxPyWs7jzgyemT2UdMno0ciko805qeIiIiIiIikSht+iuSDGoNbRY8Wm3TwkOiRiIiIiIiIiIjkO3V7FxERERERERERkbyk8FNERERERERERETyksJPERERERERERERyUsKP0VERERERERERCQvKfwUERERERERERGRvKTwU0RERERERERERPJS1uHnP//8Y5dffrntvvvuVrduXbv55pujV8zuuecen9akSRMbPHhwNDU3d9xxh+27777WuHFja9++vX355ZfRK2b33nuv7b///lavXj277rrrfBrLv/rqq/1x8Morr9iRRx7pjwcOHGhNmzb1z7388ss+DUnT0y2joKDAt7lRo0a2zz772KOPPurT//zzT6tevbrVr1/f/2iTM844w+bPn++vJy0jk3nz5i1abnDDDTdYgwYNbL/99rNhw4ZFU83eeustO+SQQ3zZffr0iaaaP991110X/X3//fc+PWk+y8O7775rzz33XPTM7Pfff7dTTz3V2/OAAw7wbcGsWbOsY8eO1rBhQ58+fvx4n85x16NHDz+2DjvsMJs8ebJPX5W8+uqrdvHFF0fPFps9e7a1bNkyepYe+3Rldc4556Tdrmy8+OKL9tVXX/njo48+2r7++mt//F/Tr18/Pwb+/fffaEru5s6d6/MJ+E7gO4rvFM4bERERERERkf+0giwNGjSooFq1agV33313wTXXXOOPX3755YKxY8f64wceeKCgV69eBTVq1CiYOnVq9KnsvPTSSwUtWrQo+OWXXwr+/fffgvHjxxfss88+BQt/vBe89dZbBQceeGDBwh/4BbNnzy5o1qxZwauvvlrw1VdfFdSrV8/fH1xwwQUF/fv3L/j+++8LGjduXPDrr78WzJw5s6BBgwY+r6TpScsYMWJEQceOHQv++ecf/+zOO+9c8NNPPxX88ccfBbVq1YqWWlDw888/Fxx22GHeBknLyGTMmDEFJ510UsFxxx0XTSkoeOGFFwqOOOKIgjlz5hR88803BXvvvXfBrFmz/HnDhg193syXdeUx69SoUaPo04slzWdZmz9/fsFzzz1XcMABBxT07t07mlrgx0/Xrl29TdnPYZ2vvvrqgjvvvNMfDx8+vKB58+b+eNiwYd427OdXXnmloEOHDj49G9Wfa7nU34rAfuK4ScVxc9BBB0XP0uMYL6k///wzelR6fvvtt4KmTZtGz4qWug6XXnppwciRI/1xu3btcv7OKKll0SbF0apVq4Iff/wxelY8fL5169bRswLfL7///rufM99++200VUREREREROS/KevKz2+++cZq1aplp5xyip1++uk+beLEiV5tufHGG1unTp18OhVMY8aM8dez9dNPP9mWW25pFSpUsDJlytgee+xh3bp1s7///tt+/vlnO+qoo2ydddaxDTbYwCszZ8yYYTVq1LANN9zQJkyY4PPgvVTYNW/e3EaNGuUVpOutt55tvvnm9r///c/GjRuXOD1pGeuvv75v72qrrWabbLKJ/7GcVBUrVvSq1SlTpiQuIxOqGTfbbLPoWaH333/fq0dpkypVqnibjB071tubZTHvcuXK2ZAhQ3y9WF+mpUqaz7K2YMECX6dKlSpFUwrRlmXLlvV/1113Xd/f2GGHHezQQw/1xzvvvLP98ccf/nj48OHWpk0bfx9VbJ9//rlXtK1oH330kR/zVKly3M+ZM8enP/bYY76eVLaeeOKJXg1MFeuNN97or7/99tt24IEH+ufuuusun5YJla89e/a0gw8+2I499thFy+FYb9asmVfzXnLJJfbXX3/59IDPHXTQQf5ZKmf5HBW3vJ9Kw1BZm7Qdd999t1fhcj49+OCDPi2OquuZM2cuqr6muph5c6w98sgjPu2NN96wiy66yE444QR75plnfBqYznFMm4TqT6ql27Zta61atbL33nvPpxW1jVSHU+lNFTTr2qVLF1//3377zdc7oDL94YcfXqpNmHbmmWf6/uLz8crrdNuT1FZFrSeGDh3qn+M9l156qVdmUknOd0b43gyS1osq8iuuuMIfg3WhDa666ir/DmGd77//fj/v+N5iPThmQHUt2832s/9DG3/xxRfWunVrPyYvu+wyO+2003x6HNX87Bva9Prrr4+mpm+j2267zY+zcA6wX1k27+PcANXf5513nn+Pcc5zPIiIiIiIiIgsK1mHn2eddZb/gAcBH+ha/eOPP3oICQK2Nddc07sw54If3nTdJDQkTCFMIChgfrxGd2jQNZZl86MZ/Gh/4YUX/DGh0nbbbedB7HfffbdEEFi5cmX74YcfEqcnLYOgcK+99rIBAwbYEUcc4d3YU0NKTJo0yV566SWrWbNm4jIy6dChg3fpjttmm208RCY8o43ZPkLib7/91p+3aNHCu0TfdNNNtvrqq/v06dOn+3SGILjlllsyzmdZI0Q++eSTvf3iCGwIg+mGT2hFIANCENqK58ccc4yH36DrfjxApW2Las/loXfv3h7Kst+33XZbD5kJsAh7Bg0aZK+//rq/j/YOCLw4j6655hr/HM+LwrbuueeeHkBVrVrVj3fCegI0ujoznzXWWMNuv/326BOLEYzRzpxTvE74Teh47bXX+noQjqXbjtdee81Gjx7ty+LYf/bZZ5cKzNk/XLBgPQjlGL6AIRWefvpp73ZNqAaW16tXLw/PAi4uEIadf/75fnyC45Nlde7c2R566KGst5EQt3v37r6+nJt33nln9Ep68TYB+4dglmWwjSNHjkzcnnRtlc16ct4RWhJqMm+es60c51tttZXPl4sBcenWKwmhJcNwdO3a1YNmziPmmYrjk31KMPrAAw/4tAsvvNA/zzAEIcxNdeWVV3q7Pv/88zZt2jRf/0z7nAsfXLRgKBDamc8SoIchOgjWN9poI9+mW2+91c4++2z75Zdf/DURERERERGR0pbzDY/4QX7BBRfYcccdZ3vvvfcSFUugQi91WlEIT6kc4gdytWrVvFKLYCxU+BUUFFj//v19mVQeEeKACibCAZbHv1Q1geehohB8norNpOnhcbplgAqmdu3a+Q9+QkZQmUj4yB+BA0Ep1aOZlpELwlDWgcCVCkKqbqn0JFggwCFIIXgI1aCECVSQEW4QUhBUEyInzWdFIQDZcccdPfxh3FbGjIxXyrG+7PsQIFGtF0d7pk5bETgmqHLjWKUqk7CcAItQiX3Ca5988skSAefUqVO9uni33Xbz50WN94ny5cv7mLPYfvvt7ddff7WPP/7Y24zQjbDwww8/9CAuFcsKFwoIqlhP7LTTTn4MEGSl2w7eS6hHeEUFIssMlYLp8H6qAtdaay2veGYe77zzjr/GeLjpLhikCpWaHJ+5bCMBbGhPjp142JxOvE1AZSUV3lQjc+GAbUnannRtlc16fvDBB7bLLrvYpptu6scI31NFrWe69Sop5gOOI8JGKjD5PuO7C0nHI99xVGo+8cQTfnGCSvNM+5yqT0JgqvO5+MH3O/uJ7y5w8YPjifYiFKb9QgWwiIiIiIiISGnLKfzkhy/dWKn64kZAIHQLlZ50NyXsofoyF4QJBAYEn1RDPfnkk/6DmQo0gi66YhIgcOOcOnXqRJ8qrALkM/yYpoqIalEQMlCBGVA9x3uTpictg/CQaiYqqaj4ImR58803/bW1117bQy7+qBSlOoqq16Rl5IpqTtqbCkLWiRCM9aC9qVzjX0IHAhmGH6ASjeo6glfanwCCdU+az4oyYsQIO/74472akzCGbaArO12AOX4IZgjdOKaoZCU4o/ozoD1p4xWNrrwct4Q/dPOlOpLAkOOE9Wa/hNAyIPiKB+N0+y8KARjHVRz7kOpeqg35I0immi4VQV9AaBxfNuEUQX267SAY5SJAmD9DKxBUZRKfN8ccy0N8HTJJ/c7IdhtTl5vuQgPHVZBpfWiTEKyn2550bZXNeia1fbbi6xUX365sMDxHHPOMr1fS/KjOpFKY85DQl676SNrn8SpW1j0VbUb1cWgzvrsJyUVERERERESWhazDT36gMp4c3VUJZKgupAKRruoEVXRtZGw6fhDHK6uyRUUVlUggPGKMUbqEUsFIUED3boKyVPwYZ7l0OQ8/7lkngku60rJuVBkRUCRNT1oGVVFUIBLoUinFncsJWzNJWkauPv30Uw8zqYqiYpAqQiptCdToxk7XddaJ4JWqsr59+9q5557r28FdxAmOa9eunTifFYVqsFCtS7UX1YdUphIqP/XUU/4e1pkghgozjjeONfAeuvcyHMKKxlAF7F+CXEIxAvgQ1tLdf+utt/YwPR7GcTwzvAPTER8Hk/0ZxjktCgEx408yRAPtyDiToe2SUIEX2vGzzz7zLs6sT7rtoFs63fep1KP6miEJMlXm0S2fbtNsK/uNYzJ+kSJJukAvyHYbOX6otgTtyboQTrJNfJ9w7hK4J+E8oS14H+3DeidtT7q2ymY9OT+5uMM+JiAkTE4dDiJVuvXiuynsB76bwnGE1LbM1LYB8yN05nsN6dqJ45WqcYYn4PuF8XhZbnH3OTi+qFwHF264CELbxXHs0daZHouIiIiIiIhkI+vwkwpH8IOXrrD80e2aMI6x++i2zrh7BJGEKrngRiV0vaTbKl0sO3Xq5JVVBAt0paRqkdfDX7hxBuh6SQgburyDqkLmSTd05sWYgFR8JU1PWgY3UyEgoKKUkJXPEShmkrSMXNE1nEpTKjgJHQhhqQAkVKObPWOQMuYpXcQJM1kWoSBBBdMIPAkjkuazohAOUelJt15CQm56Q2jNuIME7Ew/55xzfD2prqUrLmE720QFGuNHrgxob8ZY5Njghi1sCze3IsjlWOH8IAiP30SHdmcbqJ4m8GF/UTGH9u3be+CVDY4nKo05zjj/CLrCjW2SsD6E36wv4zzSvoSE6baDKjzGweWcYjrHGsdREm72wznDuUjlK+vF+LuZEAhS/Uf4lU6228iNstgWjnsumLCdHDecI5wfTOecScJ6UOXK9hLosS1J25OurbJZTyqV+R5gOp+lapvvlkzSrRf/UhHPDYZYF8ZcBucHoWoYx5R1J6jNBjctooqf84vvrtQLTFQn873MenNcc66yTsXZ5wHrTqjKdxLjj3KOpFaIEqayDzM9FhEREREREclGmYU/mgv7KorkmRqDW0WPFpt08JDo0cqFAJ4AKARaUjQuenCn81BFmCs+R3Vlujucr0jLc70IHhkLmECWm6RxozLCSREREREREZF8kfMNjyR3VKRRYZXuLz4+qPx3MYQBlaMiyxNhOzczatWqlVcGF1WRKiIiIiIiIrKqUeWn5K1VqfJTcsd4k1RI0hW8OBjPlLEms7nx1PK0sq6XiIiIiIiIyKpIlZ8iskpimIDiBp/gruMrY8C4sq6XiIiIiIiIyKpI4aeIiIiIiIiIiIjkJYWfIiIiIiIiIiIikpcUfoqIiIiIiIiIiEheUvgpIiIiIiIiIiIieUnhp4iIiIiIiIiIiOSlMgULRY8XmTdvXvRIZNW184gjo0eLfdT0yeiRiOQb7pQvIiIiIiIiEqfKTxEREREREREREclLCj9FREREREREREQkLyn8FBERERERERERkbyk8FNERERERERERETyksJPERERERERERERyUsKP0VERERERERERCQvZR1+/vPPP3bVVVdZvXr1rFGjRnb77bdHrxT67rvvrFu3brZgwYJoSm569+5tzZo1s+bNm9uJJ55oEydOjF4xe+CBB6xVq1bWpEkTu/nmm30ay7/uuuv8cfDqq69ahw4d/PEzzzxjLVu29M+98sorPg1J09MtA+mm//nnn7bjjjvafvvt53+0yXnnnWfz58/315OWkcm8efOWWC7Gjx9vRx55pC/j/vvvj6YWSvd+pJvO5+vWrbvo7/vvv49eWfbef/99e/7556NnZr///rudffbZ1rRpUzv44IN9GzFr1izr3LmzHXDAAT59woQJPp3jrlevXtaiRQtr166dTZkyxaeXtnHjxvn89913X9/X119/ve/n4pg7d649/PDD/vj111/38+K/hGOQ/YXPPvvMRo0a5Y+zcfTRR9uMGTOiZ0vLdX65uuuuu2zQoEHRs+Xnscces/vuuy96tuLxfROOf74/shU/9uM4zy+66KLomYiIiIiIiMjyk3X4SYD1xBNPWKdOnTzU69Onj40cOdJ+/vln/+F+2mmn2bPPPmt//fVX9InsERC+/PLL9tRTT9nQoUPt1FNP9T/CxHfeeccGDx7sy3766ac9+Hjttdc8XBw+fLgVFBREcymcz4EHHmg//PCD9e3bd1GgQEjKvJKmJy0jaTrWXnttXx5/L7zwgs2cOdMGDhyYuIxM3njjDbvwwgvtiy++iKaY/fbbb3bZZZd5yEvbP/fcc4tCy3TvR7rpBBhrrbWWh3vhb7PNNoteXXYIwWmX7t27ezAe0C7rrLOO77srrrjCLr/8cp9O2LLHHnvYSy+9ZGeddZZdffXVPp1jjLZlXl26dPEgtLRNmjTJ1+Occ86x0aNHW//+/e2rr76yG264IXpHbtj2eOD7X1OuXLlFYf3UqVP9PCotpT2/0sL30N9//x09W/GK8z0cx3d5cbbnv37si4iIiIiIyMon6/Bz+vTptu2223p1HiEUCI2+/fZbDy6p9iqun376yapUqWIVKlSwMmXK2O67724XXHCB//iePXu2HXHEER6YbbDBBl65SBhWvXp1q1ixor333ns+D95LMEn16JgxY7yCb7311vOgb6eddrK33347cXrSMpKmp2I96tev78FM0jIy4XObbrpp9KwQQSvzZB6ESQSrG2+8sb+W7v1IN531XR5hZypCEJa9+eabR1MKrbbaala2bFn/l3Zlf6NWrVpe8QnaLFSdES4fcsgh/j4qjr/88ssSHWvp3HrrrXb66afbnnvu6cvhWCSYXWONNaJ3mN17771elcq6PPLIIz6NarYzzjjD2rdvbwcddJBXRoPAm33BfOMIVrlwwHsJhUNAtf/++/tFBf49//zz7e677/a2oIr5119/9fc8+eSTXk1JxWy4MADmxWd4PxXT4f1xrAfv44LB448/7tMIofkc82S7QrVw0nSkmw/LI3Bv3bq1nXDCCfbxxx/7dvGYCyNcJCHQ5qLG119/bR07dvRz9KSTTiqyAjm1zVPnF0fFJhXFoY2ovqZNWV/aDlQdX3zxxb78o446yi8EpPPpp596W7KvOO+Q9FkuTtxzzz3Wtm1br0pOd5ykSteOIHA/5phjfBkcQwSq6doX6dqSKunDDjvMq5a5SMBxdMcdd/j7cfLJJ/v3U1H74aGHHvJzl4sQYL7Mk+9C1mHOnDk+Pd0xmXTsg+9T9hHbEs4VZNNmIiIiIiIiIsWVdfhJZSeBAvgBjf/973+28847W79+/bx7dnER+hAu0OWdH8+ERPygJgzlNQIBTJs2zZe9zz77+HPCgxdffNEfUw1GOEtAyI/5eOBXqVIl+/HHHxOnJy0j07LjJk+ebCNGjLAaNWokLiMTuvoSCMTR9ZdQ+NBDD/Vup4Qsq6++ur+W7v1Img8BNfMhPIyHIcsSYTEBUp06daIphQgKCYNpW0IkQkAQ3tFWhCzHH3+8de3a1adTSRtvTx4X1Z65IuwisI6rXLmyh06g6zp/dIemS++QIUPszTff9NfGjh3rASEVwnTV54IAn9t666096AkI7q688koP76jiJVglnAL7iBCVY/mDDz7w8JD5EcISgv/7779eiffoo4/6cYZQ/UjQxHnD+7faaisbNmyYTw8IjwlpOXcJNhnGgQAZzIPwjfVhe1hW0vSk+VDlR4jNezn+3nrrLZ8HNtxwQ79QQrBFSEZ4SKBL1S/HKSFrknRtTkVzfH6pCNxpB0I+vkduuukmb2/WFVRks04EpxxnXGD55Zdf/LU4AkYCTUI5huNAps/SXuwbzv2k4yTItD+YzvLYBubJRaWk9k1qSwLUevXq+XlFO1E5DSrJGUpk7733LnI/EIxyLt52223+nPOtdu3aXpnPMclxmnRMpjv2A7bv0ksv9c8RHnPsZjq3REREREREREpDzjc84oc5P2CPPfZY22uvvaKpJUOoQDdZggoCHB4TpoYKPyqgCAoIPugOvcUWW/h0Qh9+ePNDnH/5sQ+eh4pChC6pSdPD43TLSJr+xx9/eCjJH1VPVKtSGZVpGbkgyCHIoOs8QQmBAGOa5ioEUHTbp4qNwINweUVh3+6www4edFx77bVLVECCEJRQO4ROVJ3F0Z6p00qC+VGRVr58eX9OyB3262677ebBJKEO72FcV4IiKt8IKcF+32STTTyYrlmzZtrKSxCwsp09evTwCwkEbB999JG/xj5iPsyD45+KPFDdzLKokqU9CIyocGTcy1AZS1VwCG4J/1OXz7ozP4Y+oBqZauJ3333XXyPIX3/99b0Sl3OH9yZNT5oP6004euONN3rgTWVgOpwvtAHVghyPBGAffvhh9OrSMrV5koYNG3qoTLsRutOuhHWcSyB0Zx60f8+ePX1/cOEiFfue+RAABpk+y3Jpl2zWOakdQbszLbR7UvtmakuO43CBhosxfB9x8YN5cMGBcyeX/QDmGY6x7bbbrshjMgnbwsUpvh85VplPcfaziIiIiIiISC5yCj+peGJsRKryqB4qLQR8/ACnYohqwP79+3tXcqrq+LHOWIwERfxgpwIpoAqQzxBMhGpREEbFu3JSucR7k6YnLSPTsuNjflIJxfica665ZuIycrXRRht5BRf/EojwmKrCXBHItWnTxgMHggeCmvjNpJY3AlyqP+kOT8BDdS8VfYyrSnUa3d9POeUUr5QktKEbP9WfAe1JG5cW2qVatWqLgsiqVasu2q+hyz5hFBXAhPP8UQFHl2pwnGaDAIlu9WEehL+33HKLv0aQFBfvbg/aguXTHhwHDRo0iF4xrwrkuMskHsYTsBL4pmKZ6ULl+PR089l+++29PQjaqJQMlbypWEeOZd7D9vNvpjA/U5snibdjahuCfRCqb/mjinPXXXeNXl2M9UyV6bO8hmzXOWl/xKez/vyla99MbcnxEEfQymtc9OCiQq77AXz/pB5jmY7JJGHYjrji7GcRERERERGRXGQdflI5xFhwdJ2mKozutaUZolE9RNd38IOavy233NLH96N66ZprrvEf4ano/kq3Y8KBEEQR8DH+J2PQcRdxKokInpKmJy2jqGWnk7SMXBEmcAMjur7TvZZ5MsRArhiSgLsssx1UWNHFlIrGFYWqvFCtS5j7zTffeDUtAXYYVoEqMoJQwhKOtzC0AV1+CSoJTEsTYSvBFpW2oILtzjvvXBRiM9YrFXJUEFKNzF3hiwqiU4NEKuY++eQTryxl2xlTNGxvUUIIzDAChLMEtemCynS4iRTdp6k+pk0ZuiGE+FxcoNKO7eV8zjQ9aT6ct2wHwypwUYT9mIp1JZhk3zMPUInMhZQkmdo8221PRaV6GGOTfc34mOyLbGTz2WyOk0z7g3OTaVR20u58b6Rr31zakvCTm4VxDDGWbi6fzdTOmY7JXPZPUpvxnO9PJD0WERERERERyUbW4We4wQc/mumeyB/VRKWBrpx0ieSHL11OGbOPGytRBUjXT7p805Uz/FGFGfDjnjCCu7wHVOwxbiRj1zEvqlSpiEqanrSMopadTtIyckX3Z8blo0qSIQC4+Upxhhng8+uuu65/nqpaqkBD2LIicNwwxiFVn9xQhRuzEFozViDVu0znMd18qa5laAOCZ9qAIIjQsLRRMcyYm5dccol3DabdCVk5Bqm+o8KP6bQdYyQSRNF1PwkXB6jmY9zJgGOA4SJYDscqARH7IxuE3oxByrAKBGAE7PEbEWXC+cS5xTozZiTHJl2OwXzp+syNZgjleG/S9KT5MEwB3wnsN4YxSK0I56IElYWMXUnVIje0YZ8SeBHKJ0lq8/j8csW5wAUWLpgwJiU33UlXIZpONp/N5jjJtD9473HHHefBKtWUdDVPat9s25LK77lz5/p6Bdl8lgskHP9Jko7JdMd+JkltRijMcCJIeiwiIiIiIiKSjTILf6gu1Qc2jLUpsirbecTSN+H6qGnm8Pq/gnFsqTIk5IxLmi6yKghDEIiIiIiIiIgEOY35KcUzffp0u/XWW9P+xccHFRERERERERERkdKjyk/JW6r8TMY5zpiVDIkQlzRdZFWgyk8RERERERFJpcpPkf8gQqJ0AWfSdBERERERERGRVZHCTxEREREREREREclLCj9FREREREREREQkLyn8FBERERERERERkbyk8FNERERERERERETyksJPERERERERERERyUtlChaKHovklRqDW0WPFpt08JDokYiIiIiIiIiI5DtVfoqIiIiIiIiIiEheUvgpIiIiIiIiIiIieUnhp4iIiIiIiIiIiOQlhZ8iIiIiIiIiIiKSlxR+ioiIiIiIiIiISF7KOvz8559/7PLLL7fdd9/d6tatazfffHPG6bm64447bN9997XGjRtb+/bt7csvv4xeMbv33ntt//33t3r16tl1113n01jO1Vdf7Y+DV155xY488kh/PHDgQGvatKl/7uWXX/ZpSJqebhncCJ9ta9Soke2zzz726KOP+vQ///zTqlevbvXr1/c/tv2MM86w+fPn++tJy8hk3rx5i5YbZJpPuvcHt912mz322GPRM7MbbrjBGjRoYPvtt58NGzYsmrp8vPvuu/bcc89Fz8x+//13O/XUU709DzjgAHvrrbd8+qxZs6xjx47WsGFDnz5+/HifzvHVo0cPa9KkiR122GE2efJkn76iHH300VanTh3f73vvvbf/+8ADD0Svltxrr71mF1xwQfQsvRdffNG++uorf3zIIYfYd999549XRVdcccVSx/bcuXOtX79+0bPcfPrppzZy5Eh/zL76+uuv/XGuOId22WUX3798J+y111523nnn2d9//x29o/QlbXeu7XHaaafZ+++/Hz1b8Vb1Y1RERERERERWbVmHnwRYjzzyiHXu3Nl/zN555502YsSIxOm5GD58uAc6Q4YM8eDirLPOshNOOMHDxLffftueeeYZe/bZZ23o0KH++qhRo6x169Ye5BFQBsynZcuW9sMPP9jdd99tTz/9tD388MN21VVX+bySpictg3+nTZvm//IZwkZCOqy99to2duxY/yN0nTFjhj3xxBOJy8jk9ddft3POOcc+//zzaIplnE+69wdsS58+fTwcBdtDkMi/Dz30kM/n559/9teWpQULFtjgwYPtkksusZkzZ0ZTzfr372/rrruuh3w9e/a0iy66yKffc889tueee9ro0aPt/PPPt+7du/t0gjHaljY+/fTTPQhd0TjG2e9vvvmm75tevXotsY3LGsv+5ptv/HHv3r1tk0028cfL219//RU9Kl0cO/HAPBeE4yFQL6njjz/e2/qNN97w43LChAn2wgsvRK+WLtoyabtL0h7FxfdqaQW9K/IYFREREREREck6/CRsqVWrlp1yyikeQmHixImJ03Px008/2ZZbbmkVKlSwMmXK2B577GHdunXzH98EdUcddZSts846tsEGG3gVFmFYjRo1bMMNN/RAArz31VdftebNm3twSQXpeuutZ5tvvrn973//s3HjxiVOT1rG+uuv79u12mqr+Y93/tIFAhUrVvSq1SlTpiQuIxMCm8022yx6VijTfNK9H7/99ptdf/31XvEWUAFG9ShtW6VKFW9bAp1ljcCGNqxUqVI0pRBtWbZsWf+XEJT9jR122MEOPfRQf7zzzjvbH3/84Y8JtNu0aePvo/qTwDcEuysDKoDZxl9++cWrWqkO5FhgWwjN8NFHH1mnTp28opVzZM6cOT6dQJppVOReeumlXlEcR5A/e/Zsf8x+pKKPeRIE33jjjV79yfx+/PFHnycVtcyLz4XKWSoYL7vsMmvVqpW/9uSTT/r0uO+//96rbqlwZt1DdTDH00knneSVxxdeeKF/HlRXE/oeccQR/pyQnopdzr0HH3zQpyHddLaRwJtlMZ8Q4sYR0HOMU7GctF2sG23PvwHn8V133eXtyoUUUD3dtm1b3/733nvPp02dOtXPd7b1uOOOy6oqkWOWcydcOOC7plmzZr5eBPyEl1Srs01cBGK7u3Tpsmhfsy28l3ORi0VgX9IWXOjh4kt8u+OyaY9Uzz//vB100EH+HcK8kXR8xlFRT7jPxSWWmW47CUavvfbaRdvDxSEktWs4Ro855phFy2Q+XOxge5KOHxEREREREZHSkHX4STUmoQII5rDrrrsmTs/FgQce6F07+aFOV3Z+cBMKEdjxGsEM6MLKMvhxDX6gh0osKh63224723jjjf1HN4FhULlyZa+kTJqetAzCDrq7DhgwwIMeQqB0oeOkSZPspZdespo1ayYuI5MOHTp4l+64TPNJ937QRf/MM8/0MDbYZpttbMyYMV41SgBBOxE2L2uEyCeffLK3XxwhICEu3fAJiaj+BGEM28hzQhLCbxDMxQNU2qSo9lxeCMKpTCWYZd8T4my00UYeBt1666129tlneyhK5RsBLsfItttu60Em+4Iu34SIzIPnHGdFIZgndKI6ln0b3H777R5uE4wSTHFehspMwlcCMIZCYL1S8Rm68HPcEx727dvXp/OYrt+sH+H7r7/+6tPBPqBSmgpeqiI5D1l/phGuJ02nOpptZZm33HJL2gCOsJZgs2vXronbRXDOcc2/ARdDCIhbtGjhYSc47lk+lelUPoMAkKpO1pHziHOmKJ999pmHp7Q/AShhNV3R2adrrLGGrycII6laZtv5riBIpP2oRiVUppKbyuwvvvjC3892UTlMQBvf7rhs2iMV32MEoAStVEvzfZJ0fKZiXQmN+S5Nt51sB8MLcGGCYyVUYxfVruyX0CuAY4HvV86FdMeJiIiIiIiISGnJ+YZHhCiMSUhlD4FJkDQ9G4QW/Ein6qdatWreBZpgLFT4UWnUv39/nzeVjfz4B5VP/Cj/999//V8qncDzUFGI0IUzaXp4nG4ZYHzKdu3aeYjx7bff+jQqE8OYn1Ru8UOeqqdMy8hFrvOhizmVqiEYDggh2BaC2xNPPNGrdMuVKxe9uvwR/Oy4444edFBldvHFFy8R3rC+7PsQJjHmZxztkDpteaOSjf1OgEvAy3FPMESoS0BG0EawyXZRncnxQwUmx/XBBx/s++iDDz7wYHHTTTf1KliOXYLp4iKwYt7YaaedfB8zZAOo3GP9CONYVioqhQmpCWIJakMFKt36+SzCvAOq9MByCQMZ85awi4CUNsg0nf27+uqre9jNeLmZJG0X58YWW2yxxDmSTlhPjnvWgfP2448/9iCYIH7QoEGJ42Mylms4xzkuCT65wMLn2bcEg+zrDz/8cNE8qGDfbbfd/DGfYZ+yDazHWmut5ZXc7P933nnH38P2p7ugkiSpPVIR/oY2Yn1Yx6TjMxUV1qxr0nYSWHPBiH3I9zVDlmTTrlzQClWifIbjIOk4ERERERERESktOYWfYYxGqvKoMgySpmeLUIgfyvyQpuqPrrkEolQQEXRRzUVYxLh33GwmoAqQz/Cjnh/V/LgGgRKVTgFVarw3aXrSMqhApUKLkIbKPUIEAiHEx/ykYu7KK6+0NddcM3EZucp1PoSftAEhJ5VahIwEbgQU7B/GCWXbypcv79uzolD5RXUY1ZxUghEG0ZWdm0nRhXn77bf3IISxVadPn+7BENWfAe1A26xI8TE/CXE49gmmaVuq8Wh//tgfhFuEixzjBEp0Waa6jRA3HtwRThJ4J4l3704n0/yows2EikCOFSr9GOYhYJvCPFOXzxARIHzjgkHYZrqbE5QlTSd8ja9nvHIznVzbKRXbFMc5yjQutLBe/BuGk0gVxvzkj4pO/mWYDfYzN3cL20aQTxsivq6ce+GCRep0tguhHbNVnPZgeWx30vGZKqxT0nayTcwv4FxlnYpqV17nvCckpT0Z+iDpOBEREREREREpLVmHn/xQvv/++73LLRWGdKlkjLuk6bkifGFMOhB6MRbgVltt5V3q+XF/0003eVCWiioobkREt+PQ3Zvu8wSXdHnlhzmVRPyIT5qetAyqPKlApBKO7qHcuZywNZOkZeQq1/ncd999HhZTnUrlFeMN0iWW7ql0qaWCizH5Pvnkk5wrc0sTY7WGal2qzqhaozKVIPGpp57y97DOhG2MscpxxTEF3kP3X4ZDWFlw3HHMMDYnlYFUT4Jxbwl32U6GKWAfEqYRhBIKUfVJ4M8QBARahD6pQwQw3zB+bhhaIkitfuWzoZ3oos1Yipw/2SDQI/Skmo9gKgR2jMnI8QS6OKfDNtNFmjFeqdRm+AL2a9J05kn4S7sQZKfr9o6wfUnbxee5OBBCxLhMlcGErwwXQGgNjjnC66IQ0rLPGMeWgJ6hBOhaznowFmY4djmeqZAE1fBsL38sj3bluOZiSfwiTlzSuhfVHqnoRk7bsL4cZwxbkHR8JknaTtaBru9MY3vZt9m2K8ukmz/fQVwMSDpOeM45g6THIiIiIiIiItnIOvyk6gn8uKV7In/8AE6anguCOrqN002ULqaEd1TI8eOb7qFULfJ6+KNrZUB3UsLW0OUdVBcxT7qhMy/G4KOaKWl60jK4eQkBFBWlhKx8rnbt2tFS0ktaRq5Kaz50Madila6s5557roe58aqt5Y0AhUpPbnBCl3Fu3ENozbiGBOlM5072rCfVtQxtQKhOF1nGKSQ4WdnQ1Znuu+wnxq6lrQkTCeWpzGM64zVyHBH2sd1Ur7JPjz32WJ/OWIwcb3Hsf7onEwDHKy8J4ajgi99YjPOOcJt5cXMi2o9wKRtU2rGuHOMEkoRaAwYM8PCcbtsEVoS76S4+UDlIaMr5x7IZG5djLmk628i2UqHMOjPvVOxvgjvG/03aLrpHE5zR3nGE0VQXZ7o7OvOlMpELDARvHHvZYD/THpyHVHqzf6heJJhkP4Ibd7GObB8XcFh/hg7gO4XvKirI+Rzd51PFtzsum/ZIxQUC2p6xjKn8ps2Tjs8kSdvJWMtbb721rwPz55xkPtm0K8cEbcg8kHScEBAznAiSHouIiIiIiIhko8zCH9VLl06J5IEagwtvehM36eDCu4BL0Qiwqlat6pWLhLaMiUv3/ZUFQRthXKYAb3niIgx3Zg/VlSIiIiIiIiKy4in8XA6oAOMu1+m0b9++WGOCStEUfpYMlaVU5tLNmG7NVAoWNezD8sINdhgGgYrElYXCTxEREREREZGVj8JPyVsKP2V5YkxPxuDkZm0iIiIiIiIisnLI6W7vIiKSHt3vFXyKiIiIiIiIrFwUfoqIiIiIiIiIiEheUvgpIiIiIiIiIiIieUnhp4iIiIiIiIiIiOQlhZ8iIiIiIiIiIiKSlxR+ioiIiIiIiIiISF5S+CkiIiIiIiIiIiJ5qUzBQtHjRebNmxc9Ell17TziyOjRYh81fTJ6JCL5pnz58tEjERERERERkUKq/BQREREREREREZG8pPBTRERERERERERE8pLCTxEREREREREREclLCj9FREREREREREQkLyn8FBERERERERERkbyUdfj5zz//2FVXXWX16tWzRo0a2e233+7T//33X7v66qutYcOG/tevXz+fXlp69+5tzZo1s+bNm9uJJ55oEydOjF4xe+CBB6xVq1bWpEkTu/nmm30a63Xdddf54+DVV1+1Dh06+ONnnnnGWrZs6Z975ZVXfBqSpqdbBtJN//PPP23HHXe0/fbbz/9oq/POO8/mz5/vryctI8n48ePtyCOP9Hndf//90dRC3JE/vj5xd911lz355OK7mmeaz7L2/vvv2/PPPx89M/v999/t7LPPtqZNm9rBBx/s64ZZs2ZZ586d7YADDvDpEyZM8Okcd7169bIWLVpYu3btbMqUKT59RTrzzDPt1FNPjZ4Vos0ffvjh6Fl+OProo23GjBnRs8V69uxZ5PGbzXuKY+7cuYvamWPnoosu8sfZ+Oyzz2zUqFHRs8Xi0zt16mTTpk3zx8vSjTfe6OfF66+/bt26dYumFs+vv/5qRx11lA0YMCCaUii+XRyfgwYN8scrk/j+FBEREREREVkWsg4/+aH+xBNPeDhAeNenTx8bOXKkPfroo/6j+oILLrAGDRr4j/qvvvoq+lTJEJ68/PLL9tRTT9nQoUM9cOKPMPGdd96xwYMH+zo9/fTT/iP/tdde83Bx+PDhVlBQEM2lcD4HHnig/fDDD9a3b1977LHH7L777vOQlHklTU9aRtJ0rL322r48/l544QWbOXOmDRw4MHEZSX777Te77LLLPMyl7Z977jn7/vvv/bU33njDLrzwQvviiy/8eRzrRjAb5p1pPsvSggULfPu7d+9u3333XTTVfPvXWWcd30dXXHGFXX755T6dUHaPPfawl156yc466ywP1MExRhsyry5dungQujIg1GUbZPniuIqH6bmYOnWqnx+pkqYvSyeccII1btw4elYy7777ru2www7Wtm3baEqhFbFduUran3/99Vf0SERERERERKRksg4/p0+fbttuu61X5xFCYdKkSbb55pvbaaedZgcddJCHn/j777/935L66aefrEqVKlahQgUrU6aM7b777h6yMv/Zs2fbEUcc4UHaBhtsYHXr1vWQrHr16laxYkV77733fB68l2CS6tExY8bYvvvua+utt55tttlmttNOO9nbb7+dOD1pGUnTU7Ee9evX9xAiaRlJCFT5LO8tV66cB6gbb7yxv8b8Nt10U38cR9B5yy23LBGCZJrPskSoQZtwfMStttpqVrZsWf+X9mO/olatWl7xCdqGKloQIh9yyCH+PiqOv/zyS696XdGo/iTop81TEeCyLZwTPXr0WLQtcfvvv79fSODf888/3+6++27/DBXKVPKB6l0qXqmSDaE/mC+f4f1UQ4f3xxEW8zqhP/MGbUc1J5/nPA6hdJs2bfyYxocffmjnnHOOPw5YfwJ05sX68V2QKuk9VPpefPHFfv5RnThu3DifTiXilVdeaYcffrivD8flN998Y61bt/bXceedd9qDDz4YPSvERQOO/1tvvdWfs95UEvM5KtNBuM/2sS4sl5D6559/9gs27BsupATpplOh3b59e1+3Dz74wKd9/fXX1rFjR5/fSSedlPYCQrr30PZ33HFH9A6zk08+2b8LHnroIa9Ij2OfsVyOd87jVHPmzPFtpb3YZ1RHcy5wYWP06NG+3kG67fr000/9eOHiFe2NpP0T98svv/jxzrF4zDHH+HaCfcC6cMHp8ccf92mcr1T9Bvy3gmp9qnTPOOMM3z4+E/ZVfH+ybI6hU045xS8uiYiIiIiIiJSGrMNPAs7w45of7/jf//7n4Q0/qKlmpMrvsMMO8yCrNDBvfpzT5Z0fyfzAJwgiDOU1foiDbqqs0z777OPP+TH+4osv+mMqnwhtCfwIIwgBg0qVKtmPP/6YOD1pGZmWHTd58mQbMWKE1ahRI3EZSehuTPh76KGHend1Ao7VV1/dXyPAiodEAdWSBAcEskGm+SxLrAPHRZ06daIphQg/CH1pQ8IWjhkQ1NEm119/vR1//PHWtWtXn07FbLzdeJyp3ZYX9inrHEK4gHUj/OF8GDJkiLd9uu7G7BfCII5TAjYq3Qh8CPsJrBlO4tlnn/XKao4hhCo+wiLOCd6/1VZb2bBhw3x6QKjJkAgEbMyD8IvAipCLUI6QjeOVts4GQRnbwec4D996663olcWS3kO184YbbujrwPK4eEGYhk8++cSrp6lUJgzdcsstPRjnogpoGwK3OCqet956aw8BQah26aWX+nYSntE2tN9ee+3ln6cdGIqDdSCIY1iF+DzTTSe4pys24TTtD9qOUJcglXOP8DlVuvcwX6qXQVBOELj33nv781TXXHON3XTTTX68EASnHucMAbLFFlt4GxMcc46sueaadvrpp/vwG5zjQbrt+vjjj+2ee+6xe++91+eFTPsnILzdfvvt/T2ErnyWkJO2578JVHOzDwnXMxk7dqzvD45bglv2c+r+ZN9REc5yREREREREREpDzjc84ocrYcOxxx7rAUOwyy67eIUhQQwVRqWBH+V0h6aCiZCHx4xdGSr/6NrOD29+5BP8EQyAYIjAiACJfwkAwPNQaQg+T2Vo0vTwON0ykqb/8ccfHjLyR7Ue1apUiWZaRjoEMAwfQIhGwPDmm28uVSkWRyBCVWlqCJvrfJY19iFddAmXrr32Wg8/411cCRQJr0M4w5ifcbRb6rQVhSCRkC9UB+Kjjz6ynXfe2StzqW6l+jCMXxrHsc2xQRDNsU3lHahcpsKPz9IGBHoEg4zfGCpIqeClihgE+6mVn3SDZt5UHhMmMjTD+uuvb99+++2i6mzOkWy7RBNyUS3IuhJQ77rrrtEriyW9h6Cb9uHiCaEw+5qLAuAcWWONNTzQZnvBucpQF1Sgck6lq3COYzu5sMG5RVvQdnxHELpzfnLcp6u8zSTsi9C2nNN8p1GJy/lO0Mr6xSW9h5Ccc5+2J9gj9Geb0yEUveSSS3yYD86LTTbZJHqlEG0cgkzGFuY4ICTNVmhv9k+Qaf8EHOMcxyBgJZzk2KGd1lprLf/e4buf4y4T9hXbxDFSs2bNtBXLHDfxix0iUpqmW99RrazG4IV/owbaih9BW0RERERk+cgp/KQyiDEaqTiiyglUAPFjnbCBH8V0zQ2VoSVFcEGAQGUQVYL9+/f3QIcKIgIwuucSNhE41K5dO/pUYXUgn+GHfagWBT+8491VqazivUnTk5aRadnxMT+pPKMbJ9VZSctIstFGG/kNk/iXcIHHoSIuHcJPtpUAimCRkJHQLNf5LGsEr1R/0h2eoIsqXsYupQKQyjiqhqlepdsugRHhF9WfAe2WGgqtKAQ/XAjguI+H5fGQm7CJ8CtVCPuC1ECM7ae6mDZgn4XQEgwXwDGVhOMzPj/alfMyvl6pyw/SdePnvfHPsvxUSe8pX768VylyAYM/qgdDMEogmyqEn4wDSRBelHRDOFDZyLHPa1Qf54pzJY62ZhrzZRv4N/UCQqb3EBLymIpQgv0kfL9ywYTjnaErqGKNSz22CBHTHVtJUrcLmfZPED+eWIdQkZq6LryWihsaBXx3FyXdsSWyyvhtoLUmWIz9XbD0iDgrzswBdu2cxtb34CE2qdHhVi2aXFJTvjzLt7X1lylDosy8eYm2iL8ePlP4d7MV1sdHUtpxpWpDEREREVklZR1+UrVEN1rGXSS0oMKTLpxUFTKuIcHkbbfd5u+leq20EGLQ9R0EQfzRPZax7PjhT1dRQr1UVEjRxZLKq/Cjm7vRM/4nQRB3F6fiac8990ycnrSMopadTtIykhB2cWMjuhLTDZXPUlGYhLEFCVfoak24yA1VqObKdT7LGsdGqMolhKVyjQo/guowrAJVjoRwhFccb2EIAyrQqlWr5oHpyoLxXukSTHUlaFsCe9qbMIgwiRs55SoEv4R3VatW9aCdECobhFe0J1WQVCQyjADHHBV/XDgA528I7TmGQyDOsZ2K93Fcsc8I5tJ1e096D9XhYTxIvisYFoP3JOFYIEjkGKFSMZ2i2oHxfhnzkypLqjHjFdZJn800T4JdjltCWbCvw426gkzvIfzk4gT7lPFs0yEk5MIF82B8TSo72edxXGAK58Lnn3/u5wjHRiZFtVU2+4d9G44LtoPvPY5ptpW2ZT244MX7OJZC5ShDO6RuQzrp1pGKdY7ZTI9FVioEfa8+ZNvVGWKTCBf971Kz8SnB3oq2XtVSCz0X/lfZLhjcys62o+2ipf5XaOFr47+2ixpHbdG448Ivrhusr19fe8t6z2hgI6J26rvFq9Z5QvjvysLPLWxHq9W78HN1Gtug8WdFnxMRERERKZ6sw89wIwx+8DKWHX+Ebdzggwotwjdeo8snoUNpIMCjq2S7du08COFmIYQaVAfSxZIu3HTzDn9UYQYEDvyYD901QbUhQRDrzLyoXqXSKGl60jKKWnY6SctIQldoxg8kyKQbL8FIfJiBbJXWfEoLxw1jA1Lhx13duXs74TRj/1G5ynQeczMhqmjpnk2gwroThFNlubJhm0LARmBJ12WOXc4Lurdz45xcEaJWrlzZh0yg0prwnDA/G1wcYBxIKkepNOR8JHQjtGLMR9aLsOqiiy7y93Nc0uWZG9HEK/UC1p/t4NhhXUK38Lik93DccfGCixGM68iNblKrXFOxrVS70mU/FRdeCJUZVzRJp06dvK1oO6oUCfMYR5MLIVRgpt5dPGl6HN3QH3nkET8e6dIe2i4u6T108aZdM30vrrvuur5fGCuU9WbM3NT3UxHN2K28j4pyzhGqj5Nks13Z7B+q3Bm6geOD7eN45/uY72a6wTMGMccQwwQQinI+c9zx3wLGhc4kaX9yfFIFm+mxyMqDoO9Va1NniF2/eFSJhfay6w8+15r4Y4LCs6zvl6EaMoSisa7oS0xH4WtLVD4Ssi7qss7rBIOFIWTh59MHhV5puXAdbc5D1pT3LQobMy0/aZ0Dtm+IPbdt4bA/S/jtG/vCtrIa4VplhS1tu+ihfy5WeVqtQnV/v2/TzLE2yBrbWdtW8desUlu7aL3JNnHp/zSJiIiIiGStzMIfnkv1VVwZ7qYtUlI7jzgyerTYR00zh9Sy4lHtSnCWerMskaLQjV9kuSOQHG/Wd1HQmQ5B4tUe7C1+H8HjKXZthUttUu3Ci5KElE0/3yp6T+HrE7eLhaos64uqNsLDw+jzc6rbRY1vs84VUj+fYonPoqjlp1vndArn83zl3vZcCC0XGjmhlXX+ls+2tUlpXi9UuIwvahW+5sunKjRlHdN/VkREREQkOzmN+Smla/r06X638HR/8fFBRf4rqPCl67uCTxFZVXmVZdpKSrM2dWIh4m/j7HnG4IyCR1Tb9mhrY1/bpBy6ebepUxh8IqfPZ7n8JdY5B01qR13aB59i11pHuzUeXhLEevtcbYO2uDRDsFllcfWoiIiIiEgxKfxcgapUqeJdTdP96Y7H8l/E8BGMISwisqqqtu1ti8arzGjuNPs4erjYFlazxN28s/z8Mls+Crviv1i5cFzPEZXHWNN4l/xK5xa2Ea9VeGypkHix6TkFwSIiIiIi6Sj8FBERESmuSvWtjb1qL+Z6V/J1q9rSt0D71ibOqW41142eFkuWn19my6f69TEbtF5HOyXqrl9t264+dufzM1PuCL9QtUoNFq5HYbVptQpbmc2ZFo1puth2FdTlXURERESKT+GniIiISLHtZafUqm6DxqfcnKgoFepay/XidzoPoWEDa+pdvQu7fA+aEb3+20BrzU2LUsTvhr7k54tQ5PKLb6kQ07vYRyEm2xFf5swx9nG4OVJqkDxzgF07p7EduMSNpEREREREcqMbHkne0g2PRP5bdMMjWaEI9V59aImu5DtFN/IJN/axpe4IH24qFFmvY+xmP4i/3tj61jHrnHLDo4nbXWo2PvaepJsTLXXDI2RaftI6BymfjYRtLrzhUTRxocVtEW7UVDidStVwwya3RDumvCYiIiIiUgwKPyVvKfwU+W9R+Cn/LSH8TAonRUREREQE6vYuIiIiIiIiIiIieUnhp4iIiIiIiIiIiOSltN3eRfJBjcGtokciIiK5mXTwkOiRiIiIiIisylT5KSIiIpJCF9BERERERPKDwk8RERERERERERHJSwo/RUREREREREREJC8p/BQRERFJoTE/RURERETyg254JHkr3Xht+jErIiIiIiIiIvLfocpPERERERERERERyUtZh5///POPXX755bb77rtb3bp17eabb45eKTRjxgw7+eST7eqrr46mlI477rjD9t13X2vcuLG1b9/evvzyy+gVs3vvvdf2339/q1evnl133XU+jfVKXYdXXnnFjjzySH88cOBAa9q0qX/u5Zdf9mlImp5uGUg3/c8//7Tq1atb/fr1/Y+2OuOMM2z+/Pn+etIykrz11lt2yCGH+Lz69OkTTV3stttus8ceeyx6lvz+G264wRo0aGD77befDRs2LJq6fLz77rv23HPPRc/Mfv/9dzv11FNtn332sQMOOMDXGbNmzbKOHTtaw4YNffr48eN9Osddjx49rEmTJnbYYYfZ5MmTfXpxhX1z7rnnRlNK19y5c61fv37+mG1bVstZGcybN8/3S67ibXTNNdcscXyUpqOPPtq+/vrr6FluSvLZ2bNnW8uWLaNn/2/vTMBtqt4wvjQnNA+khExFmiSNlIRChTQaU2n4G4qIUqikUQOVUmiQoTIVEkoDSRGVMWVMSoZSIvfv9929ru109rnncE2n9/c857n7rLP3XsNea99nv+v91t4M/W5rGTlyZMr31nHjxrl27drZNsfPnTvXtpMpB/cP7mewLeUWQgghhBBCCCF2NkmLnwgUr776qmvatKkJbM8884wbM2ZM8KuzB3MetidOnBikbDujR4+2h/Zhw4a5sWPHuubNm7smTZqYmPj555+7t99+273zzjvu3Xfftd/Hjx/vatWqZQJfOJqf8yBG/Pzzz65Hjx7urbfecv369XOdO3e2c0WlR+URlQ777ruv++STT+yD6Ioo3L9//8g8olizZo2766673HPPPWftzHE//fRT8KuzMiBQIEBB1P6UDyGRv3369LF8V6xYYcdsT/766y83dOhQd/fdd7ulS5cGqc717dvX5cmTx3300UfugQcecG3btrV0yn3GGWe4Dz/80N15552uY8eOlo5ITBvSlrfddpsJoduCvzZh8Z6+smHDhuDbtkG9t5eYt6ux//772z0hVcJtxIQJEwIi52GSirEE9PmFCxfadjIwUZNTY0IIIYQQQgghhNiZJC1+8uBcsmRJ16xZMxOhYM6cOfb3448/Nkcebr6c5JdffnHHHHOMy5s3r8uVK5c7/fTTXfv27e2hHAHvqquucgcccIA76KCDzIGJSFa0aFF3yCGHuClTptg52BdRtlq1aiZQ4iDNly+fO+qoo9xJJ51kYm1UelQeUemxHHzwweZanT9/fmQeUSD2cSz7IjIhAB9++OH2G0Jnt27dzJ3midp/6tSpJi7RhgULFrQ2RAjZ3iBw0Sb58+cPUjLZY4893H777Wd/EUG5rnDCCSe4yy+/3LbLlCnj1q1bZ9sI17Vr17b9cBnOnDkzS/DdVhBAEfERzOm/CMS4TnHIdujQIcv5hmuWtubv//73P/fUU0/ZfriJV61aZft4EJc5F25bwAWI07Vq1armnPYghONypV++8sorQepmEGS7du1qZeH6IbDDrFmzrD1IxynrheVkypjdPvQryuNBkEaoB/LjGPbHgc3+69evd9ddd539/t1339n1o05MUnCNcO3SjhUrVrQx4tsk3EYvvvhi1iRKVPvHy5v2efTRR21Sg8kYJiLigYh35ZVXuho1arivvvrK0n744Qcbv4yXhg0bbjGpECbesV988YW75JJLrJ789W5KJiO4xpTx2WeftbRYaA8E/5o1a1q7rV692tJxb9O3uX/ecMMNWZMiTFbQdtSRe1gsnMuLz9wX/bWYMGGCa9SokU160Eaffvqp3R/YprxR5fC89NJLNna510PU/pSpSpUqdn2Y5KA/hMH5TH+jT3H9GWstWrSw/uwdqUA/4Bykb42YLoQQQgghhBBCJCJp8RNBA3ECvMvx5JNPtgfeTp06maCx9957W3pOgZhAiCyioXeWIi4g5PEb4g8QnkqZEDMAMWvEiBG2jShRokQJd9hhh5nIgTjoKVCggDkyo9Kj8kiUd5h58+a5UaNGuWLFikXmEcXixYvd8uXLXfXq1S3s9LHHHnN77rmn/YaIhhiEuOqJ2v/44483MQRBhd9pD0Tl7Q2iMK6+8uXLBymZIMog+iLAIVohqgDCGW3C92uvvdbEHFi2bNkWAiptmKjdsoPrxAcRERBoELloy/vuu8/EPtymtNWAAQNsH9qW8HWuMyIYfZ7reuyxx5qoFOaee+6xpQ9at25t36kr4wM3MqIzQjiuVxyu9FHyQLiLFaTZ/9tvvzXxt1evXlmO1zvuuMPdeuutWWK3b79kyphKPWKh3Ihw7F+4cGE3fPjw4JdMELUQwBij++yzj/2OIIpQiXDLZ+DAgSaKx7YR0N5R7R8vb5a/QDTFXczSGLRnPOj3nAfHOmIi0IaNGze268BSCoyleMQ7lr5C3yQ/+i8Oa0Ra7o+E8FNGL9rGQr/F3UyZaXOu/8aNG+2cgwcPtkkkYIzSP1laA4c5v9EXYkEs9RMoM2bMcAsWLLBriujJ+PIgPCIu4gLlfhCvHGFw1zMWe/bsad/j7c8EEAI1yxdQ57322svE9FgQW3G+I/Ay2VC3bl3b/7PPPrP+yLVmDNLfaUvc7H5sCiGEEEIIIYQQOUHKLzziYbxNmzbmmKpQoYLr3bu3O+WUU1y5cuWCPXIOHJw4gXDGIXrgRENw8M4/3F+EUVMWnJA4G8ELJQgL/MWhBXz3TkPw4c5R6X47Xh5R6TgWER/5ICLgtMRlliiPeCASIfAgBnnhDIEKAeLAAw/8l9gatT/iDmVjnVFcZbh3cYbuLBA3TjzxRBOVEENwgCHYeCgv19gLKbjOwtBusWmpgODFB0EccNwh1k2bNs2VLVvWHXHEEeZKpc8gQsGhhx5q1xEx+bjjjstyR+Iy9q7KKBgXCO9ce9qe/RF7EI9YDxbhjTTvLPTg1kNkJ0/6Pss/4MxctGiRuQ4BR2IqZdyWetBnmIQA6rFy5UrbBoRKvlNX2g4RkGtYunRpE9wQ73Adsk/4WodJ1P7x8qZNGGss9YDwxxiMh6+jb3uOYX/EOIR4yoY7Oh6xxwKTPJwDYRIREKETJykucO6DwP0nHrlz5zYnJ5QqVcrOSV25hzJeWcP3m2++sXNSd+qMKM9SGozfWBAkaTecmLQR7YdIijs1LH7GEq8ciYi3P23ItUQARRz++uuv47YjZeR+RX+jLtwXfZ+m3IwF2pkxiCue+xrlF0IIIYQQQgghcoqUxE9cZqzRiPPJh/DigMLRhZMLlxoP3+GQxm0BsZMHah6UcQO++eabJogiXiGA4YDj4Z/1A8PiK+5AjsEVheMMtyggrIRDXHE0sW9UelQeifIOr/mJww7XH47YqDyiQKhCwOAvogAuL5YZQPykToghuK4QExFNovZHaOC64SqjrAgZuLp2Frj1cN3h5sSlSlkJZX/ttddM3ENcQRTkBUgIfUceeaS5Pz20G22ZUyBaAaJqWJzGyYZgDQhUYfgtWRA+Y0GoQjTn+vFhiQIEpDAI42EnNe1BWriM4XIlU8ZU68H18LBEQZSzO7asiIMIZLj7CJ2mHevUqWNjMopE7R8vb8QyRH7CrunbhGT7/cPEtj/nIY0JFdqev1HLT8S7dvRdxjX91L9EjXYNl53yxgMRMLYeiOAsY0BfZ8x6kTH2nLjdY6EfFSpUyARcXq6G0IhoznjBbR5FvHIkIt7+3EdYU9T3YSYzWMohFu4/nqj+Fq4n+9MXhBBCCCGEEEKInCJp8RPBjbXgCJ/kYdiHnhLC+sYbb9gHERCHz4033hgcte0g7PGGcEAgYO1RHvgJwUfsILwbAS0WxBDewo4I4MPDcVIRlks4K2ISbjse4KPSo/LILu94ROURBSII4eqEqON0Q3DB2cUaiYi/hIviXLvpppss5DZqf8Ro1i3EpYVDDWcZjt2dBS5D78olJJZQXZypCGWI6ECZEd5Ys5T+5kOs2QeRPZ4QtK3QVgjttB/iC4JkbMh+smTnTCUMmVBn3Lq4mAn592tHesgbcY92oo3Yh3ojXHOdgTKGhfdtBUGRvsl4w33o1+LMDkR8xDH6NG3HEhWE8k+fPt0mHhAJSfdh2RDbRqm2P21A+DouWF6OhYswLJJHgahI6DdOWqDP+ZduZQfXgpBswvsRGqkv9eF+xPIcTIYA7vhk8QI/S0Rw7+QciMk4dLnnIiKTR+wyAx4EU+4J7E+ZcMqzHY/s+mUsifZH/OX6suwH7UKb+PGbCpSZa0GdGfPct+jTjA36IkRtCyGEEEIIIYQQyZC0+ImTEXhQJVSXD+IMD8GIFHwQGXEEJXJ4pQLCHg/yhEITLonYx8tjyJPQSBxf/O4/hLJ6CKVEnPUh74DbkHMShs65EE1w/kWlR+WRXd7xiMojCoQQwuZZI481TAkjTiRaRu1PiDnhuIR3s94j4eSpuL5yGkQSnJ6IVgg+hEPTbxDREXtIb9mypZUTFy0hxIjt1OfJJ590Dz74YHCmnAU3KdeEF7og2OGg9c6+VKCsXgCMApceIe30TfLimnGdwnANuab8zvqy1BvnHII7a1ySjuuadsspaG/6EHnjLCb/ZMC5xyQF65Ii8iO+045cM8YKExGEiePSxIUcr41SbX8vjBImjSvcu4mTgXxxfFJWROhk2xDhFMc3LlbKxsQK90DWJKVvIqLiZkakDjseE8GLzxD/aSPuqZSJSRuWZcANjOiNMzRqogTxE0c5fap48eI2uRK7JAYgLvMCLf+Suuzgnla/fv3g27/h3oWrnXsaEy8IpVy7VKFPkBf3a+rJ+ag7Iih9EaK2hRBCCCGEEEKIZMiVgQohRBpSdGiNYGsz82oOC7aEEEIIIYQQQgghRLoj8XMnQgg/b0KOx/XXX59wTVCRPRI/hRBCCCGEEEIIIf7bSPwUaYvETyGEEEIIIYQQQoj/Nkmv+SmEEEIIIYQQQgghhBC7ExI/hRBCCCGEEEIIIYQQaYnETyGEEEIIIYQQQgghRFoi8VMIIYQQQgghhBBCCJGWSPwUQgghhBBCCCGEEEKkJRI/hRBCCCG2kfmzm7tasxcF37YXi1yv8TVcm6XB1xwn8/xFh276jB/k5gepuw+TXJuhzV2vNcFXIYQQQgghNpErYxPBdhZr164NtoTYfSkzpl6wtZnpld8MtoQQ6Ubu3LmDLSF2NIhur7tilbq7pnn5jojYzHVdbT8GVHK9arZyFwTfto7M884pMcx1yx8kpQACbeU117h5p5UPUmJY+rgrOtnlQDm3xPKd+b0rXbKnG1K8YJC6CctvXPDFbfG7PyaTmLZbM8jVGtfHzQi+1i63uT04roVrvWU+QgghhBDiP42cn0IIIYQQ28LST9zgfOe6yiZ8bgYxb17NYZs+PV3bfONc0ymTgl92YfId6woHm9sOonAN18Jds6n+QVIWm36b/KNrW4n22fSp1MC5mY8Ers1JrueSc90Ya7thrtfR4bbbdNy4Ps75ti1XyQ2evNntWThvITdjycTd0LUqhBBCCCG2FxI/hRBCCCG2gbFLxjmX95gEomFBV7lAEefWLAxEuTjh67ggw6HmuBsJPw8+cUPdOWaLMO9Q2Lp9HndjQ+nmpFzcxX6LDdHHMWkuzNV9XGWOzRIbo84JQZj5bMoR+xuUd91qDnNDih8dfA+xqS1muUKuqBeMN7VfiWDTjqtYJ6s9C+cNtR1Cs6vkmntnZ/4rXdt837s5v2d+dfnPdrVXT3BjFPouhBBCCCECJH4KIYQQQmw1i9y8Nc6VzhtH4MtikRuz5HtXusCZSboqY9yNNTs4NytmDU4LGcc56UPtnRs7pZnrmrdDcMwwN6bkj66pCaoFXdOKfC/i3NGZv8eGhRcu3t1clC5fg0zHpYXGZ4bZ/+ucW4ic37uuMwmV5/cUwuXz1nHNcXTaucini5tVsnVWXTYzyfUkZD5ou/lrfozrTp21xou5R7tiYTFUCCGEEEL855H4KYQQQgix1Sx2c1Y7VyLvv9eYnDGzWeCIZP3PIu7S/EmuQxnrboxxQporNEb4JK374iKubfHN63kWLn7Ntrkg10x0w1dXcr1Ca4TaOd2PJvh6apfbujVCLzgtCGmnfVwD92RYkDVXK23XxQ0+ukOCNTwLbnaPGpnfN4uhQgghhBDiv47ETyGEEEKI7cDmNT83fcoVcl3HpfAm8gRrbw6eufllP1vy/aY8fHh6IBwGv2wVvy+Ik09OOSsz1wMdWSBwlBaY4CqHQ/jzt8pquzF5X99Ul9iQek+m81YIIYQQQogokhY///nnH9e5c2d31llnuYoVK7qnnnoq+MW5888/35144olZn8mTJwe/bDs9e/Z0VapUcdWqVXM33HCDmzNnTvCLc71793Y1atRwF1xwgXv88cctjXI9/PDDtu0ZN26cq1+/vm2//fbb7tJLL7XjPvjgA0uDqPR4eYR59tln3ZtvZr5B/O+//7b6X3jhhfahre644w73559/2u9RecTj+eefd2eeeWbW57TTTnNdunSx32jfevXqWR4vvfSSpQHfw8csW7bM0p988kl30UUXuUsuucSNHj3a0nYUU6dOdcOHDw++OffHH3+4Fi1auMqVK7uaNWtm9ZVff/3VNW3a1F188cWWPmXKFEun3z344IOuevXq7uqrr3bz5+9+rzDgWvlrEeayyy5za9ZEP7F99913bvz48bbdqFEjt2DBAttOhvfff9/Nmzcv+LYZ+gv9FOgvuxNR7ZiIRx99dIv+l678/vvvrl+/frbNmGrbtq1tp8qHH37oTj311Kx7mP8sXrw42CN7rrnmGrdkyZLgW2JWrVrlrrrqKjdgwIAgRYg0hrUoXQqi4eoFW4a5h+Dt5mNKOtd1SkwovL0VPRBbsz4hd2iq5DnWlQ42N4PTtYgrlif4upXMn/26G5yvgWsWvKW9cPHWtnbn8KX/dmwWzn/upnJkuk15oVG8tonnvBVCCCGEEAKSFj8REPr3728iDOId4tzYsWPtofuXX35x999/v4mOfIoUKRIctW0gECLiDBw40L377rvulltusQ9i4hdffOGGDh1qZXrrrbdMJProo49MXETgy8jICM6SeZ6qVau6n3/+2fXq1cu9/vrr7sUXX7Sycq6o9Kg8PPyOOOrFTdh3330tPz4jRoxwS5cudYMGDYrMI4qbbrrJTZw4MetzxhlnmEiAWHbPPfeYyMs1GTJkiAlCCFr77LPPFscceeSRbtSoUSYkUv4XXnjBde3a1a1YsSLIZfvx119/Wf07duzofvrppyDVWf0POOAAu0b33Xefu/feey0dUe7000+38jZv3jxL6KWP0YacizZBCN3d6N69uzvssMOCb8nzww8/WB/bGj777DO3aNG/HyDfeecdt2HDhuDb7sXWtGOTJk1cpUqVgm85y/r164Ot7UeyeTDeckrkZeLE38P85+ijE61luJlU2+TLL790J5xwgrvyyiuDFCF2R3BCJhFmbaHsXjTMDM0evCR4qZCFsY/L3AYTSse57lkvJZrk2oRfhrQJEwtdH1fZv5go75nu0nzhYyLIeulSEgTnDL+lPlO0/Peb7VPlXyKmhdgHIibtEc5z6QQ3w78cKWibkf4FUEsHuK6rK7mqgYjqnaASQ4UQQgghhCdp8RMhpXjx4ubOQ4QCnGU//vijiVk4C3EnIj4eeuih9vu2gqhasGBBlzdvXpcrVy5zJLVp08bEm99++83VrVvX8j7ooIPsgR2RDOH14IMPdl999ZWdg30RLHGPTpgwwZ133nkuX758JgyWLl3aff7555HpUXkAIuQTTzyR8KGdcpx99tkmYkXlkQyIGux//PHHmwDLOTnH/vvvb8IqghDlIi2Wr7/+2sQf2hABgzZEGN3eIMZQpqOOOipIyWSPPfZw++23n/2lXbmuULJkSXN8AnX17kSEFxyS7IfjePbs2W7t2rX2284Ct9pdd93latWqZeLajBmZQYEIOVdccYW5V/nrnZetWrWyvowI3LhxY3Pgtm7d2q1bt85+jwcCNRMMiMEI/4Bz+Prrr3d16tRx06ZNs7R4eXJ96SeIhWH3Z58+feyaIC4Drtpu3bpZH6ceq1dveurcBO4/JjgoJ+J1PEELMZrrxaRCjx49LI3jcfVyXO3atbPcu7ijSce9i+OXejCJwn7eNc13xjb1oDzsw9giDz+WfTvGa38mO3A4ky8O0WHDhtkx1BnnN/A7eXKPeuONNywNl+Ttt99u7cpvuNs9TBZQHvrfq6++amm0LZMPzZo1s4kR8uZ+SDkpn29DD3WZO3eubXN+nOxAWZgE4f7ZoEEDuz/deOONWc5WnOpMFFx33XX2PV5ZwjCZwn2G8wL3LtqcNsquTslCf2GSi2uOE97nFdsm8YjXNxjHTOLQ37jeQuy+ZAqZM9b82yG9ec3PTZ/J41ztcptdmBec1sHVDt68XnTcAteclw1lUd51q9TAuazjX3fFTgut+WnwEqPMc2S+uZ3vPd2lS0J58gmJpuagDN7mHvu29/hszsOfr/KSc92Y8PqjCckMbc9c83Rze1je+VsF630G5Qxe8NQNERPRdU0oz5nOta3k1xXNbJtZk4PjbO3T8JqjOeNMFUIIIYQQ6UPS4uett96a9YCKkAcnnXSSPbwjMPIwjLBBaLoXrrYVBFXCpAl55+Geh2TyQMjjt2uvvdb2IxyYMp1zzjn2HXFj5MiRto1zDtEWgRBhISwQ5s+f3y1fvjwyPVEeOBN52EcUjeL77793Y8aMcUWLFo3MIztoSwSchg0b2nfCSRGALr/8cgtHRTzYc889LZ3QVNIRCZ9++mnbHzH4008/NZcp+dEehJhvb2gX+kK5cuWClEwQmRB9aVucrIhrgHhEmyDGIRAiDgKO2XC7sZ1Mu21PEKMRbnHdEuI7aVKmO4XxgYCHYIl4GCsE4bplKQSEQ65dojD2Qw45xEQ1hCpEQ0BQJrQZofC1116ztHh5ItLTBxA56XseRDbaGFEUaEcmLHBWM8nAmEF07dSpkwmv1G+vvfbKEjc9iLgsAUG/xEmKOMt9AGEPgZ36cQ6uoRdOKTtlu/nmm20sP/bYY5YHzmkP4j71oM8iXiLOUdfBg7dcsS5e+7McBiIndeHcH3/8cbB3JojoCJ20F6Ii+SKkwyeffGL1oXyIcgjGHM+HvGlzyoObFhCWcS0j4iFgIiRyLOVnsiEM15t8aVfGsheEEa35rV27diZ04oRGqLzzzjvtd+D6IA4nKosHMfi4444zgRHIs0OHDnZ9ECcRRpM5D9CeiLH+w/0XZs6caSIm15sPbnKuK4TbJB7x+sbee+/tbrvtNvvfwX1LiN2ZCwpUinFUZr5dfcvw82GZwl4W5V23rN9auQtY4zIsKuat44Zk/e5F08zzbj5P5jk2vwwoTr4R54z7AqHYMhjhcsacL/hty3qFiTk2+Pi8eeFRvPR/1yMmdD9u2wSYwzZwiQohhBBCCLGJlF94xEM+D9U4ksqXL2+iIts81CII8jDNdk6AAEQ4NCJJoUKFbBtXl3f+4fZCyEAkIm8fmsnDOqLjxo0b7S8CEvDdOw2B4xFuo9L9dmwe1A8HpxdCw+Dm8+vkEaKP0xL3V6I8EkF7n3vuuSb2AGIDTjJEF4Qc2hvRx4tlCBK4QRFTEIsRVAoUKGDCGAI2QjDOy50F15AwV8QXxMBYZyEiKOK1d8jhNgtDu8Wm7Wi4pog9rCeJyIv7EAjh5/ojrCEkxk4CIHxRP0DwSfU6IEQB1xD3I2SXZyJy585tbmQoUaKECVvffvutXQ8cfvQXnI3Tp0+3fTwId7QBzmbqQJ9DPEVw80Ita9/iTF64cKF9Z11ghFSETQRx+ivHePEMaBNgH9zKLOOAeBvrpozX/gh/1J17E3XwyyZ4EP1pP87J2MU9TT2A8x1++OE2iVCsWDFrW/bHPcmavQiSlMG7bU8++eQsQb5ChQomDjMecTXG3hP4nXNxLP2aCQruX96lTlkROBm7CJVh8dRf70RliYI6cW/mnkN/4Zhkz8N9nfuH/7z33nuWzjVFUKecCOi0kx+74TaJR6K+IURaQCj2trxVXeQYY5eMc+7os7fq7fNCCCGEECI9SUn8xJWH2IJrB8cSHHvssSZIEt7MwzoP2z40fFtBUEAMQNjAJdi3b18TXHBqIYC1bNnShBnEAxxsHh7COQaHoXeLAgKHDysFnFXsG5UelQfipw8NRsRD0CO0F8JrfiJGEQ6Kwykqj+zgHOE1C1lSAMcYfxFx2MaphmiD64r2R/RAbMINh6CDIwsRmJeKIHjh/ttZINTi/qS/IErj4p01a5atq8pSAoS/46jFKYdQdMQRR5j700O70ZY7k1KlSpnDEGGOMGLvXqXcuIMREuM54MLiN0IgfSUV4i0nkV2eiaD/0DfD0D9YX5YJBz70b5Z3CMO4oPwerhuCHsJ0uI70PUR/YJkDT/jYMOzvCW/HEq/9ETWZDGDSAaezn3AIE1s2ygvcU2JB1MX17duB/HChgp+IAO59CIHkj+sSN2UYREgEZF78xTb9G+cl9xKO4ZpSB/Lgrw/RB59PorJEwT0glq05TxgcoYiftCtuVyakPOE2iUeiviFEelDeNSvp4r6sR+xIJrmRi4u4tsXLB9+FEEIIIYRIQfzEaUWYK+G0Bx54oLmBENdIQwTEbfjQQw/ZQ25sqPO2gKhI6DsghvE55phjLDyWh2fyRMSJBZcRoawIJF7cQBBk/U9CwAn9xvWE0BOVHpUHTi8cUQgdCHk4z3DJJSIqj0TgzCJ0njUwPbhAEXcIfV+5cqWds0yZMu7ll1+2NzxTXtxdhLeecsop9sZwyohDi9BkvuPs2lngdvOuXERb3F+4aRGqEa+AMiKoIeDQ3xCAAfGlcOHCJpjuTOiTlJVQXSYCKDv1IYya77y4iesbdrQCgpcXxxDPvXsT92PUUgSJXK7Z5Rl1bKJzIqJ+8803FpLP+RHO/XXx4PKjzvRPXKcsU0DfYtz7a0WINNeQyZGcJl77MyHChAzj4+6777a8w6I57cPL03Bb8xuCcXjCJBaWDsDhyLVB2L366qu3WD/Vw9IOCPWErrNGKWUJg8CJSEiYOO1GOV555RWbtEAQZjxQLsBBy+RSLMmWJdF1hWTPEwX9gokk1pzlPs860LF9PIpU+0Z4TERtC7GrUbh49/ih5GIHQpj9NrzdXgghhBBCpCVJi5/+JTk8qBMyyQcBEOcZobOPPPKIiVOEneaU+ImoiFuKh3QcXawXyAuXcE8RskrIN2Gm/uNfngK4UAkP5+UcHtyGCDWsfci5EEtwLEWlZ5dHKkTlkQiETQSLsGMKIQWhBUET1xnCM2Im3/PkyWPfccniAkTcIcS8bNmyJgYjjhIqHOv225HQbxDscH3iIuPt7YjTrFmIIEg625QTZyRLGCA8U1dEL8S4nQ2h64wDykroPtcSIYvQZZx1XB/Wf2R8MEHgoV6IdByPGIY4DQhxvOwmFoR7nIBRb/FOlCeCOOK/f9mOhzwZQ1HQJxnDvASIsYOgRn8Kw+QDazWSL6HcrN+KQM+9ABGU+uF45hoi/uU08drf33MY94wz2oMx5+H+wb0EwZR1QtmHcPAoECqpF+OIpSM4jrEUC+Ouffv2Vibuf4ihsRBiz9hEtGdM4gAnHB5wrbK2Kf0cYZIxGksyZWFCCkGS9VSjSLZOsWt+8mHyi2vNy6dw1TLZQpty30+GVPsGY4JlQxJtCyGEEEIIIYQQyZBr0wNzZuxnCFxBQuzulBlTL9jazPTKWydeb29wUCOMCyG2HpZtEEIIIYQQQgghwkj83IkQNspyAfHA7ZrMmqAimt1F/CT0nzFHyLkQYuuR+CmEEEIIIYQQIhaJnyJt2Z2cn0KIbUfipxBCCCGEEEKIWFJ627sQQgghhBBCCCGEEELsLkj8FEIIIYQQQgghhBBCpCUSP4UQQgghhBBCCCGEEGmJxE8hhBBCCCGEEEIIIURaIvFTCCGEEEIIIYQQQgiRlkj8FEIIIYQQQgghhBBCpCUSP4UQQgghhBBCCCGEEGlJroxNBNtCpBVFh9YItjYzr+awYEsIIYQQQgghhBBCpDtyfgohhBBCCCGEEEIIIdISiZ9CCCGEEEIIIYQQQoi0ROKnEEIIIYQQQgghhBAiLZH4KYQQQgghhBBCCCGESEskfgohhBBCCCGEEEIIIdKSpMXPf/75x917773u1FNPdWeeeaZ7/PHHg1+ce/fdd12VKlXc6aef7jp37mz7psrTTz/tzjvvPFepUiV3/fXXu9mzZwe/OPfCCy+4iy66yJ111lnu4YcftjTy79Kli217PvjgA1evXj3bHjRokKtcubId9/7771saRKXHy4MX4VPnihUrunPOOce99tprlv7333+7IkWKuLPPPts+tMntt9/u/vzzT/s9Ko9ErF27Nitfz6RJk9xll11meTz//PNBaurpEO/8O4Ivv/zSDRkyJPjm3B9//OFuueUWa8+LL77Yygy//vqra9CggTv//PMtffLkyZZOX7r//vvdBRdc4K644gr3/fffW/quxI8//uhuuukmV6FCBXfGGWdY/ZYsWRL8mjpcO/oYH8bEzoTrUL16dTdjxowgJXm+/fZbN3bsWNvu3r27GzBggG2Heeihh7boH8ni2wjo78ny+++/u5dffjn4thn6YatWrYJvqbEtx8J9992X9H0iTLh/bG07JkO88oXbcVvrL4QQQgghhBBCbE+SFj95sH711Vdd06ZNTWB75pln3JgxY9z8+fNd8+bNTehr3bq17ecFrWQZPXq0GzlypBs2bJiJJZyvSZMmJiZ+/vnn7u2333bvvPOOiaz8Pn78eFerVi333nvvmUDp4TyXXnqp+/nnn12PHj3cW2+95fr162eCLOeKSo/Kg78LFiywvxyDeIhIB/vuu6/75JNP7IPoitjVv3//yDwS8fHHH7uWLVu6mTNnBinOrVmzxt11113uueees3bmfD/99FPK6RDv/Nubv/76yw0dOtTdfffdbunSpUGqc3379nV58uRxH330kXvggQdc27ZtLZ1yIxx++OGH7s4773QdO3a0dEQX2pY2vu2220wI3ZWgns2aNTMRijpR3uOOO87dcMMNW/TNVEA837BhQ/Bt58KYQswtXbp0kBLN+vXrg61MEKqzuxfcfPPNNlGQKlvbRlyv7SUSbg/oQ8nUc2vbcWvZ3dpRCCGEEEIIIcR/l6TFz4ULF7qSJUua0IMIBXPmzHEjRowwMQvnI65JnGL8TYVffvnFHXPMMS5v3rwuV65c5iBt3769PfSvWLHCXXXVVe6AAw5wBx10kJ0bMaxo0aLukEMOcVOmTLFzsO+4ceNctWrVTLjEQZovXz531FFHuZNOOslNnDgxMj0qjwMPPNDqu8cee7jDDz/cPvGEiIMPPtjEL4TgqDwSgUh05JFHBt8yQezjnJxj//33N2GY/FNNh3jn394gjtCG+fPnD1IyoS33228/+0u/4XrDCSec4C6//HLbLlOmjFu3bp1tI77Vrl3b9sP9iYCLi3VX4Y033rDyXnvttW7vvfe2PnPHHXeYC3T58uW2D/0SZ/SFF15oYrAXCfn+v//9z5yuuJ1XrVrlXnrpJWs3+l2YH374wfoo17hhw4ZZwvY111zjWrRoYY7ZmjVrmijIxADn9GPjiy++cJdccontw9+5c+daOsciQJPGpMFXX31l6R6OZ4IBMf+bb76xiQHOS7k7dOhgzkOcuRzPecLCNGPq2WeftWPoi4B7lHoyUfLmm29a2osvvmhiPXVHoK9ataq77rrr3PTp0+33eMS2EWUgf+rPsatXr7b0119/3foMLmPEaCYhmIxgPDzyyCO2T5jffvvNhF7KgOPbQ/1pO+4tr7zySpC6JfGOpVy0E85x7ik+T9oN0Z90nOrcW2PB2c4EE9eS8kb1IY9vR4gtL2ORe6ofN9OmTTMXNcSrWzLli23HePXH5c1YoM8ytj/99FNLD8MYwfFNvbhGuIyZyBFCCCGEEEIIIXKKpMVP3JgIGYDAByeffLK5+jZu3GgPuDxE87Ce6sMrD8yEUSIaEsrOgz4iC2Iov/FwDIQXkzd5AXkhvgLuzRIlSrjDDjvMhCFEQE+BAgXMkRmVHpUHgkH58uUtXLdu3bom2sQTEefNm+dGjRrlihUrFplHIurXr58lRngWL15swgBiAGG9jz32mNtzzz1TTod459/eICLjRqP9wjRq1MjE4HPPPdccxIhWgDhCW/EdIRHxG5YtW7aFgErbZteeOxJEOgS2MLT7Pffc44444ggTARHACBGmj+y1117uqaeesv0QyxEdSS9cuLAbPny4OZ5ph549e9o+HkSkxo0bm7uUa4lo6ilevLg5ZpkQwME8cOBAc2j7MHMEUdqTfWhzXMEexi5jCCGxd+/eQWomp512mo0N8qIuhD/369fP3K30NX9+lqjgeoaXoWBi4tZbb7X+WKNGDUv7+uuvrR369OmT1QYeXNeI4TjAGYvxhDJPbBvRH3AN4zQ+9thjrT7Ui3oPHjzYnM/APYLrwpIVuNRjoV926tTJBF8c3Vwf2pt245zUl3LyWyzxjv3uu+9M1MU5zofrghCJQ5z2Y8LiiSeeiKwrrlnqwD0tqg/FEq+8iNiI8RMmTLB9aGP6QVTdkilfbDvGqz/C6qGHHmp1f/LJJ02kX7lype3vefDBB20ZFfLiHrsj3elCCCGEEEIIIf4bpPzCI8SVNm3amPuMB2oceoidrOlHSDPuLh7YUwGhhJB6nEeIQIRA83DunUqEfnJu8uzWrZsrWLCgpXvhCKGDvzjQgO/eUQg+dDQq3W/HywNwjl199dUm+iAyAvVGZOSDGINQijMvUR6pgEiCqITY5AUFBIJU03c1WKvxxBNPNLEFd1u7du22cLEh7HHtvbgTu34s7RmbtjNBJMIx7EHoo0/g+EV4YzxQP8QrBEkEwKlTp9q+OHQR/AFXdaww5KGvcR6cjIjHnNefAxDl4fjjj7fQ53322ce2Ed4Alx7nYF1bhC6/ViYgTkKpUqUi8wfcgmXLljURFNcuYw0xEai/n5BIBM5FhDuEy1jKlStn/ZW1KxHOWUM1WXLnzm0uRaAe1JsyIuYyDrg34VwN1zselAGhkfHL9eA8CJAI2DjbEYFJi3XIQrxjWSqAJRy4Xo8++qi1L32Bc9LHEclpC9YMjgeiOtcyUR+KJaq8XGd/P8AhyvWL2jfZ8oWJV38EUc5HmRHOqYN3HXsQVv3EDPUNjyUhhBBCCCGEECInSEn8xJVHOCQuMh/aiHDJQy9CKKGdhHsTxp4KiJ08zCN84vojJJbz4kxC6MJBhvjCGnM8ZHtwAXIMD9m4i3CLAgKNDwsGnGHsG5UelQcO1FmzZpkAQOj1Kaec4j777DP7LbzmJ05RXE+EPUflkSo4pnDT8Zc2RYBlmYFU03c1EF5wMOLmRJChrLi9eJkUIjriFWIMa6suWrTInLa4Pz20J228q4DbOCyGEeJNn/CiIsIczjZce3wQfXHEAU5H+kx2sA9jjMkBzsFf+rzHO3wBcTEW2ps+Stv6F4J5WLIhGRCdw6I++SD0Q7KCFXWIgmUPaDvc07QPod3JwlIDse2IqMeYpQ8xFrw4moh45UOgZkLEXz/KiJgXS7xjuVfgqKWd6tSpY/cqQJgNtyX9IB6+XRP1oViiykv9KQ9O5UKFCtl9ImrfZMsXJl79KXfXrl2zzs89OlZIZWLI58Vf8hZCCCGEEEIIIXKSpJ80eXBlrT3WZkNsIEQXpyFuHcROnE2EMLLeXmyoczLgzmKNOECwYJ05HtIJtUc8IIwboSwW1vnjRUSIJl7IwU2HcMkaf4hoiFOIB1HpUXng8sSBiGMM1xZvLvcCRhRReaQKYgVhqrQteSNe4bxLNX1Xg9Bs79bFBcYLpXDZIswQFgy8JRwhlDVL6W/0NWAfQm1ZDmFXAWER1zLXHIEQcAgTWgwIjghOLKdAnRH1fD0TEXa3Igjh5PTn5Hj/oqjsIE8EfPIlNJz+GHbaJgt9iQkK+hf1RChLZpwn69Jl/FMvRELK6icZyM+v/xpLonN74ZylF3gBFRMb3oGdbJmACR2c7DircaKzPEOsezEKrjsTMgjOtBl9nbbnOhBizrVBzE8U4g+p9KGo8iJ04kbGWctyIYn2TbZ82bUj5+/Xr59tMxHDhADnDIMYSjQBMIb8kimUyb9cLmpbCCGEEEIIIYRIhqTFT9xsgABDiCQfwqsR9ngzN45J1txr1apVUiGwYQhxJWyc8EdChgnt5eUZPPTzshbW7ON3/yH818OLOhBhfcg74CrknIShcy7Kh4sqKj0qD0QLxFAEDERWjmMdxERE5ZEqCDaE07MOHmIFYai4a1NN39VAuMHpyfqwCFOI5ojWrCGIwE46L75BdMZdy9IGiO3Uh3UDEdh3JXAFs1wCbkwchpQXoYoXwVAvrj2uYPoEAjWCES/lSQT9j3Vaw7CeJnkgriNY0V7JgHCKqxlRkf7MJAHjljGTCrht6cuUnfGAczDWRRoLeeH0Teat4Ahv3Fu4/rQXeQEvSMIBHku8NgqD0IeozrjlXkW7MUlCX0KIDK9PmgjEOdY95f5CvRlfLNuQDPRZ7i2UgSUHCPvHPU+70X4sV0DZvEs4ilT6UKLykg99078VPmrfZMqXTDty/2MtZybIcMDS/rHOZK6zd+2zJAMTXsDkDfezRNtCCCGEEEIIIUQy5Nr0AJtpVxMizSg6NPNFO2Hm1cx887jYPUA8RDDj5Woi/WEya8CAAbuUu1sIIYQQQgghxO6NxM8dACH8vEE5HjjbtmZNUJE9Ej93f3CO4hzUWpD/DSR+CiGEEEIIIYTIaSR+irRF4qcQuxeslcz6pEIIIYQQQgghRE4hO5UQQohdAgmfQgghhBBCCCFyGomfQgghhBBCCCGEEEKItETipxBCCCGEEEIIIYQQIi2R+CmEEEIIIYQQQgghhEhLJH4KIYQQQgghhBBCCCHSEomfQgghhBBCCCGEEEKItETipxBCCCGEEEIIIYQQIi3JlbGJYDuLtWvXBltC7L6UGVMv2NrM9MpvBltCiHQjd+7cwZYQQgghhBBCCJGJnJ9CCCGEEEIIIYQQQoi0ROKnEEIIIYQQQgghhBAiLZH4KYQQQgghhBBCCCGESEskfgohhBBCCCGEEEIIIdISiZ9CCCGEEEIIIYQQQoi0JGnx859//nGdO3d2Z511lqtYsaJ76qmnLP2WW25xJ5544hafli1b2m+p0LNnT1elShVXrVo1d8MNN7g5c+YEvzjXu3dvV6NGDXfBBRe4xx9/3NLI/+GHH7Ztz7hx41z9+vVt++2333aXXnqpHffBBx9YGkSlx8sD4qX//fffVs8LL7zQPrTJHXfc4f7880/7PSqPRPCG/XC+EFWmyZMnu3r16lneL730UpDq3JNPPukuuugid8kll7jRo0dbGi/z57pVrVrVVa5c2fXv39/SdxRTp051w4cPD74598cff7gWLVpYWWrWrGl1gV9//dU1bdrUXXzxxZY+ZcoUS6ffPfjgg6569eru6quvdvPnz7f0HQXX2l/nM888051xxhlZ33/66Sf7u7uyYMECd/vtt7tKlSq58847z67L0qVL7bePP/7YtW/f3rZ3Bz788EN377332vb777/v5s2bZ9s5fX3oD9ynkuH33393/fr1s+1nn33WDR482LZTZdmyZa5u3bru3Xffda+88krWOflbu3Ztt3Hjxhyrc/icqdKoUSN37rnnWv70Kf727ds3+FUIIYQQQgghhNg5JC1+ImAhnPGAiyD3/PPPu7Fjx7qGDRuaCMkHATBXrlzulFNOCY5KDgRCHt4HDhxoD/gIqnwQE7/44gs3dOhQy/utt95y48ePdx999JGJiwh8iHsezoPI9/PPP7tevXq5119/3b344otWNs4VlR6VR1Q67LvvvpYfnxEjRphoNGjQoMg8EvHpp5+6u+66y82aNStIcZF5r1mzxt1zzz0m/nJNhgwZYuLIqFGjTDBk3xdeeMF17drVrVixwo5buHChtStleuKJJ0xo3N789ddf1i4dO3Y0kdBDGQ444AC7dvfdd1+WYIWIe/rpp1s9mjdv7rp06WLp9DHalnPddNNNJoTuSPbZZ5+s63zzzTe7K6+8Muv7UUcdFeyVOuvXrw+2dg5cH9r57LPPtmsxbNgwV6hQIRt34TG1u4AoTX3gs88+c4sWLbLtnQltHBb+t5ZJkybZdWIC4IorrnCXXXaZpXN/4D68xx575Fidw+fcGpikYWwwEcX975FHHtli/AshhBBCCCGEEDuapJ9webAuXry4ufMQoQCnEaIDQiQfBDsErOuuu85+T5ZffvnFFSxY0OXNm9fE01NPPdW1adPGbdiwwf3222/mekIwO+igg8x9hxhWpEgRd/DBB7uvvvrKzsG+iIO4siZMmGBOtnz58rkjjzzSlS5d2n3++eeR6VF5RKXHQjkQJ3744YfIPBLBcUcccUTwLZOovBEzyYtz77///ia4HnbYYe7rr782txVtePTRR1sbTpw40R144IF2zRAzDj/8cNuXttreIPxQ3liBkHLst99+9pe6cb2hZMmS5vgE2gyHHSCkIPawH47j2bNnm0t2VwFnardu3exaNWnSxK1evdrScSIySYALFwEYsZN9Ea/YHxEXF2y7du2sz1511VV2vcKsWrXK6gyI6riNvSP22muvtWv+5ptvmiiGk9ZPGCBe4gLGwYdDGGEzFiYaOB/57r333tZP/ve//7ny5cu75cuXB3tl8uWXX1q5ceXyl3EflQfi/G233WZ179Chg9UfaBecpXznGOqBQNaqVSv7HXGbPst15/pWqFBhC4GYNnvjjTdsG9GctoZPPvnE7kecr3v37taGjBG2KWfU9QmD4E7fo2z3339/Vt/j+5133mm/4UbneniYUKhVq1bwzblnnnnGXJlhmPhgbNNO8O2339p5aBvGLWTXB7g3IiIyeUE5mdx45513bFLkxx9/tAknJk9yos7hc8Y6P+P1s+woXLiwjf+VK1dG1pMoAfoJfYvzjxw50tKFEEIIIYQQQoicImnx89Zbb7VwbkDgg5NOOsn+Ag/giI9t27ZN2TVEqDYPx4S8IxggHPGQjZDHbwg9QJgueZ9zzjn2HcHVPywjvCLOIu7hhEQc9OTPn98Enaj0qDwS5R3m+++/d2PGjHFFixaNzCMR11xzzRZCCkTlvWTJEhNELr/8cgsrxQG65557mhjMNUCUID/aA4cnglK5cuUs5BZRmhD6cPm2Fwi2CD3kHeb66683MZj6IYIgDAJCDG2FaNO4cWPXunVrS0f0C5eX7ezac0dCWU477TQTExHw6Y84bjt16mQOOpy5e+21l+vRo4ftP3fuXFsmgXrjED7kkENM2KLeCP4IRR4ESdqEY6ZNm2bbiHwIy4hvpUqVMiHstddes/4HXHeWjEBYpEyPPfaYhbDH8s0332QJqx76EeM3Vohn3FM2BDOEO9yBUXngOi5TpoyJoYjYXjBkWQtEeURO2obre/LJJ9uyCMBf2oJyIerSpoiyHsRQv0QCIiL1RxylPZgM8DBJQL1wgTIe412fMPz+wAMPmMBImRlbPjwd4ZJ7EvXFFfvee+9ZOhxzzDEm4vtQc86LeBcGN/dxxx1noi/MmDHDPffcc9ZGtAdk1we4nzEeuNchEHoYO5QBYZX+lBN1Dp8zfA9HCI3XzxLBBAsTF0xaHH/88QnrSf60LeWhb+wIZ7oQQgghhBBCiP8OKcc2IgTg1EFIwyXmYX1KHI84+FKFh2LCnhGLEBnYxk3mHX64zAiXxuFFODQiCiBM8DDOwzl/vTjAd+8oBI7nYTwq3W/HyyMqfd26dSY+8sEJhciI0ypRHqkSL2+EL8QwhAJEKcJdEaEQTwsUKGDiFEI1QjDijAeBhPIRRo6AurPg2p5wwgm2tiCh+d4V6UEERejx4hAutjC0SWzaziR37tzW76FEiRLmskOco0446rgWiF7Tp0+3fdjfC+iIwIia7IMYxTEI6WEQ/RCacDiz5AQuTM6FwxpxkHbCRceakt999505+BDc+Ms4pSx+CYEwCE6UxVOnTh3ry4xphK4wLE1Af2eMI6QlyoMQbc4D3vUJCJdeHMRximsZdzPiHAIibcQ9hfrFCppAfak37cuxCKwzZ860dqFvRxHv+oThnJwLwRfBj2UzyB/Ixx/LeAo7P4H7Dct1INYyNmNF41hoF4RwRGxPMn0gVbalzvFgn3j9LB44eaknkxuIsTg7qXOienIfJw/uX5SLviCEEEIIIYQQQuQUKYmfOHYQQnBtEcLowUmE6MED79aAkIeAgKCC+6hv374WSk5IK0IXD9A8sBN6iaPJgwuQY3iw9m5RILwbB6aH8rFvVHpUHonyDq/5iSDEOpyIUVF5pEpU3oceeqiJPfwltJ5txCNce4QDIwIPGDDABBBEFtqFUHG2CR8vW7asXaudBUIt7k/CYRGPcPey1ikht4RMI543a9bM3JOLFy82gQb3p4f2pI13FbgGYYci0PYsB4GYzwehl7VWgVB/D/t5hygfXHG4IcMgAiJ+0g8QuHEB099Jp41wB9NO9ANeNgOsU4owjgiFG9iL8mGKFStmYpSHMGz6cryX+XA9cB4jpBGyDlF5IPR78Z8X/ngQrcOTAvRX9kfc5dwIZNQJEQ5BM1b8RIg89thjzUlLGzEeEOPoGzgLo4h3fcLEloty+LbiWiU61oufrOuJaJ8djNlYkukDqbItdY5HVD+LR3jNT0R0/l/QJxLVM7ZfbO1kkRBCCCGEEEIIEY+kxU/WlOvTp4+FVxKOS5iifyM7IgoPz7EhzqmAo4jQd+Ahmw8hmITacu6HHnrIHupjwU3GAzfhngimcP7551sIPiHghFBSPsSoqPSoPLLLOx5ReaRKVN4ID4hNOPcIGyUv3FK4sRAVcVSxbh/fcfHh8sS1hVOL/QkvRjDeWRCe7926iLaEL+OaQ9DzyypQdoRQQn7pbz5sF9GWdQQRTHdlEAkJ30aopJ6I0r5uYbg+fh1L3Lysp8n+YVhagnORnidPHnMgsu4jIpQXh1leAGEQgRTRnEkDJifoK3fffbe1ZVhAhvr165vgjDiOGAaIVoh5YcgX8ZwJD9yX9Gf6WFQe7IO7GDifh3uDv444NtmfMlMPxGHqiYhJqDnjht9iQShlXU1eqEY+lD/q5Wq0QzIwdph4YTzRDohynDsZ6LeIjPRn73aNJbtyJNMHkmV71Tmqn2UH15OxisM3UT3pF5SDFyNxf6J8ONx9+HvUthBCCCGEEEIIkQxJi5/+BRWII7wEhA9viQZEUBw7W/v2a17KQdj41VdfbSICb9XmJT24AAmDJbSbUGH/wQnpwanGwzShmx7KwTp5DRo0sHMh0uDiikqPyiO7vOMRlUeqROXNsgAIVwidLA1AmDvCAqHkuDoRg1m38dFHHzVhhnBmBAhcg7ypnO+pvo0/J6HfIKbhmiMslhf/IFqzPiJCHOlsU37ctYTEIv5SVwRyhMRdHa434eC333679UuEIhzNsXANEfy5ZqwL2blzZ3PhheEaIhizfibgePSuXkQiQoVxXSJOIrwzEeAnIRgb9EX6S+zY5HjWYXz11VfNMY2jE5GVkGTWa/UQjsyyCzj/OA+CFuPeTzTE5sESEDhVWZMW0csL97hHEeVxSOKS5vriHqVetAF9Egcg5+deEA/ET1zVOAZxrjLBwLiIhXahHbgvZAeiHuIt9yDKxhIcjJFkoc0RcMNLTHiYJELUYx3jKJLpA8mwPesc1c+SgWUN6A+J6okjlv5Hedq3b2/lww1MX4KobSGEEEIIIYQQIhlybXo4z7R9hdiV3qYtxNZSZky9YGsz0ysnFq/FtoHDFcc2TkImTFizlpdypSu4IRGHt8X1/l+GpT0QRqME71RhYkAIIYQQQgghhAgj8XMHsGjRIltTMR64XbdmTVCRPRI/dzw4D7t3727rROIaxeG3M5dZ2J7gRsadzoutxNYh8VMIIYQQQgghxPZG4qdIWyR+CrFrwzrECJYsgZATSPwUQgghhBBCCBFLSm97F0IIIXIK1pfNKeFTCCGEEEIIIYSIh8RPIYQQQgghhBBCCCFEWiLxUwghhBBCCCGEEEIIkZZI/BRCCCGEEEIIIYQQQqQlEj+FEEIIIYQQQgghhBBpicRPIYQQQgghhBBCCCFEWpIrYxPBthBpRdGhNYKtzcyrOSzYEkIIIYQQQgghhBDpjpyfQgghhBBCCCGEEEKItETipxBCCCGEEEIIIYQQIi2R+CmEEEIIIYQQQgghhEhLJH4KIYQQQgghhBBCCCHSEomfQgghhBBCCCGEEEKItETipxBCCCGEEEIIIYQQIi1JWvz8559/3L333utOPfVUd+aZZ7rHH3/c0jds2JCVfsYZZ7hnnnnG0lPl6aefduedd56rVKmSu/76693s2bODX5x74YUX3EUXXeTOOuss9/DDD1sa+Xfp0sW2PR988IGrV6+ebQ8aNMhVrlzZjnv//fctDaLS4+URpnv37u7111+37b///tsVKVLEnX322fah7rfffrv7888/7feoPBKxdu3af+XLuU8++eSsz7Jlyyw9qqyPPPKIO/fcc92FF17o3nvvPUt79tlntzjHCSec4Dp27Gi/7Qi+/PJLN2TIkOCbc3/88Ye75ZZb3DnnnOMuvvhiN2nSJEv/9ddfXYMGDdz5559v6ZMnT7Z0+t3999/vLrjgAnfFFVe477//3tKF2B5w37n22mttHNEPn3jiieCXnOf33393L7/8cvBt+9OyZUvXrl274NtmRo8e7apVq2Zj77LLLnMjRowIfnHu+eeft/tdFFtThzVr1rj77rsva7tVq1ZZ960WLVpYGlCWn376ybaFEEIIIYQQQoitJiNJBg8enFG4cOGMHj16ZDz00EO2/f7772f0798/o1ixYhmDBg3K6N27t6VPnDgxOCo5Ro0alVG9evWMlStXZmzcuDFj8uTJGeecc07G2rVrMyZNmpRRtWrVjE0P2Rm//fZbRpUqVTLGjRuXMXfu3IyzzjrL9ve0adMmo2/fvhnLli3LqFSpUsaqVasyli5dmrHpwdrOFZUelYeH30uVKpXRq1cv+75u3bqMkiVL2jasWLEi44orrrD6R+WRiAkTJmTceOONGQ0bNgxSMvOoWLFi8G0zUWUdMWJERt26dTNWr16dsXDhwowKFSpk/Prrr8FRm2ncuHHG7Nmzg2/bjz///DNjyJAhGRdffHFGz549g9QM6z+tW7fO+Oeff+w6+zp26dIl45lnnrHt0aNHZ1SrVs2233vvPWsbrvMHH3yQUb9+fUtPhiJDLv3XR4hEcB8aOHCg9TfGcIsWLbLG/dbAedavXx9825Lly5dn1KpVK/iWPX///XewlTpr1qzJqFy5cvBtM7/88ktG2bJls+4JixYtsvvL1KlT7TvH/PHHH7Ydj1TrAI8//njGtGnTbLtTp04Z7dq1s/sdH+6hTZo0sd+WLFmSsWHDBtv2bEsbCCGEEEIIIYT4b5K083PhwoWuZMmSrlmzZu62226ztDlz5rj999/f7bfffuZ+5LPXXnu5ww47zH5Plk0P4O6YY45xefPmdbly5XKnn366a9++vblKV6xY4a666ip3wAEHuIMOOsjcjpseil3RokXdIYcc4qZMmWLnYN9x48aZg2n8+PHmIM2XL5876qij3EknneQmTpwYmR6VB+BC6tatm7vmmmvsezwOPvhgc63Onz8/Mo9E4GY88sgjg2+ZkD/HxxJV1qlTp5rblDYsWLCgteEnn3wSHJUJDkzKU6xYsSBl+/HXX39ZufLnzx+kZLLHHntYf+Fvnjx57HoDjtTLL7/ctsuUKePWrVtn27jSateubfvh/pw5c6a5ZIXYHnAv4j5Hf2MM40osXbq0/cb9pUqVKuZQvPvuu9369estfejQoeYSrVixonvqqacsDWc6LvhatWrZ+I53bOfOne03HNs4nDt06GDnYEyTBp9++qlr27ata9KkiXv77bfNnUkZ4bPPPnM33HCDbYd59913rTzkxTlxbuKSX7p06b/c8qtWrbJ7eIECBez70UcfbftwH3nppZdsDHPPB5zvjEFc2+SL0z1cB1zdd9xxh90LGcuUPRbKwv2Q+xBQF+7l++yzj324z9asWdN+4//M8uXL/9UGQgghhBBCCCFEKiQtfjZv3tweqgGBDwijvuSSS9xxxx1nD9o88PJgzMNsKlStWtXCJxENefBGKODhnQdwfiMcGn788UfLm4drQFjwIZqff/65K1GihAmvhEqGhUMe7H/++efI9ER5ENL/v//9zwTOKObNm+dGjRplomJUHomoX7++hXSHWbx4sVu0aJGrXr26LTPgw2+jynr88ce7CRMmmCCBYEB7eJEEEB0QM+KJJdsDhNmbb77ZlS9fPkjJpFGjRiZ+EOZKWOsDDzxg6fQd2orvhB0jfgOh/mEBlbbNrj2F2Fpat27tGjZs6Jo2ber69Onj9txzTxt/TDogJBLizVhnkgehE0GRpSf69+/vRo4caULoDz/8YOdiSQeWwOCeFO/Ye+65x5bPIM/vvvvOhMixY8faZ+DAgTaBACzn8eCDD7orr7zSxj/nACYzYu8bjH1Cyvv162dLbvB9wIABNp6YYKIcYcgfUZZ7COV45513TPwlHbGRMdmzZ0+3ceNGq8vgwYPdxx9/bMdyjwnXoUePHu7QQw+18j/55JMWwr5y5Urb1zN37lwrh6dx48butddes3rQJkyoefEzTLgNhBBCCCGEEEKIVEj5hUc4b9q0aWMCQYUKFczdxAMrD8U8JCPGvfXWW8HeyYGD89VXX3WvvPKKK1y4sHvuuedMGPMOv4yMDNe3b1/LExcmzka49NJLTQjgwZy/CLHAd+8oBI7HGRqV7rdj80DIOPDAA7OE0DA4E/2an4gEOC1xZCbKIxUQEXA+Ie4OHz7chGdEYYhXVsQD/rIWKAInAgaOLg/XDVcZjtGdCWsInnjiiSbI4I5jDULvoAPqwbX3DjoccWGoe2yaEDlFnTp13EcffWTrDuN2pz8i+s2YMcP6KeIhwujXX39tbusvvvjCxj6TIziaGade3MMliZsx6tgwuEvvvPNOu48++uijJhr6cYGj3jvDucchsjKZgSOSSacw06ZNc2XLlnVHHHGEuavZH5EyEaypy/2T+xz3GATWb775Jvg1E87Vu3dvc5Oz/jG/U4YwTGp89dVXVkcEWMqP2BkG5zaTZR7KikjLJBOiMBNNsQIthNtACCGEEEIIIYRIhZTET1x5hB/iIuJhFXjYRrA85ZRTzEGEWzP2wT47EDs5hvPg+nvzzTdNEEWEQOi69dZbLR+cTuXKlQuOynQBcgwP3biNcIsCD/7hF2XgFGTfqPSoPBA/OS+CIq4thDse/GHfffc1IYAPgm+nTp3c3nvvHZlHqhQvXtxcTgipOMcQUmbNmhVZVhxqXB9cWaTnzp3bXFseBFTC4nc2Y8aMMbcXbk5crYQWI4jg/mKJgVKlStnLo3gBEs5XBA//oiegPWljIXIaxE7GEGMHIRARjnsTjmnScIByH+CDeI/TkfGIaOehD/tJGz/REHVsGELYCS9n8gQBlvuaJzxhQbg4rnBEUtzT3IfCMDkQnnyhbJwzCu8y5R5To0YNu79xDyYtDM5Xlp9gTOLuZyIlFurZtWvXrHpybkTLMAii4fZC8MXhShQBL0LjRW0IyIz/MDt70kYIIYQQQgghxO5L0uInD7KIADiNcEMipvFmZFx8iHIIg4Ql8vDPmo2pwkM3a8YBD9gIEYUKFbIHYR7eH3vsMRPKYiFEkrBTQs59aDrh8ziYCAHnIRo3EuJDVHpUHi+++KIJsDiTCNe+6aabLPw/EVF5pEqvXr1svUHK9dtvv1k5TjvttMiyfvvttyaWIi4QdoszC2curF692kLzt+a65DQsieDdurjCFixYYI5VxB8vuFAX+tHhhx9u/Y2+BuxDiC0CuxA5DW5rwr5ZPgJwNuLsRIhElJ8+fbotNUHfZd1O+iviHpMvjDHc4AiHPuzdE3UseBczvzN5U69ePRMwGRfe+RkL+yEy+jVyw+CkZCKJJS84z7Bhw/619EQY7pm4rMkPCL3n3hEWXymjn4hgKQucm0y+eEe7rwNrlRJuD0QDMLlBfcOwNEm4fbhHvvDCC7YfH86Lg5b/MfFAKI0VRoUQQgghhBBCiEQkLX76l+cQckloIh9cOrgQ69ata+InjkMEQpxLqcAxhI4SYkoYOUIjLiBEA8QH3Iz87j+8eMPDC44QYX3IO+Aq5JyEoXOujh07mnMoKj27PFIhKo9U4VhEPlynhIEjbOLyjCorIjTuWxyiiKYIGjhRATGWdg07wnYWCD84PXlxC0IKIb4IMKwdiMBOesuWLa38uNpY2gAhhDZgHUEEdiG2BzgXWbsTMQ7nJx8mL3B1M4b5y9jG9Yjgd91117ljjz3W+iv3PJzVhIzHTjJEHUu/RqBknWP6N2ObyRzyx0Xv18ONhTHOmGHsx4IrmnsO50ckRdBFUI2Cewb7s8QGTlIES156xPFAHqxJjOOUSQrKx72fSR4mncJ14J7F2s2UDxcrv4ddnoCjHRHYwzjnpUq0HeIp9wNcsbHHeXDZs8yIEEIIIYQQQgiRLLk2PbhmBNtCpBVFh9YItjYzr+awYEuI3RPEQVyh2bnQd1VYDxRBFRe3EEIIIYQQQgixvZH4uQMghB9HWTx4scrWrAkqskfip0g3cJASGs6b6PPkyROk7l7wBnqW9cAFLoQQQgghhBBCbG8kfoq0ReKnELsmrGPs12gWQgghhBBCCCG2Jym97V0IIYTYViR8CiGEEEIIIYTYUUj8FEIIIYQQQgghhBBCpCUSP4UQQgghhBBCCCGEEGmJxE8hhBBCCCGEEEIIIURaIvFTCCGEEEIIIYQQQgiRlkj8FEIIIYQQQgghhBBCpCW5MjYRbGexdu3aYEuI3ZcyY+oFW5uZXvnNYEsIkW7kzp072BJCCCGEEEIIITKR81MIIYQQQgghhBBCCJGWSPwUQgghhBBCCCGEEEKkJRI/hRBCCCGEEEIIIYQQaYnETyGEEEIIIYQQQgghRFoi8VMIIYQQQgghhBBCCJGWJC1+/vPPP65z587urLPOchUrVnRPPfWUpW/YsME98MAD7uyzz3a1atVy48ePt/RU6dmzp6tSpYqrVq2au+GGG9ycOXOCX5zr3bu3q1Gjhrvgggvc448/bmnk//DDD9u2Z9y4ca5+/fq2/fbbb7tLL73Ujvvggw8sDaLS4+XBi/Cpc9WqVV3lypVd//79Lf3vv/92J554orvwwgvtQ5vccccd7s8//7Tfo/JIBG/Y9/l6nnzySXfRRRe5Sy65xI0ePTpIdW7y5MmuXr16lvdLL70UpDr7fuaZZ2Z9li1bFlmHHcXUqVPd8OHDg2/O/fHHH65FixZWlpo1a1pd4Ndff3VNmzZ1F198saVPmTLF0ul3Dz74oKtevbq7+uqr3fz58y19V4PyLViwIPiWyWWXXbZFP04E7dC2bdvgW+qQ15o1a4Jv/+bZZ591gwcPDr6lPzlR3w8//NDde++9tv3++++7efPm2XYUjEXuDYxl+sO2whhnDG8LvkxR/P77765fv362Ha5vThPORwghhBBCCCGE2JEkLX4iYCGcNWrUyES9559/3o0dO9a9+eabbsCAAa558+bupJNOci1btnRLly4NjkoOBELEhYEDB7p3333X3XLLLfZBTPziiy/c0KFDLe+33nrLxNWPPvrIxEUEQcQ9D+dB5Pv5559dr1693Ouvv+5efPFFE0k5V1R6VB78XbhwoZWJY5544gkT6WDfffe1/PiMGDHC6jxo0KDIPBLx6aefurvuusvNmjUrSHFu1KhRJgBSnhdeeMF17drVrVixwgSue+65x8RfrsmQIUNMIEHg2GeffdzEiROzPkceeWTCOmxP/vrrL2uXjh07up9++ilIdVaGAw44wK7dfffdlyW2INKcfvrpVm/6UpcuXSydPkbbcq6bbrrJhFCRXqxfvz7Y2rU444wzrC/CZ5995hYtWmTbUbzzzjs2GbT//vtvMSmxtXTv3t0ddthhwbetw5cpCsapn5wI1zenCecjhBBCCCGEEELsSJIWP3nwL168uLnzEKEAJ9RXX33lSpYs6a688kpz9CHC4cBMhV9++cUVLFjQ5c2b1+XKlcudeuqprk2bNvbQ/ttvv7m6deuaYHbQQQeZoxExrEiRIu7ggw+2/IF9ESxxj06YMMGdd955Ll++fCYAli5d2n3++eeR6VF5HHjggVbfPfbYwx1++OEmRMQTEigHztcffvghMo9EcNwRRxwRfMvk66+/dpUqVbI2Ofroo61NEDQRM8mLcyOyILhSLspLWizJ1iGnQeygTEcddVSQkgnl2G+//ewv7c31BvoQjk+gzbxbDXEZVyP74TiePXu2Oet2NxCwcbVSl1dffTVI3RL6IWMIBzVuXUDY5voh6tO3vQMYQblx48bmCm7durVbt26dpSfi22+/NVc1kxf0G1i9erXlyXlq165tgjvjt1WrVvY7ojN9zzsaK1SoYGJlvPpQHvoosA/jgPPH25f9EPGbNWtmEw/xwL3ZqVMnV6dOHStfojLHI159cR63a9fO2vKqq67KKu+MGTPsvkYfpO7kwXkRIP24Y5t73pdffumuuOIKqxN/SevTp4/1d8RD6t6kSRM7L/31mmuusbJyHf1EAJNI3bp1s+Opw7Rp0yw9DOXg3ogr+Pbbb3fXX3+9ncf3jTDxyh8uEzBRhSMV17WfXGJyhvsPLnNfXyCfO++8085HG65atcrS33vvPZt44n5Jv2MyI5bs8omFNPLjvG+88UaQKoQQQgghhBBC5AxJi5+33nqrhXMDAh/g9ERQIxQZAQChCpYvX25/k4XQbkQJQt55SCb8kgdnhD9+u/baa20/worJ+5xzzrHvPCyPHDnStnFvIs4i7iEYhYXA/PnzW5mi0qPyQPQpV66chc9ed911FhIfT2D8/vvv3ZgxY1zRokUj80gE4giCVxjEXRyhCAccT/1wbC5ZssQEkcsvv9zC3HGA7rnnnpa+ePFiS0ckfPrpp+08ydYhp0FERjQh7zAIOIjBtDniE85QQGShrRCEENEQVgAnbbi8bKfav3YUDRs2NFHNfxB74OOPP7YP14DQ32HDhpmTMBZErg4dOphbD8GN4xHdypcvb/2cZRFefvll2xcnMMstIE7SD2JD7uOBQPbcc8+ZGMkyE8BfxHXOg9BIu5988sm2XAHw95BDDnHffPONCfKnnXaamzRpUtz6UGc/8cF3+h7HRNWduuH+RfyLgnxxZLMsBWIoxCtzPPdovPriyqY+OKHpa0yyrFy50lzaiLMIsccff7yV28NkCGMKEZExzn2Q43ApI6xyTIMGDaz/evHQg9B64403Wlm5p5CnZ+PGjebsRhTt27dvkBqfTz75xK4/eSFSxobgxyt/uEzkRb967bXX7F4F3FNwnB933HEmJoeh73E/5nyFChUy0RPnOWXgHPRDrk0sqebD/wz6PW2KkMp1RjAWQgghhBBCCCFyipRfeMTDMAINQhqiDGts4nJEvCIUHocejr5UQIwgTJTjedBmm/XuvMOP0HYejHE2EQ6N8AE8nPOAzQM3f3FiAd+9oxA4HrdjVLrfjpcHIDLhdCIEG5ERcNohOvHB3YTQwz6J8kgFxNACBQqYuILwjLCLYxJH5dy5c03sQDBASEJwog0pO2IKLjccgojInnh12BlwbU844QQTwhDwED/DwhX9CFHbi1Ws+RmG9oxN21V45ZVXrN39B7EHEH9wdbIuLG46XHnxnH70IcR7+g/Xm/0YBwjI9E2uuXfEIoDRVoCgTd/IDvrqXnvtZYKYB+HJr0/JOra4iSkrTmwENgRExjpuR/LEdRxVHyYsfJ9jGQv6bqK6I7JmJ8T7MrMfbmGIV2aWdoglXn0R3smfMcVaxfQ9Ji9wtPbo0cPaGBein2CJB0s1MP4R6hCl/TWJhSUqmJA499xz7Tv3K9rD4+9XJUqUyHJWRkHfYKKJiY5ixYr9a//syk/bMaYQ1RGRv/vuu8hyA22Kcxfoi+SHoFqmTBlzk+fJk8edf/759nuYVPOhPRDNWbKD/yP0L/qaEEIIIYQQQgiRU6QkfuJa4sEfpxWOJjj00ENtzU/W6ySkFXEqLBwmAw/sPFgjFuEGxAVFKDluJ4Qu1hGdPn26hVPiPPMgiHAMgoZ3iwIiAQ5MD05B9o1Kj8qDc+JCQjzBVVW2bFlzvUF4zU8EEEJ4995778g8UgWRA1ccoi7tmzt3bisH7Y2QyV/EArYRqRBEcNAhnCGgIUzwsp1EddgZINTi/iQcHvEHdy9rneLuQywi/J1QaFxmCEcsB4D700N70sa7EwiTOIsR9/kwVgh7joXrFguuRUQkfsNJ6wkL7Ah89MfsoM/EwngNn4t+h4CPmIYDmnMjSCF8ssQE21H1oYxcVxyBCFiIfonqnswkCUJbLFFljiVefRlHuEV9eXCAIsIiMuOQRITDnYhDNQr6J22DaJnItQrhcnrx1oM7Olm4HyYiu/IznrgOjCnuGV6QjYJrw/0sDPfJcH3ivWAr1Xwg9lpyfYUQQgghhBBCiJwiafGTEFXWkCP8E0GCMEjENcIbSUPw4OEbcSGRayoKBB5C34EHZz7HHHOMhZYibDz00EMm9sWCA4xQTMJRvUCA8Mf6n4SMEyqO04uXeUSlR+WBQxIXE84lQmMJAfZuviii8kgVHFOIhDjTfvzxR/uO0xYxgXB4Qt8pE3nhxiIMlbeFUw+cdoQan3LKKVtVh+0J4fzerYtoi2MPsRwB2y+rQF0RVhDT6Ft+aQNE28KFC5tgujtB2DShwLh2cTPz1vrYsOUoEBwJi2aZANaw9C5iBHovcCFweycgeaTyQitcpb59Z86cae1+7LHHmnDVr18/W9qCMGrCoDkvvyWqDy6+Rx991Po8Qlwydc+pMicDY8ivK4mDmjU36YsIywh3ONlZvzjeOr2If+zLZAITQLygi/Eddi6HXcn0UyYdmMQB7pnhyZucJFH5KZOfSGA/2oqJHl/WcJkTwVq8TFLR3tyr402ipJoPbYhLmH7NeRGVaaNwn4jaFkIIIYQQQgghkiFp8dO/GIQHVcJX+RDai5jBy1gQPBAiWGsyVXGKl4MQ1sm5CFW9+eabTfDBBYiDjNBuBFX/wZ3pQWxBxKAMHtxnrBvJmnecC5cqTqao9Kg8eNEKdSEEHUGB7wiKiYjKI1UIDcelibiLqEn74sRiWQAEDoRR3F6EFiPo8J1QVL7jnsWRhoiwNXXYntBvEI9wfbKGIm9vR7RmTUBEPNLZpr64GQkVRpCmrgjkuGF3N3AWIl5yTbgOrMvK9U2GRo0ambjPkgW4XhHfWD+TNkJUI/Qd96K/pohHrDGZLLgYEdc5D+5l2h3REqELgYvz4sxDAGWMQqL6kE4oM0sXQDJ1z6kyJwPjhHoxrnBI8vIg3K2kt2/f3s6JqBd22QITDFwHQuRZXgJ3I+OQduE+yEQQbcV9KwwTKqw7ynmpJ2N5exBVfl8mys8yGv5FRUzSUB8msnBastZyduBg58VL3F/4y7WNnZBKNR/u9/Qr+gVrH3PvJMyetmI5EYjaFkIIIYQQQgghkiHXpgfSf8UY+rU2hdidKTOmXrC1memVNwvn6QoObcT33Yndscz/NXAXs1QAL3vC7YoIzVqupUqVCvbY+RB5IIQQQgghhBBChJH4uQNYtGiRvYQoHrhdt2ZNUJE9/0Xxk2UEGL+sR7m7sDuW+b8KL1XCKY8LGVc27vZdCYmfQgghhBBCCCFikfgp0pb/qvNTiP8qEj+FEEIIIYQQQsSS0tvehRBCCCGEEEIIIYQQYndB4qcQQgghhBBCCCGEECItkfgp0pbYEHeFvAshhBBCCCGEEEL8t9Can0IIIdICrfkphBBCCCGEECIWOT+FEEIIIYQQQgghhBBpicRPIYQQQgghhBBCCCFEWiLxUwghhBBCCCGEEEIIkZbEXfNTCCGEEEIIIYQQQgghdnfk/BRCCCGEEEIIIYQQQqQlEj+FEEIIIYQQQgghhBBpiHP/B3Ejbd+nhAlNAAAAAElFTkSuQmCC\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "flFp5C_ptODw"
   },
   "source": [
    "The following code example demonstrates this concept in action. During this example we will:\n",
    "\n",
    "1. Count the number of words in each sentence of our Hansard sample\n",
    "2. Group the data by decade\n",
    "3. Sum the total number of words per each decade based on our sentence-level word counts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Groupby in Action"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let me try to translate groupby() into plain English.  Imagine that I'm counting easter eggs.  I can count all of my eggs together, and I get one number.  \n",
    "\n",
    "But what if I want to count all of the green eggs, the red eggs, and the yellow eggs, and have separate counts for each?  Let's say, to make things easy, I sort the eggs by color, putting each color of eggs into a unique basket.  Now I have a basked of red eggs, and if I count all of the eggs in the basket, I know how many red eggs there are. \n",
    "\n",
    "Groupby() is like having a series of baskets.  \n",
    "\n",
    "Groupby() tells the computer to \"group\" your data by some variable (in the case of my example, you're \"grouping\" eggs by \"color\") before you apply any other function, for example, counting them.  \n",
    "\n",
    "With groupby(), the computer can count all the words per date or words per speaker.\n",
    "\n",
    "Without groupby(), all the computer knows how to do is to count everything -- like someone who can only count the total number of eggs, but not the number of red eggs and yellow eggs separately. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Groupby() is a function that takes as its argument the column that you are grouping the data by.  \n",
    "\n",
    "If we were counting eggs, we would groupby(['egg_color']).\n",
    "\n",
    "Because we are counting words per month, we will group_by(['month_year']).\n",
    "\n",
    "Notice please that the name of the column is in both brackets [] and single quotation marks ''.  The brackets tell Python that you are want part of the dataframe WHICH corresponds to an index.  The quotation marks tell Python that you want the column of the dataframe that corresponds to the string 'month_year.'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Groupby() is usually strung along with another command -- which tells Python what to do with the data once you group it. In this case, we want to total the number of words, so we will string together \n",
    "\n",
    "    .groupby().sum().\n",
    "\n",
    "Groupby() is also typically followed by a command that says that Python should create a new database with the grouped data: \n",
    "\n",
    "    .copy().  \n",
    "    \n",
    "Without `.copy`, the new variable you create is actually a reference to the original data frame, not a new subset that just contains word_count. You don't need to understand why, but you should know that this convention will help your data to retain its integrity.  Use .groupby().sum().copy()."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "Gg_t4VuJtOGJ",
    "outputId": "7726c60f-8dcc-4185-95c4-6a367ea24772"
   },
   "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>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1980-01-01</th>\n",
       "      <td>1043116</td>\n",
       "      <td>14107500</td>\n",
       "      <td>7125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-02-01</th>\n",
       "      <td>1664367</td>\n",
       "      <td>23039280</td>\n",
       "      <td>23272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-03-01</th>\n",
       "      <td>1655767</td>\n",
       "      <td>23120460</td>\n",
       "      <td>35031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-04-01</th>\n",
       "      <td>1508604</td>\n",
       "      <td>18883260</td>\n",
       "      <td>38148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-05-01</th>\n",
       "      <td>2193062</td>\n",
       "      <td>32808600</td>\n",
       "      <td>82850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-06-01</th>\n",
       "      <td>2373233</td>\n",
       "      <td>29911448</td>\n",
       "      <td>90276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-07-01</th>\n",
       "      <td>1864321</td>\n",
       "      <td>27505968</td>\n",
       "      <td>96852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-08-01</th>\n",
       "      <td>1607097</td>\n",
       "      <td>19625536</td>\n",
       "      <td>78976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-09-01</th>\n",
       "      <td>2703096</td>\n",
       "      <td>37992668</td>\n",
       "      <td>171999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-10-01</th>\n",
       "      <td>2995352</td>\n",
       "      <td>36895292</td>\n",
       "      <td>185590</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>101 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            word_count      year   month\n",
       "month_year                              \n",
       "1980-01-01     1043116  14107500    7125\n",
       "1980-02-01     1664367  23039280   23272\n",
       "1980-03-01     1655767  23120460   35031\n",
       "1980-04-01     1508604  18883260   38148\n",
       "1980-05-01     2193062  32808600   82850\n",
       "...                ...       ...     ...\n",
       "1988-06-01     2373233  29911448   90276\n",
       "1988-07-01     1864321  27505968   96852\n",
       "1988-08-01     1607097  19625536   78976\n",
       "1988-09-01     2703096  37992668  171999\n",
       "1988-10-01     2995352  36895292  185590\n",
       "\n",
       "[101 rows x 3 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words_per_month_counts = congress.groupby(['month_year']).sum()\n",
    "words_per_month_counts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you're wondering what happened, you might scroll up to the last instance of the table and notice that there was already a word_count column that had the number of words in every speech.  There were many speeches per month_year.  \n",
    "\n",
    "In the new word_count column, we have put all the speeches that happened in a single month_year into one basket.  For each basket, we have added together all of the word_count numbers from the individual speeches.  \n",
    "\n",
    "The sum of all the word_count numbers is the number of words that were spoken in every month.\n",
    "\n",
    "See for yourself -- look at the table above and notice that the word_count column numbers are around 100 words per speech.  Look at the word_count column in words_per_month_counts, and see that the word_count column is closer to 1 or 2 million in many cases.\n",
    "\n",
    "We have grouped the data by month and taken a count.  .groupby().sum().copy().\n",
    "\n",
    "--\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, let's play with making our dataframe more elegant.\n",
    "\n",
    "Let's say that we only want to look at one column here -- word_count.  \n",
    "\n",
    "We use double backets to select the column from the dataframe, while keeping the data classified as a dataframe (not a series object), because dataframes are more flexible.  For more on dataframes and series objects, see  https://www.educative.io/edpresso/series-vs-dataframe-in-pandas. \n",
    "\n",
    "Again, you don't really need to know about dataframes vs. series objects, but you should remember that square brackets will be more flexible when we're working with tabular data in pandas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "IwselOIltOGQ",
    "outputId": "b012d442-1e9a-49c4-8051-6766536f381f"
   },
   "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>word_count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month_year</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1980-01-01</th>\n",
       "      <td>1043116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-02-01</th>\n",
       "      <td>1664367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-03-01</th>\n",
       "      <td>1655767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-04-01</th>\n",
       "      <td>1508604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1980-05-01</th>\n",
       "      <td>2193062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-06-01</th>\n",
       "      <td>2373233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-07-01</th>\n",
       "      <td>1864321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-08-01</th>\n",
       "      <td>1607097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-09-01</th>\n",
       "      <td>2703096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1988-10-01</th>\n",
       "      <td>2995352</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>101 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            word_count\n",
       "month_year            \n",
       "1980-01-01     1043116\n",
       "1980-02-01     1664367\n",
       "1980-03-01     1655767\n",
       "1980-04-01     1508604\n",
       "1980-05-01     2193062\n",
       "...                ...\n",
       "1988-06-01     2373233\n",
       "1988-07-01     1864321\n",
       "1988-08-01     1607097\n",
       "1988-09-01     2703096\n",
       "1988-10-01     2995352\n",
       "\n",
       "[101 rows x 1 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "just_counts = words_per_month_counts[['word_count']].copy() \n",
    "just_counts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You'll notice that the month_year column is in bold.\n",
    "\n",
    "By putting the month_year column in bold, pandas is telling us that it used the month_year column to \"group\" the other data.\n",
    "\n",
    "Technically, the month_year column is an \"index\" -- we might also call it a 'grouping' or a 'label' -- not just another column of data.\n",
    "\n",
    "But we're going to want to treat month_year like another kind of data soon.  We're going to want to put it on the x axis.  So we need to take month_year out of \"index\" format and make it into a column.  This process is called \"resetting the index.\"  Again you don't need to know much about what's being done, but you need to notice that resetting the index can solve the problem of working with the data that was the \"group\" after you run a groupby() command.   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "2ovTmUqmtOGU",
    "outputId": "c95f7c2e-e9b7-45fe-acf3-dc9291b58779"
   },
   "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>month_year</th>\n",
       "      <th>word_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1980-01-01</td>\n",
       "      <td>1043116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1980-02-01</td>\n",
       "      <td>1664367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1980-03-01</td>\n",
       "      <td>1655767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1980-04-01</td>\n",
       "      <td>1508604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1980-05-01</td>\n",
       "      <td>2193062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>1988-06-01</td>\n",
       "      <td>2373233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>1988-07-01</td>\n",
       "      <td>1864321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>1988-08-01</td>\n",
       "      <td>1607097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>1988-09-01</td>\n",
       "      <td>2703096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>1988-10-01</td>\n",
       "      <td>2995352</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>101 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     month_year  word_count\n",
       "0    1980-01-01     1043116\n",
       "1    1980-02-01     1664367\n",
       "2    1980-03-01     1655767\n",
       "3    1980-04-01     1508604\n",
       "4    1980-05-01     2193062\n",
       "..          ...         ...\n",
       "96   1988-06-01     2373233\n",
       "97   1988-07-01     1864321\n",
       "98   1988-08-01     1607097\n",
       "99   1988-09-01     2703096\n",
       "100  1988-10-01     2995352\n",
       "\n",
       "[101 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "just_counts.reset_index(level=0, inplace=True)\n",
    "display(just_counts)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "AfEklsZ9tOGz"
   },
   "source": [
    "Just_counts is a dataframe grouped by month_year with column word_count.  It's perfect for showing change over time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we wanted to start the whole process over again, and count by year, we could do that by first extracting the year column. \n",
    "\n",
    "Then we'd groupby() year and sum() and copy()."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "Lp3f836qtOGb"
   },
   "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>month_year</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1980-01-01</td>\n",
       "      <td>1043116</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1980-02-01</td>\n",
       "      <td>1664367</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1980-03-01</td>\n",
       "      <td>1655767</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1980-04-01</td>\n",
       "      <td>1508604</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1980-05-01</td>\n",
       "      <td>2193062</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1980-06-01</td>\n",
       "      <td>3330363</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1980-07-01</td>\n",
       "      <td>1957658</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1980-08-01</td>\n",
       "      <td>1542431</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1980-09-01</td>\n",
       "      <td>2672507</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1980-10-01</td>\n",
       "      <td>370352</td>\n",
       "      <td>1980</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   month_year  word_count  year\n",
       "0  1980-01-01     1043116  1980\n",
       "1  1980-02-01     1664367  1980\n",
       "2  1980-03-01     1655767  1980\n",
       "3  1980-04-01     1508604  1980\n",
       "4  1980-05-01     2193062  1980\n",
       "5  1980-06-01     3330363  1980\n",
       "6  1980-07-01     1957658  1980\n",
       "7  1980-08-01     1542431  1980\n",
       "8  1980-09-01     2672507  1980\n",
       "9  1980-10-01      370352  1980"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "just_counts['year'] = pd.to_datetime(just_counts['month_year']).dt.year # now we can make a column for just the year \n",
    "\n",
    "just_counts[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What if we wanted to find the top words per year?\n",
    "\n",
    "First, we'd make sure there was a year column in all_data."
   ]
  },
  {
   "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>speech</th>\n",
       "      <th>date</th>\n",
       "      <th>speaker</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2329890</th>\n",
       "      <td>Mr. Speaker. we in Delaware are proud of the o...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. EVANS of Delaware</td>\n",
       "      <td>122</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329891</th>\n",
       "      <td>Mr. Speaker. it is logical for Americans to be...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. DERWINSKI</td>\n",
       "      <td>82</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329892</th>\n",
       "      <td>The Chair has examined the Journal of the last...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>The SPEAKER pro tempore</td>\n",
       "      <td>32</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329893</th>\n",
       "      <td>Mr. Speaker. I ask unanimous consent that the ...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. WHITE</td>\n",
       "      <td>36</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329894</th>\n",
       "      <td>Is there objection to the request of the gentl...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>The SPEAKER pro tempore</td>\n",
       "      <td>11</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329895</th>\n",
       "      <td>Mr. Speaker. it is obvious that in Iran today ...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. LUNGREN</td>\n",
       "      <td>210</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329896</th>\n",
       "      <td>Mr. Speaker. for 60 days the American people a...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. ROTH</td>\n",
       "      <td>206</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329897</th>\n",
       "      <td>The hour of 12 noon having arrived. pursuant t...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>The SPEAKER pro tempore</td>\n",
       "      <td>46</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329898</th>\n",
       "      <td>Mr. Speaker. the dawning of 1980 was not the f...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. BIAGGI</td>\n",
       "      <td>604</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329899</th>\n",
       "      <td>Mr. Speaker. I call for an allout information ...</td>\n",
       "      <td>1980-01-03</td>\n",
       "      <td>Mr. DERWINSKI</td>\n",
       "      <td>284</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    speech        date  \\\n",
       "2329890  Mr. Speaker. we in Delaware are proud of the o...  1980-01-03   \n",
       "2329891  Mr. Speaker. it is logical for Americans to be...  1980-01-03   \n",
       "2329892  The Chair has examined the Journal of the last...  1980-01-03   \n",
       "2329893  Mr. Speaker. I ask unanimous consent that the ...  1980-01-03   \n",
       "2329894  Is there objection to the request of the gentl...  1980-01-03   \n",
       "2329895  Mr. Speaker. it is obvious that in Iran today ...  1980-01-03   \n",
       "2329896  Mr. Speaker. for 60 days the American people a...  1980-01-03   \n",
       "2329897  The hour of 12 noon having arrived. pursuant t...  1980-01-03   \n",
       "2329898  Mr. Speaker. the dawning of 1980 was not the f...  1980-01-03   \n",
       "2329899  Mr. Speaker. I call for an allout information ...  1980-01-03   \n",
       "\n",
       "                         speaker  word_count  year  month  month_year  \n",
       "2329890    Mr. EVANS of Delaware         122  1980      1  1980-01-01  \n",
       "2329891            Mr. DERWINSKI          82  1980      1  1980-01-01  \n",
       "2329892  The SPEAKER pro tempore          32  1980      1  1980-01-01  \n",
       "2329893                Mr. WHITE          36  1980      1  1980-01-01  \n",
       "2329894  The SPEAKER pro tempore          11  1980      1  1980-01-01  \n",
       "2329895              Mr. LUNGREN         210  1980      1  1980-01-01  \n",
       "2329896                 Mr. ROTH         206  1980      1  1980-01-01  \n",
       "2329897  The SPEAKER pro tempore          46  1980      1  1980-01-01  \n",
       "2329898               Mr. BIAGGI         604  1980      1  1980-01-01  \n",
       "2329899            Mr. DERWINSKI         284  1980      1  1980-01-01  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress['year'] = pd.to_datetime(congress['month_year']).dt.year \n",
    "congress[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then, we'd groupby().sum().copy() with the year column instead.  \n",
    "\n",
    "We'd select just the word_count column, as before...\n",
    "\n",
    "... and reset the index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>year</th>\n",
       "      <th>word_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1980</td>\n",
       "      <td>21180872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1981</td>\n",
       "      <td>19532362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1982</td>\n",
       "      <td>19821530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1983</td>\n",
       "      <td>20990969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1984</td>\n",
       "      <td>19329275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1985</td>\n",
       "      <td>22472067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1986</td>\n",
       "      <td>20966815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1987</td>\n",
       "      <td>23052012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1988</td>\n",
       "      <td>19841019</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   year  word_count\n",
       "0  1980    21180872\n",
       "1  1981    19532362\n",
       "2  1982    19821530\n",
       "3  1983    20990969\n",
       "4  1984    19329275\n",
       "5  1985    22472067\n",
       "6  1986    20966815\n",
       "7  1987    23052012\n",
       "8  1988    19841019"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "word_per_year_counts = congress.groupby(['year']).sum().copy() \n",
    "\n",
    "per_year_counts = word_per_year_counts[['word_count']].copy()\n",
    "\n",
    "per_year_counts.reset_index(level=0, inplace=True)\n",
    "\n",
    "per_year_counts[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Out of curiosity, what was the maximum number of words spoken in Congress per year?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23052012"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(per_year_counts['word_count'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What if I want to know what year that was?  I can use .query(), and specify in the parentheses the factors I'm trying to match.  \n",
    "\n",
    "In the brackets that follow, I can specify what column I want to look up.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series([], Name: year, dtype: int64)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "per_year_counts.query('word_count>29212581')['year']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following line does the same thing:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23052012"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(per_year_counts['word_count'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The maximum number of words per year happened in 1995."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Groupby() Speaker"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, let's use groupby() to \"group\" by speakers in the Congress data, using the 'speaker' column.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's count how many times the word 'crime' appears in Congress."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>speech</th>\n",
       "      <th>date</th>\n",
       "      <th>speaker</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_year</th>\n",
       "      <th>vocabcount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2744529</th>\n",
       "      <td>Mr. President. I thank the distinguished Senat...</td>\n",
       "      <td>1982-09-30</td>\n",
       "      <td>Mr. CHILES</td>\n",
       "      <td>4392</td>\n",
       "      <td>1982</td>\n",
       "      <td>9</td>\n",
       "      <td>1982-09-01</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2930424</th>\n",
       "      <td>Mr. President. the amendment is designed to he...</td>\n",
       "      <td>1984-01-31</td>\n",
       "      <td>Mr. HUDDLESTON</td>\n",
       "      <td>2402</td>\n",
       "      <td>1984</td>\n",
       "      <td>1</td>\n",
       "      <td>1984-01-01</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2548179</th>\n",
       "      <td>Mr. President. the bill that I have just sent ...</td>\n",
       "      <td>1981-07-09</td>\n",
       "      <td>Mr. BIDEN</td>\n",
       "      <td>3608</td>\n",
       "      <td>1981</td>\n",
       "      <td>7</td>\n",
       "      <td>1981-07-01</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2528729</th>\n",
       "      <td>Mr. President. first let me thank the Senator ...</td>\n",
       "      <td>1981-05-20</td>\n",
       "      <td>Mr. NUNN</td>\n",
       "      <td>3667</td>\n",
       "      <td>1981</td>\n",
       "      <td>5</td>\n",
       "      <td>1981-05-01</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3407624</th>\n",
       "      <td>Mr. President. the subject of my speech today ...</td>\n",
       "      <td>1987-06-19</td>\n",
       "      <td>Mr. BYRD</td>\n",
       "      <td>8891</td>\n",
       "      <td>1987</td>\n",
       "      <td>6</td>\n",
       "      <td>1987-06-01</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2663843</th>\n",
       "      <td>Mr. President. many times during my service he...</td>\n",
       "      <td>1982-05-12</td>\n",
       "      <td>Mr. HEFLIN</td>\n",
       "      <td>1583</td>\n",
       "      <td>1982</td>\n",
       "      <td>5</td>\n",
       "      <td>1982-05-01</td>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2517058</th>\n",
       "      <td>Mr. President. today I am introducing the Nati...</td>\n",
       "      <td>1981-04-09</td>\n",
       "      <td>Mr. HEPLIN</td>\n",
       "      <td>2675</td>\n",
       "      <td>1981</td>\n",
       "      <td>4</td>\n",
       "      <td>1981-04-01</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2638317</th>\n",
       "      <td>Mr. President. FBI crime reports state that a ...</td>\n",
       "      <td>1982-02-24</td>\n",
       "      <td>Mr. SYMMS</td>\n",
       "      <td>2713</td>\n",
       "      <td>1982</td>\n",
       "      <td>2</td>\n",
       "      <td>1982-02-01</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2548181</th>\n",
       "      <td>Mr. President. I thank my colleague from Delaw...</td>\n",
       "      <td>1981-07-09</td>\n",
       "      <td>Mr. NUNN</td>\n",
       "      <td>2161</td>\n",
       "      <td>1981</td>\n",
       "      <td>7</td>\n",
       "      <td>1981-07-01</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3028226</th>\n",
       "      <td>Mr. President. S. 2423. the Victims of Crime A...</td>\n",
       "      <td>1984-08-10</td>\n",
       "      <td>Mr. HEINZ</td>\n",
       "      <td>1916</td>\n",
       "      <td>1984</td>\n",
       "      <td>8</td>\n",
       "      <td>1984-08-01</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    speech        date  \\\n",
       "2744529  Mr. President. I thank the distinguished Senat...  1982-09-30   \n",
       "2930424  Mr. President. the amendment is designed to he...  1984-01-31   \n",
       "2548179  Mr. President. the bill that I have just sent ...  1981-07-09   \n",
       "2528729  Mr. President. first let me thank the Senator ...  1981-05-20   \n",
       "3407624  Mr. President. the subject of my speech today ...  1987-06-19   \n",
       "2663843  Mr. President. many times during my service he...  1982-05-12   \n",
       "2517058  Mr. President. today I am introducing the Nati...  1981-04-09   \n",
       "2638317  Mr. President. FBI crime reports state that a ...  1982-02-24   \n",
       "2548181  Mr. President. I thank my colleague from Delaw...  1981-07-09   \n",
       "3028226  Mr. President. S. 2423. the Victims of Crime A...  1984-08-10   \n",
       "\n",
       "                speaker  word_count  year  month  month_year  vocabcount  \n",
       "2744529      Mr. CHILES        4392  1982      9  1982-09-01          68  \n",
       "2930424  Mr. HUDDLESTON        2402  1984      1  1984-01-01          58  \n",
       "2548179       Mr. BIDEN        3608  1981      7  1981-07-01          52  \n",
       "2528729        Mr. NUNN        3667  1981      5  1981-05-01          51  \n",
       "3407624        Mr. BYRD        8891  1987      6  1987-06-01          51  \n",
       "2663843      Mr. HEFLIN        1583  1982      5  1982-05-01          42  \n",
       "2517058      Mr. HEPLIN        2675  1981      4  1981-04-01          40  \n",
       "2638317       Mr. SYMMS        2713  1982      2  1982-02-01          40  \n",
       "2548181        Mr. NUNN        2161  1981      7  1981-07-01          37  \n",
       "3028226       Mr. HEINZ        1916  1984      8  1984-08-01          37  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congress['vocabcount'] = congress['speech'].str.count('crime')  # Create a new column for the count of the word democracydis\n",
    "congress.nlargest(10, ['vocabcount'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The speeches listed above have extraordinarily large numbers of mentions of the word 'crime' -- 99 mentions in Mr. Hatch's 2000 speech.  If we were interested in studying crime in Congress, these should be the first places we start reading."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Use .aggregate() to group and sum at the same time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll use a new function with .groupby():\n",
    "\n",
    "    .aggregate()\n",
    "    \n",
    "Aggregate is more flexible than .sum() in that it allows us to specify which column to \"group by\" and which column to \"aggregate.\"  \n",
    "\n",
    "In the case of speakers and words, we want to \"group by\" the speaker, and \"aggregate\" how many times overall each speaker invoked the word 'crime.'\n",
    "\n",
    "The way we do this has a peculiar grammar that you'll simply need to memorize, take notes on, and/or cut and paste:\n",
    "\n",
    "    .groupby().aggregate({‘new_column_name’: np.sum}) \n",
    "    \n",
    "One advantage of this method is that you can string together multiple charges, each a pair of commands enclosed in curly brackets:\n",
    "\n",
    "    .aggregate(\n",
    "        {‘column_name1’: np.sum}\n",
    "        {‘column_name2’: np.mean}\n",
    "        )\n",
    "\n",
    "For now, however, we're just going to ask the code to *group by* \"speaker\" and to *aggregate* a new per-speaker column called \"vocabcount\" where we count how many times each speaker mentioned democracy across all of their speeches."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>vocabcount</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>speaker</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Mr. CHILES</th>\n",
       "      <td>975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. PROXMIRE</th>\n",
       "      <td>919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. HUGHES</th>\n",
       "      <td>751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. THURMOND</th>\n",
       "      <td>690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. BIDEN</th>\n",
       "      <td>643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. NUNN</th>\n",
       "      <td>594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. SPECTER</th>\n",
       "      <td>580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. KENNEDY</th>\n",
       "      <td>519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. DAMATO</th>\n",
       "      <td>493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr. HATCH</th>\n",
       "      <td>492</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              vocabcount\n",
       "speaker                 \n",
       "Mr. CHILES           975\n",
       "Mr. PROXMIRE         919\n",
       "Mr. HUGHES           751\n",
       "Mr. THURMOND         690\n",
       "Mr. BIDEN            643\n",
       "Mr. NUNN             594\n",
       "Mr. SPECTER          580\n",
       "Mr. KENNEDY          519\n",
       "Mr. DAMATO           493\n",
       "Mr. HATCH            492"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "congressgrouped = congress.groupby('speaker').aggregate({'vocabcount': np.sum})  # Sum the rows by democracy_count\n",
    "topvocabspeakers = congressgrouped.nlargest(10, ['vocabcount'])\n",
    "topvocabspeakers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Mr. CHILES', 'Mr. PROXMIRE', 'Mr. HUGHES', 'Mr. THURMOND', 'Mr. BIDEN',\n",
       "       'Mr. NUNN', 'Mr. SPECTER', 'Mr. KENNEDY', 'Mr. DAMATO', 'Mr. HATCH'],\n",
       "      dtype='object', name='speaker')"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topvocabspeakers.index"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Navigate back to the top 5 longest speeches of a particular speaker"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's play with the data some more.  \n",
    "\n",
    "What if we want to get all the speeches by the speakers who spoke the most about crime?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we'll return to the 1980s data in eighties_data, but find only the speeches where the name in the 'speaker' column matches the list called top_speaker."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>speech</th>\n",
       "      <th>date</th>\n",
       "      <th>speaker</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_year</th>\n",
       "      <th>vocabcount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2329939</th>\n",
       "      <td>Mr. President. on December 20. 1979. the Senat...</td>\n",
       "      <td>1980-01-10</td>\n",
       "      <td>Mr. THURMOND</td>\n",
       "      <td>116</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2330056</th>\n",
       "      <td>I wish to compliment the able Senator from Vir...</td>\n",
       "      <td>1980-01-22</td>\n",
       "      <td>Mr. THURMOND</td>\n",
       "      <td>41</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2330058</th>\n",
       "      <td>Mr. President. the deception. cruelty. and bru...</td>\n",
       "      <td>1980-01-22</td>\n",
       "      <td>Mr. THURMOND</td>\n",
       "      <td>1074</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2330061</th>\n",
       "      <td>Mr. President. the American Legion has approxi...</td>\n",
       "      <td>1980-01-22</td>\n",
       "      <td>Mr. THURMOND</td>\n",
       "      <td>191</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2330154</th>\n",
       "      <td>Mr. President. I support the agreement on trad...</td>\n",
       "      <td>1980-01-22</td>\n",
       "      <td>Mr. KENNEDY</td>\n",
       "      <td>1259</td>\n",
       "      <td>1980</td>\n",
       "      <td>1</td>\n",
       "      <td>1980-01-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3623175</th>\n",
       "      <td>Mr. Speaker. I am going to be very brief also....</td>\n",
       "      <td>1988-10-21</td>\n",
       "      <td>Mr. HUGHES</td>\n",
       "      <td>530</td>\n",
       "      <td>1988</td>\n",
       "      <td>10</td>\n",
       "      <td>1988-10-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3623249</th>\n",
       "      <td>Mr. Speaker. section 6182 of the AntiDrug Abus...</td>\n",
       "      <td>1988-10-21</td>\n",
       "      <td>Mr. HUGHES</td>\n",
       "      <td>484</td>\n",
       "      <td>1988</td>\n",
       "      <td>10</td>\n",
       "      <td>1988-10-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3623481</th>\n",
       "      <td>Mr. Speaker. over the Fourth of July weekend. ...</td>\n",
       "      <td>1988-10-21</td>\n",
       "      <td>Mr. KENNEDY</td>\n",
       "      <td>410</td>\n",
       "      <td>1988</td>\n",
       "      <td>10</td>\n",
       "      <td>1988-10-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3623519</th>\n",
       "      <td>Mr. Speaker. this Saturday. the parishioners o...</td>\n",
       "      <td>1988-10-21</td>\n",
       "      <td>Mr. HUGHES</td>\n",
       "      <td>242</td>\n",
       "      <td>1988</td>\n",
       "      <td>10</td>\n",
       "      <td>1988-10-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3623587</th>\n",
       "      <td>Mr. Speaker. as part of the AntiDrug Abuse Act...</td>\n",
       "      <td>1988-10-21</td>\n",
       "      <td>Mr. HUGHES</td>\n",
       "      <td>154</td>\n",
       "      <td>1988</td>\n",
       "      <td>10</td>\n",
       "      <td>1988-10-01</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>49973 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    speech        date  \\\n",
       "2329939  Mr. President. on December 20. 1979. the Senat...  1980-01-10   \n",
       "2330056  I wish to compliment the able Senator from Vir...  1980-01-22   \n",
       "2330058  Mr. President. the deception. cruelty. and bru...  1980-01-22   \n",
       "2330061  Mr. President. the American Legion has approxi...  1980-01-22   \n",
       "2330154  Mr. President. I support the agreement on trad...  1980-01-22   \n",
       "...                                                    ...         ...   \n",
       "3623175  Mr. Speaker. I am going to be very brief also....  1988-10-21   \n",
       "3623249  Mr. Speaker. section 6182 of the AntiDrug Abus...  1988-10-21   \n",
       "3623481  Mr. Speaker. over the Fourth of July weekend. ...  1988-10-21   \n",
       "3623519  Mr. Speaker. this Saturday. the parishioners o...  1988-10-21   \n",
       "3623587  Mr. Speaker. as part of the AntiDrug Abuse Act...  1988-10-21   \n",
       "\n",
       "              speaker  word_count  year  month  month_year  vocabcount  \n",
       "2329939  Mr. THURMOND         116  1980      1  1980-01-01           0  \n",
       "2330056  Mr. THURMOND          41  1980      1  1980-01-01           0  \n",
       "2330058  Mr. THURMOND        1074  1980      1  1980-01-01           1  \n",
       "2330061  Mr. THURMOND         191  1980      1  1980-01-01           0  \n",
       "2330154   Mr. KENNEDY        1259  1980      1  1980-01-01           0  \n",
       "...               ...         ...   ...    ...         ...         ...  \n",
       "3623175    Mr. HUGHES         530  1988     10  1988-10-01           0  \n",
       "3623249    Mr. HUGHES         484  1988     10  1988-10-01           0  \n",
       "3623481   Mr. KENNEDY         410  1988     10  1988-10-01           0  \n",
       "3623519    Mr. HUGHES         242  1988     10  1988-10-01           0  \n",
       "3623587    Mr. HUGHES         154  1988     10  1988-10-01           0  \n",
       "\n",
       "[49973 rows x 8 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_speaker_speeches = congress[congress['speaker'].isin(topvocabspeakers.index)] # Filter for only this specific speaker's speeches\n",
    "top_speaker_speeches "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, let's use the function \n",
    "\n",
    "    .nlargest() \n",
    "\n",
    "to find the longest speeches by the size of the column 'word_count'."
   ]
  },
  {
   "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>speech</th>\n",
       "      <th>date</th>\n",
       "      <th>speaker</th>\n",
       "      <th>word_count</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>month_year</th>\n",
       "      <th>vocabcount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2744529</th>\n",
       "      <td>Mr. President. I thank the distinguished Senat...</td>\n",
       "      <td>1982-09-30</td>\n",
       "      <td>Mr. CHILES</td>\n",
       "      <td>4392</td>\n",
       "      <td>1982</td>\n",
       "      <td>9</td>\n",
       "      <td>1982-09-01</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2548179</th>\n",
       "      <td>Mr. President. the bill that I have just sent ...</td>\n",
       "      <td>1981-07-09</td>\n",
       "      <td>Mr. BIDEN</td>\n",
       "      <td>3608</td>\n",
       "      <td>1981</td>\n",
       "      <td>7</td>\n",
       "      <td>1981-07-01</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2528729</th>\n",
       "      <td>Mr. President. first let me thank the Senator ...</td>\n",
       "      <td>1981-05-20</td>\n",
       "      <td>Mr. NUNN</td>\n",
       "      <td>3667</td>\n",
       "      <td>1981</td>\n",
       "      <td>5</td>\n",
       "      <td>1981-05-01</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2548181</th>\n",
       "      <td>Mr. President. I thank my colleague from Delaw...</td>\n",
       "      <td>1981-07-09</td>\n",
       "      <td>Mr. NUNN</td>\n",
       "      <td>2161</td>\n",
       "      <td>1981</td>\n",
       "      <td>7</td>\n",
       "      <td>1981-07-01</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2675104</th>\n",
       "      <td>Mr. President. for the last week Senator NUNN ...</td>\n",
       "      <td>1982-05-26</td>\n",
       "      <td>Mr. CHILES</td>\n",
       "      <td>1096</td>\n",
       "      <td>1982</td>\n",
       "      <td>5</td>\n",
       "      <td>1982-05-01</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2744279</th>\n",
       "      <td>Mr. President. this bill is unique in the hist...</td>\n",
       "      <td>1982-09-30</td>\n",
       "      <td>Mr. SPECTER</td>\n",
       "      <td>4344</td>\n",
       "      <td>1982</td>\n",
       "      <td>9</td>\n",
       "      <td>1982-09-01</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2682023</th>\n",
       "      <td>Madam President. over the past weeks. together...</td>\n",
       "      <td>1982-06-09</td>\n",
       "      <td>Mr. CHILES</td>\n",
       "      <td>1763</td>\n",
       "      <td>1982</td>\n",
       "      <td>6</td>\n",
       "      <td>1982-06-01</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2744526</th>\n",
       "      <td>Mr. President. I yield myself as much time as ...</td>\n",
       "      <td>1982-09-30</td>\n",
       "      <td>Mr. BIDEN</td>\n",
       "      <td>1615</td>\n",
       "      <td>1982</td>\n",
       "      <td>9</td>\n",
       "      <td>1982-09-01</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2667469</th>\n",
       "      <td>Mr. President. I rise today to speak on this s...</td>\n",
       "      <td>1982-05-19</td>\n",
       "      <td>Mr. CHILES</td>\n",
       "      <td>1542</td>\n",
       "      <td>1982</td>\n",
       "      <td>5</td>\n",
       "      <td>1982-05-01</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2744535</th>\n",
       "      <td>Mr. President. I personally commend the Senato...</td>\n",
       "      <td>1982-09-30</td>\n",
       "      <td>Mr. NUNN</td>\n",
       "      <td>1545</td>\n",
       "      <td>1982</td>\n",
       "      <td>9</td>\n",
       "      <td>1982-09-01</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    speech        date  \\\n",
       "2744529  Mr. President. I thank the distinguished Senat...  1982-09-30   \n",
       "2548179  Mr. President. the bill that I have just sent ...  1981-07-09   \n",
       "2528729  Mr. President. first let me thank the Senator ...  1981-05-20   \n",
       "2548181  Mr. President. I thank my colleague from Delaw...  1981-07-09   \n",
       "2675104  Mr. President. for the last week Senator NUNN ...  1982-05-26   \n",
       "2744279  Mr. President. this bill is unique in the hist...  1982-09-30   \n",
       "2682023  Madam President. over the past weeks. together...  1982-06-09   \n",
       "2744526  Mr. President. I yield myself as much time as ...  1982-09-30   \n",
       "2667469  Mr. President. I rise today to speak on this s...  1982-05-19   \n",
       "2744535  Mr. President. I personally commend the Senato...  1982-09-30   \n",
       "\n",
       "             speaker  word_count  year  month  month_year  vocabcount  \n",
       "2744529   Mr. CHILES        4392  1982      9  1982-09-01          68  \n",
       "2548179    Mr. BIDEN        3608  1981      7  1981-07-01          52  \n",
       "2528729     Mr. NUNN        3667  1981      5  1981-05-01          51  \n",
       "2548181     Mr. NUNN        2161  1981      7  1981-07-01          37  \n",
       "2675104   Mr. CHILES        1096  1982      5  1982-05-01          35  \n",
       "2744279  Mr. SPECTER        4344  1982      9  1982-09-01          35  \n",
       "2682023   Mr. CHILES        1763  1982      6  1982-06-01          34  \n",
       "2744526    Mr. BIDEN        1615  1982      9  1982-09-01          34  \n",
       "2667469   Mr. CHILES        1542  1982      5  1982-05-01          33  \n",
       "2744535     Mr. NUNN        1545  1982      9  1982-09-01          33  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "longest_speeches = top_speaker_speeches.nlargest(n=10, columns=['vocabcount']) # Get the top 5 longest speeches by word_count\n",
    "longest_speeches "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice that this dataframe is similar tot he one we generated above -- with the speeches that mentioned the word 'crime' the most -- but this dataframe is confined to the list of top speakers about crime that we generated in a previous step.  This kind of iterative practice can be useful in refining your data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we were really interested in what was said about crime in Congress during the 1980s, this would be a great place to start reading -- the five longest speeches by the five individuals who talked the most about democracy."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Group by two factors at the same time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can group by multiple columns.  When you do, you can apply **.value_counts()** to understand their content.\n",
    "\n",
    "Notice that the first column in the grouping takes parentheses **()** while the second column takes square brackets **[].**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker                  month_year\n",
       "The PRESIDING OFFICER    1980-06-01    3609\n",
       "                         1982-12-01    3600\n",
       "                         1987-10-01    3376\n",
       "                         1986-06-01    3076\n",
       "                         1986-08-01    3069\n",
       "                         1986-10-01    2938\n",
       "                         1985-11-01    2864\n",
       "                         1988-05-01    2687\n",
       "                         1985-10-01    2637\n",
       "                         1986-09-01    2624\n",
       "                         1982-05-01    2621\n",
       "                         1984-06-01    2549\n",
       "                         1987-07-01    2510\n",
       "                         1988-07-01    2508\n",
       "                         1983-10-01    2462\n",
       "                         1984-10-01    2420\n",
       "                         1988-10-01    2414\n",
       "                         1981-07-01    2377\n",
       "                         1987-09-01    2372\n",
       "                         1981-11-01    2319\n",
       "                         1980-09-01    2317\n",
       "                         1983-11-01    2313\n",
       "The SPEAKER pro tempore  1988-09-01    2270\n",
       "The PRESIDING OFFICER    1985-05-01    2212\n",
       "                         1982-07-01    2178\n",
       "                         1982-08-01    2164\n",
       "                         1980-12-01    2106\n",
       "                         1988-03-01    2079\n",
       "                         1988-09-01    2064\n",
       "                         1980-07-01    2035\n",
       "                         1983-06-01    2004\n",
       "                         1985-06-01    1993\n",
       "                         1983-03-01    1984\n",
       "                         1987-12-01    1977\n",
       "                         1981-12-01    1976\n",
       "                         1982-09-01    1938\n",
       "                         1981-09-01    1922\n",
       "                         1987-05-01    1911\n",
       "                         1985-12-01    1849\n",
       "                         1988-06-01    1826\n",
       "                         1980-05-01    1810\n",
       "Mr. BYRD                 1988-10-01    1755\n",
       "The PRESIDING OFFICER    1983-09-01    1753\n",
       "                         1981-06-01    1753\n",
       "                         1981-05-01    1708\n",
       "                         1983-05-01    1695\n",
       "                         1983-07-01    1682\n",
       "                         1986-03-01    1664\n",
       "                         1984-09-01    1648\n",
       "                         1987-06-01    1647\n",
       "Name: month_year, dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "speaker_by_month = congress.groupby('speaker')['month_year'].value_counts().sort_values(ascending=False)[:50]\n",
    "speaker_by_month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker                  year\n",
       "The PRESIDING OFFICER    1987    19439\n",
       "                         1986    18423\n",
       "                         1980    18275\n",
       "                         1988    17950\n",
       "                         1982    17829\n",
       "                         1985    16594\n",
       "                         1981    16409\n",
       "                         1983    15954\n",
       "                         1984    14170\n",
       "Mr. BYRD                 1987    11486\n",
       "                         1988     9768\n",
       "Mr. ROBERT C. BYRD       1980     9164\n",
       "Mr. BAKER                1982     8172\n",
       "                         1983     8015\n",
       "                         1981     7481\n",
       "The SPEAKER pro tempore  1988     6747\n",
       "                         1980     6568\n",
       "Mr. BAKER                1984     6489\n",
       "Mr. DOLE                 1985     6281\n",
       "                         1986     5930\n",
       "Name: year, dtype: int64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "speaker_by_year = congress.groupby('speaker')['year'].value_counts().sort_values(ascending=False)[:20]\n",
    "speaker_by_year"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Two is not a limit. We can keep adding more columns to group by.  \n",
    "\n",
    "   * Notice the punctuation and formatting of the extra columns added below."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How much did each speaker speak about crime per year?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker       year\n",
       "Mr. CHILES    1982    631\n",
       "Mr. LUNGREN   1984    246\n",
       "Mr. HUGHES    1984    227\n",
       "Mr. NUNN      1982    220\n",
       "Mr. BIDEN     1982    190\n",
       "Mr. THURMOND  1984    165\n",
       "Mr. PROXMIRE  1983    158\n",
       "              1984    146\n",
       "Mr. GONZALEZ  1980    144\n",
       "Mr. NUNN      1981    141\n",
       "Mr. PROXMIRE  1980    141\n",
       "Mr. BIDEN     1984    141\n",
       "Mr. WALKER    1984    140\n",
       "Mr. KENNEDY   1981    139\n",
       "Mr. PROXMIRE  1982    138\n",
       "Mr. THURMOND  1982    136\n",
       "              1981    135\n",
       "Mr. PROXMIRE  1981    134\n",
       "Mr. HEFLIN    1982    129\n",
       "Mr. CHILES    1981    125\n",
       "Name: vocabcount, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "speaker_by_words = congress.groupby(['speaker', 'year'])['vocabcount'].sum().sort_index(ascending=False).sort_values(ascending = False)[:20]\n",
    "speaker_by_words"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's the same question, broken down by month."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "speaker         month_year\n",
       "Mr. CHILES      1982-06-01    209\n",
       "                1982-05-01    155\n",
       "                1982-09-01    147\n",
       "Mr. LUNGREN     1984-05-01    114\n",
       "Mr. BIDEN       1982-09-01     94\n",
       "Mr. THURMOND    1984-02-01     91\n",
       "Mr. LUNGREN     1984-09-01     81\n",
       "Mr. LEVIN       1984-02-01     79\n",
       "Mr. HUGHES      1984-09-01     73\n",
       "                1984-10-01     70\n",
       "Mr. KENNEDY     1981-07-01     65\n",
       "                1984-02-01     65\n",
       "Mr. MITCHELL    1984-02-01     64\n",
       "Mr. HEFLIN      1982-05-01     61\n",
       "Mr. GINGRICH    1984-09-01     60\n",
       "Mr. NUNN        1981-05-01     59\n",
       "                1982-09-01     59\n",
       "Mr. HUDDLESTON  1984-01-01     58\n",
       "Mr. THURMOND    1981-06-01     57\n",
       "Mr. CHILES      1982-07-01     56\n",
       "Name: vocabcount, dtype: int64"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "speaker_by_words = congress.groupby(['speaker', 'month_year'])['vocabcount'].sum().sort_index(ascending=False).sort_values(ascending = False)[:20]\n",
    "speaker_by_words"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What was happening in November 1993, that so many people were talking so much about crime?  \n",
    "\n",
    "What was happening in May 1968 and May-June 1982, and how was it different from the large number of speeches about crime in the 1990s?\n",
    "\n",
    "A good data analysis typically leads to applicable historical questions that can be answered by looking things up and by guided reading of the details of selected documents."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Assignment (working in groups)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1) Using groupby(), count the number of times per month that Mr. HATCH spoke about crime.  \n",
    "\n",
    "2) Find the month in which Mr. HATCH spoke about crime the most.  \n",
    "\n",
    "   * Find the speech where he mentions crime the most in that month.\n",
    "   \n",
    "3) Navigate to the speeches of Biden and Hatch in November 1993 and the speeches of May 1968 and May-June 1982.   What were they talking about?  How was talking about crime different in these three windows of time?  Use any relevant strategies you have learned.\n",
    "\n",
    "For each part of the assignment, take a screenshot of your code and the results.  Upload it to Canvas."
   ]
  },
  {
   "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": 4
}