running_metrics.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import os\n",
    "import json\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import torch\n",
    "from sklearn.metrics import r2_score\n",
    "from tqdm import tqdm\n",
    "\n",
    "from datatools.trace_helper import TraceHelper\n",
    "from models import load_model"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculating and Comparing Running Performance Metrics in a Match"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'dataset': 'soccer',\n",
       " 'model': 'macro_pi_lstm',\n",
       " 'target_type': 'ball',\n",
       " 'macro_type': 'team_poss',\n",
       " 'macro_weight': 50.0,\n",
       " 'phys_weight': 50.0,\n",
       " 'speed_loss': False,\n",
       " 'prev_out_aware': True,\n",
       " 'bidirectional': True,\n",
       " 'flip_pitch': True,\n",
       " 'n_features': 6,\n",
       " 'batch_size': 1792,\n",
       " 'start_lr': 0.0005,\n",
       " 'min_lr': 1e-05,\n",
       " 'seed': 100,\n",
       " 'cuda': True,\n",
       " 'best_loss': 0,\n",
       " 'type_aware': False,\n",
       " 'pred_delta': False,\n",
       " 'fix_dec_std': False,\n",
       " 'normalize': True,\n",
       " 'embed_dim': 128,\n",
       " 'macro_rnn_dim': 32,\n",
       " 'micro_rnn_dim': 256,\n",
       " 'n_layers': 2,\n",
       " 'dropout': 0,\n",
       " 'total_params': 3493128}"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trial = 707\n",
    "device = \"cuda:0\"\n",
    "save_path = f\"saved/{trial:03d}\"\n",
    "\n",
    "with open(f\"{save_path}/params.json\", \"r\") as f:\n",
    "    params = json.load(f)\n",
    "    if \"dropout\" not in params:\n",
    "        params[\"dropout\"] = 0\n",
    "    if \"speed_loss\" not in params:\n",
    "        params[\"speed_loss\"] = False\n",
    "\n",
    "state_dict = torch.load(\n",
    "    f\"{save_path}/model/{params['model']}_state_dict_best.pt\",\n",
    "    map_location=lambda storage, _: storage,\n",
    ")\n",
    "model = load_model(params[\"model\"], params).to(device)\n",
    "model.load_state_dict(state_dict)\n",
    "\n",
    "params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# match_traces = pd.read_csv(\"data/metrica_traces/match3_test.csv\", header=0)\n",
    "match_traces = pd.read_csv(\"data/gps_event_traces_gk_pred/22213-22216.csv\", header=0)\n",
    "helper = TraceHelper(match_traces)\n",
    "macro_pred, micro_pred = helper.predict(model, split=False, evaluate=True)\n",
    "macro_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "players = helper.team1_players + helper.team2_players\n",
    "cols = [\"total_dur\", \"total_dist\", \"hsr_dist\"] + [f\"zone{i}_dist\" for i in np.arange(5) + 1]\n",
    "speed_bins = [-0.1, 7, 15, 20, 25, 50]\n",
    "\n",
    "true_att_stats = pd.DataFrame(columns=cols)\n",
    "true_def_stats = pd.DataFrame(columns=cols)\n",
    "\n",
    "pred_att_stats = pd.DataFrame(columns=cols)\n",
    "pred_def_stats = pd.DataFrame(columns=cols)\n",
    "\n",
    "baseline_att_stats = pd.DataFrame(columns=cols)\n",
    "baseline_def_stats = pd.DataFrame(columns=cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calc_running_metrics(traces: pd.DataFrame, team_poss: pd.Series, p: str, attacking=True):\n",
    "    if attacking:\n",
    "        speeds = traces.loc[(team_poss.notna()) & (team_poss == p[0]), f\"{p}_speed\"]\n",
    "        total_dur = len(traces.loc[(team_poss == p[0]) & (speeds.notna())]) * 0.1\n",
    "    else:\n",
    "        speeds = traces.loc[(team_poss.notna()) & (team_poss != p[0]), f\"{p}_speed\"]\n",
    "        total_dur = len(traces.loc[(team_poss != p[0]) & (speeds.notna())]) * 0.1\n",
    "\n",
    "    speed_cats = pd.cut(speeds * 3.6, bins=speed_bins, labels=np.arange(5) + 1)\n",
    "    zone_dists = (speeds.groupby(speed_cats).sum() * 0.1).round(1).values\n",
    "    total_dist = zone_dists.sum()\n",
    "    hsr_dist = zone_dists[-2:].sum()\n",
    "\n",
    "    return np.array([total_dur, total_dist, hsr_dist] + zone_dists.tolist())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 27/27 [00:04<00:00,  6.43it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Physical stats when attacking:\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\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>total_dur</th>\n",
       "      <th>total_dist</th>\n",
       "      <th>hsr_dist</th>\n",
       "      <th>zone1_dist</th>\n",
       "      <th>zone2_dist</th>\n",
       "      <th>zone3_dist</th>\n",
       "      <th>zone4_dist</th>\n",
       "      <th>zone5_dist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A02</th>\n",
       "      <td>1647.7</td>\n",
       "      <td>3054.0</td>\n",
       "      <td>158.6</td>\n",
       "      <td>1182.0</td>\n",
       "      <td>1379.6</td>\n",
       "      <td>333.8</td>\n",
       "      <td>124.5</td>\n",
       "      <td>34.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A04</th>\n",
       "      <td>1755.6</td>\n",
       "      <td>2787.5</td>\n",
       "      <td>81.6</td>\n",
       "      <td>1239.9</td>\n",
       "      <td>1230.5</td>\n",
       "      <td>235.5</td>\n",
       "      <td>64.2</td>\n",
       "      <td>17.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A06</th>\n",
       "      <td>1755.6</td>\n",
       "      <td>3497.1</td>\n",
       "      <td>40.4</td>\n",
       "      <td>1185.9</td>\n",
       "      <td>1751.7</td>\n",
       "      <td>519.1</td>\n",
       "      <td>40.4</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A09</th>\n",
       "      <td>1755.6</td>\n",
       "      <td>2886.6</td>\n",
       "      <td>240.1</td>\n",
       "      <td>1234.3</td>\n",
       "      <td>1057.0</td>\n",
       "      <td>355.2</td>\n",
       "      <td>173.8</td>\n",
       "      <td>66.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A10</th>\n",
       "      <td>1755.6</td>\n",
       "      <td>2814.8</td>\n",
       "      <td>270.4</td>\n",
       "      <td>1224.7</td>\n",
       "      <td>1040.7</td>\n",
       "      <td>279.0</td>\n",
       "      <td>198.3</td>\n",
       "      <td>72.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_dur  total_dist  hsr_dist  zone1_dist  zone2_dist  zone3_dist  \\\n",
       "A02     1647.7      3054.0     158.6      1182.0      1379.6       333.8   \n",
       "A04     1755.6      2787.5      81.6      1239.9      1230.5       235.5   \n",
       "A06     1755.6      3497.1      40.4      1185.9      1751.7       519.1   \n",
       "A09     1755.6      2886.6     240.1      1234.3      1057.0       355.2   \n",
       "A10     1755.6      2814.8     270.4      1224.7      1040.7       279.0   \n",
       "\n",
       "     zone4_dist  zone5_dist  \n",
       "A02       124.5        34.1  \n",
       "A04        64.2        17.4  \n",
       "A06        40.4         0.0  \n",
       "A09       173.8        66.3  \n",
       "A10       198.3        72.1  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_dur</th>\n",
       "      <th>total_dist</th>\n",
       "      <th>hsr_dist</th>\n",
       "      <th>zone1_dist</th>\n",
       "      <th>zone2_dist</th>\n",
       "      <th>zone3_dist</th>\n",
       "      <th>zone4_dist</th>\n",
       "      <th>zone5_dist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A02</th>\n",
       "      <td>1554.9</td>\n",
       "      <td>3027.9</td>\n",
       "      <td>163.8</td>\n",
       "      <td>1091.3</td>\n",
       "      <td>1406.7</td>\n",
       "      <td>366.1</td>\n",
       "      <td>134.3</td>\n",
       "      <td>29.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A04</th>\n",
       "      <td>1669.8</td>\n",
       "      <td>2775.8</td>\n",
       "      <td>72.7</td>\n",
       "      <td>1242.8</td>\n",
       "      <td>1260.2</td>\n",
       "      <td>200.1</td>\n",
       "      <td>56.0</td>\n",
       "      <td>16.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A06</th>\n",
       "      <td>1669.8</td>\n",
       "      <td>3508.1</td>\n",
       "      <td>50.0</td>\n",
       "      <td>1183.3</td>\n",
       "      <td>1757.6</td>\n",
       "      <td>517.2</td>\n",
       "      <td>50.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A09</th>\n",
       "      <td>1669.8</td>\n",
       "      <td>2943.0</td>\n",
       "      <td>244.5</td>\n",
       "      <td>1180.8</td>\n",
       "      <td>1146.9</td>\n",
       "      <td>370.8</td>\n",
       "      <td>175.3</td>\n",
       "      <td>69.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A10</th>\n",
       "      <td>1669.8</td>\n",
       "      <td>2826.4</td>\n",
       "      <td>287.4</td>\n",
       "      <td>1178.3</td>\n",
       "      <td>1080.6</td>\n",
       "      <td>280.1</td>\n",
       "      <td>211.7</td>\n",
       "      <td>75.7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_dur  total_dist  hsr_dist  zone1_dist  zone2_dist  zone3_dist  \\\n",
       "A02     1554.9      3027.9     163.8      1091.3      1406.7       366.1   \n",
       "A04     1669.8      2775.8      72.7      1242.8      1260.2       200.1   \n",
       "A06     1669.8      3508.1      50.0      1183.3      1757.6       517.2   \n",
       "A09     1669.8      2943.0     244.5      1180.8      1146.9       370.8   \n",
       "A10     1669.8      2826.4     287.4      1178.3      1080.6       280.1   \n",
       "\n",
       "     zone4_dist  zone5_dist  \n",
       "A02       134.3        29.5  \n",
       "A04        56.0        16.7  \n",
       "A06        50.0         0.0  \n",
       "A09       175.3        69.2  \n",
       "A10       211.7        75.7  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_dur</th>\n",
       "      <th>total_dist</th>\n",
       "      <th>hsr_dist</th>\n",
       "      <th>zone1_dist</th>\n",
       "      <th>zone2_dist</th>\n",
       "      <th>zone3_dist</th>\n",
       "      <th>zone4_dist</th>\n",
       "      <th>zone5_dist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A02</th>\n",
       "      <td>1417.4</td>\n",
       "      <td>2744.5</td>\n",
       "      <td>195.4</td>\n",
       "      <td>957.6</td>\n",
       "      <td>1166.6</td>\n",
       "      <td>424.9</td>\n",
       "      <td>168.7</td>\n",
       "      <td>26.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A04</th>\n",
       "      <td>1561.3</td>\n",
       "      <td>2559.7</td>\n",
       "      <td>141.2</td>\n",
       "      <td>1079.5</td>\n",
       "      <td>1082.3</td>\n",
       "      <td>256.7</td>\n",
       "      <td>95.6</td>\n",
       "      <td>45.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A06</th>\n",
       "      <td>1561.3</td>\n",
       "      <td>3141.2</td>\n",
       "      <td>182.1</td>\n",
       "      <td>1046.2</td>\n",
       "      <td>1430.1</td>\n",
       "      <td>482.8</td>\n",
       "      <td>162.9</td>\n",
       "      <td>19.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A09</th>\n",
       "      <td>1561.3</td>\n",
       "      <td>2575.2</td>\n",
       "      <td>190.7</td>\n",
       "      <td>1079.6</td>\n",
       "      <td>1002.1</td>\n",
       "      <td>302.8</td>\n",
       "      <td>130.0</td>\n",
       "      <td>60.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A10</th>\n",
       "      <td>1561.3</td>\n",
       "      <td>2570.3</td>\n",
       "      <td>233.1</td>\n",
       "      <td>1049.1</td>\n",
       "      <td>1017.7</td>\n",
       "      <td>270.4</td>\n",
       "      <td>181.5</td>\n",
       "      <td>51.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_dur  total_dist  hsr_dist  zone1_dist  zone2_dist  zone3_dist  \\\n",
       "A02     1417.4      2744.5     195.4       957.6      1166.6       424.9   \n",
       "A04     1561.3      2559.7     141.2      1079.5      1082.3       256.7   \n",
       "A06     1561.3      3141.2     182.1      1046.2      1430.1       482.8   \n",
       "A09     1561.3      2575.2     190.7      1079.6      1002.1       302.8   \n",
       "A10     1561.3      2570.3     233.1      1049.1      1017.7       270.4   \n",
       "\n",
       "     zone4_dist  zone5_dist  \n",
       "A02       168.7        26.7  \n",
       "A04        95.6        45.6  \n",
       "A06       162.9        19.2  \n",
       "A09       130.0        60.7  \n",
       "A10       181.5        51.6  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Physical stats when defending:\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>total_dur</th>\n",
       "      <th>total_dist</th>\n",
       "      <th>hsr_dist</th>\n",
       "      <th>zone1_dist</th>\n",
       "      <th>zone2_dist</th>\n",
       "      <th>zone3_dist</th>\n",
       "      <th>zone4_dist</th>\n",
       "      <th>zone5_dist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A02</th>\n",
       "      <td>1204.2</td>\n",
       "      <td>2468.1</td>\n",
       "      <td>223.9</td>\n",
       "      <td>743.3</td>\n",
       "      <td>950.1</td>\n",
       "      <td>550.8</td>\n",
       "      <td>202.9</td>\n",
       "      <td>21.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A04</th>\n",
       "      <td>1385.3</td>\n",
       "      <td>2376.6</td>\n",
       "      <td>184.8</td>\n",
       "      <td>915.3</td>\n",
       "      <td>994.7</td>\n",
       "      <td>281.8</td>\n",
       "      <td>121.0</td>\n",
       "      <td>63.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A06</th>\n",
       "      <td>1385.3</td>\n",
       "      <td>2829.2</td>\n",
       "      <td>327.7</td>\n",
       "      <td>919.5</td>\n",
       "      <td>1109.5</td>\n",
       "      <td>472.5</td>\n",
       "      <td>293.4</td>\n",
       "      <td>34.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A09</th>\n",
       "      <td>1385.3</td>\n",
       "      <td>2309.8</td>\n",
       "      <td>154.4</td>\n",
       "      <td>948.7</td>\n",
       "      <td>968.8</td>\n",
       "      <td>237.9</td>\n",
       "      <td>98.4</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A10</th>\n",
       "      <td>1385.3</td>\n",
       "      <td>2323.7</td>\n",
       "      <td>199.2</td>\n",
       "      <td>895.1</td>\n",
       "      <td>999.2</td>\n",
       "      <td>230.2</td>\n",
       "      <td>158.1</td>\n",
       "      <td>41.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_dur  total_dist  hsr_dist  zone1_dist  zone2_dist  zone3_dist  \\\n",
       "A02     1204.2      2468.1     223.9       743.3       950.1       550.8   \n",
       "A04     1385.3      2376.6     184.8       915.3       994.7       281.8   \n",
       "A06     1385.3      2829.2     327.7       919.5      1109.5       472.5   \n",
       "A09     1385.3      2309.8     154.4       948.7       968.8       237.9   \n",
       "A10     1385.3      2323.7     199.2       895.1       999.2       230.2   \n",
       "\n",
       "     zone4_dist  zone5_dist  \n",
       "A02       202.9        21.0  \n",
       "A04       121.0        63.8  \n",
       "A06       293.4        34.3  \n",
       "A09        98.4        56.0  \n",
       "A10       158.1        41.1  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_dur</th>\n",
       "      <th>total_dist</th>\n",
       "      <th>hsr_dist</th>\n",
       "      <th>zone1_dist</th>\n",
       "      <th>zone2_dist</th>\n",
       "      <th>zone3_dist</th>\n",
       "      <th>zone4_dist</th>\n",
       "      <th>zone5_dist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A02</th>\n",
       "      <td>1297.0</td>\n",
       "      <td>2494.3</td>\n",
       "      <td>218.8</td>\n",
       "      <td>834.0</td>\n",
       "      <td>923.0</td>\n",
       "      <td>518.5</td>\n",
       "      <td>193.1</td>\n",
       "      <td>25.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A04</th>\n",
       "      <td>1471.1</td>\n",
       "      <td>2388.3</td>\n",
       "      <td>193.7</td>\n",
       "      <td>912.4</td>\n",
       "      <td>965.0</td>\n",
       "      <td>317.2</td>\n",
       "      <td>129.2</td>\n",
       "      <td>64.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A06</th>\n",
       "      <td>1471.1</td>\n",
       "      <td>2818.3</td>\n",
       "      <td>318.1</td>\n",
       "      <td>922.2</td>\n",
       "      <td>1103.6</td>\n",
       "      <td>474.4</td>\n",
       "      <td>283.8</td>\n",
       "      <td>34.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A09</th>\n",
       "      <td>1471.1</td>\n",
       "      <td>2253.2</td>\n",
       "      <td>150.0</td>\n",
       "      <td>1002.1</td>\n",
       "      <td>878.8</td>\n",
       "      <td>222.3</td>\n",
       "      <td>96.8</td>\n",
       "      <td>53.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A10</th>\n",
       "      <td>1471.1</td>\n",
       "      <td>2312.0</td>\n",
       "      <td>182.2</td>\n",
       "      <td>941.4</td>\n",
       "      <td>959.3</td>\n",
       "      <td>229.1</td>\n",
       "      <td>144.7</td>\n",
       "      <td>37.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_dur  total_dist  hsr_dist  zone1_dist  zone2_dist  zone3_dist  \\\n",
       "A02     1297.0      2494.3     218.8       834.0       923.0       518.5   \n",
       "A04     1471.1      2388.3     193.7       912.4       965.0       317.2   \n",
       "A06     1471.1      2818.3     318.1       922.2      1103.6       474.4   \n",
       "A09     1471.1      2253.2     150.0      1002.1       878.8       222.3   \n",
       "A10     1471.1      2312.0     182.2       941.4       959.3       229.1   \n",
       "\n",
       "     zone4_dist  zone5_dist  \n",
       "A02       193.1        25.7  \n",
       "A04       129.2        64.5  \n",
       "A06       283.8        34.3  \n",
       "A09        96.8        53.2  \n",
       "A10       144.7        37.5  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_dur</th>\n",
       "      <th>total_dist</th>\n",
       "      <th>hsr_dist</th>\n",
       "      <th>zone1_dist</th>\n",
       "      <th>zone2_dist</th>\n",
       "      <th>zone3_dist</th>\n",
       "      <th>zone4_dist</th>\n",
       "      <th>zone5_dist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A02</th>\n",
       "      <td>1434.5</td>\n",
       "      <td>2777.6</td>\n",
       "      <td>187.0</td>\n",
       "      <td>967.7</td>\n",
       "      <td>1163.1</td>\n",
       "      <td>459.8</td>\n",
       "      <td>158.6</td>\n",
       "      <td>28.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A04</th>\n",
       "      <td>1579.6</td>\n",
       "      <td>2604.4</td>\n",
       "      <td>125.3</td>\n",
       "      <td>1075.7</td>\n",
       "      <td>1142.8</td>\n",
       "      <td>260.6</td>\n",
       "      <td>89.6</td>\n",
       "      <td>35.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A06</th>\n",
       "      <td>1579.6</td>\n",
       "      <td>3185.1</td>\n",
       "      <td>186.0</td>\n",
       "      <td>1059.2</td>\n",
       "      <td>1431.1</td>\n",
       "      <td>508.8</td>\n",
       "      <td>170.9</td>\n",
       "      <td>15.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A09</th>\n",
       "      <td>1579.6</td>\n",
       "      <td>2621.3</td>\n",
       "      <td>203.8</td>\n",
       "      <td>1103.4</td>\n",
       "      <td>1023.7</td>\n",
       "      <td>290.4</td>\n",
       "      <td>142.1</td>\n",
       "      <td>61.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A10</th>\n",
       "      <td>1579.6</td>\n",
       "      <td>2568.1</td>\n",
       "      <td>236.4</td>\n",
       "      <td>1070.6</td>\n",
       "      <td>1022.2</td>\n",
       "      <td>238.9</td>\n",
       "      <td>174.9</td>\n",
       "      <td>61.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_dur  total_dist  hsr_dist  zone1_dist  zone2_dist  zone3_dist  \\\n",
       "A02     1434.5      2777.6     187.0       967.7      1163.1       459.8   \n",
       "A04     1579.6      2604.4     125.3      1075.7      1142.8       260.6   \n",
       "A06     1579.6      3185.1     186.0      1059.2      1431.1       508.8   \n",
       "A09     1579.6      2621.3     203.8      1103.4      1023.7       290.4   \n",
       "A10     1579.6      2568.1     236.4      1070.6      1022.2       238.9   \n",
       "\n",
       "     zone4_dist  zone5_dist  \n",
       "A02       158.6        28.4  \n",
       "A04        89.6        35.7  \n",
       "A06       170.9        15.1  \n",
       "A09       142.1        61.7  \n",
       "A10       174.9        61.5  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_team_poss = macro_pred.idxmax(axis=1)\n",
    "\n",
    "true_team_poss = helper.traces[\"team_poss\"].copy()\n",
    "true_team_poss.loc[pred_team_poss.isna()] = np.nan\n",
    "\n",
    "baseline_team_poss = pd.Series(np.random.choice([\"A\", \"B\"], len(helper.traces)), index=helper.traces.index)\n",
    "baseline_team_poss.loc[pred_team_poss.isna()] = np.nan\n",
    "\n",
    "for p in tqdm(players):\n",
    "    true_att_stats.loc[p] = calc_running_metrics(helper.traces, true_team_poss, p, attacking=True)\n",
    "    true_def_stats.loc[p] = calc_running_metrics(helper.traces, true_team_poss, p, attacking=False)\n",
    "\n",
    "    pred_att_stats.loc[p] = calc_running_metrics(helper.traces, pred_team_poss, p, attacking=True)\n",
    "    pred_def_stats.loc[p] = calc_running_metrics(helper.traces, pred_team_poss, p, attacking=False)\n",
    "\n",
    "    baseline_att_stats.loc[p] = calc_running_metrics(helper.traces, baseline_team_poss, p, attacking=True)\n",
    "    baseline_def_stats.loc[p] = calc_running_metrics(helper.traces, baseline_team_poss, p, attacking=False)\n",
    "\n",
    "print(\"\\nRunning metrics when attacking:\")\n",
    "display(true_att_stats[:5])\n",
    "display(pred_att_stats[:5])\n",
    "display(baseline_att_stats[:5])\n",
    "\n",
    "print(\"\\nRunning metrics when defending:\")\n",
    "display(true_def_stats[:5])\n",
    "display(pred_def_stats[:5])\n",
    "display(baseline_def_stats[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA8kAAALDCAYAAAAmIEJLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd3gUxf8H8Pe1XHqH0EOvfum9BgyKiRC6WJDQ7AqCDUEJRVAEFFF/oAJBVERCFQGpAWnSCUhvIaEnkErKlfn9kbvlLneX5JJLIbxfz3OPMrszOzO3t9nPzu6sTAghQERERERERESQl3YFiIiIiIiIiMoKBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMmCQTERERERERGTAIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiIiIDBslERCXg6tWrkMlkkMlkuHr1amlXh6hUGX8L0dHRpV0VCxEREZDJZAgKCirtqlAxiI6OlvY/e4WHh0MmkyE8PNzxFSOiMoVBMhEVmPHksSAnF6ZBYWRkpNV1NBoNFi9ejJCQEFStWhVqtRpeXl6oX78+unfvjgkTJmDTpk3IzMy0yBsUFCSVb/pxc3NDnTp1MGTIEPz9999Fam9kZKTVbbi7u6NKlSpo3bo1Ro0ahaVLlyI9Pb1I27JXUlISIiIiEBERgaSkpBLddnly584dfP755+jZsyeqVasGFxcXuLm5oWbNmujbty9++OEH9m8Z8d5770Emk8Hb2xs6nc7mekOGDMn32AMABw4ckNYr6rHiUfbcc89J/TBx4sR814+MjERERES+FziMxydeFCSiRxGDZCIqFXFxcWjZsiVGjhyJTZs24caNGwAAhUKBS5cuITo6Gp9//jlCQkJw4MABm+WoVCoEBARIn+zsbFy+fBkrVqxAr169MHr0aAghilxff39/aRtOTk64ffs2jhw5gkWLFiE8PByVKlXC1KlTodVqbdazQYMGaNCgAVQqVZHrk5SUhClTpmDKlCkM4gpBCIEZM2agVq1amDBhArZt24br169DqVRCoVAgNjYW69atw6uvvoqaNWti8eLFpV3lx1737t0BAMnJyTh27JjN9UyDt7wCuZ07dwLI+W127tzZIXV81CQmJmLt2rXSv5cuXZrnBQggJ0ieMmVKvkGy8fhUnoLkypUro0GDBqhcuXJpV4WIihmDZCIqcTqdDmFhYTh16hRcXV0xbdo0XLt2DZmZmbh37x4ePHiAgwcPIiIiAnXq1MmzrI4dO+LWrVvSJzMzE4cPH0aXLl0AAD/99JNDApxDhw5J27h37x60Wi3Onj2LBQsW4H//+x/S0tIwefJk9OzZE1lZWRb5q1atirNnz+Ls2bOoWrVqketDhSeEwNChQzFx4kQ8ePAA7dq1w6pVq3D//n2kpqYiJSUFycnJWLNmDXr37o3k5GSsX7++tKv92OvSpQsUCgUA28HvmTNncPv2bVSqVCnP9UyXtW3bFm5ubo6s6iPjl19+QXZ2NkJCQlCnTh1cv379sR5Vz8/MmTNx9uxZzJw5s7SrQkTFjEEyEZW4HTt2SCNBixYtwqRJk1C9enXpNm61Wo02bdpg8uTJuHDhAjp06FDgshUKBVq1aoV169bBz89P2oajyWQyNGjQAK+++iqOHTuGcePGAcg58X7nnXccvj1ynFmzZuHXX38FAIwdOxb79+9H//794e3tLa3j6emJvn37Yv369di1axeqVatWSrUlI09PT7Rq1QrAw1Hg3IyBb1hYGOrXr4/Y2FhcuXLFYj2NRoO9e/cCeDhC/TgyHhtffvllDB061CyNiOhxxiCZiErc8ePHpf8PCwvLc12ZTAa1Wm33Nnx8fNCuXTsAwH///Wd3fnsoFArMmTMHoaGhAHJOMs+fP2+2Tn4Td8XHx+Pdd99FkyZN4ObmBrVajSpVqqBVq1Z49913cejQIWndoKAg1KpVS/p3rVq1zJ6ZNp1wSK/XY/v27XjnnXfQvn17VKtWDU5OTvDz80O3bt2wYMECaDQaq+3KXefbt29jzJgxqFWrFpydnREQEIAhQ4bg7NmzefaPXq/HH3/8gb59+0rPnleoUAGtWrXChx9+iFOnTlnNl5qais8//xwdOnSAr68v1Go1qlevjiFDhmD//v15btOWhIQETJs2DQDw5JNPYu7cufk+Y9+1a1d88803VpdFR0dj0KBBUrv8/f3x5JNPYsmSJVZvW23WrBlkMpl0UcWWHTt2QCaTQS6X49q1axbLr169irFjx6JJkyZwd3eHq6srGjZsiDFjxlhdH3j4jH3NmjUB5ASaffv2ReXKlaFQKCwmIypK/9+/fx/vv/8+6tSpA2dnZ1SuXBmDBg3CkSNH8syXH2NAu2fPHqv9awySg4KC0K1bN7M0U4cOHZLmEcgvSN6+fTtCQ0NRoUIFODs7o1GjRpgyZYrVuRJM3b17F5MmTUKLFi3g5eUFZ2dn1K5dGyNHjrR5TMo9qdTFixcxYsQIVK9eHWq1GtWqVcPo0aNx/fr1PLddEIcOHcLJkyfh5eWFsLAwvPzyy5DJZPjzzz9x9+5di/WN+8+uXbsA5NxOnXu+hqtXr0qTWxl1797dbB3j/md04MABfPjhh+jSpQsCAwPh7OwMb29vtG/fHl988QXS0tLybcu///6L4cOHo27dunB1dYWnpycaN26MESNG2D0ynpiYiA4dOkAmk6FWrVpmx/K8Ju4yzpMREREBIQR+/PFHtGvXDp6envDw8ECHDh3wyy+/5LltjUaDuXPnonnz5nBzc4Ovry+CgoIQFRVlsQ0iKmaCiKiAJk+eLACIghw6rly5Iq27ZMkSs2WzZs2Slp0/f75QdenWrZsAILp162ZznWeeeUYAEG5uboXaxpIlS6R6XrlyJd/1Dx06JK3/ySefmC0z7Y/cZR0/flz4+PhIyxUKhfDx8REymUxKGzZsmLR+v379hL+/v7TM399fBAQESJ9+/fpZ3S4A4e7uLry8vMzSunTpIh48eGDRHtO8GzZsEBUrVhQAhKurq1Cr1dIyT09Pcfz4cat9cvfuXdG1a1ez7Xl7ewt3d3fp32FhYRb5jh07JqpVq2bWJx4eHtK/ZTKZmDFjRr7fSW6m+94///xjd35T7777rll9vL29hUKhkNJ69OghUlJSzPJ8+eWXAoCoVKmS0Gq1NssODw8XAERQUJDFsl9++cWs/9VqtXBxcZH+7eHhIf7++2+LfMb9OTAwUHz99dfS/uXl5SVUKpXZPlaU/r9y5YoIDAyU1nVychKenp7S/69bt05atnPnzoJ1tsHmzZulvAcPHrRYHhAQIACIGzduiF9++UUAEC+//LLFep999pnUdxkZGWbLjMe5bt26iVmzZgmZTCZ9v6a/ye7du9v8Drdu3Sq8vb2ldVUqlXBzczPrk6VLl1rk27lzp7TOjh07pN+Jh4eHUCqV0rIqVaqI+Ph4u/out1dffVUAEKNHj5bSjL/VOXPmWKz/+++/i4CAAKFSqaTjqulxJyAgQFy7dk2888470vcAQPj4+Jit07p1a7NyTY8Nrq6uZsdCAKJx48bi9u3bVtug1WrFO++8Y7a+m5ub2fHTy8vLZh/ndvXqVdGwYUMBQDRr1kzcuHHDbPmwYcMsjsdGxr9JkyZNEmFhYQKAUCqV0r5v/Hz66adW25KWlmZ2rMz9d2DChAnSNiZPnmy1DCJyHAbJRFRgjgqSo6OjzQKJwpzs5Rck37t3T/j5+QkA4n//+5/d5Qthf5AshJACya5du5ql5xUkP/nkkwKAaNmypdi/f7/Q6/VCCCGysrLE+fPnxezZs8WsWbMKXJ6puLg48eKLL4r169eLxMREKT01NVUsWbJEVKlSRQAQ7777rkVe0234+PiITp06iUOHDgkhhNBoNGLr1q2icuXKUqCdm0ajEZ06dZKCkS+++ELcuXNHWn79+nWxcOFCMWHCBLN8N27ckPqxf//+4vDhwyI7O1sIIcTt27fFJ598IgUMa9assdl2a55++mkBQFSoUMGufLnNnz9f6ptXXnlF3Lx5UwiRc6L71VdfSfV77rnnLNpmDKQ3bdpktewHDx5IAenixYvNlm3ZskXI5XKhVCrFBx98IK5cuSL0er3Q6/Xi7NmzYtCgQdKFi9jYWLO8xv3Z2dlZKBQKER4eLq5duyaEyAk2Ll68KNWxsP2v1WpF69atpX3mjz/+EBqNRgghxH///Se6dOliFjzaGySnpaVJQdoXX3xhtuz06dMCgKhfv74QImffByBq1KhhUU5wcLDN44fxOOft7S3kcrmYMGGCuHv3rhBCiOTkZPHpp59K9V+0aJFF/piYGOmixejRo8Xp06elYDo2Nla88cYbUgBl/D0ZmQZwPj4+ok+fPuLMmTNCiJzjwYoVK6R9Y+jQoXb1nan09HQpeDO9WLRo0SIBQDRp0sRm3oIGagX9jnv37i1WrFgh/YaEyPkNrF69WjRo0EAAMLvwZ+qDDz6QtjNixAhx7tw5aVlSUpJYu3atxW/QVpB84sQJ6XjYvXt3kZycbLG9ggTJPj4+wsvLS0RGRkoXH+Pi4kTv3r0FACGXy61eHDZetJDL5eKLL74QqampQoicC43GCwHG3w6DZKLixyCZiArMNEjOPYKQ+2M60pk7SBZCiJ49e5pdMe/QoYMYO3asWLZsWYFGl20FyVqtVhw+fFh06dJFKn/u3LmFam9hgmRju6pWrWqWnldQazyh3rdvX4HrVtAgOT/G0W83NzeLETXTbTRs2NDqaPP69euldeLi4syW/fTTTwLIGXX866+/ClynESNGCADihRdesLnO3LlzpdEeexhHR3v27GlXPlMPHjwQvr6+AoB4/vnnra7zzTffSP1y+PBhs2XGQN1W3t9++00AEC4uLmYj0TqdTtSrV08AEAsXLrRZvz59+ggAYsyYMWbppvtz//79beYvSv+vWLFC2sa2bdss8qWnp4s6deoUOkgWQoiOHTsKAOKZZ54xS//++++lwNSodu3aAoC4dOmSlJaVlSVcXV0FABEREWFRvulxzlYw0r9/fwFABAcHWyzr0aOHNPJnizHoyX0XhWkA1717d6HT6SzyGvctFxcX6QKEvZYuXSoAiDp16pilp6SkSMejAwcOWM3r6CA5L/Hx8UKtVguZTGZx0efcuXNCLpcLAOKDDz4ocJnWguTo6GjpDptBgwaJzMxMq3kLEiQDOXcB5JaZmSkF4dOnTzdbFhsbK7Vl2rRpeW6bQTJRyWCQTEQFZnryaM/HWpCclpYm3njjDWlUKPcnMDBQREREWL2aL8TDExKVSmUWnJvekghADBgwQBoFs1dhguTnnntOOoE1lVdQaxyNXbVqVYHr5qggWYiHo9/79++3uY0ff/zRal6NRiOcnJwEALFx40azZcZgJjQ0tMB1ycjIEM7OzgKAOHHihM31EhISpLrdunWrwOUbA4Dco0v2ML1d2HTkypRWq5W+13HjxpktM94GnDsINjI+JpA7iDae3Pv7+1sNnoyioqKkCxumTPfn3COYRkXt/379+gkAolOnTjbzLly4sEgB1MSJEwWQcwuyaZBoHEX/9ddfpbThw4cLAOKnn36S0v755x9p+7t27bIo33icU6vV0mhebsYgs2LFimbpxt+MUqk0u3Mjt8OHD0sXp0xv2TYN4LZu3Wo1b2xsrLTO6dOnbW4jL8bbeq1dJHjhhRcsLjaYKskgWQgh2rZtKwCI5cuXm6V//PHHAoDw8/OzuMCXl9xB8sqVK6XHF9566608f1sFCZLz2veN++PgwYPN0o0XnVxdXUVaWprVvBcvXmSQTFSCOHEXERWKyLnIZvNjbUZZU25ubvjuu+8QHx+PH374AUOHDkWjRo2kV7zExsYiIiICzZs3x6VLl2yWo9FocPv2beljfE+xTCbDd999h6ioKIe8l7g4PfvsswCAYcOGYfz48di1axcePHjgsPKzs7OxYMECPPXUU6hSpQrUarXZRDp37twBkDN5mC3GSdByUyqVqFChAgDg3r17UrpWq5UmG+vdu3eB63rkyBFpQqSnnnoKlSpVsvpp0qSJlCc2NrbA5TvC4cOHAQDVq1dH/fr1ra6jUCjQo0cPs/WN+vXrBw8PD2RkZGDVqlVmy27fvo0tW7YAyJlx2JRxNubk5GRUqVLFZt+MHj0agO1+cXFxQcuWLa0uK2r/G9tqbLs1eS0rCOPEdKmpqWYTgRknlDJO2GX6/9benezi4oL27dvb3I5xUjRrqlSpAsB8nwcefkd6vR6NGze22X+9evUCAKSnpyMxMdHqNmz95ozbtrb9grh48SL++ecfyGQyaUZrU8OGDQMA/P777w49Dtmi1+vx22+/oU+fPqhRowZcXFzMjk8HDx4EYHl82rdvHwCgZ8+ecHZ2LtS2v/vuOzz33HPIysrCZ599hvnz50MuL9qpsa3vDbC93xw9ehQA0Lp1a5uvI6tTpw6qV69epLoRUcEpS7sCRPR4q1ixIkaPHi2d2KelpWHHjh348ssvsWfPHly5cgVDhgwxm93ZVLdu3aSTXo1Gg9jYWPzwww+YPXs2PvjgAzRp0sTspLm4GU9+jK+fKohZs2bh4sWL2LlzJ+bOnYu5c+dCoVCgefPmCA0NxSuvvFLodyvfuXMHwcHBOHnypJTm7OwMf39/6YLE3bt3odfrpdl+rfHw8LC5TKnM+VNiOkt2YmKi9O/AwMAC1/fGjRvS/9++fbtAeew5kffz80N8fHyhggsj40WF/L4T42ujjOsbubq6YsCAAYiMjMSyZcvMZspdvnw5dDodKlWqhJ49e5rlM/aN8cJQfjIyMqym+/n52QwEitr/Bembor5Oq1OnTnByckJ2djaio6PRrl07nD59Gnfu3EHdunXNtp1XkGwsx5aC7PPGi3JGxv7T6/VF3n9tbd+4bQA2Z6bPy+LFiyGEQOfOnVG7dm2L5cHBwahatSquX7+OlStXSkFzcXjw4AGeffZZs1d6OTk5wdfXV7q4ee/ePWg0Govj061btwDYd3zJ7a233gIAvPnmm/j4448LXY4pe4+VAKTZxE0vgFhTtWpVxMXFFbGGRFQQHEkmojLF3d0dffr0wa5du6RXsxw+fNjstVG2qFQq1K1bF7NmzcLkyZORnp6OwYMHWwQpxSkmJgZAzlX/gvL29saOHTvwzz//4IMPPkCnTp2gVCpx5MgRTJ06FfXq1cPy5csLVZ93330XJ0+ehJ+fHxYvXoybN28iIyMDd+/exa1bt3Dr1i3pxEwIUahtWJPfa5VsMX2tT0ZGRr53LAghzF55lR/jCGhB9qfiZBwljo6ONjvpXbZsGQDghRdekC5iGBn7pl27dgXqF1vfZ+5yrW0DKJ7+dwQXFxdptM4YXBkD39wXxGrWrIkaNWogPj4eFy9eRHZ2tjQCWRzvRzb2X0BAQIG/o9yvRCpOOp0OS5cuBZDzGq3cr3CSyWRQKBTSK6aK+53Jn332GXbu3AkXFxd89dVXiI2NRWZmJhITE6Xjk/G7zr0/F/YYY+qll14CAPz444/4888/i1xeUTmiTUTkGAySiahMksvlGDVqlPTvc+fO2ZX/448/Rp06dXDnzh188sknjq6eVYcPH5ZGjwoTOHTu3BlffPEF9uzZg6SkJKxbtw7/+9//kJGRgREjRhR4ZMpIo9Fg9erVAIBvv/0Ww4cPR6VKlczW0el0SEhIsLuu+TEdCbLndmjT+hXHbdRPPvkkgJyRmz179hSqjIoVKwLI+/Z00+XG9U0FBQWhevXq0Ov1+PXXXwEAp0+flm67zH2rNfCwb4rz9vKi9r+xrXm9x9cR7/g1Brh79+6FVqs1ez9ybqajyf/++680wl4cQbKx/xISEvK8M6O0bNq0yexugfz8888/uHDhQrHV5/fffwcAfPrppxg7dixq1KhhESgaR4xzc8TvYenSpRg2bBiys7MxYMAArF27ttBlFYXxkZX8vhtH/HaIqGAYJBNRmWX6PKBarbYrr0qlwqRJkwDkjIacP3/eoXWzZsqUKQByRuqMIxSF5ezsjD59+khBbmZmpllQZ3q7rK0Rw7t370rPl7Zo0cLqOnv27JHWcSSlUom2bdsCgF0jNG3atJFugS2OkZ3hw4fD1dUVABAREVHg0XO9Xi/9f+vWrQHkBMG29iudTieNcrZp08ZiuUwmk/YR4+ix8b9NmzZFs2bNLPJ06tQJQE7QkPs5Z0cpav8b+8b09tncduzYUbjKmTAGuGlpaTh06JDV55GNjGk7d+6U6uXu7m71eykq43ek0+mwadMmh5dfVMaR4X79+iE1NTXPj/G59cWLF5uVYTz25PfbMQa7ea1nvIvC1vHp6tWruHjxotVlHTt2BABs3bq10McwuVyOxYsXY8SIEdBoNBg8eLDFPAElwdjXhw8ftnlx5fLly7zVmqgEMUgmohJ36tSpAl0R//nnn6X/t3USlZeXXnoJgYGB0Ol0UgBbHHQ6HcaPH48NGzYAAEaPHo26desWKK9WqzULwHJzcXGR/t80MPb09JT+PykpyWpeT09P6UT1xIkTVrc9ceLEAtWzMEaOHAkA2LhxIzZu3FigPG5ubnjhhRcAAF988QWuXbuW5/r2Plvs7+8vXTzZvn07xo8fn+/J/t69ezFmzBjp3z179pSeOY+IiLCaZ+HChdKo0PPPP291HeNo8enTp3H48GFpRNnaKDKQExga96t3330X2dnZeda7MM9dF7X/n3vuOQA5F19MnwM2ysjIwJdffml3vXLr0KGDNFnT999/jzt37qB27dpWJzYyBsm7du2S6tS5c2ezZ3sdpV69etJo9sSJE5GcnJzn+kV5Nt5et2/flo5Rzz33HNzd3fP8DBo0CEDOaKvpbfjGY4+t444963l5eQGwfnwCgI8++shm3vDwcCgUCiQmJmLy5Ml51iUvcrkcP/30E1555RVoNBoMGTIEf/zxR6HLK4z+/ftDLpcjPT0d8+bNs7rOZ599VqJ1InrcMUgmohIXHR2N2rVr47nnnsPKlStx8+ZNaZlxxLRPnz7SFf2BAwcWanIWpVKJ999/H0DObX2nT592TAOQMzpy4cIF/PDDD2jZsiXmzp0LIOd2XlsnOdbEx8ejXr16mD59Oo4dO2Y2EVBMTIw02ujm5mY2Subt7S1NULRkyRKLCYSAnNEy48jWuHHjsGPHDikgP3XqFEJCQnD48GGbs6kW1dChQ9G5c2cIITBgwAB8+eWXZrd237hxA1999RU+/PBDs3wzZsxAlSpVkJCQgA4dOmDZsmVITU2Vlt+9exerVq1Cv379bAagefnoo4+kYO6rr75Cp06dsGbNGqSkpEjrpKamYsOGDejfvz+6dOliNoLj4uIiBcfLly/Ha6+9Jt0K/+DBA3zzzTcYO3YsgJxgpFWrVlbr0bBhQ2nk9fXXX0dcXBwUCoUUpOamVCqxYMECKJVK7NmzB127dsX27dvNJgG6fPkyFixYgDZt2uD777+3u2+AovX/gAEDpFGxAQMGYNWqVVKAdebMGTzzzDPSJEVFoVar0aFDBwCQnte3NUFf3bp1UaVKFVy/fh27d+8GUDy3WhvNnz8f7u7uOH/+PNq3b49169aZjXRev34dy5Ytw5NPPmmx7xenn3/+GVqtFi4uLtKM+nkZPHgwAODmzZtmF7meeOIJADkXv/K62Glc79dff7U5OZlxlu/p06dj9erV0nHsypUreOGFF/DHH3/Ax8fHat66detKx/dZs2Zh1KhRZreGp6SkYMWKFejXr1++bZXJZFiwYAHeeOMNaLVavPDCC4WeB6IwAgMDpYuKn376KWbPno20tDQAOZMgjhs3DosXL4a3t3eJ1Ynosee4t0kRUXln+p7k/Ji+Yzf3e5IXLFhg8V5kZ2dn4ePjY5H+1FNPWX2XrPGdlN26dcuzHhkZGaJSpUoCgBg4cKA9zTV7r6y/v7/0LmZfX1+hUCjM6unh4SGmT59u9t5WW/1h+l5j03QAQqFQCF9fX+ndwwCEk5OTWLlypUWZ06ZNk9ZRq9WievXqIjAw0OwdwIcPHxZubm5m63l4eEjvcv35559FYGCg1e+poO9itpVfCCHu3r0runTpIpUjk8mEt7e3cHd3l9LCwsIs8p0+fVrUr19fWkculwtfX1+ztgAQwcHBNuuVF71eL6ZMmSK9N9n0ezT2j/Hj6+srfv75Z4sy3n33XbN2+fj4mL2nu3v37lb3XVPffPON2baefvrpfOu+Zs0aszqqVCrh5+cnvevV+Jk+fbpZPuP+HBgYmO82itL/ly5dEtWrVzfb57y8vKR92fQ900V5h+7UqVPN6rJ06VKb6z7//PNm6x48eNDmusbjXF7Hltzv2s1tz5490nHH+Lv28/Oz2N9GjRplV7lGhem/hg0bCiDn3fEF1bJlSwFA9O3bV0o7f/689C5tuVwuAgICRGBgoAgMDBRxcXHSesuWLTPbR6tWrSoCAwPN3iN89epVERAQIK2nVCqlfQWAmDFjRp7vZdZqteLNN98061N3d3fh4+MjZDKZACC8vLzM8uTXx2+//bb0nS1btsxsWUHek5zXO4zz2rdSU1NF586dzfYZ03ZMmjRJer/1zJkzbW6DiByDI8lEVOJeffVVnDhxAl988QXCwsJQt25dKBQKJCcnw8PDA40bN8bLL7+MjRs34u+//87zlRr5cXZ2xrhx4wAAq1atsnlbX34SEhKkdzFnZmaiQoUKaNWqFUaOHImlS5fi5s2bmDhxot23cFatWhXr16/Hu+++i/bt26Ny5cpIS0uDUqlE48aN8eabb+LUqVMYOHCgRd6PP/4Y8+bNQ+vWraFSqRAfH4/Y2FiziW5atWqFgwcPYvDgwfD394der4eHhwcGDx6Mffv2WX1PqiP5+/sjOjoav/zyC5555hlUqFAB6enpcHV1RatWrfDRRx9hxowZFvkaNWqEmJgYLFy4EE899RT8/f2RkpICIQTq1q2LQYMG4Ycffij0bZEymQyffvopLl++jBkzZqBHjx6oUqUKsrOzodVqERgYiL59++Knn37C1atXrfbT3LlzsWPHDgwYMAABAQFIS0uDh4cHunfvjsWLF2Pr1q357rvPP/+82Xu8bd1qbapv3764ePEiJk+ejLZt28Ld3R1JSUlQq9Vo1qwZRo0ahTVr1kijbIVRlP6vXbs2jh8/jnHjxqFWrVoQQsDZ2RkDBw7Evn370KdPn0LXy1Tu0eC8XvVmuszT09Pme6IdpVOnTjh//jxmz56Nrl27wtvbG0lJSVAoFGjUqBFeeukl/Prrr/j666+LtR5Ge/fuxdmzZwE8HCEuCOO6GzZskO6WqFevHnbu3Ik+ffqgQoUKSExMRGxsLGJjY83uaHnppZewbNkydO7cGa6urrh58yZiY2PNJrwLDAzE4cOHMXLkSGmWfWdnZzz77LP4+++/MWHChDzrp1Ao8O2332LPnj148cUXUaNGDWg0Gggh0LhxY4wcOdLuZ4y/+eYbvPvuu9DpdBg2bBgiIyPtyl9Y7u7u2L59O7788ks0bdoUTk5OEEKgW7duWL16NaZNmybdus4RZaLiJxPC/nd+BAUFSZNkWLNp0ybpFhpTkZGR+P7773H69Gk4OTmhffv2mDRpkjT5gjV79+7FZ599hgMHDiA7OxuNGzfGW2+9VaATCSIiIiKiR11aWhr8/PyQnZ2N3bt3o0uXLqVdJaJyrUizVgwYMMBs9lkj43NypsaOHYt58+bBxcUFTz31FDIzM7F161Zs2bIFUVFR6Nu3r0WeVatW4bnnnoNer0fXrl3h7++P7du3Y9iwYYiJicHs2bOLUn0iIiIiojJv7ty5yM7Ohq+vb7HMzE5E5oo0knzlyhXUrFkz3/W3bdsmzQa6f/9+1KtXDwCwf/9+BAUFwdXVFVeuXDG7feTevXuoVasWUlJSsGrVKvTv3x9AzuyMnTt3xsWLF7Fz585CvYuUiIiIiKisSE1NxejRoxEeHo727dtL58SxsbH49ttvMWfOHAghMHXqVHzyySelW1mix0CJPJNsnPV10qRJUoAM5LzC4bXXXkNSUpL07j6jn376CSkpKQgLC5MCZAAICAjArFmzAABz5swpgdoTERERERUfnU6HFStW4JlnnoGPjw88PT3h6emJmjVrYvbs2dJbAvJ7TpuIHKPYg+SMjAzs2LEDAKxOPGNM+/PPP83S//rrL5t5QkND4ezsjG3bthX6BfJERERERGWBu7s7vv32W4SFhaFOnTqQyWTIzMxE5cqV0bt3b6xcuRIrV64slvd7E5GlIt1uPWnSJCQmJkIul6N+/fro27cvatSoYbbu8ePH0aJFC1SoUAF37tyxKCs9PR3u7u7w8fHBvXv3pHRvb28kJyfjv//+Q+PGjS3ytWnTBocPH8aJEyfQtGlTe5tAREREREREZKFIl6OmT59u9u/33nsPn3zyidmzEteuXQMAVKtWzWoZbm5u8Pb2xv3795GamgoPDw+kpKQgOTk5z3zVqlXD4cOHERsbyyCZiIiIiIiIHKJQQXLXrl0xatQodOzYEZUrV0ZcXByioqIwffp0fPrpp/D09MSYMWMA5ExZDwCurq42y3Nzc0NSUpIUJBvz5JXPzc0NQM5EB9ZkZWUhKytL+rder8e9e/fg5+cHmUxmX4OJiIiIiIjokSWEQGpqKqpUqQK5PO+njgsVJE+dOtXs3/Xr18fHH3+M1q1b4+mnn0ZERAReeeUVuLi4FKZ4h5g5cyamTJlSatsnIiIiIiKisiUuLs7m3cpGDn36/6mnnkLr1q1x+PBh/PvvvwgKCpLeo/zgwQOb+dLT0wEAHh4eAGD27uUHDx7A09Mz3zy5TZgwAePGjZP+nZycjBo1auDKlStSHrlcDoVCAZ1OB71eL61rTNdqtTB9ZFuhUEAul9tM12g0ZnUwTq6g1WoLlK5SqaDX66HT6aQ0mUwGpVJpM91W3dkmtoltYpvYJraJbWKb2Ca2iW1im9imHKmpqahVq5bN+NFse/muYad69erh8OHDuHnzJgBIE3nFx8dbXT89PR1JSUnw8fGRKuzp6QkvLy8kJycjPj7e6sRdxvICAwOtlqtWq6FWqy3SfX19rQbdREREREREVD6pVCoAKNCjtw5/BdT9+/cBPHxmuEGDBlCr1bh79y6uX79usf7Ro0cBwGLyrWbNmpktN6XRaHDq1Ck4Ozujfv36Dq0/ERERERERPb4cGiTfvXsX//zzDwCgZcuWAAAXFxf06NEDALBy5UqLPFFRUQCA3r17m6WHhoaaLTe1YcMGZGZmIjg4GM7Ozo5rABERERERET3W7A6S9+3bh7Vr15rd3w0AV69eRb9+/ZCeno4+ffqYPQxtfDZ4+vTpuHDhgpS+f/9+LFy4EN7e3hg5cqRZeaNGjYKnpyfWrVuH1atXS+l37tzBBx98AAAYP368vdUnIiIiIiIissnuZ5LPnz+P4cOHo1KlSmjZsiW8vb0RGxuLI0eOIDMzE02aNMGPP/5olic4OBhjxozBvHnz0Lx5c/Ts2RPZ2dnYunUrhBBYsmQJvL29zfL4+vpi8eLFGDx4MAYOHIigoCD4+flh27ZtSEpKwrhx4xAUFFSUthMRERERERGZkQnTacYK4MyZM5g/fz7+/fdfxMXF4f79+3Bzc0OjRo0waNAgvP766zZf/RQZGYlvv/0WZ86cgZOTE9q3b49PPvkEHTt2tLm9vXv3Yvr06Thw4ACys7PRuHFjvPXWWxg2bJhdDU1JSZEmA+PEXURERERERI8Pe+JBu4PkRxWDZCIiIiIioseTPfGgw2e3JiIiIiIiInpUOfw9yURERERE9HjRaDQWE/sSFSeFQiG9+9jRGCQTEREREVGhpKSkICEhAVlZWaVdFXoMqdVq+Pv7O/xxWgbJRERERERkt5SUFFy/fh3u7u7w9/eHSqWCTCYr7WrRY0AIAY1Gg+TkZFy/fh0AHBooM0gmIiIiIiK7JSQkwN3dHdWqVWNwTCXOxcUFHh4eiI+PR0JCgkODZE7cRUREREREdtFoNMjKyoKXlxcDZCo1MpkMXl5eyMrKgkajcVi5DJKJiIiIiMguxkm6imviJKKCMu6Djpw4jkEyEREREREVCkeRqbQVxz7IIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiokKSyWR2fWrWrFnaVaZ88BVQRERERETkcGcaNirtKhRIo7NnipR/2LBhFml79uzBpUuX0KxZMzRv3txsmb+/f5G2R8WPQTIREREREVEhRUZGWqSFh4fj0qVL6Nu3LyIiIkq8TlQ0vN2aiIiIiIiIyIBBMhERERERUQmIjIyETCZDREQEzp8/jyFDhiAgIAByuRxr164FANSsWdPma42io6Mhk8kQHh5usUwIgeXLl6NHjx7w8fGBs7MzGjVqhIiICDx48KAYW1X+8HZrIiIiIiKiEnTu3Dm0adMGfn5+6N69O+7fvw+VSlXo8vR6PV566SUsX74c7u7uaN26NXx8fHD48GFMmTIFmzZtQnR0NFxcXBzYivKLQTIREREREVEJ+v333/HWW2/h66+/hkKhKHJ5c+bMwfLlyxEUFITly5ejUqVKAIDs7Gy88cYbWLRoEaZMmYLPP/+8yNt6HDBIJiIiKiWOnPm1qLOzEhFRyalQoQK++OILhwTIWq0Ws2bNgpubG37//XcEBARIy5ycnDB//nz89ddf+OGHHzBjxgzI5XziNj8MkomIiMqDXYcdV1a31o4ri4iILAQHB8PV1dUhZR09ehQJCQno2bOnWYBs5OLiglatWuGvv/7ChQsX0KBBA4dstzxjkExERERERFSCatSo4bCyrl69CgDYunWrzQm/jBISEhgkFwCDZCIiIiIiohLk7OxcqHx6vd5mWt26ddGpU6c88/v5+RVqu48bBslERERERERlhJOTEwAgLS0N7u7uZsvi4uIs1q9WrRoAoGHDhoiMjCz2+j0O+NQ2ERERERFRGVG5cmUAwPnz5y2Wbd261SKtTZs28PLywq5du3Dv3r1ir9/jgEEyERERERFRGdGtWzcAwMyZM6HT6aT05cuXY/ny5Rbrq9VqfPDBB0hNTUX//v1x+fJli3WuX7+OZcuWFV+lyxnebk1ERERERFRGvPnmm1iwYAGioqLQuHFjNG3aFBcuXMCpU6cwZswYfPXVVxZ5PvroI5w9exbLli1Do0aN0KJFC9SqVQvZ2dk4d+4cTp8+jaZNm2Lo0KGl0KJHD0eSiYiIiIiIyoiAgADs3r0bzz77LG7evIlNmzbBy8sLW7duRZ8+fazmkcvl+Pnnn7Fu3Tr07NkTV65cwapVq7Bnzx44Ozvj/fffx+LFi0u4JY8umRBClHYlSkJKSgq8vLyQnJwMT0/P0q4OERERzjRs5LCyGi104G10fE8yEeUjMzMTV65cQa1atQo9UzORIxR0X7QnHuRIMhEREREREZEBg2QiIiIiIiIiAwbJRERERERERAYMkomIiIiIiIgMGCQTERERERERGTBIJiIiIiIiIjJgkExERERERERkwCCZiIiIiIiIyIBBMhEREREREZEBg2QiIiIiIiIiAwbJRERERERERAZFDpITExNRsWJFyGQy1K1bN891IyMj0bZtW7i7u8PX1xchISHYt29fnnn27t2LkJAQ+Pr6wt3dHW3btsXPP/9c1GoTERERERERWShykDx+/HgkJCTku97YsWMxfPhwnDp1CsHBwWjbti22bt2Krl27Yu3atVbzrFq1Ct26dcPmzZvRtGlT9OrVCxcuXMCwYcPw3nvvFbXqRERERERERGaKFCRv374dS5cuxejRo/Ncb9u2bZg3bx78/Pxw4sQJrF27Fps3b8bu3buhUCgwfPhwJCUlmeW5d+8eRowYAZ1Oh6ioKERHRyMqKgpnz55F3bp1MWfOHERHRxel+kRERERERA4hk8nMPnK5HN7e3ujSpQt++uknCCFKtX6RkZGQyWSIiIgo1Xo8CgodJGdkZODVV19F48aN8x3VnTt3LgBg0qRJqFevnpTeoUMHvPbaa0hKSsKiRYvM8vz0009ISUlBWFgY+vfvL6UHBARg1qxZAIA5c+YUtvpEREREREQON2zYMAwbNgwvvvgiGjdujL1792L06NF44YUXSrtqVEDKwmacMmUKLl++jF27dkGlUtlcLyMjAzt27AAADBw40GL5wIED8c033+DPP//E+PHjpfS//vrLZp7Q0FA4Oztj27ZtyMzMhLOzc2GbQURERERExWHX4dKuQcF0a+3Q4iIjI83+vXXrVoSEhOD333/Hiy++iGeffdah2yPHK9RIckxMDObMmYPhw4ejS5cuea577tw5ZGVloUKFCqhWrZrF8pYtW0plmjpx4oTZclNOTk544oknkJmZifPnzxemCURERERERMWuZ8+eGDp0KADYnIuJyha7g2S9Xo9Ro0bB29tbuu05L9euXQMAqwEyALi5ucHb2xv3799HamoqACAlJQXJycl55jOmx8bG2tsEIiIiIiKiEtOiRQsAQFxcnJQmk8lQs2ZNZGdnY+rUqWjYsCHUajX69u0rrfPgwQPMnDkTLVq0gLu7O9zd3dG+fXssXbrU5rb27t2L4OBgeHh4wNvbG08//TT+/fffYmtbeWT37dbz58/HoUOHsGTJEvj5+eW7flpaGgDA1dXV5jpubm5ISkpCamoqPDw8pDx55XNzcwMAKbDOLSsrC1lZWdK/U1JSAAAajQYajQYAIJfLoVAooNPpoNfrpXWN6Vqt1uwBe4VCAblcbjPdWK6RUpnTvVqttkDpKpUKer0eOp1OSpPJZFAqlTbTbdWdbWKb2Ca2iW0q+23Sq1SQazTQKxQQyod/kmU6HeRaLfRKJYRC8TBdq4Vcp4NepYKQP7zOLTfUQSv0MJ0WRgEZ5DIZNEIPU0rIctaHsJ6eqw8e9++JbWKb2CbLdI1GAyGEtH0hhFnZMpnMcEQp+3LXHcjpS6ttkslsphsZ+8R0fWMs4uTkBCGEtL5er0dYWBj++ecfdOvWDU2bNoWvry/0ej3u3LmDp59+GjExMahUqRK6du0KIQT279+P8PBwHDp0CPPnzzery4YNGzBgwABotVq0bdsWtWrVQkxMDLp27Yrw8HCpvdbqaE9bHZVuuv8a0411LEi66fek0WigUChs/p5y7895sStIvnbtGiZNmoRu3bpJnVxWzZw5E1OmTLFI37JlixR416hRAy1atEBMTIw04g0ADRo0QMOGDXHw4EHcvXtXSm/evDkCAwOxe/dus+C8Q4cOqFixIrZs2WLW+d27d4eLiws2btxoVoeQkBBkZGRg586dUppSqURoaCgSEhKwf/9+Kd3DwwM9evRAXFwcjh8/LqVXqFABHTt2xIULF3Du3DkpnW1im9gmtoltenTa5PryUFRbtBj3unfHvZ7BUrrnwUOotGoV7oSFIaVtGyndd+s2+G/bhhsvD8WD+vWl9ICoVQCA3al3kap/WPcObn6oqHLGluRbZgFxd4+KcJErsDH5pnmbvCojQ6/DTpO28ntim9gmtslam5RKJSpVqoS0tDS4urpCq9UiPT1dWlcul8MTjwa9Xm8x8Obt7W29TZ6eyM7ORkZGhpSuVCrh7u4u/ds0IHZ1dcWDBw+wbt06ADnfUWZmJlxcXADkjCyrVCocPHgQderUgVqtRkpKClJSUvDyyy8jJiYGb7/9Nr788ktpm3fu3MGQIUPw3XffISQkBB07dgSQM3g4cuRIaLVa/PTTTxgwYACAnMByypQpmDdvHoCcwURjHfNrU2ZmptnAo7FNGRkZyM7OltLVajVcXFyQnp5uti+5uLhArVYjNTXVLCB2c3ODSqWS6mHk4eEBmUxmke7p6QkhhM3vKSMjA7t374ZWq7X9NzePQdvcZMKOuch79+6NLVu24MSJE2jYsKGUfvXqVdSqVQt16tTBxYsXzfKsX78eYWFhaNGiBY4ePWq1XB8fHyQlJSElJQUeHh5ISUmBl5cXACA5ORmenpY/sX79+mHt2rVYv349evfubbHc2khy9erVkZCQIJVXFq7CmSovVxbZJraJbWKb2KaCtel8y1YOG0lu/H2k40aSOzUvdJvK4/fENrFNbJNlemZmJuLi4lCzZk24urpaHyncfQSPAtG1lUNGko2jncbvUK/X48qVK5gxYwYiIyOhVqtx8uRJ1K1b12z9FStWYODAgWblHzt2DK1atUKbNm2wf/9+KBQKs+/62LFjaN26Nfr06YM1a9YAAJYsWYJRo0aha9euiI6ONqujRqNB3bp1ER8fj08//RSTJ08uUJsehZHkjIwMXLlyBdWrV4ezs7PN31Nqair8/Pxsxpem7BpJ3rBhA7y9vfHaa6+ZpWdmZgIArl+/jqCgIADA77//jkqVKqFGjRoAgPj4eKtlpqenIykpCT4+PvDw8ACQc6XAy8sLycnJiI+PR+PGjS3yGcsLDAy0Wq5arYZarbZIV6lUFrNxKxQKKExOQoyUSuvdYyvd1izf9qTL5XLI5ZaPittKt1V3toltsjedbWKbALbJVh3tTS9om+TGx390OsDkD7m0XKsFrNweJs91QmuklFmfakRlK93GzZD8ntgmtoltyi9dp9NJ7wIGLG85fpTYqru96UbWvhMPDw8sXbrU7HW4xrLCwsLMvneZTIZt27YBAPr27SuVZ7pOq1at4O7ujoMHD0rpe/fuBQAMGTLEoo5qtRoDBw7E119/bfa9OboP7E23tr8b1y9ourHs3HFe7t+Trf3fGrufSU5KSsKuXbusLsvMzJSWGQPnBg0aQK1W4+7du7h+/TqqVq1qlsc4uty0aVOz9GbNmmH37t04evSoRZCs0Whw6tQpODs7o77J7WZERERERESladiwYQAe3sr8v//9D/3794ePj4/FuhUrVrQ6sHf16lUAwMSJEzFx4kSb2zLGXABw48YNALYHEWvWrFnQJjz27AqSbd2Zndft1i4uLujRowc2bdqElStXYuzYsWbLo6KiAMDilunQ0FDs3r0bUVFReOmll8yWbdiwAZmZmXj22Wf5jmQiIiIiIiozcr8nOS+2YhnjbcidO3dGnTp1HFEtsoPdI8mFMW7cOGzatAnTp09HaGiodJvB/v37sXDhQnh7e2PkyJFmeUaNGoXPPvsM69atw+rVq9G/f38AOQ+qf/DBBwCA8ePHl0T1iYiIiIiISozxdbd9+/YtcMxTuXJlALZfkctX5xac3e9JLozg4GCMGTMGiYmJaN68Ofr27YuQkBB07doVWq0WS5Ysgbe3t1keX19fLF68GHK5HAMHDkSPHj0waNAgNGjQABcvXsS4ceOk55+JiIiIiIjKi549ewKANClXQXTp0gUA8Mcff1gs02q1WLVqlWMq9xgokSAZAL7++mssWbIEjRo1wtatW7F//34EBwdj9+7dZi/MNjVgwADs3r0bTz/9NI4dO4aNGzeibt26iIyMxJw5c0qq6kRERERERCWmXbt26NmzJ/bu3Ys333zT4pVIAHDixAls3rxZ+vegQYPg5+eH6OhoLF26VEoXQmDy5MlmrwujvDnkduuaNWvafF7ZVHh4uN3vV+7UqRM2bdpUyJoRERERERE9en755Rf06tUL33//PX777Tc0b94cVapUQXJyMmJiYhAXF4cxY8agV69eAHJm0F60aBEGDBiA8PBw/N///R9q166NEydO4MKFCxg9ejR+/PHHUm7Vo6HERpKJiIiIiIioYCpWrIh9+/bhm2++QePGjXHs2DFERUUhJiYGtWvXxpdffon33nvPLE9YWBh27tyJ7t2749SpU/jrr79QuXJl7Nq1Cx07diylljx6ZKIgQ8DlQEpKivTu5fxeHk1ERFQSzjRs5LCyGi1c5rCy0K2148oionIpMzMTV65cQa1atfi2GSpVBd0X7YkHOZJMREREREREZMAgmYiIiIiIiMiAQTIRERERERGRAYNkIiIiIiIiIgMGyUREREREREQGDJKJiIiIiIiIDBgkExERERERERkwSCYiIiIiIiIyYJBMREREREREZMAgmYiIiIiIiMiAQTIRERERERGRAYNkIiIiIiIiIgMGyUREREREREQGDJKJiIiIiIiIDBgkExERERERFZFMJrP4qFQqVKlSBQMGDMC+fftKu4oFEh0dDZlMhvDwcLP0yMhIyGQyRERElEq9SpKytCtARERERETlz5QpU0q7CgUyefJkh5Y3bNgw6f9TU1Nx4sQJrF69GmvWrMEvv/yCF154waHbI8djkExERERmHHli6+iTTyKisi4yMtLs33q9Hh9//DG++OILvPPOOxg0aBBUKlXpVK4I+vXrh/bt28Pf37+0q1LseLs1ERERERFRMZHL5Zg6dSqUSiUSExPx33//lXaVCsXLywsNGzZkkExERERERERF4+TkBC8vLwCAVqs1W3b8+HF88MEHaNWqFSpUqAC1Wo3atWvjjTfewI0bN6yWd+rUKbz00kuoXbs2nJ2dUaFCBTRv3hxjx47FzZs3LdY/c+YMwsPDUb16dajVagQEBGDIkCF2Bey2nkkODw+HTCZDdHQ0du/ejR49esDDwwOenp4IDQ3F6dOnbZa5efNmhIaGmrV73LhxSExMLHC9igODZCIiIiIiomJ05coVJCYmQqVSoW7dumbLPv/8c3z11VcAgM6dOyMkJARCCPzf//0fWrdubREoHzlyBG3atMGvv/4KDw8PhIWFoX379tBoNJg3bx7OnTtntv7atWvRokULLF26FP7+/ujTpw9q1aqFP/74A23btsXu3bsd0sY///wTPXr0wIMHDxASEoLKlStj48aN6Nq1K27dumWx/kcffYRnnnkG27ZtQ4MGDdCnTx8olUp89dVXaNeuHW7fvu2QehUGg2QiIiIiIqJikJaWhj179mDw4MEAgNdffx3e3t5m67z66quIj4/HkSNHsGbNGqxZswaXLl3ClClTcPPmTUyaNMls/W+++QaZmZmYPXs2Tpw4gRUrVuDPP//Ef//9hzNnzqBBgwbSulevXsVLL70ElUqFrVu34tixY1i5ciUOHDiAjRs3QqPR4KWXXkJ2dnaR2/r1118jKioKBw4cwIoVK3D69GkMGDAAiYmJ+P77783WXblyJb744gs88cQT+O+//7Bnzx6sXLkS586dw6effopLly5hzJgxRa5TYTFIJiIiIiIichDTV0B5eHigS5cuOHfuHObPn4+vv/7aYv3u3bsjICDALE0ul+PTTz9F1apVsX79erNld+/eBQAEBwdblNWwYUNUrlxZ+vfXX3+N9PR0zJw502L9Xr164fXXX0dcXBz++uuvwjZX8vzzz6Nv377SvxUKBSZMmAAAFqPVn332GQBg+fLlZiPrxtu5mzdvjqioKCQkJBS5XoXB2a2JiIiIiIgcxPQVUFlZWYiNjcW///6LqVOnok6dOnjmmWcs8iQmJmL9+vU4deoUkpKSoNPpAAAajQaJiYm4d+8efH19AQCtWrXCpk2b8Oabb2L69Ono3LkzlErrYd2WLVsAAP3797e6vEuXLvjmm29w8OBB9OvXr0jtfuqppyzS6tevDwBmz0nfuXMHJ06cQL169fDEE09Y5JHJZOjUqROOHz+OI0eO4Omnny5SvQqDQTIREREREZGD5H4FFAAcO3YM3bp1Q58+fXDq1CmzW6KXL1+OV155BWlpaTbLTE1NlYLk999/H3v27EF0dDS6d+8Od3d3dOjQAaGhoQgPD5cmCANybrcGgKpVq+ZZZ0eM2FarVs0izcPDA0DOxYLcdbpw4QJkMlmx16swGCQTEREREREVoxYtWuDVV1/F7Nmz8X//93/SbdexsbEIDw8HkHNrdGhoKKpWrQoXFxcAQMeOHbF//34IIaSyPD09sWPHDuzduxd//vknoqOjsWPHDmzduhUzZ87EP//8g3r16gHIeUczYD66bU27du2K3Ea5vGBP8hrrVKlSpXxHiQMDA4tcr8JgkExERERERFTMatWqBSBnBNVo48aNyM7OxnvvvWd1oqrLly9bLUsmk6Fz587o3LkzgJxbmMeOHYvly5dj4sSJ+OOPPwDkjO5eunQJc+bMgZ+fn6ObVCjGEWd/f3+ro+5lASfuIiIiIiIiKmbGgNfd3V1Ku3//PgDrtyrv3r27wK9BqlixovT+4lOnTknpPXv2BACsWbOmUHUuDtWqVUPDhg1x+vRpnD9/vrSrYxWDZCIiIiIiomJ07Ngx/PDDDwCAkJAQKd04sdUvv/yC9PR0Kf369et47bXXrJa1YMECXLlyxSJ948aNAIDq1atLaePHj4eLiwvee+89rF692iJPVlYWoqKiEB8fX4hWFd4nn3wCvV6PAQMG4Pjx4xbLExMT8eOPP5ZonUzxdmsiIiIiIiIHMT5jDADZ2dmIjY3FgQMHoNfr0bt3bwwdOlRa3qdPHzRp0gSHDx9G3bp10alTJ2RmZmLnzp1o3rw5OnbsiH379pmVv2DBArz++uto3LgxGjVqBKVSibNnz+LEiRNwdnbGp59+Kq1bt25dLF++HC+88AIGDBiAunXrolGjRnBzc8P169dx9OhRpKen49ixY1ZHs4vLCy+8gP/++w8zZsxAq1at0Lx5c9SpUwdCCFy6dAkxMTFwd3fH6NGjS6xOphgkExEREREROcjSpUul/5fL5fD29kbXrl0xdOhQhIeHm01w5eTkhH/++QcTJ07Epk2bsGHDBlStWhVvv/02Pv30U7NRZ6Np06Zh7dq1+Pfff7F9+3ZkZ2ejWrVqGDVqFN577z2zmbMBICwsDDExMZg7dy62bt2KrVu3QqVSoUqVKujduzf69++Pxo0bF1+H2PDZZ5/h6aefxrfffou9e/fi5MmT8PT0RNWqVfH6669j0KBBJV4nI5kwnSqtHEtJSYGXlxeSk5Ph6elZ2tUhIiLCmYaNHFZWo4XLHFbWlOi/HFbW5MmTHVYWEZUdmZmZuHLlCmrVqgVnZ+fSrg49xgq6L9oTD/KZZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMihUkDx37lz0798f9erVg5eXF9RqNQIDA/Hyyy/j5MmTNvNFRkaibdu2cHd3h6+vL0JCQize+5Xb3r17ERISAl9fX7i7u6Nt27b4+eefC1NtIiIiIiIiojwVKkieMWMGNm3aBF9fXzz55JMIDQ2Fs7Mzli1bhlatWmHDhg0WecaOHYvhw4fj1KlTCA4ORtu2bbF161Z07doVa9eutbqdVatWoVu3bti8eTOaNm2KXr164cKFCxg2bBjee++9wlSdiIiIiIiIyCZlYTKtW7cOrVq1sngP1ffff48333wTo0aNQnx8PJTKnOK3bduGefPmwc/PD/v370e9evUAAPv370dQUBCGDx+OoKAgeHt7S2Xdu3cPI0aMgE6nw6pVq9C/f38AwO3bt9G5c2fMmTMHzz77LIKCggrTBCIiIiIiKiIhRGlXgR5zxbEPFmokuVOnTlZf1PzGG2+gTp06uH37Nk6fPi2lz507FwAwadIkKUAGgA4dOuC1115DUlISFi1aZFbWTz/9hJSUFISFhUkBMgAEBARg1qxZAIA5c+YUpvpERERERFQECoUCAKDRaEq5JvS4M+6Dxn3SERw+cZdKpQIAODk5AQAyMjKwY8cOAMDAgQMt1jem/fnnn2bpf/31l808xtu7t23bhszMTMdVnoiIiIiI8qVSqaBWq5GcnMzRZCo1QggkJydDrVZLcagjFOp2a1uWLVuGc+fOoV69etKI8blz55CVlYUKFSqgWrVqFnlatmwJAIiJiTFLP3HihNlyU05OTnjiiSdw+PBhnD9/Hk2bNnVkM4iIiIiIKB/+/v64fv064uPj4eXlBZVKBZlMVtrVoseAEAIajQbJyclIS0tD1apVHVp+kYLkL7/8Ev/99x/S09Nx5swZ/Pfff6hSpQqWL18uDXdfu3YNAKwGyADg5uYGb29v3L9/H6mpqfDw8EBKSgqSk5PzzFetWjUcPnwYsbGxDJKJiIiIiEqYp6cnACAhIQHXr18v5drQ40itVqNq1arSvugoRQqS//77b2zfvl36d2BgIH7++We0atVKSktLSwMAuLq62izHzc0NSUlJUpBszJNXPjc3NwBAamqq1eVZWVnIysqS/p2SkgIg5551433rcrkcCoUCOp0Oer1eWteYrtVqzW4fUSgUkMvlNtNzP5NhnLhMq9UWKF2lUkGv10On00lpMpkMSqXSZrqturNNbBPbxDaxTWW/TXqVCnKNBnqFAkL58E+yTKeDXKuFXqmEMHnGSqbVQq7TQa9SQcgfPjElN9RBK/QwvelRARnkMhk0Qg9TSuSM9GghrKbL5eZPYxn71Vq6TCazGDkyTeffXLaJbSrfbXJxcUHt2rWRlZUFjUYjlf8otwkof99TeWuTTCaDQqGAi4sL9Hq9WTm22pS7zLwUKUjetm0bACApKQknT57E1KlT0a1bN0yfPh0TJ04sStFFNnPmTEyZMsUifcuWLVLgXaNGDbRo0QIxMTHSiDcANGjQAA0bNsTBgwdx9+5dKb158+YIDAzE7t27zYLzDh06oGLFitiyZYtZ53fv3h0uLi7YuHGjWR1CQkKQkZGBnTt3SmlKpRKhoaFISEjA/v37pXQPDw/06NEDcXFxOH78uJReoUIFdOzYERcuXMC5c+ekdLaJbWKb2Ca26dFpk+vLQ1Ft0WLc694d93oGS+meBw+h0qpVuBMWhpS2baR0363b4L9tG268PBQP6teX0gOiVgEAdqfeRar+Yd07uPmhosoZW5JvmQXE3T0qwkWuwMbkm+Zt8qqMDL3O7A4tnU6HkydPwsPDA3Xq1JHSMzIycO7cOfj4+KBGjRpSekpKCi5fvoyAgABUqlRJ6rdH+Xsqj/se28Q2ObpNycnJVtsUGxtrtU1nz5612qZjx45ZbdO+ffustmnHjh1W27R9+3arbdqyZUuB23Tnzh226RFoU4H/5uYxaJubTDjwSXuNRoMOHTrg6NGj+Pfff9GmTRusX78eYWFhaNGiBY4ePWo1n4+PD5KSkpCSkiLdbu3l5QUASE5Otjp83q9fP6xduxbr169H7969LZZbG0muXr06EhISpPLKypUQo/J0dYdtYpvYJraJbcq/TedbtnLYSHLj7yMdNpI8/Z/NZulFGUn+6KOPpLyP6vdUHvc9toltYpvYpsetTampqfDz87MZX5ptL8+ldlKpVHjuuedw5MgR/Pnnn2jTpo10dTk+Pt5qnvT0dCQlJcHHxwceHh4Acp5v8PLyQnJyMuLj49G4cWOLfMbyAgMDrZarVquhVqut1jH3zGcKhQIKheWU4cYvo6DptmZUsyddLpdbnITklW6r7mwT22RvOtvENgFsk6062pte0DbJjbci63SAyR9yablWC1i5PUye66TCSCmz/tIKla10WJ9gx/TEJ790IYTVmW2N6fyba4ltYpvYJrYJYJts1dHe9IK2yVYdrHH4K6D8/f0BQBpmb9CgAdRqNe7evWv1gX7j6HLuybeaNWtmttyURqPBqVOn4OzsjPomt5sRERERERERFYXDg+Rdu3YBgPTckouLC3r06AEAWLlypcX6UVFRAGBxy3RoaKjZclMbNmxAZmYmgoOD4ezs7LjKExERERER0WPN7iB579692Lx5s8UtVxqNBvPnz8eyZcvg4uKC5557Tlo2btw4AMD06dNx4cIFKX3//v1YuHAhvL29MXLkSLPyRo0aBU9PT6xbtw6rV6+W0u/cuYMPPvgAADB+/Hh7q09ERERERERkk93PJF+4cAHDhw+Hv78/WrVqBT8/PyQkJODkyZO4efMmnJ2dERkZierVq0t5goODMWbMGMybNw/NmzdHz549kZ2dja1bt0IIgSVLlsDb29tsO76+vli8eDEGDx6MgQMHIigoCH5+fti2bRuSkpIwbtw4BAUFFbX9RERERERERBK7g+Ru3brh448/xq5duxATE4OEhAQ4OTmhZs2aGDhwIN555x3UrVvXIt/XX3+N5s2b49tvv8XWrVvh5OSE4OBgfPLJJ+jYsaPVbQ0YMAC7d+/G9OnTceDAAWRnZ6Nx48Z46623MGzYMPtbS0REj45dhx1XVrfWjiuLiIiIyjW7g+RatWrhs88+K9TGwsPDER4ebleeTp06YdOmTYXaHhEREREREZE9HD5xFxEREREREdGjikEyERERERERkQGDZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMrD7PclERES2nGnYyGFlNVq4zGFlERERERUUR5KJiIiIiIiIDBgkExERERERERkwSCYiIiIiIiIyYJBMREREREREZMAgmYiIiIiIiMiAQTIRERERERGRAYNkIiIiIiIiIgMGyUREREREREQGDJKJiIiIiIiIDBgkExERERERERkwSCYiIiIiIiIyYJBMREREREREZKAs7QoQERERERGRHXYddlxZ3Vo7rqxygkEyEVFJ4R80IiIiojKPt1sTERERERERGTBIJiIiIiIiIjLg7dZERHk407CRw8pqtHCZw8oiIiIiouLBkWQiIiIiIiIiAwbJRERERERERAa83ZqIiMq9KVOmOKysyZMnO6wsIiIiKns4kkxERERERERkwCCZiIiIiIiIyIC3WxMRERERERUjR74tA+AbM4obR5KJiIiIiIiIDBgkExERERERERkwSCYiIiIiIiIyYJBMREREREREZMAgmYiIiIiIiMiAQTIRERERERGRAYNkIiIiIiIiIgO7g+QHDx5g7dq1GDlyJBo0aABnZ2e4ubmhWbNmmDp1KtLS0mzmjYyMRNu2beHu7g5fX1+EhIRg3759eW5v7969CAkJga+vL9zd3dG2bVv8/PPP9labiIiIiIiIKF92B8m//fYb+vXrh8WLF0OhUKBPnz7o0qULrly5gsmTJ6NNmza4c+eORb6xY8di+PDhOHXqFIKDg9G2bVts3boVXbt2xdq1a61ua9WqVejWrRs2b96Mpk2bolevXrhw4QKGDRuG9957z+7GEhEREREREeXF7iBZpVLhlVdewenTp3H69Gn88ccf2Lx5M86dO4cWLVrg7NmzGDt2rFmebdu2Yd68efDz88OJEyewdu1abN68Gbt374ZCocDw4cORlJRklufevXsYMWIEdDodoqKiEB0djaioKJw9exZ169bFnDlzEB0dXYSmExEREREREZmzO0geNmwYFi5ciEaNGpmlV65cGd999x0AYPXq1cjOzpaWzZ07FwAwadIk1KtXT0rv0KEDXnvtNSQlJWHRokVm5f30009ISUlBWFgY+vfvL6UHBARg1qxZAIA5c+bYW30iIiIiIiIimxw6cVezZs0AAFlZWUhMTAQAZGRkYMeOHQCAgQMHWuQxpv35559m6X/99ZfNPKGhoXB2dsa2bduQmZnpuAYQERERERHRY03pyMIuX74MIOeWbF9fXwDAuXPnkJWVhQoVKqBatWoWeVq2bAkAiImJMUs/ceKE2XJTTk5OeOKJJ3D48GGcP38eTZs2dWQziIiIiIjoETRlyhSHlTV58mSHlUWPFoeOJM+bNw8A0KtXL6jVagDAtWvXAMBqgAwAbm5u8Pb2xv3795GamgoASElJQXJycp75jOmxsbGOawARERERERE91hw2krxx40YsWrQIKpUK06ZNk9KNr4RydXW1mdfNzQ1JSUlITU2Fh4eH2WukbOVzc3MDACmwzi0rKwtZWVnSv1NSUgAAGo0GGo0GACCXy6FQKKDT6aDX66V1jelarRZCCCldoVBALpfbTDeWa6RU5nSvVqstULpKpYJer4dOp5PSZDIZlEqlzXRbdWeb2Ca2yTFtAgC9UgmhUDxcX6uFXKeDXqWCkD+81ijXaiHT6aB3coKQyR6mazSQ6fXQiIdlA4ASOetoIQqUrpLJoRcCOgjA0D9l7XsSMhkgBPSGC6VS+VlZgEwGvZOTWboiKwtCLodepXpYdyEgz85+2FZjOgClTA6dENCbpMshg0Ims5muFXrpuwQAIQSEEGZpAKT+yC/d2GZH7Ht6lQpyjQZ6hQJC+fBPskyng1yrtWvfAwCt0JvtNQrIIJfJ7N737OkbmUwGmcn+njudf3PZJraJbSrJNpnmM1XQY7xpuiPbpDP5u1jY8wjTMjRCX7TzCGNdkBMQPg77Xu68eXFIkHz27Fm89NJLEELgyy+/lJ5NLk0zZ860ervFli1bpMC7Ro0aaNGiBWJiYqQRbwBo0KABGjZsiIMHD+Lu3btSevPmzREYGIjdu3ebBecdOnRAxYoVsWXLFrPO7969O1xcXLBx40azOoSEhCAjIwM7d+6U0pRKJUJDQ5GQkID9+/dL6R4eHujRowfi4uJw/PhxKb1ChQro2LEjLly4gHPnzknpbBPbxDY5tk3OAO6EhSGlbRsp3XfrNvhv24YbLw/Fg/r1pfSAqFXwOnQI1956E9kBAVJ61UWL4Hb+ArYk3zL7g9XdoyJc5ApsTL5p3iavysjQ67Az9eHr9JSQIdS7ChK0WdifnghsvFnoNhXn9xRYsQKUScm4NNX8+Fvn08nQenshdtw4KU2emYm6kyPwoG4dXB85Ukp3un0bNed+hbjsBziekfSwTUo1Orr740JmKs5lPaxLDSdXtHD1QUxGEq5lP3jYJrUHGrp44mD6PbPHcq5du4Z79+6hXr16cHFxkdIvXbqE1NRUNGnSBAqTwPTs2bPIzs6WyjDug47Y91xfHopqixbjXvfuuNczWEr3PHgIlVatsmvfA4DdqXeRqn/4fXRw80NFlbPd+55pf+l0Opw8eRIeHh6oU6eOlJ6RkYFz587Bx8cHNWrUkNJTUlJw+fJlBAQEoFKlSlJ/lddjBNvENrFNZatNHh4eBTqWG8XExMDJyQkNGzaU0ozHPYe2yeTvYmHPI658PAF6Z2cAwKXkm0U7jzC2Sa5ED+Cx2PfyGrTNTSasXXqxw/Xr19GpUyfExsZi3LhxFjNOr1+/HmFhYWjRogWOHj1qtQwfHx8kJSUhJSUFHh4eSElJgZeXFwAgOTkZnp6eFnn69euHtWvXYv369ejdu7fFcmsjydWrV0dCQoJU3qN8JaQ8Xt1hm9imstim802ecNhIct0FS83rXpQrwJ1aFLpNxfk9XWjazGEjyQ0W/OywkeSZ//wtpRd1JPmjjz4C4Jh973zLVg4bSW78faTDRpKn/7O5wH2T30iysb/K6zGCbWKb2Kay1aYZM2ZI+UwVZiT5k08+cVibzrRs9XB9B4wkN5j3g+NGkoPaPhb7XmpqKvz8/GzGl2bby3NpPu7du4ennnoKsbGxGD58OGbPnm2xjvHqcnx8vNUy0tPTkZSUBB8fH3h4eAAAPD094eXlheTkZMTHx6Nx48YW+YzlBQYGWi1XrVZLz0WbUqlUUJmcjAE5X5LplSYjpdJ699hKz11uYdLlcrnFjzSvdFt1d3SbimMShNJuU1HSy+r3VJR0tsl23eVaLWDlFh15rgO7lG7yCjyzOsqsTwOhgqzA6XKZDHLIgFztLSvfk8zwB1BhcpFSIoTVdJlebzVdamvuustkUNiRrpTJzf6QG1lLK0h67jYXZd8z7kNynQ4w+UMuLbdz31Pa2sfs3Pfs6RvjRQdb6fyba4ltYpvYplx133X4Ybrhk5utoMVWemGP8aYc+j1Z+ztn53mEaRmmx/VCnUfkTn8M9j1bdbCm0EFyWloannnmGZw+fRr9+/fHjz/+aHElGcgZeler1bh79y6uX7+OqlWrmi03ji7nvvWhWbNm2L17N44ePWoRJGs0Gpw6dQrOzs6ob3KLAhEREVFJ42y6RETlS6Fmt87KykJYWBgOHjyIp59+GsuXL7d6JQEAXFxc0KNHDwDAypUrLZZHRUUBgMUt06GhoWbLTW3YsAGZmZkIDg6Gs+G+fCIiIiIiIqKisnskWafT4fnnn8eOHTvQpUsXrF69Gk65njHLbdy4cdi0aROmT5+O0NBQ1KtXDwCwf/9+LFy4EN7e3hhpMmkLAIwaNQqfffYZ1q1bh9WrV6N///4AgDt37uCDDz4AAIwfP97e6hMRlQscuSIiIiIqHnYHyd9++y3WrFkDAPD398cbb7xhdb3Zs2fD398fABAcHIwxY8Zg3rx5aN68OXr27Ins7Gxs3boVQggsWbIE3t7eZvl9fX2xePFiDB48GAMHDkRQUBD8/Pywbds2JCUlYdy4cQgKCrK3+kREREREREQ22R0k379/X/p/Y7BsTUREhBQkA8DXX3+N5s2b49tvv8XWrVvh5OSE4OBgfPLJJ+jYsaPVMgYMGIDdu3dj+vTpOHDgALKzs9G4cWO89dZbGDZsmL1VJyIiIiIiIsqT3UFyREQEIiIiCrWx8PBwhIeH25WnU6dO2LRpU6G2R0RERERERGSPIr0CiojIUfiMLRERERGVBYWa3ZqIiIiIiIioPGKQTERERERERGTAIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiIiIDzm5NRERERFRG8e0PRCWPI8lEREREREREBgySiYiIiIiIiAwYJBMREREREREZ8Jnk8mzX4dKuARERERER0SOFQTIREREREdFjipPDWeLt1kREREREREQGHEkuY840bOSwshotXOawsoiIiIiIiB4HDJKJiIiIiMguHNih8oy3WxMREREREREZMEgmIiIiIiIiMmCQTERERERERGTAIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAyUpV0BovJqypQpDitr8uTJDiuLiIiIiIhs40gyERERERERkQGDZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMmCQTERERERERGTAIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZ2B0kHzlyBJ9//jn69++PatWqQSaTQSaT5ZsvMjISbdu2hbu7O3x9fRESEoJ9+/blmWfv3r0ICQmBr68v3N3d0bZtW/z888/2VpmIiIiIiIioQJT2Zpg2bRrWrVtnV56xY8di3rx5cHFxwVNPPYXMzExs3boVW7ZsQVRUFPr27WuRZ9WqVXjuueeg1+vRtWtX+Pv7Y/v27Rg2bBhiYmIwe/Zse6tORERERERElCe7g+QOHTqgadOmaNOmDdq0aYOaNWsiKyvL5vrbtm3DvHnz4Ofnh/3796NevXoAgP379yMoKAjDhw9HUFAQvL29pTz37t3DiBEjoNPpsGrVKvTv3x8AcPv2bXTu3Blz5szBs88+i6CgIHurT0RERERERGST3bdbf/jhh5g6dSp69+6NSpUq5bv+3LlzAQCTJk2SAmQgJ9h+7bXXkJSUhEWLFpnl+emnn5CSkoKwsDApQAaAgIAAzJo1CwAwZ84ce6tORERERERElKdinbgrIyMDO3bsAAAMHDjQYrkx7c8//zRL/+uvv2zmCQ0NhbOzM7Zt24bMzExHV5mIiIiIiIgeY8UaJJ87dw5ZWVmoUKECqlWrZrG8ZcuWAICYmBiz9BMnTpgtN+Xk5IQnnngCmZmZOH/+fDHUmoiIiIiIiB5XxRokX7t2DQCsBsgA4ObmBm9vb9y/fx+pqakAgJSUFCQnJ+eZz5geGxvr6CoTERERERHRY8zuibvskZaWBgBwdXW1uY6bmxuSkpKQmpoKDw8PKU9e+dzc3ABACqytycrKMptQLCUlBQCg0Wig0WgAAHK5HAqFAjqdDnq9XlrXmK7VaiGEkNIVCgXkcrnNdGO5RkplTvdqtdoCpatUKgi5HHqVSkqTCQF5djaEQgG98uHXJdPrIddooFcoIEzTdTrItVrolUpohEmbIINCJoNW6PGw5oACMsjzSDeWIZfnXE8x9pPx30b2pGu1WiiVSuj1euh0uod1l8mgVCptfh9l6XuyVXfTdNM+y/2qNCEEhBA20631Y1lok2m6o78nR+5jGo3GYW0CAL1SCaFQPFxfq4Vcp4Pe8JuV1tdqIdPpoHdygjD5XuUaDWR6vdlvEgCUyFlHa/brs52uksmhFwI6mO8j1vYxW+nW9j2NRuOw35OQyQAhoFerzdLlWVmATAa9k5NZuiIry+Zxz9hWKR2AUiaHTgiYHrGMxzdb6VqhN+uvvH5nQP77mLHNjvg96VWqfI/lBd33AOR7LDfKb9+zp2/y2/fK8t9cRx33CnosL8i+Z9quR/FYnju9LH1Pj2qbjP8tynmEMd2RbdKZHOfzOpYX5BxWI/QFOpYX5BxWyleE8whjuiP3PdP+Kux5hGkZGqEv0nmEVBeTNhT2PMI0vSz/nnLnzUuxBsmlaebMmZgyZYpF+pYtW6Tgu0aNGmjRogViYmKkUW8AaNCgARo2bIiDBw/i7t27Unrz5s0RGBiI3bt3mwXoHTp0QMWKFbFlyxazzu/evTtcXFywceNGszqEhIQgIyMDO3fulNKUSiVCQ0PxoG4dXB85Ukp3un0bNed+hZSWLXF74AAp3fX8eVRbtBj3unfHvZ7BUrrnwUOotGoV7oSF4WLyzYdtUnugoYsnDqbfw13tw4sHzV28Eah2w+7Uu0jVP6x7Bzc/VFQ5Y0vyLWgh0LRpUwDA2bNnkZ2dLf3bKCYmBk5OTmjYsKGUptPpcPLkSXh4eKBOnTpS+u7du9GjRw/ExcXh+PHjUnqFChXQsWNHXLhwAefOnZPSy+L3lJCQgP3790vpHh4eFm1q2rQpUlJScPnyZQQEBJhNdJeYmIi4uDhUq1YNfn5+UvqtW7dw69Yt1KxZE56enlJ6XFxcmWgTUHzfk3GfunTpElJTU9GkSRMoTIIDe/a9LVu2OKxNzgDuhIUhpW0bKd136zb4b9uGGy8PxYP69aX0gKhV8Dp0CNfeehPZAQFSetVFi+B2/oL0ezLq7lERLnIFNpr8VgEgxKsyMvQ67Ey9I6UpIUOodxUkaLOwPz1R6oeMjAycO3cOPj4+qFGjhrS+Pfvexo0bHfZ7CqxYAcqkZFyaan78rfPpZGi9vRA7bpyUJs/MRN3JETaPe3HZD3A8I+nh96RUo6O7Py5kpuJc1sO61HByRQtXH8RkJOFa9gMp3fS4Z7rfXLt2Dffu3UO9evXg4uIipRd03zP+rhzxe3J9eWi+x/KC7nsA8j2WG+W375n2l61jeUH3PWN/ParHciD/415Bj+UF2fdM6/8oHsuNyuL39Ki2CUCRzyOM+55D22Q4zud3LC/IOeyl5JsFOpYX5BzWw8OjyOcRxuOeQ/c9k7+LhT2PuPLxBOidnaU+K8p5hNQmeU44WJTzCODhvleWf095DdzmJhOmIX0hODs7IysrC9aKWb9+PcLCwtCiRQscPXrUan4fHx8kJSUhJSUFHh4eSElJgZeXFwAgOTnZ7Adu1K9fP6xduxbr169H7969rZZrbSS5evXqSEhIkMosi1cWTzdu4rCR5PrfPpw1vKgjyZ//8zcAx4zyTZgw4bG4Avz5559LfVDUK8ATJ04sE20yTXf092TaX8Z8puxJ/+ijjxzWpvNNnnDYSHLdBUvN6l6UK8DG36SxD4pyBfijjz5y2O/pQtNmDhtJbrDgZ4eNJM806a+ijiR/9NFHABzzezrfspXDRpIbfx/psJHk6f9sLnDf5LfvGfvrUT2Wm6bbqvvUqVMdNpJs7K/SblN5/J4e1TZ99tlnDhtJnjhxosPadK71wwt4RR1JbjDvB4eNJM/YtSknnwNGkj/55BOH7XtnWrZ6uL4DRpIbzPvBYSPJn+3a5LCRZFv7WFn4PaWmpsLPz89mjGm2vTyXFpHxakR8fLzV5enp6UhKSoKPjw88PDwAAJ6envDy8kJycjLi4+PRuHFji3zG8gIDA21uW61WQ53rJA3I6UiVyQ8YyPmSTK80GSmV1rvHVnrucguTLtProbDy3mmZTgeFyZdsJNfpAGvpWi1UMstHzpVW0vJKN5Zh+oO39m970o39J5fLLQ5IgO3voyx9T7bqbppu2mbjgSM3W+m5+9FYZmm3yZSjvydH7mPG9jmqTXKtFrByi44814FdSs/Otppu7TcJACrICpwul8kgh8yiH+zdx0zTTfeHov6eZIYyrR3HIIT145uN456xrbkpZDIo7EhXyuRW95vC7mO521yU35NxH8rrWG7Pvpffsdwi3ca+Z0/f5LePleW/uY46RhT0WF6QdGv1fJSO5QWpo73pbFPRzyOMHNmm3MftopzDmh6j8jqWW2MrvSjnEVIdHbnvWfs7Z+d5hGkZpn1WmPOI3IpyHmGqLP+ebNXBmmINkhs0aAC1Wo27d+/i+vXrqFq1qtly4+hy7tsemjVrht27d+Po0aMWQbJGo8GpU6fg7OyM+ia3JxAREVH5dqZhI4eV1ejsGYeVRURE5Uuxzm7t4uKCHj16AABWrlxpsTwqKgoALG6ZDg0NNVtuasOGDcjMzERwcDCcDffkExERERERETlCsQbJADDOMEnL9OnTceHCBSl9//79WLhwIby9vTHS5CF/ABg1ahQ8PT2xbt06rF69Wkq/c+cOPvjgAwDA+PHji7vqRERERERE9JixO0j+66+/0L59e+mTbbhn3jTtr7/+ktYPDg7GmDFjkJiYiObNm6Nv374ICQlB165dodVqsWTJEnh7e5ttw9fXF4sXL4ZcLsfAgQPRo0cPDBo0CA0aNMDFixcxbtw4BAUFFanhRERERERERLnZ/Uzy3bt38e+//1qkm6aZTvsNAF9//TWaN2+Ob7/9Flu3boWTkxOCg4PxySefoGPHjla3M2DAAOzevRvTp0/HgQMHkJ2djcaNG+Ott97CsGHD7K02ERER0UO7Dpd2DYiIqIyyO0gODw9HeHi43RsqTL5OnTph06ZNdm+LiIiIiIiIqDCK/ZlkIiIiIiIiokdFsb4Ciqi4OfJ1IABfCUJERERE9LjjSDIRERERERGRAYNkIiIiIiIiIgMGyUREREREREQGDJKJiIiIiIiIDDhxF5EpvjeTiIjoseTIyUA5ESjRo41BMhERERGRI/GiO9EjjbdbExERERERERkwSCYiIiIiIiIyYJBMREREREREZMAgmYiIiIiIiMiAQTIRERERERGRAWe3JiIiIqK8OXK25m6tHVcWEVEx4EgyERERERERkQFHkomIiIioxEyZMsVhZU2ePNlhZRERGXEkmYiIiIiIiMiAI8lERERE5dCZho0cVlajhcscVhYRUVnHkWQiIiIiIiIiA44kE1HhOXK2UyIiIiKiMoAjyUREREREREQGDJKJiIiIiIiIDBgkExERERERERkwSCYiIiIiIiIy4MRdRI8ZvhKEiIiIiMg2jiQTERERERERGTBIJiIiIiIiIjJgkExERERERERkwCCZiIiIiIiIyIBBMhEREREREZEBg2QiIiIiIiIiAwbJRERERERERAYMkomIiIiIiIgMGCQTERERERERGTBIJiIiIiIiIjJgkExERERERERkwCCZiIiIiIiIyIBBMhEREREREZFBmQ+SMzIy8Omnn6J+/fpwdnZGlSpVMGLECFy/fr20q0ZERERERETlTJkOkjMzM9GjRw9MmzYNaWlpCAsLQ/Xq1bFkyRK0aNECly9fLu0qEhERERERUTlSpoPk6dOn48CBA+jQoQPOnz+PFStW4N9//8WcOXNw9+5djBgxorSrSEREREREROVImQ2Ss7Oz8e233wIAvvvuO7i7u0vLxo0bh6ZNm2LXrl04cuRIaVWRiIiIiIiIypkyGyTv3bsXycnJqFOnDlq0aGGxfODAgQCAP//8s6SrRkREREREROVUmQ2ST5w4AQBo2bKl1eXG9JiYmBKrExEREREREZVvZTZIvnbtGgCgWrVqVpcb02NjY0usTkRERERERFS+KUu7ArakpaUBAFxdXa0ud3NzAwCkpqZaXZ6VlYWsrCzp38nJyQCAe/fuQaPRAADkcjkUCgV0Oh30er20rjFdq9VCCCGlKxQKyOVym+nGco2Uypzu1Wq1BUpXqVRIFQJ6lUpKkwkBeXY2hEIBvfLh1yXT6yHXaKBXKCBM03U6yLVa6JVKJKalPGwTZFDIZNAKPR7WHFBABnke6RqR0y/Z2dkAIPWTXG5+fcWe9Hv37kGpVEKv10On0z2su0wGpVJp8/uwlp6m00GvUkGYlC/XaiHT6aB3coKQyR6mazSQ6fXQqdVmdZRnZwNCQK9Wm/WZEjl5tWY9A6hkcuiFgM4kXQZAmSvdtM9kMhlkJnURQkAIYTM9dz8mJSU5bN9L1euhd3IyS1dkZUHI5Xbve9b2MZ0QMN2bCrrvOXIfS0xMhEqlsrmP2bPvpel00CuVEArFw/W1WsgLse+Z9hdgex8ryL5n7C9jH+Tel2ylW9v3EhMTHXbcS9Xrpd+TKXlWFiCT2bXvJaWlWv2d2drH8tr3TPvL1u+soPtYYmIigLyP5QXd91Lk8nyP5QXd91LS0/I9lhvlt++Z9ld+fZPfvmfsL0f8zU1WKgt0LDdLt7HvpaSnFehYbppuax/Lysoq0LG8IPuesb+Awp1HmO5jyUplkc4jTPe9+2mpRTqPkOoOGTIzM4t0jDemm+5fQOHOI0z3vWRDPxTlPMIoMS2lyOcRxvTMzMwin0cY0+/du+ewc9hkk/2msOcRxn0vMS2lyOcRRpmZmTn5HLCPJSUlFfk8wrjvmfZXYc8jdAXcx+zd96ztY8Y+sHffs7WPFWf8VNC/uca40bQeNokyavTo0QKAmDhxotXlFy5cEABEvXr1rC6fPHmyAMAPP/zwww8//PDDDz/88MMPPwKAiIuLyzcWLbMjycbZrB88eGB1eXp6OgDAw8PD6vIJEyZg3Lhx0r/1ej3u3bsHPz8/i6skj7OUlBRUr14dcXFx8PT0LO3qlHnsL/uwv+zD/rIP+8s+7C/7sc/sw/6yD/vLPuwv+7C/LAkhkJqaiipVquS7bpkNkmvUqAEAiI+Pt7rcmB4YGGh1uVqthjrX7TDe3t6Oq2A54+npyR+QHdhf9mF/2Yf9ZR/2l33YX/Zjn9mH/WUf9pd92F/2YX+Z8/LyKtB6ZXbirmbNmgEAjh49anW5Mb1p06YlViciIiIiIiIq38pskNypUyd4eXnh0qVLOH78uMXyqKgoAEDv3r1LuGZERERERERUXpXZINnJyQlvvfUWAODNN9+UnkEGgLlz5yImJgbdunVDq1atSquK5YJarcbkyZMtbk0n69hf9mF/2Yf9ZR/2l33YX/Zjn9mH/WUf9pd92F/2YX8VjUyIgsyBXToyMzMRFBSEf//9F5UrV0aXLl0QGxuLf//9FxUqVMCBAwdQu3bt0q4mERERERERlRNlOkgGgIyMDMycORO//fYb4uLi4Ovri169emHatGmoVq1aaVePiIiIiIiIypEyHyQTERERERERlZQy+0wyERERERERUUljkFxOHDx4EDKZDDKZDFOnTs13/cjISLRt2xbu7u7w9fVFSEgI9u3bZ3Xds2fP4osvvkD37t3h7+8PlUqFSpUqoX///vjnn38c3ZQSUZz9Zc20adOk7f3yyy9FqXqpKKn+Wrt2LXr16oUKFSrA2dkZ1atXR79+/bBnzx5HNKPEFHd/6XQ6zJ8/H61atYKbmxu8vLzQtWtXrF692pHNKDEF7a+4uDh8//33CA8PR6NGjSCXyyGTyRAdHZ3vNvbu3YuQkBD4+vrC3d0dbdu2xc8//+zAVpSc4uyv9PR0LFu2DG+//TbatWsHtVoNmUyGiIgIxzekhBRnf0VHR0tlW/u0b9++GFpUvEri9wiUn+M9UPA+W79+PYYNG4b//e9/0vlUxYoVERISgg0bNljNExMTg7feegvt27dHlSpVoFar4eXlhQ4dOmD+/PnQaDTF1axiU9D+ioiIyPP39dFHHxVoe4/LOVhh+6u8nVM4jKBy4a233hIABABRv379PNcdM2aMACBcXFxEWFiYePrpp4VSqRQKhUKsWbPGYv2qVasKAMLd3V0EBweLwYMHiyeeeEIAEDKZTHz11VfF06hiVJz9ldvZs2eFWq0WMplMABDLli1zUCtKTnH3l06nEyNGjBAAhJubm3j66afFc889Jzp06CCcnJzEtGnTiqFVxac4+0ur1Ypnn31W+k0+9dRTIjg4WLi6ugoAYvLkycXTqGJU0P766quvpPVMPzt37syz/KioKKFQKIRMJhPdunUTAwYMEN7e3gKAGD9+vINbU/yKs7+OHTtmNc+juF8ZFWd/7dy5UwAQderUEcOGDbP4TJ06tRhaVLyK+/dY3o73QhS8zwYMGCBkMpl44oknREhIiHjuuedEu3btpLwTJkywyDN//nwBQAQGBoonn3xSDBkyRDz55JPC2dlZABDdunUTWVlZxdk8hytof02ePFkAEJ06dbL6+/rjjz/y3dbjdA5WmP4qj+cUjsIguRzIzs4W/v7+AoCoVKmSACAOHDhgdd2tW7cKAMLPz0+cP39eSt+3b59wcnIS3t7e4v79+2Z5nnzySfHzzz+LjIwMs/QFCxYIAEKhUIj//vvP4e0qLsXdX6b0er3o2rWrCAgIEGFhYY/kAbok+st4YO/du7dITEw0W3bv3j2zssq64u6v2bNnCwCiZs2a4uLFi1L6mTNnRJUqVQQAsW/fvmJpW3Gwp7/WrVsnxo4dK3799Vdx/vx58dRTT+V7Up6YmCg8PT0FALFq1Sop/datW6Ju3boFOqkvS4q7vy5evChGjhwpFixYII4cOSKmTp36SJ8oFXd/GYPkYcOGFU8DSlhx95cQ5et4L4R9fXb06FGRkJBgkX7gwAHh7u4uZDKZiImJMVt26dIlcenSJYs8t27dkgYs5s+f75jGlAB7+su4ryxZsqRQ23rczsEK01/l7ZzCkRgklwPr16+XrhwZT2jefPNNq+s+88wzAoDV0d933nlHABCzZ88u8LaNfxQjIiIKW/0SV5L99cMPPwgA4pdffhHDhg17JA/Qxd1fcXFxwsnJSdSoUUM8ePCgOJpQooq7v+rUqSMAiF9//dUij3F/69u3r0PaUhLs6a/cnn766XxPyr/44gsBQISFhVksW716tQAgnn322ULWvuQVd3/lNnPmzEc6SC7u/ipvQXJx91d5O94LUbQ+MzVy5EgBQMybN6/AeZYtWyYAiH79+tm9vdJiT38VNUh+3M7BCtNf5e2cwpH4THI5YHy+4qWXXsJLL70EAFixYoXFcyoZGRnYsWMHAGDgwIEW5RjT/vzzzwJvu1mzZgCAGzdu2F/xUlJS/XXr1i188MEHePLJJ/Hiiy86rP4lrbj7a+nSpcjOzsaoUaPg4uLi8PqXtOLsr+TkZFy6dAkAEBQUZJGne/fuAIC///4bWVlZRWxJyShofxXWX3/9BcB6H4eGhsLZ2Rnbtm1DZmamQ7ZX3Iq7v8ob9pd9iru/ytvxHnBcn6lUKgCAk5NTseYpbSX1m3zczsEKozyeUzgSg+RHXHJyMtavXw8nJycMHjwYtWrVQseOHZGQkIDNmzebrXvu3DlkZWWhQoUKVt8x3bJlSwA5k0QU1OXLlwEAlSpVKkIrSk5J9tc777yDjIwM/N///Z/jG1JCSqK/jIFix44dcfPmTcyePRuvvfYaPvzwQ2zevBniEXpLXXH3V3p6uvT/Pj4+Fnn8/PwA5ATg58+fd0ibipM9/VVYJ06cAPCwP005OTnhiSeeQGZmJvurHCrJ/rpw4QImTJiAV155BR9//DE2btwIvV7v0G0Ut5Lor/J0vAcc12cnT57EihUroFKp0LNnzwLluX//PubMmQMg54Lfo6Cw/bVjxw6MHTsWr732GqZPn44jR47ku63H7RzMVEH7q7ydUzgag+RHXFRUFDIzM/HMM8/A19cXAKQrTcuWLTNb99q1awBg9YQcANzc3ODt7Y379+8jNTU1321funRJmo2xT58+hW5DSSqp/tqwYQNWrlyJjz/+GPXq1XN0M0pMSfTX6dOnpf82atQI77//PhYuXIhZs2bhmWeeQY8ePZCUlOTophWL4u4vX19fKBQKAEBsbKxFnitXrkj/b215WWNPfxVGSkoKkpOTAdjuZ2M6+6v8Kcn+2rdvHz7//HP8+OOPmDlzJkJDQ9G8eXNcuHDBodspTiXRX+XpeA8Uvs/+/PNPhIeH48UXX0SXLl3QvHlzPHjwAD/++CPq1KljNc+FCxcQHh6Ol19+GU8//TRq1KiBQ4cO4bXXXntkRkoL21/Lli3DvHnzsHDhQnzyySdo3bo1Bg4ciLS0NKvrP47nYKYK2l/l7ZzC4Ur7fm8qmm7dugkAYuXKlVJaQkKCUKlUwtnZWSQlJUnpv/76q/Rcgy3GmayvX7+e53Y1Go3o3LmzACCee+65ojekhJREf6Wmporq1auL+vXri8zMTCn9UXwepiT6S61WCwBCqVSKLl26iKNHj4qUlBSxbds2UatWLQFADBw4sHga6GAl0V/t27cXAMSHH35osf6bb74pzYD522+/OahVxcee/rImv2cgr1+/LvWHRqOxus6LL75o83mssqa4+8uaR/mZ5JLor6NHj4r3339fHDhwQCQmJorExESxfft26Xdas2bNfLdTVpREf5Wn470Qhe+zadOmSccmIOftBj/88IPQ6XQ2t/XPP/+Y5QEg3nnnHZGcnOzwdhUXe/tr2bJlYvbs2eK///4TaWlpIi4uTvz666/S30Zrz8o+rudgQhSuv8rTOYWjMUh+hMXGxgqZTCa8vb3NDgRCCGkWvx9//FFKc2SQ/PrrrwsAonbt2hazU5ZVJdVfb7/9tgAgtm/fbrbuo3aALqn+UqlUAoCoUKGCSEtLM1v/5MmT0msbzp0756CWFY+S6q+1a9dKJ5lffvmluHnzprh+/bqYPn26UCgUQqlUCgDi999/d3wjHcje/rLmcQqSS6K/rHlUg+TS6i8jrVYrunTpIgCIGTNm2J2/pJVUf5WX470QjumzjIwMcfLkSfHGG29Ikwjm9zonrVYrLl++LObMmSM8PT1FvXr1xJUrV4ranGLniP4yunHjhvDz8xMAxP79+82WPa7nYHnJq7/KyzlFceDt1o+wX3/9FUIIDBw4EGq12myZ8XYM05emu7u7AwAePHhgs0zj8wkeHh421/nss8/wf//3fwgICMDff/8t3QJS1pVEfx08eBDfffcdhg4dih49eji0/iWtpPYvY75BgwbBzc3NbP0nnngCbdq0AQDs3r27sE0pESXVX2FhYfjiiy8ghMD777+PypUro2rVqpg0aRJGjhyJFi1aALD+fFFZYm9/FYaxjwHb/VyQY15ZUBL9VZ6Udn8pFAp8+OGHAHImvSnrSqq/ysvxHnBMnzk7O+OJJ57Ad999h7fffhsbNmzA/Pnz88yjUChQq1YtjBs3DkuWLMGFCxfw9ttvF60xJcCR+1jlypUxfPhwADB7NvdxPgfLi63+AsrPOUVxUJZ2BajwjM8jREdHo3PnzmbLsrOzAeT8oYmNjUVgYCBq1KgBAIiPj7daXnp6OpKSkuDj42PzhHHBggWYNGkSvLy8sHnzZtStW9dRzSl2JdFfxslaTp48aTFT4NmzZwHkXGT46aef0KtXL3z00UcOa5+jldT+FRgYiPv376NmzZpW89WsWRMHDx7EnTt3itqkYlWSv8cPPvgA/fr1Q1RUFK5evQovLy+EhoaiW7du0jO2TZo0cWj7HM3e/ioMT09PeHl5ITk5GfHx8WjcuLHFOsb+L+w2SkpJ9Fd5Uhb6y/gs5M2bN4ulfEcqqf4qL8d7wPF9NnToUMyfPx/r1q3D+PHjC1SHfv36wd3dHZs3b0Z2dnaZnuXa0f1l7ff1OJ+D5Sev41F5OKcoFqUyfk1FdvjwYYtnU2x9PvvsMyGEEA8ePJCeB4qPj7coc/fu3QKA6Natm9VtLl++XMjlcuHq6ir27NlTnM1zuJLqL+M76gryKcvv1SzJ/Wv48OECgJgwYYLVugQHBwvAvndHlrTS+D1aExsbKwCIunXrOqppxaIw/WVNQW7v7Nq1q81b7LKzs4Wzs7NwdnYWGRkZjmhasSjJ/srtUbzdujT7y9SBAwcEANGiRYtCtqRklGR/lYfjvRCO6zNTly9fFgBEw4YN7apLjRo1BABx69atwjSlRBRHf33++ecCgHj33XeltMf5HCw/1vorP4/KOUVx4e3WjyjjLRbvvfceRM6z5Raf6Ohos3VdXFyk209WrlxpUWZUVBQAoHfv3hbLNm7ciJdffhlKpRJr1qxBp06diqNZxaak+isiIsJm+cOGDQOQc3VQCIHIyMjiam6RleT+ZZwZfdeuXRZ50tLScPToUQCQbvkpi0r692iL8Ta9V155pdBtKQmF6a/CMr4axdifpjZs2IDMzEwEBwfD2dm5SNspTiXZX+VBWemvVatWAbD++rGypCT7qzwc74Hi6TNjn9ia3dqay5cvIy4uDp6envD397evESXI0f0lhMCaNWsAmP++HudzsLzY6q/8PCrnFMXG3qiaSp9WqxUBAQECgDhy5IjN9XQ6nTTxz+HDh4UQQmzdulUAEH5+fuL8+fPSuvv27RNqtVp4e3uL+/fvm5WzZ88e4eLiIpRKpVizZk1xNKlYlXR/2fKoTBpR0v2l1WpFo0aNBADx3XffmaWPHj1aABBPPPGE0Ov1jm2og5R0f6WlpYnTp09blL9gwQKhUChEgwYNLCb5KEuK0l+5FWTkKjExUXh6egoAYtWqVVL67du3Rd26dYs0UlgSSrq/cnvURpJLur+++uorce3aNbM0vV4vFixYIJRKpZDJZDbLLwtKur8e9eO9EIXvszt37ogffvhBpKenW6y7ZcsWUaFCBQFAREVFmS375ptvxM2bNy3ynD17VrRr104AEG+99VbRG1ZMitJf3377rUhJSTFbLzU1Vbz66qsCgKhUqZLV/rSmvJ+DFba/HvVziuLEIPkRtHHjRgFA1K9fP991x40bJwCIMWPGSGljxowRAISrq6sICwsTzzzzjFAqlUKhUFgNgr29vQUAUatWLTFs2DCrn4LOsFcaSrq/bHlUDtCl0V/Hjh2TAplmzZqJAQMGiNq1a0sBZExMjINa53gl3V9XrlwRAESTJk1E3759xaBBg6S+qlmzprh8+bIDW+d4RemvGzduiHbt2kkfLy8vAUA0atRISps6dapFOVFRUUIulwuZTCa6d+8uBg4cKB3Xxo0b5+gmOlRp9Fffvn2l5dWrVxcARNWqVaU0a68RKStKur8CAwOFQqEQbdq0EYMHDxZ9+vSRXmUkl8vF/Pnzi6OZDlMa+9ejfLwXovB9Zjx2u7q6ii5duojnn39e9OnTRzRs2FC6ddbarbCBgYFCLpeLFi1aiEGDBomBAweKNm3aCLlcLgCIrl27itTU1OJoqkMUtb/c3d1F9+7dxQsvvCB69uwpzdLs7e1t16N/5f0crLD99aifUxQnBsmPoOeff77AV/YPHTokAIiKFSuavQJlyZIlolWrVsLV1VV4e3uLXr16ib1791otw3jwzutTlp/tKOn+suVROUCXVn9dvnxZvPzyy6JSpUpCpVKJatWqiVGjRomrV68WtUnFqqT7Kzk5Wbz22muicePGwsPDQ7i6uoomTZqIyZMnl+kTJaOi9Jfxj3lhjkV79uwRvXr1Et7e3sLV1VW0bt1aREZGOrZxxaA0+iswMDDPPIGBgQ5vp6OUdH9988034tlnnxW1atUSbm5uwsnJSQQGBoqXXnpJHDx4sHga6UCl9Xt8VI/3QhS+z5KTk8WsWbNESEiICAwMFC4uLkKtVouaNWuKIUOG2ByB/+WXX8QLL7wg6tevLzw9PYVKpRIBAQHi6aefFpGRkXm+W7ksKGx/paSkiA8//FB069ZNVK1aVajVaunv3fjx463O5ZGX8n4OVtj+etTPKYqTTAghQERERERERETgxF1EREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMmCQTERERERERGTAIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMmCQTERERERERGTAIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMmCQTERERERERGTAIJmIiIiIiIjIgEEyERERERERkQGDZCIiIiIiIiIDBslEREREREREBgySiYiIiIiIiAwYJBMREREREREZMEgmIiIiIiIiMmCQTETkQFevXoVMJoNMJsPVq1dLuzpEpcr4W4iOji7tqliIiIiATCZDUFBQaVflsbBkyRJ06NABnp6e0n7x9ddfl3a1CiQyMhIymQw1a9a0WMb9iKh8YpBMRBaMf/RlMlm+65oGhZGRkVbX0Wg0WLx4MUJCQlC1alWo1Wp4eXmhfv366N69OyZMmIBNmzYhMzPTIm9QUJBUvunHzc0NderUwZAhQ/D3338Xqb3GE6DcH3d3d1SpUgWtW7fGqFGjsHTpUqSnpxdpW/ZKSkpCREQEIiIikJSUVKLbLk/u3LmDzz//HD179kS1atXg4uICNzc31KxZE3379sUPP/zA/i0j3nvvPchkMnh7e0On09lcb8iQIfkeewDgwIED0npFPVY8KkyPy6YfZ2dnVKxYEY0bN8aQIUMwd+5cXL9+vdjrM2fOHIwYMQIHDhxARkYGKlasiICAALi5uRX7tomICoNBMhEVq7i4OLRs2RIjR47Epk2bcOPGDQCAQqHApUuXEB0djc8//xwhISE4cOCAzXJUKhUCAgKkT3Z2Ni5fvowVK1agV69eGD16NIQQRa6vv7+/tA0nJyfcvn0bR44cwaJFixAeHo5KlSph6tSp0Gq1NuvZoEEDNGjQACqVqsj1SUpKwpQpUzBlyhQGcYUghMCMGTNQq1YtTJgwAdu2bcP169ehVCqhUCgQGxuLdevW4dVXX0XNmjWxePHi0q7yY6979+4AgOTkZBw7dszmeqaj03mNVO/cuRNAzm+zc+fODqnjo8TT01M6pnl4eCApKQlnzpzBihUrMH78eAQGBuLFF19EQkJCsdXhyy+/BAC88847ePDgAW7fvo1bt25h9OjRxbbNkuLv748GDRqgRo0apV0VInIgBslEVGx0Oh3CwsJw6tQpuLq6Ytq0abh27RoyMzNx7949PHjwAAcPHkRERATq1KmTZ1kdO3bErVu3pE9mZiYOHz6MLl26AAB++uknhwQ4hw4dkrZx7949aLVanD17FgsWLMD//vc/pKWlYfLkyejZsyeysrIs8letWhVnz57F2bNnUbVq1SLXhwpPCIGhQ4di4sSJePDgAdq1a4dVq1bh/v37SE1NRUpKCpKTk7FmzRr07t0bycnJWL9+fWlX+7HXpUsXKBQKALaD3zNnzuD27duoVKlSnuuZLmvbtu1jOXI5b9486Zh29+5dZGdn48aNG1i1ahWeeeYZ6HQ6/Pbbb2jWrFmxPCJy9+5d3L59GwAwevRoh1w8LEveeustnD17Fj///HNpV4WIHIhBMhEVmx07dkgjQYsWLcKkSZNQvXp16TZutVqNNm3aYPLkybhw4QI6dOhQ4LIVCgVatWqFdevWwc/PT9qGo8lkMjRo0ACvvvoqjh07hnHjxgHIOfF+5513HL49cpxZs2bh119/BQCMHTsW+/fvR//+/eHt7S2t4+npib59+2L9+vXYtWsXqlWrVkq1JSNPT0+0atUKwMNR4NyMgW9YWBjq16+P2NhYXLlyxWI9jUaDvXv3Ang4Qk1A5cqV0b9/f2zcuBErVqyASqXCjRs3EBoaavMumcJ68OCB9P/u7u4OLZuIqLgwSCaiYnP8+HHp/8PCwvJcVyaTQa1W270NHx8ftGvXDgDw33//2Z3fHgqFAnPmzEFoaCiAnKD8/PnzZuvkN3FXfHw83n33XTRp0gRubm5Qq9WoUqUKWrVqhXfffReHDh2S1g0KCkKtWrWkf9eqVcvs+ULTiWL0ej22b9+Od955B+3bt0e1atXg5OQEPz8/dOvWDQsWLIBGo7Hartx1vn37NsaMGYNatWrB2dkZAQEBGDJkCM6ePZtn/+j1evzxxx/o27ev9Ox5hQoV0KpVK3z44Yc4deqU1Xypqan4/PPP0aFDB/j6+kKtVqN69eoYMmQI9u/fn+c2bUlISMC0adMAAE8++STmzp2b7zP2Xbt2xTfffGN1WXR0NAYNGiS1y9/fH08++SSWLFli9bnZZs2aQSaTSRdVbNmxYwdkMhnkcjmuXbtmsfzq1asYO3YsmjRpAnd3d7i6uqJhw4YYM2aM1fUBy0mGdu7cib59+6Jy5cpQKBQIDw83W78o/X///n28//77qFOnDpydnVG5cmUMGjQIR44cyTNffowB7Z49e6z2rzFIDgoKQrdu3czSTB06dEiaRyC/IHn79u0IDQ1FhQoV4OzsjEaNGmHKlClW50owdffuXUyaNAktWrSAl5cXnJ2dUbt2bYwcOdLmMSk6Otps3oeLFy9ixIgRqF69OtRqNapVq4bRo0eXyPPCgwcPxowZMwAAp0+fxtKlS22uu3fvXrz00ksIDAyEs7MzvLy80LZtW3zxxRdIS0szW9fYRtPJrkyPYdYmwXLE/n7kyBEMHjwYlStXhlqtRu3atTFu3Djcv38/z344cOAA+vbtC39/f7i4uKBBgwaYOHGiRbtyy2virvDwcMhkMuk3FxUVhaCgIPj6+sLV1RXNmzfHvHnzoNfrbZYvhJAmPfPw8ICXlxfatWuHH374AeL/27vv+Kaq/g/gn4wuutJFGYWyl/zYIJuCRYTKruBAy1JxPIKgCCoyRBEEFFEfUBAEeVApe0rZsi2rDBkySousFjrpTM7vj+ZekiZpkzZpS/m8X6++oN9778k5Jye395tzhxAmr0FEdiKIiPKZPHmyACCs2UVcvXpVXnfJkiVGy2bNmiUvu3jxYpHq0qVLFwFAdOnSxeI6PXv2FACEu7t7kV5jyZIlcj2vXr1a6Pp//fWXvP6kSZOMlhn2R/6yTp48KXx8fOTlKpVK+Pj4CIVCIcciIiLk9fv37y/8/f3lZf7+/iIwMFD+6d+/v9nXBSA8PDyEt7e3UaxTp07iwYMHJu0x3HbTpk2iYsWKAoCoUKGCcHFxkZd5eXmJkydPmu2Tu3fvis6dOxu9nkajER4eHvLvffv2NdnuxIkTIigoyKhPPD095d8VCoX4/PPPC31P8jMce3/++afN2xt69913jeqj0WiESqWSY926dRMpKSlG23z55ZcCgKhUqZLIzc21WPbQoUMFABESEmKy7JdffjHqfxcXF+Hm5ib/7unpKf744w+T7aTxHBwcLL7++mt5fHl7ewsnJyejMVac/r969aoIDg6W13V2dhZeXl7y/9evXy8v2717t3Wdrbdt2zZ526NHj5osDwwMFADEv//+K3755RcBQLzyyism63322Wdy32VkZBgtk/ZzXbp0EbNmzRIKhUJ+fw0/k127drX4HkZFRQmNRiOv6+TkJNzd3Y365OeffzbZbvfu3fI6u3btkj8nnp6eQq1Wy8uqVKki4uPjbeo7IQreL5uTkZEh72s6depkslyr1Yp33nnHZB9j+DmoX7++uHbtmrzNgQMHRGBgoMV9WKtWrYxewx7jfcWKFcLJyUke70qlUt7+iSeeEKmpqWbbv3jxYqN1vb29hbOzswAgGjRoIObOnSu/Rn6G4yi/iIgIeb/+1ltvCQBCqVQajRlLY1cIIXJzc8XgwYONPo8+Pj5yXV944QWj1yAi+2GSTEQm7JUk79mzxyiRKMrBXmFJ8r1794Sfn58AIP7v//7P5vKFsD1JFkLIiWTnzp2N4gUlyU899ZQAIFq0aCEOHTokdDqdEEKIrKwscfHiRTF79mwxa9Ysq8szFBcXJ1566SWxYcMGkZiYKMdTU1PFkiVLRJUqVQQA8e6775psa/gaPj4+okOHDuKvv/4SQgiRk5MjoqKiROXKlS0eQOfk5IgOHTrIB7YzZ84Ud+7ckZffuHFDLFy4UEycONFou3///VfuxwEDBojo6GiRnZ0thBDi9u3bYtKkSXLCsHbtWottN6dHjx4CgAgICLBpu/zmz58v981rr70mbt68KYQQIi0tTXz11Vdy/QYPHmzSNimB2Lp1q9myHzx4ICekP/30k9Gy7du3C6VSKdRqtRg/fry4evWq0Ol0QqfTifPnz4vnnntO/uIiNjbWaFtpPLu6ugqVSiWGDh0qrl+/LoTIO+j+559/5DoWtf9zc3NFq1at5DHz+++/i5ycHCGEEGfPnhWdOnUySgRsTZLT0tLkZGfmzJlGy86dOycAiHr16gkh8sY+AFG9enWTckJDQy3uP6T9nEajEUqlUkycOFHcvXtXCCFEcnKy+OSTT+T6L1682GT7mJgYOYl79dVXxblz5+RkOjY2Vrz55psCgFCr1fLnSWKYJPv4+Ig+ffqIv//+WwiRtz/47bff5LHx8ssv29R3QtieJAshxKBBg+TEPv8XCh9//LEAICpWrCi+++47eR+TnZ0tdu/eLZo3by7v27RarcW6WNqH2WO8S1/qjRw5Uh7v6enp4ttvv5XHUv4vNYUQ4tixY/I4DwkJkd+H7OxssXLlSqHRaOSxXNQk2cfHRzg7O4u5c+eK5ORkIYQQCQkJYuTIkXLf7Ny502T7GTNmyMvHjh0rEhIShBB54/Pzzz+Xk2YmyUT2xySZiEwYJsmGM5fmfgxnCcwdjHXv3l1erlKpRLt27cSYMWPE8uXLrZpdtpQk5+bmiujoaNGpUye5/Llz5xapvUVJkqV2Va1a1She0AGhdEB98OBBq+tmbZJcGGn2293d3eQA2PA1GjRoYHa2ecOGDfI6cXFxRssWLVokz3Js3rzZ6joNHz5cABAvvviixXWkGZymTZtaXa4QQp4d7d69u03bGXrw4IHw9fWVZ2zM+eabb+R+iY6ONlomJeqWtv3f//4nAAg3NzejmWitVivq1q0rAIiFCxdarF+fPn0EADF69GijuOF4HjBggMXti9P/v/32m/waO3bsMNkuPT1d1K5du8hJshBCtG/fXgAQPXv2NIp///33cmIqqVWrlgAgLl++LMeysrJEhQoVBAAxZcoUk/IN93OTJ082W4cBAwYIACI0NNRkWbdu3QQAky9/DEmzr/nPojBMkrt27WqSWArxcGy5ubnJX0BYqyhJsjTrDkBcunTJqCyVSiXc3NwsnkmSkpIif+byf6FS2D7MnuPdUqI4duxYAUDUqVPHZJl0JlK9evXM7vsMz2ooapJc0PvQsmVLAUCMHDnSKJ6WliafmTFixAiz2xqOYSbJRPbFJJmITBj+4bXlx9xBQFpamnjzzTflb/Lz/wQHB4spU6bI367nJyXJTk5ORsm54SmJAMTAgQPlWTBbFSVJlk6Bc3NzM4oXdEAozcauXr3a6rrZK0kW4uHs96FDhyy+xo8//mh225ycHPn0wy1bthgtk5KZsLAwq+uSkZEhXF1dBQBx6tQpi+slJCTIdbt165bV5UtfSOSf4bWF4enCFy5cMLtObm6u/L6OHTvWaJl0GnD+JFgiHZznT6KlBMrf399s8iSJjIyUv9gwZDie889gSorb//379xcARIcOHSxuu3DhwmIlyR999JEA8k6zNUwSpVnFFStWyLFhw4YJAGLRokVy7M8//5Rff+/evSblS/s5FxcXi6fh/vzzzwLIm0E1JH1m1Gq10Zkb+UVHR8tfThmesm2YJEdFRZndNjY2Vl7n3LlzFl/DnKIkyf/973/lbY4cOSLHpX4yvLzDnLffflsAEKNGjbJYF3P7MHuOd8Pk3tDevXvlddLT0+X4/fv35VOXLe37hBCiXbt2xUqSq1WrJp85lN/UqVMFANGmTRuj+Jo1a+Q6X7lyxey2KSkp8hdBTJKJ7EsNIqICiEKePXzt2jWjm0vl5+7uju+++w6TJ0/G+vXr8eeffyI6OhoXL16EVqtFbGwspkyZgp9//hlRUVEWHwWVk5MjP0bEkEKhwLfffos333zTtoaVgmeffRY//vgjIiIicODAAfTp0wetW7dGhQoV7FJ+dnY2fvrpJ6xZswZnzpxBYmIisrOzTdaLj4+3WIZ0E7T81Go1AgICcOPGDdy7d0+O5+bmyjcb6927t9V1PXbsmHxDpKefftqqbWJjYxEYGGj1axRXdHQ0AKBatWqoV6+e2XVUKhW6deuGFStWyOtL+vfvD09PT6SmpmL16tVGN9a5ffs2tm/fDgB45ZVXjLaT7sacnJyMKlWqWKyf9N7GxsaaXe7m5oYWLVqYXVbc/pfa2q1bN4vrF7TMGiEhIfjss8+QmpqKY8eOyWNz7969ACDfsEv6/5IlS7Bnzx6MGDECwMMbebm5uaFt27YWX0e6SZQ5Uv8bjnng4Xuk0+nQqFEji2VLNx1LT09HYmIiKlasaLKOpc+c4Xuf//VLktTW7du3y4/cMke6wZWl8VhY+cUd776+vqhTp47ZZYbl3r9/X97nHj9+XL5pVmFjuag3EQSA1q1bW7xxoKUxdvz4cQBA9erVLf6N9fT0RMuWLfHnn38WuW5EZB6TZCIqERUrVsSrr76KV199FUDeAdWuXbvw5ZdfYv/+/bh69Sqef/55o7s7G+rSpYt80JuTk4PY2Fj88MMPmD17NsaPH48nnnjC6KDZ0aQDGunxU9aYNWsW/vnnH+zevRtz587F3LlzoVKp0KxZM4SFheG1114r8rOV79y5g9DQUJw+fVqOubq6wt/fX37m7N27d6HT6eS7/Zrj6elpcZlanfcnw/Au2YmJifLvwcHBVtf333//lf9v7ssPcwwfJVMYPz8/xMfHFyu5uHPnDgAU+p5Ij42S1pdUqFABAwcOxNKlS7F8+XKjJHnlypXQarWoVKkSunfvbrSd1DeWvhjKLyMjw2zcz88PSqX5h1gUt/+t6ZviPk6rQ4cOcHZ2RnZ2Nvbs2YMnn3wS586dw507d1CnTh2j1zZ3h2vp/1I5llgz5vM/FknqP51OV+zxa+n1pdcGYPHO9PZk+Fkx3K9JbU1PTy9w3yGx5XNqWH5xx7s176P0OhLDz6wjx7Kt+1Ugb38NoMAvDoDC909EVDR8BBQRlQoPDw/06dMHe/fulR/NEh0dbfTYKEucnJxQp04dzJo1C5MnT0Z6ejoGDRpkkqQ4UkxMDABYnPk2R6PRYNeuXfjzzz8xfvx4dOjQAWq1GseOHcO0adNQt25drFy5skj1effdd3H69Gn4+fnhp59+ws2bN5GRkYG7d+/i1q1buHXrlnywVdjZAbYo7LFKlhg+1icjIwMi7/KfAn/MPWLFkieeeAIArBpPjiTNEu/ZswdxcXFyfPny5QCAF198Uf4SQyL1zZNPPmlVv1h6P/OXa+41AMf0vz24ubnJs6zS85KlxDf/F2I1atRA9erVER8fj3/++QfZ2dk4ePAgAMc8H1nqv8DAQKvfI3OPPCpLTp06BSDv+fWGiZfU1g8++MCqdpp7FFdB7DXey6ui7mOJqHiYJBNRqVIqlRg5cqT8+4ULF2za/sMPP0Tt2rVx584dTJo0yd7VMys6Olqe8ShK4tCxY0fMnDkT+/fvR1JSEtavX4//+7//Q0ZGBoYPH271zJQkJycHa9asAQB8++23GDZsmMlpkVqtFgkJCTbXtTC+vr5wcnICYNtplob1s/X0TGs89dRTAPJmY/bv31+kMqRTYws6Pd1wublTaUNCQlCtWjXodDqsWLECQN6zaKVTKfOfag087BtH9Ev+1yjq60htLeg5vvZ4xq+U4B44cAC5ublGz0fOz3A2+ciRI/KMoyOSZKn/EhISrJpdLesyMzOxa9cuAEDbtm3h6uoqL3P0eCyJ8W6J4WfW0WPZVgEBAQCMz/owpzTqRvQ4YJJMRKXO8HpAFxcXm7Z1cnLCxx9/DABYvHgxLl68aNe6mTN16lQAeTN1Q4YMKVZZrq6u6NOnj5zkZmZmGiV1hqfLWppBuXv3rnx9afPmzc2us3//fnkde1Kr1WjTpg0AYOPGjVZv17p1a/kUWFu2s9awYcPk6w6nTJli9eyTdH0iALRq1QpAXhJsaVxptVp5lrN169YmyxUKhTxGpNlj6d8mTZqgadOmJtt06NABAHDr1i2T65ztpbj9L/WN1HZzpKSrOKQENy0tDX/99ZfZ65ElUmz37t1yvTw8PMy+L8UlvUdarRZbt261e/kl7dtvv5W/RDO8LAB42NYdO3Y4ZB9SEuPdkhYtWsj7WEePZVtJ9xOIjY3FtWvXzK6TlpaGY8eOlWCtiB4fTJKJyGHOnDlj1bfcy5Ytk/9vKckryJAhQxAcHAytVisnsI6g1Woxbtw4bNq0CQDw6quvWrxRTH65ublGCVh+bm5u8v8NE2MvLy/5/0lJSWa39fLykk/Jk06ZzP/aH330kVX1LArpRklbtmzBli1brNrG3d0dL774IgBg5syZuH79eoHr23ptsb+/v/zlyc6dOzFu3LhCE+UDBw5g9OjR8u/du3eXr82cMmWK2W0WLlwoz/S88MILZteRZovPnTuH6OhoeUbZ3CwykJcYSuPq3XffNXvzNUNFue66uP0/ePBgAHlfvpg7vTYjIwNffvmlzfXKr127dvKs5vfff487d+6gVq1aqFatmsm6UpK8d+9euU4dO3Y0uh7VXurWrSvPZn/00UdITk4ucP3SvPFWYX7//Xd8+OGHAIDGjRubfPE3fPhwqNVqJCQkYPLkyQWWlZ2dLd/Ay1olMd4t0Wg08o3rZs+ebfZLgB07dsin7pekp59+Wt7/f/7552bX+eqrr2y+BpyIrMMkmYgcZs+ePahVqxYGDx6MVatW4ebNm/Iyaca0T58+WL16NQAgPDzcpps/SdRqNd5//30AwK+//opz587ZpwHIm729dOkSfvjhB7Ro0QJz584FkHc677x586wuJz4+HnXr1sX06dNx4sQJoxsBxcTEyAem7u7uRrNkGo1Gvj5wyZIlJjcQAvJmy6TZmLFjx2LXrl1yQn7mzBn06tUL0dHRcHd3t7H11nn55ZfRsWNHCCEwcOBAfPnll0andv/777/46quv8MEHHxht9/nnn6NKlSpISEhAu3btsHz5cqSmpsrL7969i9WrV6N///4WE9CCTJgwQU7mvvrqK3To0AFr165FSkqKvE5qaio2bdqEAQMGoFOnTkbXDbu5ucnJ8cqVKzFq1Cj5VPgHDx7gm2++wZgxYwDkJY0tW7Y0W48GDRrIM69vvPEG4uLioFKp5CQ1P7VajQULFkCtVmP//v3o3Lkzdu7caXRjnytXrmDBggVo3bo1vv/+e5v7Bihe/w8cOFCe6Ro4cCBWr14tX1v6999/o2fPnvKNh4rDxcUF7dq1AwD5en1LN+irU6cOqlSpghs3bmDfvn0AHHOqtWT+/Pnw8PDAxYsX0bZtW6xfv94oybpx4waWL1+Op556ymTsl7Zbt25hzZo1CAsLw+DBg5GTk4OqVati06ZNJl8q1K5dW76UZdasWXjllVdw5swZeXlubi5OnjyJadOmoU6dOjbfB6Ckxrsln376KVQqFc6fP4+wsDD5kp/c3Fz8/vvvGDRoEDQajV1f0xru7u7yuPnxxx8xfvx4+QuC1NRUzJw5E1OmTIGPj0+J143osVD8p0gRUXlj+JzkwhT0PM4FCxaYPBfZ1dVV+Pj4mMSffvpps8+SlZ6TbO4ZlIYyMjJEpUqVBAARHh5uS3ONnrPp7+8vP4vZ19dXqFQqo3p6enqK6dOnGz231VJ/GD4T1DAOQKhUKuHr6ys/exiAcHZ2FqtWrTIp89NPP5XXcXFxEdWqVRPBwcFGzwCOjo4W7u7uRut5enrKz3JdtmyZCA4ONvs+WfssZkvbCyHE3bt3RadOneRyFAqF0Gg0wsPDQ4717dvXZLtz586JevXqyesolUrh6+tr1BYAIjQ01GK9CqLT6cTUqVPl5yYbvo9S/0g/vr6+YtmyZSZlvPvuu0bt8vHxMXpOd9euXc2OXUPffPON0Wv16NGj0LqvXbvWqI5OTk7Cz89PuLi4GJU1ffp0o+2k8Wzuma75Faf/L1++LKpVq2Y05ry9veWxbPic6aI8J1kybdo0o7r8/PPPFtd94YUXjNY9evSoxXULer6txPB5xubs379f3u9In2s/Pz+T8TZy5EibypUUtf8MP9NeXl7yPi0gIMBonyPV+eWXXy7wec86nU5MmjRJKBQKeTs3Nzfh5+dnso/cv3+/xboUtH9x5HgvrA4LFy40apu3t7f8ug0aNBBz5861+BrWPCe5oGcYF1T/nJwcER4ebvT59PHxkfv85ZdfFq+88ooAIF5//XWLr0FEtuNMMhE5zOuvv45Tp05h5syZ6Nu3L+rUqQOVSoXk5GR4enqiUaNGeOWVV7Blyxb88ccfBT4mozCurq4YO3YsAGD16tVmTzu2RkJCAm7fvo3bt28jMzMTAQEBaNmyJUaMGIGff/4ZN2/exEcffWTzKZxVq1bFhg0b8O6776Jt27aoXLky0tLSoFar0ahRI7z11ls4c+YMwsPDTbb98MMPMW/ePLRq1QpOTk6Ij49HbGwsbt26Ja/TsmVLHD16FIMGDYK/vz90Oh08PT0xaNAgHDx4EC+//HKR+sNa/v7+2LNnD3755Rf07NkTAQEBSE9PR4UKFdCyZUtMmDDB7CmDDRs2RExMDBYuXIinn34a/v7+SElJgRACderUwXPPPYcffvgBv//+e5HqpVAo8Mknn+DKlSv4/PPP0a1bN1SpUgXZ2dnIzc1FcHAw+vXrh0WLFuHatWtm+2nu3LnYtWsXBg4ciMDAQKSlpcHT0xNdu3bFTz/9hKioqELH7gsvvCDf4AywfKq1oX79+uGff/7B5MmT0aZNG3h4eCApKQkuLi5o2rQpRo4cibVr18pnURRFcfq/Vq1aOHnyJMaOHYuaNWtCCAFXV1eEh4fj4MGD6NOnT5HrZSj/bHBBj3ozXObl5WXxOdH20qFDB1y8eBGzZ89G586dodFokJSUBJVKhYYNG2LIkCFYsWIFvv76a4fWoyApKSnyPi05ORleXl5o2LAhBg8ejLlz5+L69etYtmwZfH19LZahUCgwbdo0xMTE4M0330TDhg3lfbmPjw/at2+P999/HwcPHpTParFVSYx3S1577TUcOHAAvXv3hq+vL7KyshAcHIyJEyfi6NGjpTZbq1ar8fvvv2PRokVo06YN3NzckJubi1atWmHRokVYtmyZfBlOacx2E5VnCiFsv5d+SEiIfPMMc7Zu3YpnnnnGJL506VJ8//33OHfuHJydndG2bVt8/PHHaN++vcWyDhw4gM8++wyHDx9GdnY2GjVqhLffftuqAwwiIiIiovJICCE/+mzZsmUO/zKU6HFSrLtZDBw40OiutBJzDzYfM2YM5s2bBzc3Nzz99NPIzMxEVFQUtm/fjsjISPTr189km9WrV2Pw4MHQ6XTo3Lkz/P39sXPnTkRERCAmJgazZ88uTvWJiIiIiB5Jy5cvR3x8PNRqNUJDQ0u7OkTlSrFmkq9evYoaNWoUuv6OHTvku4QeOnQIdevWBQAcOnQIISEhqFChAq5evWp0qsi9e/dQs2ZNpKSkYPXq1RgwYAAA4Pbt2+jYsSP++ecf7N69u0jPKCUiIiIiKuteeOEFDBw4ECEhIfD39weQdyy8ZMkSTJ48GdnZ2Rg+fDgWL15cyjUlKl9K5Jpk6W6wH3/8sZwgA3mPdhg1ahSSkpJMPtyLFi1CSkoK+vbtKyfIABAYGIhZs2YBAObMmVMCtSciIiIiKnlbt27Fc889h4CAALi7u0Oj0aBSpUqYOHEisrOz0alTJ3z11VelXU2icsfhSXJGRob8EHZzN6SRYhs3bjSKb9682eI2YWFhcHV1ddiD7YmIiIiISts333yD559/HvXr14eLiwsePHiAgIAAdO/eHYsXL8bOnTvl5ykTkf0U65rkxYsXIzExEUqlEvXq1UO/fv1QvXp1o3UuXLiArKwsBAQEICgoyKQM6c6TMTExRnHpzrTm7kzp7OyMxo0bIzo6GhcvXkSTJk2K0wwiIiIiojLnlVde4c1qiUpBsZLk6dOnG/3+3nvvYdKkSfJD5wHg+vXrAGA2QQYgnzpy//59pKamwtPTEykpKUhOTi5wu6CgIERHRyM2NpZJMhEREREREdlFkZLkzp07Y+TIkWjfvj0qV66MuLg4REZGYvr06fjkk0/g5eWF0aNHAwDS0tIAABUqVLBYnru7O5KSkuQkWdqmoO3c3d0BAKmpqWaXZ2VlISsrS/5dp9Ph3r178PPzg0KhsK3BRERERERE9MgSQiA1NRVVqlSBUlnwVcdFSpKnTZtm9Hu9evXw4YcfolWrVujRowemTJmC1157DW5ubkUp3i5mzJiBqVOnltrrExERERERUdkSFxdn8WxlSbFOt87v6aefRqtWrRAdHY0jR44gJCREfo7ygwcPLG6Xnp4OAPD09AQAo2cvP3jwwOwNCfJvk9/EiRMxduxY+ffk5GRUr14dV69elbdRKpVQqVTQarXQ6XTyulI8NzcXhk/IUqlUUCqVFuM5OTlGdVCr87o3NzfXqriTkxN0Oh20Wq0cUygUUKvVFuOW6s42sU1sE9vENrFNbBPbxDaxTWwT28Q25UlNTUXNmjUt5o9Gr1foGjaqW7cuoqOjcfPmTQCQb+QVHx9vdv309HQkJSXBx8dHrrCXlxe8vb2RnJyM+Ph4NGrUyGQ7qbzg4GCz5bq4uMDFxcUk7uvry7sAEhERERERPUacnJwAwKpLb+3+CKj79+8DeHjNsHTL+rt37+LGjRsm6x8/fhwATG6+1bRpU6PlhnJycnDmzBm4urqiXr16dq0/ERERERERPb7smiTfvXsXf/75J4CHj25yc3NDt27dAACrVq0y2SYyMhIA0Lt3b6N4WFiY0XJDmzZtQmZmJkJDQ+Hq6mq/BhAREREREdFjzeYk+eDBg1i3bp3R+d0AcO3aNfTv3x/p6eno06eP0cXQ0rXB06dPx6VLl+T4oUOHsHDhQmg0GowYMcKovJEjR8LLywvr16/HmjVr5PidO3cwfvx4AMC4ceNsrT4RERERERGRRTZfk3zx4kUMGzYMlSpVQosWLaDRaBAbG4tjx44hMzMTTzzxBH788UejbUJDQzF69GjMmzcPzZo1Q/fu3ZGdnY2oqCgIIbBkyRJoNBqjbXx9ffHTTz9h0KBBCA8PR0hICPz8/LBjxw4kJSVh7NixCAkJKU7biYiIiIiIiIwohOFtxqzw999/Y/78+Thy5Aji4uJw//59uLu7o2HDhnjuuefwxhtvWHz009KlS/Htt9/i77//hrOzM9q2bYtJkyahffv2Fl/vwIEDmD59Og4fPozs7Gw0atQIb7/9NiIiImxqaEpKinwzMN64i4iIiIiI6PFhSz5oc5L8qGKSTERERERE9HiyJR+0+92tiYiIiIiIiB5Vdn9OMhERERERPV5ycnJMbuxL5EgqlUp+9rG9MUkmIiIiIqIiSUlJQUJCArKyskq7KvQYcnFxgb+/v90vp2WSTERERERENktJScGNGzfg4eEBf39/ODk5QaFQlHa16DEghEBOTg6Sk5Nx48YNALBroswkmYiIiIiIbJaQkAAPDw8EBQUxOaYS5+bmBk9PT8THxyMhIcGuSTJv3EVERERERDbJyclBVlYWvL29mSBTqVEoFPD29kZWVhZycnLsVi6TZCIiIiIisol0ky5H3TiJyFrSGLTnjeOYJBMRERERUZFwFplKmyPGIJNkIiIiIiIiIj0myURERERERER6TJKJiIiIiIiKSKFQ2PRTo0aN0q4yFYKPgCIiIiIiIiqiiIgIk9j+/ftx+fJlNG3aFM2aNTNa5u/vX0I1o6JikkxERERERHZXY8Lm0q6CVa59EVas7ZcuXWoSGzp0KC5fvox+/fphypQpxSqfSh5PtyYiIiIiIiLSY5JMRERERERUApYuXQqFQoEpU6bg4sWLeP755xEYGAilUol169YBAGrUqGHxsUZ79uyBQqHA0KFDTZYJIbBy5Up069YNPj4+cHV1RcOGDTFlyhQ8ePDAga0qf3i6NRERERERUQm6cOECWrduDT8/P3Tt2hX379+Hk5NTkcvT6XQYMmQIVq5cCQ8PD7Rq1Qo+Pj6Ijo7G1KlTsXXrVuzZswdubm52bEX5xSSZiIiIiIioBP366694++238fXXX0OlUhW7vDlz5mDlypUICQnBypUrUalSJQBAdnY23nzzTSxevBhTp07FF198UezXehzwdGsiIiIiIqISFBAQgJkzZ9olQc7NzcWsWbPg7u6OX3/9VU6QAcDZ2Rnz589HpUqV8MMPP0Cn0xX79R4HTJKJiIiIiIhKUGhoKCpUqGCXso4fP46EhAS0b98egYGBJsvd3NzQsmVL3L9/H5cuXbLLa5Z3PN2aiIiIiIioBFWvXt1uZV27dg0AEBUVZfGGX5KEhATUr1/fbq9dXjFJJiIiIiIiKkGurq5F2s7c6dJSrE6dOujQoUOB2/v5+RXpdR83TJKJiIiIiIjKCGdnZwBAWloaPDw8jJbFxcWZrB8UFAQAaNCgAZYuXerw+j0OeE0yERERERFRGVG5cmUAwMWLF02WRUVFmcRat24Nb29v7N27F/fu3XN4/R4HTJKJiIiIiIjKiC5dugAAZsyYAa1WK8dXrlyJlStXmqzv4uKC8ePHIzU1FQMGDMCVK1dM1rlx4waWL1/uuEqXMzzdmoiIiIiIqIx46623sGDBAkRGRqJRo0Zo0qQJLl26hDNnzmD06NH46quvTLaZMGECzp8/j+XLl6Nhw4Zo3rw5atasiezsbFy4cAHnzp1DkyZN8PLLL5dCix49nEkmIiIiIiIqIwIDA7Fv3z48++yzuHnzJrZu3Qpvb29ERUWhT58+ZrdRKpVYtmwZ1q9fj+7du+Pq1atYvXo19u/fD1dXV7z//vv46aefSrgljy6FEEKUdiVKQkpKCry9vZGcnAwvL6/Srg4RERER0SMrMzMTV69eRc2aNYt8p2Yie7B2LNqSD3ImmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj0myURERERERER6TJKJiIiIiIiI9JgkExEREREREekxSSYiIiIiIiLSY5JMREREREREpMckmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIr1iJ8mJiYmoWLEiFAoF6tSpU+C6S5cuRZs2beDh4QFfX1/06tULBw8eLHCbAwcOoFevXvD19YWHhwfatGmDZcuWFbfaRERERERERCaKnSSPGzcOCQkJha43ZswYDBs2DGfOnEFoaCjatGmDqKgodO7cGevWrTO7zerVq9GlSxds27YNTZo0wTPPPINLly4hIiIC7733XnGrTkRERERERGSkWEnyzp078fPPP+PVV18tcL0dO3Zg3rx58PPzw6lTp7Bu3Tps27YN+/btg0qlwrBhw5CUlGS0zb179zB8+HBotVpERkZiz549iIyMxPnz51GnTh3MmTMHe/bsKU71iYiIiIiIiIwUOUnOyMjA66+/jkaNGhU6qzt37lwAwMcff4y6devK8Xbt2mHUqFFISkrC4sWLjbZZtGgRUlJS0LdvXwwYMECOBwYGYtasWQCAOXPmFLX6REREREREdqNQKIx+lEolNBoNOnXqhEWLFkEIUar1W7p0KRQKBaZMmVKq9XgUqIu64dSpU3HlyhXs3bsXTk5OFtfLyMjArl27AADh4eEmy8PDw/HNN99g48aNGDdunBzfvHmzxW3CwsLg6uqKHTt2IDMzE66urkVtBhEREREROcAXJwq/JLMsmNDc367lRUREAAC0Wi0uX76MAwcOYP/+/di5cydWrlxp19cixyjSTHJMTAzmzJmDYcOGoVOnTgWue+HCBWRlZSEgIABBQUEmy1u0aCGXaejUqVNGyw05OzujcePGyMzMxMWLF4vSBCIiIiIiIrtbunQpli5diuXLl+PgwYP4448/oFar8euvv2LTpk2lXT2ygs1Jsk6nw8iRI6HRaOTTngty/fp1ADCbIAOAu7s7NBoN7t+/j9TUVABASkoKkpOTC9xOisfGxtraBCIiIiIiohLRvXt3vPzyywBg8YbFVLbYfLr1/Pnz8ddff2HJkiXw8/MrdP20tDQAQIUKFSyu4+7ujqSkJKSmpsLT01PepqDt3N3dAUBOrPPLyspCVlaW/HtKSgoAICcnBzk5OQAApVIJlUoFrVYLnU4nryvFc3Nzja4dUKlUUCqVFuNSuRK1Oq97c3NzrYo7OTlBp9NBq9XKMYVCAbVabTFuqe5sE9vENrFNbBPbxDaxTWwT2+SoNuXk5EAIIb++EMKobIVCgUdF/roDeX1prk0KhcJiXCL1ieH6zZo1A5A3gSiEkJcFBwfj/PnzmDlzJlauXImrV6+iZ8+eWLNmDQDgwYMH+Oabb7Bq1SpcunQJANC4cWO8/vrrGDp0qFx/QwcPHsSUKVNw5MgRqFQqtGnTBlOnTjVqr7k62tJWe8UNx68UN9cmS3HD9yknJwcqlcri5yn/eC6ITUny9evX8fHHH6NLly7ym1JWzZgxw2gwSLZv3y4n3tWrV0fz5s0RExMjz3gDQP369dGgQQMcPXoUd+/elePNmjVDcHAw9u3bZ5Sct2vXDhUrVsT27duNOr9r165wc3PDli1bjOrQq1cvZGRkYPfu3XJMrVYjLCwMCQkJOHTokBz39PREt27dEBcXh5MnT8rxgIAAtG/fHpcuXcKFCxfkONvENrFNbBPbxDaxTWwT28Q2ObpNarUalSpVQlpaGipUqIDc3Fykp6fL6yqVxX7SbInR6XQmE28ajcZsm7y8vJCdnY2MjAw5rlar4eHhIf8uTc45OzujQoUKyMjIkN8PlUqFzMxMuLm5AchL3Hr37o1Dhw6hU6dOaNKkCTw9PZGSkoK7d++if//+OHv2LCpVqoT27dtDCIGjR49i+PDhOHbsGL755hv59QBg27ZtePnll5Gbm4uWLVsiODgYZ8+eRUhIiJy/ZWVlydsU1qbMzEyjiUfDNmVnZ8txFxcXuLm5IT093Wgsubm5wcXFBampqUYJsbu7O5ycnIzqDuR9FhQKhUncy8sLQgiL71NGRgb27duH3Nxci5+ngiZt81MIG26z1rt3b2zfvh2nTp1CgwYN5Pi1a9dQs2ZN1K5dG//884/RNhs2bEDfvn3RvHlzHD9+3Gy5Pj4+SEpKQkpKijwovL29AQDJycnw8vIy2aZ///5Yt24dNmzYgN69e5ssNzeTXK1aNSQkJMjllYVv4QyVl28W2Sa2iW1im9gmtoltYpvYpvLdpszMTMTFxaFGjRqoUKGC2ZnCmScT8Sj4oJmfXWaSpdlO6T00nC3t2LEjDh06hA8//BDTp083Wr9OnTrYtWsXgoKCjMp/9tlnsXXrVrzzzjuYNWuWfLPk27dvo0+fPoiOjsaWLVvQo0cPAHln2NatWxd3797F4sWLjWaaP/zwQ/lS2U8++QSTJ0+2qk2PwkxyRkYGrl69imrVqsHV1dXi5yk1NRV+fn4W80uj17MlSVYoFNBoNGjatKlRPDMzE0eOHIGrqyuefPJJAMCvv/6KSpUq4eTJk2jevDkCAgJw584dkzLT09Ph4eEBHx8f3Lt3T45rNBokJyfj7NmzaNSokcl2rVu3RnR0NE6dOoUmTZoUWncp8bamU4iIiIiIyLLMzExcvXoVNWvWtPikmcft7tb5EzmtVosrV67g888/x9KlS+Hi4oKzZ8+idu3aRuuvWrXK5Ik+Ug7VunVrHD582GRm/sSJE2jRogX69OmD9evXAwCWLFmC4cOHo3Pnzti7d6/R+jk5OahVqxbi4+MxefLkcvUYKGvGImBbPmjzNclJSUkmnW5YQWlZZmYmgLzTPlxcXHD37l3cuHEDVatWNdpGml3On+g2bdoU+/btw/Hjx02S5JycHJw5cwaurq6oV6+erU0gIiIiIiJyCHPXY3t6euLnn3+WE2TDdc2dFbt9+3YAQL9+/cyeut68eXN4eHjg6NGjcuzPP/8EADz//PMm6zs5OSE8PBxff/21TW15XNl0sYA0TZ7/5+rVqwCA2rVry7EaNWoAyDsPvVu3bgDyviXJLzIyEgBMBkdYWJjRckObNm1CZmYmQkND+YxkIiIiIiIqMyIiIhAREYFhw4Zh9OjRWLRoEWJjY9G/f3+TdStWrAgXFxeT+LVr1wAAH330kdHp3IY/aWlpSEh4OFv/77//AgCCg4PN1kvKz6hwNs8kF8XYsWOxdetWTJ8+HWFhYahbty4A4NChQ1i4cCE0Gg1GjBhhtM3IkSPx2WefYf369VizZg0GDBgAALhz5w7Gjx8PABg3blxJVJ+IiIiIiMgqS5cutXpdSxN+0rW6HTt2NJl9JscrkSQ5NDQUo0ePxrx589CsWTN0794d2dnZiIqKghACS5YsgUajMdrG19cXP/30EwYNGoTw8HCEhITAz88PO3bsQFJSEsaOHYuQkJCSqD4REREREVGJCQoKApB3urW1E4OVK1cGAMTGxppdbilOpkrs3uxff/01lixZgoYNGyIqKgqHDh1CaGgo9u3bh379+pndZuDAgdi3bx969OiBEydOYMuWLahTpw6WLl2KOXPmlFTViYiIiIiISkz37t0BAGvXrrV6m06dOgEAfv/9d5Nlubm5WL16tX0q9xiwS5Jco0YNCCFMHv+U39ChQxEdHY309HTcv38fW7duRfv27QvcpkOHDti6dSvu37+P9PR0/PXXX4iIiLBHtYmIiIiIiMqcJ598Et27d8eBAwfw1ltvmTw3GABOnTqFbdu2yb8/99xz8PPzw549e/Dzzz/LcSEEJk+ebPRMbSrYo/OUbyIiIiIiosfEL7/8gubNm+P7779HcHAwunbtipdeegnPPvssqlevjmbNmhklyZ6enli8eDFUKhWGDh2Ktm3b4sUXX0Tjxo3x5Zdf4tVXXy3F1jxamCQTERERERGVMRUrVsTBgwfxzTffoFGjRjhx4gQiIyMRExODWrVq4csvv8R7771ntE3fvn2xe/dudO3aFWfOnMHmzZtRuXJl7N27t9AzeOkhhZCedl3O2fLwaCIiIiIisiwzMxNXr15FzZo1+UhWKlXWjkVb8kHOJBMRERERERHpMUkmIiIiIiIi0mOSTERERERERKTHJJmIiIiIiIhIj0kyERERERERkR6TZCIiIiIiIiI9JslEREREREREekySiYiIiIiIiPSYJBMRERERERHpMUkmIiIiIiIi0lOXdgWIiIiobJk6dardypo8ebLdyiIiIioJnEkmIiIiIiIi0mOSTERERERERKTHJJmIiIiIiKiYFAqFyY+TkxOqVKmCgQMH4uDBg6VdRavs2bMHCoUCQ4cONYovXboUCoUCU6ZMKZV6lSRek0xERERERHZnz/sbOJK9750QEREh/z81NRWnTp3CmjVrsHbtWvzyyy948cUX7fp6ZH9MkomIiIiIiOxk6dKlRr/rdDp8+OGHmDlzJt555x0899xzcHJyKp3KFUP//v3Rtm1b+Pv7l3ZVHI6nWxMRERERETmIUqnEtGnToFarkZiYiLNnz5Z2lYrE29sbDRo0YJJMRERERERExePs7Axvb28AQG5urtGykydPYvz48WjZsiUCAgLg4uKCWrVq4c0338S///5rtrwzZ85gyJAhqFWrFlxdXREQEIBmzZphzJgxuHnzpsn6f//9N4YOHYpq1arBxcUFgYGBeP75521K2C1dkzx06FAoFArs2bMH+/btQ7du3eDp6QkvLy+EhYXh3LlzFsvctm0bwsLCjNo9duxYJCYmWl0vR2CSTERERERE5EBXr15FYmIinJycUKdOHaNlX3zxBb766isAQMeOHdGrVy8IIfDf//4XrVq1MkmUjx07htatW2PFihXw9PRE37590bZtW+Tk5GDevHm4cOGC0frr1q1D8+bN8fPPP8Pf3x99+vRBzZo18fvvv6NNmzbYt2+fXdq4ceNGdOvWDQ8ePECvXr1QuXJlbNmyBZ07d8atW7dM1p8wYQJ69uyJHTt2oH79+ujTpw/UajW++uorPPnkk7h9+7Zd6lUUTJKJiIiIiIgcIC0tDfv378egQYMAAG+88QY0Go3ROq+//jri4+Nx7NgxrF27FmvXrsXly5cxdepU3Lx5Ex9//LHR+t988w0yMzMxe/ZsnDp1Cr/99hs2btyIs2fP4u+//0b9+vXlda9du4YhQ4bAyckJUVFROHHiBFatWoXDhw9jy5YtyMnJwZAhQ5CdnV3stn799deIjIzE4cOH8dtvv+HcuXMYOHAgEhMT8f333xutu2rVKsycORONGzfG2bNnsX//fqxatQoXLlzAJ598gsuXL2P06NHFrlNRMUkmIiIiIiKyE8NHQHl6eqJTp064cOEC5s+fj6+//tpk/a5duyIwMNAoplQq8cknn6Bq1arYsGGD0bK7d+8CAEJDQ03KatCgASpXriz//vXXXyM9PR0zZswwWf+ZZ57BG2+8gbi4OGzevLmozZW98MIL6Nevn/y7SqXCxIkTAcBktvqzzz4DAKxcudJoZl06nbtZs2aIjIxEQkJCsetVFLy7NRERERERkZ0YPgIqKysLsbGxOHLkCKZNm4batWujZ8+eJtskJiZiw4YNOHPmDJKSkqDVagEAOTk5SExMxL179+Dr6wsAaNmyJbZu3Yq33noL06dPR8eOHaFWm0/rtm/fDgAYMGCA2eWdOnXCN998g6NHj6J///7FavfTTz9tEqtXrx4AGF0nfefOHZw6dQp169ZF48aNTbZRKBTo0KEDTp48iWPHjqFHjx7FqldRMEkmIiIiIiKyk/yPgAKAEydOoEuXLujTpw/OnDljdEr0ypUr8dprryEtLc1imampqXKS/P7772P//v3Ys2cPunbtCg8PD7Rr1w5hYWEYOnSofIMwIO90awCoWrVqgXW2x4xtUFCQSczT0xNA3pcF+et06dIlKBQKh9erKJgkExEREREROVDz5s3x+uuvY/bs2fjvf/8rn3YdGxuLoUOHAsg7NTosLAxVq1aFm5sbAKB9+/Y4dOgQhBByWV5eXti1axcOHDiAjRs3Ys+ePdi1axeioqIwY8YM/Pnnn6hbty6AvGc0A8az2+Y8+eSTxW6jUmndlbxSnSpVqlToLHFwcHCx61UUTJKJiIiIiIgcrGbNmgDyZlAlW7ZsQXZ2Nt577z2zN6q6cuWK2bIUCgU6duyIjh07Asg7hXnMmDFYuXIlPvroI/z+++8A8mZ3L1++jDlz5sDPz8/eTSoSacbZ39/f7Kx7WcAbdxERERERETmYlPB6eHjIsfv37wMwf6ryvn37rH4MUsWKFeXnF585c0aOd+/eHQCwdu3aItXZEYKCgtCgQQOcO3cOFy9eLO3qmMUkmYiIiIiIyIFOnDiBH374AQDQq1cvOS7d2OqXX35Benq6HL9x4wZGjRpltqwFCxbg6tWrJvEtW7YAAKpVqybHxo0bBzc3N7z33ntYs2aNyTZZWVmIjIxEfHx8EVpVdJMmTYJOp8PAgQNx8uRJk+WJiYn48ccfS7ROhni6NRERERERkZ1I1xgDQHZ2NmJjY3H48GHodDr07t0bL7/8sry8T58+eOKJJxAdHY06deqgQ4cOyMzMxO7du9GsWTO0b98eBw8eNCp/wYIFeOONN9CoUSM0bNgQarUa58+fx6lTp+Dq6opPPvlEXrdOnTpYuXIlXnzxRQwcOBB16tRBw4YN4e7ujhs3buD48eNIT0/HiRMnzM5mO8qLL76Is2fP4vPPP0fLli3RrFkz1K5dG0IIXL58GTExMfDw8MCrr75aYnUyxCSZ6DFTY0Lxn4MnufZFmN3KIiIiIioPfv75Z/n/SqUSGo0GnTt3xssvv4yhQ4ca3eDK2dkZf/75Jz766CNs3boVmzZtQtWqVfGf//wHn3zyidGss+TTTz/FunXrcOTIEezcuRPZ2dkICgrCyJEj8d577xndORsA+vbti5iYGMydOxdRUVGIioqCk5MTqlSpgt69e2PAgAFo1KiR4zrEgs8++ww9evTAt99+iwMHDuD06dPw8vJC1apV8cYbb+C5554r8TpJFMLwVmnlWEpKCry9vZGcnAwvL6/Srg5RqWGSTESFmTp1qt3Kmjx5st3KIqKyIzMzE1evXkXNmjXh6upa2tWhx5i1Y9GWfJDXJBMRERERERHpMUkmIiIiIiIi0mOSTERERERERKTHG3cRUZF9cSLBbmVNaO5vt7KIiIiIiIqKM8lEREREREREekVKkufOnYsBAwagbt268Pb2houLC4KDg/HKK6/g9OnTFrdbunQp2rRpAw8PD/j6+qJXr14mz/3K78CBA+jVqxd8fX3h4eGBNm3aYNmyZUWpNhEREREREVGBipQkf/7559i6dSt8fX3x1FNPISwsDK6urli+fDlatmyJTZs2mWwzZswYDBs2DGfOnEFoaCjatGmDqKgodO7cGevWrTP7OqtXr0aXLl2wbds2NGnSBM888wwuXbqEiIgIvPfee0WpOhEREREREZFFRbomef369WjZsqXJc6i+//57vPXWWxg5ciTi4+OhVucVv2PHDsybNw9+fn44dOgQ6tatCwA4dOgQQkJCMGzYMISEhECj0chl3bt3D8OHD4dWq8Xq1asxYMAAAMDt27fRsWNHzJkzB88++yxCQkKK0gQiIiIiIiomIURpV4Eec44Yg0WaSe7QoYPZBzW/+eabqF27Nm7fvo1z587J8blz5wIAPv74YzlBBoB27dph1KhRSEpKwuLFi43KWrRoEVJSUtC3b185QQaAwMBAzJo1CwAwZ86colSfiIiIiIiKQaVSAQBycnJKuSb0uJPGoDQm7cHuN+5ycnICADg7OwMAMjIysGvXLgBAeHi4yfpSbOPGjUbxzZs3W9xGOr17x44dyMzMtF/liYiIiIioUE5OTnBxcUFycjJnk6nUCCGQnJwMFxcXOQ+1B7s+Amr58uW4cOEC6tatK88YX7hwAVlZWQgICEBQUJDJNi1atAAAxMTEGMVPnTpltNyQs7MzGjdujOjoaFy8eBFNmjSxZzOIiIionOMj7IiKz9/fHzdu3EB8fDy8vb3h5OQEhUJR2tWix4AQAjk5OUhOTkZaWhqqVq1q1/KLlSR/+eWXOHv2LNLT0/H333/j7NmzqFKlClauXClPd1+/fh0AzCbIAODu7g6NRoP79+8jNTUVnp6eSElJQXJycoHbBQUFITo6GrGxsUySiYiIiIhKmJeXFwAgISEBN27cKOXa0OPIxcUFVatWlceivRQrSf7jjz+wc+dO+ffg4GAsW7YMLVu2lGNpaWkAgAoVKlgsx93dHUlJSXKSLG1T0Hbu7u4AgNTUVLPLs7KykJWVJf+ekpICIO+cdem8daVSCZVKBa1WC51OJ68rxXNzc41OH1GpVFAqlRbj+a/JkG5clpuba1XcyckJOp0OWq1WjikUCqjVaotxS3Vnm9gmS21SQMA53yUbWVoFlBBwMogLAWTrFFApBNRK83Gl7mH5AgoIpQoKnRYKPKyjUCghFEqTuE6hBAzihteT8H1im9im0m2TUml8NZZUB3NxhUJhMnNkGC+rf3MBAEJAKR6+H9J+DEIHpdCZxBVCB4WZeFlpU3kYe2zTo9cmNzc31KpVC1lZWcjJyZHLf5TbBJS/96m8tUmhUEClUsHNzQ06nc6oHEttMvt3wIJiJck7duwAACQlJeH06dOYNm0aunTpgunTp+Ojjz4qTtHFNmPGDEydOtUkvn37djnxrl69Opo3b46YmBh5xhsA6tevjwYNGuDo0aO4e/euHG/WrBmCg4Oxb98+o+S8Xbt2qFixIrZv327U+V27doWbmxu2bNliVIdevXohIyMDu3fvlmNqtRphYWFISEjAoUOH5Linpye6deuGuLg4nDx5Uo4HBASgffv2uHTpEi5cuCDH2Sa2qbA2BboBE5s93GFk5gIf/KVGPW+BNxo93NndfAB8cUqN1gECL9R+GP87SYEFf6vQvaoO9eOPyPEk94q46VcXle5fgSb9jhy/61UNCZrqCEo4D4/MpIfl+9ZGkkcl1Lx9Ci45GdgSX/Q2lcf3iW1im0qzTYZnaGm1Wpw+fRqenp6oXbu2HM/IyMCFCxfg4+OD6tWry/GUlBRcuXIFgYGBqFSpkvwapd2m/O8TqjwJ98wkVL/78EajWU5uuFK5BTTpd1D53mU5nuaqQVzFJ+CXHI+AlDg5Lu33ykqbysPYY5se3TYlJyebbVNsbKzZNp0/f95sm06cOGG2TQcPHjTbpl27dplt086dO822afv27Va36c6dO2zTI9Amaz9PBU3a5qcQdrzSPicnB+3atcPx48dx5MgRtG7dGhs2bEDfvn3RvHlzHD9+3Ox2Pj4+SEpKQkpKiny6tbe3NwAgOTnZ7PR5//79sW7dOmzYsAG9e/c2WW5uJrlatWpISEiQyysr34RIytO3O2xT2W1TrYmb7TaT/Hp4q4fxYs4kv9vEr8htMhd/1N8ntoltKs02TZ8+3ShenJnkCRMm2K1Njaf8gRwtoIMCLirjw5dsLSAAuJjs3wAFYLLfGxbe1m4zye838eHYY5vYJraJbSrjbUpNTYWfn5/F/NLo9QpcaiMnJycMHjwYx44dw8aNG9G6dWv52+X4+Hiz26SnpyMpKQk+Pj7w9PQEkHd9g7e3N5KTkxEfH49GjRqZbCeVFxwcbLZcFxcXuLi4mK1j/jufqVQqqFSmtwyX3gxr45buqGZLXKlUmhyEFBS3VHe2iW2yFBdQIEtruq7OQlwrFNBaiOuUpvURShWE6eqFxvPX9XF/nyzF2Sa2CXB8mwwPfAqLCyHM3tlWitvzb26WVmH2/4bM7ceEhTgUCugUZl5Xocz7Ii9/Ofov/fLj2GObbI2zTWwTwDZZqqOtcWvbZKkO5tj9EVD+/nl3eJSm2evXrw8XFxfcvXvX7AX90uxy/ptvNW3a1Gi5oZycHJw5cwaurq6oV6+eXetPREREREREjy+7J8l79+4FAPm6JTc3N3Tr1g0AsGrVKpP1IyMjAcDklOmwsDCj5YY2bdqEzMxMhIaGwtXV1X6VJyIiIiIioseazUnygQMHsG3bNpNTrnJycjB//nwsX74cbm5uGDx4sLxs7NixAPKucbp06ZIcP3ToEBYuXAiNRoMRI0YYlTdy5Eh4eXlh/fr1WLNmjRy/c+cOxo8fDwAYN26crdUnIiIiIiIissjma5IvXbqEYcOGwd/fHy1btoSfnx8SEhJw+vRp3Lx5E66urli6dCmqVasmbxMaGorRo0dj3rx5aNasGbp3747s7GxERUVBCIElS5ZAo9EYvY6vry9++uknDBo0COHh4QgJCYGfnx927NiBpKQkjB07FiEhIcVtPxEREREREZHM5iS5S5cu+PDDD7F3717ExMQgISEBzs7OqFGjBsLDw/HOO++gTp06Jtt9/fXXaNasGb799ltERUXB2dkZoaGhmDRpEtq3b2/2tQYOHIh9+/Zh+vTpOHz4MLKzs9GoUSO8/fbbiIiIsL21REREREQO9sWJBLuVNaG5v93KIiLr2Jwk16xZE5999lmRXmzo0KEYOnSoTdt06NABW7duLdLrERERERFZo8aEzXYra9TgJ+1WFhGVPLvfuIuIiIiIiIjoUcUkmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj0myURERERERER6TJKJiIiIiIiI9JgkExEREREREekxSSYiIiIiIiLSY5JMREREREREpMckmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj11aVeAiIiIiu+LEwmlXQUiIqJygTPJRERERERERHqcSSYiIiolNSZstltZowY/abeyiIiIHmecSSYiIiIiIiLSY5JMREREREREpMckmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj0myURERERERER6TJKJiIiIiIiI9JgkExEREREREekxSSYiIiIiIiLSY5JMREREREREpMckmYiIiIiIiEhPXdoVICIicrSpU6farazJkyfbrSwiIiIqeziTTERERERERKTHJJmIiIiIiIhIj0kyERERERERkR6TZCIiIiIiIiI93riLiIiIiIhsUmPCZruVde2LMLuVRWQPnEkmIiIiIiIi0uNMchnDb+WIiIiIiIhKD2eSiYiIiIiIiPQ4k1yOfXEiwW5lTWjub7eyiIiIiIiIyirOJBMRERERERHpMUkmIiIiIiIi0rM5SX7w4AHWrVuHESNGoH79+nB1dYW7uzuaNm2KadOmIS0tzeK2S5cuRZs2beDh4QFfX1/06tULBw8eLPD1Dhw4gF69esHX1xceHh5o06YNli1bZmu1iYiIiIiIiAplc5L8v//9D/3798dPP/0ElUqFPn36oFOnTrh69SomT56M1q1b486dOybbjRkzBsOGDcOZM2cQGhqKNm3aICoqCp07d8a6devMvtbq1avRpUsXbNu2DU2aNMEzzzyDS5cuISIiAu+9957NjSUiIiIiIiIqiM1JspOTE1577TWcO3cO586dw++//45t27bhwoULaN68Oc6fP48xY8YYbbNjxw7MmzcPfn5+OHXqFNatW4dt27Zh3759UKlUGDZsGJKSkoy2uXfvHoYPHw6tVovIyEjs2bMHkZGROH/+POrUqYM5c+Zgz549xWg6ERERERERkTGbk+SIiAgsXLgQDRs2NIpXrlwZ3333HQBgzZo1yM7OlpfNnTsXAPDxxx+jbt26crxdu3YYNWoUkpKSsHjxYqPyFi1ahJSUFPTt2xcDBgyQ44GBgZg1axYAYM6cObZWn4iIiIiIiMgiu964q2nTpgCArKwsJCYmAgAyMjKwa9cuAEB4eLjJNlJs48aNRvHNmzdb3CYsLAyurq7YsWMHMjMz7dcAIiIiIiIieqzZNUm+cuUKgLxTsn19fQEAFy5cQFZWFgICAhAUFGSyTYsWLQAAMTExRvFTp04ZLTfk7OyMxo0bIzMzExcvXrRnE4iIiIiIiOgxprZnYfPmzQMAPPPMM3BxcQEAXL9+HQDMJsgA4O7uDo1Gg/v37yM1NRWenp5ISUlBcnJygdsFBQUhOjoasbGxaNKkiT2bQURU5k2dOtVuZU2ePNluZRERERE96uyWJG/ZsgWLFy+Gk5MTPv30UzkuPRKqQoUKFrd1d3dHUlKSnCQbPkbK0nbu7u4AgNTUVLPLs7KykJWVJf+ekpICAMjJyUFOTg4AQKlUQqVSQavVQqfTyetK8dzcXAgh5LhKpYJSqbQYl8qVqNV53Zubm2tV3MnJCUoIOKkexoQAsnUKqBQCamXhcZ0AcnQKOCkFlLqH5QuFEkKhhEKnhQIP665TKAGzcRWgUMhlSG0rSpt0Oh20Wq0cUygUUKvVFuOW3o+y9D49ym1SQMBZZRRGllZRpLFnNMaggFCqTMaStWNPagPfJ+vapFQ+fFN0Oh0UCgUUCoXR+ubiQggIIYziOTk5ZaJNEke8T4b9JfWBYUzqL2m7guKF7Q9taZOzUhS+Lzd4W3N1gFYo4KwUUOSLAyh0X24UB6AUWvNxG/qmsLFnz7+5LiqBHC2ggwIuKmH0mtlaQABwMdm/AQrAZL8HABDCqA+k/RiEDkqhM4krhA4KM/GyuI+QlKf9nqPbJI0pS58zW8aeUpdr+XOmVNs09gCUyffJsB+KehwhxQ3/Dj2OY49tKpk25d+2IHZJks+fP48hQ4ZACIEvv/xSvja5NM2YMcPsTMv27dvlxLt69epo3rw5YmJi5BlvAKhfvz4aNGiAo0eP4u7du3K8WbNmCA4Oxr59+4yS83bt2qFixYrYvn27Ued37doVbm5u2LJli1EdevXqhYyMDOzevVuOqdVqhIWFoZ63wBuNHg64mw+AL06p0TpA4IXaD+N/Jymw4G8VulfVoWe1hwPu0G0Ffr2iwsAaOtSPPyLH73pVQ4KmOoISzsMjM+lh+b61keRRCTVvn4JLToYcvx7QCOluPqhzIxoqocWW+KK3KSEhAYcOHZLjnp6e6NatG+Li4nDy5Ek5HhAQgPbt2+PSpUu4cOGCHC+L79Oj3KZAN2Bis4c7jMxc4IO/1EUae4ZjLMm9Im761UWl+1egSX/4GDhrx540xvg+Wdcm6QyajIwMXLhwAT4+Pqhevbq8fkpKCq5cuYLAwEBUqlRJjicmJiIuLg5BQUHw8/MDkPclp73aNOOkCvezgVltjA8Kxx9VwcfZ/Nhr4K0zO/b2vtHYbu+T4RlH169fx71791C3bl24ubnJ8cuXLyM1NRVPPPEEVKqHR3rnz59Hdna2XIb0ftlj7A2vX/i+vF3gw/jWOAW2xaswvL4ODTUP4ysv5x1pFrYvl9taqTly1c5Gn2EAuBD0JNS52Who0F9arRanT5+Gp6cnateuLcetHXtSf9nj8zSrDfDfc0qcT1ZgWgstXA2OYmwde1cBuGcmofrdc3I8y8kNVyq3gCb9DirfuyzH01w1iKv4BPyS4xGQEifHpf3ezz//LH+eAODWrVu4desWatWqBS8vr4fvh37s1a9f3+zY+7//+z+jscf9Xsm2aVabvPjKy0ocvqPA2P/TorLBXI1NYy/+iPx5qn3rhBzWKlS4WK2tTWMPCCyT75PUX8U5jpD2e1u2bCkTbZLw81Q+21TQpG1+CmGY0hfBjRs30KFDB8TGxmLs2LEmd5zesGED+vbti+bNm+P48eNmy/Dx8UFSUhJSUlLk0629vb0BAMnJyUZ/YCT9+/fHunXrsGHDBvTu3dtkubmZ5GrVqiEhIUEuryx+E1Jrwia7zSS/OrDVw/WLOZP8bhO/Irfpcf7Gqiy2qdbEzXabSX493GCMFXMmWRpjfJ+sa9MXX3whx4s7kzxhwgS7tanepD9sms0raOxd+byn3d6nGTNmmPRBUWeSJ0yYAMA+Y++JyX/YbSb51UFt7TaTnLN5gdV9U9jYk/rLHp+nxlP+sNtM8rDwtnabSc7e+L3Zz1n+/rJm7En9BXC/V9Jtajzlj7w62GEmecSAVnabSf6gZWCZfJ+k/gKKP5N8ZkqPMtEmQ/w8lb82paamws/Pz2J+afR6BS4txL179/D0008jNjYWw4YNw+zZs03Wkb5djo+PN1tGeno6kpKS4OPjA09PTwCAl5cXvL29kZycjPj4eDRq1MhkO6m84OBgs+W6uLjI10UbcnJygpOTk1FMpVIZfXMrkd4Ma+P5yy1KXAcFsrSm62qFAlob4jk6Rd5OOB+hVEGYrm4xLpWRv662tEmpVJocEBQUt/R+lKX36VFuk7Awxooy9uwxxqR4ccaYpfij/D5Zikt1N/zDBDw8+M7Pmrjh6xS3TdKpgebGkrAQtzT27Pk+5e8vAGZj1sStGavWjr1snf5UygL25eZkW4gXti83iSssxG3om8LGmD3/5mZpFWb/b8iWsQeFwnwfKJR5X+TlL0f/pZ9J3EIfFGWMmRtP3O+VTJvyjylLnzNrxp7hZ878GLNt7JXF9yl/PxTnGNbwdR7HsWdtnG0qXpss1cGcIifJaWlp6NmzJ86dO4cBAwbgxx9/NPkmGcibendxccHdu3dx48YNVK1a1Wi5NLuc/+ZbTZs2xb59+3D8+HGTJDknJwdnzpyBq6sr6tWrV9QmEJn44kSC3cqa0NzfbmUREREREVHJKNIjoLKystC3b18cPXoUPXr0wMqVK81+kwAAbm5u6NatGwBg1apVJssjIyMBwOSU6bCwMKPlhjZt2oTMzEyEhobC1dW1KE0gIiIiIiIiMmFzkqzVavHCCy9g165d6NSpE9asWQNnZ+cCtxk7diwAYPr06bh06ZIcP3ToEBYuXAiNRoMRI0YYbTNy5Eh4eXlh/fr1WLNmjRy/c+cOxo8fDwAYN26crdUnIiIiIiIissjm062//fZbrF27FgDg7++PN9980+x6s2fPhr9/3ummoaGhGD16NObNm4dmzZqhe/fuyM7ORlRUFIQQWLJkCTQajdH2vr6++OmnnzBo0CCEh4cjJCQEfn5+2LFjB5KSkjB27FiEhITYWn0iIiIiIipDeLkblTU2J8n379+X/y8ly+ZMmTJFTpIB4Ouvv0azZs3w7bffIioqCs7OzggNDcWkSZPQvn17s2UMHDgQ+/btw/Tp03H48GFkZ2ejUaNGePvttxEREWFr1akYzD1Oq6gmT55st7Ko/OAYIyIiMsW/j0Qlz+YkecqUKZgyZUqRXmzo0KEYOnSoTdt06NABW7duLdLrUflXY8Jmu5Y3avCTdi2PiIiIiIgeLcV6BBQREZGj2PP0OyIiIiJrFenu1kRERERERETlEWeSiYiIiIioXOA13GQPnEkmIiIiIiIi0mOSTERERERERKTHJJmIiIiIiIhIj0kyERERERERkR6TZCIiIiIiIiI9JslEREREREREekySiYiIiIiIiPSYJBMRERERERHpMUkmIiIiIiIi0mOSTERERERERKSnLu0KEBERERERUemYOnWq3cqaPHmy3coqTZxJJiIiIiIiItJjkkxERERERESkx9OtiYgKUGPCZruVde2LMLuVRURERESOwZlkIiIiIiIiIj3OJBMRlZAvTiSUdhWIiIiIqBCcSSYiIiIiIiLSY5JMREREREREpMckmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj0myURERERERER6TJKJiIiIiIiI9JgkExEREREREekxSSYiIiIiIiLSY5JMREREREREpMckmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj11aVeAqLyaOnWq3cqaPHmy3coiIiIiIiLLOJNMREREREREpMckmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj0myURERERERER6NifJx44dwxdffIEBAwYgKCgICoUCCoWi0O2WLl2KNm3awMPDA76+vujVqxcOHjxY4DYHDhxAr1694OvrCw8PD7Rp0wbLli2ztcpEREREREREVrH5EVCffvop1q9fb9M2Y8aMwbx58+Dm5oann34amZmZiIqKwvbt2xEZGYl+/fqZbLN69WoMHjwYOp0OnTt3hr+/P3bu3ImIiAjExMRg9uzZtladiIiIiIiIqEA2J8nt2rVDkyZN0Lp1a7Ru3Ro1atRAVlaWxfV37NiBefPmwc/PD4cOHULdunUBAIcOHUJISAiGDRuGkJAQaDQaeZt79+5h+PDh0Gq1WL16NQYMGAAAuH37Njp27Ig5c+bg2WefRUhIiK3VJyIiIiIiIrLI5tOtP/jgA0ybNg29e/dGpUqVCl1/7ty5AICPP/5YTpCBvGR71KhRSEpKwuLFi422WbRoEVJSUtC3b185QQaAwMBAzJo1CwAwZ84cW6tOREREREREVCCH3rgrIyMDu3btAgCEh4ebLJdiGzduNIpv3rzZ4jZhYWFwdXXFjh07kJmZae8qExERERER0WPMoUnyhQsXkJWVhYCAAAQFBZksb9GiBQAgJibGKH7q1Cmj5YacnZ3RuHFjZGZm4uLFiw6oNRERERERET2ubL4m2RbXr18HALMJMgC4u7tDo9Hg/v37SE1NhaenJ1JSUpCcnFzgdkFBQYiOjkZsbCyaNGlidp2srCyja6VTUlIAADk5OcjJyQEAKJVKqFQqaLVa6HQ6eV0pnpubCyGEHFepVFAqlRbjUrkStTqve3Nzc62KOzk5QQkBJ9XDmBBAtk4BlUJArSw8rhNAjk4BJ6WAUvewfKFQQiiUUOi0UOBh3XUKJWA2rgIUCrkMpTLvRaR+kn6X17chnpubC7VaDZ1OB61WK8cVCgXUarXF98NcHACclQKGN1jP1QFaoTCJ52gBHRRwUT1sJwBkawEBwEUFoz7TKfLeCKXQGq2vU6oBIYziAgoIpQoQOiiFcZt1Op3JXeCFEBBCWIyb60d7jT0FBJxVRmFkaRVFGntGY0zfB/nHkrVjz55jLCcnB05OThbHmC1jD0De58mKMVbY2DPsr7w+sDDGrBh7hu01N8Ysxc2NvZycHLvt9xQQ8ufJUJYWUAA2jT3DzxNgMMaEDgpz8QLGnmF/FfQ5AwofY1KbC9qXWzv2nJWi8H25lWMPQKH7cqM4Chh7NvRNYWPPnn9zXVTCqn25IUtjD4BV+3LDuMWxZ+W+3JqxZ/iZKspxhD32e2Xt2Kik2iSNqeIcR8hl63KLfRwhxaW2Fec4Qorb830y7IeiHkdIcaUut9jHEfkV5zhCij8unyfp36IeRxjGy0qbzL1P+bctiEOT5LS0NABAhQoVLK7j7u6OpKQkOUmWtiloO3d3dwBAamqqxXJnzJiBqVOnmsS3b98ul1u9enU0b94cMTExckIPAPXr10eDBg1w9OhR3L17V443a9YMwcHB2Ldvn9Frt2vXDhUrVsT27duNOr9r165wc3PDli1bjOrQq1cvZGRkYPfu3XJMrVYjLCwM9bwF3mj08EN08wHwxSk1WgcIvFD7YfzvJAUW/K1C96o69Kz2cMAduq3Ar1dUGFhDh/rxR+T4Xa9qSNBUR1DCeXhkJj0s37c2kjwqoebtU3DJyZDj1wMaId3NB3VuREMltID+y4jz588jOzvb5MuJmJgYODs7o0GDBnJMq9Xi9OnT8PT0RO3ateX4vn370K1bN8TFxeHkyZNyPCAgAO3bt8elS5dw4cIFOV7Q+wQAw+vr0FDzsA9WXlbi8B0Fxv6fFpUNhtB/zylxPlmBaS20cDUY+TNOqnA/G5jVRgsY9NmFoCehzs1G7VsnHrZJocLFam3hnpmE6nfPyfEsJzdcqdwCmvQ7qHzvcl6wSROkpKTgypUrCAwMNLqGPzExEXFxcQgKCoKfn58cv3XrFm7duoUaNWrAy8tLjsfFxdlt7AW6ARObPdxhZOYCH/ylLtLYMxxjSe4VcdOvLirdvwJN+h05bvXY04+py5cvIzU1FU888QRUqod/bW0Ze9u3b0dYWBgSEhJw6NAhOe7p6Wnz2AOAgTV0aBf4cIxtjVNgW7zK5rEnf570Lldqjly1s1E/AlaOPX0/ZGRk4MKFC/Dx8ZHrC8Cmsbdlyxa77fcC3Qw+TwbGH1XBx9m2sWf0eQKQ5qpBXMUn4Jccj4CUODlu1dgzGDfXr1/HvXv3ULduXbi5uT18P6wce9LnqqB9ubVjb3j9wvfl1o49AIXvy6W2FjL2Ghr0l6V9ubVjT+ove/zNndXGyn25AUtj7ypg3b4chY89a/fl1ow9w/12UY4j7LHfK2vHRiXVpllt8uLFOo6QxB8p/nEE8sbeP0CxjyOksWfP90nqr+IcR8j7vfgjxT+O0Lvs6Vns4whpv/e4fJ4AFOs4Ang49spKm8y9TwXlpPkphGFKXwSurq7IysqCuWL+97//4aWXXkKHDh2wf/9+s9sHBQXhxo0buHHjBqpUqYJ///0XVatWBZD3bb30rYGhIUOGYMWKFVixYgVefPFFs+Wam0muVq0aEhIS5J1GWfx2p9aETXabSX51YKuH6xdzJjlry495cTvM8k2cONFu38LV/mibXWeSRwx42GfF/QbYsM+K+w3wRx99ZLexV2viZrvNJL8ebjDGivkNsD3H2IQJE+w2+1D7o212m0l+87mWxnUvxkyy1F9SHxTnG+AJEybYbb9Xb9IfdptJHjWotd1mkrO3/GDSB0UdYxMmTABgn2+1n5j8h91mkl8d1NZuM8k5mxdY3TeFjT2pv+zxN7fxlD/sNpM8LLyt3WaSszd+b7eZZKm/gLI96yopS7NExW1T4yl/5NXBDjPJIwa0sttMcuamBXabSf7oo4/s9j5J/QUUfyZ5xIBWdptJzti0EIB9ZpInTZr0WHyePvvsM7vNJFsaY2VhH5Gamgo/Pz8kJycbfYlkjkNnkj08PAAADx48sLhOeno6gIffYkjbSNuZa0D+bcxxcXGBi4uLSdzJyQlOTk5GMZVKZfRNk8Rcgl5QPH+5RYnroECW1nRdrVBAa0M8R6fI2wnnI5QqMyekWI5LZRh+sM39bktc6j+lUmmyQwIsvx+W4tk6hUmsoHiW1lIcZvtMpzDzfisUFuLKvJ02jNss7TjysxTP349SP9lj7AkLY6woY88eY0yK23OMSe22NMZsHXs5No4xS3Fz/QVYGGOW4vqxl78fbB1jhnHDcVLc/Z50aqC5sSQsxC2NPcPPk1E5+gMmk3gBY8zcuCnqGMvfZnOfM2vHnjRWCtqXm2NpjBW2LzeJWxp7NvRNYWPMnn9zDfffBe3LTepiIW7NvtyoHEtjz8p9uTVxc+PJln28vfZ7ZenYqKTalH9MFeU4QmL4mSvqcYSh4h5HSOz5PuXvh+Icwxr2V1GPI/IrznGE5HH6PBXnOMJQWWpT/rilOpjj0Bt3SVP28fHxZpenp6cjKSkJPj4+csLr5eUFb2/vAreT4sHBwfauMhERERERET3GHJok169fHy4uLrh79y5u3Lhhsvz48eMAYHJtQNOmTY2WG8rJycGZM2fg6uqKevXqOaDWRERERERE9Lhy6OnWbm5u6NatG7Zu3YpVq1ZhzJgxRssjIyMBAL179zaKh4WFYd++fYiMjMSQIUOMlm3atAmZmZl49tln4erq6sjqExERERERFVuNCZvtWt61L8LsWh4Zc+hMMgCMHTsWADB9+nRcunRJjh86dAgLFy6ERqPBiBEjjLYZOXIkvLy8sH79eqxZs0aO37lzB+PHjwcAjBs3ztFVJyIiIiIioseMzUny5s2b0bZtW/knOzsbAIximzc//KYkNDQUo0ePRmJiIpo1a4Z+/fqhV69e6Ny5M3Jzc7FkyRJoNBqj1/D19cVPP/0EpVKJ8PBwdOvWDc899xzq16+Pf/75B2PHjkVISEixGk5ERERERESUn82nW9+9exdHjhwxiRvGDJ+NBQBff/01mjVrhm+//RZRUVFwdnZGaGgoJk2ahPbt25t9nYEDB2Lfvn2YPn06Dh8+jOzsbDRq1Ahvv/02IiIibK02ERERERERUaFsTpKHDh2KoUOH2vxCRdmuQ4cO2Lp1q82vRURERERERFQUDr8mmYiIiIiIiOhRwSSZiIiIiIiISI9JMhEREREREZEek2QiIiIiIiIiPZtv3EVERERERESl54sTCaVdhXKNM8lEREREREREekySiYiIiIiIiPSYJBMRERERERHpMUkmIiIiIiIi0mOSTERERERERKTHJJmIiIiIiIhIj0kyERERERERkR6TZCIiIiIiIiI9JslEREREREREekySiYiIiIiIiPTUpV0BIiIiIrK/GhM2262sa1+E2a0sIqKyjjPJRERERERERHqcSSYiIiKiEjN16lS7lTV58mS7lUVEJOFMMhEREREREZEek2QiIiIiIiIiPZ5uTUREREQF+uJEQmlXgYioxHAmmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj0myURERERERER6TJKJiIiIiIiI9JgkExEREREREekxSSYiIiIiIiLSY5JMREREREREpMckmYiIiIiIiEiPSTIRERERERGRHpNkIiIiIiIiIj0myURERERERER6TJKJiIiIiIiI9JgkExEREREREekxSSYiIiIiIiLSY5JMREREREREpFfmk+SMjAx88sknqFevHlxdXVGlShUMHz4cN27cKO2qERERERERUTlTppPkzMxMdOvWDZ9++inS0tLQt29fVKtWDUuWLEHz5s1x5cqV0q4iERERERERlSNlOkmePn06Dh8+jHbt2uHixYv47bffcOTIEcyZMwd3797F8OHDS7uKREREREREVI6U2SQ5Ozsb3377LQDgu+++g4eHh7xs7NixaNKkCfbu3Ytjx46VVhWJiIiIiIionCmzSfKBAweQnJyM2rVro3nz5ibLw8PDAQAbN24s6aoRERERERFROVVmk+RTp04BAFq0aGF2uRSPiYkpsToRERERERFR+VZmk+Tr168DAIKCgswul+KxsbElViciIiIiIiIq39SlXQFL0tLSAAAVKlQwu9zd3R0AkJqaanZ5VlYWsrKy5N+Tk5MBAPfu3UNOTg4AQKlUQqVSQavVQqfTyetK8dzcXAgh5LhKpYJSqbQYl8qVqNV53Zubm2tV3MnJCchKh1r1MCYEkKNTQKUQUCmtjzspBbJT7j+MK5QQCiUUOi0UeFh3nUIJmI2rAIUCSl1eHbOzs/Pi+n5SKo2/X7Elfu/ePajVauh0Omi1WjmuUCigVqstvh/m4rqsB3BSCigUD19TqwO0QmESz9UCOijgrHrYTgDI0QICgLMKRn2mU+S9EUqhNVpfp1QDQhjFBRQQShUgdFAKnUmfKRQKKAwqI4SAEMJiPH8/JiUl2W3siax0OKmMwsjWKqCEsHnsGY0xfR/kH0vWjj17jrHExEQ4OTlZHGO2jD1bxlhhY8+wv/L6wMIYs2LsSf0l9UH+sWQpbm7sJSYm2m2/J7LS5c+ToWwtoABsGnuZqcny5wkwGGNCB4W5eAFjz7C/LH3OrB1jiYmJAArel1s79lQ56YXuy60de5lpqYXuy43isDz2cgz6q7C+KWzsSf1lj7+56tx0q/blhiyNvcy0VKv25YZxS2MvOyvLqn25NWNP6i+gaMcRhmNMnZterOMIw7c1KzW5WMcRhvGszMxi7eOluOH4Aop2HGE49tS56QCKdxwhyU65X+zjCCmemZlZ7OMIKX7v3j27HcNK/QUU/ThCimen3C/2cYQkMzMTQPGOI6R4UlJSsY8jpLFn2F9FPY4wHHsFjjEbx565MSb1ga1jz9IYc2T+ZO3fXClvNKyHRaKMevXVVwUA8dFHH5ldfunSJQFA1K1b1+zyyZMnC+Tts/jDH/7whz/84Q9/+MMf/vCHP/wRcXFxheaiZXYmWbqb9YMHD8wuT0/P+zbG09PT7PKJEydi7Nix8u86nQ737t2Dn5+fybckj7OUlBRUq1YNcXFx8PLyKu3qlHnsL9uwv2zD/rIN+8s27C/bsc9sw/6yDfvLNuwv27C/TAkhkJqaiipVqhS6bplNkqtXrw4AiI+PN7tcigcHB5td7uLiAhcXF6OYRqOxXwXLGS8vL36AbMD+sg37yzbsL9uwv2zD/rId+8w27C/bsL9sw/6yDfvLmLe3t1XrldkbdzVt2hQAcPz4cbPLpXiTJk1KrE5ERERERERUvpXZJLlDhw7w9vbG5cuXcfLkSZPlkZGRAIDevXuXcM2IiIiIiIiovCqzSbKzszPefvttAMBbb70lX4MMAHPnzkVMTAy6dOmCli1bllYVywUXFxdMnjzZ5NR0Mo/9ZRv2l23YX7Zhf9mG/WU79plt2F+2YX/Zhv1lG/ZX8SiEsOYe2KUjMzMTISEhOHLkCCpXroxOnTohNjYWR44cQUBAAA4fPoxatWqVdjWJiIiIiIionCjTSTIAZGRkYMaMGfjf//6HuLg4+Pr64plnnsGnn36KoKCg0q4eERERERERlSNlPkkmIiIiIiIiKill9ppkIiIiIiIiopLGJLmcOHr0KBQKBRQKBaZNm1bo+kuXLkWbNm3g4eEBX19f9OrVCwcPHjS77vnz5zFz5kx07doV/v7+cHJyQqVKlTBgwAD8+eef9m5KiXBkf5nz6aefyq/3yy+/FKfqpaKk+mvdunV45plnEBAQAFdXV1SrVg39+/fH/v377dGMEuPo/tJqtZg/fz5atmwJd3d3eHt7o3PnzlizZo09m1FirO2vuLg4fP/99xg6dCgaNmwIpVIJhUKBPXv2FPoaBw4cQK9eveDr6wsPDw+0adMGy5Yts2MrSo4j+ys9PR3Lly/Hf/7zHzz55JNwcXGBQqHAlClT7N+QEuLI/tqzZ49ctrmftm3bOqBFjlUSn0eg/OzvAev7bMOGDYiIiMD//d//ycdTFStWRK9evbBp0yaz28TExODtt99G27ZtUaVKFbi4uMDb2xvt2rXD/PnzkZOT46hmOYy1/TVlypQCP18TJkyw6vUel2OwovZXeTumsBtB5cLbb78tAAgAol69egWuO3r0aAFAuLm5ib59+4oePXoItVotVCqVWLt2rcn6VatWFQCEh4eHCA0NFYMGDRKNGzcWAIRCoRBfffWVYxrlQI7sr/zOnz8vXFxchEKhEADE8uXL7dSKkuPo/tJqtWL48OECgHB3dxc9evQQgwcPFu3atRPOzs7i008/dUCrHMeR/ZWbmyueffZZ+TP59NNPi9DQUFGhQgUBQEyePNkxjXIga/vrq6++ktcz/Nm9e3eB5UdGRgqVSiUUCoXo0qWLGDhwoNBoNAKAGDdunJ1b43iO7K8TJ06Y3eZRHFcSR/bX7t27BQBRu3ZtERERYfIzbdo0B7TIsRz9eSxv+3shrO+zgQMHCoVCIRo3bix69eolBg8eLJ588kl524kTJ5psM3/+fAFABAcHi6eeeko8//zz4qmnnhKurq4CgOjSpYvIyspyZPPsztr+mjx5sgAgOnToYPbz9fvvvxf6Wo/TMVhR+qs8HlPYC5PkciA7O1v4+/sLAKJSpUoCgDh8+LDZdaOiogQA4efnJy5evCjHDx48KJydnYVGoxH379832uapp54Sy5YtExkZGUbxBQsWCABCpVKJs2fP2r1djuLo/jKk0+lE586dRWBgoOjbt+8juYMuif6Sduy9e/cWiYmJRsvu3btnVFZZ5+j+mj17tgAgatSoIf755x85/vfff4sqVaoIAOLgwYMOaZsj2NJf69evF2PGjBErVqwQFy9eFE8//XShB+WJiYnCy8tLABCrV6+W47du3RJ16tSx6qC+LHF0f/3zzz9ixIgRYsGCBeLYsWNi2rRpj/SBkqP7S0qSIyIiHNOAEubo/hKifO3vhbCtz44fPy4SEhJM4ocPHxYeHh5CoVCImJgYo2WXL18Wly9fNtnm1q1b8oTF/Pnz7dOYEmBLf0ljZcmSJUV6rcftGKwo/VXejinsiUlyObBhwwb5myPpgOatt94yu27Pnj0FALOzv++8844AIGbPnm31a0t/FKdMmVLU6pe4kuyvH374QQAQv/zyi4iIiHgkd9CO7q+4uDjh7OwsqlevLh48eOCIJpQoR/dX7dq1BQCxYsUKk22k8davXz+7tKUk2NJf+fXo0aPQg/KZM2cKAKJv374my9asWSMAiGeffbaItS95ju6v/GbMmPFIJ8mO7q/yliQ7ur/K2/5eiOL1maERI0YIAGLevHlWb7N8+XIBQPTv39/m1ysttvRXcZPkx+0YrCj9Vd6OKeyJ1ySXA9L1FUOGDMGQIUMAAL/99pvJdSoZGRnYtWsXACA8PNykHCm2ceNGq1+7adOmAIB///3X9oqXkpLqr1u3bmH8+PF46qmn8NJLL9mt/iXN0f31888/Izs7GyNHjoSbm5vd61/SHNlfycnJuHz5MgAgJCTEZJuuXbsCAP744w9kZWUVsyUlw9r+KqrNmzcDMN/HYWFhcHV1xY4dO5CZmWmX13M0R/dXecP+so2j+6u87e8B+/WZk5MTAMDZ2dmh25S2kvpMPm7HYEVRHo8p7IlJ8iMuOTkZGzZsgLOzMwYNGoSaNWuiffv2SEhIwLZt24zWvXDhArKyshAQEGD2GdMtWrQAkHeTCGtduXIFAFCpUqVitKLklGR/vfPOO8jIyMB///tf+zekhJREf0mJYvv27XHz5k3Mnj0bo0aNwgcffIBt27ZBPEJPqXN0f6Wnp8v/9/HxMdnGz88PQF4CfvHiRbu0yZFs6a+iOnXqFICH/WnI2dkZjRs3RmZmJvurHCrJ/rp06RImTpyI1157DR9++CG2bNkCnU5n19dwtJLor/K0vwfs12enT5/Gb7/9BicnJ3Tv3t2qbe7fv485c+YAyPvC71FQ1P7atWsXxowZg1GjRmH69Ok4duxYoa/1uB2DGbK2v8rbMYW9MUl+xEVGRiIzMxM9e/aEr68vAMjfNC1fvtxo3evXrwOA2QNyAHB3d4dGo8H9+/eRmppa6GtfvnxZvhtjnz59ityGklRS/bVp0yasWrUKH374IerWrWvvZpSYkuivc+fOyf82bNgQ77//PhYuXIhZs2ahZ8+e6NatG5KSkuzdNIdwdH/5+vpCpVIBAGJjY022uXr1qvx/c8vLGlv6qyhSUlKQnJwMwHI/S3H2V/lTkv118OBBfPHFF/jxxx8xY8YMhIWFoVmzZrh06ZJdX8eRSqK/ytP+Hih6n23cuBFDhw7FSy+9hE6dOqFZs2Z48OABfvzxR9SuXdvsNpcuXcLQoUPxyiuvoEePHqhevTr++usvjBo16pGZKS1qfy1fvhzz5s3DwoULMWnSJLRq1Qrh4eFIS0szu/7jeAxmyNr+Km/HFHZX2ud7U/F06dJFABCrVq2SYwkJCcLJyUm4urqKpKQkOb5ixQr5ugZLpDtZ37hxo8DXzcnJER07dhQAxODBg4vfkBJSEv2VmpoqqlWrJurVqycyMzPl+KN4PUxJ9JeLi4sAINRqtejUqZM4fvy4SElJETt27BA1a9YUAER4eLhjGmhnJdFfbdu2FQDEBx98YLL+W2+9Jd8B83//+5+dWuU4tvSXOYVdA3njxg25P3Jycsyu89JLL1m8HquscXR/mfMoX5NcEv11/Phx8f7774vDhw+LxMREkZiYKHbu3Cl/TmvUqFHo65QVJdFf5Wl/L0TR++zTTz+V901A3tMNfvjhB6HVai2+1p9//mm0DQDxzjvviOTkZLu3y1Fs7a/ly5eL2bNni7Nnz4q0tDQRFxcnVqxYIf9tNHet7ON6DCZE0fqrPB1T2BuT5EdYbGysUCgUQqPRGO0IhBDyXfx+/PFHOWbPJPmNN94QAEStWrVM7k5ZVpVUf/3nP/8RAMTOnTuN1n3UdtAl1V9OTk4CgAgICBBpaWlG658+fVp+bMOFCxfs1DLHKKn+WrdunXyQ+eWXX4qbN2+KGzduiOnTpwuVSiXUarUAIH799Vf7N9KObO0vcx6nJLkk+sucRzVJLq3+kuTm5opOnToJAOLzzz+3efuSVlL9VV7290LYp88yMjLE6dOnxZtvvinfRLCwxznl5uaKK1euiDlz5ggvLy9Rt25dcfXq1eI2x+Hs0V+Sf//9V/j5+QkA4tChQ0bLHtdjsIIU1F/l5ZjCEXi69SNsxYoVEEIgPDwcLi4uRsuk0zEMH5ru4eEBAHjw4IHFMqXrEzw9PS2u89lnn+G///0vAgMD8ccff8ingJR1JdFfR48exXfffYeXX34Z3bp1s2v9S1pJjS9pu+eeew7u7u5G6zdu3BitW7cGAOzbt6+oTSkRJdVfffv2xcyZMyGEwPvvv4/KlSujatWq+PjjjzFixAg0b94cgPnri8oSW/urKKQ+Biz3szX7vLKgJPqrPCnt/lKpVPjggw8A5N30pqwrqf4qL/t7wD595urqisaNG+O7777Df/7zH2zatAnz588vcBuVSoWaNWti7NixWLJkCS5duoT//Oc/xWtMCbDnGKtcuTKGDRsGAEbX5j7Ox2AFsdRfQPk5pnAEdWlXgIpOuh5hz5496Nixo9Gy7OxsAHl/aGJjYxEcHIzq1asDAOLj482Wl56ejqSkJPj4+Fg8YFywYAE+/vhjeHt7Y9u2bahTp469muNwJdFf0s1aTp8+bXKnwPPnzwPI+5Jh0aJFeOaZZzBhwgS7tc/eSmp8BQcH4/79+6hRo4bZ7WrUqIGjR4/izp07xW2SQ5Xk53H8+PHo378/IiMjce3aNXh7eyMsLAxdunSRr7F94okn7No+e7O1v4rCy8sL3t7eSE5ORnx8PBo1amSyjtT/RX2NklIS/VWelIX+kq6FvHnzpkPKt6eS6q/ysr8H7N9nL7/8MubPn4/169dj3LhxVtWhf//+8PDwwLZt25CdnV2m73Jt7/4y9/l6nI/BClPQ/qg8HFM4RKnMX1OxRUdHm1ybYunns88+E0II8eDBA/l6oPj4eJMy9+3bJwCILl26mH3NlStXCqVSKSpUqCD279/vyObZXUn1l/SMOmt+yvJzNUtyfA0bNkwAEBMnTjRbl9DQUAHY9uzIklYan0dzYmNjBQBRp04dezXNIYrSX+ZYc3pn586dLZ5il52dLVxdXYWrq6vIyMiwR9McoiT7K79H8XTr0uwvQ4cPHxYARPPmzYvYkpJRkv1VHvb3QtivzwxduXJFABANGjSwqS7Vq1cXAMStW7eK0pQS4Yj++uKLLwQA8e6778qxx/kYrDDm+qswj8oxhaPwdOtHlHSKxXvvvQeRd225yc+ePXuM1nVzc5NPP1m1apVJmZGRkQCA3r17myzbsmULXnnlFajVaqxduxYdOnRwRLMcpqT6a8qUKRbLj4iIAJD37aAQAkuXLnVUc4utJMeXdGf0vXv3mmyTlpaG48ePA4B8yk9ZVNKfR0uk0/Ree+21IrelJBSlv4pKejSK1J+GNm3ahMzMTISGhsLV1bVYr+NIJdlf5UFZ6a/Vq1cDMP/4sbKkJPurPOzvAcf0mdQnlu5ubc6VK1cQFxcHLy8v+Pv729aIEmTv/hJCYO3atQCMP1+P8zFYQSz1V2EelWMKh7E1q6bSl5ubKwIDAwUAcezYMYvrabVa+cY/0dHRQgghoqKiBADh5+cnLl68KK978OBB4eLiIjQajbh//75ROfv37xdubm5CrVaLtWvXOqJJDlXS/WXJo3LTiJLur9zcXNGwYUMBQHz33XdG8VdffVUAEI0bNxY6nc6+DbWTku6vtLQ0ce7cOZPyFyxYIFQqlahfv77JTT7KkuL0V37WzFwlJiYKLy8vAUCsXr1ajt++fVvUqVOnWDOFJaGk+yu/R20muaT766uvvhLXr183iul0OrFgwQKhVquFQqGwWH5ZUNL99ajv74Uoep/duXNH/PDDDyI9Pd1k3e3bt4uAgAABQERGRhot++abb8TNmzdNtjl//rx48sknBQDx9ttvF79hDlKc/vr2229FSkqK0Xqpqani9ddfFwBEpUqVzPanOeX9GKyo/fWoH1M4EpPkR9CWLVsEAFGvXr1C1x07dqwAIEaPHi3HRo8eLQCIChUqiL59+4qePXsKtVotVCqV2SRYo9EIAKJmzZoiIiLC7I+1d9grDSXdX5Y8Kjvo0uivEydOyIlM06ZNxcCBA0WtWrXkBDImJsZOrbO/ku6vq1evCgDiiSeeEP369RPPPfec3Fc1atQQV65csWPr7K84/fXvv/+KJ598Uv7x9vYWAETDhg3l2LRp00zKiYyMFEqlUigUCtG1a1cRHh4u79fGjh1r7ybaVWn0V79+/eTl1apVEwBE1apV5Zi5x4iUFSXdX8HBwUKlUonWrVuLQYMGiT59+siPMlIqlWL+/PmOaKbdlMb4epT390IUvc+kfXeFChVEp06dxAsvvCD69OkjGjRoIJ86a+5U2ODgYKFUKkXz5s3Fc889J8LDw0Xr1q2FUqkUAETnzp1FamqqI5pqF8XtLw8PD9G1a1fx4osviu7du8t3adZoNDZd+lfej8GK2l+PhcjU2wAAAflJREFU+jGFIzFJfgS98MILVn+z/9dffwkAomLFikaPQFmyZIlo2bKlqFChgtBoNOKZZ54RBw4cMFuGtPMu6KcsX9tR0v1lyaOygy6t/rpy5Yp45ZVXRKVKlYSTk5MICgoSI0eOFNeuXStukxyqpPsrOTlZjBo1SjRq1Eh4enqKChUqiCeeeEJMnjy5TB8oSYrTX9If86Lsi/bv3y+eeeYZodFoRIUKFUSrVq3E0qVL7ds4ByiN/goODi5wm+DgYLu3015Kur+++eYb8eyzz4qaNWsKd3d34ezsLIKDg8WQIUPE0aNHHdNIOyqtz+Ojur8Xouh9lpycLGbNmiV69eolgoODhZubm3BxcRE1atQQzz//vMUZ+F9++UW8+OKLol69esLLy0s4OTmJwMBA0aNHD7F06dICn61cFhS1v1JSUsQHH3wgunTpIqpWrSpcXFzkv3fjxo0zey+PgpT3Y7Ci9tejfkzhSAohhAARERERERERgTfuIiIiIiIiItJjkkxERERERESkxySZiIiIiIiISI9JMhEREREREZEek2QiIiIiIiIiPSbJRERERERERHpMkomIiIiIiIj0mCQTERERERER6TFJJiIiIiIiItJjkkxERERERESkxySZiIiIiIiISI9JMhEREREREZEek2QiIiIiIiIivf8HRfIAlVe7pDgAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1000x700 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(nrows=2, ncols=1, figsize=(10, 7))\n",
    "plt.rcParams.update({\"font.size\": 15})\n",
    "plt.subplots_adjust(left=0.05, bottom=0.05, right=0.95, top=0.95, wspace=0.3, hspace=0.3)\n",
    "\n",
    "\n",
    "col = \"hsr_dist\"\n",
    "team_players = [p for p in true_att_stats.index if p[0] == \"A\" and true_att_stats.loc[p, col] > 0]\n",
    "team_players.sort()\n",
    "team_true_att_td = true_att_stats.loc[team_players, col]\n",
    "team_true_def_td = true_def_stats.loc[team_players, col]\n",
    "team_pred_att_td = pred_att_stats.loc[team_players, col]\n",
    "team_pred_def_td = pred_def_stats.loc[team_players, col]\n",
    "team_baseline_att_td = baseline_att_stats.loc[team_players, col]\n",
    "team_baseline_def_td = baseline_def_stats.loc[team_players, col]\n",
    "\n",
    "axes[0].set_title(\"HSR Distance Covered When Attacking\")\n",
    "axes[0].bar(np.arange(len(team_players)) - 0.25, team_true_att_td, width=0.25, color=\"tab:red\", label=\"True\")\n",
    "axes[0].bar(np.arange(len(team_players)), team_pred_att_td, width=0.25, color=\"pink\", label=\"Pred\")\n",
    "axes[0].bar(np.arange(len(team_players)) + 0.25, team_baseline_att_td, width=0.25, color=\"grey\", label=\"Baseline\")\n",
    "axes[0].set_xticks(np.arange(len(team_players)), labels=team_players)\n",
    "axes[0].set_ylim(0, 500)\n",
    "axes[0].grid(axis=\"y\", linestyle=\"--\")\n",
    "axes[0].legend(loc=\"upper right\")\n",
    "\n",
    "axes[1].set_title(\"HSR Distance Covered When Defending\")\n",
    "axes[1].bar(np.arange(len(team_players)) - 0.25, team_true_def_td, width=0.25, color=\"tab:blue\", label=\"True\")\n",
    "axes[1].bar(np.arange(len(team_players)), team_pred_def_td, width=0.25, color=\"skyblue\", label=\"Pred\")\n",
    "axes[1].bar(np.arange(len(team_players)) + 0.25, team_baseline_def_td, width=0.25, color=\"grey\", label=\"Baseline\")\n",
    "axes[1].set_xticks(np.arange(len(team_players)), labels=team_players)\n",
    "axes[1].set_ylim(0, 500)\n",
    "axes[1].grid(axis=\"y\", linestyle=\"--\")\n",
    "axes[1].legend(loc=\"upper right\")\n",
    "\n",
    "plt.savefig(\"img/rp_metrics.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "- TD when attacking: 0.0169\n",
      "- TD when defending: 0.0156\n",
      "- HSRD when attacking: 0.1164\n",
      "- HSRD when defending: 0.0684\n"
     ]
    }
   ],
   "source": [
    "# Mean absolute percentage error\n",
    "mape_att_td = ((pred_att_stats[\"total_dist\"] - true_att_stats[\"total_dist\"]).abs() / true_att_stats[\"total_dist\"]).mean().round(4)\n",
    "mape_def_td = ((pred_def_stats[\"total_dist\"] - true_def_stats[\"total_dist\"]).abs() / true_def_stats[\"total_dist\"]).mean().round(4)\n",
    "mape_att_hsrd = ((pred_att_stats[\"hsr_dist\"] - true_att_stats[\"hsr_dist\"]).abs() / true_att_stats[\"hsr_dist\"]).mean().round(4)\n",
    "mape_def_hsrd = ((pred_def_stats[\"hsr_dist\"] - true_def_stats[\"hsr_dist\"]).abs() / true_def_stats[\"hsr_dist\"]).mean().round(4)\n",
    "\n",
    "print(f\"- TD when attacking: {mape_att_td}\")\n",
    "print(f\"- TD when defending: {mape_def_td}\")\n",
    "print(f\"- HSRD when attacking: {mape_att_hsrd}\")\n",
    "print(f\"- HSRD when defending: {mape_def_hsrd}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.557709170359725"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "((baseline_att_stats[\"total_dist\"] - true_att_stats[\"total_dist\"]).abs() / true_att_stats[\"total_dist\"]).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R2 scores for baseline predictions:\n",
      "- TD when attacking: 0.9281\n",
      "- TD when defending: 0.9356\n",
      "- HSRD when attacking: 0.758\n",
      "- HSRD when defending: 0.7922\n"
     ]
    }
   ],
   "source": [
    "baseline_att_td_r2 = r2_score(true_att_stats[\"total_dist\"], baseline_att_stats[\"total_dist\"]).round(4)\n",
    "baseline_def_td_r2 = r2_score(true_def_stats[\"total_dist\"], baseline_def_stats[\"total_dist\"]).round(4)\n",
    "baseline_att_hsrd_r2 = r2_score(true_att_stats[\"hsr_dist\"], baseline_att_stats[\"hsr_dist\"]).round(4)\n",
    "baseline_def_hsrd_r2 = r2_score(true_def_stats[\"hsr_dist\"], baseline_def_stats[\"hsr_dist\"]).round(4)\n",
    "\n",
    "print(\"R2 scores for baseline predictions:\")\n",
    "print(f\"- TD when attacking: {baseline_att_td_r2}\")\n",
    "print(f\"- TD when defending: {baseline_def_td_r2}\")\n",
    "print(f\"- HSRD when attacking: {baseline_att_hsrd_r2}\")\n",
    "print(f\"- HSRD when defending: {baseline_def_hsrd_r2}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculating MAPE for All Test Matches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['data/metrica_traces/match3_test.csv',\n",
       " 'data/gps_event_traces_gk_pred/20862-20875.csv',\n",
       " 'data/gps_event_traces_gk_pred/21689-21704.csv',\n",
       " 'data/gps_event_traces_gk_pred/22213-22216.csv']"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrica_paths = [\"data/metrica_traces/match3_test.csv\"]\n",
    "\n",
    "gps_files = os.listdir(\"data/gps_event_traces_gk_pred\")\n",
    "gps_paths = [f\"data/gps_event_traces_gk_pred/{f}\" for f in gps_files]\n",
    "gps_paths.sort()\n",
    "\n",
    "test_paths = metrica_paths + gps_paths[-3:]\n",
    "test_paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[1] data/metrica_traces/match3_test.csv\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Phase 2: 100%|██████████| 10/10 [00:08<00:00,  1.15it/s]\n",
      "Phase 3: 100%|██████████| 2/2 [00:01<00:00,  1.93it/s]\n",
      "Phase 4: 0it [00:00, ?it/s]\n",
      "Phase 5: 0it [00:00, ?it/s]\n",
      "Phase 6: 100%|██████████| 1/1 [00:01<00:00,  1.79s/it]\n",
      "Phase 7: 100%|██████████| 2/2 [00:01<00:00,  1.26it/s]\n",
      "Phase 8: 100%|██████████| 4/4 [00:03<00:00,  1.32it/s]\n",
      "Phase 9: 100%|██████████| 4/4 [00:04<00:00,  1.25s/it]\n",
      "Phase 10: 100%|██████████| 6/6 [00:03<00:00,  1.57it/s]\n",
      "Phase 11: 100%|██████████| 7/7 [00:05<00:00,  1.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "macro_acc: 0.9077\n",
      "micro_pos_error: 5.7287\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Calculating stats: 100%|██████████| 32/32 [00:03<00:00,  8.74it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction\n",
      "- TD when attacking: 0.036\n",
      "- TD when defending: 0.034\n",
      "- HSRD when attacking: 0.2058\n",
      "- HSRD when defending: 0.1251\n",
      "\n",
      "Baseline\n",
      "- TD when attacking: 0.0881\n",
      "- TD when defending: 0.0771\n",
      "- HSRD when attacking: 0.514\n",
      "- HSRD when defending: 0.2091\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "cols = [\"total_dur\", \"total_dist\", \"hsr_dist\"] + [f\"zone{i}_dist\" for i in np.arange(5) + 1]\n",
    "speed_bins = [-0.1, 7, 15, 20, 25, 50]\n",
    "\n",
    "true_att_stats_list = []\n",
    "true_def_stats_list = []\n",
    "pred_att_stats_list = []\n",
    "pred_def_stats_list = []\n",
    "base_att_stats_list = []\n",
    "base_def_stats_list = []\n",
    "\n",
    "for i, data_path in enumerate(test_paths[:1]):\n",
    "    print(f\"\\n[{i + 1}] {data_path}\")\n",
    "    match_traces = pd.read_csv(data_path, header=0)\n",
    "\n",
    "    helper = TraceHelper(match_traces)\n",
    "    macro_pred, _ = helper.predict(model, split=True, evaluate=True)\n",
    "    if macro_pred is None:\n",
    "        continue\n",
    "\n",
    "    players = helper.team1_players + helper.team2_players\n",
    "    match_true_att_stats = pd.DataFrame(columns=cols)\n",
    "    match_true_def_stats = pd.DataFrame(columns=cols)\n",
    "    match_pred_att_stats = pd.DataFrame(columns=cols)\n",
    "    match_pred_def_stats = pd.DataFrame(columns=cols)\n",
    "    match_base_att_stats = pd.DataFrame(columns=cols)\n",
    "    match_base_def_stats = pd.DataFrame(columns=cols)\n",
    "\n",
    "    pred_team_poss = macro_pred.idxmax(axis=1)\n",
    "    true_team_poss = helper.traces[\"team_poss\"].copy()\n",
    "    true_team_poss.loc[pred_team_poss.isna()] = np.nan\n",
    "    base_team_poss = pd.Series(np.random.choice([\"A\", \"B\"], len(helper.traces)), index=helper.traces.index)\n",
    "    base_team_poss.loc[pred_team_poss.isna()] = np.nan\n",
    "\n",
    "    for p in tqdm(players, desc=\"Calculating stats\"):\n",
    "        match_true_att_stats.loc[f\"M{i}-{p}\"] = calc_running_metrics(helper.traces, true_team_poss, p, attacking=True)\n",
    "        match_true_def_stats.loc[f\"M{i}-{p}\"] = calc_running_metrics(helper.traces, true_team_poss, p, attacking=False)\n",
    "        match_pred_att_stats.loc[f\"M{i}-{p}\"] = calc_running_metrics(helper.traces, pred_team_poss, p, attacking=True)\n",
    "        match_pred_def_stats.loc[f\"M{i}-{p}\"] = calc_running_metrics(helper.traces, pred_team_poss, p, attacking=False)\n",
    "        match_base_att_stats.loc[f\"M{i}-{p}\"] = calc_running_metrics(helper.traces, base_team_poss, p, attacking=True)\n",
    "        match_base_def_stats.loc[f\"M{i}-{p}\"] = calc_running_metrics(helper.traces, base_team_poss, p, attacking=False)\n",
    "    \n",
    "    true_att_stats_list.append(match_true_att_stats)\n",
    "    true_def_stats_list.append(match_true_def_stats)\n",
    "    pred_att_stats_list.append(match_pred_att_stats)\n",
    "    pred_def_stats_list.append(match_pred_def_stats)\n",
    "    base_att_stats_list.append(match_base_att_stats)\n",
    "    base_def_stats_list.append(match_base_def_stats)\n",
    "\n",
    "true_att_stats = pd.concat(true_att_stats_list)\n",
    "true_def_stats = pd.concat(true_def_stats_list)\n",
    "pred_att_stats = pd.concat(pred_att_stats_list)\n",
    "pred_def_stats = pd.concat(pred_def_stats_list)\n",
    "base_att_stats = pd.concat(base_att_stats_list)\n",
    "base_def_stats = pd.concat(base_def_stats_list)\n",
    "\n",
    "true_att_stats = true_att_stats[(true_att_stats[\"hsr_dist\"] > 0) & (true_def_stats[\"hsr_dist\"] > 0)]\n",
    "true_def_stats = true_def_stats.loc[true_att_stats.index]\n",
    "pred_att_stats = pred_att_stats.loc[true_att_stats.index]\n",
    "pred_def_stats = pred_def_stats.loc[true_att_stats.index]\n",
    "base_att_stats = base_att_stats.loc[true_att_stats.index]\n",
    "base_def_stats = base_def_stats.loc[true_att_stats.index]\n",
    "\n",
    "pred_mape_att_td = ((pred_att_stats[\"total_dist\"] - true_att_stats[\"total_dist\"]).abs() / true_att_stats[\"total_dist\"]).mean().round(4)\n",
    "pred_mape_def_td = ((pred_def_stats[\"total_dist\"] - true_def_stats[\"total_dist\"]).abs() / true_def_stats[\"total_dist\"]).mean().round(4)\n",
    "pred_mape_att_hsrd = ((pred_att_stats[\"hsr_dist\"] - true_att_stats[\"hsr_dist\"]).abs() / true_att_stats[\"hsr_dist\"]).mean().round(4)\n",
    "pred_mape_def_hsrd = ((pred_def_stats[\"hsr_dist\"] - true_def_stats[\"hsr_dist\"]).abs() / true_def_stats[\"hsr_dist\"]).mean().round(4)\n",
    "\n",
    "baseline_mape_att_td = ((base_att_stats[\"total_dist\"] - true_att_stats[\"total_dist\"]).abs() / true_att_stats[\"total_dist\"]).mean().round(4)\n",
    "baseline_mape_def_td = ((base_def_stats[\"total_dist\"] - true_def_stats[\"total_dist\"]).abs() / true_def_stats[\"total_dist\"]).mean().round(4)\n",
    "baseline_mape_att_hsrd = ((base_att_stats[\"hsr_dist\"] - true_att_stats[\"hsr_dist\"]).abs() / true_att_stats[\"hsr_dist\"]).mean().round(4)\n",
    "baseline_mape_def_hsrd = ((base_def_stats[\"hsr_dist\"] - true_def_stats[\"hsr_dist\"]).abs() / true_def_stats[\"hsr_dist\"]).mean().round(4)\n",
    "\n",
    "print(\"Prediction\")\n",
    "print(f\"- TD when attacking: {pred_mape_att_td}\")\n",
    "print(f\"- TD when defending: {pred_mape_def_td}\")\n",
    "print(f\"- HSRD when attacking: {pred_mape_att_hsrd}\")\n",
    "print(f\"- HSRD when defending: {pred_mape_def_hsrd}\")\n",
    "\n",
    "print(\"\\nBaseline\")\n",
    "print(f\"- TD when attacking: {baseline_mape_att_td}\")\n",
    "print(f\"- TD when defending: {baseline_mape_def_td}\")\n",
    "print(f\"- HSRD when attacking: {baseline_mape_att_hsrd}\")\n",
    "print(f\"- HSRD when defending: {baseline_mape_def_hsrd}\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.8"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "d4d1e4263499bec80672ea0156c357c1ee493ec2b1c70f0acce89fc37c4a6abe"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}