automation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "import gc\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy.stats as stats\n",
    "import matplotlib.pyplot as plt \n",
    "import seaborn as sns\n",
    "\n",
    "from py.tap_trial import tap_trial\n",
    "from py.central_node_selector import central_node_selector"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is written to be run on a **Linux** machine. First pass evaluations are done on the Sioux Falls network for speed and per standard.\n",
    "\n",
    "# Examine runtime of different networks\n",
    "\n",
    "Candidate networks need to have node coordinates and run relatively quickly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "friedrichshain-center\n",
      "728609.3041808965\n",
      "\n",
      "berlin-mitte-center\n",
      "1051175.4763035253\n",
      "\n",
      "berlin-mitte-prenzlauerberg-friedrichshain-center\n",
      "2362496.729620472\n",
      "\n",
      "berlin-prenzlauerberg-center\n",
      "1399925.4585168122\n",
      "\n",
      "berlin-tiergarten\n",
      "716823.6964993116\n",
      "\n",
      "SiouxFalls\n",
      "7480171.248174683\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "friedrichshain-center                                0.645014\n",
       "berlin-mitte-center                                  0.923601\n",
       "berlin-mitte-prenzlauerberg-friedrichshain-center    2.940272\n",
       "berlin-prenzlauerberg-center                         0.905502\n",
       "berlin-tiergarten                                    0.774335\n",
       "SiouxFalls                                           0.109808\n",
       "dtype: float64"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "networks = [\n",
    "#     'berlin-center',\n",
    "    'friedrichshain-center',\n",
    "    'berlin-mitte-center',\n",
    "    'berlin-mitte-prenzlauerberg-friedrichshain-center',\n",
    "    'berlin-prenzlauerberg-center',\n",
    "    'berlin-tiergarten',\n",
    "#     'Birmingham',\n",
    "#     'ChicagoSketch',\n",
    "#     'Philadelphia',\n",
    "    'SiouxFalls',\n",
    "#     'ChicagoRegional'\n",
    "]\n",
    "\n",
    "tap_times = {}\n",
    "for net in networks:\n",
    "    print(net)\n",
    "    start = time.time()\n",
    "    try:\n",
    "        tstt, _, _ = tap_trial('inputs/{}_net.tntp'.format(net), 'inputs/{}_trips.tntp'.format(net), norm_mean=1, norm_sd=0)\n",
    "        if tstt == 0:\n",
    "            print('TAP ISSUE with', net)\n",
    "            tap_times[net] = np.nan\n",
    "        else:\n",
    "            tap_times[net] = time.time()-start\n",
    "            print(tstt)\n",
    "        print('')\n",
    "        del tstt\n",
    "    except OSError:\n",
    "        print('MEMORY ISSUE with', net)\n",
    "        tap_times[net] = np.nan\n",
    "    gc.collect()\n",
    "pd.Series(tap_times)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Analyze a specific network\n",
    "\n",
    "## Scale iterations to match and find unperturbed (\"true\") network metrics."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True TSTT: 7480171.248174683\n",
      "      WVC: 1.4740301437798424\n",
      "      VMT: 3419126.307475\n",
      "   Demand: 360600.0\n"
     ]
    }
   ],
   "source": [
    "net = 'SiouxFalls'\n",
    "#     'friedrichshain-center'\n",
    "#     'berlin-mitte-center'\n",
    "#     'berlin-mitte-prenzlauerberg-friedrichshain-center'\n",
    "#     'berlin-prenzlauerberg-center'\n",
    "#     'berlin-tiergarten'\n",
    "#     'SiouxFalls'\n",
    "\n",
    "netFilename = 'inputs/{}_net.tntp'.format(net)\n",
    "tripsFilename = 'inputs/{}_trips.tntp'.format(net)\n",
    "    \n",
    "n = 500\n",
    "if tap_times[net] > 1:\n",
    "    n = 100\n",
    "elif tap_times[net] > .25:\n",
    "    n = 200\n",
    "\n",
    "true_df, true_tstt, true_wvc, true_vmt, true_demand, _ = tap_trial(netFilename,\n",
    "                                                                   tripsFilename,\n",
    "                                                                   norm_mean=1,\n",
    "                                                                   norm_sd=0,\n",
    "                                                                   returnDataFrame=True,\n",
    "                                                                   returnODinfo=True)\n",
    "print('True TSTT:', true_tstt)\n",
    "print('      WVC:', true_wvc)\n",
    "print('      VMT:', true_vmt)\n",
    "print('   Demand:', true_demand)\n",
    "\n",
    "true_costs = true_df['cost']\n",
    "true_flows = true_df['flow']\n",
    "# vc = true_df['flow']/true_df['capacity']\n",
    "# true_most_congested_link_id = vc.idxmax()\n",
    "# true_most_congested_link_vc = vc.max()\n",
    "# del vc\n",
    "\n",
    "# print('')\n",
    "# print('Most congested link:', true_most_congested_link_id)\n",
    "# print('           Link V/C:', true_most_congested_link_vc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run trials for several perturbation mean and standard deviation combinations\n",
    "\n",
    "(Except N(1,0) perturbations, since they only need one trial each.)\n",
    "\n",
    "Avoid redoing the sampling by reading from a pre-exported csv. If it doesn't exist, then run the trials."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.004, 0.02 , 0.04 , 0.08 , 0.2  , 0.4  , 0.8  , 2.   ])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array([0.001, 0.005, 0.01, 0.02, 0.05, 0.1, 0.2, 0.5])/.25"
   ]
  },
  {
   "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>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>tstt</th>\n",
       "      <th>wvc</th>\n",
       "      <th>vmt</th>\n",
       "      <th>costs_rmse</th>\n",
       "      <th>cost_err25</th>\n",
       "      <th>cost_err75</th>\n",
       "      <th>flows_rmse</th>\n",
       "      <th>flow_err25</th>\n",
       "      <th>flow_err75</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>7.188583e+06</td>\n",
       "      <td>1.454891</td>\n",
       "      <td>3.386991e+06</td>\n",
       "      <td>1.793009</td>\n",
       "      <td>-0.747304</td>\n",
       "      <td>0.333359</td>\n",
       "      <td>1245.525701</td>\n",
       "      <td>-754.643099</td>\n",
       "      <td>611.069795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>7.008864e+06</td>\n",
       "      <td>1.438517</td>\n",
       "      <td>3.347382e+06</td>\n",
       "      <td>2.277803</td>\n",
       "      <td>-1.484768</td>\n",
       "      <td>0.221406</td>\n",
       "      <td>1553.510582</td>\n",
       "      <td>-1023.401674</td>\n",
       "      <td>517.422141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>7.354342e+06</td>\n",
       "      <td>1.455456</td>\n",
       "      <td>3.381127e+06</td>\n",
       "      <td>1.481547</td>\n",
       "      <td>-0.870643</td>\n",
       "      <td>0.112251</td>\n",
       "      <td>978.445622</td>\n",
       "      <td>-759.886026</td>\n",
       "      <td>312.904665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>6.804394e+06</td>\n",
       "      <td>1.429417</td>\n",
       "      <td>3.295416e+06</td>\n",
       "      <td>2.017559</td>\n",
       "      <td>-1.446902</td>\n",
       "      <td>0.004870</td>\n",
       "      <td>1288.414099</td>\n",
       "      <td>-1117.310915</td>\n",
       "      <td>143.474060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>8.272373e+06</td>\n",
       "      <td>1.494986</td>\n",
       "      <td>3.519640e+06</td>\n",
       "      <td>2.166542</td>\n",
       "      <td>-0.500441</td>\n",
       "      <td>1.169936</td>\n",
       "      <td>1295.564747</td>\n",
       "      <td>-433.883377</td>\n",
       "      <td>1353.147350</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean   sd          tstt       wvc           vmt  costs_rmse  cost_err25  \\\n",
       "0   1.0  0.5  7.188583e+06  1.454891  3.386991e+06    1.793009   -0.747304   \n",
       "1   1.0  0.5  7.008864e+06  1.438517  3.347382e+06    2.277803   -1.484768   \n",
       "2   1.0  0.5  7.354342e+06  1.455456  3.381127e+06    1.481547   -0.870643   \n",
       "3   1.0  0.5  6.804394e+06  1.429417  3.295416e+06    2.017559   -1.446902   \n",
       "4   1.0  0.5  8.272373e+06  1.494986  3.519640e+06    2.166542   -0.500441   \n",
       "\n",
       "   cost_err75   flows_rmse   flow_err25   flow_err75  \n",
       "0    0.333359  1245.525701  -754.643099   611.069795  \n",
       "1    0.221406  1553.510582 -1023.401674   517.422141  \n",
       "2    0.112251   978.445622  -759.886026   312.904665  \n",
       "3    0.004870  1288.414099 -1117.310915   143.474060  \n",
       "4    1.169936  1295.564747  -433.883377  1353.147350  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def track_ci_width_evo(netFilename, tripsFilename, true_costs, true_flows, n=1000):#, true_most_congested_link_id, n=1000):\n",
    "#     if n<=30:\n",
    "#         raise ValueError('Sample size needs to be greater than 30.')\n",
    "    \n",
    "    means = np.array([1]) # np.arange(1.75, 0, -.25)\n",
    "    sds = [0.001, 0.005, 0.01, 0.02, 0.05, 0.1, 0.2, 0.5][::-1]\n",
    "\n",
    "    results = {}\n",
    "\n",
    "    for mean in means:\n",
    "        results[mean] = {}\n",
    "        for sd in sds:\n",
    "            results[mean][sd] = {\n",
    "                'tstts': np.zeros(n)*np.nan,\n",
    "                'wvcs': np.zeros(n)*np.nan,\n",
    "                'vmts': np.zeros(n)*np.nan,\n",
    "                'costs_rmses': np.zeros(n)*np.nan,\n",
    "                'cost_err25': np.zeros(n)*np.nan,\n",
    "                'cost_err75': np.zeros(n)*np.nan,\n",
    "                'flows_rmses': np.zeros(n)*np.nan,\n",
    "                'flow_err25': np.zeros(n)*np.nan,\n",
    "                'flow_err75': np.zeros(n)*np.nan,\n",
    "#                 'most_congested_link_vcs': np.zeros(n)*np.nan,\n",
    "#                 'tstt_widths': np.zeros(n)*np.nan,\n",
    "#                 'wvc_widths': np.zeros(n)*np.nan,\n",
    "#                 'vmt_widths': np.zeros(n)*np.nan,\n",
    "#                 'flows_rmse_widths': np.zeros(n)*np.nan,\n",
    "#                 'most_congested_link_vc_widths': np.zeros(n)*np.nan\n",
    "            }\n",
    "            r = results[mean][sd]\n",
    "            for i in range(n):\n",
    "                r['tstts'][i], r['wvcs'][i], r['vmts'][i], r['costs_rmses'][i], r['cost_err25'][i], r['cost_err75'][i], r['flows_rmses'][i], r['flow_err25'][i], r['flow_err75'][i] = tap_trial(\n",
    "                    netFilename, tripsFilename, norm_mean=mean, norm_sd=sd,\n",
    "                    true_costs=true_costs, true_flows=true_flows)#, focus_link=true_most_congested_link_id)\n",
    "#                 ni = np.sqrt(i+1)\n",
    "#                 results[mean][sd]['tstt_widths'][i] = 4*results[mean][sd]['tstts'][0:i+1].std()/ni\n",
    "#                 results[mean][sd]['wvc_widths'][i] = 4*results[mean][sd]['wvcs'][0:i+1].std()/ni\n",
    "#                 results[mean][sd]['vmt_widths'][i] = 4*results[mean][sd]['vmts'][0:i+1].std()/ni\n",
    "#                 results[mean][sd]['flows_rmse_widths'][i] = 4*results[mean][sd]['flows_rmses'][0:i+1].std()/ni\n",
    "#                 results[mean][sd]['most_congested_link_vc_widths'][i] = 4*results[mean][sd]['most_congested_link_vcs'][0:i+1].std()/ni\n",
    "                if i>0 and results[mean][sd]['tstts'][i] > results[mean][sd]['tstts'][i-1]*1000:\n",
    "                    raise ValueError('There may be a float/initialization error here.')\n",
    "    \n",
    "    # format the results into a DataFrame for easy querying and csv export\n",
    "    df_trials = pd.DataFrame(columns=['mean', 'sd', 'tstt', 'wvc', 'vmt', 'costs_rmse', 'cost_err25', 'cost_err75', 'flows_rmse', 'flow_err25', 'flow_err75'])\n",
    "#     df_ci_evo = pd.DataFrame(columns=['mean', 'sd', 'n', 'tstt_ci_width', 'congestion_ci_width', 'vmt_ci_width'])\n",
    "    for mean in results:\n",
    "        for sd in results[mean]:\n",
    "            df_trials = pd.concat([df_trials, pd.DataFrame({\n",
    "                'mean': np.ones(n)*mean,\n",
    "                'sd': np.ones(n)*sd,\n",
    "                'tstt': results[mean][sd]['tstts'],\n",
    "                'wvc': results[mean][sd]['wvcs'],\n",
    "                'vmt': results[mean][sd]['vmts'],\n",
    "                'costs_rmse': results[mean][sd]['costs_rmses'],\n",
    "                'cost_err25': results[mean][sd]['cost_err25'],\n",
    "                'cost_err75': results[mean][sd]['cost_err75'],\n",
    "                'flows_rmse': results[mean][sd]['flows_rmses'],\n",
    "                'flow_err25': results[mean][sd]['flow_err25'],\n",
    "                'flow_err75': results[mean][sd]['flow_err75'],\n",
    "#                 'most_congested_link_vc': results[mean][sd]['most_congested_link_vcs']\n",
    "            })], ignore_index=True, sort=False)\n",
    "#             df_ci_evo = pd.concat([df_ci_evo, pd.DataFrame({\n",
    "#                 'mean': np.ones(n)*mean,\n",
    "#                 'sd': np.ones(n)*sd,\n",
    "#                 'n': np.arange(1,n+1),\n",
    "#                 'tstt_ci_width': results[mean][sd]['tstt_widths'],\n",
    "#                 'congestion_ci_width': results[mean][sd]['wvc_widths'],\n",
    "#                 'vmt_ci_width': results[mean][sd]['vmt_widths'],\n",
    "#                 'flows_rmse_ci_width': results[mean][sd]['flows_rmse_widths'],\n",
    "#                 'most_congested_link_vc_ci_width': results[mean][sd]['most_congested_link_vc_widths']\n",
    "#             })], ignore_index=True, sort=False)\n",
    "    df_trials.to_csv('output/data/{}_trials.csv'.format(net), index=False)\n",
    "#     df_ci_evo.to_csv('output/data/{}_ci_evo.csv'.format(net), index=False)\n",
    "    return df_trials#, df_ci_evo\n",
    "\n",
    "try:\n",
    "    # avoid redoing trials by reading from exports if they exist\n",
    "    df_trials = pd.read_csv('output/data/{}_trials.csv'.format(net))\n",
    "#     df_ci_evo = pd.read_csv('output/data/{}_ci_evo.csv'.format(net))\n",
    "#     n = df_ci_evo['n'].max()+1\n",
    "    n = df_trials[(df_trials['mean']==1) & (df_trials['sd']==0.1)].shape[0]\n",
    "except FileNotFoundError:\n",
    "    df_trials = track_ci_width_evo(netFilename, tripsFilename, true_costs, true_flows, n=n)\n",
    "df_trials.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compare results for different means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>mean</th>\n",
       "      <th>tstt</th>\n",
       "      <th>wvc</th>\n",
       "      <th>vmt</th>\n",
       "      <th>costs_rmse</th>\n",
       "      <th>cost_err25</th>\n",
       "      <th>cost_err75</th>\n",
       "      <th>flows_rmse</th>\n",
       "      <th>flow_err25</th>\n",
       "      <th>flow_err75</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>6.244605</td>\n",
       "      <td>-7.479121</td>\n",
       "      <td>-1.200594</td>\n",
       "      <td>12466.195912</td>\n",
       "      <td>-15292.871544</td>\n",
       "      <td>-8382.829875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.25</td>\n",
       "      <td>8.277958e+05</td>\n",
       "      <td>0.514646</td>\n",
       "      <td>7.961663e+05</td>\n",
       "      <td>6.101839</td>\n",
       "      <td>-7.360098</td>\n",
       "      <td>-1.180327</td>\n",
       "      <td>9521.677677</td>\n",
       "      <td>-11911.621544</td>\n",
       "      <td>-5650.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.50</td>\n",
       "      <td>1.870592e+06</td>\n",
       "      <td>0.829003</td>\n",
       "      <td>1.623629e+06</td>\n",
       "      <td>5.544822</td>\n",
       "      <td>-6.893024</td>\n",
       "      <td>-1.077085</td>\n",
       "      <td>6584.562265</td>\n",
       "      <td>-8262.457785</td>\n",
       "      <td>-3475.187360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.75</td>\n",
       "      <td>3.654464e+06</td>\n",
       "      <td>1.145810</td>\n",
       "      <td>2.493965e+06</td>\n",
       "      <td>3.897694</td>\n",
       "      <td>-4.863088</td>\n",
       "      <td>-0.778839</td>\n",
       "      <td>3399.198727</td>\n",
       "      <td>-3942.319025</td>\n",
       "      <td>-1768.643446</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.00</td>\n",
       "      <td>7.480171e+06</td>\n",
       "      <td>1.474030</td>\n",
       "      <td>3.419126e+06</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.25</td>\n",
       "      <td>1.567504e+07</td>\n",
       "      <td>1.790134</td>\n",
       "      <td>4.408337e+06</td>\n",
       "      <td>7.861444</td>\n",
       "      <td>1.164602</td>\n",
       "      <td>9.631130</td>\n",
       "      <td>3838.708583</td>\n",
       "      <td>1750.000000</td>\n",
       "      <td>4985.245509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.50</td>\n",
       "      <td>3.286836e+07</td>\n",
       "      <td>2.129701</td>\n",
       "      <td>5.357056e+06</td>\n",
       "      <td>22.435459</td>\n",
       "      <td>4.676741</td>\n",
       "      <td>27.591739</td>\n",
       "      <td>7183.991432</td>\n",
       "      <td>3904.694325</td>\n",
       "      <td>8745.275521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.75</td>\n",
       "      <td>6.554625e+07</td>\n",
       "      <td>2.479270</td>\n",
       "      <td>6.278202e+06</td>\n",
       "      <td>46.646506</td>\n",
       "      <td>9.232242</td>\n",
       "      <td>57.163937</td>\n",
       "      <td>10463.053891</td>\n",
       "      <td>5735.050706</td>\n",
       "      <td>12980.236887</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean          tstt       wvc           vmt  costs_rmse  cost_err25  \\\n",
       "0  0.00  0.000000e+00  0.000000  0.000000e+00    6.244605   -7.479121   \n",
       "1  0.25  8.277958e+05  0.514646  7.961663e+05    6.101839   -7.360098   \n",
       "2  0.50  1.870592e+06  0.829003  1.623629e+06    5.544822   -6.893024   \n",
       "3  0.75  3.654464e+06  1.145810  2.493965e+06    3.897694   -4.863088   \n",
       "4  1.00  7.480171e+06  1.474030  3.419126e+06    0.000000    0.000000   \n",
       "5  1.25  1.567504e+07  1.790134  4.408337e+06    7.861444    1.164602   \n",
       "6  1.50  3.286836e+07  2.129701  5.357056e+06   22.435459    4.676741   \n",
       "7  1.75  6.554625e+07  2.479270  6.278202e+06   46.646506    9.232242   \n",
       "\n",
       "   cost_err75    flows_rmse    flow_err25    flow_err75  \n",
       "0   -1.200594  12466.195912 -15292.871544  -8382.829875  \n",
       "1   -1.180327   9521.677677 -11911.621544  -5650.000000  \n",
       "2   -1.077085   6584.562265  -8262.457785  -3475.187360  \n",
       "3   -0.778839   3399.198727  -3942.319025  -1768.643446  \n",
       "4    0.000000      0.000000      0.000000      0.000000  \n",
       "5    9.631130   3838.708583   1750.000000   4985.245509  \n",
       "6   27.591739   7183.991432   3904.694325   8745.275521  \n",
       "7   57.163937  10463.053891   5735.050706  12980.236887  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "means = np.arange(0,2,.25)\n",
    "reses = [tap_trial(netFilename, tripsFilename, norm_mean=mean, norm_sd=0,  \n",
    "                   true_costs=true_costs, true_flows=true_flows, returnDataFrame=False) for mean in means]\n",
    "\n",
    "# TODO: vmt, pc, most congested link analysis?\n",
    "dfmeans = pd.DataFrame({\n",
    "    'mean': means,\n",
    "    'tstt': [r[0] for r in reses],\n",
    "    'wvc': [r[1] for r in reses],\n",
    "    'vmt': [r[2] for r in reses],\n",
    "    'costs_rmse': [r[3] for r in reses],\n",
    "    'cost_err25': [r[4] for r in reses],\n",
    "    'cost_err75': [r[5] for r in reses],\n",
    "    'flows_rmse': [r[6] for r in reses],\n",
    "    'flow_err25': [r[7] for r in reses],\n",
    "    'flow_err75': [r[8] for r in reses],\n",
    "})\n",
    "dfmeans.to_csv('output/data/{}_trials_means_only.csv'.format(net), index=False)\n",
    "dfmeans.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams.update({'font.size': 12})\n",
    "\n",
    "def tstt_wvc_plot(df, input_key, input_name, net, log_input=False, log_tstt=False, tstt_top=None, wvc_top=None, output_fig_name=None):\n",
    "    fig, ax1 = plt.subplots()\n",
    "    ax1.set_xlabel(input_name)\n",
    "    ax1.set_ylabel('TSTT or VMT scaled to true')\n",
    "    ax1.plot(df[input_key], df['tstt']/(df.loc[df['mean']==1,'tstt'].values if input_key=='mean' else 1), c='k', label='TSTT')\n",
    "    ax1.plot(df[input_key], df['vmt']/(df.loc[df['mean']==1,'vmt'].values if input_key=='mean' else 1), c='k', linestyle='-.', label='VMT')\n",
    "    if log_input:\n",
    "        ax1.set_xscale('log')\n",
    "    if log_tstt:\n",
    "        ax1.set_yscale('log')\n",
    "    ax1.set_ylim(top=tstt_top)\n",
    "\n",
    "    ax2 = ax1.twinx()\n",
    "    ax2.set_ylabel('WVC')\n",
    "    ax2.plot(df[input_key], df['wvc'], linestyle='--', c='k', label='WVC')\n",
    "    ax2.set_ylim(bottom=0, top=wvc_top)\n",
    "\n",
    "    fig.legend(loc=1, bbox_to_anchor=(0.25,1), bbox_transform=ax1.transAxes)\n",
    "#     ax1.set_title('TSTT and congestion for varied {}\\n[{}]'.format(input_name,net), y=1.05)\n",
    "    ax1.set_title('{}'.format(net), y=1.05)\n",
    "    fig.tight_layout()\n",
    "    if output_fig_name:\n",
    "        plt.savefig(output_fig_name, dpi='figure')\n",
    "    plt.show()\n",
    "\n",
    "tstt_wvc_plot(dfmeans, 'mean', 'Demand scaling factor', net, log_tstt=False,\n",
    "              tstt_top=9, wvc_top=2.5,\n",
    "              output_fig_name='output/plots/{}_mean_scaling.png'.format(net))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compare results for different standard deviations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sds = df_trials['sd'].unique()\n",
    "tstts_scaled = np.array([df_trials[(df_trials['mean']==1) & (df_trials['sd']==sd)]['tstt'].mean() for sd in sds])/true_tstt\n",
    "congestions = np.array([df_trials[(df_trials['mean']==1) & (df_trials['sd']==sd)]['wvc'].mean() for sd in sds])\n",
    "vmts_scaled = np.array([df_trials[(df_trials['mean']==1) & (df_trials['sd']==sd)]['vmt'].mean() for sd in sds])/true_vmt\n",
    "# tstt_wvc_plot(sds[:-2], tstts_scaled[:-2], congestions[:-2], 'OD matrix perturbation SD', net, log_input=True, output_fig_name='output/plots/{}_stdv_scaling_clipped.png'.format(net))\n",
    "tstt_wvc_plot(pd.DataFrame({\n",
    "    'sd': sds,\n",
    "    'tstt': tstts_scaled,\n",
    "    'wvc': congestions,\n",
    "    'vmt': vmts_scaled\n",
    "}), 'sd', 'OD matrix perturbation SD', net, log_input=True, output_fig_name='output/plots/{}_stdv_scaling.png'.format(net))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize Monte Carlo trials"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  true mean: 7480171.248174683\n",
      "sample mean: 7480052.478109291\n",
      "sample stdv: 1328.8163655396295\n",
      "  shapiro p: 0.20807474851608276\n",
      "% in 95% CI: 0.944\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  true mean: 1.4740301437798424\n",
      "sample mean: 1.4740235411137674\n",
      "sample stdv: 7.156937657910632e-05\n",
      "  shapiro p: 0.47453516721725464\n",
      "% in 95% CI: 0.952\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>tstt_mean</th>\n",
       "      <th>tstt_sd</th>\n",
       "      <th>tstt_p</th>\n",
       "      <th>wvc_mean</th>\n",
       "      <th>wvc_sd</th>\n",
       "      <th>wvc_p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.001</td>\n",
       "      <td>7.480052e+06</td>\n",
       "      <td>1328.816366</td>\n",
       "      <td>0.208075</td>\n",
       "      <td>1.474024</td>\n",
       "      <td>0.000072</td>\n",
       "      <td>0.474535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.005</td>\n",
       "      <td>7.479971e+06</td>\n",
       "      <td>6667.739101</td>\n",
       "      <td>0.282825</td>\n",
       "      <td>1.474010</td>\n",
       "      <td>0.000360</td>\n",
       "      <td>0.040673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.010</td>\n",
       "      <td>7.480355e+06</td>\n",
       "      <td>13235.637989</td>\n",
       "      <td>0.937507</td>\n",
       "      <td>1.474018</td>\n",
       "      <td>0.000730</td>\n",
       "      <td>0.362552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.020</td>\n",
       "      <td>7.478847e+06</td>\n",
       "      <td>26331.054793</td>\n",
       "      <td>0.562073</td>\n",
       "      <td>1.473865</td>\n",
       "      <td>0.001479</td>\n",
       "      <td>0.459469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.050</td>\n",
       "      <td>7.480273e+06</td>\n",
       "      <td>65948.749634</td>\n",
       "      <td>0.273677</td>\n",
       "      <td>1.473665</td>\n",
       "      <td>0.003783</td>\n",
       "      <td>0.741436</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean     sd     tstt_mean       tstt_sd    tstt_p  wvc_mean    wvc_sd  \\\n",
       "0   1.0  0.001  7.480052e+06   1328.816366  0.208075  1.474024  0.000072   \n",
       "1   1.0  0.005  7.479971e+06   6667.739101  0.282825  1.474010  0.000360   \n",
       "2   1.0  0.010  7.480355e+06  13235.637989  0.937507  1.474018  0.000730   \n",
       "3   1.0  0.020  7.478847e+06  26331.054793  0.562073  1.473865  0.001479   \n",
       "4   1.0  0.050  7.480273e+06  65948.749634  0.273677  1.473665  0.003783   \n",
       "\n",
       "      wvc_p  \n",
       "0  0.474535  \n",
       "1  0.040673  \n",
       "2  0.362552  \n",
       "3  0.459469  \n",
       "4  0.741436  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def visualize(data, mean_true, plot_true_mean=True, xlabel='TSTT scaled to true', img_title=None, img_filename=None, verbose_output=False):\n",
    "    mean_sample = data.mean()\n",
    "    sd = data.std()\n",
    "    plt.figure()\n",
    "    \n",
    "    mean_sample_normed = mean_sample/mean_true\n",
    "    sd_normed = sd/mean_true\n",
    "    xsnorm = np.linspace(mean_sample_normed - 4*sd_normed, mean_sample_normed + 4*sd_normed, 100)\n",
    "\n",
    "    (data/mean_true).hist(bins=min([len(data)//10,100]), label='Sample', color='lightblue', density=True)\n",
    "    if plot_true_mean:\n",
    "        plt.axvline(x=1, color='orange')\n",
    "        xtnorm = np.linspace(1 - 4*sd_normed, 1 + 4*sd_normed, 100)\n",
    "        plt.plot(xtnorm, stats.norm.pdf(xtnorm, 1, sd_normed), color='orange', label='True mean')\n",
    "    plt.axvline(x=mean_sample_normed, color='steelblue')\n",
    "    plt.plot(xsnorm, stats.norm.pdf(xsnorm, mean_sample_normed, sd_normed), color='steelblue', label='Sample mean')\n",
    "    \n",
    "    if img_title:\n",
    "        plt.title(img_title)\n",
    "    plt.ylabel('Density')\n",
    "    plt.xlabel(xlabel)\n",
    "    plt.legend()\n",
    "    plt.tight_layout()\n",
    "    if img_filename:\n",
    "        plt.savefig(img_filename, dpi='figure')\n",
    "    \n",
    "    if verbose_output:\n",
    "        plt.show()\n",
    "        print('  true mean:', mean_true)\n",
    "        print('sample mean:', mean_sample)\n",
    "        print('sample stdv:', sd)\n",
    "    else:\n",
    "        plt.close()\n",
    "    \n",
    "    return mean_sample, sd\n",
    "\n",
    "def test_perturbation_pair(\n",
    "    df,\n",
    "    \n",
    "    norm_mean,\n",
    "    norm_sd,\n",
    "    \n",
    "    tstt_true,\n",
    "    wvc_true,\n",
    "    \n",
    "    net_name=None,\n",
    "    plot_true_mean=True,\n",
    "    verbose_output=False\n",
    "):\n",
    "    \n",
    "    dft = df.loc[np.isclose(df_trials['mean'],norm_mean) & np.isclose(df_trials['sd'],norm_sd)]\n",
    "    n = len(dft)\n",
    "    \n",
    "    tstts = dft['tstt']\n",
    "    tstt_mean, tstt_sd = visualize(tstts, tstt_true, plot_true_mean, \n",
    "                                   img_title='TSTT dist for N({},{}) perturbation vs. true\\n[{}, $n={}$]'.format(norm_mean, norm_sd, net_name, n), \n",
    "                                   img_filename='output/plots/mcs/{}_tstt_N({},{})_n{}.png'.format(net_name, norm_mean, norm_sd, n), verbose_output=verbose_output)\n",
    "    tstt_pvalue = stats.shapiro(tstts)[1]\n",
    "    if verbose_output:\n",
    "        print('  shapiro p:', tstt_pvalue)\n",
    "        print('% in 95% CI:', ((tstts > tstt_mean-2*tstt_sd) & (tstts < tstt_mean+2*tstt_sd)).sum()/len(tstts))\n",
    "        print('')\n",
    "    \n",
    "    wvcs = dft['wvc']\n",
    "    wvc_mean, wvc_sd = visualize(wvcs, wvc_true, plot_true_mean, xlabel='Weighted V/C scaled to true',\n",
    "                                 img_title='w v/c dist for N({},{}) perturbation vs. true\\n[{}, $n={}$]'.format(norm_mean, norm_sd, net_name, n), \n",
    "                                 img_filename='output/plots/mcs/{}_wvc_N({},{})_n{}.png'.format(net_name, norm_mean, norm_sd, n), verbose_output=verbose_output)\n",
    "    wvc_pvalue = stats.shapiro(wvcs)[1]\n",
    "    if verbose_output:\n",
    "        print('  shapiro p:', wvc_pvalue)\n",
    "        print('% in 95% CI:', ((wvcs > wvc_mean-2*wvc_sd) & (wvcs < wvc_mean+2*wvc_sd)).sum()/len(wvcs))\n",
    "    \n",
    "    return tstt_mean, tstt_sd, tstt_pvalue, wvc_mean, wvc_sd, wvc_pvalue\n",
    "\n",
    "def plot_all_pairs(df_trials, true_tstt, true_wvc, net):\n",
    "    means = df_trials['mean'].unique()\n",
    "    means.sort()\n",
    "    sds = df_trials['sd'].unique()\n",
    "    sds.sort()\n",
    "    \n",
    "    df_pairs = pd.DataFrame(columns=['mean', 'sd', 'tstt_mean', 'tstt_sd', 'tstt_p', 'wvc_mean', 'wvc_sd', 'wvc_p'])\n",
    "    \n",
    "    k=0\n",
    "    for i,mean in enumerate(means):\n",
    "        for sd in sds:\n",
    "            tstt_mn, tstt_sd, tstt_p, wvc_mn, wvc_sd, wvc_p = test_perturbation_pair(df_trials,\n",
    "                                                                                     mean, sd,\n",
    "                                                                                     true_tstt, true_wvc,\n",
    "                                                                                     net_name=net,\n",
    "                                                                                     plot_true_mean=(mean==1),\n",
    "                                                                                     verbose_output=(k==0))\n",
    "            df_pairs.loc[k] = {\n",
    "                'mean': mean,\n",
    "                'sd': sd,\n",
    "                'tstt_mean': tstt_mn,\n",
    "                'tstt_sd': tstt_sd,\n",
    "                'tstt_p': tstt_p,\n",
    "                'wvc_mean': wvc_mn,\n",
    "                'wvc_sd': wvc_sd,\n",
    "                'wvc_p': wvc_p\n",
    "            }\n",
    "            k+=1\n",
    "    \n",
    "#     df_pairs.to_csv('output/data/{}_monte_carlo.csv'.format(net), index=False)\n",
    "    \n",
    "    return df_pairs\n",
    "\n",
    "df_pairs = plot_all_pairs(df_trials, true_tstt, true_wvc, net)\n",
    "df_pairs.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAASsAAAEvCAYAAAAdNeeiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3Xe8XFW5//HPN7QAAYlSRJCqqBQp0lQEFFDEAope0StFBOwNropcRIp6hSuCelEMRZpi+QGCDRtdRAUVJShFSKSDoUgggCTP749nTbIzmXNmzsycM2effN+v17zOzNptzZy9n1l7zd7rUURgZjbeTRp0BczMOuFgZWa14GBlZrXgYGVmteBgZWa14GC1mJO0o6SQtOag62IjJ2md8v/bbtB1GW1tg5WkM8qHEZKeljRT0smSnjUWFRxN5f3sN+h6DNjVwOrA3aO5EUnLSnpC0vuayg8u+1ar8ickrS3pKUnvH2K9e0qaJ+n5lbLdJF0saZakOZJuKvvsBqPz7iY2SWuW/9GOo7Dujo/BTltWV5I79DrAh4E9gbO6qRyA0lLdLj/eSJokaYlB16MbEfFURNwbEfNGeTtzgN8AOzVNehXwjyHKr46ImcAPgAOHWPWBwGURcQuApCOAHwK3AW8GXgjsDzwFfLb3dzIykpYe621OWBEx7AM4A/hlU9l/A3OBZcvrzwF/BR4H7gBOBp5RmX8/4GnglcAfyR3n9cC6wPnkt/rjwF+AvZu2dRlwGrmj3Q88XLY3CTgCuA94APhc03JLAkcCtwNPANOB91SmzwCi+qhMewnwc2B2Wff5wNqV6UcCtwJvA/5W3tvGwEbAz0odHyufyd7DfLaNz2XnUr8ngN8BW1TmmQqcQx7Qc4CbgEMAVeYZdrvAAaXsCWAWcAWwZpm2Y3n/za93KfM9DtwIvKap7psD15R13gy8pXymhw/zfg8v259UXi8BPFLq16r8sPJ6p1KnLZvWtza5H+5V+b8FcOgQ25/abj8HDgJmAv8CLgRWaZpv3/J5PAncSe6XS7bYX48B7gEeqOxvxwBfL+/tfuCDwDLAV4GHgLuADzZt7yPAn8h98V7gO8DqlenrlPe8XaVsVeCb5LHxRNln9iePmdsan2tl/uXL+91viM8mmh4zKtN2AX5N7pt3le0+q5N9k2GOwZb16DJYHVxWvkJlJ3xF+eB2Ig/gM5sOynnA78lvzPWAVYBNgA8ALwbWBz5ECWpN//xHgGOBDcqHHsBPgONK2b6l7LVN9f4z8GoyKL6tfGDvLtNXKdv6CPBs4NmlfMOyYxxFfitvAnyfPCAnV4LV48DlwLalDiuU7X27rGM94LXA69sEq3nAH4AdyufwI3InX67M82zgk8AW5X28s9TvXZX1DLld8gB+GtiHPLg3IYNDu2B1PbAr8HyyFf0wsFKZZ7lSxx+WOm9Lnk4+zvDB6qVl3VuU19uSQWpJ4MGm8gC2Ka8F3AJ8o2l9R5NfJkuX1yeSB8TS7fbrIfbzR4BzyS+el5FBq7ofv44Mjp8q//O3kUHmmKb99VHyC3tDYJPKgfkweew8jzxm5pH7caPsU6Vsw6ZgtXP537+0fM6XDxWsgGXJgPCHstx65DHQCOifAv7Owl927y51W26Iz2bzso03k/vjKqX8VeV//qGyn2wFXEp+yamDfbPlMdi3YFU2+nfgmmGWeRP5zdP4ptyvvNlXdLC9C4FTmv75f2qaZzrwl6ay64Evlufrln/6C5vmOaK6rvJB7dc0zxnAd5rKlin/lD0qwWoesFbTfI80r6/Ne218LjtVyqaSweiAYZb7MvCLTrZb/hePACsOMX1HWgerN1fmeXYpe015fWCpY7X1/MIyz3DBaslSl4+X14cB55fnP2gqfxhYorLsJ8hv/ynl9RJky+aLlXl+Avy508+/xf/9AWCZStmhwD2V11cC32ta7iNkq6IRMC8jv9gmNc03A/hB5fWk8n5+2FT2EE2tq6b1NALHGuX1OiwcrN5NtqbWHGL51cgzm50rZb8BThpmm2uWbezYVH4Z8IWmsrXKvJt1ckzQ4hgc6tFpn9WOkmZLmgPcQDYl39GYKOnNkq6QdLek2cC3gKXJnbzq99UXkpaT9AVJ0yU9WJbdjWwBVF3f9PpeMmI3l61anm9JfhtfW+o9u6z7MPIbYDhbAW9qWm4WMLlp2fsi4h9Ny34ROFXSZZKOlLRFm201/KbxJCIeIr8ZN4T5/WGHSvqTpH+W+ryXhT+j4bb7C/L/dbuk70g6SNLKHdTpT5U63Uu2KFYrRRsCf42IRyrz/I0MMEOKiKfJb92dS9FOwCXl+SVN5ZdGxNzK4t8kvzTeVl6/FlgDOKUyj8gDpVt/jYgnK6/vYsF7hjyluaJpmcvJfWP9Stl10boPcP5+XKY/QGU/LmX3s2A/bvxa+zNJd0h6FLiqTGo+RhpeAtwYEXe2mhgR95ENggPL+jciW7KntJq/ja2AjzYdKzeWaY1jpdtjYhGdBqvfApsBLyL7qXaJiNsAJG1DniZdQX6Lb0EeTJABq2FuRDzRtN7/JU9rjib7szYjvx2bOyX/3fQ6hihrvJ/G35eVdTYeG5OnLcOZBJzdtNxmZLP/1Mp8jzUvGBHHlPm+V7Z1jaRuOnVVeX4I2XT/Ktk/sFmpx/zPaLjtRsRsMni/ifzGfy9wq6SXtKnDUy3KqvtLt0Hhl8B2klYk/z+NYHVpU/mvqgtFxAPABWSfEuXv5RFxU2W2m4Dn9dCp3fyeg4X/F42yKrUoX2TfKEa0H0taizweZgB7kf/HN5b5hnuP7f43JwN7SFqFDFq/j4g/tVmmlUlk90zzsfJ84KfQ12Oi42A1JyJujYgZTd88ANsB/4yIwyPitxFxM9ls7MT2wLci4rsRcT3ZAujHz8vXlb9rlXpXH3+vzPcUeTpRdS0Z0P7eYtmH2m04Im6LiK9FxFvI0873tVuG/GYDQNJK5CnVX0vR9sDFEXFaRPwxIm6lRetwuO1GxNyIuCIijiC/ee+h0jLuwo3AiyQ9o1LvFwArdbDsr8g+r4OBhyOi8U18A3mQH0y2VH7ZYtlvAFtLei3ZAp/WNP2cyroXIWlqB/UbznSyb7Fqe/I08LYe193KVmQf1Ecj4tclMK/WZpnrgI3aXDd3CfmDzUHA3rRvVTWCeKtjZaMWx8mt5UsSaHtMtDoGW+rHRaE3AatIerek9STtA7S8JmaIZXeXtLWkDcmd7zm9Vqgc0KcDp0jaW9LzJG0qaX9Jn6zMejvwSknPqZwafZ5sQZ5T6rWupFdK+rKk9YbapqQpkk6S9KqyzOZkB/WNQy3TqC5wnKTtJW1CdmY/RnZKQn5GO5Y6bFC+lbbpdLuSdpf0MUkvKd/UewDP7aBew/kW2Wd1lqQXl9b1aeRBO+y3ekTcQP5KdTDZmmqUB9kHcjBwVzmtbF72UrJ1+C3ylPO8punXkq30z5fPZHvldVovlXQCGex68T/AnuW0fANJ/0H2Xx4fEa1aor26hfw8Dyn/2z3Ig30455I/DFwkaeey3E6SGqfPjc96WlnX0mWZ4fyT/H+/WtKzK0H/CPL4PUHSZpLWl7SrpNOU19V1cky0OgZb6jlYRcSPyEsJPk9eerAX8PEOF/8Y+cFeSn7j3gX8v17rVBwEnEBeZnFjWf++LPwNeAjZ0rid7D8gIv5KnoZMIX9yvZH85lmW4ftkniY7x08jW0U/Iw/Kdi2YeWRf2jfIb6rVgddFRONU4hiyX+RCsm9rKvCVEWz3IeANwMXkgX4c+XP76W3qNaSIeJxs2axG9kOeQ/4SN5vs3G3nEvLX00uayi8t5b9aZIkFppHv96wWrXwi4jPA7mTr80Iy2J9B9ncd1kHdhhQRPyF/jd6XbAmeAHyN/OW47yLiz+Qvbe8h98P/Aj7aZpnHydbfDeRlDn8FTiL336pvkqew3662goZY5zzyV/v/IC9N+mMpv5T8RXAT8seHP5OfyaPk6W0nx8Qix+BQGj8v2gAor9w9NSKWHHRdeiVpbbJv5Y0R8cMBV8faKGcy08lr165rN/94UPuDxAZD0jvJlvDt5C9Tx5Gt5J8Psl42PEnLkL+i/g/5A0UtAhX4Rmbr3rPIXyX/RvZ5/APYvtWpmY0rbyfvvliPPL2sDZ8GmlktuGVlZrXgYGVmteBgZWa1ULtgVe4xagwG+PoRLHeGpFZXRddCc/3r/n5scCR9sXIMHT7o+nSqdsGq+DZ58eQvYP4olMdIukU5MuQsSb+X9OHKMh8B3jqIymrh0Varj70GUZ9BKze0tvo8ntc0327KG7iflDRD0lC30XQ033jSbZ3bLdfhZ3sUefy0vNl5vKrrdVZzykgADV8nb4T+CHln+4rkUBprNWaojhAwIFeSVwBXDTtKwQQ3gxyfqWr+FcyStiSvQD+e/Ll9G+BkSY9HxMkjnW886bbOI1huBsN8thHxKPCopLnUSSfjyIynB3n/2KlNZQ8zzBhAsWC8ouq4XEsBXyAvbHyKvJ3hHR1s63DKSInktUZ3AF+uTF+VvFH42KG23aJuu5RtPUiO/3M5sHWb+je/3o4csfHR8rieptE923w+U8n70P6THEXjX+StEe/pdB0j2NaRwK1t5vk2Oaxxtex/gdu7ma/Dem1IDij4WPkf7k0OijcLeGYf339Xde5kuU4+28q8Mxhm/LHx9qjraWCze4BdJT1zBMt8nhwe46Pk0BXnkDcvN48FPqSImEUe3O+X9AZJIoeXmUkGtU5NIe/f2pa8L/EW4GJ1mJRDOf77ReRQPluUx5HkgIGd2rz8PZj8LDYjb6r+P0nLt9jmYaqMYzTEY7h78daUdGd5/FTSy5qmv5y8n7HqYmAdLTyiQKfzDUt5E/k15AWTW5D3s51I/h9PjIgHm+bv5f13W+dOl2v32dZSXU8Dmx1Afus8IGk6udP9GLgoyldIlaTlyMQXH4uI75fiz0vairzxebgbaRcSEVcoR0L4JnAm2TTfPCKaxynaUTk4WcN9EbF+WccFTfU7iEzKsSs5wkA7K5Ito4uiJE4gA95IbE7eeLpXLEi+cBZ58+xUFh2j6WRyjKLhPDhE+W/JYZb/BjyDHDLkSkm7RsQvyjyrkwMqVt1bmXbnCOdr53hyXKePAUg6ndwXXkPrARt7ef/d1rmT5Tr5bGtpQgSriPi1pPWBrclz9e3J4UN+KumNLQLW88ihMVqN+vipLqpwDLlTH0we7Le3mOe35N36DU83nkhalxza5KXkaeQkclymoUaDXEhEPCTpVOBnki4p7+OCWHhgunY2J0fnrAa555Ots0XSdJWWxlAHY7v6/rSp6EpJa5CjdXRyQHV620VH85UW7M7kF0RDY8iXL0SLUQl6ef9tdHtLSUBfPttxa6KcBhIRT0fE1RFxfETsTo5v/noycA25WNPr5mFx57HoSJGtUoitTg4aOJehBw9sDGDYeMyoTPsR+WPAB8hTwc3I4W07HvEyIg4kh9r4BWWIEEkjufdrczIZQdUW5JjmiwzR24fTwGa/IccTb7iHRYfFbgw8d28X8w1nQ/L//NtK2QvK3+YB/oCe33+3de52uebPtpYmRMtqCI2RNldtMe1WMqHFDuQwGQ3bN72+n0UHA1xoDGlJk8g+nunkWD7fl3RpRFxFB8q3+obAbhHxs1K25hD1HlbkwHY3AF+SdDI5plfbAeckLUsenNc2TdqCzJLSSi+nQa1sTv5Y0fBrsrV6dKVsV2BmLDy+eKfzDacxwunTML8P8NhSNtT4XL28/27r3O1yzZ9tPQ26h3+kD1r/Qnc5Obb4luSp007kt+RDwMplnjNY+Nez48hfed5Knu4cRrakqplmDiAHEfsP8tTxUPKXxxmVeT5NjqTYyA7zVbKDfWplnoW23VT3SWRQPJ9slb2UvMzhMeDIodZRfV3qdiz5i+DaZR3TgbM7/Ey3IVuUqzeV30tJXdbn/+GXWJCSbTPyx4V5wBsq82xVPvvPkcM870OORPrepnV1NF+b+qxR3v8JZV+YRgbuINNYLdXn99+2zmROwb91sVzbz7Yy7wxq9GvgwCvQxT/6MhYNVoeWA/x+8pvwH2Rrp5p/rflg7+TShaXIX4QayVVPIr/VZpTpLys7zxsryyxDXjZw3lDbbvGedijLNBJS7km2/o4cpv7zX5OnoeeTHaxPkn1Mp1BSZbEg5dc6Q2z/veRQwtWyxgG8+Sj8D8+t1PV+crz1V7WY73Xlc3mS/AI4eIj1DTtfu/df5jmYDM5PkiNaTiV/3JgLrD8Kn0G7Oh9Ji6SfHSzX0Wdb5p1BjYJV7YaIkXQZeR3JAYOuS11IOpoMgJtGpsNarCzu738okmaQX/xdZZsZa3XtYN+3dGDuOuiK1MTryYtmF9cDdXF//wuR9LlyGc1abWceR+rYslqDBYPf3x05QL6Zdaj8qNPIUDMrOkgxNx7ULliZ2eKprqeBZraYcbAys1qYyBeFWh9o2aWCFZYZm409+iQx59/NdwyYAQ5W1s4Ky8CeG4/Nts67YWy2Y7Xk00AzqwUHKzOrBQcrM6sFByszqwUHKzOrBQcrM6sFByszqwUHKzOrBQercUDSMyVdIOkxSTMlvWOI+STpWGXG6VmSjivpvxrTN5N0naTHy9/NKtNeKelSSY+UcYzMasXBanw4iRytdDUyD+HXJW3UYr6DgD2ATYEXk+M0vQdA0tJktt5zyOE/zgQuLOWQwySfTmY5MasdB6sBKwlE9wQ+HRGzIxNNXERmA262L3B8RNwZEXeRue72K9N2JG+fOjEinoyIr5AZW14FEBG/i4izgdtG8/2YjRYHq8HbAJgbETdXyq4HWrWsNirTWs23EZk2qzpA2Z+HWI9Z7fhG5sGbAjzSVPYIsEIH8z4CTCn9ViNZz7BKRuiDcosdpy40G1VuWQ3ebDL9e9WKwKMdzLsiMLu0pkaynmFFxLSI2DIitmTZVjldzcaeg9Xg3QwsKen5lbJNWTjZasP0Mq3VfNOBF1d/HSQ74Vutx6x2HKwGLCIeI3P+HS1peUkvB3YHzm4x+1nAwZLWkPQc4BAyfyBkPsW5wIclLSPpg6X8EsjM0ZImk7kQJWly5ZdCs3HPwWp8eD+Zsed+Mknl+yJiuqRXlJRJDd8Afgj8hUwT/+NSRkQ8RV7WsA+ZkHV/YI9SDrA9mb33J2QKpjnAz0f5fZn1jbPb2LC06pQYy5FC4/7ZHtbYWnLLysxqwcHKzGrBwcrMasHBysxqwcHKzGrBwcrMasHBysxqwddZ2bDW0+T4LGuPybYOZya3xRO+zspacsvKzGrBwcrMasHBysxqwcHKzGrBwcrMasHBysxqwcHKzGrBwcrMasHBysxqwcFqHBij9PFHSvq3pNmVx3pj8f7M+sHBanwYi/TxAN+NiCmVh7MzW204WA3YWKWPN6s7B6vBG8v08W+Q9KCk6ZLe13vVzcaO08cP3lilj/8eMA24D9gGOE/SwxFxbvNGqunjV/YuYuOEW1aDNybp4yPixoi4OyLmRsTVwJeBt7SqUDV9/AosMeI3ZDYaHKwGb1Dp44Ps0zKrBQerARvD9PG7S5paLn/YGvgw+euhWS04WI0PY5E+fi/gVvK08Czg2Ig4c7TfmFm/eFhjG5aHNbbxwi0rM6sFByszqwUHKzOrBQcrM6sFByszqwUHKzOrBd/41UeSFgr+ETFvUHUxm2h8nVWPJG1Bjkf1YmByoxiIiKj9jXVadUqw58Zjs7HzbiDun+3rrKwlt6x6dyZ5Vfn+wOMDrovZhOVg1bu1gf8ON1HNRpU72Ht3AfDqQVfCbKJzy6p3k4ELJF0F3FudEBH7DKZKZhOPg1XvbiwPMxtFDlY9ioijBl0Hs8WBg1UfSHolmY1mDeAu4JyIuGSwtTKbWNzB3iNJBwDfJfurzgfuAb4t6cCBVsxsgnHLqnefAHaJiPkpsiR9FzgPOGVgtTKbYNyy6t2zWLSD/Sbgmf3aQB/Ty0+TdJOkeZL261f9zMaCg1XvrgK+JGk5mJ9h+X+Bq/u4jZ7TyxfXk+O9/6GPdTMbEw5WvXsvGRgekXQfmaxhUxYOEl3rY3p5IuKkiPgV8EQ/6mY2ltxn1aOIuAfYQdJzgdWBuyPizj5uYqj08ju0mHe49PJmteZg1QVJatwLWBkW5q7ymF/WpyFi+pJefiT3LlbTxzNl6RFV1my0+DSwO9WA8DTw76ZHo6wf+pVevmPV9PEsu9SIKms2Wtyy6k711GrdUd7W/PTyEXFLKWuXXv53beYzqx23rLoQEXdUXr41ImY2P8hO8X5sq1/p5ZG0tKTJ5OCAS0ma3Dy6qdl45R21d0cMUX54H7fRc3r54ufAHOBlwLTyfPs+1tNs1Pg0sEuSXlWeLlHuDawOx7serfuUuhIRD5LXTzWXX0l2qjdeB3lF/SeGWM+O/aqT2VhzsOreaeXvZOD0SnmQ9wl+aMxrZDaBOVh1KSLWBZB0lgfZMxt97rPqkQOV2dhwy6pHklYEjiSvKF+ZSt9VRKw1oGqZTThuWfXua8AWwNHkSAsfAv4BnDDISplNNG5Z9e7VwIsiYpakuRFxoaRryUsIHLDM+sQtq95NYsHtN7MlrUSOFvq8wVXJbOJxy6p3jREQfgVcSY49NZu8TcbM+sQtq94dCMwozz9MXhW+EuBfCc36yC2rHkXEbZXnDwAHDLA6ZhOWg1UXJO0dEWeX5/sPNV9EnD7UNDMbGQer7rydBaMetBpeGPK2Gwcrsz5xsOpCROxWef7KQdbFbHHhDvYeSTpR0laDrofZROdg1TsBF0q6RdJRkl4w6AqZTUQOVj2KiI8Aa5ID5D0XuEbSdZIOHmzNzCYWB6s+iIh5EfGLiNgf2BiYRSY6NbM+cbDqA0lTJL1T0o/JK9efJhOOmlmfOFj1SNL3yZFBDwJ+BKwdEbtFxDl93MYzJV0g6TFJMyW9Y4j5JOlYSbPK4zhJKtM2kHShpAckPSjpZ+5fszpxsOrdtcCGEbF9RHw9Iv45Cts4CXgKWA34T+DrklplWj6IHKt9UzKl/etZkMZ+JTLt/AvKen4HXDgKdTUbFRph/ktrQdJSwLbAcyLiu5KWh/lptHpd9/LAQ8DGjRTyks4G7oqIQ5vmvRo4IyKmldfvBg6MiG1brPeZZN/ayhExa8jtrzol2HPjXt9GZ867gbh/ttrPaIsjt6x6JGkTsp/qFBYkkdiB/l29vgEwtxGoiutZONFqw0ZlWrv5IFNw3dsqUEk6SNK1kq5lTr8SS5v1xsGqd18HjoiIF7IgZfzlwHZ9Wv8UFk5XT3m9QgfzPgJMafRbNUhakzy1bHl5hdPH23jkYNW7jYBGZ3rA/NO/Zfu0/tnAik1lK9I6L2HzvCsCs6Nyri9pFTLZ6dci4tw+1dFs1DlY9W4G8JJqgaStgVv7tP6bgSUlPb9StikwvcW808u0lvNJmkoGqosi4nN9qp/ZmHCw6t2ngR9LOgpYWtKngO/Tp/TxpZV2PnC0pOUlvRzYnQWjPlSdBRwsaQ1JzwEOAc6A+Vl4fgb8urlj3qwOHKx6FBE/Al4LrEL2Va0NvDkift7HzbyfPK28HzgXeF9ETJf0CkmzK/N9g0xU8RfgBuDHpQzgTcBWwLskza48nC7MasGXLtiwfOmCjRcez6oLko7uZL6IOGK062K2uHCw6s5zK88nA3sCvwdmAmsBWwPnDaBeZhOWg1UXIuJdjeeSvgO8PSLOq5S9GXjrIOpmNlG5g713rwV+0FR2IbBbi3nNrEsOVr27FfhAU9n7gb8PoC5mE5ZPA3t3AHCBpE8AdwFrkONZvXmgtTKbYBysehQRfyxXl28LPAe4B/hNRPgOYLM+crDqgxKYrhx0PcwmMvdZmVktOFiZWS04WJlZLbjPqguSOgryETFvtOtitrhwsOrO05SB9tpYYrQrYra4cLDqzrqV568D3gL8D3lv4NrAJ/G9gWZ95WDVhYiY2Xhe0sRvGREPl6KbJV1Lpuj6+iDqZzYRuYO9d88AlmsqW66Um1mfuGXVuzOBX0o6EbiDHD7mw6XczPrELavefQL4CvA24EvAXsD/lfK+6Ef6+DI9yjoaQxqf2q86mo02t6x6IGkJ4DPA5yLi5FHcVDV9/GZkgorrI6I5w001fXwAvwBuA6p12zQi+pV5x2zMuGXVg4iYSw4PM2o3LZf08XsCn46I2RFxFXARsHeL2fcFjo+IOyPiLuB4YL/RqpvZWHKw6t2ZwHtHcf39Th9/haR7JZ0vaZ1+VtRsNDlY9W5r4MuSZki6UtIVjUef1t/P9PE7AOsALwTuBn4kaZGuAEkHSbpW0rXM8Ug3Nj64z6p3p5THaOlb+viIaATQpyR9BPgX8CIyz+B8ETENmAYlFZfZOOBg1aOIGO1LFOanj4+IW0pZu/Txv2szX0MAztNnteBg1QeSViNPB1emcvBHxOm9rjsiHpPUSB9/APlr4O7Ay1rM3kgf/xMyEB0CfLXUcSNgKbIVtSzwWXIY5r/2WkezseBg1SNJewDnALeQndnTgY2Bq4Ceg1Xx/rKu+4FZVNLHAz+NiCllvm8A67HgtO5UFqSPX428/WdN4DHgauD1Hn7Z6sLp43sk6QbgqIj4vqSHImKqpHcBG0XEfw26fr1y+ngbL/xrYO/WiojvN5WdCewziMqYTVQOVr27v/RZAcyQ9FJgfTyWlVlfOVj17hRgu/L8BOBS8mLMrw2sRmYTkDvYexQRx1aenyXpMmD5iPCvbGZ95GDVZxHxj0HXwWwicrDqgqQ76GAM9ohYawyqY7ZYcLDqzjsrz7ciRzv4CgvGYP8geYGmmfWJg1UXIuLyxnNJJwGvKUOyNMp+ClxMDtFiZn3gXwN79xzyBuKq2cAaA6iL2YTlYNW7i4CLJO0i6UWSXg1cUMrNrE8crHr3XuA35NDBfyh/f8voDshntthxn1WPIuIJ4NDyMLNR4mDVB5JeQI4dNaVa3o8hYswsOVj1SNJhwBHkLTaPVyYF/Rsixmyx52DVu48CW0fEnwddEbOJzB3svZsD/G3QlTCb6Bysevdp4KuSVpc0qfoYdMXMJhIfUL07AzgQuJNMdvpv4GlGMfFp1QhSy79S0qWSHpE0YyzqZtZP7rPq3boD3n6nqeW5RKqKAAAHwklEQVQfIzv8zwUOG9sqmvXOwapHETFzUNuupJbfOCJmA1dJaqSWX+i6r4j4HfA7STuPfU3Neudg1QeS3khmO25OxTXa47APlVp+h1HertmYc59VjyR9hkx3NQl4K5kq6zXAw2Ow+ZGklu+Y08fbeORg1bv9gV0i4mPAU+XvG4B1xmDbI0kt37GImBYRW0bEliy7VC+rMusbB6verRQRN5TnT0laqvQPjcWp2PzU8pWydinjzWrJwap3fy+p2QFuAN4naW/godHecEQ8BjRSyy8v6eVkavmzm+ct135NJlPIS9JkSUuPdh3N+sUd7L07HHhWeX4o8G2yL+kDY7T9TlPLb0+mCWuYA1wO7DhG9TTridPH27CcPt7GC58G9kjSg0OU3z/WdTGbyByserfIz2WSlsLp4836yn1WXZJ0JTlm1WRJVzRNXhO4euxrZTZxOVh171TyavWtgNMq5QHcB1wyiEqZTVQOVl2KiDMBJF0TER7PymyUuc+qS5JeImnjRqCStIqkb0m6XtLJkqa0W4eZdc7BqnsnAs+uvD6VvLF4GrAxcNwgKmU2Ufk0sHsvAq4EkLQS8FpyqJabyzAtV5MXbJpZH7hl1b0lyUHvALYF7m0M1RIRdwArDapiZhORg1X3ppNDwgDsBfyyMUHSGiw6dIuZ9cCngd37JPBDSScDc4HtKtPeBvx6ILUym6AcrLoUEVdJWovsVL85IqpjSP0Y+M5gamY2MTlY9aAEqOtalN80gOqYTWjuszKzWnCwMrNacLAys1pwsDKzWnCwqrkRpI//uKQbJD0q6XZJHx/rupr1wr8G1l+n6eMF7AP8GVgf+LmkOyLCl1hYLbhlVWOV9PGfjojZEXEV0Egfv5CIOC4i/hART5dLKy4EXj62NTbrnoNVvQ2VPn6jIeYHMg8X8AqcX9BqxKeB9dZt+vgjyS+qb7aaKOkg4KDcglML2vjgllW9jTh9vKQPkn1Xr4uIJ1vN4/TxNh45WNXbiNLHS9qfTMS6U0TcOQb1M+sbB6saG2H6+P8EPg/sEhG3jW1NzXrnYFV/7weWJdPHn0slfbyk2ZX5Pkumuf+9pNnlcfIA6mvWFXew11xEPAjs0aL8SrIDvvF63bGsl1m/uWVlZrXgYGVmteBgZWa14GBlZrXgYGVmteBgZWa14GBlZrXgYGVmteBgZWa14GBlZrXgYGVmteBgZWa14GBlZrXgYGVmteBgZWa14GBlZrXgYGVmteBgVXMjSB9/pKR/V4Y0ni1pvbGur1m3PKxx/XWaPh7guxHxzjGtnVmfuGVVYyNJH29Wdw5W9TbS9PFvkPSgpOmS3jf61TPrH58G1ttI0sd/D5gG3AdsA5wn6eGIOLd5RqePt/HILat66zh9fETcGBF3R8TciLga+DLwllYrdfp4G48crOptROnjmwSgUamV2ShwsKqxEaaP313SVKWtgQ8DF45tjc2652BVf52mj98LuJU8RTwLODYizhzz2pp1yR3sNTeC9PFvH8t6mfWbW1ZmVgsOVmZWCw5WZlYLDlZmVgsOVmZWCw5WZlYLDlZmVgsOVmZWC4qIQdfBxjFJDwAzu1h0ZeCfI1xm7YhYpYtt2WLAwcpGhaRrI2LLQdfDJg6fBppZLThYmVktOFjZaJk26ArYxOI+KzOrBbeszKwWHKzMrBYcrKwrkj4o6VpJT0o6o828H5N0r6RHJJ0uaZkxqqZNIA5W1q27gc8Cpw83k6TXAIcCOwHrAOsBR4125WzicbCyrkTE+RHxA2BWm1n3BU6LiOkR8RBwDLDfaNfPJh4HKxttG5FZohuuB1aT9KwB1cdqysHKRltz1ujG81ZZo82G5GBlo605a3Tj+SJZo82G42Blo206mSW6YVPgvoho19dlthAHK+uKpCUlTQaWAJaQNFlSqzyUZwHvlrShpKnA4cAZY1hVmyAcrKxbhwNzyMsS3lmeHy5pLUmzJa0FEBEXA8cBl5LjYs0EPjOYKlud+d5AM6sFt6zMrBYcrMysFhyszKwWHKzMrBYcrMysFhyszKwWHKzMrBYcrGxIkmZIekrSyk3lf5IUktYZTM1sceRgZe3cDry98ULSJsCyg6uOLa4crKyds4F9Kq/3Je/3A0DSMpK+KOkfku6TdLKkZcu0qZJ+JOkBSQ+V52tWlr1M0jGSfi3pUUk/b27FmTU4WFk71wArSnqRpCWAtwHnVKYfC2wAbAY8D1gDOKJMmwR8E1gbWIu8f/D/mtb/DuBdwKrA0sB/jc7bsLpzsLJONFpXuwB/A+4q5QIOBD4WEQ9GxKPA54G9ACJiVkScFxGPl2mfA3ZoWvc3I+LmiJgDfI8MemaLaDWkh1mzs4ErgHWpnAICqwDLAddJapSJHDYGScsBJwC7AlPL9BUkLRERc8vreyvre5wcWdRsEW5ZWVsRMZPsaN8NOL8y6Z/kqd1GEbFSeTwjIhoB5xDgBcA2EbEisH0pF2Yj5GBlnXo38KqIeKxSNg84BThB0qoAktYo6bcgx1mfAzws6Zl4HCvrgYOVdSQi/h4R17aY9EngVuAaSf8Cfkm2pgBOJC9z+CfZUX/xWNTVJiYPvmdmteCWlZnVgoOVmdWCg5WZ1YKDlZnVgoOVmdWCg5WZ1YKDlZnVgoOVmdWCg5WZ1cL/ByQyD3oZPWm8AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def normalcy_test_summary(df, match, alpha, net, n):\n",
    "    means = df['mean'].unique()\n",
    "    means.sort()\n",
    "    sds = df_trials['sd'].unique()\n",
    "    sds.sort()\n",
    "    \n",
    "    dat = np.zeros((len(sds), len(means)))\n",
    "    \n",
    "    for j,mean in enumerate(means):\n",
    "        for i,sd in enumerate(sds):\n",
    "            dat[i,j] = df.loc[(df['mean']==mean) & (df['sd']==sd), match.lower() + '_p'] > alpha\n",
    "    plt.xticks(range(len(means)), labels=means)\n",
    "    plt.yticks(range(len(sds)), labels=sds)\n",
    "    plt.xlabel('Mean')\n",
    "    plt.ylabel('Standard deviation')\n",
    "    plt.title('Parameters passing {} normalcy test\\n'.format(match)+r'[{}, $n={}$, $\\alpha ={}$]'.format(net, n, alpha))\n",
    "    if dat.all():\n",
    "        plt.imshow(dat, cmap='ocean')\n",
    "    else:\n",
    "        plt.imshow(dat, cmap='RdYlGn')\n",
    "    plt.savefig('output/plots/{}_normalcy_{}_a{}.png'.format(net, match.lower(), alpha), dpi='figure')\n",
    "    plt.show()\n",
    "\n",
    "normalcy_test_summary(df_pairs, 'TSTT', 0.05, net, n)\n",
    "normalcy_test_summary(df_pairs, 'WVC', 0.05, net, n)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Perturb a subset of nodes\n",
    "\n",
    "How does TSTT change when we perturb only a subset of all nodes?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>mean</th>\n",
       "      <th>p</th>\n",
       "      <th>method</th>\n",
       "      <th>tstt</th>\n",
       "      <th>wvc</th>\n",
       "      <th>vmt</th>\n",
       "      <th>costs_rmse</th>\n",
       "      <th>cost_err25</th>\n",
       "      <th>cost_err75</th>\n",
       "      <th>flows_rmses</th>\n",
       "      <th>flow_err25</th>\n",
       "      <th>flow_err75</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.25</td>\n",
       "      <td>By origin</td>\n",
       "      <td>4.429737e+06</td>\n",
       "      <td>1.228819</td>\n",
       "      <td>2.690839e+06</td>\n",
       "      <td>3.553178</td>\n",
       "      <td>-4.216486</td>\n",
       "      <td>-0.133485</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-3074.739448</td>\n",
       "      <td>-696.865251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.25</td>\n",
       "      <td>By destination</td>\n",
       "      <td>4.430935e+06</td>\n",
       "      <td>1.229201</td>\n",
       "      <td>2.689873e+06</td>\n",
       "      <td>3.560649</td>\n",
       "      <td>-4.241476</td>\n",
       "      <td>-0.126547</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-3034.491615</td>\n",
       "      <td>-685.665704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.25</td>\n",
       "      <td>Both</td>\n",
       "      <td>2.949053e+06</td>\n",
       "      <td>1.037014</td>\n",
       "      <td>2.201476e+06</td>\n",
       "      <td>4.593235</td>\n",
       "      <td>-6.014135</td>\n",
       "      <td>-0.671673</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-5063.134254</td>\n",
       "      <td>-2102.847820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.50</td>\n",
       "      <td>By origin</td>\n",
       "      <td>3.241227e+06</td>\n",
       "      <td>1.077114</td>\n",
       "      <td>2.309530e+06</td>\n",
       "      <td>4.481714</td>\n",
       "      <td>-5.509276</td>\n",
       "      <td>-0.198472</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-4905.500750</td>\n",
       "      <td>-1639.048031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.5</td>\n",
       "      <td>0.50</td>\n",
       "      <td>By destination</td>\n",
       "      <td>3.241529e+06</td>\n",
       "      <td>1.077144</td>\n",
       "      <td>2.309123e+06</td>\n",
       "      <td>4.485378</td>\n",
       "      <td>-5.594514</td>\n",
       "      <td>-0.203936</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-4962.360305</td>\n",
       "      <td>-1649.373501</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean     p          method          tstt       wvc           vmt  \\\n",
       "0   0.5  0.25       By origin  4.429737e+06  1.228819  2.690839e+06   \n",
       "1   0.5  0.25  By destination  4.430935e+06  1.229201  2.689873e+06   \n",
       "2   0.5  0.25            Both  2.949053e+06  1.037014  2.201476e+06   \n",
       "3   0.5  0.50       By origin  3.241227e+06  1.077114  2.309530e+06   \n",
       "4   0.5  0.50  By destination  3.241529e+06  1.077144  2.309123e+06   \n",
       "\n",
       "   costs_rmse  cost_err25  cost_err75  flows_rmses   flow_err25   flow_err75  \n",
       "0    3.553178   -4.216486   -0.133485          NaN -3074.739448  -696.865251  \n",
       "1    3.560649   -4.241476   -0.126547          NaN -3034.491615  -685.665704  \n",
       "2    4.593235   -6.014135   -0.671673          NaN -5063.134254 -2102.847820  \n",
       "3    4.481714   -5.509276   -0.198472          NaN -4905.500750 -1639.048031  \n",
       "4    4.485378   -5.594514   -0.203936          NaN -4962.360305 -1649.373501  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def partial_perturber(netFilename, tripsFilename, nodeFilename, proportion, true_costs, true_flows,\n",
    "                      norm_mean=1, norm_sd=0, show_plot=False):\n",
    "    '''\n",
    "    Perturb <proportion> of zones closest to geographic center by Euclidean distance.\n",
    "    Note that we perturb only zones (nodes with origin and destination flow), but that\n",
    "    we measure by proportion of all nodes and only then filter for just zones.\n",
    "    '''\n",
    "    nodes_to_perturb = central_node_selector(nodeFilename, proportion, make_plot=show_plot, show_plot=show_plot)\n",
    "    if show_plot:\n",
    "        nodes_to_perturb = nodes_to_perturb[0]\n",
    "    \n",
    "    tstt = [None]*3\n",
    "    congestion = [None]*3\n",
    "    vmt = [None]*3\n",
    "    costs_rmse = [None]*3\n",
    "    cost_err25 = [None]*3 \n",
    "    cost_err75 = [None]*3\n",
    "    flows_rmses = [None]*3\n",
    "    flow_err25 = [None]*3\n",
    "    flow_err75 = [None]*3\n",
    "    #most_congested_link_vc = [None]*3\n",
    "    \n",
    "    tstt[0], congestion[0], vmt[0], costs_rmse[0], cost_err25[0], cost_err75[0], flows_rmses[0], flow_err25[0], flow_err75[0] = tap_trial(\n",
    "                                        netFilename, tripsFilename, norm_mean=norm_mean, norm_sd=norm_sd, returnDataFrame=False,\n",
    "                                        nodesPerturbedIfOrig=nodes_to_perturb, true_costs=true_costs, true_flows=true_flows)\n",
    "    tstt[1], congestion[1], vmt[1], costs_rmse[1], cost_err25[1], cost_err75[1], flows_rmses[1], flow_err25[1], flow_err75[1] = tap_trial(\n",
    "                                        netFilename, tripsFilename, norm_mean=norm_mean, norm_sd=norm_sd, returnDataFrame=False,\n",
    "                                        nodesPerturbedIfDest=nodes_to_perturb, true_costs=true_costs, true_flows=true_flows)\n",
    "    tstt[2], congestion[2], vmt[2], costs_rmse[2], cost_err25[2], cost_err75[2], flows_rmses[2], flow_err25[2], flow_err75[2] = tap_trial(\n",
    "                                        netFilename, tripsFilename, norm_mean=norm_mean, norm_sd=norm_sd, returnDataFrame=False,\n",
    "                                        nodesPerturbedAlways=nodes_to_perturb, true_costs=true_costs, true_flows=true_flows)\n",
    "    return tstt, congestion, vmt, costs_rmse, cost_err25, cost_err75, flows_rmses, flow_err25, flow_err75\n",
    "\n",
    "\n",
    "def partial_perturber_sim(netFilename, tripsFilename, nodeFilename, net, true_costs, true_flows):\n",
    "    \n",
    "    df = pd.DataFrame(columns=['mean', 'p', 'method', 'tstt', 'wvc', 'vmt', 'costs_rmse',\n",
    "                              'cost_err25', 'cost_err75', 'flows_rmses', 'flow_err25', 'flow_err75'])\n",
    "    counter = 0\n",
    "    \n",
    "    means = np.arange(.5, 1.6, .25)\n",
    "    means = means[means != 1]\n",
    "    ps = np.arange(.25, 1.1, .25)\n",
    "    methods = ['By origin', 'By destination', 'Both']\n",
    "    \n",
    "    for k,mean in enumerate(means):\n",
    "        if mean==1:\n",
    "            continue\n",
    "        for p in ps:\n",
    "            tstt, congestion, vmt, costs_rmse, cost_err25, cost_err75, flows_rmses, flow_err25, flow_err75 = partial_perturber(\n",
    "                netFilename, tripsFilename, nodeFilename, p, true_costs, true_flows, norm_mean=mean, show_plot=(k==0))\n",
    "            for i,method in enumerate(methods):\n",
    "                df.loc[counter]={\n",
    "                    'mean': mean,\n",
    "                    'p': p,\n",
    "                    'method': method,\n",
    "                    'tstt': tstt[i],\n",
    "                    'wvc': congestion[i],\n",
    "                    'vmt': vmt[i],\n",
    "                    'costs_rmse': costs_rmse[i],\n",
    "                    'cost_err25': cost_err25[i],\n",
    "                    'cost_err75': cost_err75[i],\n",
    "                    'flows_rmse': flows_rmses[i],\n",
    "                    'flow_err25': flow_err25[i],\n",
    "                    'flow_err75': flow_err75[i]\n",
    "                }\n",
    "                counter += 1\n",
    "    \n",
    "    df.to_csv('output/data/{}_trials_partial_perturb.csv'.format(net), index=False)\n",
    "    \n",
    "    return df\n",
    "\n",
    "nodeFilename = 'inputs/{}_node.tntp'.format(net)\n",
    "try:\n",
    "    df_partial = pd.read_csv('output/data/{}_trials_partial_perturb.csv'.format(net))\n",
    "except FileNotFoundError:\n",
    "    df_partial = partial_perturber_sim(netFilename, tripsFilename, nodeFilename, net, true_costs, true_flows)#, true_most_congested_link_id,)\n",
    "df_partial.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>mean</th>\n",
       "      <th>tstt</th>\n",
       "      <th>wvc</th>\n",
       "      <th>vmt</th>\n",
       "      <th>costs_rmse</th>\n",
       "      <th>cost_err25</th>\n",
       "      <th>cost_err75</th>\n",
       "      <th>flows_rmse</th>\n",
       "      <th>flow_err25</th>\n",
       "      <th>flow_err75</th>\n",
       "      <th>p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>7.480171e+06</td>\n",
       "      <td>1.47403</td>\n",
       "      <td>3.419126e+06</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean          tstt      wvc           vmt  costs_rmse  cost_err25  \\\n",
       "4   1.0  7.480171e+06  1.47403  3.419126e+06         0.0         0.0   \n",
       "\n",
       "   cost_err75  flows_rmse  flow_err25  flow_err75  p  \n",
       "4         0.0         0.0         0.0         0.0  1  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = dfmeans[dfmeans['mean']==1].copy()\n",
    "a.loc[:,'p'] = [1]\n",
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA5wAAANUCAYAAADfCSIIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xd4VNX28PHvIiGEkEJoIQRDDb0kJiCiSCheBJRuQEBAQATpKEVFRV4QLlIuIEV6EWmiFBW5igRFuCK9KRAgSK+hhITU/f4xE34xpgFJJmV9nmceM+fss886gzNn1uwmxhiUUkoppZRSSqmMls/WASillFJKKaWUyp004VRKKaWUUkoplSk04VRKKaWUUkoplSk04VRKKaWUUkoplSk04VRKKaWUUkoplSk04VRKKaWUUkoplSk04VQ5gogsEZEfEz0fIyIhWXDeUBEZndnnUUoppXIzvY8rlXdpwplHWD/ojYhMS2afEZGutojrMUwG6tkygESvaWqPQGvZ10Rkr4jcEZG7IvKHiMx/iHqC01GmrA1fDqXSRUR6iUiMiLgk2X4ole23RSRSRIqkUOcmEdmR6LmTiIy2HhshIjdF5DcRGSgiTplzZUplLr2PZzwRaWt97aqlsH+miJwVkXwi0sNa9rKI5E9SrriIRFn3P5uobGqPMVlykUplA5pw5i2RQH8RqZTRFYuIQ0bXmRpjTLgx5npWnjMZgwHPRI9QYEqSbTtFpAcwF1gMBAD+wCjA7iHqCUryPA4YkmTbuUy7UqUyzo+APdAwYYOIFAOqA5eS2V4DGGjd9GrSykTEC2gOzLM+dwV+tR4zC6iP5T03Gcv76F8ZfUFKZSG9j2esTcBF4PWkO0TEEegKLDTGxFs3xwGxwEtJir+G5fMrwWr+fn/+AtiVZNvkDLsKpbI5TTjzlp3AXuCT1AqJiKeIrBKRW9ZWhWARCUi0P9D661xLEdkhIveBPtZf9GJFpJGIHLYeu11ESonIcyKyX0TuiciP1i+JCfWVE5GvROSitTXisIj844tlkhj/1hUnrVY/EXEWkekicsF6jv0i0i5JnbVFZKeI3BeREyISlFoMxpjbxpjLCQ8sN6LwxNuMMdFAG2CTMeZTY8wJ62ODMabnQ9RzNUkZgNtJysSlFq9S2YEx5ixwCmiSaHNj4AiwIZntAmwFviSZL4VAT+AusNb6fDxQBahnjPnMGHPAGHPGGLMWeA4IzrirUSrL6X08Y+/jscAi4FURKZBk98uAC7AwyfZFJPosEhEBeicuZ4yJTHLPjgSik9yzw1OLTancRBPOvGco8JKINEpup/WDcz2WL2wvAnWBK8AP1taGxKYAk4Cq1mPA8v/Uh1g+fJ8BSmH5pW8s0A94FigNTE1UjzOWL5QvADWxtFQsTinGFCT+1bAUllaUP4Ar1mvaBNQGOmJpMZkDrBKRJtbrLgh8B9wCngK6A8OBEg8RQ0ouAQGZ8Yu0UjnUVv6ZWP5kfSTd/qcx5gLwGVBdRJ5O2Cki+YBewHJjTKT1eWdghTHmTNKTGotbGX41SmUtvY9n7H18AeAOtEuy/XXgO+vnT9LyTUWkjPV5I6A4lh/FlFLJsLd1ACprGWP+JyKrgaki4p+om0iCxlhuTtWNMccARKQblm6eb2K54SQYb4zZmPDEcj9AgCHGmAPWbfOw3MwCjDF7rds+A95LFNNh4HCiemeKSFMsXxy3pfO6Elr9EJHxQC0sLRyRYhlH+TTgYYy5bS02T0TqYel2txXoArgBXYwxYdZ6XksS16P6CMsN+LiInAV+s57zc2NMRAbUr1ROsxV4XURKGGOuYvnceRvYAVRLsn0LgDFmh4gcw/IlcJe1nueBMli70wLFgCLAsSy7EqWymN7HM/Y+bow5KyJbsHy2rLQeVxlowD+7zmKM+UtEfsDSu+JDoA+wHEsrplIqGdrCmTeNwvLLZ49k9lUHbiTcpACMMVFYkqTqScruTuZ4w98/3BNuIIeSbCsqInbwYIKPiSJyVCyTe4QDLbB8kXwo1i48w4A2iVo46gAOwAURCU94YBmb4WMtUw34I+EmBWCMOQLc5jFZu848az3HBOAelpv3URHJiBZUpXKan6z/bSIipYHywHZjzE0snxUJ232wtHIkmAd0FMs4TbAmn9b3Kli+KIPlc0ip3Ezv4xl7H/8MCBSRitbnrwPngc0plJ8H9BQRD6AtMD/dF6hUHqQJZx5kHUM1DRgnIs7JFUlmmySz/V4y5eKTjCU01nPGJFN/wpfDT7DcNMZi6Zrii6VbzENNYCAiz2K5CfQwxuxKtCsflhuOb5JHNSyTjSTEkqlfUo0xf1jHlPUE/LB0SeqXmedUKjuyThRyEEv32SbAvkStFtsSbY/j72Mul2F5r3axftFrxf+1bgJcA8L455dqpXIVvY9n+H38G6yTB4ll8qTuWCYLSmluhG+sMa3A8vmVEb2hlMq1NOHMuyZg+fcfmWT7UaCYJJoi3DqQvq51X2Z4DsuYq9XGmIPAaeChxjuKSHnga2CcMWZ1kt17gMKAozEmJMnjL2uZo1i68hVOVGd1LN1zMkMoEEHGjBFVKidKGMeZMH4zwbZE2/ckSkSxtlysxdL60APLe2hNov3xWGaD7CIi5ZKeUCwy6z2tVFbT+3gG3cetieVCLJ8rQVi65iedLChx+YTJhpqgrZtKpUkTzjzKGHMXeB94K8mun7B0sflCRJ4RkRpYWhUcsQzQzwzHgdYiUtd6g5yHZcKAdLFOFPANlnFdC0WkZKKHHZZr+hH4SixrbpUXEX+xrMmXMNPcF1hmuvzcOstdPSw3k8cekyEic0TkQxFpICJlRMQfWAq48n+TNCiV12wFymLpjpY44fwZeMK6fWsyx32GpYfASJIfB/0ecBL4n4j0sb6fy4lIW2A7ltYXpXI8vY9n+H18AVAUmAFsNsaktdTYWCyTBS1N73UqlVdpwpm3LcTyxewBY4zBsozHn8C3wO9ASeB5k3nrZQ0FzmJp2dgKXODhZnvzwDLD3ktYZoRN/HjCek2tgK+wzKqXcG0tsSzPgPVLawssN5vdWLrJTAOuPt6lAfADlnUAVwInsHQz8gRaGGN+yID6lcqJfgZisHwJ3pGw0RhzB8uyDy78ffxmwv6dWJZQcefv3WkT9t/GMrnIbGAQ8D9gH5Yxb6uxTkKkVC6h9/EMuo9bE8zNpPDZkkz5GGPMdV2STKm0ieU9rJRSSimllFJKZSxt4VRKKaWUUkoplSk04VRKKaWUUkoplSk04VRKKaWUUkoplSk04VRKKaWUUkoplSk04VRKKaVUtiUiA0Rkj4hEiciSNMqWF5FvROSuiFwXkUlZFKZSSqkU2Ns6gNykWLFipmzZsqmWuXfvHoUKFcqagDKIxpw1cmvMe/fuvW6MKZ5FIWUafX9nHxpz1shG7++LwDigGVAwpUIi4oBlGapZQEcgDqiUnhPo+zv7yGkx57R4IVu9t1UeoQlnBipbtix79uxJtUxwcDCBgYFZE1AG0ZizRm6NWUTOZk00mUvf39mHxpw1ssv72xjzlfVcAUDpVIr2AC4aY6Ym2nYoPefQ93f2kdNizmnxQvZ5b6u8Q7vUKqWUUio3qAeEishma3faYBGpaeuglFIqr9MWTqWUUkrlBqWBRkArYCswGNggIlWMMdFJC4tIH6APgIeHB8HBwalWHh4enmaZ7EZjznw5LV7ImTGrnE0TTqWUUkrlBpHADmPMZgARmQyMBqoCB5MWNsbMA+YBBAQEmLS6GObWrpPZTU6LOafFCzkzZpWzaZdapZRSSuUGhwBj6yCUUkr9nSacSimllMq2RMReRBwBO8BORBxFJLkeWp8D9USkqYjYAUOA68AfWRiuUkqpJDThVEoppVR2NhpLd9lRQFfr36NFxFtEwkXEG8AYc9y6fy4QBrQGWiU3flMppVTW0TGcSimllMq2jDFjgDEp7HZOUvYr4KtMDkkppdRD0BZOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhNOpZRSSimllFKZQhPONIiIj4jcF5HPbR2LUkoppZRSSuUkmnCmbRbwu62DUCovE5ECIrJQRM6KyF0R2S8izVMpP1RELovIbRFZJCIFEu0rKyLbRCRCRP4UkaZZcxVKKaWUUnmPJpypEJFOwC1gq61jUSqPswfOAQ0BN+B9YI2IlE1aUESaAaOAJkBZoDzwUaIiK4H9QFHgPeBLESmeeaErpZRSSuVdmnCmQERcgbHAW7aORam8zhhzzxgzxhgTaoyJN8Z8A5wB/JMp3h1YaIw5aowJA/4f0ANARCoBTwIfGmMijTHrgMNA+yy5EKWUUkqpPMbe1gFkY/8Py5fWcyKSYiER6QP0AfDw8CA4ODjVSsPDw9Msk91ozFlDY04/EfEAKgFHk9ldHdiQ6PlBwENEilr3nTbG3E2yv3oK59H3dzakMWeNnBizUkqp7EcTzmSIiC/QFPBLq6wxZh4wDyAgIMAEBgamWj44OJi0ymQ3GnPW0JjTR0TyAyuApcaYP5Mp4gzcTvQ84W+XZPYl7PdK7lz6/s6eNOaskRNjVkoplf1owpm8QCxjv/6ytm46A3YiUs0Y86QN41IqTxORfMByIBoYkEKxcMA10fOEv+8msy9h/12UUkoppVSG0zGcyZsHVAB8rY+5wLdAM1sGpVReJpZffxYCHkB7Y0xMCkWPArUTPa8NXDHG3LDuKy8iLkn2J9c1VymllFJKPSZNOJNhjIkwxlxOeGBpFblvjLlm69iUysPmAFWBl4wxkamUWwb0EpFqIuIOjAaWABhjTgAHgA9FxFFE2gK1gHWZGrlSSimlVB6lXWrTwRgzxtYxKJWXiUgZ4A0gCricaCKvN4BfgGNANWPMX8aY70VkErANKIglmfwwUXWdsCSgYcBfQAf9MUkppZRSKnNowqmUyvaMMWeBlKeLtoyzTlx+KjA1hbpCsYzTVkrlACIyAMvSRjWBlcaYHuk45iegEZDfGBObqQEqpZRKlSacSimllMrOLgLjsMyjUDCtwiLSBf1+o5RS2YZ+ICullFLqgd27d3P79m3y589v61AAMMZ8BSAiAUDp1MqKiBuWLvTdgF2ZH51SSqm0aMKplFJKKaKiohg7diwTJ07Ez8+PTz75xNYhPYqPsUwwdjmtgiLSB+gD4OHhQXBwcKrlw8PD0yyT3WjMmS+nxQs5M2aVs2nCqZRSSuVx+/bto3v37hw5coTXXnuNqVOncuDAAVuH9VCsLaDPAINJoyUUwBgzD8syaAQEBJjAwMBUywcHB5NWmexGY858OSnea9euMWTIECpUqMDYsWNtHY7KQ3RZFKWUUioPW7ZsGU899RQ3btxg06ZNLFq0iMKFC9s6rIciIvmA2cBgnSRIqb8zxrB8+XKqVq3K2rVruXnzpq1DUnmMJpxKKaVUHmSMAaB+/fp069aNI0eO8OKLL9o4qkfmCgQAq0XkMvC7dft5EWlgu7CUsq3Q0FCaN29Ot27dqFSpEgcOHKBDhw62DkvlMZpwKqWUUnnM5MmT6dSpE8YYKlasyMKFCylSpIitw0qWiNiLiCNgB9iJiKOIJB0SdBsoBfhaHy2s2/2B37IsWKWymfPnz7Nr1y4+/fRTduzYQbVq1WwdksqDdAynUkoplUdFRUXh6Oho6zDSMhrLzLMJugIficgi4BhQzRjzF4kmCrImqABXtIutymsOHz5McHAwAwcO5Nlnn+Wvv/7Czc3N1mGpPEwTTqWUUiqXi4uLY9q0aZQvX5527drx1ltvISK2DitdjDFjgDEp7HZO4ZhQIGdcoFIZbN68eaxdu5bu3bvj6uqqyaayOe1Sq5RSSuViJ06coEGDBgwfPpzNmzcD5JhkUymVPjt27GDPnj0AfPzxxxw5cgRXV1cbR6WUhSacSimlVC4UHx/P9OnT8fX15c8//2TFihXMmzfP1mEppTLQnTt3ePPNN2nQoAEffmjpee7i4kKxYsVsHJlS/0cTTqWUUiqXOXXqFIGBgQwZMoQmTZpw9OhROnfurC2bSuUiGzdupFq1anz22WcMHTqU1atX2zokpZKlCadSSimVS8THxzN79mxq1arFoUOHWLJkCRs3bsTT09PWoSmlMsiVK1fo2LEjrVu3xt3dnV27djF16lScnZMd0qyUzWnCqZRSSuUS/fr1o3///jRo0IAjR47QvXt3bdVUKpcwxrBkyRKqVq3K+vXrGTduHHv37qVu3bq2Dk2pVOkstUoppVQOZowhJiYGBwcHXnvtNQICAujdu7cmmkrlMl26dGHlypU8++yzzJ8/nypVqtg6JKXSRRNOpZRSKoeKjY2lVatWVKhQgZkzZ1KvXj3q1atn67CUUhkkNjYWEcHOzo727dvToEED3njjDfLl006KKufQhFMppZTKoezt7fHz88PLy8vWoSilMtitW7do2rQpnTt3ZtiwYbRv397WISn1SPTnEaWUUioHuXjxIq1bt2bHjh0AjB8/njfffNPGUSmlMooxBgA3Nzd8fX0pW7asbQNS6jFpwqmUUkrlAMYYVqxYQY0aNfjhhx84c+aMrUNSSmWw7du34+fnx6lTpxARFixYQLt27WwdllKPRbvUKqWUUtnclStX6Nu3L+vXr+fpp59myZIlVKpUydZhKaUyyK1btxg5ciTz5s2jfPny3LhxgwoVKtg6rGTt27evmb29/YfGmJJo45WCOGBHbGzs6/7+/tHJFdCEUymllMrG1qxZw5tvvkl4eDiffPIJQ4cOxc7OztZhKaUyyNdff03//v25cuUKw4cPZ8yYMTg5Odk6rGTt27evWYECBT4tW7ZsdMGCBcPy5ctnbB2Tsq34+Hg5e/bss7du3eoHTE+ujP4qoZTKEURkgIjsEZEoEVmSSrm5IhKe6BElIncT7Q8WkfuJ9h/PkgtQ6iFdv36doKAgOnbsSPny5dm3bx9vv/12liSbUVFRxMfHZ/p5lMrLLl26RPv27WnXrh0eHh7s3r2bSZMmZdtkE8De3v7DsmXLRhcqVChSk00FkC9fPlOqVKlwOzu7HimWybpwlFLqsVwExgGLUitkjOlrjHFOeAArgbVJig1IVKZyJsWr1CMzxtC4cWPWr1/P+PHj2blzJ9WqVcv08546dYrAwEA8PDz47bffMv18SuVFxhgWLFhA1apV+e6775g4cSK7d+/G39/f1qGlyRhTsmDBgvdtHYfKXhwcHGKMMW4p7dcutUqpHMEY8xWAiAQApdNzjIgUAtoDL2ZiaEplmLCwMFxcXLC3t2fatGkUL16cWrVqZeo5jTFs2bKFZcuWsWbNGuLi4qhQoQLu7u6Zel6l8qpz584xcOBAnnrqKebPn4+Pj4+tQ3oY+bRlUyUlIpBKQ6a2cCqlcrP2wDXg5yTbJ4jIdRH5VUQCsz4spf7p8uXLVK9encmTJwPQpEmTTE02o6KiWLZsGR4eHjRv3pxvvvmGQYMGcfDgQUJCQqhSpUqmnVupvCY2NpYvv/wSYwze3t789ttv/PTTTzkt2VTqkWgLp1IqN+sOLDMJi5pZjASOAdFAJ2CTiPgaY04lPVhE+gB9ADw8PAgODk71ZOHh4WmWyW405qyRWsxxcXHY2dlhjKFRo0YUKVIkU6/v1KlTzJw5k7Nnz3Lr1i2KFy9Ow4YNGTZsGK6urty8eZPg4OAc+TorlV0tXbqU3r1788svv/Dss89mes+FlBhjOHfunE3OndVExP/w4cNHatSoEfW4dZ08edKhdu3a1e/cubPf3j7npE/Hjx93qFKlSs3o6Oi9+fPnf+z6HvU1zTmvmFJKPQQReQJoCLyeeLsxJvHAtKUi8grQApiZtA5jzDxgHkBAQIAJDAxM9ZzBwcGkVSa70ZizRkoxb9myhYEDB7Jp0yYqV65Mo0aNMuX8xhh27NjB7NmzWbt2LXFxcdStW5dx48bRtGnThO5Q6YpZKZU+ERERHD9+HD8/P7p3746XlxfPPvusTWK5f/8+ixYtYvjw4URERFC7du0MTXq9vLxq3rhxI3++fPlMwYIF4xs1anR74cKFf7m5uT307GMzZswounTp0mJ79+612aR+Xl5eNWfNmhXapk2buwA+Pj7RERER+20VT3oljTu70C61Sqncqhuw0xhzOo1yBvjnt22lMtHdu3fp06cPL7zwAvb29ty/nzlzcERFRbFw4UJcXFx47rnn2Lx5MwMHDmTbtm389ttvPP/888kmm0qpx7N161Zq1qxJ8+bNiYiIwN7enhdeeCHL41i3bh1t2rTBy8uL/v37Y2dnR/PmzSldOl1TITyUVatWnYyIiNi/Z8+eYwcPHiz0zjvveD5sHTExMY8dR0bUoTKWJpxKqRxBROxFxBGwA+xExFFEUuul0Q1YkqSOwiLSLOFYEekCPAdsybTAlUrip59+ombNmixYsIDhw4ezb98+ateunaHn2L9/Px07dsTb25vevXvj4OBAt27dOH/+PNOmTdOWS6UySVhYGD179qRp06bky5ePlStXZvkyJxEREaxdu5YmTZrQoUMHNmzYQOPGjdm6dSu3bt1ixIgRFClSJNPOX65cuZjGjRvf/uOPPwoC3Lhxwy4oKKhM8eLFa5UoUaLWoEGDSsXGxgKW1swnn3yySq9evZ5wc3PzffHFFysMHz68zIEDB5ydnJz8XFxcfAHq1q1beerUqcUSzjFjxoyi/v7+D2aZFxH/CRMmFC9TpkyNsmXL1kzYvn79erfSpUvXdHd3r/3GG2+UjouLA+Do0aMF6tWrV6lw4cK+7u7utVu1alXu+vXrdgBt2rQpd+nSJYdOnTr5ODk5+Y0ePdrj+PHjDiLin5DMhoaG5m/cuHFFNzc3X29v7xpTpkx5ENuwYcNKtWjRonzbtm3LFipUyK9ixYrVf/755xT/JxAR/4kTJxYvU6ZMjUKFCvkNHjy41NGjRwv4+vpWcXZ29mvRokX5+/fvP/hlcOXKlW5VqlSp5uLi4uvn51flt99+K5hS3AnHzJ07t6inp2dNd3f32iNHjiyZsD0yMlJ69uz5RIkSJWqVKFGiVs+ePZ+IjIx8cK7333/fI+Hf7T//+U/Rh/jf4G804UyBiHwuIpdE5I6InBCR3raOSak8bjQQCYwCulr/Hi0i3tb1NL0TCorI01hmsk26HEp+LEurXAOuAwOBNsYYXYtTZbrw8HAGDBhAkyZNcHBwYMeOHUyaNAlHR8cMqd8Yw86dO3nllVfw9/dnzZo1+Pv788MPP3Djxg2WLl2Ks7NzhpwrKz3EGrzdRWSv9b59XkQmpfGjlFIZxhjD2rVrqVq1KsuWLWPkyJEcOnQo07rJJ+fMmTO0adMGFxcXgoKCOHXqFG+99RbHjx9n7dq1NG7cmHz5Mv+rf0hISP6tW7e61apVKwKgY8eOZe3t7Tl16tSR/fv3H9u2bZvbtGnTHiRohw4dKlS+fPmo69evH1i3bt3pTz755Kyvr294RETE/rt37x5I73k3bdpUePfu3X8cP378SOJte/fuPbZ79+4/tmzZUnj69OnFwPLvNXLkyMuXLl06ePTo0aMXL150GDFiRCmA9evXn/H09IxOaLEdN27claTnevnll8uXKlUq+tKlSwdXrVp1aty4cV4bNmxwSdi/devWwp06dQq7ffv2/mbNmt0aOHCgd9I6Evvvf//rtn///mPbt2//Y86cOSV79+5d5osvvjh99uzZQ8ePHy84f/78IgA7duxw6t+/f9nZs2efDQsLO9CzZ89r7dq1qxgZGSmpxf3rr786nzx58sh33313Ytq0aaX27dvnCPDOO+947t27t9D+/fuPHThw4Nj+/fsLjRo1yhPgyy+/dJ09e3bJ77777kRISMiRrVu3uqb33yIpTThTNgEoa4xxBVoB40Qk+y+QpFQuZYwZY4yRJI8xxpi/rOtp/pWo7C5jTCFjzN0kdVwzxtQxxrgYYwobY+oZY37I+qtRec3BgwepXbs2s2fPZsiQIRw4cID69etnSN1RUVGMGzcOZ2dnnnnmGTZv3kyvXr348ccf+e6771Ico5mDpGsNXsAJGAIUA54CmgBvZ25oSsGFCxdo27YtQUFBeHl58fvvvzNx4kQKFiyY6ecODw9nyJAhNGzYkAoVKrBx40ZKlSrF0qVLOXXqFJMnT6ZSpUqZHgdA586dK7q4uPg+99xzVerVq3d3/Pjxl86dO2f/888/u82bN+8vV1fXeC8vr9gBAwZc+fLLLx80sRYvXjz6vffeu5o/f36cnZ0fecmVUaNGXfbw8IhLXMfw4cMve3h4xPn4+ET37dv3ytq1a4sA1KhRI6pt27Z3ChYsaEqVKhU7ePDgK7t27XJJufb/ExISkn/fvn3OM2fOPO/k5GTq168f2blz5+vLli170ALo7+8f3rFjx9v29vb07NnzxvHjx1Nt5h41atSlIkWKxAcEBNz38fGJbNy48Z1q1apFFy1aNK5x48a39+/f7wQwZ86cYq+++uq1xo0b37O3t2fgwIE38ufPb3766adCqdU/fvz4i87Ozubpp5+OrFy5cuSePXsKAqxbt67Iu+++e8nLyyu2VKlSsaNHj7745ZdfFgVYvXp1kY4dO16vU6fOfVdX1/jx48dfTM/rkxz95S8FxpijiZ9aHxWAvbaJSCmlVE4UFhbGO++8g6enJ8HBwTz33HMZUu/evXtZvHgxa9as4dq1axQoUIBhw4bx0Ucf5ciWzJSkdw1eY8ycRE8viMgKIOual1SeFBwcTOvWrYmJieGTTz5hyJAhZMUspmfOnGHZsmXMnTuXy5cv4+LiwgcffECvXr144oknMv38yfniiy9Ckk5W8/vvvzvExsaKp6fng3EDxhgpWbJkdMJzT0/PDBl0Wa5cueik28qWLRud+O8rV67kB7hw4YL9G2+84b17927niIgIu/j4eFxdXePSc56//vrLwdXVNdbd3f3BhEhlypSJTkgKAYoXL/7gmpydneOjoqIkJiaGlGaKLVWqVGzC346OjvH5NmajAAAgAElEQVQeHh4Pji9YsGB8Qtznz593+Oqrr4ouWrSoRML+2NhYOX/+vENqMXt7e/+tvvDwcDuAa9euOVSoUOHBjLPly5ePvnr1an6Ay5cv53/yySfvJezz8fH5x+ubXppwpkJEZgM9gILAfuC7ZMrosgnZkMacNXJizEplBWMMv/zyCw0aNMDd3Z2PP/6Y3r17P3YiaIxh165d/Oc//2HtWkuP8ZdeeolBgwbRpEmTnN6SmdGeA46mtFPv39lTTovZyckJHx8fhg4dipeXFzt27Mi0c8XHx7Nnzx4mTZrEjRs3EBHq1KlDhw4daN26NdZuq5w69Y9Vvv4mK1/j8uXLxzg4OJibN28eSCnZEhGT5Pk/yjg5OcVFREQ86Jl5+fLlf1SW3HGhoaEOAQEB9wHOnj3rkJDIDR061EtEzKFDh46WLFkybvny5YXffvvtVLu9JvD29o6+c+eOfVhYWL6EpPOvv/5yyKjEOTVeXl4xgwYNuvTvf//7ckbUV7x48ehTp04VSHiNzpw541CiRIkYAA8Pj5hz5849SGRDQkJSTWpTowlnKowxb4rIQOBpIBD4x5ozumxC9qQxZ42cGLNSWWHTpk20bt2aDRs20KpVK3x9fR8r2bx//z6jRo1ixYoVXL9+HTc3N1q3bs3gwYOzdIxYTiEirwEBQIrzL+j9O3vKCTHPnj2b8+fP8/HHHwMwYsSITD3fnj17+Oijjzhy5AihoaE4OzvTqFEjFixYQPny5R+6vqx8jcuUKRPzzDPP3O7Tp88TU6ZMueDm5hb/559/FggNDc3fsmXL8OSO8fT0jLl8+bLD/fv3xdHR0QDUrFkzcuPGje6DBw++fvbs2fwrVqwoVqxYsTQTvClTppRs2LDhvdu3b+ebO3duif79+18BCA8Pt3N1dY0rVqxY3JkzZ/JPmzatZOLjihUrFhMSElIA+MfyIhUrVozx9fUNHzx4cOm5c+eeO3z4sOPKlSuLLViw4MwjvUgPoW/fvteCgoIqNmvW7G5gYOC98PDwfN99951Ls2bN7rq7u8enFndy2rZte3PixImeDRo0uCcijB8/3rN9+/Y3AIKCgm6++eab5Xr16nWjUqVK0aNHjy71qHHrGM40GGPijDE7sHTj6WfreJRSSmVPMTExHDp0CICWLVuyaNEimjdv/lh17t27l/feew9vb2+mT5/OnTt3mDx5MufPn2f9+vWabCZDRNoAE4Hmxpjrto5H5T5//PEHBw4cIGGm1cwQExPDDz/8QFBQEE899RTffPMNJUuWZPXq1dy4cYOffvrpkZJNW1izZk1odHS0VK1atUbhwoV9O3ToUOHChQvJN3cCL7744l0fH59IDw+P2u7u7rUB3n333Sv58+ePL1myZO1u3bqVa9++/c30nLtly5a3/Pz8qgUEBFRv2rTp7SFDhlwHGDt27MXDhw87ubq6+jVv3tznpZdeCkt83PDhwy9PmTLF08XFxfeDDz7wSFrvmjVrTp87d87B09OzdocOHSqMHDnyYtu2be883Cvz8J577rmITz/9NHTQoEHebm5uvhUqVKixdOnSB2NH04o7qYkTJ16qXbv2vdq1a1erVatWtZo1a0ZMnDjxEkBQUNCdPn36XGnWrFnlChUq1GjUqNEjX58Y88hjc/MUEVkA3DPGDE6pTEBAgNmzZ0+q9eSEX+6S0pizRm6NWUT2GmMCsiaizKPv7+wjO8b8yy+/0K9fPy5fvszp06dxdf37ZH4PE3NCt9kPPviArVu3AtCqVSvefPNNnn/++SyZZRKy3/tbRMYBpY0xPVIp8wKwHGhpjNmd3rr1/Z19ZMeY4+PjmTVrFv7+/tSvX5/o6Gjy58+PiGR4vDdu3GDy5MlMnjyZ2NhYihQpwiuvvEL79u0z7Aemx3lvHzx4MLR27dr6Q476h4MHDxarXbt22eT2aZfaZIhICaAx8A2WpReaAq8AnW0Zl1JKqezl+vXrjBgxgsWLF1OmTBkWL178j2QzvSIiIhg0aBDbtm17kLTWq1ePSZMm0aBBgwyOPOewLm1iT6I1eIFYY0xsknKNgRVA24dJNpVKTUhICD179uSXX36hf//+1K9fHweHRx7Klqz4+HimT5/Opk2b2LlzJ1FRURQrVox+/frx7rvvZtjSSUrZiiacyTNYus/OxdLt+CwwxBizwaZRKaWUyhbi4+NZvHgxI0aM4M6dO4waNYrRo0dTqFCqM9MnKyQkhCVLljB37lxu3LiBi4sLc+bMoWvXrrlqttnHMBr4MNHzrsBHIrIIOAZUsy6L9D7gBnyXaPKQX4wxj9evWeVJ8fHxzJw5k3feeQcHBweWLFlCt27dMvQcV69eZc2aNcydO5ejR49iZ2dHv379eOONN6hRo0aGnkspW9KEMxnGmGtAQ1vHoVR2dfjwYebMmUP79u1tHYpSWe7w4cP07duXnTt30qBBA+bMmUP16tUfqg5jDDt37qR///4cPHgQsHSbffXVV2nXrl2WdZvNCYwxY4AxKex2TlROB7SqDHHy5El69uzJjh07aNmyJZ999hleXl4ZUrcxhj179jBw4EB2796NMYY6deowceJEXn/9dYoUKZJ2JUrlMJpwKqUeyh9//EGTJk1wcHDQCUtUnjNr1iwGDx6Mu7s7ixcvpnv37g+1FMmtW7d466232LNnD4cOHcLZ2Rl/f38WLFiAr69vJkaulEpLXFwcM2bMeNCNdenSpbz66qsZstzQhQsXGDZsGH/88QeHDx/G0dGR6tWrM2PGDL2XqlxPE06lVLpFRETwr3/9i3z58vHTTz9x8eJFW4ekVJZIWLC7Tp06vPbaa0ycOJGiRYumfaDV1atXGT16NLNmzeLWrVuULFmSuXPn0rVr10fqhquUyngvv/wyX3/9NS+++CKfffYZpUo98ioQgKU18/fff2fJkiUsXbqUiIgISpcuzezZs+nSpcsjj/dWKqfRhFMplW5OTk5MmTKFatWqUalSJU04Va4XHx9Phw4dKF68OJ999hl169albt266T72p59+onfv3pw9exYRoVWrVrzwwgv06dNHu80qlQ3ExcUBYGdnR/fu3WnXrh1dunR5rFbNyMhIVq1axaBBgwgPD8fR0ZGOHTvSokULXn755QxpMVUqJ9GEUymVpgsXLnDs2DGef/55goKCbB2OUpkuLi4OOzs78uXLR40aNShcuHC6jz106BATJ05k9+7dnDp1igIFClC9enU2bdpEuXLlMjFqpdTDuHv3Ls2aNaNt27YMHz6c1q1bP1Z9GzZs4D//+Q8HDhzg1q1bFCtWjFatWjFz5kwdm6nyNP15VSmVprfffpvOnTtz9+5dW4eiVKb7+eefqV27Ntu2bQNg7NixDBs2LNVjYmJi+P7772nfvj2+vr6sXLmSEiVK8Pnnn3Pr1i0+/fRTTTaVymacnZ2pWrUqTzzxxCPXcffuXb744gsCAwNp06YNwcHBNG7cmODgYK5evcqKFSs02VR5niacSqk0zZkzh82bN+Pi4mLrUJTKNNeuXeO1116jYcOGhIeHY4xJ85gLFy7w5ptvUrBgQZo3b87PP/9M37592bx5Mzt37qRLly66hp5S2UjCxHcnT55ERFi4cCGdOnV6pHo6d+5M69at6dKlC+fOnePDDz/k+PHjrFu3joYNG2rX2Txqzpw5RZ555hmfjC6bk2nCqZRKVlhYGMOGDSMyMpLChQsTEBBg65CUyhTx8fEsWLCAKlWq8PnnnzNq1CiOHj1K48aNky0fExPDBx98wLPPPou3tzdz5szB3d2d8ePHc/78eWbPns0LL7yQxVehlEpNbGwskyZNws/PjwMHDhAaGvrQdYSFhdG9e3eqV69OtWrVWLNmDcWLF2fevHmcPHmSMWPGUKlSpYwPXj00Ly+vmo6Ojk86OTn5ubq6+gYGBlYMCQnJnxXn7tev381ff/31ZEaXzcl0DKdS6h/u3LnDCy+8wIEDB+jQoQP169e3dUhKZYqHWVPzxIkTrFixgoULF3LhwgUKFCjAiBEj6N27NxUqVMjiyJVS6XXs2DFee+01du/eTdu2bZk9ezYlS5ZM17Hx8fFs2bKFr7/+mi+++IJ79+7h5ubG5MmTefXVVzl27BiBgYGZewHqkaxatepkmzZt7kZEREj37t3L9O3b1/vHH388lZnnTJjRXP1dtmjhFJHhIrJLRMJE5JaI7BCRF5KU6SEiJplH00RlaovIHhG5JyLbRMQ7SR1TRGROVl2XUjnRvXv3aNmyJfv27WPNmjWabKpc6d69ewwfPhw/Pz9OnDjB4sWL2b59+z+SzZiYGNavX0+FChWoXLkyY8eOpWbNmsycOZObN28yYcIETTaVyqZiY2OZOHEifn5+nDp1ilWrVrFu3bp0JZs3btxg+vTpFC1alBYtWvD5558TFBTEpk2buHnzJm+99RYlSpTIgqtQj8vJycl06NAhLCQkpCDA9u3bnYoWLVo7JibmQZklS5YUrlKlSrXkjr9x44Zd27Zty7q7u9cuVapUzREjRngmzG48Y8aMok8++WSVXr16PeHm5ub71ltvlZoxY0ZRf3//ygnHf/XVV65ly5at4eLi4tu1a1fvOnXqVJ46dWqxhOMTlxUR/0mTJhUvU6ZMDVdXV99XX33VOz4+PnNemCyUXVo4GwOLgN+BSOB14BsRaWiM+TVRuTigdJJjbyb6eyHwK9AJ+DcwBXgZQESeAtoCtTPjApTKDSIjI2nVqhU7d+5k1apVjz1jn1LZVVxcHCtXrkxxTc2dO3cyYsQIQkJCuHLlCu7u7jRo0ID58+dTuXLlFGpVSmUXR48epUePHuzZs4d27doxe/ZsPDw8Uj0mPj6eefPmMWHCBC5dukRMTAyVKlWiVatWTJs2TSf/SUPPnj2fOHLkiFNmnqNGjRoRixYtOvcwx9y9ezff6tWr3f38/MIBGjZsGFG4cOHYr7/+2jUoKOgOwIoVK4p27NjxRnLH9+7d+4k7d+7YnT59+vDVq1ftmzVrVsnT0zNm6NCh1wEOHTpUqH379jevX79+ICoqShYtWvTgf5RLly7Zd+vWrcKsWbPOdO7c+dbEiRNLrFq1qtgrr7yS7LkANm/e7LZ3794/wsLC7OrWrVv1q6++utWhQ4c7D3PN2c0jJZwiEgycBq4CvQEHYA0w0BgT+bD1GWOaJ9n0tog0A9phSSATl72cSlVVga7GmBARWQBMtsbrgCUZ7WeM0Wk2lUpGVFQU7dq1Y9u2bSxbtoyXX37Z1iEplaHOnj3L5MmTmTJlCq6urhw9ehQ3N7cH+8PDw/n6669Zvnw5P/zwAwBPPfUU8+fPp3nz5tjbZ5ffaJVSqbl79y7PPvss9vb2rF69Os21L3ft2sWGDRtYvXo1oaGhiAgdO3bk3XffpWbNmlkYucpInTt3rmhnZ2ciIyPtihQpEvPNN988GCvZqVOnG59//nnRoKCgO1euXLHbsWOH64IFC/5KWkdsbCzffvttkV27dh1zd3ePd3d3j+7fv//llStXFk1IOIsXLx793nvvXQXInz//32abW7dunVvFihUju3fvfgvgvffeuzpr1qxUm9hHjRp1uVixYnHFihWLe/rpp+/u27fPKU8mnFYdgNVAA6AiloQuAhgkIu8C76Zx/MfGmI+T2yEi+QAX4HqSXXYichooCBwHJhtjvkm0/yDQXEROAM2B/dbtHwK7jTFb0ntxSuUlMTExBAUF8f333zN//ny6du1q65CUynCHDh1i8eLFvPrqq9StW/dBsnny5ElmzZrFjBkzMMbg7e3NmDFjaNmypU6WpVQOEhoaSpkyZXBxcWHFihUEBASk2O01JiaGTZs2MXv2bLZu3QpA06ZN+fjjj2nZsiWurq5ZGXqu8LAtj5ntiy++CGnTps3d2NhYVqxYUfj555+vfPjw4SPe3t6xvXv3vlGzZs0at2/fzrd06dIi/v7+4WXKlIlJWselS5fsY2JixMfHJzphW7ly5aKvXLnyYKCmp6fnP45LcPHixfylSpV6cGy+fPkoWbJkdErlAby8vB7UV7Bgwfjw8PBsMQTycTzOBdwE+hpj/jDGbAJGA2+ISCFgLuCbxmNuKnW/CxQGlifadhzojqXVsx1wANgkIr0SlekJvASEAt5YWkr9gK7ASBGZKiIh1vGdOo2YUlh+vevSpQsbN27k008/pXfv3rYOSakM88svvzBv3jwAXnrpJc6cOUPdunW5f/8+Q4cOpWLFilSqVIlPP/2UKlWq8O9//5vTp0/z4YcfarKpVA7yv//9j0qVKrFmzRoAWrRokWyyefz4cerXr4+bmxvt27fnzz//5KWXXuKXX37hhx9+4JVXXtFkM5ext7ene/fut/Lly2e2bt3qAlCuXLkYX1/f8M8//7zwqlWrinbu3DnZLq6enp6x9vb25uTJkw4J20JDQx08PDweJIUikuIaWp6enjGXLl16cGx8fDyXL192SKl8bvU4LZy7jTFxiZ7/iqVrbQVjzCH+PrYy3UTkTSwJZytjzPmE7caYXcCuREV3iUgRYCSW1lWMMX9iGQ+aUJc9sBkYgGUspz9QHUs34OXAU48So1K5SWRkJOfPn2fKlCn079/f1uEolSGuX7/OiBEjWLx4MZUqVaJHjx44ODhw8OBBvvvuO5YuXcrNmzdxcHBg3Lhx9OzZE09PT1uHrZR6SJGRkRQsWJA6deowatSoZJczunLlChMmTGDv3r3s2LEDEeGJJ55gzpw5NGvWDDs7OxtErrJKfHw8X3zxReG7d+/a16xZ88HQvy5dutyYNm1ayYsXLxbo2rXrreSOtbe3p0WLFmGjRo3yWrNmzZlr167Zz5o1y2PgwIFX0nPuDh063B41apT38uXLC3fq1OnWJ598Uvz69et5bnxGRjbRPugcLyLvikh4Go9/dLkVkbeBT7Akmz+m45w7gbKp7H8HOGZtgW0KrDHGRAFLgboioqvYqzwrPj6e+/fv4+Liwvbt2xk2bJitQ0qViAywzkIdJSJLUinXQ0TiknzeBCbaX9bayyFCRP5MPNO1yvni4+NZuHAhlStXZvny5YwcOZJff/2VVatW4ePjw/PPP8/MmTNp0qQJ69ev5969e7z33nuabCqVw8TExDB27Fh8fHy4du0adnZ2jB07luLFiwOWz4Ldu3fTt29fvL29mT59OhcuXGDSpElcvHiRs2fP0qJFC002c7FOnTr5ODk5+bm4uPh99NFHXjNnzjwTEBBwP2F/165db128eLHAv/71rzBXV9cUp4JdsGDBX05OTvHly5ev+dxzz1Vp3779zcGDBycd9pcsT0/P2KVLl556//33SxcpUsT32LFjBWvUqBFRoECBnD/17EN4nAy7jojYJWrlfBqIBk5h6S67Jo3j/9YCKiJjgaFAC2PM9nTG4Ack219cRKpjme3Wz7opH5DQ39oh0Tal8qRhw4Zx8OBBvv/+ewoUKGDrcNLjIjAOaIZlHHdqdhljnk1h30osvSVaWB9fioiPMeZahkWqbOLw4cP069ePX3/9lQYNGtC7d28mT57MrFmzCA8Pp1y5crRs2ZKpU6fq4uxK5WAHDx6kR48eHDhwgM6dO/8tabx58yZTp05l8uTJREVFUbBgQdq0aUNgYCBvvPEG+fLpV7+84MKFC4fTKuPi4hLv7u4e061btxRnjAUoXrx43IYNG84kt2/QoEE3Bg0adCO1bR06dLjToUOHI2CZId3T07OWt7d3THJljTF7E9e1bt260LSuIyd4nISzKDBLRKYD5YH/B8w3xtwD7vEQXWpF5D/AG8ArwHERSZi9KdIYc9taZgywGzgBFMAyaVFvYFAy9dlhWWZlmDEm4R/xZ6CniPyAZSzowYS6lcqL6tSpQ4ECBXBwyBlDCYwxXwGISAD/XB4pXaxjt58E/mWdUXudiAwB2pP6uHKVjd27d4+PPvqIadOm4erqSvv27Tl37hzdu3cHICAggE8++YSGDRumOlOlUip7i46OZsKECYwbN46iRYvy9ddf06ZNG+Li4pg0aRJbtmzh119/JSoqChcXF/r168eYMWP+Nhu1UgmWLFlSWER46aWXMnUFi3Xr1rk2bNjwXqFCheI//PDDksYYGjVqFJ6Z58xuHifh/BK4C+zA0mK4FhjxiHUNtv736yTblwI9rH+7ArOAkljW6vwTCDLGrEumvqHAOWPMl4m2zcayBudOLEu6dHvEWJXK0U6ePImPjw9dunShS5cutg4ns/iJyHUsP3wtByYYY2KxjOE+nWR5pIPW7SqHGjJkCAsWLKBKlSocP36cdevWUa1aNaZNm0ZQUBClSpWydYhKqcd04MABevTowcGDB+nSpQvTp08nLCyM8ePHs2DBAkJDQ7G3t6dfv3706tWL2rV12XWVsrp161YOCQlxnD9//pnM7lb966+/Ovfq1at8TEyMVKhQIXL16tWnnJ2dU5xoKDd6nIQz3hgzHBj+uEEYY9L8ydkYMwxI1yAzY8zkZLbdx9KyqVSeNX78eD766CN27dqFv7+/rcPJLD8DNYCzWBLJ1UAsMAFwBpL2bLgNeCVXkYj0AfoAeHh4EBwcnOqJw8PD0yyT3eTUmFetWkVkZCQbN25kyxbLilenT5/Gz8+P5557jlatWiEinDhxghMnTtg44pz7Oue0mFXuEx0dzfjx4/n4448pVqwYX375JXZ2djz99NOcPGlZVrFx48YMGDCAXr16UbhwYRtHrHKC3bt3H8+qc02dOvXi1KlTL2bV+bKjPDdLklJ51dSpUxk9ejRdu3bF19fX1uFkGmPM6URPD1vHhw/HknCGY+ktkZgrlt4aydU1D5gHEBAQYAIDA1M9d3BwMGmVyW5yWszx8fG8/fbbzJ07l+joaOLi4nB0dGTs2LEMGDAAd3d3W4eYrJz2OkPOjFnlPocOHWLcuHE0bdqU69ev069fP65du0aRIkV45plnmDNnDjVr1rR1mEqpVOjIaaXygFmzZvHWW2/x8ssvs3jx4rw2K5/h/2bRPgqUTzJDdW3rdpWNXblyhWHDhlG5cmWmTZtGTEwMHTp0YNeuXdy7d4/3338/2yab6vGkd4Zqa9mhInJZRG6LyCIRyREzoqm/i4qK4ttvv+XatWv8/vvv+Pn58d///pd9+/ZRsWJFNm3axJUrV9ixY4cmm0rlAI/UwmmMCczgOJRSmWThwoUMGDCAVq1asWLFCuztc2bHBuu6uvaAHWAnIo5ArHVsZuJyzYF9xpgrIlIFeB/LGHOMMSdE5ADwoYiMBpoDtbBMGqSymejoaDZu3Ej//v25evUqAOXKlaN58+bMmzeP0qUfae4olfOka4ZqEWkGjMKyHvdFLPNCfGTdpnIIYwzDhg1j9uzZD7b5+PgwceJEWrZsSY0aNWwYnVLqUeTMb55KqXRZsWIFr7/+Os2aNWPNmjXkz58/7YOyr9HAh4medwU+EpFFwDGgmjHmL6AJsEREnIErwOfAx4mO6wQsAcKAv4AOuiRK9mGMYeXKlSxatIj9+/dz86ZlwnMR4fXXX2fatGns3r1bk8085CFmqO4OLDTGHLWW/3/ACjThzBEuXLhA//79+fnnnwkLC8PBwYFatWrRs2dP+vbtqzNMK5WDacKpVC715Zdf0q1bNwIDA/nqq69yylqbKTLGjAHGpLDbOVG5t4G3U6knFAjMuMhURjhy5AhbtmxhyZIlHDlyBOBBF9nGjRsze/ZsKleubMsQVfZXHdiQ6PlBwENEiiZaIu0BnRTM9sLDw1m9ejX79u3j2LFjABQoUIBBgwbx/PPP4+xs+Wjfvj29y7Nnvez+GicnJ8ascjZNOJXKha5cuUK3bt14+umn2bhxI05OTrYOSal/iIqKYtOmTUyYMIF9+/YBljUzn3rqKXbv3o2joyOrVq0iKChIWzdUeiSdhTrhbxfgHwmnTgpmG7GxsezYsYMVK1awfPlyoqKiAMsPTGPHjqVGjRrZLubUZMfXOC05MebsZMaMGUWXLl1abO/evRk60+2oUaNKnjlzpsDq1avPZmS9AN9//71z3759y4aGhh7J6LrTQycNUioX8vDwYOPGjXz77bcPfiFWKjswxrB161Zq1KhBsWLFePnll7l48SJPP/00S5cuJTQ0lD179jB48GD+/PNPOnbsqMmmSq+ks1An/J2pi7qr9Pnzzz/p168fjo6ONGrUiGXLlhEfH4+dnR3vv/8+Fy9eZMCAAbYOU+USXl5eNR0dHZ90cnLyc3V19Q0MDKwYEhKSbcYVffPNNy4eHh61Em+bOHHi5YxKNkXE/8iRIw+6tr3wwgvhtko2QVs4lcpVgoODCQsLo23btjRt2tTW4Sj1wJ49e1i2bBlbt2590HXO39+fjz/+mHr16uHq6kpsbCzbt29n0KBBumi7ehRHscw6vcb6vDZwJbnutCprXL58mQEDBrB//35Onz6NiODu7k6BAgW4dOkSTZs2Zfbs2fj4+Ng6VJULrVq16mSbNm3uRkRESPfu3cv07dvX+8cffzxl67jyIm3hVCqXMMYwYcIExo4dS2xsbNoHKJXJIiIiWL16NS1atKBOnTrMnDmTwoULM2/ePK5fv86ePXs4deoUVatW5datW9jb27Nw4UJNNtXfiIi9dVbqBzNUW2etTmoZ0EtEqomIO5aJxpZkYagKuHnz5oMZZUuXLs26desICwtj6tSpDB8+nFu3bj2YHOy///2vJpsq0zk5OZkOHTqEhYSEFATYvn27U9GiRWvHxMQ8KLNkyZLCVapUqZbc8ZcvX7Zr3LhxRWdnZ7+aNWtWPXXq1N8mxdi/f79j/fr1fdzc3HzLli1bY8GCBQ/W6Fq9erVbhQoVqhcqVMivRIkStT744AOPO3fu5OvQoYPPtWvX8js5Ofk5OTn5hYaG5h82bFip1q1blwM4fvy4g4j4z5w5s6inp2dNd3f32iNHjiyZUO+2bducfH19q7i4uPgWL+pNhZcAACAASURBVF68Vrdu3bzv378vAAEBAZUB6tSpU83Jyclv/vz57klbVPft2+dYt27dyi4uLr4VK1asvmLFCreEfe3bty/76quvegcGBlYsVOj/s3ff8TWe/x/HX1cSkshAhFRiRM0aTZCmtVcpapaixCiKGk21ZomtI1rU/Jrla/vaTWu0Ijal1IgRI8OIiBhNJDKv3x/BL0gIcnIyPs/HI4/HOee+7vt+n8NJzufc17Cq+vbbb1fw9/d/rYlApOAUIodQSrF+/Xq2bduWbZc+Edmf1ppDhw5Rr149bGxs6NSpE6dPn6Z79+7s2LGD/fv307t3b6ysrAB47733aNGihZFTiyxuNBBD8myzHg9vj1ZKlVBKRSmlSgBorbcB3sAuIPjhz9jUDykyUmJiIr6+vvTs2RMHBwdGjhzJyZMnGTp0KPv37yc8PJzBgwfj7u7O559/ztmzZ+nUqZN0l8+B3N3dy8+YMaMQQGxsrHJ3dy8/Z84cO4DIyEgTd3f38gsWLCgIEBERYeru7l5+6dKlBQBCQ0PN3N3dy69cuTI/QEhIiJm7u3v5devW2QK8apfYyMhIkzVr1hSsWrVqFEC9evWiCxQokLBx48bHXfBXrFhRqGPHjqn2hujdu3dJCwuLpOvXr59YvHhx4KpVq+wfbfv3339NmjVrVq5jx463b9269c/y5csvDx06tMTRo0ctAAYOHFhy9uzZwffv3z/u7+/v36RJk0hbW9ukdevWXShcuHB8dHT08ejo6OPOzs7xqZ17//791hcuXDj9+++/B0ybNs3x2LFjFgBmZmb89NNPV27fvv3Pvn37zu3bt8/G29u7MMDRo0fPAxw5cuRMdHT08c8+++xOymPGxsaqNm3alGnYsOG98PDwE1OnTg3p06fPmydOnHhcVG7ZssVu7Nix1+/evXvc2dk5dvjw4U6v8to/IgWnENncP//8Q7t27YiKisLa2hoHBwdjRxK50IkTJ2jSpAlly5alRo0aHDp0CGdnZ9avX09QUBBLliyhcePGBAQE0KRJE/r16wdA1apVmTdvHgUKFDDyMxBZldZ6nNZaPfUzTmsdorW2frgc0qO2U7XWDlprW631p1rrWGNmz+lOnz5Nhw4dMDc3p1GjRqxbt47mzZszZcoUAgMDGTNmDBMmTMDb2xuAdu3aMWvWLHm/i0zRuXPnMjY2Nq4FCxasum/fPtuRI0eGPdrWqVOniOXLlxcCCAsLM923b59tz549bz99jISEBLZt21Zg8uTJ121tbZPeeeedBx06dHhcmK5Zsya/k5NTrKenZ0SePHmoXbt2dLNmze6uWrWqIICZmZk+deqUxe3bt00KFy6cWLt27eiXeQ6TJ0++bm1trWvUqBFTvnz5mKNHj1oC1KlTJ7pRo0b38+TJQ/ny5eN69OgRvnfvXpv0HHPXrl1W0dHRppMnT75hYWGhW7VqFdmwYcO7S5cuLfSoTdOmTe80aNAgOk+ePHTp0uW2v79/mmsgp4dcBkmFUsocmAO8D9gBF4FvtNZbjRpMiKf4+/vTuHFjLC0tuX37tkwQJDLVrVu3WL9+PRs2bGDHjh0AVKhQgUWLFtG+fXtsbf9//pbo6Gi+++47vL29sbS0pE2bNsaKLYR4DRcuXGDo0KGcO3eO8+fPY2pqip2dHSNHjqRfv35YWlqitUYphZmZGUWLFsXOzs7YsUUm+euvvx7P3Gpubq5T3rexsUlKeb9QoUKJKe8XLVo0IeX9EiVKPHG/TJkyqV4FTMvKlSsvtmnTJjIhIYEVK1YUaNy4cflTp06dLlGiRELv3r0jqlSpUvnevXsmS5cutatevXpUyZIlnzn+9evXzRITE1Xp0qXjHj1WsmTJ2IMHD1oDBAcH5z158qSVjY2N66PtiYmJqm3bthEAq1evvjRhwoSiEydOLFa+fPmY77777ur7779/P73PoUSJEo8zWVpaJkVFRZkCnDx50tzT07P4qVOnrB48eGCSmJhIxYoV01XMXrlyJc8bb7wRZ2pq+vix4sWLx12/fv3xFWQHB4fH57WyskqKjo425TXIFc7UmQFXgHpAfsALWKuUcjZiJiGeEBAQQKNGjciTJw++vr6UKFEi086ttWb79u0EBgZm2jlF1qC15sCBA/Tq1YsiRYrQr18/zp8/z5gxY9i3bx9nz56lZ8+eTxSbPj4+VKpUiUmTJtGhQwfOnTvHgAEDjPgshBAvIzw8nBkzZtC0aVPKly/P5s2buX//PjNmzCA0NJSbN28yePBgLC0t8fX1xcXFhfPnk+uEX375hb59+xr5GYjczMzMjO7du981MTHRO3futAEoVapUvKura9Ty5csLrF69ulDnzp1T7U7r6OiYYGpqqi9dupT30WMhISGPu54WL148/p133omMjIz859FPdHT08RUrVoRAcvfdnTt3XgoPDz/RokWLOx4eHqUBlFL6dZ5T3759S5YtW/bBhQsXTkVFRR0fOXLktfTuW7x48fgbN27kTUxMfPzYlStX8jo6Or5UQf8ypOBMhdb6/sPuOkFa6ySttQ8QCFQ3djYhAC5fvkzDhg1JSkpi586dlClTJtPOvXPnTpo0aULTpk353//+l2nnFcYVEhJCixYtKFiwILVq1WL16tW4u7sza9YsLl++zPjx46lVq9YT+wQHB9OmTRtatmxJvnz52LVrF8uWLeONN95I4yxCiKwiMTGRHTt20LVrVxwcHPD09OT8+fOMGjWKXbt2ceXKFQYNGkThwoWB5PWfPTw8aNSoEdHR0dy5c+cFZxAicyQlJbF8+fICkZGRZlWqVIl59HiXLl0ifv755zcCAgIsPTw87qa2r5mZGR988MFdLy8vx8jISJO///7bYu3atY+7nnbo0OFuUFCQxezZs+1iY2NVbGys2r17d75jx45ZPHjwQM2dO9cuIiLC1NzcXNva2iaZmppqSC5k7927ZxYREfFKVw6joqJMbW1tE/Pnz590/Phxi8WLFxdJub1QoUIJAQEBqU70U79+/fuWlpaJXl5eb8TGxiofHx8bX1/fAl27dn2mS3FGkS616aCUcgDKkTzl+tPb+gB9IHntQz8/v+ceKyoq6oVtshrJnDnSm/nmzZt4enoSHR3N1KlTCQsLIyws7IX7va4bN24wceJEzpw5g7W1NQMGDKBRo0bZ7nUW6RcWFsakSZM4c+YMu3btQmuNnZ0dixYt4uOPP8bGJu3hIt7e3owbNw6lFD/88ANffvklefPmTbO9ECJrOHHiBEOHDmXPnj3ExsZSoEABGjZsSOvWrRkwYAAmJk9eq0hMTGT+/PmMHDmSmJgYvLy8GDlyJJaWrzXkS4jX1qlTp7ImJiZaKYWjo2PczJkzA93c3B482u7h4XF36NChJZs0aXLH1tY2Ka3jLFiwIKRLly7ORYsWdSlVqtSDTp063dq3b58NQMGCBZO2bt0a4OnpWXz06NHFtdaqQoUK0VOnTr0CsHLlykLDhw8vkZSUpJydnR8sXrw4EKBq1aoPWrZsebt06dJVEhMTOXXq1DM1xvN4e3tf6d+/f8k5c+a88dZbb0W3adPmdsoxnMOGDbvet29f5+7du5tMmzYt+I033ni8fIGFhYXeuHHjxc8//7zkzJkz3yhSpEj83LlzA6tWrfog9bO9Pik4X0AplQdYASzVWp97ervWej4wH8DNzU3Xr1//ucfz8/PjRW2yGsmcOdKTOTQ0lM8++4zo6Gh8fX2pXt3wF92DgoLw9vZm8eLFQPKsoqtXr6ZkyZLZ8nUWz5eYmMjevXtZtmwZK1asIDY2FkdHR8aOHUvXrl15880303WcK1eu0KxZM6ZNm5ap3b2FEC/vxIkTeHl5cfnyZfz9/TE1NcXe3p4ff/yR9u3bY2Fhkep+x48fp1+/fvz11180bNiQOXPmUL58+UxOL8Szrl27dupFbWxsbJIKFiwY361bt+eu1evo6Jiwa9eui2ltd3FxifXz80t1+969ey+ktd///ve/oJT3p06dev3R7fLly8dprf9OuT3leNZmzZpFBQYGplmkDhs2LHzYsGHhKR8LCws7+ei2m5vbgyNHjpx/dk9Yv379E7latGgRmXLfVyEF53MopUyAZUAcMNDIcUQuFx4eTqNGjQgNDeWPP/4weLEZGxvLzz//zIgRI9Ba06NHDyZMmEDx4sUNel5hHMHBwUyfPp1Zs2aRkJCAtbU17dq1o1atWvTt25eUkwukJiIigkGDBtG/f39q167NtGnTZHkeIbKwGzdusHHjRtavX8/OnTsBeOutt5gzZw4dOnSgUKFCae7777//MmbMGGbOnIm9vT3Lly+nc+fOssyJyFaWLFlSQClFy5YtI42dJaeTTwNpUMm/NRcBDkBzrbXBBtIKkR6Wlpa8+eabzJ07lxo1ahjsPAkJCUyYMIFly5YRFBRElSpVGDVqFB07djTYOYVxREZG4u3tzYABAzhz5gyQPDSgV69efPPNN4/XykwPCwsL/v77b86ePUvt2rWl2BQiC0pISOCPP/5g4cKFbNiwAYDSpUvj5eVFrVq1+OCDD9J1nDt37rBo0SL69u3L5MmTKViw4It3EiILcXd3L3/x4kWLBQsWBL7oC1Xx+uQTQdrmAm8B72utY17UWAhDuXfvHiYmJtjY2ODj42PQc+3atYsePXoQEhJCuXLl2LFjB40bNzboOUXmunDhAr/88gvnzp1j+/btREdHY2VlxcSJE+natSslS5ZM97EOHjzIjz/+yIoVK7CysuLUqVMyTlOILEZrzbFjxxgwYACnTp0iOjoaOzs73N3d6du3L59++mm6rkxevnyZJUuWMH78eEqWLMnly5cfTxgkRHaTsnuqMDwpOFOhlCoJ9AVigRspfhH31VqvMFowketorWnVqhVKKXbt2mWw7kqbNm3ip59+Yt++fRQrVox+/foxY8YM8uTJ8+KdRZZ37NgxDh06xMaNG9m5cydaa4oWLUqPHj0oXrw4w4YNe2YikOe5desWI0aMYNGiRTg5OXHp0iUqVaokxaYQWchff/3FhAkTuHHjBmfPnsXExIRixYoxY8YMmjVr9tLv199//51p06bRtWtXypYtK8WmECLdpOBMhdY6GJCBCMLolFJ8+eWXjxfRzmjXrl3Dy8uLX375BTMzM7y9vRk0aFCaE0SI7EFrTUBAAJs2bWLevHmP10stV64cPXv2pF69enTu3BlTU1P8/PzSXWwmJSWxePFihg8fzr///svQoUMZM2YM1tbWhnw6Qoh0unz5Mtu3b2fVqlXs3bsXgHfffZf58+fTrl077OzsXup4fn5+3L17lzZt2vD555/z0Ucf4ejoaIjoQogcTApOIbKgBw8e8Ndff1G3bl3atm2b4ce/du0avXr1Yvfu3SQlJdGpUyfGjx9PuXLlMvxcInM86jY3b948lixZQnx88rDzypUr07RpU7y8vKhRo8Yrf3Fx/Phx+vfvz6FDh6hbty5z5syhUqVKGfkUhBCvICoqii1btjBjxgwOHz4MJE/+M3z4cIoUKcJXX3310se8efMmQ4YMYdmyZbzzzju0bt0aU1NTKTaFEK9ECk4hspi4uDg+/vhjtm/fzoULF15qTF16jv1ovbSoqCjef/995s+fT6lSpTLsHCLzJCQk8OeffzJ06FBCQ0OJiIhAKYWdnR1du3Zl8ODBr70kyb179/Dy8mL27NnY29vz3//+Fw8PD5mNUggjiouLY9OmTYwYMYJr164RFxeHk5MT1atXZ9y4cXz44YcopV56neSkpCQWLlzI8OHDuX//PqNGjeKbb76R97sQ4rVIwSlEFpKQkMAnn3yCj48Pc+fOzbBiMykpieHDh7NixQpCQ0OpV68en332GV26dMmQ44vMExMTw8SJEzlw4ACnT59+XGRWqFCBKVOm0LJlS+zt7TPsfH5+fsyaNYv+/fszadIkChQokGHHFkKkX1JSEgsWLGDz5s0cPnyY27dvY2pqyrvvvsv3339PrVq1Xmos9tNOnDhBv379OHToEPXr12fOnDm89dZbGfgMhBC5lRScQmQRiYmJdOvWjQ0bNjBt2jT69euXIcfdt28fQ4YM4fDhw1hbW/Pbb7/RrFkz+cY6G7l48SILFy7k4sWLbNu2jfv372NqakrHjh1p27YtH3zwATY2Nhl2vrNnz3L8+HE6d+5Mq1atOHv2rCzmLoQRaK3x9fV9PC7z6tWrKKX45JNP6Ny5M++//z7m5uavdY7IyEjGjRvHzz//jJ2dnfRiEOIp7dq1c3ZycoqbMWPGdWNnya6k4BQiC0hKSuLHH39k27ZtfPfdd3z55ZevfcwdO3bQp08fgoODKVq0KFOmTGHgwIEyIVA2cfToUY4ePcrGjRv5888/SUpKokiRInh4eNCgQQNatWqFpaWlQc49YcIE9uzZw0cffYSFhYUUm0JkskuXLrFq1Sp+/vlnbt26hampKc2aNWPw4MF07NgRJyenDDtX586d8fHxoW/fvnz77bcvPbGQEFmRk5NTlYiIiDwmJibazMxMV6tWLWrhwoXBZcqUiX/efjNmzCi0dOlS+7///luWTclAUnAKYWSRkZF4enqybds2xo4dy4gRI17reGFhYYwfP5558+aRlJTEZ599xrRp07CyssqgxMIQtNZcuHCBzZs3M3/+fC5evAgkL8r+6aefUrt2bbp27YohFqjWWrNp0ybeeustKlSowPTp01FKyZcTQmSisLAwpk6dyty5c4mMjATA1dWVBg0aMGXKlAwdzx8YGEj+/Pmxs7NjwoQJfPPNN9SoUSPDji9EVrB69eoLbdq0iYyOjlbdu3cv2a9fvxJ//vnnJWPnyo1evbO/EOK1bdu2jfLly/PLL7/QtWtXxo4d+8rHioyMpFGjRpQoUYIFCxbw+eefExAQwPz583NEsamUGqiUOqqUilVKLXlOu+5Kqb+VUv8qpa4qpbyVUmYptvsppR4opaIe/hjtW0ytNcePH2fAgAGPryQOGzYMc3NzPvjgA/z8/Lhw4QILFy6kR48eBik2r1+/TosWLfjoo4+YNm0aAA4ODhQpUiTDzyWEeNKtW7fo378/NWrUwNHREW9vb+Li4ujXrx/BwcEcP36ctWvXZmixeefOHVxcXBgzZgwAVatWlWJT5Gj58uXT7du3v3Px4kVLgIiICNO2bds6FyxY0MXR0bHKsGHDiiYmJnLs2DGLoUOHlvznn3+s8+XLV9XGxsb10THu3LljVr9+/TJWVlZV33777Qr+/v6v15c9l5ErnEIYwaN1NQsWLEjx4sXZuHEjMTExrzRmJiEhgcWLFzN27Fhu3LhByZIl2bFjR05c4uQ6MAn4AHheX9J8wJfAYaAwsAUYAnyfos1ArfVCA+V8rsTERHbv3o2np+czM8t26tSJIUOG4OzsbPAcsbGxeHt7M2nSJPLmzcvUqVMZOHCgwc8rRG4XFRXF6tWr2b59O1u2bCEuLg5bW1tGjhzJJ598YrDlhs6fP0/58uUpWLAgs2bNokGDBgY5j8jd3N3dXzgGo2nTpncnTJgQ9qi9h4fHrS+++CIiNDTUrHXr1qWft+9ff/310l8SR0ZGmqxZs6Zg1apVowB69+5d/N9//zW9fPnyqZs3b5p98MEH5YoWLRo/ePDgW1OmTAlOrUvtli1b7DZt2hRQu3bt6Hbt2pUaPny4k4+Pz+WXzZJbScEpRCbSWjN48GDi4+OZPXs27777LocOHXql6eu11kydOpXRo0fz4MEDatasyZo1a6hbt65hwhuZ1noDgFLKDSj2nHZzU9y9ppRaARj1k9WDBw+YPHkyu3fv5ty5c4SHh6OUomzZsvzwww+0bNkyU68o+vn50a9fP86fP0+9evVYsWJFho4JE0I8KTExET8/P1auXMnSpUtJTEykSJEi9O3bl5o1a9K+fXvMzAzzkSw8PJxhw4axZMkSdu/eTd26denWrZtBziVEVtK5c+cypqamOiYmxtTOzi7ex8fnQkJCAr/99pvdwYMHzxQsWDCpYMGCcQMGDLixatWqQoMHD76V1rGaNm16p0GDBtEAXbp0uT1ixIg0P4eIZ0nBKUQmeHRFUylFnjx50Fo/8djLOnjwICNGjGDPnj3kzZuXKVOm8PXXX8usgqmrC/g/9dh3SqnvgfPAKK21X2o7KqX6AH0guZvpi74UiIqKetzm2rVr/Prrr4SFhXH48OHHV7AbNGhA7dq1cXd3f9zV+cyZM5w5c+aVn2B63b17l7lz57Jjxw4cHR354YcfqFixIhcuXODChQsGP39GSfk6ZxeSOffRWnP06FG8vLzw9fUlPj4eGxsb6tSpQ/369Rk1apTBikxInozut99+o127dvz777+MHDkSNzc3g51PCHj5K5Ap2xctWjThVa5gpmXlypUX27RpE5mQkMCKFSsKNG7cuPyRI0fOxMfHq7Jly8Y9aleqVKm4sLCwPM87loODw+PJhqysrJKio6MzfoxLDiYFpxAGdujQIQYNGsSUKVOoX78+3t7er1wYXrx4kcaNGxMUFESRIkWYM2cOvXr1Im/evBmcOmdQSn0KuAG9Uzw8HDgDxAGdgF+VUq5a62cmEtBazwfmA7i5uen69es/93z/+c9/MDExYdOmTWzfvp2kpCTs7e3x8PCgUaNGtGzZknz58mXMk3tJcXFxlC1bltDQUL755htGjx6NpaUlfn5+vOh5ZTWSOXNklcxKKTtgEdAEuAWM1FqvTKWdOfAz0BbIA+wH+mmtr2ViXPbv38/YsWO5fPkygYGBmJqaUqhQIb777js++eQTg80undLx48cZNGgQ+/fvp27dusyZM8dgXXWFyOrMzMzo3r373S+//FLv3r3byszMTF+4cCFv9erVHwAEBQXlfVRQyhf3hiEFpxAGEhYWxogRI1iyZAlFixYlOjoaeLVfZiEhIUydOpU5c+agtaZu3br8+uuv2NraZnTsHEMp1YbkcZvva60fd5PRWh9O0WypUuoToDkw81XPdffuXWrUqMG5c+cAKFWqFD179qRmzZp07drVoFcxXiQwMBBnZ2fy5s3LTz/9RMWKFalYsaLR8gjxCmaT/AWRA+AK/KaUOqG1frrngidQA3gbuAcsIPl9/ZGhA546dYoNGzawZcsWjh07BkCVKlVYuHAh7dq1o0CBAoaOAMDNmzcZNWoUixYtwt7enuHDh/Pdd9/Jh2iRqyUlJbFy5coCkZGRZlWrVo1p3rz5nREjRjitXbs2MDw83Gz27NkOgwYNCgMoWrRo/I0bN/I+ePBAWVhYaGNnzymk4BQigz0anzl27FhiYmIYNmwYo0ePxsbG5qWPFRMTg6enJwsWLEApRe/evRk3bhyOjo4GSJ5zKKWakvxh80Ot9akXNNfAa30ay58/P87OzlhbW/P999/TsGHDLPEB7/Dhw9SpU4elS5fyySef0L59e2NHEuKlKKWsgHZAZa11FLBPKbUF6Ao8vYZUKWC71jrs4b6rgamGynb79m3Wr1/Pf//7X/bt2weAm5sbP/30E7Vq1eLdd9811KlTtWfPHlq2bEl0dDSDBw/Gy8uLf/75J0v8LhLCGDp16lTWxMREK6VwdHSMmzlzZqCbm9uDhQsXhvTu3bvEm2++WcXc3Fx7eHiEe3p63gJo0aJF5KxZs2IcHBxcTExM9J07d04Y+3nkBFJwCpGBdu3axaBBg/D396dJkybMmDGD8uVfOGHbM+Li4vjPf/7DlClTuHr1Kk5OTixYsIBmzZoZIHX28HBpEzPAFDBVSlkACVrrhKfaNQRWAG211n89ta0A8C6wG0gAOpI8xvPL18zG1q1bs0wXxLCwMBwcHHBzc2PUqFG8//77xo4kxKsqByRqrQNSPHYCqJdK20XAz0opR+Au0AXYmpFhoqOj+fXXXxk5ciRBQUForSlbtiwtW7akb9++fPjhhxl5uhfSWnP79m0KFSqEq6srLVu2xMvL65X+7giRk1y7di3NL5sLFy6cuHnz5sDUtllYWGg/P7+LKR9bv359UMr7LVq0iAwLCzuZIUFzCSk4hcgA4eHhDBw4kLVr1+Ls7MzGjRtp3br1S3+zrLVmx44dtG/fnqioKNzc3Fi+fDn16qX22SrXGQ2kXKjUAxivlFpM8pjMilrrEMALyA/8nuL136u1bkbyuK5JQAUgETgHtNFaG20tzox048YNBg8ezB9//MHZs2cpXLjwa63tKkQWYE1y99iU7gGpdRkJAEKAayS/v08Baa71k95JwRISEtiyZQs+Pj7cuHGDmJgYLC0tKVOmDF9//TXlypV7/Ls+sydZmjZtGidOnGDRokWYmprSu3dvQkNDCQ0NBbLnxE/ZLXN2ywvZM7PI3qTgFCIDmJubc/ToUcaPH8/QoUNfaVKIPXv2MGnSJHbu3EnhwoXp2bMn06ZNw8TExACJsx+t9ThgXBqbrVO0S3MJFK11OPBOhgbLApKSkpg3bx4jR44kJiaGUaNGyfhekVNEAU//Z7YFIlNpOxewAAoB94FhJF/hTLVva3onBTt48CAzZyYP8e7cuTO9e/embt26mJoaZ5LKO3fuYG5uTr58+YiOjiYgIIC6deuSJ8+zk2xmlV4XLyO7Zc5ueSF7ZhbZmxScQrwiX19fpk+fzvr167G1teXs2bMvPVtsfHw8O3bswNvbmz179mBtbc306dPp168f5ubmBkoucpITJ07Qt29fDh8+TMOGDZk7dy7lypUzdiwhMkoAYKaUKqu1frR2jwvPLnX06PFRWuvbAEqpmcAEpZR9yonDXtZ7773H2rVryZs3L61bt37Vw7y2xMREFixYwOjRoxk4cCDjxo2jefPmNG/e3GiZhBAiPeTSiRCv6P79+wQEBHD16lWAdBebSUlJ/PHHH1SqVImCBQvSokULTp48iYuLC6dPn8bT01OKTfFCUVFRDBkyhOrVq3P58mWWLVvGn3/+KcWmyFG01veBDSQXjlZKqVpA20AdVwAAIABJREFUa2BZKs2PAN2UUvmVUnmA/sD11yk2IXmM9scff0z+/Plf5zCvZdeuXVSrVo3PP/+cypUr06ZNG6NlEUKIlyUFpxDpFB0djZeXF5MnTwagRYsWnDp1ilKlSr1wX601q1atok2bNpQsWZImTZpw7tw53nzzTX799VfCwsKYPn06JUuWNPTTEDmAn58flSpV4qeffqJnz56cO3cODw8PmY1S5FT9AUvgJrAK+Fxr7a+UqqOUikrRbgjwALgAhJO83FHbzA6bkQIDA2nfvj0NGzbk3r17rFu3jl27duHq6mrsaCL3SkxKSpI/NuIJD/9PJKW1XbrUCvECWmvWr1/PV199xZUrV+jVqxdaa5RSqY6ZSWn//v38+eefrFy5koCA5EkWW7RowZQpU2jRogXW1tbP3V+I1FhZWVGgQAFWrFhB7dq1jR1HCIN62EX2mUt6Wuu9PDl+O4LkmWmzvaioKL7//nt+/PFHTE1NmThxIl9//fUrzQ8gRAbbFxwcXNvR0TEqb9688fJFp0hKSlLh4eH5gdNptZGCU4jnOHPmDIMGDcLX1xcXFxdWrFhBnTp1nrvP1atXWbNmDdOnT3/c3bZBgwZ0796dDh06UKZMmcyILnKY2bNnc+nSJaZOnco777zD8ePHZUIpIXKgmzdvUrVqVa5fv06XLl34/vvvKVasmLFjCQFAQkLCZ3fv3v08MjKyh9baDuktKZKvbJ5OSEjonVYDKTiFSMW9e/cYP348M2fOxNramtmzZ9OnTx/MzFJ/y0RERDBr1iymTZvGvXvJM/hXqlSJatWq8f333/PWW29lZnyRAwUFBXH+/HkSEhIwMzOTYlOIHCY0NJSiRYtSpEgRPDw8aN26NTVr1jR2LCGeUL169Tjg54c/QqSLFJxCpKC1ZtmyZQwbNoybN2/Su3dvJk+eTOHChZ9pe+fOHby8vDh+/Dh//fUXCQkJ5M2bl27dujF69GjKli1rhGcgcop///0XLy8vWrduTcOGDfn2228xMzOTcZpC5EALFy5k4MCB+Pv7U7p0aX744QdjRxJCiAwjBacQKSilWLduHc7Ozvj4+ODm5vbE9qioKFatWoWvry+bN28mJiaGfPnyMXjwYD755BNcXV2lIBCvRWvNhg0b+OKLLwgNDeWNN96gYcOGLxwvLITIXh48eEBERAROTk40b96cIUOGUKRIEWPHEkKIDCcFp8j1bt++zejRo/nqq68oU6YMy5Ytw8bG5nGXxcTERPbs2cOqVav45ZdfSEhIoFChQnTv3p369evTrl27NLvaCvEygoKCGDhwIL/99hsuLi5s3LgRd3d3Y8cSQmQgrTWbNm3i66+/plixYuzevRtHR0cmTZpk7GhCCGEQ8ilZ5HqxsbGsXr2a6tWrU6ZMGfLnz4/WmqNHjzJmzBj++OMPEhISsLKyom7dutSpU4dvvvkm3etuCvEi8fHxTJs2jfHjx6OU4qeffuKLL76QLzKEyGFOnjzJl19+ya5du6hcuTJjx46VXjFCiBxPPs2IXGn//v0sW7aMuXPnUrRoUQIDA8mfPz979uzBy8uLkJAQgoKCMDMzw97enrFjx9KtWzfy5ctn7Ogihzl48CB9+/bl1KlTtG7dmhkzZlCiRAljxxJCZKBbt24xZswY5s2bR4ECBV44EZ0QQuQk8ptO5CqhoaEMHz6cZcuWUaxYMa5cuUJoaCibN29m27ZtHD9+HIDKlSuzcOFCPvroIwoWLGjk1CKniouLo2PHjmit2bhxI23aPLPUoBAiG4uPj2fu3LmMHTuWyMhIBgwYwLhx47CzszN2NCGEyDRScIpcIT4+nhkzZjB+/HhiY2Px9PTE2dmZTz75hAMHDgDg7u7O9OnTadiwIVWqVDFyYpFTaa3ZvHkzzZs3J2/evGzZsoXSpUtjY2Nj7GhCiAyktaZevXocPHiQxo0bM23aNCpVqmTsWEIIkemk4EyDUmog0AOoAqzSWvcwaiDxyv7880+++OILzp49i6urKzdu3GDGjBlorXnrrbfo2LEjffr0oWHDhsaOKnKBgwcP0rZtW+bPn89nn32Gq6ursSMJITJQYGAgJUuWxMTEhH79+jFixAhatmwpYzWFELmWFJxpuw5MAj4ALI2cRbyC4OBgBg8ezMaNG8mTJw958uThn3/+oUCBArz33nvMmTMHFxcX+RAgDC42NpZDhw5Rr149atasiY+PD02bNjV2LCFEBjt27BjvvfceixYtomvXrnTr1s3YkYQQwuhMjB0gq9Jab9BabwIijJ1FvJy4uDhWr15N+/bt2bhxI5DcpbZTp07s27ePiIgIDhw4IGtmikyxe/duXF1dadKkCdeuXQPgww8/xNTU1MjJhBAZITExEX9/fwBcXV3x8vKicePGRk4lhBBZh1zhFNnezZs3Wbx4MdevX2fLli3cuHGD2NhYzM3NadeuHd27d+f999/H0lIuVIvMc+vWLYYOHcqSJUtwdnZm06ZNODk5GTuWECID7du3D09PTy5dusSlS5coVKgQXl5exo4lhBBZihScr0kp1QfoA+Dg4ICfn99z20dFRb2wTVaTlTInJSURGBjIunXruHPnDsHBwdy4cePxdgsLC5ycnKhSpQo9e/bE1tYWgMOHDxsrcrplpdc5vbJjZkPTWrN06VKGDBnCvXv3GDFiBF5eXrKkjhA5SEhICMOHD2f16tUUK1aMuXPnysyzQgiRBik4X5PWej4wH8DNzU3Xr1//ue39/Px4UZusxpiZ7969y+HDh/nxxx+5fv06165d4969e0+0MTc3Z/z48XTu3JlixYqhlJLXOZNkx8yGFBISQoMGDdi9ezc1a9Zk3rx5VK5c2dixhBAZJDo6milTpvDDDz+gtWbMmDEMGzYMKysrY0cTQogsSwpOkWUkJSVx9uxZDh48yJw5cwgKCuLu3btorYHkwtLc3BwAU1NTGjduTKdOnWjdujUFChQwZnQhWL16Nb169cLa2pr58+fTq1cvTExkmLwQOYHWGl9fX7p168aVK1fo0KED3t7elCxZ0tjRhBAiy5OCMw1KKTOSXx9TwFQpZQEkaK0TjJss57hy5Qr+/v4cPHiQ5cuXExQURFJSEpDcNdbOzo6PP/6YrVu3cuXKFeLj46lTpw4dO3akbdu2FCpUyMjPQIj/V7NmTRo3bsySJUsoUqSIseMIITLQV199xfTp06latSrLly+nbt26xo4khBDZhhScaRsNjE1x3wMYD4wzSppsLi4uDh8fH8LDwzl8+PDj2wAmJiY4OTlRoUIFPD09OXPmDA0bNqRVq1Zcv36dgIAARo4cSbt27eSDfC72MmvjKqUGA8NJXtJoPfC51jr24TZn4BfgXSAEGKi1/vN185UoUYJhw4bJ/1EhcqDu3btjZmbG999/LzNMCyHES5KCMw1a63FIcfnKgoODWbJkCVFRUfzzzz/s27ePBw8eAFCoUCEqVKiAi4sLX3zxBc7Ozly7do2mTZuitaZMmTLkyZOHVq1a4ejoyK5du4z8bEQWka61cZVSHwAjgIYP99lI8pdFIx42WQUcBJo//FmnlCqrtQ43XHQhRHbm6urK3bt3pdgUQohXIAWneG2JiYkcOnSIOXPmEB0dzZkzZwgICABAKYWLiwsff/wxlpaWfPbZZ1SvXp2rV6+ydu1aJk6cyJEjRyhcuDChoaGYmppy6tQpmdFTPENrvQFAKeUGFHtO0+7AIq21/8P2E4EVwAilVDmgGtBEax0DrFdKfQm0A/5jyPxCCCGEELmRFJzipd26dYs9e/YwdepUoqKiuHTpElFRUQBYWVnRqFEjunTpgo2NDZ07d8bBwQGA69evs27dOjw9PTlw4AAA1atX54cffqBDhw6PvzmWYlO8pkrA5hT3TwAOSqlCD7dd1lpHPrW9UibmE0K8BKWUHbAIaALcAkZqrVem0bYaMJ3kL5buA99qrX/OrKxCCCGeJQWneK74+HguXrzImTNnmD59Onfu3OHWrVuPtxctWpTu3bvz3nvvUbx4cerUqfPMzJyHDh1ixIgR7NmzB601b7/9NpMnT6ZDhw6UKVMms5+SyPmsgZRr5zy6bZPKtkfbnVI7kKyzmzVJ5syRhTLPBuIAB8AV+E0pdeJRL4ZHlFL2wDZgMLAOyMvze0MIIYTIBFJwCmJjYwkMDGTz5s1cvHgRS0tLLl68yJ49e4iOjn68LEnevHlxdnZmyJAh1KhRg7fffjvV5UhiYmJYvnw5Li4uuLu7Y2FhQXh4OOPGjaNDhw5UqFAhs5+iyF2iANsU9x/djkxl26PtkaRC1tnNmiRz5sgKmZVSViR3ea+stY4C9imltgBd+f9x2Y98BWzXWq94eD8WOJtpYYUQQqRKCs5cIioqil27dnHkyBGsra25ePEiO3bsICwsjNjY2MdFJYCNjQ1lypShdOnS2NvbU6NGDXr16oWzszNKqVSPf/fuXYKCgnB1dcXExIQhQ4bQp08f3N3dcXFxwd/fP9X9hDAAf8AFWPvwvgsQprWOUEr5A28qpWxSdKt1AVLtnieEMLpyQKLWOiDFYyeAeqm0fQ84pZQ6AJQBDgMDtNYhho8phBAiLVJw5hBaa4KCgvDz88Pc3PxxQXn69GksLCwICwt7on3hwoWxsrKiePHidOnShdKlS2NtbU358uWpUKHCE4Wln58fpUqVeuac//77L1u2bGHNmjVs376d0qVLc+bMGczNzTl9+jTFiiX3ZEqrSBXiZbzE2rj/BZYopVYAoSQvcbQEQGsdoJT6BxirlBoNNAPeJvkKihAi60mrG7xNKm2LkTx2szFwCvAmeVbqWqkdWLrMZ03ZLXN2ywvZM7PI3qTgzEYSEhI4fvw4kZGRBAUF4evry+7du7G3tycwMJB79578m1ywYEGUUjRt2pQKFSpgZWWFvb09zZs3J3/+/K+UISoqCh8fH9asWcPWrVuJjY2lePHifPHFF3Ts2PFxu+LFi7/WcxUiFamujauUWgycASpqrUO01tuUUt7ALv5/Hc6U+3UiuQC9Q/I6nO1lSRQhsqyX6QYfA2zUWh8BUEqNB24ppfJrrZ8uWqXLfBaV3TJnt7yQPTOL7E0KziwmJiaGkJAQgoODOXDgAL/++iuFChXi+vXrBAQEEB8f/7itiYkJpqamlCpVCg8PD2xtbTEzM6Nt27ZUqFABS8s0lyp8aUeOHGHKlCn4+PgQExND0aJF6du3Lx07duS99957ZqIgITLaC9bGtX6q7VRgahrHCQLqZ1wyIYQBBQBmD9fKvfDwMReSu84/7SSgU9x/dFu62QghhBFJwWkEMTExBAYGcvz4cdauXYuVlRW3b9/G39+fq1evPtPe2dkZFxcXatSoQVRUFG3btsXNzY0SJUpgZmaYf8I7d+4wZ84cmjdvDkBERAS7d+/m008/pWPHjtSqVUsWwBZCCGFQWuv7SqkNwASlVG+SZ6ltDdRMpfkvJK+tO4PkgtQL2Ke1vptpgYUQQjxDCs5M9NFHH7F582aSkpKeeNzS0pKKFStSvXp1SpQoQbNmzahbty6lSpXC0dHRYIWd1ppr165x7Nixxz+NGjXC09MTpRReXl4UKFCASpUq0bhxY65du2awAlcIIYRIQ39gMXATiAA+11r7K6XqAFu11tYAWmtfpdQ3wG9APmAf0NlImYUQQjwk1UMmcnJy4o033qBx48Y0adKEUqVKUbhwYUqXLp0pE+sEBgY+Liz//vtvjh07Rnh48tA1ExMTKlSoQIMGDQAoUKAA9+7dw8bGBj8/P7maKYQQwii01reBNqk8vpdnu9PPBeZmUjQhhBDpIAVnJpo5cybt2rXLtIHav/32G7dv36Zr164A1KlT5/FVykqVKtGiRQuqVatGtWrVcHFxwcrK6on9bWxSmwRQCCGEEEIIIdJHCs5sLiEhgXPnzj2+chkSEsKGDRsAWLJkCf7+/o8LzoULF2Jvb0/lypWxsLAwZmwhhBBCCCFELiAFZzYSGxuLv7//E2MuT5w4wYMHDwDIly8frq6uxMTEYGlpydy5c59Y/qRp06bGii6EEEIIIYTIhaTgzMKuXLmCj48PHh4e2NjY8O233zJhwgQAbG1tqVq1Kp9//jnVq1enWrVqlCtX7omxlvb29saKLoQQQgghhBBScGYFkZGRnDhx4vFkPv369aNGjRqcPn2a/v37U6lSJerWrUuHDh2oVKkS1apV480335S1L4UQQgghhBBZmhScmSwyMhJfX98nusUGBASgdfL61A4ODnz44YfUqFGDevXqERQURIkSJQCoVKkSlSpVMmZ8IYQQQgghhEg3KTgz0dWrV2nVqtXj+yVKlKBatWp06dLl8WyxRYsWfbw9X758lCxZ0hhRhRBCCCGEEOK1ScGZiZycnOjXrx9t27alatWqFC5c2NiRhBBCCCGEEMJgpODMREopOnbsmGnrcAohhBBCCCGEMcmsM0IIIYQQQgghDEIKTiGEEEIIIYQQBiEFpxBCCCGEEEIIg5CCUwghhBBCCCGEQUjBKYQQQgghhBDCIKTgFEIIIYQQQghhEFJwCiGEEEIIIYQwCCk4hRBCCCGEEEIYhBScQgghhBBCCCEMQgpOIYQQQgghhBAGIQWnEEIIIYQQQgiDkIJTCCGEEEIIIYRBKK21sTPkGEqpcCD4Bc3sgVuZECcjSebMkVMzl9RaF86MMIYk7+8sRTJnDnl/Pymn/htmNdktc3bLC7novS2yBik4M5lS6qjW2s3YOV6GZM4ckjn7y46vh2TOHJI5+8uOr4dkNrzslheyZ2aRvUmXWiGEEEIIIYQQBiEFpxBCCCGEEEIIg5CCM/PNN3aAVyCZM4dkzv6y4+shmTOHZM7+suPrIZkNL7vlheyZWWRjMoZTCCGEEEIIIYRByBVOIYQQQgghhBAGIQWnEEIIIYQQQgiDkIJTCCGEEEIIIYRBSMEphBBCCCGEEMIgpOAUQgghhBBCCGEQUnAKIYQQQgghhDAIKTiFEEIIIYQQQhiEFJxCCCGEEEIIIQxCCk4hhBBCCCGEEAYhBacQQgghhBBCCIOQglMIIYQQQgghhEFIwSmEEEIIIYQQwiCk4BRCCCGEEEIIYRBScAohhBAiy1JKDVRKHVVKxSqllryg7ZtKKR+lVKRS6pZSyjuTYgohhEiDmbED5CT29vba2dn5uW3u37+PlZVV5gTKIJI5c+TUzH///fctrXXhTIpkMPL+zjokc+bIQu/v68Ak4APAMq1GSqm8wB/AbKAjkAiUS88J5P2ddWS3zNktL2Sp97bIJaTgzEDOzs4cPXr0uW38/PyoX79+5gTKIJI5c+TUzEqp4MxJY1jy/s46JHPmyCrvb631hofncgOKPadpD+C61npqisdOpucc8v7OOrJb5uyWF7LOe1vkHtKlVgghhBA5wXtAkFJq68PutH5KqSrGDiWEELmdXOEUQgghRE5QDGgAtAJ2Ap7AZqVUBa113NONlVJ9gD4ADg4O+Pn5PffgUVFRL2yT1Uhmw8tueSF7ZhbZmxScQgghhMgJYoB9WuutAEqpH4HRwFvAiacba63nA/MB3Nzc9Iu6GObUrpNZTXbLnN3yQvbMLLI36VIrhBBCiJzgJKCNHUIIIcSTpOAUQgghRJallDJTSlkApoCpUspCKZVaD63lwHtKqfeVUqbAl8At4GwmxhVCCPEUKTiFEEIIkZWNJrm77AjA4+Ht0UqpEkqpKKVUCQCt9fmH2/8D3AFaA61SG78phBAi88gYTiGEEEJkWVrrccC4NDZbP9V2A7DBwJGEEEK8BLnCKYQQQgghhBDCIKTgFEIIIYQQQghhEFJwCiGEEEIIIYQwCCk4hRBCCCGEEEIYhBScQgghhBBCCCEMQgpOIYQQQgghhBAGIQWnEEIIIYQQQgiDkIJTCCGEEADs27ePK1euGDuGEMJA9u/fz40bN4wdQ+QyUnAKIYQQghUrVtCoUSO++uorY0cRQhjA0qVLadiwIXPnzjV2FJHLSMEphBBC5GJaayZNmoSHhwc1atRg3rx5xo4khMhAiYmJDBs2jB49elCnTh2GDBli7Egil5GCUwghhMil4uLi6NmzJ15eXnTt2pXt27djZ2dn7FhCiAwSGRlJ27ZtmTJlCv3792fr1q3Y2NgYO5bIZaTgfAGlVFml1AOl1HJjZxFCCCEyyt27d2nWrBlLlixh7NixLF26FHNzc2PHEkJkkODgYGrVqsXvv//OrFmzmD17Nnny5DF2LJELScH5YrOBI8YOIURuppQyV0otUkoFK6UilVLHlVLNntN+sFLqhlLqnlJqsVLKPMU2Z6XULqVUtFLqnFLq/cx5FkJkHUFBQdSsWZO9e/eyZMkSxo0bh1LK2LGEEBnkwIEDuLu7ExISwtatWxkwYICxI4lcTArO51BKdQLuAjuNnUWIXM4MuALUA/IDXsBapZTz0w2VUh8AI4BGgDPwJjA+RZNVwHGgEDAKWKeUKmy46EJkLVeuXOHdd98lNDSU7du30717d2NHEkJkoNu3b9O0aVNsbW05dOgQjRs3NnYkkctJwZkGpZQtMAH42thZhMjttNb3tdbjtNZBWuskrbUPEAhUT6V5d2CR1tpfa30HmAj0AFBKlQOqAWO11jFa6/XAKaBdpjwRIbKAYsWK8emnn3LgwAEaNGhg7DhCiAyitQbAzs6ONWvWcPjwYSpUqGDkVEIkXzUQqZtI8ofWK8/rZqSU6gP0AXBwcMDPz++5B42Kinphm6xGMmcOyZx+SikHoBzgn8rmSsDmFPdPAA5KqUIPt13WWkc+tb1SGueR93cWJJlfntaaTZs24e7ujpOTE02bNiUsLIywsLA09zF2ZiFE+t2/fx8PDw86depEx44dadYszVEnQmQ6KThToZRyBd4Hqr6ordZ6PjAfwM3NTdevX/+57f38/HhRm6xGMmcOyZw+Sqk8wApgqdb6XCpNrIF7Ke4/um2TyrZH251SO5e8v7Mmyfzybt68yccff4ypqSldunRJ1z7GziyESD8zMzPu3r1LRESEsaMI8QwpOFNXn+SxXyEPr25aA6ZKqYpa62pGzCVErqaUMgGWAXHAwDSaRQG2Ke4/uh2ZyrZH2yMRIgeKiYnBwsKCIkWKcPjwYUqVKmXsSEKIDPTXX39RunRpChUqxJ9//ompqamxIwnxDBnDmbr5QGnA9eHPf4DfgA+MGUqI3Ewlf/uzCHAA2mmt49No6g+4pLjvAoRprSMebntTKWXz1PbUuuYKka1dv36dWrVqMXHiRABKly6NiYn82Rcip1ixYgV169ZlyJAhAFJsiixL/vKkQmsdrbW+8eiH5KsiD7TW4cbOJkQuNhd4C2iptY55Trv/Ar2UUhWVUgWB0cASAK11APAPMFYpZaGUagu8Daw3aHIhMtnJkyd59913uXDhAu+8846x4wghMlBSUhKjRo3Cw8ODGjVq8OOPPxo7khDPJV1q00FrPc7YGYTIzZRSJYG+QCxwI8VEXn2BvcAZoKLWOkRrvU0p5Q3sAixJLibHpjhcJ5IL0DtACNBevkwSOcn27dv5+OOPsbW1Ze/evbi6uho7khAig9y/f5+uXbuyceNGPvvsM2bNmkXevHmNHUuI55KCUwiR5Wmtg4HnrUpv/VT7qcDUNI4VRPI4bSFynPnz59O/f38qV66Mj48PxYoVM3ak16aUGkjy0kZVgFVa6x7p2McXaADk0VonGDSgEJnkypUrtGrVipMnTzJ9+nS++OILnreSghBZhRScQgghRDaXlJTEyJEj8fb2plmzZqxZswYbG5sX75g9XAcmkTyPguWLGiuluiCfb0QOc/jwYVq3bk1MTAy//fYbTZs2NXYkIdJNfiELIV6a1ppffvkFc3NzY0cRIteLiYmhe/fu/O9//6Nfv37MnDkTM7Oc8+dda70BQCnlBjz3kq1SKj/JXei7AQcNn04Iwzty5Aj16tXDyckJX19fKlasaOxIQryUnPMXSQiRaZRS9OzZk507dxo7ihC53u3btzlw4AA//vgjX331VW7vYvctyROM3XhRQ6VUH6APgIODA35+fs9tHxUV9cI2WY1kNrzMyJuYmEi7du3o0KEDN2/e5ObNm691vOz2GovsTwpOIcRL+f3333FwcKB69eoyBbsQRhQcHEyxYsVwcnLi7NmzGdaFNj4+npkzZ1K2bNkMOV5meXgFtBbgyQuuhAJoreeTvAwabm5uun79+s9t7+fnx4vaZDWS2fAMlff+/fsMGzYMLy8v3njjDRo1apRhx85ur7HI/mRZFCFEummtGTFiBF9++SVaa2PHESLXunbtGlWrVmXChAkAGVZsbtiwATs7O77++mt8fX0z5JiZQSllAswBPGWSIJETXL58mWXLlsmVSJEjyBVOIUS6KaXYs2cPd+7cye3d9oQwKkdHR4YPH06HDh0y5HjR0dGMGzeOn376CaUU48ePp27duhly7ExiC7gBax7+bnrU/eKqUupjrfVeoyUT4iWEhIRQokQJqlSpwuXLl7G3tzd2JCFem1zhFEKky9WrV4mPj6dAgQKUKlXK2HGEyHW01nz33XecPHkSpRTDhw/PkPfizJkzKVasGFOmTKFnz57cuHGDMWPGZEDijKGUMlP/x959h0V1vG0c/44Ye4/YooklKvYeS4yKmp/Ghg3RYMcuGAt2sXcsUVFjV+zYe08wGsCSWKMCscaCiiKCFCnz/oH6EkVBWFjA53Nde8XdM2fOzUbcfc6ZM6NUBqKKSBOlVAal1NsnzP2BAkDFV48mr16vApxKsrBCJMCmTZsoWbIka9euBZBiU6QaUnAKIWIVHh5Os2bNaNmypbGjCPFJevnyJd26dWPUqFGsW7fOIH0+e/aMXr16MWDAAAICAti1axfLli1Ljl9yxwDBwAig46s/j1FKfamUClRKfamj+Lx+AI9f7ftQa/3SSLmFiJPIyEjGjRtHhw4dqFZlaBMvAAAgAElEQVStWqIueaK1xsvLK9H6FyImUnAKIWK1YMECLly4gI2NjbGjCPHJ8fPzo1GjRqxZs4aJEycyY8aMBPfp4OBA8eLFWbFiBQMHDuTevXu0aNHCAGkNT2s9Xmut3nqM11rf0Vpn0VrfiWGfW6/ayf2cIlkLCgrCysqKiRMn0q1bN44cOYKpqWmiHOv69evUr1+f3r174+bmlijHECImcg+nEOKD7t69y9ixY2nSpAmtWrUydhwhPik3btygadOmbyYQ6dixY4L68/HxoW/fvuzcuZPs2bNz6tQpqlataqC0QoiPce/ePSwsLPjrr79wdHRkyJAhiTI/wr179xgyZAg7d+7ks88+o3Xr1vJ7L5KUFJxCiA8aOHAg4eHhODk5yURBQiQhDw8PWrRoQXh4OEeOHEnQJD6RkZEMHTqUlStXEhwczIABA5gyZQpZsmQxYGIhRFydPXuWFi1aEBAQwO7du2nWrJnBj6G1Zvv27XTt2pXAwEDatm3LvHnz8PLyIl26dAY/nhDvI0NqhRDvtX//frZt24aDg4NMFCREEtq2bRvm5uZkzZoVd3f3BBWbN27coHLlysyZM4d8+fJx4cIF5s2bJ8WmEEayZ88evvvuO9KlS4ebm1uiFJsuLi589913tG3blkKFCrFs2TK2bNlCgQIFDH4sIWIjBacQIkZBQUHY2tpSqlQp7O3tjR1HiE/G33//jaWlJZUqVcLDw4OSJUvGq5+wsDBGjx5NuXLluH79Ot27d+fixYvx7k8IYRjFihXD3Nyc06dPU65cOYP2/fz5c+zs7LCysuLMmTMsWLCAixcv0qNHD4MeR4iPIUNqhRAxmjp1Kjdv3uS3336ToTdCJKEyZcqwceNGWrRoQcaMGePVx6VLl6hfvz6+vr40btyYpUuXUqhQIQMnFULEVXBwMOvXr8fGxobSpUuzf/9+g/b/+gTT2rVrefjwIT/88ANOTk4ULVrUoMcRIj7kCqcQ4h03b95k5syZdOrUiXr16hk7jhCpXkBAAG3atOHUqaglI62srOJVbAYEBDBq1CgqV65MeHg4ffv2Zd++fVJsCmFkq1evplevXpw9e9bgfZ8/fx4zMzMcHR3JlSsXHh4e7N+/X4pNkWxIwSmEeEfhwoVZtWoVs2bNMnYUIT4JISEh/P3333h6esa7j6NHj2Jqasq0adPo0KED3t7eLFq0iDRp5KNeCGN5+TJqGdjevXvj7u5OtWrVDNb3lStX6NChA1WqVMHf35/+/ftz8eJFvvnmG4MdQwhDkE8hIcR/hIeHo5TC2tqaPHnyGDuOEKnatWvXCAsLw9TUlAsXLtC5c+eP7uP1PVv/+9//SJcuHePHj8fZ2ZncuXMnQmIhRFxt2bKFkiVLcvv2bdKkSUP16tUN0m9ERARLly6lfPnybNq0if79++Pt7Y2TkxMmJiYGOYYQhiQFpxDiDT8/P8zMzHBxcTF2FCFSvf3791O1alXGjBkDQPr06T+6j3nz5pE7d26cnJywtbXl3r17jBs3ztBRhRAfQWvNpEmTaNeuHQUKFCBDhgwG63vZsmV888039O7dmzJlyrBnzx7mz59Pzpw5DXYMIQxNJg0SQrwRHBxMqVKlKF68uLGjCJGqLVq0CDs7OypUqMBPP/300fv7+voycOBA1q9fT/r06Vm1ahVdu3Y1fFAhxEcJDg6me/fubNq0iU6dOrF06VKDFJwPHz6kS5cuHDp0iBw5crBhwwbat28v62OLFEEKTiHEGwUKFGDPnj3GjiFEqhUZGcmwYcOYPXs2zZo1Y+PGjR+1HqbWmsGDB7NkyRLCw8MZO3YsI0eONOgVFCFE/Dx48AALCwvOnj3L9OnTGTZsWIILwsDAQCZPnszixYsJCgqiUaNGrF27FlNTUwOlFiLxScEphCA8PJyhQ4cyYMAAihQpYuw4QqRKQUFBdOrUie3bt2Nra8vPP//8Ufdb/fvvv/Tp04f9+/eTKVMm/vjjD4NOQCKEiL+//vqLFi1a8OzZM7Zv307Lli0T3KerqysWFhY8f/6cBg0a4OTkhJmZmQHSCpG05B5OIQQLFy7k559/TpTp2g1FKWWrlDqrlApVSq3+QLtflFKB0R6hSqmAaNtdlVIh0bbHf1pQIeLo6dOnmJubs2PHDubOncv8+fPjXGyGh4fToUMHSpYsiaurK46Ojjx9+lSKTSGSiW3btlG7dm3SpEnDH3/8keBi88SJE1haWmJubk6mTJkYP348R44ckWJTpFhyhVOIT9z9+/dxcHCgcePGtG3b1thxPuQ+MBloBLx3gUKtdR+gz+vnr4rTyLea2WqtlydCRiHe8c8//9C/f3/8/f0/+srHtWvX6Nq1K6dOnaJgwYL8/vvvMgpBiGTG29ubChUqsGPHDvLlyxfvfkJDQxk3bhwzZswgbdq0jBs3juHDh8drTV4hkhO5winEJ27QoEGEhYXh5OSUrCcf0Fpv11rvBJ7EdR+lVGagDbAm0YIJEYvcuXOTP39+jh8/HudiMzg4GEtLS8qXL4+XlxezZs3i9u3bUmwKkUy8fPmSy5cvAzB8+HCOHz8e72JTa83y5cspV64cM2bMoEKFCpw8eZLx48dLsSlSBSk4hfiEHTp0CBcXF0aNGkWxYsWMHScxtAEeA7+/9fo0pZSvUuoPpVS9pI8lUjutNStWrCA4OJgcOXIwZ86cOA+BPXv2LGZmZmzdupUaNWpw9epVhgwZQpo08pEtRHJhZ2dHnTp1eP78OUop0qVLF69+bty4QalSpejZsyfh4eEcPHiQ8+fPG2zNTiGSAxlSK8QnKjg4mP79+1OiRAmGDRtm7DiJpQvgrLXW0V4bDlwBXgLtgT1KqYpa6+tv76yU6gX0AsibNy+urq4fPFhgYGCsbZIbyZw4rl27Rr9+/bh8+TIWFhZxyuzv78/ixYs5cuQIOXPmpFu3bnTu3JmrV69y9erVpAkeTUp4n4UwlpEjR2Jubk62bNnitb+Pjw+zZ89mwYIFpEmTBgsLCzZs2ECmTJkMnFQI45OCU4hP1PTp07l+/TrHjh2L14LzyZ1SqhBQF+gZ/XWt9aloT9copToATYAFb/ehtV4KLAWoWrWqrlev3geP6erqSmxtkhvJbFgvXrwgc+bM1KtXj/Lly1OzZk2UUrFmPnbsGNbW1rx48YLu3bszZ84csmfPnnTBY5Cc32chjOHUqVM4OzuzYMECChcuTOHChT/6pIzWGhcXF6ytrYmIiKBDhw44OjryxRdfJE5oIZIBGZ8jxCfIy8uL6dOnY21tTf369Y0dJ7F0Bty01jdiaaeB5Hvzqkgxjh49StGiRfn1118BqFWrVqz3Rd++fZsuXbrQsGFDcubMyaxZs1ixYoXRi00hxH9t376devXqceDAAR49ehSvPlxdXWnUqBHt27cnb968/PLLL2zYsEGKTZHqScEpxCfo7Nmz5MqVi9mzZxs7SpwppdIqpTIAJoCJUiqDUupDozQ6A6vf6iOHUqrR632VUtZAHeBQogUXqZ7Wmjlz5tCoUSNMTU358ssv47Tf6tWrKVq0KGvXrmX48OF4eXkxZMiQRE4rhPgYr3+/27ZtS4UKFfDw8PjoyYECAgJo2bIl5ubmuLm5MX/+fG7fvk3v3r0TKbUQyYsUnO+hlFqnlHqglHqulPJSSvUwdiYhDOXHH3/kxo0b5M2b19hRPsYYIBgYAXR89ecxSqkvX62n+eZbvlKqJlAQ2PJWH58RtbTKY8AXsANaaq1lLU4RL0FBQXTq1IkhQ4bQsmVLPDw8+Prrrz+4z507d2jdujXdunUjT548rFmzhunTp8tslO/xEWvwdlFK/fnqc/uuUmpmLCelhPig8PBwbG1tGTJkCK1bt+a3334jT548H7X/4sWLKVmyJLt27aJ8+fL89ddf2NnZkTat/NUUnw4pON9vGlBYa50NaAFMVkpVMXImIRLk2bNn7N27FyDFfbnVWo/XWqu3HuO11ne01lm01neitXXXWmfWWge81cdjrXU1rXVWrXUOrXUNrfWRpP9pRGpw+/ZtateuzYYNG5g8eTJbt24lS5Ys722vtWbw4MEULlyY/fv3M336dO7cuUOnTp2SMHWK9HoN3pWxtMsEDARyA9WBBoB94kYTqVVgYCAWFhYsWrQIe3t7XFxcPupz88KFC+TJk4d+/fpRoEABPDw8uHDhAiVKlEjE1EIkT3J65T201n9Hf/rqUQz40ziJhEi4BQsWMH78eK5du0bx4sWNHUeIFMvV1RVLS0tevnzJnj17aNq06Qfb37t3jwYNGvDbb7+RM2dOtmzZQoMGDZIobcqmtd4OoJSqStTIhfe1Wxzt6T2l1HrAPJHjiVTo3r17NGvWjIsXL7J48WL69OkT531v3LjB7Nmz+eWXX8iUKRM9evTgl19+wcTEJBETC5G8ScH5AUqpRUBXICNwDtgfQxtZNiEZkswxq1mzJjNnzuTevXvcu3cvwf2lxPdZiIRavnw5ffr0oXjx4uzcuZOSJUu+t21YWBhWVlbs2rWLLFmysGTJEnr06CFraiaNOsDf79son9/JU3LIPGbMGDw9PZk6dSpmZmYfzPM6b2RkJM7OzqxZswalFBYWFnTr1o1s2bJx4sSJpAsfB8nhPRafFik4P0Br3U8pZQfUBOoBoTG0kWUTkiHJ/F8REREEBgaSPXt2GjZsaLB+U+L7LERCffXVV7Ro0YLVq1d/cA2+Y8eOMWDAAK5cuULOnDn5888/KVKkSBIm/XQppboBVYH3zr8gn9/JkzEza61RSrFlyxYePnxIxYoVY93ndbE5YsQIzpw5Q/78+Vm8eDEWFhZJkDh+UuLfC5GyySnWWGitI7TWJ4kaxtPX2HmEiI/FixdjZmbGv//+a+woQqRId+/eZe3atQB8//33bN++/b3F5vnz5/nyyy9p2LAhwcHBbN++na1bt0qxmUSUUi2B6cAPWmtfY+cRKcPy5ctp3bo14eHh5M+fP07F5uPHj+nRowcNGjTg33//Zd26ddy7dy9ZF5tCGIMUnHGXlqh7OIVIUR48eMDo0aMpV64cBQu+9/YnIcQHTJs2DTs7O3x931+/hISEMHnyZGrWrMndu3dp2bIlV65coVWrVjKENokopRoDy4DmWutLxs4jUo7w8HBCQkIIDX1nMNs7QkJCcHJyokSJEty8eZOaNWty8eJFrK2tY117V4hPkXwCxkAplUcp1V4plUUpZaKUagR0AH41djYhPtaQIUMIDQ1l4cKF8kEoxEd6/vw5AI6Ojnh4eJA7d+532mitmThxIqampjg4ONCsWTM8PT3ZsWMHGTJkSOrIqU5c1+BVStUH1gNttNankzqnSHlCQkI4deoUAH369GHfvn1kzpz5g/scO3aMnDlzYmdnR6VKlVi5ciVubm6YmpomRWQhUiQpOGOmiRo+exfwA2YBA7XWu4yaSoiPdOTIETZu3MiIESNkVlohPkJoaCi9e/emVq1aBAYGkilTJszMzN5p5+XlRdOmTRk3bhwvX75k06ZNbNmyRX7fDCuua/A6ANmB/a9eD1RKHTBOZJHc+fr60qBBAxo2bMjjx48BPjgS4eHDh/Tq1Yvvv/+eNGnSMHToUI4dO8ZXX32VVJGFSLFk0qAYaK0fA3WNnUOIhAgJCaF///58/fXXjBgxwthxhEgxHjx4QJs2bXB3d2fkyJExrr335MkTmjdvzunTp8mUKROzZs2iT58+sV4dER9Paz0eGP+ezVmitZMlUESceHt706RJkzf3Zn/o6mRkZCR2dnb88ssvAAwaNIhx48Z9cMIwIcR/ScEpRCo1c+ZMvL29OXz4sAzrEyKOPDw8aN26Nf7+/ri4uGBpafmf7VprNm3ahL29Pffv36dMmTIcPXqUfPnyGSmxEOJjnDhxgpYtW2JiYsJvv/1GjRo13tv29OnT2NnZcfr0aXLlysWmTZv4/vvvkzCtEKmDDKkVIhX6559/mDp1KlZWVvLhKEQcrVy5krp165IhQwbc3d3fKTZ37txJvnz5+PHHH8mXLx+//vorly9flmJTiBRiw4YNNGzYEFNTUzw8PN5bbD558oTSpUtTvXp1bt++zdq1a/H19ZXPUyHiSQpOIVIhW1tb0qdPz5w5cxKlf601gwYN4tIlmQRSpHxhYWHY2tpiY2ND3bp1OXPmDOXLl3+z3c/PDzs7O1q3bs3jx48ZOXIkp0+fxtxcRnAKkRJorZkyZQrW1tbUqFEDNzc3ihYt+k67iIgIVq5ciZmZGVevXqVq1apcu3aNjh07yqR7QiSADKkVIhUaNWoU9+/fp0CBAgbvW2uNjY0Nq1atwsLCAjs7O4MfQ4ikEhQUROPGjTlx4gT29vZMmzaNtGmjPhojIiKwsbFh69atBAcH07dvX4YPH86XX34ZS69CiOSkf//+LF68GGtra1asWEH69OnfabNr1y46duxIYGAgtWrV4ujRo1SoUMEIaYVIfaTgFCIV0VqjlKJOnTqJ1v+AAQNYtWoVnTt3plOnTolyHCGSSsaMGSlXrhx9+vThxx9/fPP66dOnsbW15cyZM+TIkYMTJ05QqVIlIyYVQsTXt99+y+eff87EiRPfuVL57Nkzxo4di5OT05vZZ6dPny5r5wphQFJwCpGKDBkyBKUUs2bNMvjwn4iICKpXr86ff/6Jvb09M2fO5Pjx4wY9hhBJZd26dVSuXJnSpUuzcOHCN69fuXKFNm3acO3aNfLnz8/SpUuxsbGRL59CpDC3b9/m4sWLNG/eHGtr63e2a63p06cPa9as4eXLl/Tr14+JEyeSK1cuI6QVInWTT1AhUgmtNWFhYbx8+dLgxWZkZCS9evXizz//pHLlysycOVPuZxEp1vPnz7G3t2f27NlvXgsPD2fevHlUr16da9eu0bx5czw9PenZs6cUm0KkQMOGDaNnz568ePHinW2XLl2iTp06LF26lHTp0uHq6oqTk5MUm0IkErnCKUQqoZRiwYIFaK0N2m9ERAQ9e/Zk1apVDB06lGnTpkmxKVIkPz8/smfPTrZs2fj9998pUqQIAPPmzWPatGk8fPiQ//3vf0ycOJHq1asbOa0QIj4iIiIwMTFh8eLFPHz48D9r496/f58mTZpw6dIlcubMydKlS+nevTsmJiZGTCxE6pcsTtsqpcYrpXQMj6/falddKeWmlApRSj1QSk1TSplE2/6VUuo3pVSgUuqsUqrCW/vbKaUOJNXPJURS2bp1Kx4eHgAGLQYjIiIoX748q1atYvTo0cyYMUM+mEWKdP78eSpVqsT48eMBKFGiBD4+PlhZWTFw4ECePHnCxo0bOXjwoBSbQqRQCxYsoGHDhoSGhpIrVy5KlSoFRI0A2rBhA5UrV+bChQtUrVoVLy8vevbsKZ9pQiSBZFFwvnILyP/W4+brjUqpQsARwBOoAvQFegNTovUxG/AFKgLuwPJo+xcGRgC9Eu0nEMIIfHx86NGjB2PHjjVovxEREXTr1o0rV65Qo0YNJk2aJFc2RYq0adMmatWqRXh4OM2bN+f58+c0atSIEiVKsHv3bkaNGoWPjw/t27eXv+NCpEAREREMHDiQAQMGkD17diIiIt5s27lzJwUKFMDa2ppChQpx4sQJTp06JcNnhUhC8So4lVKuSqmVSqnpSilfpdRzpdRypVTGBGSJ0Fr7vPWIiLa9L/AcsNFa/6213gk4AHZKqdfjJUoBK7TW/wCLgNLR9l8GTNBa/5uAjEIkO/b29gQHB+Pk5GSwPl++fImlpSVr165l4sSJuLu7yxdxkeJEREQwfPhwOnToQJUqVfjzzz959OgRFSpU4PDhwxQpUoSrV68yZcoUPv/8c2PHFULEw4sXL2jTpg3z5s1j4MCBbNu2jUyZMhEQEMDQoUNp06YNDx8+xMHBAQ8PD2rXrm3syEJ8chJyD2dbYDPwHfA1sAIIAgYopUYBo2LZf6rWemq05wWVUndf/fkSMElr7RZt+7fAYa11ZLTXDgJOQCXgJHABaKyUOgw0Ac4BKKV6ElVcL/von1KIZOzXX39l/fr1ODg4UKJECYP0GR4eTsWKFbl69SojR47EwcHBIP0KkZSePn1Khw4dOHz4MH379qVFixZ89913eHt7Y2ZmxsaNG2nfvr2xYwohEsDHx4fmzZvz119/sWDBAmxtbYmMjOSnn35i7dq1+Pn50b17d8aMGfPmnm0hRNJLSMH5FOjz6irkVaXUGGCBUmok8AvgEof9XzsFdAauAdmJupp5QinVWGt95FWb/MAfb/XhE20bgD2wmKjhuV5Ad6XUF8B4oDYwUinVBfAH7LTWp+L+4wqRvISGhtKvXz+KFi3KyJEjDdJnWFgYHTt25OrVqzRq1IgpU6bEvpMQycylS5do2bIl//77L05OTty/f59mzZq9ueI5ceJE0qVLZ+yYQogE+Pvvv2nSpAm+vr7s2rWLZs2ace3aNfr378+vv/5K1qxZcXNzo2bNmsaOKsQnLyEF5+m3hrz+AaQDimmtL/LfgvKDtNZvT+Rz4lWhOJSo+zbfu2v0/2qt7wMW0RsopfYC04kabtsZqE7U1dKtSqliWuuXcc0pRHLi6OiIp6cnBw4cIGPGhIxmjxISEkLDhg35448/cHR0xN7e3gAphUhax44dw8LCgmzZstGqVSvGjx+Pr68v1tbW2NvbU7FiRWNHFEIk0NGjR2nTpg2ZM2fmxIkTfPHFF3z77becPn2azJkzM336dAYNGiQnloRIJgw5adCbG7yUUqNezRT7oUdsQ27dgcLRnj8A8r3V5vVzH2KglOpE1BVTJ6AhsEdr/UxrvQ/ICJSM808nRDJy48YNpkyZgqWlJY0bN05wfy9fvqRGjRr88ccfDBgwIFkWm0op21ezT4cqpVZ/oF1XpVTEW//e1Iu2vfCr2ayDlFLXlFINkyK/SBplypShZs2afPnll7i4uBAWFsbJkydZt26dFJtCpBJOTk589dVXuLu7c/PmTSpUqICbmxu1atXCy8uL4cOHS7EpRDKSkCuc1ZRSJtGuctYEXgLX+fghtTGpBESf4OcPoJNSKk20+zgbE3Xf6Lm3d1ZK5QWmAfW11loplYZXBbaKmv0kLclrll4h4kRrja2tLWnTpmXu3LkJ7u/ly5e0a9eOCxcu0LVrV+bNm2eAlIniPjAZaETUCaMPcddav29miI1EndBq8uqxVSlVXGv92GBJRZLy9/dn7ty5dOrUiU6dOnHq1Cly5MjBlClTsLe3ly+eQqQCWmueP39O9uzZWbt2LUeOHKFZs2ZcvnyZ8uXL4+TkRNu2bY0dUwgRg4QUnJ8DC5VS84CiwCRgmdb6BfCCjxhSq5SaA+wl6t7LbEBP4Hv+Ozx2MWALLHvVvtirYy54dcy3LQR+1lp7vXr+O1H3mK4l6n7OEKKWWBEiRbl37x5nzpxh0qRJfPHFFwnq68WLF1SuXBkvLy+cnJzo37+/gVIantZ6O4BSqipQMD59KKVKAJWB/2mtg4FtSqmBQBuiTpSJFGj//v1MnjwZR0dHgoKCaNq0KWvWrJGZZ4VIRXr37s1ff/3F4cOHmTt3LlOnTkVrzcyZMxk0aBBp0ybkK60QIjEl5LdzKxBA1Oyw6YAtwLB49pUfcAZMiZrQ5yLQUGv96+sGWut/lVL/A+YAfwLPgKXAmLc7U0q1AQoBVtFe3kHUvZtHgCfAj1rrkHjmFcJoChYsiKenJ9myZUtQPyEhIXz//fd4eXnRrVu3ZF1sxkMlpZQvUSe+1gLTtNbhQBnghtY6IFrbC69eFymMj48Phw4dYvLkyURERFChQgWGDRtGy5YtjR1NCGFgzZo1w8vLi7Jly/LgwQMsLS0ZM2YM5cuXN3Y0IUQsElJwRmqthxI1sU+CaK07xLGdB1ArDu22Adveek0DQ149hEiRfv/9d2rVqpXgBatDQkJo1aoV7u7uTJw4MbUtffI7UBa4TVQhuRkIJ2qIfRaiTmpF5w/EeKlYKdUL6AWQN29eXF1dP3jgwMDAWNskNykx8/Pnz+nQoQNbtmwhIiKCNGnSMGLECP73v/+hlEqWP09KfJ9TYmaRuvzzzz+cPn2a6tWr4+TkxPHjx8mVKxfHjx+nTp06xo4nhIgjGX8gRArh7e2Nubk5EyZMYMyYdy7sx5m/vz9mZmb4+PiwfPlybGxsDJjS+LTWN6I9vaSUmkjUibFpQCBRw/ajy0bUaI2Y+lpK1EgKqlatquvVq/fBY7u6uhJbm+QmpWV+/PgxpUuXxtfXF6UU3bp1Y+LEiRQsGK9R1kkmpb3PkDIzi9TDzc2NFi1aEBgYiNaadOnSMXDgQKZMmUKmTJmMHU8I8RFk0hwhUoivv/6azZs3J2joa1BQEM2bN8fHx4dOnTqlumLzPTT/P4v230BRpVTWaNsrvHpdJHPOzs4ULFgQX19fTE1NuXz5MitXrkz2xaZImLjOUP2q7SCllI9Syl8ptVIplT6JYgoD2rJlC/Xq1cPPz4/Q0FDq16+Pp6cnc+fOlWJTiBQoXgWn1rqe1rqHocMIIWIWGhqKUoq2bduSM2fOePXh5+dH06ZNOXnyJMuWLcPZ2dnAKROXUiqtUioDYAKYKKUyKKXeGaWhlPrh1SzVKKXMAAdgF8CrScTOA+Ne7d8KKM9bQ/BF8nLkyBG+/vprunTpQnh4+JuTJqVLlzZ2NJE0Xs9QvfJDjZRSjYARQAOillUrCkxI7HDCcLTWzJ07l3bt2hEWFkaRIkWYO3cuBw4coECBAsaOJ4SIJ7nCKUQy9+jRI4oVK8bmzZvj3Yefnx/FihXD1dWVtWvX0qNHijxfNAYIJuoLZcdXfx6jlPry1VqbX75q1wC4qJR6AewHtgNTo/XTHqgK+AHTgbayJEry9OzZMwYPHkyjRo24fv06efLk4fLlywwePJg0aeTj61Ohtd6utd5J1IR/H9IFWKG1/ltr7UfUTPZdEzufMIyAgAC++eYbdu/eDfEg8CIAACAASURBVMDkyZO5cuUKAwcONHIyIURCyT2cQiRzQ4cO5dGjR1SoUCFe+wcEBNC8eXP8/Pzo2bMn1tbWBk6YNLTW44Hx79mcJVo7e8D+A/3cAuoZLpkwtJCQELp168bBgwd59uwZAObm5uzatYusWbPy8OFDIycUyVQZXo1meOUCkFcp9bnW+p1iVSYFSx4iIyOZP38++/fvJywsDFNTU8aOHUvZsmVxc3MzdrxYpYT3+G0pMbNI2aTgFCIZc3V1xdnZmVGjRmFmZvbR+/v4+NC8eXPOnTvH5s2badeuXSKkFMJwDh06xIABA/Dy8uKrr75i+PDhBAcHM27cOLmqKWLz9izUr/+clRiujsqkYMbn7u5O165d8fLyQinFiBEjaNSoUbLO/Lbk/h7HJCVmFimbFJxCJFMvX76kX79+FClShNGjR3/0/v7+/pQuXRo/Pz82bNggxaZI1g4cOMCgQYPw9PSkYMGCdO/enWXLlkmRKT7G27NQv/5zjLNQC+P5/fffGTduHK6uruTLl4+WLVsyfPhwatSoIVfehEiFpOAUIpmaPXs2V69eZd++fR89K9+zZ89o1KgRz58/Z9CgQXToEKelboVIck+fPmXChAk4OTkRGRnJmDFjOH/+PCdOnCA8PJx06dIZO6JIOf4matZpl1fPKwAPYxpOK4zD39+fKVOm4OjoCED9+vXZvXs3mTNnNnIyIURikoJTiGTo5s2bTJw4kdatW9OkSZOP2vfOnTs0bNiQW7dusW3bNiwsLBIppRDxFxwcTOfOnTl48CBBQUF07doVGxsbatWqxdOnT6XYFG+8mo06LdFmqAbCtdbhbzV1BlYrpdYDD4iaaGx1UmYVMQsODsbGxobDhw/z9OlTLCwsCAkJoUSJErLMiRCfABmrJEQyo7XGzs6OtGnTMm/evI/a9+nTp1SrVg1vb2+WLl0qxaZIdrTW7N+/nwoVKrB161Zy5MjByZMnefjwIfb29oSFhZErVy7y5Mlj7Kgi+YjTDNVa64PATOA34ParxzjjRBYQ9fu+b98+SpUqxcaNG8mSJQs7duxg586d7N27l/nz56OUir0jIUSKJgWnEMnMzp072bdvHxMmTPioBe2fPHlCgwYN8PPzY+rUqXTt2jXxQgoRD3v37qVw4cI0bdoUpRRr1qzh6NGj2NjYcOjQITp16sRnn31m7JgimdFaj9daq7ce47XWd7TWWbTWd6K1naO1zqu1zqa17qa1DjVm9k+Zi4sLZcuWpVmzZqRLl47OnTvj6+vLggULAEibVgbZCfGpkIIzBkqp9EqpFUqp20qpAKXUOaXUD8bOJT4NZcqUoW/fvgwYMCDO+1y/fp1SpUpx5coVdu/ezciRIxMxoRAfx9fXl/79+9OiRQvu3LnDsGHDuHTpEvnz56dGjRo8evSII0eO0LdvX2NHFUIk0IMHD+jRowdWVlZcvXoVR0dH2rZti7OzM2XLlsXZ2dnYEYUQSUwKzpilBf4F6gLZAQfARSlV2IiZxCeiRIkSLFq0KM5nfx8/foy5uTmPHz9m5syZNG7cOJETChE3wcHBtGrVii+//JIlS5bQt29fPD09mT59OosWLaJx48YUKlSIM2fOyBT9QqRwvr6+NGjQgGLFiuHs7Ezv3r05d+4cx48fZ9q0aXTv3p3jx49ToEABY0cVQiQxKThjoLV+8Wq4zi2tdaTWei9wE6hi7Gwi9bp+/Tpt2rTBx8cnzvs8fPgQc3NzfH19Wb9+PT/99FMiJhQibrTW7N69m3LlyrFz505y5MjBxYsXWbhwIV999RXdu3dn0KBBWFhY4ObmRpEiRYwdWQgRT5GRkaxdu5ayZcvy66+/UqJECa5evcrAgQOxsrLi4MGDLFy4kOXLl5M+fXpjxxVCGIEMoI8DpVReoARRU66/va0X0Asgb968sa4fFRgYmOLWmJLMScPT05M//viD06dPky1btljb37p1i759+xIREcH06dMpUKBAkv/MKfF9Folr9+7d9O7dGx8fH8zMzNiyZQtt2rRBKcWzZ89o0qQJ7u7ujBs3jrFjx8o6m0KkYAsWLGDatGk8ePCAqlWr4uTkRNu2bdm7dy/W1takT5+eo0ePUrduXWNHFUIYkRScsVBKfQasB9Zora+9vV1rvRRYClC1alUd27AwV1fXFDd0TDInncmTJ8dpKYj79+/TvXt3QkNDcXR0ZPDgwUmQ7l0p9X0Whvfo0SMcHBxYtmwZAIMGDWLGjBn/mQQoW7ZsFC5cmMGDB9O2bVtjRRVCJJC3tzfDhw9nx44dmJiYsHTpUmxsbEiTJg1nz56lRYsWVKpUiR07dvDll18aO64Qwsik4PwApVQaYC3wErA1chyRSu3YsYN79+5RtmzZOBWb//zzD02aNOHRo0ccO3YMc3PzJEgpRMxCQkKwsrLi0KFDREREMGDAAIYMGUKhQoXetHFxcaFmzZoUKlSIDRs2GDGtECIhbt26RevWrbl48SIZM2Zk7Nix2NnZkTt3brTWAFSpUoVly5bx448/kjFjRiMnFkIkBzKW6T1U1MJQK4C8QButdZiRI4lURmvNlClTaN26NRs2bCAiIiLWfS5dukTp0qW5ffs2hw4dkmJTGI3Wmu3bt1OmTBl2795Njhw5uHTpEj///PN/ik1fX1969uzJ9OnTjZhWCJEQL1++ZN68eVSqVIlz585RtWpVvL29mTBhArlz5+bGjRt8++23XLlyBaUUNjY2UmwKId6QgvP9FgOlgOZa62BjhxGpS3BwMD/++CNjxozB2tqaY8eOYWJi8sF97ty5g4WFBQBz587l22+/TYqoQrxj9+7dmJqa0qZNGzJmzMiePXt48OABZmZmb9oEBQWhtSZ37twcP36cuXPnGjGxECI+tNaMHTuW3LlzM3DgQKpVq4aHhwceHh7ky5fvTbt06dLh7++Pr6+vEdMKIZIrKThjoJT6CugNVAR8lFKBrx7WRo4mUoF79+5Rp04dNm/ezLRp01i7dm2sZ4JPnz7Nd999x5MnTzhx4gT9+vVLorRC/L8HDx5gY2ODhYUFfn5+2Nracv78eZo1a0bUoJAo3t7eVKlShfnz5wNQsWLFOA0XF0IkH3/99Rfm5uZMmjSJ0NBQ1q1bx6FDh6hevToQVYy6uLgQERFBwYIFuXTpEnXq1DFyaiFEciT3cMZAa30bULE2FOIjnTlzBgsLCwICAti5cyctWrSIdR9vb29q166N1po//viDb775JgmSCvH/QkJCaNeuHQcPHgRgyJAhDB8+HFNT03faHj58GCsrK0xMTKhYsWJSRxVCJNC1a9do1qwZ169fx9TUFCcnJ2xsbMiQIcObNkFBQdjY2LBp0ybWr1/Pjz/+KDNOCyHeS/51ECKJbNy4kTp16pA+fXrc3NziVGzeuHGDhg0bki5dOpYtWybFpkhSr69gmJmZsWfPHnLlysX58+eZNWvWO8Wm1pq5c+fyww8/UKhQIc6cOSNLIQiRggQEBDBu3DgqVarEjRs3MDc3x9vbm/79+/+n2Lx16xbffvstmzdvZvr06XTo0MGIqYUQKYFc4RQiCTx79gw7OzuqVavGtm3bYrwy9DZXV1csLS2JjIzkxIkTVKpUKQmSChFl7969dOzYEX9/f8qXL8/hw4f5/vvvY2wbGhpKnz59WL16Na1bt2bNmjVkyZIliRMLIeIjIiKCQYMG8csvvxAWFoaVlRWTJ0/m66+/fqftb7/9hqWlJeHh4ezbt48ffvjBCImFECmNFJxCJKKgoCAyZsxIjhw5OH78OMWLF4/TvWze3t788MMPhIaG4u7uLsWmSDJ3795l9OjRODs7Y2JiQt++fVmwYMF7J7V68OABrVu3xsPDg/Hjx+Pg4CBD64RIIY4cOcLQoUO5cOECWbJkYfPmzbRq1eqddlprFixYwODBgylRogS7du2iePHiRkgshEiJ5FuBEInk2bNn1KxZkylTpgBQpkyZOBWbnp6e1K1bl0yZMrF58+Y3EzQIkZiCgoKwsLDgq6++YuPGjQwfPhxfX18WLVr03mLz3LlzVKtWjYsXL7J161bGjRsnxaYQKcCFCxfImzcv//vf//D392fTpk34+/vHWGyGhITQrVs3fvrpJ5o1a4aHh4cUm0KIjyJXOIVIJNmzZ+e7776jWrVqcd7nwIEDWFpakjFjRo4fP07ZsmUTMaEQEBkZyfr16xkxYgR3794lX758HDx4kAoVKsS6b7Zs2cibNy/79u2LU3shhHH5+PgwZcoUFi9eTJo0abCwsGDTpk3/uUfzbfb29qxZs0ZGMAgh4k0KTiEMzNnZmRo1alCiRAmcnJzivN+tW7ews7MjKCiInTt3SrEpEt2BAwdo27YtQUFBVKpUiXXr1sU60U9ERASbN2+mQ4cOFCtWjLNnz/5nSRQhRPITGhpK7969cXZ2RilFnz59GD9+/AfnE9Bao5TCwcGBxo0b06xZsyRMLIRITeQ0lRAGEhERwZAhQ+jSpQuzZ8/+qH0vXbrEoEGDyJw5M8ePH6dhw4aJlDLlUkrZKqXOKqVClVKrP9Cui1LqT6XUc6XUXaXUTKVU2mjbXZVSIdHW1/VMkh8gGfH09KRdu3Y0adKEly9f0qtXrzjPKrtt2zasra05dOgQgBSbQiRjr0cwlCpVijVr1mBqasqhQ4dYuHDhB4vNFStW0LRpU8LDw8mbN68Um0KIBJErnEIYgL+/P+3bt+fgwYPY2dkxZ86cOO0XFBREjx492Lx5M9myZeP48eOULFkykdOmWPeByUAjIOMH2mUCBgKnAFNgN2APTI/WxlZrvTyRciZbvr6+WFpa4urqSsaMGRk7dizffPMNTZs2jXXfsLAwPvvsMywtLcmZM+d7Z6wVQiQP7u7utGjRghcvXlCuXLkPzjT9NhMTE7TWBAcHkzVr1kROKoRI7eQKpxAJ5O3tTY0aNTh69ChLlixh/vz5pE374XM5ERERrFixgpIlS7Jx40ZMTU0ZP368FJsfoLXerrXeCTyJpd1irfUJrfVLrfU9YD3wbZKETKaePXvGpEmTKFas2JvZks+ePcuECRPInDlzrPsfOnSIkiVL4u3tjVJKik0hkrGTJ0/Stm1batWqRUREBF26dOHcuXOx/t76+Phw8OBBALp27cr+/ful2BRCGIRc4RQiAY4dO4alpSVKKY4cOUK9evXitI+FhQUvXrygcuXKb+6bc3V1TfS8n6g6wN9vvTZNKTUd8ARGa61dY9pRKdUL6AWQN2/eWP8fBQYGJqv/jxEREezevRsnJyciIyOpVasWPXv2pHDhwjx69IhHjx59MLPWmi1btrBkyRIKFy7Mn3/+yb1795L2h4hBcnuf40Iyi8R2584dunTpgqurK5kyZWLcuHF88803NGnSJNZ9T58+TatWrQgNDeXWrVtkyZJFhssLIQxGCk4h4mnRokUMGDAAMzMzdu/eTdGiRT/Y/uzZs0yePJldu3aRNWtWevXq9WamQJE4lFLdgKpAj2gvDweuAC+B9sAepVRFrfX1t/fXWi8FlgJUrVpVx3ZCwdXVNU4nHRKb1pr169czc+ZMLl26RN68eRk2bBiDBw9+p+37MoeEhNCnTx/WrFlDmzZtWL16NVmyZEmC9LFLLu/zx5DM8aeUygWsAP4H+AIjtdYbYmiXHpgHtAI+A/4A+rwa6ZBq/f3338yaNYsNG6LekipVqrBu3TrMzMzidMJg9erV9OnTh/z587N///5k83suhEg95JuuEPHg6upK//79+eGHH3Bzc/tgsfnkyROaNGlCtWrVOHLkCJMnT8bHx4clS5ZIsZmIlFItibpv8wette/r17XWp7TWAVrrUK31GqK+lMZ+CSCFOHfuHMWLF6dTp048f/6czZs38+DBgxiLzfd58OAB9erVe7MUgouLi3wJFca0kKgTRHkBa2CxUqpMDO1+AmoC5YECwDNgQVKFTGr+/v6MHj2acuXKsXr1ajp37sw///zD2bNnMTMzi3X/sLAwfvrpJ7p160bt2rU5e/asLG8khEgUcoVTiI/wepr4unXrsm3bNiwsLDAxMYmxbUBAALNnz+bnn3/m+fPnlCpVis2bN1OuXLkkTv3pUUo1BpYBTbXWl2JproEUP3bs1KlTzJo1i61bt5I9e3aaN2+Os7MzOXLk+Kh+zpw5Q8uWLfH392fbtm20bt06kRILETulVGagDVBWax0InFRK7QY6ASPeal4EOKS1fvhq301A3GZwS0GePn2Kra0tBw8exM/Pj9q1azN69GgaN24c5z4eP35Mu3btcHV1ZfDgwcyYMSPWuQeEECK+5PKKEHHk6elJ5cqVuXTpEkopWrduHWOxqbVm8+bN5M6dmwkTJlCzZk0uXrzIlStXpNhMAKVUWqVUBsAEMFFKZYi+3Em0dvWJmiiojdb69FvbciilGr3eVyllTdQ9noeS4mdIDH5+fgwYMIAaNWqwY8cORowYwa1bt9i9e/dHF5sbNmzgu+++I126dLi5uUmxKZKDEkCE1tor2msXgJiucK4AvlVKFVBKZSLqauiBJMiYJMLCwvjll18oWrQoGzduxMzMjL/++osTJ058VLF57tw5qlatiru7O87OzsyePVuKTSFEopJ/YYSIo6xZs5ImTRqCgoLe28bFxYV58+bh5ubGF198Qe/evXFwcEjClKnaGGBctOcdgQlKqZVE3ZNZWmt9B3AAsgP7o016cUJr/QNR93VNBsyACOAa0FJrneLW4gwICGDo0KG4uLjw7NkzateuzfTp0/n22/hPyPvkyRNq1KjB1q1byZ07twHTChFvWQD/t17zB2KaPtULuAPcI+r3+xJg+76OU8qkYOHh4SxcuJATJ07w5MkTSpcuTZcuXWjVqhX+/v4fzPR25uDgYNq3b0/69OmZN28ehQoVSnYTQ6W0yapSWl5ImZlFyiYFpxAfoLVm27ZttGrVigIFCnD27NkYZ+67efMmnTt35uTJk+TMmZNly5bRrVu39w63FR9Paz0eGP+ezVmitTP/QB+PgWoGDZbEIiMj2bRpE/379+fZs2d89913LFiwIN73XgUGBuLm5katWrWwtbWlb9++crVDJCeBQLa3XssGBMTQdjGQAfgceAEMI+oKZ/WYOk7uk4Jprdm7dy+jRo3i8uXLZM+enb1799KkSZM4zyD7OnNkZOSbOQO2bt1K+fLlyZs3b2LGj7fkMllVXKW0vJAyM4uUTYbUCvEeoaGhdO/eHUtLS9avXw/wzof8nTt36NGjB2ZmZpw9exZzc3MuX75Mjx49pNgUBvfzzz9TtmxZrK2tKViwIJMnT+b48eMJmuhj7ty5NG/enICAAJRSUmyK5MYLSKuUKh7ttQq8u9TR69dXa62faq1DiZow6BulVIq7XL9kyRLy5MlDixYtCAkJwcnJCV9fX5o2bfrRy5UEBATwww8/sHz5cgC+//77ZFtsCiFSJ/lmIUQMHj58SOvWrXFzc2P8+PF07NjxP9vDwsJYsmQJgwYNIjw8nM6dOzNlyhQKFixopMQiNbt8+TKDBw/myJEjZMqUCWdnZ6ytrRM0y/HrCbD69OlD3rx5ZYF3kSxprV8opbYDE5VSPYCKgAVQK4bmZ4DOSilXIAjoB9yPPkt1cvfnn38yevRoDh06RJo0aRg/fjyjRo3is88+i3efmTJlIn369DIruhDCaKTgFOItFy5coEWLFjx+/BgXFxcsLS3fbIuMjGTq1KmsXbsWLy8vKlSowJAhQ+jUqZMRE4vU6q+//qJv376cOXOGbNmyYWtry4QJE8iVK1eC+l24cCFHjx5l69atmJqaUqdOHQMlFiJR9ANWAo+AJ0BfrfXfSqnvgANa69dD6u2B+YA3kA64TNSanMne+fPnad68OXfv3uXzzz/H0dGR7t27J+h3fdeuXYSHh2NiYsKuXbs++sqoEEIYihScQkSzY8cOOnbsSK5cuTh58iSVK1d+s+3cuXN069aNCxcukC9fPnbt2kXz5s3lQ1wY3PPnz5kxYwaOjo6EhYXRtWtXZs2axeeff56gfsPCwhg4cCCLFi2iRYsWhIaGGiixEIlHa/0UaBnD6yf47/3bT4iamTbF8PLyYsaMGaxatYo0adLw/fffs3XrVrJle/u21bgLCwtj1KhRzJo1CwsLC9q0aSOfU0IIo5KCUwiihhdOmTIFBwcHqlevzo4dO8ifPz/w/0OcDh8+TM6cOencuTOLFi0ic+bMRk4tUpugoCC6d+/OgQMHeP78Oe3ataNPnz6Ym793HqQ48/Pzo127dhw9epRhw4YxdepUuc9YCCN59OgR1tbWHD16lM8++4yffvqJkSNHkidPngT1e+fOHaysrPDw8KBv3760bPlOnS6EEElOCk4hgKlTp+Lg4EDHjh1ZtmwZGTJkIDAwkJkzZzJlyhQiIyOxt7dnzJgxZM+e3dhxRSrzejbkESNGcP36dfLnz8+xY8eoWrWqQfr38vKiefPm3Lx5k1WrVtG1a1eD9CuE+Dh3795l/vz5LF68mBcvXlC8eHE2bdr0n9E08bV37146d+5MeHg4mzdvpl27drL0hRAiWZCCUwigV69eZM+enf79+xMWFkb37t3Zt28fjx49onHjxowYMYK6desaO6ZIhZYvX87IkSPx9fWlTJkyrFu3jg4dOhhsgo9jx47Rtm1b0qZNy6+//krt2rUN0q8QIu6Cg4OZM2cODg4OaK2xtLRk0qRJlCxZMsF9h4WFMXr0aBwdHalYsSJbtmzh66+/NkBqIYQwDJmyTHyyTp8+jbW1NWFhYZiammJra8uRI0coU6YMq1atInPmzLi7u3PgwAEpNoXBXbt2jVatWtGzZ0+ePn3KxIkTOX/+fIJnn43ul19+oVGjRnzxxRecPn1aik0hklhwcDBDhw7l66+/ZsyYMZQsWZK1a9fi4uJikGLz33//pW7dujg6OtK3b1/c3d2l2BRCJDtyhVN8sjw9PXF3d+f+/ftcuHABBwcHLl68SJEiRZg0aRKjRo2SaeSFwd2+fZumTZty5coVsmTJwqRJk+jduzempqYGPU54eDjr16+nUaNGbNy4MUGTkAghPk5kZCSbN2+mf//++Pn5UaVKFTZu3GjwGaE3bNjApUuX2LhxI+3btzdo30IIYShScIpPSmRkJBcuXKBSpUp06tSJ2rVrM23aNJYsWYJSimnTpjFo0CDSp09v7KgilQkICGDu3LnMmDGD4OBgqlSpwr59+xI8Scjbnj17htaanDlzsmfPHrJmzSqTAwmRRCIjI5k0aRIbN27E09OTUqVK0a9fPyZOnGiwE5hhYWH8888/lCpVCnt7e9q3b89XX31lkL6FECIxSMEpPhmBgYF07NiRgwcP4ubmxsiRIzl58iQvX77ExsaGoUOHGmSIkxDRhYeHM3jwYBYvXkx4eDht2rRh0qRJlCpVyuDHioiIoF69euTLl48DBw6QI0cOgx9DCBGz33//nWHDhnHq1CmyZcvGhg0bsLKyMvhImX79+rFr1y68vLzIkSOHFJtCiGRPCk7xSbh16xYWFhZcvnyZH3/8kaZNm+Lj40PZsmXZtm0bJUqUMHZEkcpERkayY8cOxowZw7Vr18iaNSsrVqzA0tIy0Y5pYmLC8OHDKVCggKy7J0QS2bx5Mw4ODnh7e1OgQAFGjBjBmDFjDL50VmRkJGnSpGHIkCHUr19fTigJIVIMKThFqnfy5ElatWrFixcvyJEjB+vWraNKlSr8/PPPWFlZGTueSIXc3Nxo0qQJ/v7+lChRgh07dmBhYZFoReDSpUvJli0b7du3p0OHDolyDCHEf3l6euLg4MCWLVtQSuHg4MDIkSPJmDGjQY8THh6Og4MDd+/exdnZGTMzM8zMzAx6DCGESEwyI8p7KKVslVJnlVKhSqnVxs4j4mfFihXUr1+fkJAQgoOD8ff3Z+nSpZw+fVqKTWFwXl5eWFlZ8e233xIWFoa1tTWXL1+mZcuWiVJshoeHM3DgQHr37o2Liwtaa4MfQwjxX3///TelSpWidOnS7N+/nxEjRnDz5k0mTpxo8GLz3r17mJubM336dDJmzEhERIRB+xdCiKQgVzjf7z4wGWgEGPYTRCS6iIgIbGxsWLlyJQDp0qVj3Lhx2NnZ8fnnnxs5nUhtnj9/jpWVFQcPHiRDhgyMHTsWe3t7smbNmmjH9Pf3p3379hw8eJCBAwfi6Ogow2iFSER+fn4MGjSIhQsXEhYWRp06ddiyZYvBJ/567eDBg3Tq1Ing4GDWrVuHtbV1ohxHCCESmxSc76G13g6glKrK/7F33+FRVWsbh3+LUEIJHYPSVUA6SAgoSAkgiIKHQ1MIRSkqRTmgiNKLqEgRVEoogkhTqlRRIBBADF1BESSE3kEkkBCSrO+PAF+AVCAzmeS5r2uuk5n97tnPxDNM3tlrrwUFnRxHkuD69et06dKFoKAgAKpWrcrSpUvx9PR0cjJJrTJmzMi+ffsoXLgwy5Yto3z58sl6vEOHDtG4cWMOHjzI5MmT6dKlS7IeTyQt+/fff2nTpg0rV67EWstrr71G3759KV68eLIcLyIigkGDBjFixAjKlSvH999/rwntRMSlqeGUVCE8PJzFixczduxYdu/ezfXr1ylTpgzTp0/H29vb2fEklXN3d2f//v0EBgYme7O5YcMGmjVrRlRUFGvWrKFOnTrJejyRtC4oKIjly5eTJ08eVqxYQdWqVZPtWCdOnODVV18lICCATp06MW7cOLJkyZJsxxMRcQQ1nA/IGNMF6ALg6emJv79/vPUhISEJ1qQ0KTVzaGgoS5YsYdmyZVy6dImwsDDSpUtH7ty56dmzJw0bNuTatWspMntsUurvOT6umDm5OOKPwmnTpvHWW2/x+OOPs2zZsmQ7wyIi/69ixYoEBQVx5MiRZG029+zZQ7169QgNDWXWrFn4+vom27FERBxJDecDstb6AX4AXl5etnbt2vHWFv4KzwAAIABJREFU+/v7k1BNSpOSMgcHBzN8+HCCgoIICAggIiICYwyNGzcmb9689OzZk3LlyqWozImlzBKfEydO8Pbbb1OnTh3mz5+vJRFEHKhYsWIcOXIkWY9RvHhx6taty+DBgzULrYikKmo4JcX79ddfWb16NRs2bMDf3/+OmThLly7Ntm3bNORIUq3r16+TKVMmChQoQEBAAOXLlyd9ev3TLZIanDx5kv79+zNu3Dg8PDyYN2+esyOJiDx0WhYlDsaY9MYYd8ANcDPGuBtj9Feeg2zfvp0RI0ZQuXJlqlWrxuDBg/n111+x1pIhQwbatm3L1q1b2bt3r5pNSbXOnj1L5cqVmTJlCgBPP/20mk2RVOTgwYMsXLiQXbt2OTuKiEiyUcMZt/5AKNAX8L35c3+nJkrFrLUEBgbywQcf4OHhQZUqVejXrx+nT5/G3d0dgPz58zNy5EhOnjzJN998Q9WqVbUMRBqSlLVxjTH/M8acNsZcNsZMN8ZkirGtqDFmvTHmmjFmvzGmXrKHv0+5c+emfPnyPP74486OIiIPSWRkJGvXrgWgVq1aBAcHU7NmTSenEhFJPmo442CtHWytNXfdBjs7V2oSERHBunXr8PHxwd3dnapVqzJy5EgeffRRWrVqxdatWzl37hz169dn1apVHDx4kPfee4+8efM6O7o4x621cafHV2SMaUD0F0V1gaLA48CQGCVzgV1AHqAfsMAYky8Z8t63OXPmcOrUKdKnT8+cOXOoW7eusyOJyENw6tQp6tWrR/369dm7dy8AuXLlcnIqEZHkpYZTHCo0NJTx48dTvHhx8uXLR926ddm0aRO5c+dmwoQJPPvsszzxxBPMmzePqlWrcvLkSX744QcaNmxIunT6v2taZq1dZK1dAlxIoLQ9MM1au89aewkYBnQAMMaUAJ4GBllrQ621C4HfgWbJlzzxIiMj6dOnD23atOGzzz5zdhwReYh++uknKlasSGBgIDNmzKBs2bLOjiQi4hC6GEiS3YkTJ27PLLtp0yauXbuGMYYGDRrg4+PD4cOHGTduHBkyZCBr1qxkyJDh9r46myn3oQywNMb9PYCnMSbPzW1B1tord20v48B8sbpy5Qpt2rRh2bJldO3alU8//dTZkURSBGNMbmAa8DxwHvjAWjsnjtqngc+J/mLpKjDCWjvOUVljExkZyZAhQxg+fDilSpVi/fr1lC5d2pmRREQcSg2nJItdu3axbNkyNm3axLp164iMjMTDw4P27dvTsGFDTp06xbRp0+jTpw9Zs2alc+fOVKpUiXbt2jk7uri+bMDlGPdv/ewRy7Zb2wvE9kSOWmf39OnT9OvXj+DgYN5++22aNm3K5s2bk/w898MV11JVZsdIQZm/AsIBT6AisMIYs8dauy9mkTEmL7Aa+B+wAMgIFHRw1jucOnWKNm3asH79ejp06MCXX35J1qxZnRlJRMTh1HDKQ7N7927WrFnDokWL+PXXX4HodcV69+5N+fLlqVy5MlOnTqVDhw5cunSJUqVK8cUXX9C2bVty5Mjh5PSSioQA2WPcv/XzlVi23dp+hVg4Yp3dzZs38/bbbxMeHs7q1aupX79+kvZ/UK64lqoyO0ZKyGyMyUr0kPey1toQYJMx5gegLdHXasfUC/jRWjv75v3rwJ8OC3uXtWvX0qZNG/7991++/vprOnTo4KwoIiJOpYvi5L5Za9m5cycDBgwgZ86cVKpUiffff58bN27Qpk0bli5dyp9//kn16tWZNWsWpUqVYty4cdSrV4/169ezb98+unfvrmZTHrZ9QIUY9ysAZ6y1F25ue9wY43HX9jvOlDjKN998g4+PDzly5GDr1q0ObzZFXEAJINJaeyDGY3ENg68GXDTGbDHGnDXGLDPGFHZIyruEhYXRvn17cufOzbZt29RsikiapjOckiSRkZFs2rSJjz/+mPXr1xMeHk66dOkoWbIkPj4+jBkzhqJFixIaGkrmzJmx1vLhhx9y6dIlhgwZQqdOnXjsscec/TLEBd1cBzc9MdbGBSKstRF3lX4DzDDGzAZOEb2c0QwAa+0BY8xuYJAxpj/wAlAeJ0watGbNGtq3b0+dOnVYsGABuXPndnQEEVcQ1zB4j1hqCxJ97WZ9oicDG0n0rNTVY3vi5Bgyf+nSJbJnz46bmxvDhg0jf/78nDt3zmlDk1PQsOhEc7XMrpYXXDOzuDY1nJKg8PBwJk+ezJgxYzh//jwXL14kffr05M6dm759++Lr60u+fP+/qsTEiRMZMGAAwcHBZMuWjWXLllGoUCEtWC8Pqj8wKMZ9X2CIMWY68AdQ2lp71Fq72hgzElgPZAYW3rXfK0Q3oJeAo0Bza+05B+S/Q7169Zg8eTKvvfbaHRNlicgdkjIMPhRYbK3dBmCMGQKcN8bksNbe3bQ+9CHzZ8+epXz58nTp0oWhQ4c6fTgypIxh0UnlapldLS+4ZmZxbeoA5A5RUVEEBQWxZcsWpk6dirWWHTt2EBoaCkCdOnV46623aNiwIR4e0V8wX7lyhUmTJlGzZk1Kly6Nl5cXvr6+hIWFkS1bNooVK+bMlySpxM11cAfHsTnbXbVjgDFxPE8wUPvhJUu8o0eP0rlzZ/z8/ChSpAhdunRxRgwRV3IASG+MKW6tPXjzsbiGwf8G2Bj3b/1skjHfbY888ghdu3blv//9ryMOJyLiMtRwpmFnzpxhx44dHD16lB07djB//nyuX79OeHj47RpPT098fHzo1q0b1atXJ3v2//+ied++fUycOJFvvvmGK1euMGTIEAYOHEiVKlWoUqWKM16SSIp27do1/vrrL4KDgylSpIiz44ikeNbaq8aYRcBQY0wnomepfRl4Npbyr4GFxpjxRDekA4BN1tp/kivfmTNn6NSpEx999BHly5dn4MCByXUoERGXpYYzDbDWcuzYMb7++muCg4MJCQlh9+7d/P3337drcuTIQcaMGalatSqvvPIKFSpU4IknniBXrlx3DL0IDw9n8eLFTJw4kQ0bNpApUyZatWpF165d8fb2dtIrFEnZtmzZwjPPPMNTTz3FgQMHyJgxo7MjibiSrsB04CxwAXjLWrvPGPMcsMpamw3AWrvOGPMhsALIAmwCWidXqPXr19O6dWv++ecf2rdvT/ny5ZPrUCIiLk0NZypz7do19u/fz549e5g+fTpHjx7lypUrXLp06XbNE088QYUKFShbtize3t60bt2awoULY0zco46OHTuGn58fU6ZM4cyZMxQrVoyRI0fy2muvkTdvXke8NBGXExUVxcCBA/noo4+YPXs2rVu3VrMpkkTW2ovAf2J5PIB7h9NPBCYmZ57IyEhGjBjB4MGDKVGiBGvWrKFcuXLJeUgREZemhtOF7d27l59++okbN26wZ88eVq9ezcWLF29vd3NzI3v27LRs2ZIKFSqQJ08eatSokehZYk+dOkXOnDkBGDt2LJ9//jkvvvgiXbt2pUGDBqRLp1V1ROJy9epV2rVrx6JFi+jUqRPNmzd3diQReUBnzpzB19eXn3/+GV9fXyZOnEi2bNkS3lFEJA1Tw+kCbty4wV9//cWiRYtYtWoVHh4e/Pbbb5w5c+Z2TaFChShSpAilS5emR48eVKpUiSeeeCLRTeHly5fZsWMHhQsX5sknnyQgIICaNWuyatUq3N3deffdd+nRo4cmABJJhOPHj9OkSRP27NnD2LFjeeedd+IdQSAiKd/u3bt59dVX+eeff5g6dSqvv/663tciIomghjOFCQoK4tChQ+zbt4+VK1eyadMmIiIiuHHjxu2asmXL0qhRIzw9PXnkkUdo06YNjzzySKKPcf36dfbs2UNgYCDbtm0jMDCQ/fv3AzBo0CAGDx5MxYoVGTt2LKVKleLw4cNaO1MkkQIDA3n55Ze5evUqy5Yto1GjRs6OJCIPaNSoUbz//vsUL15cQ2hFRJJIDaeTREVFsWvXLmbPnk1ERASHDx9m69atnD9//nZNrly5yJIlC02bNqV27dqULFmS0qVLkyVLliQf78aNG/Ts2ZPAwED27Nlzu4H19PTE29ubNm3aUKVKldsT/3h4eNCzZ08ADh8+/BBesUjqt27dOj777DMeffRRfv75Z8qUKePsSCLyEOTPn586deqwePHi20uCiYhI4qjhdKAvvviCUaNGkTVrVo4cOcK1a9cAMMZQunRpatasSVhYGK+++ir169fH09MzSc9vreXSpUvkzp0bgLZt25IxY0amTZtGhgwZ8Pf3J3/+/PTq1Qtvb2+qVKlCwYIFNSRI5CGYNGkSw4YN47nnnmPRokWaTEskFfH19aVAgQJqNkVE7oMaTgeaNWsWx44do0SJEnTu3Jly5crh7u7Oiy++eHtynqS4cOHC7SGxt/43Y8aMHDt2DICiRYveMSPm3r171VyKJJMGDRrQvHlzZs+erZloRVIhfX6KiNwfNZwO9NNPP7Fjxw58fHzua/9bs9IGBgYSGBhIUFAQEP0h+NRTT/HCCy/g7e1NVFQU6dKlY9iwYXfsrw9LkeRTrFgxunXrpmZTREREJAY1nA6UI0eOJC0lsnPnTiZOnMjo0aPJnj0733//PUOHDqVQoUJ4e3vTpUsXvL29qVy5MtmzZ0/G5CIiIiIiIkmnhtPJrLUEBQXdMSx22LBh1KlTh3PnzrFw4UK6detGxYoV6dq1K2+99Rb58+d3dmwREREREZEEqeF0sIsXL7Js2bLbDea2bdu4ePEiAJkzZ+bpp58mIiICgHr16nHhwoXbQ2GTOomQiIiIiIiIM6nhdKBjx47RrFkzANzc3ChbtizNmjW7vRxJmTJlSJ/+//+TuLm5OSuqiIiIiIjIA1PD6UAFCxbk7bffpmXLllSqVOm+1tMUERERERFxFWo4HcgYQ9OmTalevbqzo4iIiIiIiCS7xE+ZKiIiIiIiIpIEajhFREREREQkWajhFBERERERkWShhlNERERERESShRpOERERERERSRZqOEVERERERCRZqOEUERERERGRZKGGU0RERERERJKFsdY6O0OqYYw5BxxJoCwvcN4BcR4mZXaM1Jq5iLU2nyPCJCe9v1MUZXYMvb/vlFr/G6Y0rpbZ1fJCGnpvS8qghtPBjDHbrbVezs6RFMrsGMrs+lzx96HMjqHMrs8Vfx/KnPxcLS+4ZmZxbRpSKyIiIiIiIslCDaeIiIiIiIgkCzWcjufn7AD3QZkdQ5ldnyv+PpTZMZTZ9bni70OZk5+r5QXXzCwuTNdwioiIiIiISLLQGU4RERERERFJFmo4RUREREREJFmo4RQREREREZFkoYZTREREREREkoUaThEREREREUkWajhFREREREQkWajhFBERERERkWShhlNERERERESShRpOERERERERSRZqOEVERERERCRZqOEUERERERGRZKGGU0RERERERJKFGk4RERERERFJFmo4RUREJMUyxnQ3xmw3xlw3xsxIoPZxY8xyY8wVY8x5Y8xIB8UUEZE4pHd2gNQkb968tmjRovHWXL16laxZszom0EOizI6RWjPv2LHjvLU2n4MiJRu9v1MOZXaMFPT+PgkMBxoAmeMqMsZkBH4CvgJaAZFAicQcQO/vlMPVMrtaXkhR721JI9RwPkRFixZl+/bt8db4+/tTu3ZtxwR6SJTZMVJrZmPMEcekSV56f6ccyuwYKeX9ba1ddPNYXkDBeEo7ACettWNiPPZbYo6h93fK4WqZXS0vpJz3tqQdGlIrIiIiqUE1INgYs+rmcFp/Y0w5Z4cSEUnrdIZTREREUoOCQB2gCbAWeAdYaox5ylobfnexMaYL0AXA09MTf3//eJ88JCQkwZqURpmTn6vlBdfMLK5NDaeIiIikBqHAJmvtKgBjzCigP1AK2HN3sbXWD/AD8PLysgkNMUytQydTGlfL7Gp5wTUzi2vTkFoRERFJDX4DrLNDiIjIndRwioiISIpljElvjHEH3AA3Y4y7MSa2EVrfAtWMMfWMMW5AT+A88KcD44qIyF3UcIqIiEhK1p/o4bJ9Ad+bP/c3xhQ2xoQYYwoDWGv/url9EnAJeBloEtv1myIi4ji6hlNERERui4iI4Nq1a86OcZu1djAwOI7N2e6qXQQsSuZIIiKSBDrDKSIiIgBcv36dVq1a0bBhQyIiIpwdR0QeooiICLp06cKKFSucHUXSGDWcIiIiwtWrV2ncuDGLFi2iZcuWpE+vQVAiqUVYWBgtW7ZkypQpnD171tlxJI3Rp4mIiIjQu3dv1q5dy/Tp03nttde0Tp9IKnLs2DE2btzIuHHjKF++vLPjSBqjhlNEkuzff//lk08+wdvb29lRROQhGTZsGE2aNKFRo0bOjiIiD8m///6Lh4cHxYsX5+DBg+TKlUtfJonDaUitiCTZ6dOn+fnnnzlz5oyzo4jIAzh69Chdu3YlPDycfPnyqdkUSUXOnTtH5cqV+eSTTwDIlSuXkxNJWqWGU0SSrESJEvz666+ULFnS2VFE5AFs3LiRuXPncvDgQWdHEZGHLE+ePLzwwgvUrl3b2VEkjVPDKSJJsmbNGs6fP48xxtlRROQ+hYaGAuDr68vBgwcpU6aMkxOJyMOyZcsWDh8+TLp06Rg/fjzPPPOMsyNJGqeGU0QS7erVq7Rs2ZJevXo5O4qI3KfNmzfz+OOPs2nTJgDy5s3r5EQi8rAsX76cunXr0r17d2dHEblNkwaJSKJlzZqVX375BXd3d2dHEZH78OOPP9K0aVMKFSpE4cKFnR1HRB6ib775htdff52KFSsyY8YMZ8cRuU1nOEUkUaKiogAoVaoUxYoVc3IaEUmqBQsW0LhxY0qWLElAQIAaTpFUZMyYMbRv357atWuzfv168uXL5+xIIrep4RSRROnVqxevvPLK7cZTRFzH9OnTadWqFd7e3qxfv55HHnkk3vqIiAgHJRORB2GtpW/fvvTu3ZvmzZuzYsUKPDw8nB1L5A5qOBNgjClujAkzxnzr7CwiznL8+HEmTpxI1qxZSZdO/2yIuJKxY8fSsWNH6tevz5o1a8iZM2e89Xv27OGJJ55g165dDkooIvcjIiKCTp068emnn/LGG28wb948MmXK5OxYIvfQX44J+wrY5uwQIs700UcfYa1lwIABTjm+MSaTMWaaMeaIMeaKMWaXMeaFeOr/Z4w5bYy5bIyZbozJFGNbUWPMemPMNWPMfmNMPce8ChHHstYycOBAevXqRfPmzfnhhx/IkiVLvPssXLiQ6tWrc+3aNTJkyOCgpCKSVDdu3KBFixZMnz6dAQMGMHHiRNzc3BK1761ZqkUcRQ1nPIwxrwD/AGudnUXEWQ4fPsy0adPo1KkTRYsWdVaM9MAxoBaQAxgAfGeMuSeQMaYB0BeoCxQFHgeGxCiZC+wC8gD9gAXGGF3sIqnOzp07GT58OK+//jrz5s0jY8aMcdaGh4dTrVo1mjdvTrly5dizZw9ly5Z1YFoRSYr06dNTqFAhxo0bx9ChQxO1VJm1lqlTp/Lqq6/yxx9/OCClSDTNUhsHY0x2YCjRf7R2dHIcEacZNmwY6dKlo1+/fk7LYK29CgyO8dByY8xhoDIQfFd5e2CatXYfgDFmGDAb6GuMKQE8DTxvrQ0FFhpjegLNgEnJ+iJEHKxy5cps2rSJZ555Jt4/Ri9fvkzr1q359ddfKVOmDP7+/mTKlIkDBw44MK2IJMaZM2f4559/KFmyJOPHj0/0fiEhIVSsWJFDhw5RuXJlTSokDqWGM27DiP6j9Vh8H9TGmC5AFwBPT0/8/f3jfdKQkJAEa1IaZXaMlJj5+PHjzJw5k6ZNm3Lw4EEOHjx4x3ZnZTbGeAIlgH2xbC4DLI1xfw/gaYzJc3NbkLX2yl3bY131Xu/vlEmZ4xYeHs4nn3xCw4YN8fb2BmDDhg1x1m/ZsoUJEyZw+vRpunbtSvPmzfnll18cmllEEsdaS/PmzTl37hx79+4lffrE/RkfFBREs2bNOHToELVr16Z///5qOMWh1HDGwhhTEagHVEqo1lrrB/gBeHl52dq1a8db7+/vT0I1KY0yO0ZKzOzr64u7uztffvkl+fPnv2e7MzIbYzIQfcZyprV2fywl2YDLMe7f+tkjlm23theI7Vh6f6dMyhy3y5cvc/nyZTw8PBI83vz58+nXrx/u7u6sXbuWWrVq3bHdFX/PIqmZMYavvvqKa9euJbrZ7Nu3LxMmTCB9+vQsW7aMl156SV8kicOp4YxdbaKv/Tp68+xmNsDNGFPaWvu0E3OJOMwff/zBnDlzeO+992JtNp3BGJMOmAWEA93jKAsBsse4f+vnK7Fsu7X9CiIu7NKlS7i7u5MjRw5++eWXeK/XtNby8ccf079/f4oWLcp3331HlSpVHJhWRJIiICCAn3/+mcGDB1O+fPlE7RMREcG7777LuHHjyJs3L9u2bXPmPAySxmnSoNj5AU8AFW/eJgErgAbODCXiSDt27CBv3ry89957zo4CgIn+9mca4Ak0s9beiKN0H1Ahxv0KwBlr7YWb2x43xnjctT22obkiLuH06dPUqlWLdu3aAcTbbJ48eZJChQrRr18/XnnlFfbt26dmUyQFW7ZsGc8//zzz58/nypXEfTd68OBB6tevz7hx424PpVWzKc6khjMW1tpr1trTt25EnxUJs9aec3Y2EUdp27YtR44cIW/evM6OcstEoBTQ+OaEP3H5BuhojCltjMkF9AdmAFhrDwC7gUHGGHdjTFOgPLAwWZOLJJPg4GBq1KhBUFAQb775Zry1QUFB1K9fnxMnTtCyZUtmz56d4DIpIuI8M2bMoGnTppQtW5aAgACyZ797gM695syZw1NPPcXmzZuZOXMmCxYsSNR+IslJDWciWGsHW2t9nZ1DxFH27NkDQObMmZ2cJJoxpgjwBtEjDk4bY0Ju3toYYwrf/LkwgLV2NTASWA8cuXkbFOPpXgG8gEvAJ0BzfZkkrujPP/+kRo0aXLx4kZ9//pm6devGWTt+/Hi8vLw4deoUy5cvZ/78+YlaRkFEnGPUqFG89tpr1K5dm3Xr1iU4yY+1ljFjxtC2bVuyZs3Kd999d3vUg4izqeEUkTts376dihUrMmPGDGdHuc1ae8Raa6y17tbabDFus621R2/+fDRG/Rhrrae1Nru19jVr7fUY24KttbWttZmttSWttT8751WJ3L8dO3bw3HPPERkZyYYNG6hWrVqsddZaevfuzTvvvEOmTJnYtm0bL774ooPTPhhjTHdjzHZjzHVjzIxE7rPOGGONMZqrQlyKtZY+ffrw3nvv0aJFC1asWIGHh0e8+xw5coQSJUrQu3dvXn75ZY4dO8Z//vMfByUWSZgaThG5Q9myZfnyyy9p1qyZs6OISCw2btxInTp1yJYtGwEBAZQrVy7WutDQUNq1a8eYMWOoVKkSu3fv5oknnnBw2ofiJDAcmJ6YYmNMGzQporigiIgIOnbsyGeffcabb77J3LlzyZQpU7z7/P7771SvXp2///6bzp07s3DhQnLkyOGgxCKJo4ZTRO7g7u5Ot27dEvxGVUQcb8WKFTRo0IACBQqwadMmnnzyyVjrNm/ejKenJ99++y1Dhw5l+/bteHp6Ojjtw2GtXWStXQJcSKjWGJOD6CH0fZI9mMhD9vbbb/P1118zcOBAJkyYgJubW7z1I0eOpGrVqkRGRrJo0SL8/Pw0VF5SJH0DKCK3dejQgRdeeIFWrVo5O4qI3MVay8iRIylTpgyrVq2K85qugIAA/vOf/xASEsKIESP44IMPHJzUqUYQPcHY6YQKjTFdgC4Anp6eCa5NGBIS4nLrFypz8nuYeZ955hnc3d2pU6cOGzZsiLMuPDyc3r17s3fvXkqWLMmIESPIlStXonO42u9YXJ8aThEBYP369cycOZOKFSs6O4qI3CUiIoL06dOzZMkS3NzcYp11Mioqivfff5/PP/+cYsWKsW7dOipUqBDLs6VOxhgvoDrwDlAwoXprrR/Ry6Dh5eVla9euHW+9v78/CdWkNMqc/B407+nTp/n666/p27cvxhjatm0bb31wcDAtWrRg79691KpVix9//DHBYbcPO7NIUmlIrYhgrWXAgAEUKFAgwaUVRMSxRo8eTYMGDQgLCyNXrlyxNpvXr1/Hx8eHUaNG4eXlRWBgYFprNtMBE4B3rLURzs4jkljffvstw4cP5+DBgwnWDho0iJIlS3LgwAGWLFmCv79/kptNEWdQwykirFmzhs2bN9OvXz/c3d2dHUdEYvD09CR//vxxXs916tQpfHx82LBhAy+88AIbN24kZ86cDk7pdNmJXu5ovjHmNLDt5uPHjTHPOS+WSOwiIqK/F+nduze//fYbJUqUiLM2MjKSgQMHMnToUNzc3Fi/fj0vv/yyo6KKPDA1nCJp3K2zm0WKFKFjx47OjiMiRA+P3bVrFwC+vr58++23ZMiQ4Z66KVOmUKxYMXbt2sX8+fNZuXJlrHWuzBiT3hjjDrgBbsYY91iWO7kMPEb0Wr0VgUY3H68M/OqwsCKJsHHjRkqXLs2BAwcwxsQ7e/Qff/xB9erVGTZsGB06dODUqVM8/fTTDkwr8uDUcIqkccuXL2fbtm0MGDCAjBkzOjuOSJp348YN2rVrR7Vq1W4Ps4tt5skZM2bQtWtXIiMjWbBgAS1btnR0VEfpD4QCfQHfmz/3N8YUNsaEGGMK22inb92Aczf3PWOtDXdSbpF7/PDDDzRo0AA3NzcyZ84cb+3WrVupVKkSgYGBTJ48ma+//lpLnohLUsMpkoZFRUUxcOBAnnzySdq1a+fsOCJpXlhYGM2aNWP27NkMGTKE4sWL31Nz9epVWrVqxWuvvUbNmjU5duwYjRo1iuXZUgdr7WBrrbnrNthae9Ram81aezSWfYJv1ul6TkkxZsyYwX//+1/Kly9PQEAAhQoVirUuKiqK8ePHU7NmTfLly8fcuXPp0qWLg9OKPDyapVYkDVu8eDG7d+9m1qy4y5VuAAAgAElEQVRZqW4YnoiruXLlCi+//DLr16/nq6++omvXrvfUnD9/nkqVKnH8+HE6d+7MhAkTSJ9eH+UiKd1nn31Gnz59eP7551m4cCHZsmWLte7ixYtUqFCB48eP07hxY2bOnEmuXLkcnFbk4dIZTpE0bO7cuZQqVYpXX331vvY9d+5cwoUikqALFy5Qt25dNm7cyLfffhtrs7l79268vLw4e/Ys3bt3x8/PT82mSApnreW9996jT58+tGrVimXLlsXZbP75559Ur16d48eP06hRI5YsWaJmU1IFNZwiadj8+fNZvXp1nLNfxmXXrl20bduWqVOnJlMykbTj5MmT1KxZk99++43FixfTpk2be2q6deuGl5cXN27cYNOmTXzxxRdOSCoiSREREcHrr7/OqFGj6NatG7Nnz45zroQePXpQuXJlLly4wM8//8yKFStIl05/pkvqoK9GRdKgiIgIrl27Rvbs2SlcuHCS9r1+/Trt2rXjkUceoVu3bsmUUCRtCAoKol69epw7d45Vq1ZRp06dO7ZHRkby4YcfMmHCBLJnz87GjRvjndFSRFKOJUuWMGPGDAYPHszAgQNjnfwrPDycrl27Mm3aNB599FG2bdtGgQIFnJBWJPnoqxORNGjWrFkUK1aMQ4cOJXnfBg0asHfvXqZOnRrrAvQiknh79+7l6tWrrF279p5m89ChQ9SqVYuRI0fSpUsXTp06pWZTxAVYawFo1qwZmzZtYtCgQbE2m/v27aNWrVpMmzYNX19fDh06pGZTUiU1nCJpUKVKlWjXrh2PP/54kvbbuHEjGzZsoGTJkql6VkyR5Hb58mUAmjRpwqFDh/D29r5j+x9//EG5cuXYvHkzX331FZMnTyZLlizOiCoiSXD69Gnq1KnDb7/9hjGG6tWrx1rn5+dHuXLl2L17N9999x2zZs1KcJkUEVelhlMkDapYsSJjx46N9RvXuFy9epWOHTtSpEgRNmzYkIzpRFK3bdu2UaxYMVatWgVwzwQiS5YsoWrVqmTOnJlJkybFOoGQiKRM169f5/Tp03FOqhcVFcXw4cN54403yJEjB0uXLqVFixYOTiniWLqGUyQNCQsLo2/fvvTs2ZOiRYsmad+WLVvy999/s379ejw9PZMnoEga8NRTT9G4cWMqVKhwx+MRERH4+PgQEBCAl5cXixcvpmDBgk5KKSJJERQURNGiRSlSpAh79+6NdQbpAwcO8Pzzz3PkyBHatGnD5MmTyZo1qxPSijiWznCKpCFTpkxh3LhxSb52c+bMmaxcuZJq1apRu3bt5AknksqtXLmSkJAQPDw8mDlzJo899tjtbVeuXKF58+YEBARQvHhxNm7cqGZTxEVs2LCBSpUqMXToUIBYm83t27fz3HPPceTIEXr27MmsWbPUbEqaoYZTJI24du0aH330EbVq1cLHxyfR+12+fJmBAwdSsGBBli1blowJRVKvJUuW8NJLLzFixIh7tq1ZswYvLy+WL1/O6NGj2b9/v67lEnERmzZtokGDBhQoUICOHTvesz0qKoohQ4ZQvXp13N3dWblyZZIvaRFxdRpSK5JGTJgwgTNnzvD9998n6YPurbfe4vjx42zZsoW8efMmY0KR1MdayyeffMK4ceNo3LgxAwcOvGP70qVLadq0KenTp+fHH3+kbt26TkoqIkn1zTffMGjQILy8vFi5ciV58uS5Y/u1a9eoUaMGu3btokaNGixevFifo5Im6QynSBpw5coVPv30U55//nmee+65RO83aNAg5s6dS5s2bahatWoyJhRJfSIjI+nRowcffvghdevWZeHChbi7uwPRjeinn35K06ZNKVq0KOvWrVOzKeJCJkyYQPv27alYsSJr1669p9k8ePAg1apVY9euXTRo0IB169ap2ZQ0S2c4RdKAL774gvPnzzNs2LBE73PhwgUmTpxInjx5+PLLL5MxXeIYY7oDHYBywFxrbYc46iYBvjEeygCEW2s9bm73B6oBETe3n7DWlkye1JJWXbt2jdatW7N06VJ69+5No0aNyJAhAwDnzp3Dy8uLo0eP0rJlS6ZPn65ruURcyKeffkrfvn1p0qQJ3bp1u2em6d69e/PFF1/g4eHB6tWradCggZOSiqQMOsMpksr9888/fPbZZ7z00kv3rPUXn27duvHPP/+wdu1asmfPnowJE+0kMByYHl+RtfZNa222WzdgLvD9XWXdY9So2ZSH6ty5c/j4+PDDDz8wfvx4Ro0aRbp00R+3wcHB1KtXj6NHj9K4cWPmzZunZlPERVhr6devH3379uXVV19lwYIFZMyY8fb2Gzdu8O677zJmzBgyZcpEQECAmk0R1HCKpHpjx47ln3/+uT17XmL07NmT+fPn07dv33uWbnAWa+0ia+0S4EJi9zHGZAWaATOTLZhIDMeOHePZZ59lz549LFy4kB49etze9uWXX+Ll5cWRI0f44Ycf+OGHHzRxiIgLuXz5MvPmzaNz587MmjXr9qgFgB07dlC1alVGjx5N165dOXXqFKVLl3ZiWpGUQ0NqRVKxsLAwvvzyS5o1a0alSpUStc/p06eZMmUK2bJl48MPP0zmhMmuGXAO2HjX4x8bYz4B/gL6WWv9HR1MUqd8+fJRvnx5vvnmG5555hkg+qzI9OnTmTVrFnny5CEwMJASJUo4OamIJFZkZCTWWnLmzMmvv/5Knjx57viyyN/fn/r16xMZGcmsWbPw9fWN59lE0h41nCKpmLu7O4GBgYk+i2KtpXPnzkRFRREYGHh7ghMX1h74xlprYzz2PvAHEA68AiwzxlS01t6zOKkxpgvQBcDT0xN/f/94DxYSEpJgTUqjzA9HYGAgpUqVwsPDgx49enD9+nX8/f0JCwtj3LhxrF69muLFi/PRRx9x8uRJTp486ezICUqJv2cRR4uKiqJ169a4u7szY8aMOyb+iYyMZM6cOUyfPp2CBQvy5Zdf8tJLLzkxrUjKpIZTJJWKjIzEzc2NJ554ItH79OjRg+XLlzN27FjKlCmTjOmSnzGmEFAL6BzzcWvtrzHuzjTGvAo0Ar64+zmstX6AH4CXl5etXbt2vMf09/cnoZqURpkf3OnTp2nUqBGdOnVi/Pjxtx//8ccf+e9//0toaCht27ZlxowZt6/ldAUp7fcs4gzp0qWjbNmyZM6c+Y4vb8+dO0eFChU4deoULVu2ZOrUqXh4eDgxqUjKpYZTJJXq06cPhw4dYuHChbi5uSVYf/jwYSZOnEjOnDnvuO7MhbUDtlhrgxKos4AupJMks9ZijCF//vysXLnyjkm5Zs+eTefOnQkPD+fTTz/F29vbpZpNkbQuJCSEw4cPU65cOQYMGHDHtoMHD/Lyyy9z6tQpatSowbx583Q9tkg89OknkkoVKFCAokWLJqrZjIqKolOnTmTOnJlVq1Ylah9HM8akN8a4A26AmzHG3RgT35dm7YAZdz1HTmNMg1v7GmPaADWBH5MtuKRKYWFhtG7dmjlz5gBQu3ZtsmTJwoULF6hZsya+vr5UrlyZoKAg+vTp4+S0IpIUly5don79+tSrV4+QkJA7tg0aNIjKlStz9uxZfv75Z4YNG6ZmUyQBajjjYIz51hhzyhjzrzHmgDGmk7MziSRFr169+PzzzxNV++GHH7Ju3TrGjh1LtWrVkjnZfesPhAJ9iV5nMxTob4wpbIwJMcYUvlVojHkGKMi9y6FkIHpplXPAeaAH8B9r7V8OyC+pxMWLF2nQoAHz5s3j9OnTtx/fv38/5cuXJyAggPbt27N+/XoKFy4czzNJYhhjuhtjthtjrhtjZsRT194Ys+Pm5/ZxY8zIBL6UErnH2bNnqVOnDjt37mTy5Mm319iMiorigw8+YOjQoWTMmJHt27dTt25dJ6cVcQ1qOOP2MVDUWpsdaAIMN8ZUdnImkQQdPXqUefPmERUVlaj6gIAAPv30Ux5//HE6dUq536tYawdba81dt8HW2qM319M8GqP2F2ttVmvtlbue45y1toq11sNam9NaW81a+5PjX424quDgYGrUqMHWrVuZO3cuvXr1AmDy5Ml4eXkRFhbG2LFjmTFjBunTq9d5SBK1Bi+QBegJ5AWqAnWBd5M3mqQmx48fp2bNmhw4cIDly5fzn//8B4j+kql58+Z88sknPP/88xw4cICiRYs6N6yIC9GnYRystfti3r15ewLY4ZxEIokzfPhwZs6cSY0aNShYsGC8tZGRkfTp04csWbLw3XffaViQSDx27tzJiy++SFhYGGvWrKFWrVqEhoby7LPPsnv3bqpVq8b333+f4PtOksZauwjAGONF9MiFuOomxrh7whgzG6iTzPEklTh06BD16tXj4sWLrFmzhho1agCwbt06XnjhBSIiIhg7dizvvPOOPitFkkgNZzyMMROADkBmYBewMpYaLZuQAqXVzCdPnmT69Ok0adKEv//+m7///jve+m+//ZatW7fSr18/rly5kuTju+LvWeR+rFq1ihYtWpAnTx7Wrl1L6dKl2b9/Py1btuT333/n2WefZe3atalhKaHUpCawL66N+vxOmZyROTg4mHfffZcbN24wcuRIIiIi8Pf355dffmH48OFERUXRtWtXKlasyIYNG5ye90G5YmZxcdZa3eK5ET1BSQ2irx/LEF9t5cqVbULWr1+fYE1Ko8yO8TAyt2/f3rq7u9uTJ08mWLtw4UIL2Fq1atmoqKj7Ol5iMgPbbQp4Lz/oTe/vlMPRmf38/Kybm5utWLGiPXHihLXW2jfffNNmzJjR5smTx65atSrB50itv2dHvr+JHlY7I5G1rwHHgbyJqdf7O+VwdOYdO3bYPHny2Pz589u9e/daa629ceOGfemllyxgn376aXvkyJE490+tv+PU8tmtW8q46RrOBFhrI621m4gexvOWs/OIxGX//v3MmjWLbt268eijj8ZbGx4eTv/+/cmYMSMTJkzQ8CCROJw8eZKePXtSr149Nm7cSK5cuejUqROTJk0ic+bM7Ny5k4YNGzo7psRgjPkP8AnwgrX2vLPzSMq2ZcsWsmXLRkBAAGXKlOHff/+lTp06LF++nOrVq7Np0yZN/iXygNRwJl56oq/hFEmRhgwZQubMmXn//fcTrP3oo4/4888/mT9/PqVLl3ZAOhHXcmvSrccee4yAgACWLVtGQEAATz/9NNOmTaN3796cPn1af4imMMaYhsAUoLG19ndn55GU69ZyJ927d+e3337jySefZM+ePXh7e/PLL7/wv//9j40bN5I5c2YnJxVxfWo4Y2GMecQY84oxJpsxxs0Y0wB4FVjn7Gwisdm7dy/z58/n7bffJl++fPHWzpo1i2HDhtG6devbM/CJyP8LCQmhQYMGTJ48GYCnn36ab775hhdffJGgoCBWrVrFqFGjdL2mgyR2DV5jjA8wG2hmrQ10dE5xHT/++CPFihVj165dAGTPnp1hw4ZRqVIlzp49y9q1axkzZgzp0unPZJGHQe+k2Fmih88eBy4Bo4Ce1tqlTk0lEodBgwbh4eHBu+/GvwJAaGgo7777LunSpWPEiBEOSifiWtzd3cmaNSvu7u6EhITQuXNnOnXqROnSpdm8ebOG0DpeYtfgHQDkAFbefDzEGLPKOZElJStbtiw+Pj4UKVKEqKgoBg8ezMCBA8mRIwerV6+mVq1azo4okqpoltpYWGvPAfrXRlzC7t27WbRoEYMHDyZ37tzx1g4YMICzZ8+yaNEiihQp4qCEIq5h79695MmTh0cffZTFixezfv168uXLR1hY2O0F37W2puNZawcDg+PYnC1GnZZAkXj99NNP+Pj4UKBAAebPn8+RI0eoWbMm+/bto0OHDkycOFEjF0SSgc5wiri4MmXK4OfnR8+ePeOtmzJlCqNHj+att96iadOmDkon4hrWrVtH9erVeeONNwCYM2cOjRs3JjIykhEjRjBixAg1myIubPTo0Tz//PN89dVXAPz5559UqVKFffv28eGHHzJ9+nQ1myLJRJ+eIi4uQ4YMdO7cOd6aK1eu0KNHDzJkyMAnn3zioGQirmHWrFl07NiRkiVLMnz4cKpUqcL27dt57rnnmDt3LgUKFHB2RBG5T9ZahgwZwpAhQ2jevDlvvvkmc+fOpUuXLmTOnJmvv/6aDh06ODumSKqmM5wiLqx9+/bMnDkzwbo+ffpw/fp1Jk2aRPbs2R2QTCTls9YyYsQI2rVrR40aNZg2bRpNmjRh+/bttGrVinXr1qnZFHFh1lreffddhgwZQvv27Zk9ezaNGjWidevWlChRgp07d6rZFHEAneEUcVEhISEEBwdz/nz8y8zNmzePSZMm0bt3b15//XUHpRNJ2SIiIujWrRt+fn74+vpSsWJFfHx8yJw5M35+fgmOGhCRlC0yMpK33nqLKVOm0K1bN4YOHUrTpk1Zu3YtTz31FOvWrSNHjhzOjimSJqjhFHFR2bJlw9/f//Z6gbE5cuQIvr6+5MmTh+HDhzswnUjKFRISwiuvvMKKFSt477338Pf359tvv6VKlSosXrxYZzVFXNyNGzfo0KEDc+bMoW/fvlStWpVSpUpx8eJFJkyYwJtvvokxxtkxRdIMNZwiLmjfvn3kypWLxx57DDc3tzjr+vfvT1RUFKNHj9ZkCCLAP//8Q7169di1axdDhgxhwYIF/P777zz33HOsWbNG7xMRF3f9+nVatWrF0qVLGTFiBCVLlqRFixZERUXx448/Uq9ePWdHFElz1HCKuBhrLZ06deLixYvs378/zm9plyxZwrfffsvAgQNp3769g1OKpEzZs2enfPnyPPLIIwwbNoycOXOyevVqGjRo4OxoIvIQXLlyhYMHDzJmzBhOnDhBs2bN8PLyYtKkSVSuXNnZ8UTSJDWcIi5m1apVbN26lcmTJ8fZbP7555+0aNGCkiVL0q9fPwcnFEl5Nm3aRKFChciXLx/p0qVj1apVZM+enV9++YUnn3zS2fFE5AH9+++/uLu7kzdvXn766ScqVarE2bNnef3115kwYQKZMmVydkSRNCtFzFJrjAk2xthYbvti1HSIo6ZejJoKxpjtxpirxpj1xpjCdx1ntDFmoiNfm8jDZK1lwIABFCtWjNdeey3Omu7duxMREcHHH39MxowZHZxSJGUJDQ2lRYsWvPLKK1SoUIFp06bxwQcfcObMGTWbIqlAeHg4Pj4+dOzYkb1791KzZk0uXLiAr68v06ZNU7Mp4mQp5QxnFSDmhWhZgd+BeXfVRQIF73rsYoyfpwGbgVeAT4HRQAsAY0xVoClQ4aGlFnGwJUuWsHPnTr7++msyZMgQa83cuXNZt24dH3/8MU2bNnVwQpGUw1oLQObMmXnjjTcYMmQI6dOnZ9WqVTRs2NDJ6UTkYcmYMSO+vr74+/vj5eVFrly52LhxI88++6yzo4kI99lwGmP8gSDgLNAJyAh8B/Sw1oYm9fmstefuev7OQAaiG8i7a0/H81SlAF9r7d/GmKnAqJvPl/Hmc71lrb2S1HwiKUFUVBQDBw6kRIkS+Pr6xlqzfft2OnbsSLVq1XjvvfccnFAk5YiMjKRXr17kzJmT48ePM336dCpVqoSfnx9eXl7OjiciD0FwcDCnTp3C29ubkydPsnTpUnLlysWOHTt47LHHnB1PRG56kDOczYH5wHPAk0Q3dNeAt40xHwIfJrD/CGvtiDi2vQEss9aevOtxN2NMEJAZ+AsYZa1dHmP7HuAFY8wB4AVg183HBwGB1tofE/fSRFKe77//nr179zJnzhzSp7/3rWutpXXr1oSFhfHRRx/FO3utSGoWGhpKmzZtWLx4MW5ubkRGRtKvXz8GDx4c63tHRFzP/v37qVevHsYYSpYsydq1a/H19WXChAl4eHg4O56IxPAgn7wXgTettZHAn8aY/sAXxpgPgElEn/FMaP97GGO8gMrA3TOd/AW0B34juuFsBSwzxnSy1t46E/o6MAH4H7ATeNMYUwnwBbyMMWOAJsAx4A1r7YGkvGARZ4mIiGDQoEGUKVOGVq1axVozdepUDh48yIABA/Dx8XFwQpGU4fz58zRp0oStW7eSIUMG0qVLx9ChQ/nww4S+AxURV7Fnzx7q169PREQE//77LydPnsTPz4/OnTs7O5qIxOJBGs7Am83mLZuJHlr7hLX2N+JoKBPhDeAwsCbmg9baX4BfYjz0izEmN/A+N4feWmv3A7f/0jbGpAdWAd2JvpazMlCG6GHAs4Cq95lRxKHmzJnDX3/9xYIFC0iX7t65vrZs2ULPnj2pW7cugwcPdnxAkRTg0KFDPP/88xw+fBhrLc888wxz5syhQIECzo4mIg/J1q1beeGFF3BzcyM0NJQsWbIwduxYOnbs6OxoIhKHhzlL7e31GYwxHxpjQhK43fN1szEmO/Aq4GdvzfYQvy1A0Xi2fwD8Ya1dBtQDvrPWXgdmAt7GGI25EJfg5eVF7969Y50EKCoqipdeeonQ0FD8/PxibUhTA2NM95uzUF83xsyIp66DMSbyrn9vasfYXvTmLNbXjDH7Y850La4rMDAQb29vgoODsdbSpEkT1q5dq2ZTJBVZv3499erVIzQ0lAsXLlC5cmX++usvNZsiKdyDnOGsYoxxi3GW8xkgHDjE/Q+p9SX6LOnXicxQiejhsfcwxpQBOt+sgejm+ta0nhljPCaS4pUuXZpRo0bFum38+PFcunSJHj168Pjjjzs4mUOdBIYDDYgeVh+fX6y1NeLYNpfo0RKNbt4WGGOK3z15mbiOH374gebNmxMREUHOnDn58ssvad26tbNjichDtGLFCpo3b066dOm4fv06VatWZd26dVr6S8QFPEjDlQf4yhhTyhjzIjAMmGKtvWqtvWit/TuBW2wN5xvAEmvtmbs3GGMGG2MaGWOeNMaUMcYMInpo7JhYat2A6UAva+2Fmw9vBF6/2Yj2BfZYay8/wOsXSXbXr1+nW7duHDx4MNbtv/32Gx988AEvvfQS48aNc3A6x7LWLrLWLgEuJFgcB2NMCeBpYJC1NtRau5DoJZiaPaSY4mDnz5+nRYsW3Lhxg9KlS/P777+r2RRJZb7//ntefvlloqKiuHHjBqNHj2br1q1qNkVcxIM0nAuAK8AmotfLXAn0ud8nM8ZUA8oDk+MoyQ58RfQfhwFEn+Voaa39Kpba/wHHrLULYjw2gehZa7cA9YF295tVxFF2797NzJkzCQoKumdbWFjY7TXGpkyZgjHmnpo0rJIx5rwx5oAxZsDN67kh+hruoLuWR9pz83FxIVFRUfzxxx/UqVOH8PBw6tSpw/bt2zWEViQVmj17NpGRkWTJkoVNmzbRq1cvZ0cSkSR4kCG1Udba94CHstiftXYrMa4DjWV7LyBR/8JYa+8Ze2itDSN6llsRl1G1alWOHj1Krly57tn2ySefcPXqVd555x3y58/vhHQp1kagLHCE6EZyPhABfAxkA+4e2XAZiLVLMcZ0AboAeHp64u/vH++BQ0JCEqxJaVwx8+XLlylcuDAnTpwgR44cjBw5kipVqrB161ZnR4uTK/6eXTGzpC7BwcF89dVXLF26lAIFCrBq1SrKlSvn7FgikkRakEwkhTpy5AiFCxcmd+7c92zbvXs3I0aMoFWrVnz++edOSJdyWWtjng7+3RgzlOgvxj4GQogeLRFTdqJHa8T2XH6AH4CXl5etXbt2vMf29/cnoZqUxtUyX7t2jebNm3PixAnc3d3ZvHkzZcqk/BPUrvZ7BtfMLKnHRx99xIABA7DW0r17d0aPHq0htCIuSpPmiKRAV69exdvbm3feeeeebf/++y8+Pj7kyJGDr76KbUS53MXy/6Mn9gGP3zVDdYWbj/8fe/cZHVX1vn38uxNK6CJgUFCC0nsJAQV+VGlSpCkqKE1EqggoCkpVBAUBEaQkNCnCHwMIEVBC6JEqKCC9ShcIhJ5kPy8CeSihJzkzyfVZi7Vm5txncs2YY+aes8/e4uImT55Mnjx5WLhwIT179iQsLMwtmk15PA86Q/X12i7GmGPGmDBjTIAxJmUCxZQ49OeffzJs2DCstXz44Yd89913ajZF3NgjNZzW2orW2tZxHUZEoo0cOZITJ07QpEmTO7Z9+OGHnDlzhvfff59MmTI5kM4ZxphkxhgvwBPwNMZ43XRt5s11NY0x3tdv5wM+A+YCWGt3An8Cva/vX5/oa8dnJ9TrkEczZMgQ3nnnHf79918GDBjAgAED9AE06bgxQ3XAvYqMMdWJnhSwCtFLpj0P9I3vcBJ3oqKi6NOnDy+++CJeXl789ttvDBkyxOlYIvKYNKRWxMWcO3eOwYMHU6NGjZhJgW4IDQ1lwoQJNG3alH79+jmU0DG9gN433W8K9DXGBADbgALW2oNEf9icaIxJCxwHfgS+vGm/JsBE4AxwEGikJVFc17lz52jevDmBgYEkS5aMiRMnamKgJMZa+zOAMcYXyH6P0ncAf2vt1uv1/YGpRDeh4uIuXbpErly5OHLkCDly5GDt2rU89dRTTscSkTighlPExQwfPpzTp0/Tv3//Wx4/deoUDRs2JHv27ElyKK21tg/Q5y6b095U1w3odo/n2Q9UjLtkEl927txJkSJFuHLlCk888QShoaHkzZtXE9nI3RTk+miG6zYD3saYTDctkRZDk4K5joMHD9K+fXvCw8PJlCkTo0ePZtu2bWzbts3paPflLu/xzdwxs7g3NZwiLuTMmTMMGTKEevXq4evre8u2Bg0acOTIEcaOHUv69LfPeyOSuCxYsIDXXnuNK1eu4O3tzZYtW3S2Q+7n9lmob9xORyzr92pSMNfQrVs3hg4dirWWli1b0qxZM5fPfDN3eI9v546Zxb1p0iARFzJkyBDCwsLuGC67dOlSVqxYQb169Xj33XcdSicS/y5fvky5cuWoXbs2Fy9e5KWXXmLnzp1qNuVB3D4L9Y3bsc5CLc66evUqrVq1YsiQIVhrGTRoEP7+/k7HEpF4oIZTxEWEhYUxfKFxhokAACAASURBVPhwGjduTJEiRWIeP3LkCO+88w65c+dm2rRpDiYUiV9Hjx6lUqVKrFq1CoCGDRsSHBysM/ryoLYSPev0DUWB47ENpxVnrVy5knz58hEQEEDKlClZvHgxH330kdOxRCSeaEitiIuYPn06Fy9epG/fWydVrFy5MocOHSI4OJjUqVM7lE4kfk2cOJGPP/6Y8PBwKleuTP78+Rk+fDienp5ORxOHXZ+NOhk3zVANRFhrI24rnUz0hGFTgaNETzQ2MSGzyv3NmTOHhg0bEhUVRZYsWVixYgV58+Z1OpaIxCM1nCIu4umnn6ZTp07kz58/5rGgoCB27NhBpUqVqFSpkoPpROJHVFQU7dq1Y8yYMWTJkoW1a9dSoEABAIwx99lbkogHmqHaWrvQGDMYWAqkInq5o953PJs44ty5c3z66ad8//335MuXj2zZsjFlyhSefvppp6OJSDxTwyniIurVq3fLRfynT5+mdevWFCpUiF9//dW5YCLx5L///uPtt98mKCiIDBkykCxZMnLmzKlGU27xoDNUX68dCgyN50jykP7++2/8/Py4dOkSHTp0YMiQIVpHVyQJUcMZC2NMSmAUUBV4EtgNfGqt1ad+iXOjRo3i2rVrFC1a9JbHS5UqxfHjx1mwYAEpU6Z0KJ1I/Jg0aRLvvfceUVFRjBo1igoVKrBz504NGxdJZGbNmkXr1q2JiIgeAZ0/f341myJJjBrO2CUDDgEViF4YvhYw0xhT+PoafiJxwlpLcHAwERERFC5cOObxadOmsXfvXipXrkzx4sUdTCgSt6y1fP/993zwwQcAvPrqq7Rt2xZjTMxQWhFxf2fPnqVixYps3ryZMmXKMGPGDA4fPsxLL73kdDQRSWCapTYW1toL1to+1tr91tooa+18YB9Q0ulskjhYawkLC8MYw9SpU5k1axYeHtGH4/Hjx+nUqRMlS5ZkwYIFDicViTtHjx7lf//7Hx07dqRMmTKkT5+elStXcuzYMaejiUgc2rVrF2XLlmXz5s1kyJCB3r17kyNHDsqWLash8yJJkM5wPgBjjDeQh+gp12/f1gZoA+Dt7U1ISMg9nys8PPy+Na5GmeNWVFQUI0aM4M8//+T7778nTZo0QHTm4OBgWrZsSVhYGB06dCA0NNThtPfmyu+zuJa//vqLChUqcObMGV5//XXmzZvHM888w+LFizVpiEgi0r9/fwYPHkyyZMnw9vYmLCyMq1evOh1LRBykhvM+jDHJganAJGvtP7dvt9aOBcYC+Pr62psnfYlNSEgI96txNcocdyIiImjZsiVz586lW7du1KpVK+bb3pCQEFavXs2BAweoV68ezZs3dzbsA3DV91lcy/jx4+nUqRPp06enadOmzJgxg8KFC/Prr7/i7e3tdDwRiQOXLl2ibt26/P777/j4+HD+/HkiIiIIDg7mxRdfdDqeiDhIDec9GGM8gCnAVaCDw3HEzV25coU33niDwMBA+vfvT8+ePW8ZWnTixAkGDRqEr68vM2fOdDCpSNy4dOkSpUqVYuvWrVSsWJEKFSrQt29fKleuTGBgIOnTp3c6oojEga1bt9KkSRP+/vtvihcvzo4dO3jqqadYuHCh1tgUEV3DeTcmuhPwB7yBhtbaaw5HEjd24cIF6tSpQ2BgIMOHD6dXr163NJtRUVH06dOHyMhIZsyYoRn8xO3t2rWLF198ka1bt1KmTBlKlChB3759ady4MUFBQWo2RRKJjh07UqRIEY4dO0aXLl3YsmULefPmZc2aNWo2RQRQw3kvo4H8QB1r7SWnw4j7Onv2LNWqVWPJkiUEBATQqVOnO2qaN2/O9u3bef/993nhhRccSCkSdz7++GOKFi3KoUOHCAoKokqVKgwdOpT27dszffp0LfMjkghcuHCBFi1aMHLkSNKmTUvjxo359ttvqVKlCsuWLSNr1qxORxQRF6EhtbEwxuQA3gOuAMduOhP1nrV2qmPBxO2cPHmSatWqsXXrVn766ScaNWp0y/aoqCi++eYbpk6dynPPPcegQYMcSiry+K5evUr37t0ZMWIEGTJkYNOmTTz33HPky5ePjBkz8uGHH2qGSpFEYO7cubRr146jR4/y+eefU7duXUqXLs3bb7/N+PHjSZ48udMRRcSFqOGMhbX2AKBPRfLYhg0bxo4dO5g3bx41atS4Zdu+ffuoUqUK+/bt47XXXuOdd96JWRpFxN2sX7+e9u3bs3btWlq2bMnnn3/Ojz/+SI8ePciZMyddu3Z1OqKIPCZrLQEBAbRp0waAX375hVdeeQWANWvW4Ovrqy+VROQO+nQrEg+stQD07duX0NDQO5rNDRs28OKLL7Jv3z7at2/PjBkzSJ06tRNRRR7bTz/9hJ+fH5s2bWLmzJn4+/sTFBRE3759+euvv5yOJyJx4OjRo7z++uu0bt2acuXKERwczBdffMHcuXMBKFWqlJpNEYmVGk6ROPb3339ToUIFjh49SrJkyShSpEjMtqioKHr37s1LL71EypQpmT9/PiNHjtQfaXFLkZGRfP7557zxxht4e3szd+5c6tevD0Dbtm3566+/KFq0qMMpReRxbdiwAR8fH2bNmkW/fv0IDg7G19dX12OLyANRwykSx8LDwzl+/Dhnz56943FfX1/69etHmTJl2LhxY8xQJBF3s23bNp555hn69+9P8+bN2bNnDxkzZiR//vxs3boVYwx58uRxOqaIPAZrLWPGjKFs2bKkSpWKoUOHUqdOHS5cuECaNGkIDg6mXr16TscUERenhlMkjhw6dAiAMmXKsHXrVvLnzx+zbdu2bfj5+fHnn39Ss2ZNlixZQqZMmZyKKvJYVqxYQaVKlThx4gTt2rUjICCAkJAQKleujLWWVKlSOR1RRB7TkSNHyJkzJ23btqVixYrs3LmTvHnzUrZsWbp06QKg0Tki8kDUcIrEgaCgIPLkycPkyZMBSJbs/8/H1b59e4oUKcJ///3HkiVLCAoKumW7iLuIioqidevWVKpUiQwZMhAaGsr333/PlClTqFu3Lvny5WPVqlU8//zzTkcVkcewceNGypUrx4EDB6hbty5BQUHMnz+funXrkjdvXr744gunI4qIG1HDKfKYZs2aRb169cifPz81a9aMefzy5cu0bduWUaNGkSZNGlauXEmlSpUcTCry6M6cOUOpUqXw9/fnxRdfZP369ZQuXZohQ4bw9ttvU6FCBUJCQvD29nY6qog8ohtfKpUpU4Zr166xZMkS5syZw4ABA2jVqpXW2BSRR6KGU+QxBAQE0KRJE0qXLs3SpUvJkiULACEhIRQrVowxY8bw0Ucfcfz4cXLnzu1wWvdmjOlgjFlvjLlijJl4j7p3jDEbjDHnjDGHjTGDjTHJbtoeYoy5bIwJv/5vR4K8ADe2du1aSpQowZYtW3j77bdZtmwZ6dKl46OPPqJbt240btyYoKAg0qdP73RUEXlEZ8+epWLFivj7+1OgQAE2bdrE//73P9577z169+7N22+/zfz580mXLp3TUUXEzajhFHlEw4YNo1WrVlStWpVFixaRIUMGAObMmUPVqlXZtWsXc+fOZdCgQXh5eTmcNlE4AgwAAu5Tlxr4AMgMlAaqAN1uq+lgrU17/V/eOE+aSFhradWqVczZjhUrVjBp0iQiIyNp0aIFX3/9Ne3atWP69OmarVLEjS1fvpzixYuzZs0aWrZsyfr160mVKhX169dn3Lhx9OzZk4kTJ5I8eXKno4qIG1LDKfKQrLX079+fLl260KBBA+bNm0eaNGm4ePEiXbt2pX79+uTNm5elS5dSt25dp+MmGtban621c4D/7lM32lq7wlp71Vr7LzAVKJsgIROR8PBwmjZtSkBAAJkyZSI4OJgyZcoA0dcsT5o0iX79+jFy5Eg8PT0dTisij8JayzvvvEOFChVivlTy9/fn3LlzVK5cmaCgIEaPHs2AAQM0QZCIPDLNXCLyEKy1dO/ePea6NX9/f5IlS8auXbvw9fXl3LlztGvXjqFDh+qMj+v4H7D1tscGGmO+AnYAPa21IbHtaIxpA7QB8Pb2JiQk1rIY4eHh961xNbFlDg0NZciQIZw+fZpWrVrxxhtvcOTIEf7991+MMWTIkIHRo0eTL18+li1b5hKZXZ0yi6s5ffo0zZs355dffiFbtmwsW7aMF154AYA0adKQOXNmfv75Zy17IiKPTQ2nyEM4ePAg48ePp0OHDgwfPhwPDw9+//133njjDS5evEj79u0ZOXKk0zHlOmNMC8AXaH3Twx8D24CrQBPgF2NMMWvtntv3t9aOBcYC+Pr62ooVK97z54WEhHC/Gldze+Yff/yRXr16Ya1l3rx5MWvFHj58mEaNGjFy5Eh8fX0dfZ2J4X12B66S2RjzJOAPVANOAZ9Ya6fFUpcSGA7UB5IDq4C210c6yE3Gjx9Pt27duHjxIsOGDaNTp04YY9i4cSM5cuQgU6ZMzJ8/X2c1RSROaEityAOIjIwEIEeOHGzevJkRI0YQFRVF9erVefnll3nqqafYsmWLmk0XYox5FfgKqGmtPXXjcWvtH9ba89baK9baSUR/KK3lVE5Xce7cOVq0aEGzZs0oVaoUoaGhMc0mgIeHB5cuXSI8PNzBlJJEfU/0F0TewFvAaGNMwVjqOgMvAkWAZ4CzwHcJFdIdREVFMXjwYNq0acOFCxdYuHAhnTt3xhjD+fPnqVatGh07dgS0xqaIxB2d4RS5jytXrtC4cWNefPFFPvnkE3LkyMGpU6do2LAhy5cvp1ixYqxcuZI0adI4HVWuM8bUAMYBr1hr/7pPuQWS9Cer3bt3U7x4ccLDw/noo4/44osvYtaK3bZtG3ny5OGZZ55h06ZNeHjoe0pJOMaYNEBDoJC1NhxYaYyZBzQDetxWnhNYZK09fn3fGcDQhMzrynbt2sW7777LsmXLqF+/PkOHDsXHxydme7p06Zg6dSpFixZ1LqSIJEr65CByH8mTJ+eJJ57giSeeAKJnoS1evDh//PEHAwYMYMOGDWo2E4AxJpkxxgvwBDyNMV43L3dyU11loicKamitXXvbtieMMdVv7GuMeYvoazwXJcRrcEUrV67E19eXyMhIPv74YwYNGhTTbP7666+UKlWKvn37AqjZFCfkASKttTtvemwzENsZTn+grDHmGWNMaqLPhv6aABld3vLlyylYsCDLly9n5MiRzJ49Gx8fH6y19O3blylTpgBQvXp1rbEpInFOZzhF7uL06dNcuHCBZ599lkmTJgHQvn17Ro0ahbe3N6tXr6ZEiRIOp0xSegG9b7rfFOhrjAkg+prMAtbag8BnQAYg6KYhYSustTWJvq5rAJAPiAT+AV611ia5tTgvX75MpUqVCA0NpWTJksyaNYucOXPGbP/xxx9p0aIFhQsXpkOHDg4mlSQuLRB222NhQGyLQe4EDgL/En18/wXc9Zc3KUwKFhYWRqtWrZg4cSJPPvkkbdq0oWDBgixbtozIyEi+/fZbFixYwCuvvMKzzz7rdFzA/d5nd8sL7plZ3JsaTpFYHD9+nGrVqmGtZdOmTYSHh9OqVStmz55N7ty5Wbx48S1DkST+WWv7AH3usjntTXWV7vEcJ4FScRrMDR06dIhGjRqxdu1aXnjhBVatWnXLrMpDhw6la9euVK5cmcDAQNKnT+9gWkniwoHbfwHTA+djqR0NeAGZgAvAR0Sf4Swd2xMn9knB9uzZw6uvvkpYWBhNmjRhzJgxMcfyhQsXeP3111mwYAG9evWiX79+LnPNpru9z+6WF9wzs7g3jY8Suc3BgwcpX748u3fvZujQocyePZvs2bMTGBjIN998w44dO9RsitsaOnQoRYsWZfv27fj7+zN+/PiYZtNay0cffUTXrl1p1KgRQUFBajbFaTuBZMaY3Dc9VpQ7lzq68fhEa+1pa+0VoicM8jPGZE6AnC5l0aJFlC1blnPnzvHWW28xbdq0mGP5xIkTVKpUiV9//ZUffviB/v37u0yzKSKJkxpOkZvs3LmTcuXKceLECX777TcOHjxI06ZNuXjxIpMmTaJr1676wyxu6dq1a7Rq1YquXbvi5eXFhg0baNmy5S3bW7Rowddff83777/PjBkztJasOM5aewH4GehnjEljjCkL1AOmxFK+DnjbGJPBGJMcaAccuXmW6sTu9OnTlC5dmho1avDkk08yduxYfvzxx5i/W7t37+all17ir7/+4ueff+a9995zOLGIJAVqOEWu27JlC+XLl+fy5csEBgYyYMAAWrVqxf/+9z/27t1L06ZNnY4o8kgOHjxIpUqVCAgIoGLFimzZsoXcuf//CaOLFy9Sv359Jk2aRJ8+ffj+++/x9PR0MLHILdoBqYATwHTgfWvtVmNMeWPMzev0dAMuA7uAk0Qvd1Q/ocM6Zf369ZQsWZK1a9dSrVo1NmzYQK5cuWK2r1u3jpdeeomzZ88SHBxMvXr1HEwrIkmJruEUAUJDQ6lZsyZp06Zl7NixNGrUiNOnT9OtWze++uorffgWt/Xtt9/SvXt3UqZMybRp03jjjTfuqLl06RL79u1j9OjRtG3b1oGUIndnrT0NvBrL4yu49frt/4iemTZJuXLlCh07dmTChAl4e3szefJkmjVrdkvNxYsXqV27NmnSpGHhwoXkzZvXobQikhSp4ZQkLzg4mLp165I1a1Y6d+7Ma6+9RvLkyRkwYAA9e/Z0Op7II4mIiODzzz9n4MCBpEqVilmzZlGrVq1bao4ePcq1a9fIlCkTmzZtIkWKFA6lFZFHsXv3bipVqsThw4epXbs2kydPJmPGjHfUpU6dmpkzZ5I3b14teyIiCU5DaiXJCwwMxMfHBy8vLzp16kShQoXYvHmzmk1xW5s3byZfvnwMHDiQd999l1OnTt3RbIaHh/Piiy8yfPhwADWbIm4kKiqKkSNHUqxYMc6dO0f79u2ZN2/eLc2mtZZJkyYxatQoACpUqKBmU0QcoTOckmRdunSJVKlS0aVLF1avXs3GjRspXrw4wcHBpEqVyul4Io9k0aJFNGjQgIsXLzJ06FC6dOkSa13atGnp3r27Gk0RN3P8+HF8fX05fPgwlStXZuLEibGuoRkVFcXOnTvx8PDAWqsJ70TEMTrDKUnSxIkTKViwIN26daN48eLs3r2bWbNmsXHjRjWb4pauXr3KBx98QM2aNcmZMyfz58+PtdkMDAxk2bJlALRv3/6WyYNExLXNnTuXQoUKcezYMerWrcuiRYvuaDbPnDnDkSNH8PT0pHfv3owfP17Npog4Smc4JUkqUKAAXl5eDBkyhPTp098xm5+IOzly5AglSpTg+PHjvPPOO4waNYrUqVPfUTdq1Cg6dOhA9erVqVChggNJReRRnDhxggoVKvDPP/9QrFgxli1bRoECBe6o27lzJ3Xq1CFjxoysWbOGFClSqNkUEcfpDKckGdZalixZwrZt2/j444/Zvn07jRo1Yt++fWo2xW399ttvFCtWjLNnz9K8eXMmTJhwR7NpreXTTz+lffv2vPLKK8yePduhtCLysFatWoWfnx///PMPFSpU4I8//oi12VyyZAmlS5fm9OnTfP3112o0RcRlqOGUJCEqKooPPviAqlWrUqhQIVauXMmECROYNWsWTz75pNPxRB7atWvXqFKlCtWqVSNLlixs3LiRCRMm3PEh89q1a7Ro0SJmAqHAwMBYz36KiGu5ePEi1atXp3z58nh4ePD7778TEhIS63XXo0ePpnr16mTLlo21a9dSvnx5BxKLiMROQ2ol0YuMjKR169ZMnDgRgAwZMjB+/HgaNmzobDCRR3T06FHefPNNQkJCyJUrF2vXriVNmjR31IWHh9OoUSMWLVpE3759+eyzz3TWQ8QNbNu2jXr16sUsezJ37lzSpUt3R11ERAQffPAB33//Pa+88grTpk0jffr0DiQWEbk7neGURO3q1avUrVs3ptls3LgxBw4cULMpbmvEiBEULVqUP/74gx9++IGdO3fG2mweP36cihUr8ttvvzFu3Dg+//xzNZsiLi4yMpKePXtSokQJzp49y+DBgwkODo612Txz5gy1atXi+++/p2vXrsydO1fNpoi4JJ3hvAtjTAegOVAYmG6tbe5oIHloFy9e5OWXX2b16tUAtGrVinHjxulDt7ilyMhIevTowTfffEPGjBlZt24dBQsWjLX2ypUrlC9fnsOHDzN37lxq166dwGlF5GEdPnyY8uXLs3//fqpUqcLUqVPx9vaOtfbUqVOUK1eOvXv34u/vT8uWLRM4rYjIg1PDeXdHgAFAdUDrZLiZ8+fPkzdvXg4fPsyTTz7JTz/9RNWqVZ2OJfJIDh06RPPmzQkODqZy5cpMmTKFZ5555q71KVOm5JNPPiF//vyUKVMmAZOKyKOYOnUqHTp04MqVK7zxxhtMmTIFT0/Pu9ZnypSJqlWr8tprr/G///0vAZOKiDw8NZx3Ya39GcAY4wtkdziOPIQDBw7w2muvcfnyZXLlykVoaCiZMmVyOpbIIwkICKBNmzZ4enri7+9PixYt7nqWPigoCGMMNWvWpEWLFgmcVEQe1unTp/Hz82PPnj2UKVOGKVOm3HPW9PHjx1OxYkVy5crFyJEjEzCpiMij0zWckmiEhobStm1bSpcuzeXLlyldujTbt29XsyluKTIykv79+9O6dWu8vLyYPHkyLVu2vGuzGRUVRd++fRk4cCDW2gROKyIPa8mSJRQpUoS9e/dSoUIFli9ffs9m89SpU3z88ceMGDEiAVOKiDw+neF8TMaYNkAbAG9vb0JCQu5ZHx4eft8aV+PKmSMiIlizZg3+/v4cOHAAAF9fX9q3b0/58uVZuXKlwwkfnCu/z3fjjpndwY4dO6hZsyb79u3jrbfe4ocffiBt2rSx1lpruXbtGilSpOCXX34hVapUuk5ZxIWFhYXx8ssvs27dOvLmzcvatWvx9fW9a314eDhp0qQhc+bMrFmzhhdeeCEB04qIPD41nI/JWjsWGAvg6+trK1aseM/6kJAQ7lfjalwx8+7du3nnnXf4448/iIyM5IknnsDHx4fRo0dTo0YNl8x8P8osAMuWLaNevXqEhYXxySef8MUXX9y1gYyMjKRjx44cOHCAuXPn8tRTTyVwWhF5GJs2beKNN95gx44dlClThiVLltxzXdxdu3ZRp04d3n77bT799FPy5MmTgGlFROKGhtSK2zh79izt2rUjf/785M6dm9WrV5M6dWqmTJnCyZMn2bdvHzVq1HA6psgjiYyMpGvXrlSuXJmnnnqK+fPn8+WXX9612bx06RKNGjVi9OjRFC5c+J4TjIiIs65du8abb76Jn58f586dY+bMmaxZs+aezebSpUspXbp0zIy0IiLuSmc478IYk4zo98cT8DTGeAER1toIZ5MlPZs3b+aHH35gwoQJXLlyBYAnnniCLl260KVLl1jXJxNxJydPnsTPz4/9+/dTv359Jk2adM/f69OnT1OnTh3WrFnDiBEj6NixYwKmFZGHsXfvXl555RX++ecfypYty9y5c+87t8CYMWPo0KEDefLk4ZdffuH5559PoLQiInFPZzjvrhdwCegBNL1+u5ejiZKQ8PBwxo4dS9q0aSlWrBg//PADV65cwcfHh59++omTJ0/y+eefq9lMQowxHYwx640xV4wxE+9T28UYc8wYE2aMCTDGpLxpm48xZqkx5qIx5h9jjKPr5axYsYJixYrx77//8vrrr/N///d/9/y9PnDgAGXLlmXDhg3MnDlTzaaIi7LW8u2331K0aFGOHDlC165dWb58+T2bzYiICDp37kzbtm15+eWXWbNmjZpNEXF7OsN5F9baPkAfh2MkKdZapk6dyrhx49i4cSPh4eEAGGOoW7cuPXr00JqCSdsDrY1rjKlO9BdFla/vEwj0vf4YwHRgDVDr+r//M8bkttaejL/od4qKiqJevXosWLCAXLlysWDBAooVK3bPfTZv3kzNmjW5ePEiixcv1vp7Ii7q5MmTlCtXjp07d1K2bFmmTZvGc889d899zp49S5MmTVi0aBFdunTh66+/1lB5EUkU1HCK4w4ePMi8efMYN24cW7ZsAaBp06aUKFGCQ4cO0alTJ3x8fJwNKY57iLVx3wH8rbVbr9f3B6YCPYwxeYASQDVr7SVgtjHmA6Ah8EN85r/ZqVOnaNasGQsXLiRbtmyEhoby5JNP3nOf4OBg6tevT/r06Vm5ciWFChVKoLQi8jB++eUXWrduzenTp6lZsyZz5swhRYoU99znxrDb3bt3M27cOFq3bp1AaUVE4p8aTnGEtZYVK1bQt29fgoODAShYsCDGGLp27crXX3/tcEJxYwWBuTfd3wx4G2MyXd+211p7/rbtBRMq3Pjx4+nVqxdnzpzhu+++o127dnh43PvqBmstX375Jc8++yy//vorzz77bAKlFXGeMeZJwB+oBpwCPrHWTrtLbQlgGNFfLF0AvrTWDk+InOfOnaNs2bL8/fffFClShN9//53ChQs/0L7JkyfH09OT3377TTN/i0iio4ZTEtT+/fvp0KEDW7Zs4dChQyRPnpwMGTIwefJk6taty6pVqyhdurTTMcW9pQXCbrp/43a6WLbd2J4ttieKy3V2o6KimDp1KgEBAaRMmZKRI0eSO3duli9ffs/nvHr1KilSpKBz584A7Nmzhz179txzn4fhjmupKnPCcKHM3wNXAW+gGLDAGLP5xiiGG4wxmYGFQBfg/4AU3Hs0RJwJDQ2lWbNm7N69m5IlS7Jy5Uq8vLzuu9/ChQupVq0azz77LFu2bLnvl08iIu5IDafEu4iICGbOnElgYCBz5swhIiKCVKlS4enpSWRkJNWqVaNWrVoAlC1b1uG0kgiEA+lvun/j9vlYtt3Yfp5YxNU6u4cPH+a9994jKCiIypUr4+/v/0DDxHv06EFoaCiLFi0iZcqU961/FO64lqoyJwxXyGyMSUP0kPdC1tpwYKUx5w01zwAAIABJREFUZh7QjP9/XfYNHwKLrLVTr9+/AmyPz3yXL1+mdu3aBAcH89xzz7F06dIHfs+WL19OzZo18ff3p2XLlmo2RSTRUsMp8ebw4cNMmDCBgQMHcunSJdKmTUvWrFk5fPgwnp6edOzYkU6dOpEzZ06no0rishUoCsy8fr8ocNxa+58xZivwvDEm3U3DaosCsQ7PiwuBgYE0btwYgJEjR9KuXbu7rq15uyJFinDp0iWSJ08eX/FEXF0eINJau/OmxzYDFWKpLQP8ZYxZDeQC/gDaW2sPxkewf/75h9dee42//vqLYsWKsWzZMtKnv/37rDtZazHGUL58eWbNmkX9+vXjI56IiMtQwylx6tq1awwbNoxhw4Zx9OhRrLVky5aNCxcucPbsWTJmzMg333xD69atyZAhg9NxxY08xNq4k4GJxpipwFGilzOaCGCt3WmM+RPobYzpBdQEihB9BiVORUVFMWTIEHr06EHq1KkZPnw4LVu2vO9+YWFhrFu3jqpVq/Lmm2/y5ptvxnU0EXdyt2Hwsa0dlJ3oazdfBv4CBhM9K3WsQ2cedch8VFQUo0ePZt68eXh5edG5c2fq1avHxo0b7/ti/v33XwYMGMBHH31Ezpw5yZw5MytWrLjvfo/KhYZFPzB3y+xuecE9M4t7U8MpcWLp0qX88ssvTJ8+nWPHjmGM4d133+W9996jXLlyFCpUiK5du9KgQQOdrZFH1QvofdP9pkBfY0wAsA0oYK09aK1daIwZDCwlevmU2bft14ToBvQMcBBoFNdLouzfv5/KlSuzb98+GjZsiL+//wN9wXLkyBFq1qzJnj172L9/P5kzZ47LWCLu6GGGwV8CAq216wCMMX2BU8aYDNba25vWRxoynydPHmrUqMFff/2Fn58fc+bM4emnn36gFxISEkKnTp0AyJUrF+XLl3+g/R6HKwyLfljultnd8oJ7Zhb3poZTHtnly5cJDAzku+++Y82aNRhjqF27Nl5eXhw+fJgffvgBYwxbt27Fx8fngYcRisTmPmvjpr2tdigw9C7Psx+oGHfJbhUaGsqrr77K8ePHadOmTcxxcD/bt2+nRo0anD59msDAQDWbItF2Asmur5W76/pjRYkeOn+7LYC96f6N23Hyx2fx4sU0bNiQixcv0qxZMyZMmPDA62SOGzeOdu3akStXLn755Rdy5coVF5FERNyCrlCXh7Z161ZKlSpFxowZefPNNzly5AhlypQhJCSEefPmUatWLXx9fbl69SoAOXPmVLMpid6NYXbly5cnderUBAUFMWbMmAf63V+1ahVly5blypUrLFu2jJdffjkBEou4PmvtBeBnoJ8xJo0xpixQD5gSS/kEoL4xppgxJjnwGbDSWnv2cTKEhYVRuHBhBg4cSM6cOdm8eTOTJ09+oGYzIiKCLl260KZNG6pUqUJoaKiaTRFJctRwygM5efIk3bp146WXXqJQoUJs2LCBrFmz0qlTJ7y8vAgNDWXnzug5HZo3b86IESPibVZNEVcTFhZGkSJFmDlzJhUqVGDjxo3UrFnzgfadM2cOVatWJXPmzKxevZoSJUrEc1oRt9OO6OHxJ4i+JvN9a+1WY0x5Y0z4jSJrbTDwKbDgem0u4LEvgt6/fz+7du2iYMGCMcNqH0RYWBh16tRh2LBhdO7cmfnz52vuAhFJkjSkVu7q5MmTbNmyheHDh7No0SKuXr1Krly56N27N+fOnWPy5MmMGDGCkiVLMm3aNBo1auR0ZBFHpEiRguTJk1OuXDkWLVr0wMPsfvjhB9q3b0+pUqWYP3++htGKxMJaexp4NZbHV3DncPrRwOi4/PlFixbl0KFDbN26lbRp095/B6LXy61Tpw67du1izJgxtGnTJi4jiYi4FTWcAsC5c+eYPXs2x48fZ8eOHSxdupQDBw4AkDJlSurWrUvu3Lk5cuQIAwcO5Nq1a9SrV48PP/yQcuXKacisJGmpUqVi/fr1rFix4oGbzf79+/P5559Tu3ZtZsyYQZo0aeI5pYg8qixZsjxU/apVqzh+/DiLFy+mUqVK8ZRKRMQ9qOFMgq5du8a6desYM2YM58+fZ//+/WzZsoXIyEgg+g9rkSJFyJEjB02bNuXpp59m2bJlDBw4kNSpU9OmTRs6d+6s61BEbvKgjeYNxYoV47333mPkyJEkS6b/FYskBvv378fHx4e3336bV155hUyZMjkdSUTEcfqUk8hFRkbyzz//sHbtWoYNG8aFCxc4fPgwV65cAcDLy4vy5cvTvXt3IiIiePPNNylWrBjbt2/n888/p3z58hw7doyqVauSKVMm2rRpw5NPPunwqxJxT+Hh4axYsYKaNWtSp04d6tSp43QkEYkjo0eP5sMPP2TDhg0UKFBAzaaIyHVqOBORyMhINm3axM6dO1m/fj2TJ08mLCyMiIgIADw8PMiePTsdOnSgVKlSZMmShXTp0rFhwwbWrl3L2rVryZs3L8WLFydZsmRs2rSJo0ePYoyhevXqVK9e3eFXKOLe+vTpw3fffcfevXvJli2b03FEJA41bNiQEydOPPCkQiIiSYUaTjdlrWXdunX89ttvnD9/nvXr17Ns2bKY5tLLy4uMGTPi5+dH27ZtKVmyJB4eHmzcuJG1a9cyfPhwNm3axOXLlwHInDkzfn5+PPXUUwDkyZOHPXv2ANELBIvI4+vTpw+1atVSsymSSOzdu5evv/6aESNG8NRTT9G7d2+nI4mIuBw1nG7i6NGjTJs2jcWLF+Ph4cH69es5deoUAMmSJaNo0aJUqFCBHDly0KlTJwoWLEhwcDCRkZHUrFmTq1evkj59eq5cuUKqVKkoWbIk7dq1w8/PDz8/P3x8fDTxj0g8WL9+Pb179+ann34ibdq0VK5c2elIIhIHli1bRsOGDYmKiqJjx44UKFDA6UgiIi5JDacLOnXqFOvXr2fmzJksXLgQiG44byhQoAB169bl6aef5oUXXqBOnTps27aNtWvXcvr0aYoWLQpAv379AKhZsyYpUqRg+vTpPP/88xQsWFCTlIgkgF9//ZXGjRuTJUsWTpw48cBLKoiIa/P396dt27a88MIL/PLLL+TOndvpSCIiLktdh8NurO21ZcsWFixYwOrVq2OGxQIkT56cV155hYoVK5IvXz4KFy7MiRMnYq65nDdvHq1btyYqKgqAfPnyMWDAADw8PJgyZUrMEFmA+vXrJ/jrE0mqJk6cSOvWrSlcuDBBQUE8/fTTTkcSkccUGRlJ9+7d+fbbb6lWrRo//fQTTzzxhNOxRERcmhrOBHTmzBnmzp3L0qVL2b17NytXruTgwYMx27Nnz07WrFl5/fXXqV27NsWLF+e///7Dx8cHDw8PvvrqK+rXr8+lS5cAyJQpE35+fjRo0AA/P7+YiYBuyJkzZ4K/RpGkzlrLl19+Sa9evahatSqzZ88mffr0TscSkcd04cIF6tatS1BQEB07dmTo0KEaLSQi8gD0f8oEVKpUqZiJeLJnz06xYsXInTs3b775JvXq1SMqKoq1a9dSokQJnn76aSZNmkTz5s35559/yJs3LwULFqRt27Yx113mzJlT112KuJDIyEg6dOjAqFGjeOuttwgICCBFihROxxKRx7R37146dOjAoUOHGD16NG3btnU6koiI21DDmYC++uor5s+fT7du3ciZM2fMjLGLFi2if//+7N+/H4CAgABatGhBpUqVGDNmTMxaXlq3T8R1Xbp0iT59+rBy5Uq6d+/OV199hYeHh9OxRCQO9OzZk//++4/Fixdr4i8RkYekhjMBVahQgYkTJ/LWW2/x999/x1x3mSNHDvz8/Gjfvj1+fn6ULFkSgOeee442bdo4GVlEHtDgwYNZtWoVw4YNo3Pnzk7HEZE4NGrUKBYsWKBmU0TkEajhTEBp06Zl9erV+Pn5Ua9evZjrLr29vZ2OJiKP6eOPPyZNmjRqNkUSoYwZM5I9e3anY4iIuCU1nAkoVapUzJ49m0qVKjkdRUTimJeXF76+vk7HEBEREXEpusAogWmSHxERERERSSrUcIqIiIiIiEi8UMMpIiIiIiIi8UINp4iIiIiIiMQLNZwiIiIiIiISL9RwioiIiIiISLxQwykiIiIiIiLxwlhrnc6QaBhjTgIH7lOWGTiVAHHikjInjMSaOYe1NktChIlPOr5dijInDB3ft0qs/w1djbtldre8kISObXENajgTmDFmvbXWrVaHV+aEoczuzx3fD2VOGMrs/tzx/VDm+OduecE9M4t705BaERERERERiRdqOEVERERERCReqOFMeGOdDvAIlDlhKLP7c8f3Q5kThjK7P3d8P5Q5/rlbXnDPzOLGdA2niIiIiIiIxAud4RQREREREZF4oYZTRERERERE4oUaThEREREREYkXajhFREREREQkXqjhFBERERERkXihhlNERERERETihRpOERERERERiRdqOEVERERERCReqOEUERERERGReKGGU0REREREROKFGk4RERERERGJF2o4RUREREREJF6o4RQREREREZF4oYZTREREXJYxpoMxZr0x5ooxZuJ9ap83xsw3xpw3xpwyxgxOoJgiInIXyZwOkJhkzpzZ+vj43LPmwoULpEmTJmECxRFlThiJNfOGDRtOWWuzJFCkeKPj23Uoc8JwoeP7CDAAqA6kuluRMSYF8BvwPfA6EAnkeZAfoOPbdbhbZnfLCy51bEsSoYYzDvn4+LB+/fp71oSEhFCxYsWECRRHlDlhJNbMxpgDCZMmfun4dh3KnDBc5fi21v58/Wf5AtnvUdocOGKtHXrTY1se5Gfo+HYd7pbZ3fKC6xzbknRoSK2IiIgkBmWA/caYX68Ppw0xxhR2OpSISFKnM5wiIiKSGGQHKgF1gSVAZ2CuMSaftfbq7cXGmDZAGwBvb29CQkLu+eTh4eH3rXE1yhz/3C0vuGdmcW9qOEVERCQxuASstNb+CmCM+QboBeQHNt9ebK0dC4wF8PX1tfcbYphYh066GnfL7G55wT0zi3vTkFoRERFJDLYA1ukQIiJyKzWcIvJIrNXnOpHEKioqyukIMYwxyYwxXoAn4GmM8TLGxDZC60egjDGmqjHGE/gAOAVsT8C4Ii7PlY5vSRrUcIrIQ4uMjKR48eLMnj3b6SgiEsfmzp2Lj48Pf/75p9NRbuhF9HDZHkDT67d7GWOeM8aEG2OeA7DW7ri+/QfgDFAPqBvb9ZsiSdWnn35K06ZN2b9/v9NRJAnRNZwi8tDatm3L5s2bSZcundNRRCQOBQcHU79+fay1REREOB0HAGttH6DPXTanva32Z+DneI4k4pa++OILBg4ciIeHzjdJwtJvnIg8lKioKKZPn46HhwfdunVzOo6IxJF58+bx8ssvAzBw4EB8fX0dTiQiceXDDz+kV69epEyZkq+//hofHx+nI0kSooZTRB7K4MGDuXDhAk2aNCFDhgxOxxGRODBo0CDq1asHwOzZs+nRo4fDiUQkrrz00kt8++23pEuXjh07dlCiRAmnI0kSo4ZTRB5YZGQkvXr1whjD6NGjnY4jInEgNDSUzz77DIDJkydTv359hxOJSFwZNGgQa9asIXny5OzcuZMcOXI4HUmSIDWcIvLA5s6dGzNhUPr06Z2OIyKPac6cObz88ss899xzrF+/nrfeesvpSCISByIjI3n//ffp0aMHDRs25NixY2TNmtXpWJJEadIgEXlg3333Hc888wwrVqxwOoqIPKZ3332X8ePHxxzTTz/9tNORRCQOREZGkjVrVk6dOsVbb73FpEmT8PT0dDqWJGFqOEXkgQwaNIiQkBCGDx9O6tSpnY4jIo8hMDCQCRMmkCxZMmbOnKlmUySRuHr1Ks2aNePUqVNkz55dzaa4BA2pFZEHcuMar3feecfhJCLyON59910aNWqEn58fJ0+epGzZsk5HEpE4cPDgQQoUKMDMmTMZPHgwhw4dUrMpLkFnOEXkvlavXs21a9eoXbu2ZqYVcWPVq1dn8eLFZMmShcWLF5M2bdr77yQiLu/AgQPkzp2ba9eu0b17d7p37+50JJEYOsMpIvfVt29fsmTJwowZM5yOIiKPqEGDBixevJg0adKwadMmNZsiicSWLVvIly8f165d49VXX2Xw4MFORxK5hRpOEbmnXr16sXjxYl577TXSpEnjdBwReQQFChQgMDCQZ555hmPHjpEtWzanI4lIHJg1axbFixfn8uXLfPnllwQGBjodSeQOGlIrIvc0YsQIAD744AOHk4jIw4qKiqJkyZJs376dJ554gj179uDl5eV0LBGJA/PmzeP111/HWsuAAQP45JNPnI4kEiud4bwPY0xuY8xlY8yPTmcRSWg//vgj58+fp0qVKuTKlcvpOCLyEKKioihRogR//vknPj4+HD9+XM2mSCIREBDAq6++ijGGcePG0bNnT6cjidyVGs77+x5Y53QIESd8+OGHGGOYOHGiozmMMSmNMf7GmAPGmPPGmE3GmJr3qO9ijDlmjAkzxgQYY1LetM3HGLPUGHPRGPOPMaZqwrwKkYRz6dIlMmXKxObNmylevDh79uwhRYoUTscSkTjQvn17WrVqhYeHB7///jutW7d2OpLIPanhvAdjTBPgLLDE6SwiCW3QoEGcPHmSAgUKkD17dqfjJAMOARWADMBnwExjjM/thcaY6kAPoArgAzwP9L2pZDqwCcgE9AT+zxiTJf6iiySsa9eu8eabb3L27Fmee+451q9fj4eH/tyLJAZz5sxh7NixeHh4sGDBAipVquR0JJH70l+guzDGpAf6AV2dziLihOHDhwMwbdo0h5OAtfaCtbaPtXa/tTbKWjsf2AeUjKX8HcDfWrvVWnsG6A80BzDG5AFKAL2ttZestbOBv4CGCfJCROLZyZMnadCgAXPmzKFnz54cOHBAzaZIItG+fXsaNGhAyZIlOX78ONWrV3c6ksgD0aRBd9ef6A+th4wxdy0yxrQB2gB4e3sTEhJyzycNDw+/b42rUeaE4UqZV6xYwdGjRylSpAinT5++ay6nMhtjvIE8wNZYNhcE5t50fzPgbYzJdH3bXmvt+du2F7zLz9Hx7YKUOXbnz5+nQYMGRERE0KFDB6pWrfpYP9Md32eRxKp27dosWLCAHDly8Pvvv2tZI3ErajhjYYwpBlQFit+v1lo7FhgL4OvraytWrHjP+pCQEO5X42qUOWG4UuZXX30VgJ9//pkXXnjhrnVOZDbGJAemApOstf/EUpIWCLvp/o3b6WLZdmN7rGtE6Ph2Tcp8p/Pnz1O7dm0iIiIoW7Ys33333WM/pzu+zyKJjbWWPn36sGDBArJmzcq6devUbIrbUcMZu4pEX/t18PrZzbSApzGmgLW2hIO5ROLd0qVLCQsLI1u2bPdsNp1gjPEApgBXgQ53KQsH0t90/8bt87Fsu7H9PCJu6s8//6Rx48bs27eP6dOn06RJE6cjiUgcuHr1Krlz5+bgwYO0bNmSMWPGkCyZPrqL+9FvbezGAjNuut+N6Ab0fUfSiCSg8ePHkypVKoKCgpyOcgsT/e2PP+AN1LLWXrtL6VagKDDz+v2iwHFr7X/GmK3A88aYdDcNqy0KOH+hqsgj2Lp1KyVLliQqKorJkyer2RRJJC5evEiuXLk4evQouXPnZvz48dzrEi8RV6aZBGJhrb1orT124x/RZ0UuW2tPOp1NJD6tWLGCadOm0bFjR4oUKeJ0nNuNBvIDday1l+5RNxloZYwpYIzJCPQCJgJYa3cCfwK9jTFexpj6QBFgdrwmF4kH69evp0SJEkRFRdGiRQuaNWvmdCQRiQOnT58mZ86cHD16lNKlS7N9+3Y1m+LW1HA+gOuzYzZ1OodIfKtVqxYAnTt3djjJrYwxOYD3gGLAMWNM+PV/bxljnrt++zkAa+1CYDCwFDhw/V/vm56uCeALnAG+AhrpyyRxN7NmzaJ06dJcvXqVESNGEBAQ4HQkEYkDW7duxdvbmxMnTtCkSRNCQ0Px9PR0OpbIY9GQWhEBYPfu3YSHh/P888/zzDPPOB3nFtbaA8C9vt69ZQYFa+1QYOhdnms/0ddpi7ilRYsW8frrr2Ot5bvvvqNDh7tdzpw4GGM6EL20UWFgurW2+QPsEwxUApJbayPiNaBIHNm+fTu+vr5ERERQvXp1pk+f7nQkkTihM5wiAsDAgQPx8vJi1apVTkcRkbv4+eefY0YifPvtt4m+2bzuCDAAeKDTuMaYt9AX6uJmFi1aRNGiRbl8+TKfffYZCxcudDqSSJxRwykizJw5k4CAAN5++22yZs3qdBwRiUW/fv1o2LAhAPPnz+eDDz5wOFHCsNb+bK2dA/x3v1pjTAaih9B/FO/BROLIggULqFGjBteuXeO7776jX79+TkcSiVP6BlBEaNeuHQD169d3OImIxGblypUMGDAAgOnTp8ec5ZQ7fEn0BGPH7ldojGkDtAHw9vYmJCTknvXh4eH3rXE1yhz/Hjfvzp076d69O56enrRo0YJChQrF++t3t/dY3J8aTpEkbs2aNfz333+88MIL1KhRw+k4InKb6dOn07p1a3LmzMnPP/9MwYIFnY7kkowxvkBZoDOQ/X711tqxRC+Dhq+vr61YseI960NCQrhfjatR5vj3OHm7dOnC2LFjyZw5M+vWrSNPnjxxG+4u3O09FvenhlMkiXv33XcBmDp1qsNJROR2LVq0YOLEiWTLlo3ly5fj7e3tdCSXZIzxAEYBna21EVpCQlzdjBkzGDZsGJ6enqxcuZJnn33W6Ugi8UbXcIokYUFBQWzdupUcOXJQunRpp+OIyE1mzZrF5MmTSZEiBXPmzFGzeW/piV7u6CdjzDFg3fXHDxtjyjsXS+RO06ZNo1mzZvj4+LB8+XI1m5Lo6QynSBLWpUsXAL7++muHk4jIzd566y2mTZtGuXLlWLBgAenTp3c6kmOMMcmI/rziCXgaY7yAiNuWOwkDbl7P6VlgLVAS0Dq74jJKlizJxo0bqVChAvPmzUvSx7YkHWo4RZKoP//8k507d5InTx4aN27sdBwRua5KlSoEBweTJUsWFi5cSJo0aZyO5LReRM88e0NToK8xJgDYBhSw1h7kpomCrjelAMe1Dqe4gqioKKpWrcrGjRtJmTKlmk1JUtRwiiRRr732GgCjR492OMn/Y+++w6Oq8j+Ov7+EJlVACKAgHRVkIyKCCgQQAcG2FJUmShFZsCviwlpBsMCKUhaBpbOoC4irFBGjUqRLW4rSQUCKCQklgcz5/ZHBX8wGQpnMnUk+r+eZh5l7z9z5zCUnk+/cc88VkbNatmzJwoULKViwIOvWrVOxCTjnXgVePcfqAud4zk5AJ3JKSDhz5gy1a9dmzZo1XHvttaxdu1bFpmQrKjhFsqHly5fz008/UahQIRo1auR1HJFsLzk5meuuu46ff/6ZMmXKsHnzZvLly+d1LBG5TCdPniQyMpL4+HhuvPFGVq9eTc6c+vNbshdNGiSSDX366acAfPjhhx4nERGfz8dNN93Ezz//TNGiRfn5559VbIpkAbGxsVSsWJH4+HjKlSvHmjVrVGxKtqSCUySbOXToEMOHD6d9+/Z07NjR6zgi2dqZM2eoUaMG69evp3Llyuzfv5/cuXN7HUtELtPWrVu59tpr2b9/Pw888AA7duwgIiLC61gintDXLCLZzC233MKJEyfo27ev11FEsrWEhARKly5NfHw8tWvXZunSpeTIoe+BRcLdkSNHqF69OqdPn+bRRx9l3LhxXkcS8ZQ+2USykaNHj7J7927y589PtWrVvI4jkm0lJSXRrl074uPjqVChgopNkSxi37591K9fH+ccbdq0UbEpggpOkWzl73//O845Fi1a5HUUkWzrwIED3HPPPXz++ecMGDCAbdu2qdgUyQImTpxI5cqV2bVrF/Pnz+fjjz/2OpJISNCQWpFsYt26dbz77ru0bt2aqKgor+OIZEuxsbGUKVOGM2fOMGLECJ544gmvI4lIAKxbt44uXbpw5swZpk6dSsOGDb2OJBIy9JWqSDZx3333cfLkSVq2bOl1FJFsKS4ujpYtW3LmzBkaNmyoYlMki/j2229p0KABxYsXZ/LkyTz88MNeRxIJKSo4RbKB3bt3s3PnTgoVKsQjjzzidRyRbOe///0vUVFRLF++nE8//ZSFCxd6HUlEAuCtt94iOjqaQoUKsXTpUtq3b+91JJGQoyG1ItlAly5dABgxYoTHSUSyn3Xr1tGrVy+cc0yZMoVWrVp5HUlEAuDFF19k/vz5mBkTJkzg2muv9TqSSEhSwSmSxW3dupUFCxZQrFgxffMqEmTfffcdd955J845Hn/8cdq1a+d1JBEJgG7dujFmzBiuuOIKNmzYQIUKFbyOJBKyNKRWJIs7W2TqfDGR4Bo4cCDR0dGcOXOGZ555hlGjRnkdSUQCoEmTJowZM4YCBQowceJEFZsiGdARTpEs7Ndff2XVqlUULVqUN954w+s4ItnGoEGD+Otf/wrAxx9/TPHixT1OJCKB0KFDBxYsWECePHnYuHEj27dv9zqSSMjTEU6RLKxTp04453jrrbe8jnLZzKyXma00s0QzG3+edqPMLCHVLdHM4lOtjzGzU6nWbwnKG5Bso1u3bvTt25e8efMyb9482rRp43UkEQmA+vXrM2XKFIoXL86BAwcoW7as15FEwoKOcIpkUbt372bevHlcccUVdO/e3es4gfAL8CbQFLjiXI2ccz2AHmcf+4tTX5pmvZxzYzIho2Rjzjmuv/56tmzZwpVXXsnatWv1B6lIFvHnP/+Z77//nsKFC7N9+3YKFCjgdSSRsKGCUySLmjBhApB1zt10zs0AMLNawDUX8hwzyw+0AnTxUclUycnJPPnkk2zZsoX8+fOzbds2ihYt6nUsEQmAN998k5kzZ1KpUiVWrlypYlPkIqngFMmCTp06xciRI2nYsCHvvfee13G81Ao4BHyXZvlbZjYI2AKbghtlAAAgAElEQVT81TkXk96Tzaw70B0gMjKSmJh0m/0uISEhwzahRpkv36FDh3jppZfYvn07Dz74IN27d2fdunV/aBNqmS9EOGYWCaTk5GSqVq3Ktm3b6NSpE+PGjSMiIsLrWCJhRwWnSBZ0zz33sH//fqZMmeJ1FK89Akx0zrlUy/oA/wWSgIeAz80syjm3Le2TnXOjgdEAtWrVctHR0ed9sZiYGDJqE2qU+fIcOXKEsmXLcuLECV588UUGDx6cbrtQynyhwjGzSKA453j00UfZtm0bJUqU4J///Cc5cmjqE5FLoZ4jksWcOnWKr7/+moiICBo0aOB1HM+YWRmgATAx9XLn3DLnXLxzLtE5NwFYDNztRUYJbzt37uT222/n9OnTtG7d+pzFpoiEl+TkZHr37s2kSZNo1aoVe/bsUbEpchl0hFMkixk/fjzOOYYOHZrdPyA7AUuccxnNWe8AC0IeyUIGDRpEv379yJ8/P19//TX16tXzOpKIBEBSUhIlSpQgLi6OZ599lnfffRczfUSIXI5s/dfo+ZjZZDPbb2bHzGyrmXX1OpNIRuLj43nzzTepW7cuvXr18jpOQJlZTjPLC0QAEWaW18zO96VZJ2B8mm1caWZNzz7XzNoD9YF5mRZcspzBgwfTt29fkpOTGTVqlIrNTHYRl0R6xMxW+T+395rZ2xn8jhD5g6SkJK677jri4uK45ppreOedd1RsigSACs5zewso55wrBNwLvGlmN3ucSeS8mjZtyr59+2jfvn1W/JDsB5wEXgI6+O/3M7Oy/utp/n79CTOrS8pMtp+k2UYuUi6tcgg4DPQG7nfO6VqcckG6dOnCSy+9RJ48eViwYAEPP/yw15Gyg7OXRBqXQbt8wNPAVcCtQGPg+cyNJllFXFwclSpVYseOHURHR2sYrUgA6Zu/c3DObUz90H+rCKzyJpHI+Z04cYJly5YRERFBjx49Mn5CmHHOvQq8eo7Vf5ij3jm3FMifzjYOAbcEOptkfT6fj+rVq7Np0yYKFy7MunXrdI3NILnQSyI550amerjPzKYADTM5nmQB8fHxlCxZklOnTtGsWTPmzJnjdSSRLEUF53mY2QigMykXmV8DfJlOG102IQRlx8zvvPMOPp+PTp068f333wcu2HmE434WuVhJSUnUqlWLTZs2kT9/fn766SeKFy/udSzJWH1g47lW6vM7NAU7c0JCAh07duTUqVNUqFCBPn36XNTrax+LZEwF53k453qaWW+gLhANJKbTRpdNCEHZLXNCQgLz5s3jiiuuCOrU7eG4n0Uuxr59+6hRowZHjx6lVq1aLF26lJw59dEZ6szsUaAWcM75F/T5HZqCmXn37t3UqFGDuLg42rVrd0mXEtM+FsmYBqdnwDmX7JxbRMownie8ziOSnj//+c8kJyfTsmVLnXMiEiBHjx7l+uuv5+jRo9x1112sWLFCxWYYMLP7gUFAc+fcYa/zSGhKSEigatWqxMXF0bFjR123WiQT6S/TC5eTlHM4RUJKUlISMTEx5M6dWx+YIgGya9cu7rjjDk6ePMlDDz3EvHmayDgcmFkz4CPgHufceq/zSGiKi4ujadOmJCYm0rx5cyZOnJjxk0TkkqngTIeZlTCzh8ysgJlFmFlT4GFgodfZRNJ67rnnOH36NL179yZXrlxexxEJe4MHD6ZChQrs27ePBQsWMG3aNK8jZWsXekkkM2sETAFaOeeWBzunhIfVq1dTrlw5li1bxieffMKXX/7P9BwiEmAqONPnSBk+uxf4DXgXeNo595mnqUTSSEpKYvjw4eTMmZNBgwZ5HUck7H399df0798fn8/HkCFDaNCggdeR5MIvidQfKAx86V+eYGaablR+d+jQIaKjo4mNjaVv3760atXK60gi2YJORkmH/9IJ+itDQt6wYcNwznHnnXfq3DKRy/Taa68xYMAAqlatyr/+9S+qVavmdSThwi+J5JzTJVDknPbv30+TJk1ITEykX79+vPHGG15HEsk29BeqSJjy+XxMnDiRKlWqMHv2bK/jiIQtn89HVFQU69evp2rVqnz//fdceeWVXscSkQD54osvuPfee8mVKxdz586lYUN9NyESTCo4RcJUv379WL9+PZMnT9a5myKXKCkpiVtuuYX169dToEABFZsiWczixYu577778Pl8PPXUUyo2RTygczhFwpDP52PQoEGYGW3btvU6jkhY+uWXX4iMjGTdunXcfPPN/PbbbxQvXtzrWCISIPPmzaNBgwb4fD5GjBjB4MGDvY4kki2p4BQJQ1988QXOOdq2baujmyKXYNu2bVSsWJHY2Fjq16/PypUrdR60SBYyduxYmjVrhs/nY9KkSTzxhC6lLuIVFZwiYcY5x+uvv06FChWYNGmS13FEws6SJUu44YYbOHXqFG3atOHbb7/1OpKIBNDMmTPp1q0bAO+++y7t27f3OJFI9qaCUyTMPProo6xcuZLu3bvr6KbIRXrnnXe4/fbbSUpKYujQoXz88cdeRxKRAPr73/9Oq1atMDO++OILnn32Wa8jiWR7Gj8kEkZ8Ph9Tp04FoHv37h6nEQkvCxcupH///kDKJYV69+7tcSIRCaQpU6bwzDPPYGbMnz+fxo0bex1JRNARTpGwMnjwYE6fPk3btm0pUqSI13FEwsbLL79Ms2bNqFSpElu3blWxKZLFLF68mCeeeIJ8+fIxffp0FZsiISToBaeZ1Tezz8xsl5k5M+t3jna3mtkSMztlZvvN7C0zi0jTppSZfWxmx/y3f5lZiVTrc5vZP80szsx2mtmDaZ5/kz+H5sCXkJecnMzAgQOJiIjgo48+8jqOSFhwztG4cWPeeustrrnmGhYtWkTlypW9jiUiAdS7d28aN25MyZIl2bJlC23atPE6koik4sURzgLAf4EXgQPpNTCzMsBXwBbgZuAJ4HFgQKo2OYD/AOWBJsBdQBVglpmZv1l3oC5wB9AfGG9mxfzPzwmMA550zsUG9i2KBN5f/vIXEhISqF+/PoUKFfI6jkjI8/l8PPPMMyxcuJDixYvzww8/6BqbIlnMZ599xocffsjp06eJiYnhmmuu8TqSiKSRYcFpZjFmNs7MBpnZYf+RxDFmdsWlvKBz7kvnXF/n3HQg8RzNngCOAV2ccxudc7NIKRh7m1l+f5s7gZpAB+fcMufcD0BHUgrMBv421wOznHPrnXOT/Nus4F/XF9jknPvsUt6HSLBNnToVM/v9HE4RObcDBw5QokQJ3n//fZ555pnfH4tI1vGf//yHtm3bUq5cOb777jtKly7tdSQRSceFHuFsDRQD6gHtgXuBwQBm9rKZJWRwe/kic90OzHfO+VItmwvkA25K1WaHc27L2QbOuY3AXlKOaAKsBeqbWX4zuxXIC/xkZtVIOfr55EXmEvHE+++/T3x8PPfffz8lS5b0Oo5ISNuxYwcVK1bkyJEjNGzYkCFDhpAjh6YsEMlKHnjgAe69916qV6/OqlWruP32272OJCLncKGz1B4FejjnkoFN/vMuPzCzvsAoIKN55Y9eZK5SwOI0yw6kWnf23/SG5B5I1WYcUB1YD8QD7fz/jgOeBW41swFAbuA959zYi8wpEhR9+/bFzBgzZozXUURC2pIlS2jUqBGJiYm0a9eOKVOmeB1JRALsySefZNasWeTIkYNZs2ZRtGhRryOJyHlcaMG53F9snrWYlCKtonNuHRdfUF4Kl+bfDNs6586QchTz9yOZZvYCKUdBY4CtpAy/PQr8aGY/+I+SioSMUaNGcfLkSapVq6YPVZHzGDJkCM8999zv95955hmPE4lIoHXt2pWxY8dSqFAhVqxYQZkyZbyOJCIZuNQxRvb7ncwZUrsfSDtu8OzjA+dpAxDJuScjqgw8BfQk5VzPn51z65xzZwvQRheZUyTTzZo1iyJFivDFF194HcVTZtbLzFaaWaKZjT9Pu85mlpzmd1B0qvXlzOwbMzthZpvN7M5g5JfMpWJTJOurU6cOY8eOpUiRIvz8889UqVLF60gicgEu9AjnLWYWkeooZ10gCdhG5gypXQx0NLMcqc7jbAacANakavM3M6vsnPsJwMyuB8oAi9Ju0D9z7VjgJefcQf8st7lSNcmNrksqIWbBggXMmzePwYMHc+2113odx2u/AG8CTYGMJi1b6py74xzrpgFLgbv9t0/9v0cOBSypBFW7du2YNm0aefLkISYmhjp16ngdSUQCrE2bNixbtozcuXOzYcMGihcv7nUkEblAF1pwFgOGm9n7pMzy+gbwkXPuOHCciygozawAUMn/MDdQ0syigATn3M/+5SOBXsBHZjYEqOh/zQ/8rwmwAFgNTDaz3qQcdR0O/AB8m85L9wKOOecm+x8vB6qaWSvgCNAYeOVC34dIMLRo0YIcOXLQs2dPr6N4zjk3A8DMagGXNO+9mVUhZXbru5xzJ4F/m9nTQCtSvjyTMNOyZUu++OIL8uXLx4YNGyhfvrzXkUQkwJo0acKCBQsoWbIkmzdvpnDhwl5HEpGLcKEF56ekTLaziJQi8RNSrqN5KWoB36R6/Bf/7VsgGsA5t8fM7gKGAKuAWGA00O/sk5xzPjNrCQwDviblvM05QG/n3B/O8zSzckAf4NZUz//FzLoCfwfyAP2dc2sQCRHLli0jKSmJG2+8kQIFCngdJ9zcZGaHSfkybBLwlv+c7mrAdudcfKq2a/3LJYz4fD6ef/55vvjiC0qVKsXq1as1g7NIFlSjRg3Wr19PyZIl2bZtG/ny5fM6kohcpAstOH3OuReAFy73BZ1zMaQ6B/Q87X4AbsugzX6gzQVsayfpHBFxzk0BNIWhhKSBAwdSpEgRFi36nxHicn7fkTI79S5SCsnpwBngLaAAEJemfRxwdXobMrPupFxCicjISGJiYs77wgkJCRm2CTXhmHnPnj0ULlyYhIQEWrVqRc+ePdm8eTObN2/2Oto5heN+DsfMknU45+jfvz/r16+nWLFi/PTTTyo2RcLUhRacIhJEQ4cOZfbs2fztb3+jUKFCXscJK8657akerjez10n5suwtIAFIu0MLkTKCI71tjSZldAW1atVy0dHR533tmJgYMmoTasItc2xsLJ07dyYhIYG6devyySefkHKKfmgLt/0M4ZlZsobk5GTq1q3LihUr6Nq1K6NGjSIiIsLrWCJyiTRJjkgI+utf/wpA27ZtPU6SJTj+f1TFRqCCmRVMtf5P/uUS4jZs2EC9evXYt28fAwcOZMmSJWFRbMrludAZqv1tnzGzA2YWZ2bjzCxPkGJKgPh8PurUqcOKFSuIioriH//4h4pNkTCXYcHpnIt2znUNRhgRgRkzZnDy5EmioqKoVk2nFp5lZjnNLC8QAUSYWV4z+59RGmbW3Mwi/fevA/oDnwE457YCPwKv+J//AFAD+Hew3odcmgkTJnDjjTeyefNmBg0aRN++fb2OJMFzdobqcedrZGZNgZdImQSwHCmTHL6W2eEkcHw+H0OGDGHlypXUrl2blStXkiOHjo2IhDv1YpEQ8/TTTwPw6aefepwk5PQDTpLyB2UH//1+ZlbWf63Nsv52jYF1ZnYc+BKYAQxMtZ2HSJm87DdgENBal0QJbdOmTePRRx8F4G9/+xs333yzx4kkmJxzM5xzs0iZUf58HgHGOuc2Oud+I2V2+86ZnU8C4+TJk0RGRvLFF1/w8ssv88MPP+jIpkgWoXM4RULIhx9+yJ49e4iKiqJixYpexwkpzrlXgVfPsbpAqnbPA8+fZzs78c+ILaHvpZdeYvDgwURERPD555/TvHlzTWQj51IN/2gGv7VApJkVc879T7GqScFCR0JCAh06dCAuLo4yZcrQpEkTvv02vSvchZ5w2cephWNmCW8qOEVCyJtvvgnAyJEjPU4i4r1evXoxfPhwIiIiWLx4MbfeemvGT5LsLO0s1GfvFySdo6OaFCw0bN++nZo1axIXF0ezZs3o06dPyGdOLRz2cVrhmFnCm4bUioSIZcuWcfDgQWrXrk2dOnW8jiPiqR49ejB8+HBy587NokWLVGzKhUg7C/XZ++nOQi3emz9/PpUqVSIuLo7evXszZ84cryOJSCbQEU6REPHOO+8AMG7ceefFEMnybrrpJn788UeKFCnChg0bKF26tNeRJDxsJGXW6Y/9j/8EHExvOK14b968ebRo0QLnHF26dGHYsGFeRxKRTKIjnCIhYPLkyRw5coTSpUtrZlrJtpxz9OzZkx9//JH8+fOzc+dOFZtywTNUAxOBLmZ2g5kVIWWisfFBjCoXaODAgTRv3hyfz8f48eMZM2aM15FEJBOp4BQJAfPnz8fMmDBhgtdRRDyRlJREz549GTlyJE2bNuXQoUMUKlQo4ydKdnBBM1Q75+YCbwPfALv8t1e8iSznMmLEiN+vNT1v3jweeeQRjxOJSGbTkFoRj3Xp0oVJkybx4IMPcuedd3odRyTofvvtN6655hpOnDhBnz59eOuttzAzr2NJiLjQGar9bYcAQzI5klyiqVOn8tRTT1GkSBGmT59OkyZNvI4kIkGggjMdZpYHGAHcCRQFfgZeds7pbHYJGOccN954Ixs3bqRevXp069bN60giQXf8+HHatm3LiRMnuO666xg0aJDXkUQkE0RFRbF27VoaNGjAZ599RuHChb2OJCJBoiG16csJ7AEaAIWB/sDHZlbOw0yShTjnePHFF9m4cSOlSpVi3rx5usC1ZDurV6/m9ttvZ+HChYwePZpNmzZ5HUlEAuzsudlr166lYMGCzJ07V8WmSDajgjMdzrnjzrlXnXM7nXM+59x/gB3AzV5nk/B3+vRpqlWrxrvvvkvPnj3Zu3cvV1xxhdexRIJqxYoV1KpVi7Vr1zJ9+nQd4RfJgk6cOEHnzp0ZOXIkDzzwAAcPHiRv3rxexxKRINOQ2gtgZpFAFVKmXE+7rjvQHSAyMpKYmJjzbishISHDNqFGmQPnzJkzPPzwwxw+fJhKlSrRunVrvvvuOyB0M59POGYW73399dc0a9YM5xw9evSgdevWXkcSkQDbu3cvFStWJCkpiVdeeYVXXnlF52aLZFMqODNgZrmAKcAE59zmtOudc6OB0QC1atVy0dHR591eTEwMGbUJNcocGKdOnaJatWocPnyY6667jg0bNvxhGG0oZs5IOGYWb7399tu89NJLAIwfP14zVIpkQdu2bSMqKoqkpCRq1arFq6++6nUkEfGQhtSeh5nlACYBSUAvj+NIGIuLi6NEiRJs376dxo0bs2nTJp2zKdnOBx98QJ8+fXDOMXv2bBWbIlnQ559/zg033EBCQgLPPfccK1as8DqSiHhMBec5WMq4j7FAJNDKOXfa40gSpo4dO0alSpWIj4+nRo0aLFiwwOtIIkH3yiuv8OSTT5IrVy5mzJhBy5YtvY4kIgE2c+ZM7r33XpKSkhg6dCjvvvuu15FEJARoSO25jQSuB+50zp30OoyEpyNHjnD99ddz+PBhWrZsyeeff+51JJGga9y4MQsXLiRfvnysXLmS66+/3utIIhJgo0ePpkePHpgZgwYN4umnn/Y6koiECBWc6TCza4HHgUTgQKqT3B93zk3xLJiElR07dlClShXOnDlDx44dmThxoteRRIJu2LBhLFy4kJw5c7Jx40bKlSvndSQRCbB27doxbdo0cuXKxfz583Vuv4j8gQrOdDjndgGaSk0u2bFjx3j44Yc5c+YMDRs2VLEp2U5ycjJPPfUUw4cP59577+Wjjz6iRIkSXscSkQAbO3Ys06ZNI0eOHKxevZrq1at7HUlEQozO4RQJsHXr1tGwYUNWrlzJlClTWLhwodeRRILK5/Nx9dVXM3z4cFq3bs2///1vFZsiWYzP52PAgAF07dqV2rVrs2nTJhWbIpIuHeEUCaC9e/cSFRWFc44ZM2bwwAMPeB1JJKiSkpLo3LkzBw8e5Oqrr+Zf//qXZmQWyWKSk5O5+uqrOXjwIK1bt2by5MnkyZPH61giEqJ0hFMkQA4ePEjz5s0B6NChg4pNyXZ2795N1apVmTZtGoMHD2bv3r0qNkWymKSkJB555BEOHjxIsWLFmDJliopNETkvFZwiATBnzhxq1KjB9u3b+eqrr5g0aZLXkUSCaseOHVSqVImdO3fy9NNP8+KLL3odSUQCbNu2bdx6661MmTKFAQMG8Ouvv5I7d26vY4lIiNOQWpHLtGjRIlq0aIFzjv/85z80btzY60giQfXjjz9Sp04dTp8+TatWrRg6dKjXkUQkwDZv3kz16tVJTk7mvffe49lnn/U6koiECR3hFLkMCxcuJDo6GuccL7/8Mi1atPA6kkhQTZ48mZo1a5KYmMjgwYP59NNPvY4kIgG2ePFiatSoQXJyMq1bt1axKSIXRQWnyCUaM2YMd955Jz6fj8mTJzNgwACvI2VpZtbLzFaaWaKZjT9Pu0fMbJWZHTOzvWb2tpnlTLU+xsxOmVmC/7YlKG8gC/rkk0/o1KkTzjkGDhyoYbQiWdCQIUOoV68eZ86cYfjw4XzyySdeRxKRMKOCU+QSzJw5k27duuGcY/z48bRv397rSNnBL8CbwLgM2uUDngauAm4FGgPPp2nTyzlXwH+rGvCk2cAHH3xA27ZtyZEjB1OnTqVv375eRxKRAPvwww957rnncM4xZcoUevbs6XUkEQlDOodT5CJNmzaN9u3bkyNHDsaMGUOnTp28jpQtOOdmAJhZLeCa87QbmerhPjObAjTM5HjZyqOPPsr48ePJmTMn3333HXXr1vU6kogEWN++fRk0aBC5cuXiX//6F3/+85+9jiQiYUoFp8hF6NevHwMGDCAiIoL58+fTqFEjryNJxuoDG9Mse8vMBgFbgL8652LSe6KZdQe6A0RGRhITk26z3yUkJGTYJtRcbOavvvqKCRMmkCNHDoYMGUJiYmLQ33N22M+hIBwzS2DcdtttLF26lPz587Nq1SqqVtVAEBG5dCo4RS7QokWLePvttwH4+OOPVWyGATN7FKgFdE21uA/wXyAJeAj43MyinHPb0j7fOTcaGA1Qq1YtFx0dfd7Xi4mJIaM2oeZiMrdt25ZPPvmE6OhoPvvsMwoVKpS54c4hq+/nUBEqmc2sKDAWuAs4DPR1zk1Np10e4H3gASAXsBjo4ZzbF8S4Ye3s+dhLly4lV65crF+/nvLly3sdS0TCnM7hFLkAU6dOpWnTppQvX56tW7dqaFEYMLP7gUFAc+fc4bPLnXPLnHPxzrlE59wEUv4ovdurnOHAOUejRo345JNPqFChAnPmzPGs2JRsaTgpXxBFAu2BkWZWLZ12TwF1gRpAaSAW+CBYIcPdqVOnaNGiBf369aNDhw7Ex8er2BSRgNARTpEM9OzZk5EjRxIZGcm3335LyZIlvY4kGTCzZsBHQAvn3PoMmjvAMj9VePL5fDz11FN88803VKxYkR9//JG8efN6HUuyCTPLD7QCqjvnEoBFZjYb6Ai8lKZ5eWCec+6g/7n/AoYEM2+4SkxM5JprruHIkSM8/PDDvw+bFxEJBP02ETmP2bNn89FHH5E7d25mzZqlYtNDZpbTzPICEUCEmeVNfbmTVO0aAVOAVs655WnWXWlmTc8+18zak3KO57xgvIdwExcXR/HixX+fqfKnn36iQIECXseS7KUKkOyc25pq2VogvSOcY4Hbzay0meUj5WjonCBkDGvHjh3j7rvv5siRI0RFRTF16lQVmyISUDrCKXIOXbp0Yfz48dSqVYu5c+dSpEgRryNld/2AV1I97gC8ZmbjSDkn8wbn3G6gP1AY+NLs9wOX3zvnmpNyXtebwHVAMrAZuN85p2txpnH48GEqV65MbGwsNWvW5J133iHV/hQJlgJAXJplcUDBdNpuBXYD+0jp3+uBXufasCYFgzVr1vDyyy9z6tQpXnrpJZo2ber5ewy3/RxueSE8M0t4U8Epko42bdrw6aefUqhQIb766iudrxYCnHOvAq+eY3WBVO3OeQkU59wh4JaABsuCdu7cSY0aNYiPj+e+++5j1qxZXkeS7CsBSPsLuBAQn07bkUBeoBhwHHiRlCOct6a34ew+KVhMTAzPP/88Pp+Pfv368cYbbwQ33DmE234Ot7wQnpklvGnMhEganTt35tNPP+WKK65gzZo1KjYlW1mwYAEVKlQgPj6eZ599VsWmeG0rkNPMKqda9if+91JHZ5ePd84ddc4lkjJhUG0zuyoIOcPKtGnTaNy4Mc45Xn/99ZApNkUka9IRTpFUateuzYoVK7jqqqvYsmULRYsW9TqSSNDExMTQvHlznHN06dKF9957z+tIks05546b2QzgdTPrCkQB9wG3pdN8BdDJzGKAE0BP4JfUs1TL/0+EFxERwWeffUbLli29jiQiWZyOcIr4NWvWjBUrVpAvXz62bdumYlOyleHDh9O4cWOSk5MZM2YMY8aM8TqSyFk9gSuAX4FpwBPOuY1mVs/MElK1ex44BfwEHCLlckcPBDtsKHvuuecYOXIkZsasWbNUbIpIUOgIpwjQpEkTFixYQGRkJFu2bNEwWslWxo4dS69evX7/I/Tee+/1OpLI75xzR4H701n+PX88f/sIKTPTSjpatGjBl19+ScGCBVm1ahWVK1fO+EkiIgGgglOyteTkZMqXL8+ePXsoX748//3vf3WNQclWli1bxmuvvUa+fPmYMGGCik2RLMY5x80338yaNWsoWrQoW7Zs4aqrdFqriASPCk7JtpxzPPvss+zZs4fChQuzefNmcufO7XUskaBp0aIFc+bMISoqirlz51KiRAmvI4lIACUnJ9OjRw/WrFlDqVKl2Lp1q66lKyJBp3M4JVtKSkqie/fuDBs2jC5duvDrr7+q2JRs5fXXX+fLL78kZ86cxMTEqNgUyWIOHDhAq1atGD16NH379mXfvn0qNkXEEzrCKdmOz+cjMjKS2NhYevfuzfvvv68L2ku2cfr0afr06cPQoUO57bbbePbZZ3XOsoD5GysAAB/vSURBVEgW89tvv1GnTh3i4uJo1qwZAwcO9DqSiGRjOsIp2cqZM2fo3LkzsbGxXHPNNQwdOlTFpmQbhw8f5sorr2To0KH85S9/4bvvvqNYsWJexxKRAFq7di233XYbv/zyCw899BBz5szxOpKIZHMqOCXbiIuL44477mDSpEm88cYb7Nmzh4iICK9jiQTF7t27qVy5MidOnKBq1aoMGzZMP/8iWUyfPn2Iiopi9+7dLFiwgMcff9zrSCIiKjglezh27BilSpVi2bJl9O7dm379+nkdSSRo5s+fT5UqVYiNjaVbt25s3ryZHDn0618kK/nLX/7C22+/jZkxfPhw6tev73UkERFA53BKNnD06FGqVq3KyZMnufnmmxk2bJjXkUSCZsaMGbRq1QqAgQMH0rdvX48TiUggJScnc/PNN7N27VoKFy7M8uXLqVKlitexRER+p6+4JUvbsWMHZcqU4fDhw3Tq1ImVK1d6HUkkaEaMGEHr1q0xM958800VmyJZzNnRO2vXrqVUqVLs3r1bxaaIhBwd4TwHM+sFdAZuBKY55zp7Gkgu2t69e7n++utJTEykTZs2TJgwwetIIkHz5z//mZkzZ5IrVy4WLFig4XUiWcyWLVuoXbs2x44do3r16vz44486L1tEQpKOcJ7bL8CbwDivg8jFO3DgANdddx2JiYm0bduWjz/+2OtIIkHz97//nZkzZxIREcH69etVbIpkMR999BHXX389x44do3fv3qxfv17FpoiELBWc5+Ccm+GcmwUc8TqLXJwffviB9u3bc/z4cfr27cv06dO9jiQSFMnJybz44os888wz1KtXj+3bt1O1alWvY4lIAMXExPDkk0/inKN///6al0BEQp6G1EqWcfz4ccaPH0///v3x+Xw89thjuti1ZBsnT56kRIkSJCQk0K1bN0aMGEHOnPoVL5KVPP/88wwbNoxKlSoxefJkatas6XUkEZEM6a+Ry2Rm3YHuAJGRkcTExJy3fUJCQoZtQk2oZz569Ci9evXi4MGD+Hw+brjhBrp3706zZs1COndaob6f0xOOmbOi+Ph4WrduTUJCAuXLl2fkyJEaXieShfh8Pm655RZWr15N1apVWbJkCVdeeaXXsURELogKzsvknBsNjAaoVauWi46OPm/7mJgYMmoTakI18/jx4/n++++ZPHkySUlJXHHFFcydO5f69euHbObzUWa5FN988w1du3Zl165djBkzhi5dungdSUQC6OTJk3Tu3JnVq1dTunRpli1bRuHChb2OJSJywVRwSljx+XwMHTqU1157jfj4eHLmzEn37t3p3bs31113ndfxRIJqzpw53H333QDMnDmT+++/3+NEIhJI3377Lc2bN+fkyZO8/fbbPP/885iZ17FERC6KCs5zMLOcpOyfCCDCzPICZ5xzZ7xNlj3FxcXRunVrvv/+exITEzEzypcvz7Rp07j11lu9jicSdGPHjqVbt26YGX/9619VbIpkMbNnz+aBBx7A5/PRs2dPXnjhBa8jiYhcEs1Se279gJPAS0AH//1+nibKhvbs2cM999xDsWLFWLBgAadPn6ZTp04cPXqU7du3q9jMRsysl5mtNLNEMxufQdtnzOyAmcWZ2Tgzy5NqXTkz+8bMTpjZZjO7M9PDB9hDDz1E165diYiI4KuvvuKNN97wOpKIBNDrr7/OfffdB8C7777L8OHDPU4kInLpdITzHJxzrwKvehwj29qzZw/R0dFs374dgCuvvJIOHTrw3nvvkTt3bo/TiUfOXhu3KXDFuRqZWVNSvihq5H/OTOA1/zKAacBS4G7/7VMzq+ycO5R50QOnU6dOTJ8+nRw5crBq1Spq1KjhdSQRCaA77riDxYsXkzt3bmJiYqhbt67XkURELosKTgkpgwYNYs2aNcyYMYMzZ86QL18+hg0bpolQBOfcDAAzqwVcc56mjwBjnXMb/e3fAKYAL5lZFaAmcJdz7iTwbzN7GmgFjMrM/JcrOTmZhg0b8v3331OsWDGWLVtGxYoVvY4lIgFy+vRpbrvtNlauXEmePHn48ccfNTeBiGQJKjjFc8455s6dy9NPP83WrVvJmTMnTz75JE899RRly5b1Op6En2rAZ6kerwUizayYf91251x8mvXVgpjvoiUlJREZGUlsbCwVK1Zk3bp15MuXz+tYIkFhZkWBscBdwGGgr3Nu6jna1gT+TsoXS8eBgc6594OV9VLt2LGDmjVrEhsby4033sjy5cvJmzev17FERAJCBad4JjY2lpYtW7Jhwwbi4uIoVaoUUVFRTJgwQcME5XIUAOJSPT57v2A6686uvzq9DYXCdXZPnjzJ66+/TmxsLMWLF+cf//gHy5cvD9j2w/FaqsocHCGUeTiQBEQCUcAXZrb27CiGs8zsKmAu8AzwKZCb84+GCAm//PILf/rTn4iPj6dJkybMnz/f60giIgGlglOCbtOmTTz55JMsXLgQn89HREQEEydO5MEHH9T5mRIICUChVI/P3o9PZ93Z9fGkw+vr7C5evJjHHnuMXbt2MXLkSHr06BGwbZ8VjtdSVebgCIXMZpaflCHv1Z1zCcAiM5sNdOT/z8s+61lgnnNuiv9xIrApaGEvwerVq7nvvvs4c+YMPXr0YOTIkV5HEhEJOM1SK0GzcOFCrr76am644QYWLFhA0aJF+dvf/kZiYiIdO3ZUsSmBshH4U6rHfwIOOueO+NdVMLOCadb/4UhJKNi+fTsNGjRgx44dTJgwIVOKTZEwUAVIds5tTbXsXMPg6wBHzWyJmf1qZp+bWciel9GhQwduvvlmkpKSWLJkiYpNEcmydIRTMl2fPn2YOHEiBw4cAKBQoUK8//77dO7c2dtgElYu4tq4E4HxZjYF2E/K5YzGAzjntprZj8ArZtYPaA7UIOUISshYtWoVd999N/ny5eOJJ56gQ4cOXkcS8cq5hsEXTKftNaScu9kEWA+8Tcqs1Lent2Evh8zPmDGDKVOmYGa88MILxMbGZsrw5RAaFn3Bwi1zuOWF8Mws4U0Fp2SK5ORkZs+eTf/+/dm4MeXg0W233cbw4cOJioryOJ2EqX7AK6kedwBeM7NxwH+BG5xzu51zc83sbeAbUi6f8u80z3uIlAL0N2A30DqULonSuXNnJk6cSJkyZfj22281S6VkdxczDP4kMNM5twLAzF4DDptZYedc2qLVkyHzCQkJNG/enEWLFnHvvfcyadIkChVK+/YCJxSGRV+scMscbnkhPDNLeFPBKQF14sQJGjduzI8//sipU6coX748d955J6NGjdIlHOSyZHBt3AJp2g4BhpxjOzuB6MAlC5zu3bszYcIEzIwZM2ao2BSBrUBO/7Vyf/IvO9cw+HWAS/X47H3LxHwX7JdffqFy5cqcOHGCe++9lxkzZhAREeF1LBGRTKeCUwJi+fLlzJ49m1GjRnHkyBEiIiKYOnUqbdu21QeqSAaSk5O56667WLhwIUWKFGHJkiUqNkUA59xxM5sBvG5mXUmZpfY+4LZ0mv+TlGvrDiOlIO0PLHLOxQYt8DksW7aM6OhoTp06RZ06dfjss88yfpKISBahSYPkssyePZvSpUtz6623MmDAAOrVq8e8efNITEzk4YcfVrEpkoGkpCRKlizJwoULKVu2LLt371axKfJHPUkZHv8rKedkPuGc22hm9cws4Wwj59xC4GXgC3/bSkA7D/L+wWuvvUadOnU4deoUAwcOZOnSpV5HEhEJKh3hlIuWnJxM7969mT59OkePHgWgSJEijBw5kgcffNDjdCLh4+jRo1SvXp3Dhw9TqlQpfvrpJ83WLJKGc+4ocH86y7/nf4fTjwRCZrrXvn37MmjQIADGjh3LY4895nEiEZHgU8EpF+zEiRO8+OKLjBs3jpMnTwLQqFEjRowYQdWqVT1OJxJeVq1aRb169Th58iQPPPAAM2bM8DqSiARIcnIyzZo1Y8GCBeTPn5+vvvqKunXreh1LRMQTKjglQ8ePH+fWW29lxYoVOOfInTs3zZs3Z8yYMZQuXdrreCJhZ9euXdStW5fTp0/zxBNPMGLECK8jiUiAnD59mrJly3LgwAFKly7N+vXrKVq0qNexREQ8o4JT/ofP5+M///kP7733HldffTWzZ8/m+PHjRERE8MILL/DGG2+QM6d+dEQuxdq1a2nevDl58uShV69eDBmS7mS6IhKGYmNjadOmDQcOHKBMmTL8/PPPGiYvItmeqoZszufzsWrVKsaOHcvs2bO58sor2bt3L/HxKZc4y507N3fccQcDBgygTp06HqcVCW89evRg9OjRlCpVih9++IFq1ap5HUlEAmT69Ol069aNU6dO8c9//pPOnTt7HUlEJCSo4Mxmdu7cyccff8yXX37JDz/8QFJSEs79/2XLkpKS6NixI9WqVcPM6NKlC0uWLFGxKXKZpk6dyujRowGYMmWKik2RLGTp0qW0b9+e5ORkxowZo2JTRCQVFZxZ2MGDB/n66685evQoixYtYvr06f/Tpnjx4jRo0IBmzZpxzz33UKJECQ+SimRdPp+Pp59+mg8++ID69eszYcIEypUr53UsEQmQjz76iN69e1O2bFneeOMN2rdv73UkEZGQooIzizhz5gxjxoxh0aJFFC5cmOXLl7Ny5crf15csWZKCBQtSrlw5OnTowIMPPsi1117rYWKRrO/MmTOULVuW/fv306BBA+bOnUvevHm9jiUiAeDz+ahatSo///wzNWvWZN68eVx11VVexxIRCTkqOMPUypUree+991i+fDm//fYbv/322+/rChQoQO3atbnpppu47rrrePvtt7n66qsxMw8Ti2QvsbGx1KhRg/3791OiRAnmz5+vyUNEsoiEhASioqLYtm0b+fPnZ86cOSo2RUTOQQVnGIiLi2PlypW8/vrrrFixgsTERHw+3+/r8+XLR/Xq1alYsSLNmjWjW7duREREeJhYJHtbt24ddevW5cSJE7Ro0YLZs2eTI0cOr2OJSACsXLmS6Ohojh8/TnR0NF9//bX6t4jIeajgDDGHDh1i06ZNrFu3jn/84x9s2LDhf9pce+211KpVi3r16vHggw9SsmRJD5KKSHpWrFhB3bp1SU5OpmPHjkycONHrSCISIF999RVNmzbFOUfXrl356KOPvI4kIhLyVHB6KCkpiXHjxrFs2TJWrVrF1q1bSUxM/H19wYIFyZMnD506daJt27bUrFlTF48WCWGLFi3i1VdfJTk5md69ezNs2DCvI4lIgMycOZMPPvgAgGeffZb33nvP40QiIuFBBWcQrV+/nv79+5OQkMCuXbv+cN4lQKFChShatChdunTh8ccf13mXImHk6aef5v333ydHjhxMnz6dtm3beh1JRAKkUaNGfPPNN+TKlYu5c+fSqFEjryOJiIQNFZxBdN9997Fjxw4g5bzLG264gWLFivHYY4/Rrl07TSgiEqYOHDjA8OHDAfj3v//N/fff73EiEQmU7777jm+++YaIiAjWrFmja+iKiFwkFZxBNGrUKKZOncqgQYN03qVIFlKyZEmmTZtGcnKyik2RLKZevXqMGTOGEiVKqNgUEbkEmlYtiO666y46d+6sYlMkC2rdujWRkZFexxCRADMzunTpQsGCBb2OIiISllRwioiIiIiISKZQwSkiIiIiIiKZQgWniIiIiIiIZAoVnCIiIiIiIpIpVHCKiIiIiIhIplDBKSIiIiIiIpnCnHNeZ8gyzOwQsCuDZlcBh4MQJ5CUOTiyauZrnXPFgxEmM6l/hxRlDg717z/Kqv+HoSbcModbXshGfVtCgwrOIDOzlc65Wl7nuBjKHBzKHP7CcX8oc3Aoc/gLx/2hzJkv3PJCeGaW8KYhtSIiIiIiIpIpVHCKiIiIiIhIplDBGXyjvQ5wCZQ5OJQ5/IXj/lDm4FDm8BeO+0OZM1+45YXwzCxhTOdwioiIiIiISKbQEU4RERERERHJFCo4RUREREREJFOo4AwAMytqZjPN7LiZ7TKzdudo96qZnTazhFS3CqnWR5nZKjM74f83yuO8c9JkTTKz9anW7zSzk6nWz8+MvP7X6mVmK80s0czGZ9D2GTM7YGZxZjbOzPKkWlfOzL7x7+PNZnan15nN7BH///cxM9trZm+bWc5U62PM7FSq/bwlBDJ3NrPkND8f0anWB20/Z6Zw69sXmVn9OwiZ1b9Dl/q3+vflZlb/FrkwKjgDYziQBEQC7YGRZlbtHG2nO+cKpLptBzCz3MBnwGSgCDAB+My/3JO8zrnmqbMCS4BP0jS7J1WbuzIh61m/AG8C487XyMyaAi8BjYFyQAXgtVRNpgFrgGLAX4FPzSyzLmx8QZmBfMDTpFyI+VZSsj+fpk2vVPu5asCT/r8LzQywNM3PckyqdcHcz5kp3Pr2BWdW/75s6t/q3+rfF0b9W/1bsjkVnJfJzPIDrYD+zrkE59wiYDbQ8SI3FQ3kBP7unEt0zg0DDGgUCnnNrBxQD5gUyDwXyjk3wzk3CziSQdNHgLHOuY3Oud+AN4DOAGZWBagJvOKcO+mc+zewnpT94Vlm59xI59z3zrkk59w+YApwe2ZkyshF7OdzCvZ+zizh1rdB/Rv17/NS//5/6t/Bo/4dHOrfEspUcF6+KkCyc25rqmVrgXN9S3qPmR01s41m9kSq5dWAdc79YdrgdefZzqW62LxndQK+d87tSLN8ipkdMrP5ZvanQAa9RNVIeT9nrQUizayYf91251x8mvWB3seXqz6wMc2yt8zssJktTj30xWM3+TNtNbP+qYYRhct+zki49W1Q/w6Hnzv179Cg/q3+nRnUv0XSoYLz8hUA4tIsiwMKptP2Y+B6oDjQDfibmT18Cdu5HJf6Op2A8WmWtSdl2Mu1wDfAPDO78vIjXpa07+/s/YLprDu7PtD7+JKZ2aNALeDdVIv7kDK06GpSrp31uZlV9CBeat8B1YESpHzz+TDwgn9dyO/nCxRufftyXkv9OwjUv0OK+rf6d0Cpf4ucmwrOy5cAFEqzrBAQn7ahc+6/zrlfnHPJzrklwPtA64vdTrDynmVmdwAlgU9TL3fOLfYPuTjhnHsLiCVl2I6X0r6/s/fj01l3dn2g9/ElMbP7gUFAc+fc4bPLnXPLnHPx/uFaE4DFwN1e5fRn2u6c2+Gc8znn1gOvE/yf5cwWbn37kl5L/Ts41L9Djvq3+nfAqH+LnJ8Kzsu3FchpZpVTLfsT/zukIj2OlHM98LevYWaWan2NC9zOxbiUvI8AM5xzCRlsO/X78cpGUt7PWX8CDjrnjvjXVTCzgmnWB3ofXzQzawZ8RMokDuszaB4K+zmttD/LIbmfL1K49W1Q/w7Jnzv175Ck/v1HofBzp/7tjazYvyXUOOd0u8wb8C9SZvXKT8rJ4nFAtXTa3UfKLHYG1Ab2AY/41+UGdgFPAXmAXv7Hub3K6297BSnffDZKs7ys/7m5gbykDMc4BBTLpH2c0/86b5Ey8UFeIGc67ZoBB4Ab/Pt6ITAo1fofSBnukhd4wP/einucuREpJ/nXT2fdlUDTs88lZRjUcaCqx5mbA5H++9cBG0iZZCDo+zkzb+HWty8ms7+t+nfmZ1b/DtGb+rf6dwAyq3/rptsF3DwPkBVuQFFglv8XyW6gnX95PSAhVbtp/l9MCcBm4Mk027kJWAWcBFYDN3mZ17/sYf+Hp6VZXo2UiRGO+9/T10CtTNzHr5LyLVzq26v+D84EoGyqts8CB4FjwD+BPKnWlQNi/Pt4C3Cn15lJOX/mjH/Z2dsc/7riwApShrPE+j8ImoRA5nf9+/g4sJ2UITm5vNjPmXkLt759MZn9y9S/Mzmz+nfo3tS/1b8vN7P6t266XdjNnHOIiIiIiIiIBJrO4RQREREREZFMoYJTREREREREMoUKThEREREREckUKjhFREREREQkU6jgFBERERERkUyhglNEREREREQyhQpOEZEwZmbOzCoFaFsxZtb1AttGm9neQLxuZjKzzma2KIiv9/s+NLP2ZjY/k16nqpmtMbN4M3syM15DREQkEFRwiohcIDPbaWYnzSzBzA6a2T/NrMAlbiuohZAEn3NuinPurkza/ItAjHOuoHNu2KVuxMzGm9mbAcwlIiLyByo4RUQuzj3OuQJATeAWoN/FbsDMcl5uiEBsQ8LatcBGr0OYWYTXGUREJLSp4BQRuQTOuX3AHKA6gJkVNrOxZrbfzPaZ2Ztn/xj3H81cbGZDzewoMB0YBdT1Hy2N9bf7w5DWtEdB/cNn/2JmPwE/pYpzt5ltN7PDZvaOmeVI9ZzHzGyTmf1mZvPM7NpU65qY2WYzizOzDwE71/s1syv8R8N+M7P/klJsp15f2sz+bWaHzGxH6mGeZvaqmX1iZpP9Q0DXm1kVM+trZr+a2R4zuytV+0f9meP97+vxVOuizWyvmT3nf+5+M3s01fpiZjbbzI6Z2XKg4nneU15/piNmFmtmK8ws0r+uqP8I9i/+9zzLv7yImf3H/z5/89+/5hzbT+//r4eZ/eR/7nAzM/+6CDN7z/9/uMPMevnb/88XC2a2EGgIfOj/+aliZi38Q2yP+ffnq2mec4eZLfG/zz3+bN2B9sCL/u187m97vf9nMdbMNprZvam2M97MRprZl2Z23J9DRETknFRwiohcAjMrA9z9f+3dfajeZR3H8ffXNcWDZ57ZVnPrbJHWYlNckS2KniPbcrUWjh4nKaGFukGUVGA50GYkU0ypTErmcpjmQ21liDXMnkBaKirBHHZqZ9OzB47p1h789Mf3+tW1e+e+O7fHw4R9XnDgvn+/3/Xwe4DD9/e9rusG/lI23QwcAE4F3gR8EKjnQ84HngReBXwGuBD4g6QTJPV10fTiUtecatvHgLeQWdePAueVPi4GvgYsAaYCDwC3ln1TgDvIDO0UYDPwjg7tfoMM3k4BzgLObXaUAPfnwF+BGcD7gRURcVZVfhGwBphMXrN7yf9BM4CVwPerY58GzgYmAZ8DVkfEm6v904ATS9nzgesjYnLZdz2wFzi5XIfzOpzTuaWefuCV5D3ZU/atAXqAueQ9W122HwP8iMwwzizHf7dDG63OJoP1M4Cl5LUE+DywAJhH3sfF7SqQ9D7yXl5Unp+/Ac8By4A+4MPAF8r9JyJmki9HriOfg3nAJkk/ANYC3y71LIqIieS9/HU574uBtRExu+rCp4ArgF7Aw8LNzKwjB5xmZt25q2QkfwdsBK4sWbEFwApJz0l6mgxQPlGV2yrpOkkHJO05vNpR+5aknS11XFW2/R24Bvhk2X5BOf5xSQeAK4F5Jcu5EHhM0u2S9pdy2zq0uxS4orQzANTzBs8EpkpaKWmfpCeBG1vO/wFJ95Z+/JQMfFaVttcBr42IPgBJ6yVtVtpIBj/vrOraD6yUtF/SBuBfwOzIjPLHgcvKfXiUfBHQzn4y0DxV0kFJD0kajoiTyft5oaRdpZ2NpW87JN0h6XlJz5KB17s7tNFqlaTd5V79hgz+mut7raR/SNoFrOqiTiT9VtIjkl6Q9DD5YqHp16eB+yTdWs5lh6RNbap6G3BC6ec+SfcDv+B/zxTA3ZIeLG3t7aafZmZ29PEcIDOz7iyWdF+9ISJOByYCg2WEJOQLvYHqsPrzWIxUT73tKWB6+TwLuDYirq72B5kZnF6Xk6SI6NTHQ44v7TRmAdNLIN6YQGbhGturz3uAIUkHq++Qgc7uiFhAZlTfQF7HHuCRqvyOErg2ni9lp5L/19r1s9UaMru5rgS7twBfL9t2lsDvEBHRQ75M+BCZrQXojYgJ1fl0Ugf1Tb/h8Ovb1fMSEfPJIPU04FjgODKwhzyfzaOsajowIOmFattT5DPzovpmZmZHN2c4zczGbgD4NzBFUl/5myRpbnWMWsq0foccFtlTfZ82wjEjleuvPs8Etlb9uqDqU5+k4yX9Hhisy5W5hHU9rQY5vJ3GALClpZ1eSQs71DeiiDiOHOr7HeDVZbjxBjrML608Qw5rbtfPQ5Rs3+WS5gBvJ4e7Livnc1KTcW3xJWA2MF/SJOBdTddH0b9OBoF6LminezGSnwD3AP2STiTnCDd9GqD9XNbW52kr0B/VPGDyGv6zQxkzM7O2HHCamY2RpEFy2OfVETEpIo6JiFMiotNQy+3AayLi2GrbJmBJRPRE/rbm+aPswpfLYjb9wHJyUSLIoOOrETEX/ruw0Tll33pgbkQsKQvTXMLIAW7jtlLX5LJIzsXVvj8DwxFxaeTiQhMi4rSIOHPkqjpqsnPPAAdKtnNUPy1SMow/A75ZruEcqrmmrSLivRFxehmKO0wOsT1Y7ucvgRvK+U6MiCaw7CUzsrsj4iQyE/tSuA1YHhEzSqB7aZfle8ms7N6IeCs5z7KxFvhARCyNiFdELqzUDOXdDryuOvZP5IuPr5Tzfg85/3bdizgnMzMzB5xmZi+RZWSw9BiwC7idXLimnfvJn7XYFhFDZdtqYB8ZBNxMBgqjcTfwEBmwrgduApB0J3AVOWR0GHiUnJuIpCHgHHIY5g7g9cCDHdq4nBxauYUMrtc0O0qgt4icj7gFGAJ+SC7I05UyL/ISMgDbRQZO93RRxUXkMNVtwI/JBX7amUbep2HgcXJO7i1l32fJAPQJchGjFWX7NcDx5Dn+EfhVF33r5Ebyuj5MLqq0gczWjmaYLsAXgZUR8SxwGXn9ACjzRReS2dmd5HNyRtl9EzCnrEh7l6R9wEfI52QIuAFYJumJsZ2emZkdrULyyBgzM7OXk5LZ/Z6kWf/3YDMzs5cxZzjNzMyOsDIUeWEZ8jqDHKp755Hul5mZ2Vg5w2lmZnaEldVvNwJvJOeIrgeWSxo+oh0zMzMbIwecZmZmZmZmNi48pNbMzMzMzMzGhQNOMzMzMzMzGxcOOM3MzMzMzGxcOOA0MzMzMzOzceGA08zMzMzMzMaFA04zMzMzMzMbF/8BXufaEHuuIzkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 936x864 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def partial_perturb_plotter(df, net, plotFilename=None):    \n",
    "    means = np.append(df['mean'].unique(), 1)\n",
    "    means.sort()\n",
    "    ps = df['p'].unique()\n",
    "    ps.sort()\n",
    "    methods = df['method'].unique()\n",
    "    methods.sort()\n",
    "    methods = np.flip(methods)\n",
    "    \n",
    "    linestyles = ['-', ':', '-.']\n",
    "    \n",
    "    fig, axes = plt.subplots(nrows=len(ps), ncols=3, sharex=True, sharey='col', figsize=(13,len(ps)*3))\n",
    "    \n",
    "    axes[0][0].set_title('Normalized TSTT')\n",
    "    axes[0][1].set_title('WVC')\n",
    "    axes[0][2].set_title('Normalized VMT')\n",
    "    for i,p in enumerate(ps):\n",
    "        true_row = dfmeans[dfmeans['mean']==1].copy()\n",
    "        true_row['p'] = p\n",
    "        for j,method in enumerate(methods):\n",
    "            true_row['method'] = method\n",
    "            df = df.append(true_row, sort=False)\n",
    "            df.sort_values('mean', inplace=True)\n",
    "            axes[i][0].plot(means, df.loc[(df['p']==p) & (df['method']==method),'tstt']/true_tstt,\n",
    "                                           linestyle=linestyles[j], c='k', label=method)\n",
    "            axes[i][1].plot(means, df.loc[(df['p']==p) & (df['method']==method),'wvc'], linestyle=linestyles[j], c='k', label=method)\n",
    "            axes[i][2].plot(means, df.loc[(df['p']==p) & (df['method']==method),'vmt']/true_vmt,\n",
    "                                           linestyle=linestyles[j], c='k', label=method)\n",
    "            axes[i][0].ticklabel_format(style='sci', scilimits=(0,0))\n",
    "            axes[i][1].ticklabel_format(style='sci', scilimits=(0,0))\n",
    "            axes[i][2].ticklabel_format(style='sci', scilimits=(0,0))\n",
    "            axes[i][0].grid()\n",
    "            axes[i][1].grid()\n",
    "            axes[i][2].grid()\n",
    "            axes[i][0].set_ylabel('p={:.0%}'.format(round(p,2)), rotation=0, size='large', labelpad=30)\n",
    "    axes[0][2].legend(title='Perturbation method', loc='center left', bbox_to_anchor=(1,0.5))\n",
    "    \n",
    "#     fig.suptitle('Proportion of nodes perturbed vs. TSTT and weighted V/C\\n[{}]'.format(net))\n",
    "#     fig.text(0.5, 0.01, 'Perturbed demand scaling factor', ha='center', fontsize=14)\n",
    "    axes[len(ps)-1,1].set_xlabel('Perturbed demand scaling factor')\n",
    "    fig.tight_layout()\n",
    "#     fig.subplots_adjust(bottom=0.05)\n",
    "    if plotFilename:\n",
    "        plt.savefig(plotFilename, dpi='figure')\n",
    "    plt.show()\n",
    "\n",
    "partial_perturb_plotter(df_partial, net, plotFilename='output/plots/{}_partial_perturb.png'.format(net))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sectional perturbation vs. global perturbation with matching demand\n",
    "\n",
    "Suppose we perturb the 50% of nodes closest to the center of a net, then compare with an OD matrix that has been globally scaled to match the demand. How do the output metrics work out?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>proportion</th>\n",
       "      <th>demand</th>\n",
       "      <th>tstt_partial</th>\n",
       "      <th>tstt_comp</th>\n",
       "      <th>wvc_partial</th>\n",
       "      <th>wvc_comp</th>\n",
       "      <th>vmt_partial</th>\n",
       "      <th>vmt_comp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.25</td>\n",
       "      <td>0.25</td>\n",
       "      <td>244650.0</td>\n",
       "      <td>3.639556e+06</td>\n",
       "      <td>2.997721e+06</td>\n",
       "      <td>1.138362</td>\n",
       "      <td>1.052570</td>\n",
       "      <td>2.350273e+06</td>\n",
       "      <td>2.243948e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.25</td>\n",
       "      <td>0.50</td>\n",
       "      <td>183375.0</td>\n",
       "      <td>2.338017e+06</td>\n",
       "      <td>1.913946e+06</td>\n",
       "      <td>0.939877</td>\n",
       "      <td>0.839818</td>\n",
       "      <td>1.783363e+06</td>\n",
       "      <td>1.652788e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.25</td>\n",
       "      <td>0.75</td>\n",
       "      <td>132450.0</td>\n",
       "      <td>1.411532e+06</td>\n",
       "      <td>1.265175e+06</td>\n",
       "      <td>0.697063</td>\n",
       "      <td>0.656492</td>\n",
       "      <td>1.268043e+06</td>\n",
       "      <td>1.181013e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.25</td>\n",
       "      <td>1.00</td>\n",
       "      <td>90150.0</td>\n",
       "      <td>8.277958e+05</td>\n",
       "      <td>8.277958e+05</td>\n",
       "      <td>0.514646</td>\n",
       "      <td>0.514646</td>\n",
       "      <td>7.961663e+05</td>\n",
       "      <td>7.961663e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.50</td>\n",
       "      <td>0.25</td>\n",
       "      <td>283300.0</td>\n",
       "      <td>4.429737e+06</td>\n",
       "      <td>4.052286e+06</td>\n",
       "      <td>1.228819</td>\n",
       "      <td>1.195250</td>\n",
       "      <td>2.690839e+06</td>\n",
       "      <td>2.617764e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean  proportion    demand  tstt_partial     tstt_comp  wvc_partial  \\\n",
       "0  0.25        0.25  244650.0  3.639556e+06  2.997721e+06     1.138362   \n",
       "1  0.25        0.50  183375.0  2.338017e+06  1.913946e+06     0.939877   \n",
       "2  0.25        0.75  132450.0  1.411532e+06  1.265175e+06     0.697063   \n",
       "3  0.25        1.00   90150.0  8.277958e+05  8.277958e+05     0.514646   \n",
       "4  0.50        0.25  283300.0  4.429737e+06  4.052286e+06     1.228819   \n",
       "\n",
       "   wvc_comp   vmt_partial      vmt_comp  \n",
       "0  1.052570  2.350273e+06  2.243948e+06  \n",
       "1  0.839818  1.783363e+06  1.652788e+06  \n",
       "2  0.656492  1.268043e+06  1.181013e+06  \n",
       "3  0.514646  7.961663e+05  7.961663e+05  \n",
       "4  1.195250  2.690839e+06  2.617764e+06  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def progressive_partial_perturbation(netFilename, tripsFilename, nodeFilename, true_demand, net, verbose=False, saveplot=False):\n",
    "    df = pd.DataFrame(columns=['mean', 'proportion', 'demand', 'tstt_partial', 'tstt_comp', 'wvc_partial', 'wvc_comp', 'vmt_partial', 'vmt_comp'])\n",
    "    \n",
    "    i = 0\n",
    "    for m in np.arange(.25, 2, .25):\n",
    "        for p in np.arange(.25, 1.1, .25):\n",
    "            nodes_to_perturb = central_node_selector(nodeFilename, p, make_plot=False)\n",
    "            # TODO: determine whether to incorporate VMT, % within 1%, most congested link if\n",
    "            dfp, tsttp, wvcp, vmtp, demand, _ = tap_trial(netFilename, tripsFilename, norm_mean=m, \n",
    "                                                       norm_sd=0, returnDataFrame=True, returnODinfo=True, \n",
    "                                                       nodesPerturbedIfOrig=nodes_to_perturb)\n",
    "            dfc, tsttc, wvcc, vmtc = tap_trial(netFilename, tripsFilename, norm_mean=demand/true_demand,\n",
    "                                            norm_sd=0, returnDataFrame=True, returnODinfo=False)\n",
    "            df.loc[i] = {\n",
    "                'mean': m,\n",
    "                'proportion': p,\n",
    "                'demand': demand,\n",
    "                'tstt_partial': tsttp,\n",
    "                'tstt_comp': tsttc,\n",
    "                'wvc_partial': wvcp,\n",
    "                'wvc_comp': wvcc,\n",
    "                'vmt_partial': vmtp,\n",
    "                'vmt_comp': vmtc\n",
    "            }\n",
    "            i += 1\n",
    "            \n",
    "            if saveplot or i==1:\n",
    "                ((dfp['flow']-dfc['flow'])/dfc['capacity']).hist(density=True)\n",
    "                plt.title('Link-by-link difference in V/C between a {}x\\npartial scaling and a commensurate global scaling\\n[{}, p={}]'.format(m, net, p))\n",
    "                plt.xlabel('V/C difference')\n",
    "                plt.ylabel('Density of links')\n",
    "                plt.grid(b=None)\n",
    "                if saveplot:\n",
    "                    plt.savefig('output/plots/{}_{}x_p{}.png'.format(net, m, p), dpi='figure')\n",
    "                if verbose or i==1:\n",
    "                    plt.show()\n",
    "                else:\n",
    "                    plt.close()\n",
    "    df.to_csv('output/data/{}_ppp.csv'.format(net), index=False)\n",
    "    \n",
    "    return df\n",
    "\n",
    "try:\n",
    "    df_ppp = pd.read_csv('output/data/{}_ppp.csv'.format(net))\n",
    "except FileNotFoundError:\n",
    "    df_ppp = progressive_partial_perturbation(netFilename, tripsFilename, nodeFilename, true_demand, net, verbose=False, saveplot=False)\n",
    "df_ppp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def ppp_plotter(df_ppp, net):\n",
    "    fig, axes = plt.subplots(ncols=3, figsize=(13,3))\n",
    "    props = df_ppp['proportion'].unique()\n",
    "    props.sort()\n",
    "    means = df_ppp['mean'].unique()\n",
    "    means.sort()\n",
    "    cm = plt.get_cmap('winter')\n",
    "    linestyles = ['-', ':', '--', '-.']\n",
    "    for i,p in enumerate(props):\n",
    "        dftemp = df_ppp[df_ppp['proportion']==p].sort_values(by=['mean'])\n",
    "        axes[0].plot(means, (dftemp['tstt_partial']-dftemp['tstt_comp'])/true_tstt, c='k', linestyle=linestyles[i], label=p)\n",
    "        axes[1].plot(means, (dftemp['wvc_partial']-dftemp['wvc_comp'])/true_wvc, c='k', linestyle=linestyles[i], label=p)\n",
    "        axes[2].plot(means, (dftemp['vmt_partial']-dftemp['vmt_comp'])/true_vmt, c='k', linestyle=linestyles[i], label=p)\n",
    "    axes[0].set_ylabel('Difference relative to true TSTT')\n",
    "    axes[0].set_xlabel('Demand scaling factor')\n",
    "    axes[0].grid()\n",
    "    axes[1].set_ylabel('Difference relative to true WVC')\n",
    "    axes[1].set_xlabel('Demand scaling factor')\n",
    "    axes[1].grid()\n",
    "    axes[2].set_ylabel('Difference relative to true VMT')\n",
    "    axes[2].set_xlabel('Demand scaling factor')\n",
    "    axes[2].grid()\n",
    "#     fig.suptitle('Difference between partial and global O-D matrix perturbation\\n[{}]'.format(net))\n",
    "    axes[2].legend(title='Proportion\\nof nodes\\nperturbed', loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    fig.tight_layout()\n",
    "#     fig.subplots_adjust(top=0.8)\n",
    "    plt.savefig('output/plots/{}_partdiff.png'.format(net), dpi='figure')\n",
    "    plt.show()\n",
    "\n",
    "ppp_plotter(df_ppp, net)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}