9 Evaluating distributional forecast accuracy.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluating distributional forecast accuracy\n",
    "***\n",
    "The preceding measures all measure point forecast accuracy. When evaluating distributional forecasts, we need to use some other measures.\n",
    "\n",
    "## Quantile scores\n",
    "\n",
    "Consider the Google stock price example from the previous section. Figure 23 shows an 80% prediction interval for the forecasts from the Auto ARIMA method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import the libraries that we are going to use for the analysis:\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random\n",
    "from itertools import product\n",
    "from ipywidgets import interact\n",
    "\n",
    "from statsforecast import StatsForecast\n",
    "from statsforecast.models import __all__\n",
    "\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a dataframe from a csv file:\n",
    "google_stock = pd.read_csv(\"Assets/GOOGL.csv\")\n",
    "\n",
    "# Filter the year of interest:\n",
    "google_mask = google_stock[\"Date\"] <= \"2015-12-31\"\n",
    "google_2015 = google_stock[google_mask]\n",
    "\n",
    "# Filter the year of interest:\n",
    "google_mask = google_stock[\"Date\"] > \"2015-12-31\"\n",
    "google_2016 = google_stock[google_mask]\n",
    "\n",
    "# Define a new dataframe with daily close prices and dates:\n",
    "google_train = google_2015[[\"Date\", \"Close\"]]\n",
    "google_test = google_2016[[\"Date\", \"Close\"]]\n",
    "\n",
    "# Rename columns:\n",
    "google_train.columns = [\"ds\", \"y\"]\n",
    "google_test.columns = [\"ds\", \"y\"]\n",
    "\n",
    "# Create a new column in the dataframe, unique_id:\n",
    "google_train[\"unique_id\"] = \"Close price\"\n",
    "google_test[\"unique_id\"] = \"Close price\"\n",
    "\n",
    "# Convert string Date time into Python Date time object:\n",
    "google_train['ds'] = pd.to_datetime(google_train['ds'])\n",
    "google_test['ds'] = pd.to_datetime(google_test['ds'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import the model that we are going to use for the analysis:\n",
    "from statsforecast.models import AutoARIMA\n",
    "\n",
    "# Since we are dealing with daily data, it would be benefitial to use 5 as seasonality.\n",
    "models = [AutoARIMA(season_length=5, approximation=True)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the model, fit and predict:\n",
    "fcst = StatsForecast(df=google_train, \n",
    "                     models=models,\n",
    "                     freq=\"D\",\n",
    "                     n_jobs=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We can define the level of the forecast intervals we want to produce. StatsForecast will produce these levels:\n",
    "levels = [80]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The forecast function converts the forecast distributions into intervals:\n",
    "forecasts = fcst.forecast(h=30, level=levels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reset the index of the DataFrame, and use the default one instead:\n",
    "forecasts = forecasts.reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>y</th>\n",
       "      <th>unique_id</th>\n",
       "      <th>unique_id_x</th>\n",
       "      <th>unique_id_y</th>\n",
       "      <th>AutoARIMA</th>\n",
       "      <th>AutoARIMA-lo-80</th>\n",
       "      <th>AutoARIMA-hi-80</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ds</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-01-02</th>\n",
       "      <td>26.477501</td>\n",
       "      <td>Close price</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-05</th>\n",
       "      <td>25.973000</td>\n",
       "      <td>Close price</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-06</th>\n",
       "      <td>25.332001</td>\n",
       "      <td>Close price</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-07</th>\n",
       "      <td>25.257500</td>\n",
       "      <td>Close price</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-01-08</th>\n",
       "      <td>25.345501</td>\n",
       "      <td>Close price</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-22</th>\n",
       "      <td>37.272999</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>35.041546</td>\n",
       "      <td>42.615097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-25</th>\n",
       "      <td>36.681000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.789295</td>\n",
       "      <td>42.867348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-26</th>\n",
       "      <td>36.689499</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.708645</td>\n",
       "      <td>42.947998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-27</th>\n",
       "      <td>35.879002</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.629539</td>\n",
       "      <td>43.027103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-28</th>\n",
       "      <td>37.415001</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.551899</td>\n",
       "      <td>43.104740</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>270 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    y    unique_id  unique_id_x  unique_id_y  AutoARIMA  \\\n",
       "ds                                                                        \n",
       "2015-01-02  26.477501  Close price          NaN          NaN        NaN   \n",
       "2015-01-05  25.973000  Close price          NaN          NaN        NaN   \n",
       "2015-01-06  25.332001  Close price          NaN          NaN        NaN   \n",
       "2015-01-07  25.257500  Close price          NaN          NaN        NaN   \n",
       "2015-01-08  25.345501  Close price          NaN          NaN        NaN   \n",
       "...               ...          ...          ...          ...        ...   \n",
       "2016-01-22  37.272999          NaN  Close price  Close price   38.82832   \n",
       "2016-01-25  36.681000          NaN  Close price  Close price   38.82832   \n",
       "2016-01-26  36.689499          NaN  Close price  Close price   38.82832   \n",
       "2016-01-27  35.879002          NaN  Close price  Close price   38.82832   \n",
       "2016-01-28  37.415001          NaN  Close price  Close price   38.82832   \n",
       "\n",
       "            AutoARIMA-lo-80  AutoARIMA-hi-80  \n",
       "ds                                            \n",
       "2015-01-02              NaN              NaN  \n",
       "2015-01-05              NaN              NaN  \n",
       "2015-01-06              NaN              NaN  \n",
       "2015-01-07              NaN              NaN  \n",
       "2015-01-08              NaN              NaN  \n",
       "...                     ...              ...  \n",
       "2016-01-22        35.041546        42.615097  \n",
       "2016-01-25        34.789295        42.867348  \n",
       "2016-01-26        34.708645        42.947998  \n",
       "2016-01-27        34.629539        43.027103  \n",
       "2016-01-28        34.551899        43.104740  \n",
       "\n",
       "[270 rows x 7 columns]"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Merge test dataframe with forecast predicted\n",
    "df_plot_1 = google_test.merge(forecasts, left_on='ds', right_on='ds')\n",
    "\n",
    "# Concatenate train dataframe with the previous dataframe\n",
    "df_plot_2 = pd.concat([google_train, df_plot_1]).set_index('ds')\n",
    "\n",
    "df_plot_2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1c62f31e170>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABmMAAAJ+CAYAAABPWXS6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy89olMNAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd3QU1f/G8WfTQ0IChE6A0HtHOgKCFOlNURAQREBsKBYQBQW/ICJgQUWkCgoIUlWKdJBepEkvAqEH0hvZ+f2R364JKaRssiF5v87ZY3bnzp3PzG7Gwz6595oMwzAEAAAAAAAAAACADOFg7wIAAAAAAAAAAACyM8IYAAAAAAAAAACADEQYAwAAAAAAAAAAkIEIYwAAAAAAAAAAADIQYQwAAAAAAAAAAEAGIowBAAAAAAAAAADIQIQxAAAAAAAAAAAAGYgwBgAAAAAAAAAAIAMRxgAAAAAAAAAAAGQgwhgAAADkSP3795fJZJKfn5+9S0kRk8kkk8mksWPHZkj/c+fOtR7j4sWLCbY/atfrQQ87P3vasmWLtbYtW7bYuxykUEb/TtqLn5+fTCaT+vfvb+9SAAAAshUnexcAAACAjHf//n2tXLlS69at065du3Tjxg3dvXtXuXLlUoECBVStWjU1atRIPXr0UKlSpexdLgAAAAAA2QojYwAAALK5VatWqVKlSurRo4dmzpypY8eO6datW7p//76CgoJ07tw5rVixQu+8845Kly6tDh066NixY/YuG0AOxKgMAAAAZFeMjAEAAMjGxo8frw8//FCGYUiSmjdvrg4dOqh69ery8fFRWFiYrl27pm3btmnNmjW6ePGifvvtN/n6+uq7776zc/XITP379+cLcDtp3ry59XcUsLesNo0fAABAdkEYAwAAkE3Nnj1bH3zwgSSpUKFCWrRokZo3b55o2549e2ratGlatGiRRo0alYlVAgAAAACQ/RHGAAAAZEOXL1/WsGHDJEleXl7asWOHypYtm+w+jo6O6t27t9q3b6/t27dnRpkAAAAAAOQIrBkDAACQDU2ZMkURERGSpE8++eShQUxcefLkUceOHZNts2PHDj3//PPy8/OTm5ub8uTJo1q1amn06NG6detWio6zevVq9ejRQ76+vnJ1dZWPj48aNmyoiRMnKiQk5KH7h4WFady4capevbo8PDzk4+OjJk2aaPbs2TIMQ1u2bJHJZJLJZNKWLVtSVFNiAgMDNWHCBDVu3FgFChSQi4uLihQpoo4dO2rp0qU2m17qp59+UvPmzZU3b155enqqatWqGjNmjO7du5ei/Y8dO6bx48erTZs21mvq6empcuXKqV+/ftq9e3ey+8+dO9d6vVIzTdGXX35p3e9hx5Ck7t27y2QyKV++fNbPaGqZzWb9/PPP6t69u0qUKCF3d3e5u7urfPny6t27t5YuXaro6Og09X3r1i2NHj1atWrVUp48eeTm5iY/Pz89//zz2rFjx0P337Rpk5599lmVKlVK7u7uypUrl0qWLKkGDRpoxIgR2rRpU4J9HvZZ7d+/v0wmk/z8/CRJ9+7d04cffqgqVarIw8NDefLk0eOPP66FCxem6BxXr16ttm3bqkCBAsqVK5fKly+vt99+W9evX5dkm3Vb/P399d5776l27dry9vaWs7OzChUqpGrVqunZZ5/V3LlzFRQUZG3fvHlzmUwmXbp0SZI0b9486zWxPJIa2WeL+9HFixf17rvvqk6dOvLx8ZGzs7Py58+vpk2bauzYsTp//nyqr4HZbNbQoUOt9b/yyiupul9cvHjRuu/cuXMlSb/88otatWqlggULyt3dXRUrVtTIkSOTvU+MHTvW2o8Ue08bN26c9TMet38p5e9/eq5ZZt1XAQAAshQDAAAA2YrZbDZ8fHwMSUbu3LmNkJAQm/UdExNjDBs2zJCU5MPb29tYv359kn2Eh4cbXbt2TbaPokWLGocOHUqyj8uXLxvlypVLcv8OHToY69evtz7fvHlzgj769etnSDJKliyZ5HH+/PNP67VM6vHUU08ZwcHBqbiK8UVHRxs9e/ZMsv/SpUsb58+ftz4fM2ZMgj42b96cbI2Wx3vvvZdkHXPmzLG2u3DhQoLtSV2vO3fuGK6uroYkY/Dgwcme661btwxnZ2dDkjFs2LCUXJ4ELly4YNSsWfOh5/rge/6w8zMMw1i3bp3h5eWVbL/Dhg0zYmJiEt3/jTfeeGhdPj4+CfaL+/497LN68uRJw8/PL9n6kvPyyy8nuW/hwoWNgwcPGiVLljQkGf369Uu2r6Rs27btoddRkrF69WrrPs2aNXto+2bNmsU7ji3uR4ZhGJ999pn1c5nSYxuGkezvZFRUlNGrVy9rm9GjR6f6Ol64cMG6/5w5c4wBAwYke8/8559/Eu1nzJgx1nanT59O9PMzZ84ca/uUvP9pvWaGkTn3VQAAgKyIacoAAACymWPHjunOnTuSpKZNm8rDw8Nmfb/33nuaPn26JKlUqVJ69913Vbt2bYWGhmrVqlX6+uuvFRgYqA4dOmjv3r2qUaNGgj769eun5cuXS5Jq1Kiht956S5UqVVJAQIAWLVqkuXPnyt/fXy1bttSRI0dUrFixePtHR0erffv2OnPmjCSpffv2GjRokHx9fXXlyhV9//33WrNmTYr/Ij4pO3fuVLt27RQdHa1ChQrp1VdfVY0aNVS0aFH5+/tr8eLFWrBggX7//Xf169dPy5YtS9NxRowYoV9++UWSVKFCBb3zzjuqXr26AgMD9csvv2jmzJl65plnku3j/v378vDwUPv27fXEE0+oYsWK8vLy0s2bN3X8+HF9+eWXunTpkiZOnKjy5cvrhRdeSFOticmXL5+6deumn3/+WYsWLdLUqVPl7u6eaNuFCxdaR6wMGDAg1ce6ceOGGjduLH9/f0nSE088oX79+qlixYoymUy6cOGCNm3aZL2eqXH48GF17NhRUVFRcnZ21iuvvKJOnTrJw8NDhw4d0sSJE3XhwgVNnz5dHh4e+vTTT+Ptv2bNGk2bNk2SVL16dQ0dOlSVKlWSt7e37t27p+PHj+vPP//U3r17U12bRVhYmDp27Kg7d+5o9OjRatWqlTw9PXXo0CF99NFHunLliqZPn66OHTuqTZs2CfafNGmSvvnmG0lS8eLF9d5776lu3bqKjIzUunXrNGXKFPXo0UNhYWFprjEyMlK9evVSUFCQcufOraFDh6pFixYqWLCgoqKidOHCBf3111/We4DFnDlzFBoaqjZt2sjf31+dO3fW+PHj47V58F5mi/vRuHHj9OGHH0qKHRX48ssvq0WLFvLx8dG9e/d08OBB/frrr9ZRJSkRFham7t27a+3atTKZTJoyZYreeOONFO+fmG+++Ub79u1TvXr1NHz4cJUrV043b97U3LlztWTJEvn7+6tNmzY6duyYcufOnWQ/PXr00NWrV/Xqq6+qU6dOyps3r86cOaOSJUumuJb0XLPMuq8CAABkSfZOgwAAAGBbCxYsSNdfYyflyJEjhoODgyHJqFq1qnH37t0Ebf744w9rm3r16iXYvmbNGmttLVu2NCIjIxO0+f77761tnn766QTbp02bZt3+xhtvJFrrK6+8kuwoCcNIfmRMVFSU9a/H27Zta4SGhiZ6nLi1Puyv7xMT95rWrl070b8EnzdvXrxzSeyv8G/dupXo+2ERGRlpPPnkk9bzvX//foI2aR0ZYxiGsXHjRuu+CxcuTLKOGjVqGJKMGjVqJNkmOXFHVH366adJtgsODjYCAgLivfaw83vssccMSYajo6Oxbt26BNsDAgKMypUrG5IMBwcH49ixY/G2P//889brk9xf9N+5cyfBaykdGaP/H+nx4LENwzDOnDljuLm5GZKMTp06Jdh+7do16/ayZcsat27dStBm586dhouLi/VYaRkZE/ezEHfky4Oio6ONwMDABK+ndFSOLe5HBw8etG4vX768cfny5SSP9++//yZ4LbHfybt37xqNGze2fpbijjhJrbgjY/T/o0Wio6MTtPv444+tbd5+++0E2+OOjHFwcEj08x1Xcu9Beq5ZZt1XAQAAsirWjAEAAMhmbt++bf25QIECSbYzm806duxYko8H19z49ttvZTabJUk//PCD8uTJk6DPtm3bWkc87N27V/v27Yu33fJX7M7OzpozZ45cXFwS9DFo0CC1atVKkvTrr7/q2rVr8bZ/9913kiRfX19NnDgx0XObNGmSihYtmuS5P8yiRYt08eJFubm5af78+cqVK1ei7QYNGqR69epJUrw1F1Lqu+++s17T77//Xp6engna9O3bV+3atUu2n/z58yf6fli4uLjos88+kyRdunRJhw8fTnWtyWnRooXKlCkjKXaEQ2IOHjyov//+W1LaRsWcOnVKK1askCR16dJF77zzTpJtPT09lTdv3hT3HfezOmjQILVu3TpBm7x58+r777+XFPu7YxlhYmFZb6V27dqJvo8W+fLlS3FdiRk3bpyqVKmS4PWyZcuqS5cukpTo2jbz5s2zrtEzbdo05c+fP0GbRo0aadiwYemqz3IdJOnxxx9Psp2Tk5O8vLzSfBxb3I8+++wzmc1mmUwmLVq0SL6+vkker3jx4g+t6caNG2revLl27twpV1dXLV26NF3r7sTl6uqqmTNnyskp4eQW77//vqpWrSpJmjVrlqKiopLsp3///ol+vlMqPdcss+6rAAAAWRVhDAAAQDYTHBxs/Tm5KcqCgoJUrVq1JB9Xr16N1/7PP/+UJFWpUkX169dPst9BgwYl2EeKnUpr69atkqTWrVsn++WmpY/79+/HW9D86tWrOnnypCSpZ8+ecnV1TXR/d3d39ezZM8n+H2bVqlWSpGbNmiUbaEn/feG8a9euVB/Hcn2qVaumOnXqJNkuteFFZGSk/v33X504ccIarhlxFsS2hCK2YjKZrDVu2rRJ//77b4I2lpDGxcVFvXv3TvUxfvvtN+s5DB8+PB3VJhT3czpw4MAk2zVu3FiVKlVKsI8kFSlSRJK0bds2nTt3zqb1WZhMJj333HNJbrd8hgICAhIs6G6pN3/+/MmGe3379k1XjZbrICUdzNlCeu9HZrNZf/zxhySpefPmqlWrVrrquXjxopo0aaK///5bnp6e+v33363hmC20bt06yYDZwcFB/fr1kxT73h88eDDJftLyu2eR3muWWfdVAACArIowBgAAIJuJu15AaGioTfqMjIy0rtGS3BefklSrVi05OztLil2/xuL8+fPWtSge1kfc7XH7iPtzcuGFJNWtWzfZ7cnZv3+/JGndunUymUzJPiZPniwp/oiAlIh7TR977LFk21r+Sjw5oaGhmjBhgmrUqCEPDw+VLFlSVapUsYZrcb84jTt6ylb69+8vR0dHmc1mzZs3L962yMhI/fTTT5Kkzp07y8fHJ9X9Hzp0SFLsqKoGDRqkv+A4LJ8rFxcX1axZM9m2ls/mmTNn4o1AsIQYd+7cUdWqVdWrVy/NmTNHZ8+etVmd+fPnT/baxR11EzeUlf47x5o1a8rBIel/BlarVi3REWsp1aRJE5UuXVqS9MYbb6hevXqaMGGCdu7cmeyIjdSwxf3owoUL1sCqadOm6arnn3/+UePGjXX27Fn5+Pho48aNeuKJJ9LV54NSc484evRoku2qV6+e5hrSe80y474KAACQlRHGAAAAZDNxv6xNbhH7PHnyyDCMeA/LX1c/6O7du9afCxYsmOzxnZ2drTUEBARYX4/788P6KFy4cKL7xa3jYX9Z/bDtybl582aq9wkPD09V+7t371pHejzsehQqVCjZ7RcvXlS1atU0atQoHTlyRDExMTatNSWKFi2qp556SlLs1EJxR+KsXLnS+j6mZYoy6b8AKV++fOkKCxJjqS1fvnyJTgMVl+WzaRhGvM9jy5Yt9fXXX8vd3V0RERFavHixBgwYoHLlysnX11dDhgxJ94ikpKZ1sogbsjz4GbDU+rDfC0dHx3RNpebs7KzVq1dbRxDt27dPo0aNUpMmTZQnTx61bdtWP/3000M/o8mxxf0obiAZdzRPWixZskT+/v6SYqdPS0l4mlqpuUfEPc8HpWb6vgel95plxn0VAAAgK0v+XxoAAAB45NSoUcP6s2U0gS2ZTKYs0UdGsnxR3K5dO02aNCnDj5fe6/H888/rwoULMplMeuGFF9SrVy9VqlRJBQoUkIuLi0wmk8xmsxwdHSUpXlBiSy+++KJWr16t8+fPa9u2bWrWrJmk/6ar8vX1Tdd6FRktve/DsGHD1LNnT/3000/asGGDdu7cqcDAQF29elUzZszQ999/r1GjRmn8+PE2qjhrqly5so4eParVq1dr9erV2rZtm86ePavw8HCtW7dO69at05QpU/T7778/NGR4mKxwL2nTpo127Nih0NBQvfLKK6pSpYoqV65s02PY6jwt9wB7yOz7KgAAQFZDGAMAAJDNVK1aVT4+Prpz5462b9+usLCwh/5F/cPE/WvqGzduJNv2/v37unPnjqT40ybF/flhfcSdmibufnHrSG7UT0q2J8fHx0f+/v6KioqyLoxta3EXHH/Y9Uhu+8mTJ60Ltif3RX9yfy1vK+3bt1eRIkV07do1zZkzR82aNdPVq1e1fv16SVK/fv2SnSIrOZYF5wMCAhQVFWXT0TGWz9idO3d0//79ZEfHWD6bJpMp0VEGBQsW1BtvvKE33nhDZrNZhw8f1vLly/X111/r3r17+uSTT/TYY4+pc+fONqs/JfLmzavr168/9PciJiYm3siTtHJ0dFSXLl2s66Zcu3ZNa9eu1fTp03XgwAEdOHBAgwcP1vLly1Pdty3uR5bPk6W29GjQoIFGjhypp556Sjdv3lTLli21ZcsWVahQIV39xpWae0R6RjYlJ73XLDPuqwAAAFkZ05QBAABkMyaTSX369JEkBQUFJVi/Iy1cXV1Vrlw5SdKePXuSbXvo0CFFR0dLUrwv3EqXLm0NhR7Wx969e60/x+2jSpUq1p8PHDiQbB+W9QnSwrK+yv79+222zsWD3NzcrNd03759ybZNbvvx48etPz/zzDNJtkvP9UgpR0dH9e/fX5K0dOlShYSEaN68eTKbzdZRO2lVu3ZtSVJ0dLTNF/W2fMaioqJ0+PDhZNtaPpvlypV7aCDk4OCg2rVra9y4cdq4caP19SVLlqSv4DSw/O4cPnxYZrM5yXZHjx5VZGSkzY9fpEgRvfDCC9q1a5f1vVyzZk2CaahSMgLEFvejUqVKWQPRbdu2pfg8ktKsWTOtXr1a7u7uun79ulq0aGFd18YWUnOPyKigI73XLDPuqwAAAFkZYQwAAEA29Oabb8rNzU2SNHLkSF24cCHdfbZq1UpS7Jf/ccOSB/3www8J9pEkJycn67RVGzZs0JUrVx7ah5OTk5o3b2593dfXV+XLl5ck/fLLL0l+aRwREaFffvnlIWeUtE6dOkmSAgMDrVNsZQTL9Tl69GiyU8rNnj07yW3379+3/hwaGppku++++y4NFabewIEDZTKZFBoaqsWLF2vu3LmSpMcff1xlypRJc7/t27e3flE/bdo0G1T6n7if0+Su9a5du3TixIkE+6RE7dq1rSM64q69kVlatmxpPfYff/yRZLv58+dnaB3Ozs7W+8D9+/etC8JbWO5bDwuE0ns/cnBwUPv27SVJW7dutcmUjk888YRWrlwpNzc3Xbt2TS1atNC5c+fS3a8krV+/PsnRKGaz2Rq6582b1xp22Vp6r1lm3VcBAACyKsIYAACAbKhEiRL68ssvJcV+8dWkSRPrVFZJMQwjwRejcQ0dOtQ6xdRLL72koKCgBG3Wr1+vWbNmSZLq1aunxx57LN72YcOGSYodgTBw4EDrX6zHNXv2bOu0Vt26dUuwUPTgwYMlSVeuXNF7772XaK1vv/22dUHttOjXr5+KFy8uSRoxYsRD/wp8x44d2rp1a6qPM3jwYGvA8NJLLyUapixcuFC///57kn1YRghIsgYfD/r222+1cuXKVNeXFmXKlLEGaKNHj7aODhgwYEC6+i1fvry6du0qSVqxYoU+++yzJNuGhoamaqqtevXqqW7dupKkmTNnxhvFYhEYGGj97Dk4OGjo0KHxti9evDjZxcb3799vralUqVIprs1W+vXrJ1dXV0nSG2+8kWggtGvXLk2fPj1dx9m+fbvOnj2b5PaoqCjr74qnp6cKFCgQb7vl9/1hIYYt7kcjRoyQg4ODDMNQr169kg2Ik9sW15NPPqkVK1bI1dVVV69eVYsWLXT+/PkU7ZucyMhIDR482LruSlwTJ07U0aNHJcX+nlne54yQnmuWWfdVAACArIowBgAAIJsaNGiQxowZI0ny9/dX06ZN1bJlS02bNk0bN27UoUOHtH//fq1Zs0Yff/yxqlWrZv3C3tHRMcEUTNWqVdNbb70lSfr7779Vu3ZtzZw5U/v379fWrVs1YsQIdejQQTExMXJxcdGMGTMS1NS+fXv17NlTUuwXpQ0aNNDChQt14MAB/fnnn3rxxRf14osvSopd92DKlCkJ+njllVes0/BMmzZNHTt21KpVq3Tw4EGtWrVKHTp00Ndff6169epZ90nt4teurq5asmSJXF1dFRISoieeeEJ9+vTR0qVLdeDAAe3bt0+rVq3SmDFjVL16dTVt2tT6ZWhq1KhRwxpQ7d+/X3Xr1tXcuXN14MABbdq0SUOHDlXfvn2tQUFiatWqZb0eM2bM0DPPPKM1a9bowIEDWrlypXr27KmXX35ZjRs3TnV9aWV5Dy3rq3h5ealHjx7p7vebb75R0aJFJUnvvPOOWrZsqR9//FH79u3T/v37tXTpUg0bNkwlSpTQ33//naq+Z86cKRcXF92/f19PPfWURowYoa1bt2r//v2aOXOmateubX2PR4wYkWAqqHfffVdFixZV//79NXv2bO3YsUOHDh3Sn3/+qbFjx6pNmzaSYn+3LNcnMxUtWtR6Pzh79qzq1Kmjb7/9Vvv27dOOHTv0wQcfqGXLlipatKg1IEnLovEbN25UhQoV1Lx5c3322Wdat26dDh48qJ07d2rOnDlq2rSpDh48KCl2FNWD6/M0atRIUuy0WxMnTtTff/+ts2fP6uzZs7p69aq1nS3uRzVr1tRHH30kSTp9+rSqVaum0aNHa+PGjTp8+LC2bNmiadOm6fHHH9fzzz+f4mvQpk0b/frrr3JxcdHly5f1xBNP6NKlS6m7kA+oW7euVq9ercaNG2vx4sU6ePCg1q5dq2effVbvv/++pNiRgx988EG6jvMw6blmmXVfBQAAyLIMAAAAZGu//vqrUbp0aUPSQx8mk8lo27atcfTo0UT7iomJMV5++eVk+/D29jbWrVuXZD3h4eFG165dk+2jaNGixqFDh5Ls49KlS0aZMmWS3L9169bGH3/8YX2+e/fuBH3069fPkGSULFkyyePs2rXLKF68eIqu3bx585LsJzlRUVFGt27dkuy3VKlSxrlz56zPx4wZk6CPQ4cOGXnz5k2yj2rVqhn+/v7J9jFnzhzr9gsXLqTpelmEh4fHq2fQoEGpvzBJOHfunFG1atWHvh+bN2+Ot9/Dzs8wDGPdunWGl5dXsv0OGzbMiImJSbBvyZIlH1qTq6urMWfOnAT7bt68Ocm6DSPl1/5h52g2m43BgwcnWV/+/PmNffv2WT/zQ4YMSfZ4iRkzZkyKfl86d+5shIWFJdj/ypUrRr58+RLdp1mzZvHa2uJ+ZBiG8cknnxhOTk7J9vPgsQ3DSPb3yTAMY9WqVYazs7P19/jff/9N6WU0DMMwLly4YD3GnDlzjP79+ydZX5EiRYzjx48n2k/c9yQlLJ/lfv36JdkmrdfMMDLnvgoAAJAVMTIGAAAgm+vatatOnTqlJUuWaODAgapcubLy588vJycneXl5qVSpUurUqZMmTJigc+fO6Y8//khyAWgHBwdNnz5d27ZtU+/evVWiRAm5urrKy8tLNWvW1KhRo3TmzBm1bt06yXrc3Nz066+/atWqVerWrZuKFi0qFxcX5c2bV/Xr19eECRN06tQp1axZM8k+LCMfPvroI1WtWlXu7u7KkyePGjRooG+++UZ//PGHIiIirO29vb3TdO0aNGigM2fO6LvvvlP79u2ttbq5ual48eJq3bq1PvnkE508eVJ9+/ZN0zGcnZ21bNky/fjjj2ratKm8vb2VK1cuVapUSaNGjdKBAwdUunTpZPuoWbOmDh8+rCFDhqhkyZJydnZWvnz5VK9ePU2ePFl79+5NMN1bRnJzc7OOgJLSP0VZXKVLl9bhw4c1d+5ctW/fXkWKFJGzs7Pc3d1Vvnx59e3bVytXrlTTpk1T3Xfr1q119uxZjRo1SjVr1pSXl5dcXV1VokQJ9e7dW9u3b9fXX39tnR4rrs2bN+uLL75Q9+7dVa1aNRUoUMD6O1arVi2NGDFCJ06cUP/+/W1wFdLGZDLpu+++08qVK9W6dWvly5dPbm5uKlu2rF577TUdOnRIdevWtU75lZbfmxEjRmjZsmUaOnSoGjRooBIlSsjNzU1ubm7y8/PT008/rTVr1mjFihVyd3dPsH+xYsW0d+9eDRw4UGXLlrWuIZMYW9yPJGnUqFE6ceKE3njjDVWtWlVeXl5ycnJSgQIF1KxZM40fP14//vhjqq9Fx44dtWTJEjk7O+vChQtq0aJFiqc7S8ycOXP0008/qXnz5vLx8ZGrq6vKly+vd955R8ePH1flypXT3HdqpeeaZcZ9FQAAICsyGYZh2LsIAAAAwNbGjx+vDz74QE5OTgoODk72S13YXuPGjfXXX3+pcuXKOn78uL3LQQpduXLFuq7HDz/8oIEDB9q5opzr4sWL1vWF5syZY9cgDwAAAOnHyBgAAABkO4ZhaPHixZJiR40QxGSuU6dO6a+//pJk21ExyHg///yz9ecGDRrYsRIAAAAgeyGMAQAAwCPn4sWLun//fpLbP/zwQx07dkyS1K9fv8wqC//v008/lRQ7XRl/zZ91hIaG6tq1a0luP3TokMaNGydJqlOnjqpUqZJZpQEAAADZnpO9CwAAAABSa+7cuZozZ46ee+45NW7cWEWLFlV0dLT++ecfzZs3T1u2bJEkVa5cWYMGDbJvsTlAeHi4rl69qrCwMK1YsUJz586VJL300kvy8fGxb3GwunXrlipVqqQuXbqobdu2qlChglxdXeXv76+1a9dq1qxZCg8Pl8lk0pQpU+xdLgAAAJCtEMYAAADgkfTvv/9q4sSJSW6vWLGifvvtN7m6umZiVTnTnj171KJFi3ivFS9eXGPHjrVPQUhSRESEFi1apEWLFiW63cXFRTNnztTjjz+eyZUBAAAA2RthDAAAAB45AwcOlLe3t9avX6+zZ8/q1q1bCgsLU758+VSjRg117dpVAwYMkIuLi71LzVFMJpOKFCmiJ554Qp988ony5s1r75IQR7FixbR48WKtXbtW+/bt061btxQQEKBcuXLJz89PrVq10quvvqqSJUvau1QAAAAg2zEZhmHYuwgAAAAAAAAAAIDsipExqWA2m+Xv76/cuXPLZDLZuxwAAAAAAAAAAGBHhmEoODhYRYsWlYODQ5LtCGNSwd/fX8WLF7d3GQAAAAAAAAAAIAu5fPmyfH19k9xOGJMKuXPnlhR7Ub28vOxcDQAAAAAAAAAAsKegoCAVL17cmh8khTAmFSxTk3l5eRHGAAAAAAAAAAAASXro0iZJT2AGAAAAAAAAAACAdCOMAQAAAAAAAAAAyECEMQAAAAAAAAAAABmIMAYAAAAAAAAAACADEcYAAAAAAAAAAABkIMIYAAAAAAAAAACADORk7wIAAAAAAAAAIDnR0dGKiYmxdxkAsjlHR0c5OztnSN+EMQAAAAAAAACypKCgIN2+fVuRkZH2LgVADuHq6qr8+fPLy8vLpv0SxgAAAAAAAADIcoKCgnT16lV5enoqf/78cnZ2lslksndZALIpwzAUHR2twMBAXb16VZJsGsgQxgAAAAAAAADIcm7fvi1PT0/5+voSwgDIFO7u7sqdO7euXLmi27dv2zSMcbBZTwAAAAAAAABgA9HR0YqMjJS3tzdBDIBMZTKZ5O3trcjISEVHR9usX8IYAAAAAAAAAFlKTEyMJGXYQtoAkBzLvcdyL7IFwhgAAAAAAAAAWRKjYgDYQ0bcewhjAAAAAAAAAAAAMhBhDAAAAAAAAAAAQAYijAEAAAAAAAAAAMhAhDEAAAAAAAAAHjlhYdLdu1n/ERZm7yuVs82dO1cmk0ljx461dynpYjKZ5OfnZ+8yUswW9QYGBmrUqFGqUqWKcuXKJTc3N1WoUEHDhw/XzZs3k9xv586deuqpp5QvXz55enqqXr16mj9/frpqsQUnexcAAAAAAAAAAKkRFiYtXCgFBNi7kofLl0/q3VvKlcvelTzcli1b1KJFC/Xr109z5861dznIwW7fvq1GjRrpzJkzKly4sJ588klJ0t69ezVt2jQtXrxYu3btUsmSJePtt2zZMj3zzDMym816/PHHlT9/fm3cuFH9+vXTkSNHNHnyZHucjiTCGAAAAAAAAACPmMjI2CDG3T32kVWFh8fWGRn5aIQx2VHXrl3VoEED5c+f396lIBX+97//6cyZM+rUqZMWL14sNzc3SVJERISee+45LV++XB9++KHmzZtn3ScgIEADBgxQTEyMli1bpm7dukmSbty4oSZNmujzzz9Xhw4d1Lx5c3ucEmEMAAAAAAAAgEeTu7vk6WnvKpIXHm7vCnI2b29veXt727sMpNK2bdskSSNHjrQGMZLk5uamDz74QMuXL9e+ffvi7fPDDz8oKChInTt3tgYxklSoUCFNmjRJ3bp10+eff263MIY1YwAAAAAAAADgEfDbb79pwIABqlSpkry8vOTh4aEaNWrof//7nyIjI+O1HTt2rEwmU5LTjfn5+clkMlmf9+/fXy1atJAkzZs3TyaTyfp4cL2VXbt2qXPnzipQoIBcXV3l5+enl19+Wf7+/uk6v8OHD+udd95RnTp1rH2XLl06yb4vXrwok8mk5s2bKygoSG+++aZKlSolZ2dnvfHGG5KSXjOmf//+MplM2rJli/788089/vjjyp07twoWLKhBgwYpMDBQknTz5k0NHjxYxYoVk5ubm+rVq6ctW7YkqCUiIkKzZs1S586dVbp0abm7uytPnjx6/PHHtWjRonRdl4f5559/1L9/fxUvXlyurq4qVKiQevXqpePHj8drN2XKFJlMJr377rtJ9tW9e3eZTCatWrUq3usBAQEaOXKkKleuLHd3d3l7e+uJJ57QmjVrMuScXF1dH9rGx8cn3vPffvtNktSjR48Ebdu3by83Nzf9+eefioiIsE2RqUQYAwAAAAAAAACPgIEDB2rZsmXKly+f2rVrp6ZNm+ry5ct6//339dRTTykmJibNfTdp0kRt2rSRJJUpU0b9+vWzPmrWrGltt2DBAjVt2lSrVq1ShQoV1K1bN7m6uurbb79V7dq1dfLkyTTXMHHiRE2dOtVaz1NPPSXDMPTtt9+qbt26SYY94eHhatasmebOnauaNWuqU6dOyps3b4qOuXz5crVt21aGYaht27ZydXXVDz/8oM6dO+v27dtq2LCh1q1bp6ZNm6pmzZrat2+f2rZtq6NHj8br5+LFi3rxxRe1f/9++fn5qXPnzqpZs6Z2796tZ599NkEYZCsrVqxQrVq1NG/ePOXPn1+dOnVSqVKltGTJEtWrV886wkSSevXqJQcHBy1atEiGYSToKzAwUL/99pt8fHzUrl076+unT59WzZo1NXHiRIWFhatNmzaqW7eu9uzZo44dO2bIOiytW7eWFPuZiBueREREaNy4cZJifx/i+vvvvyVJtWvXTtCfi4uLqlatqoiICJ0+fdrm9aYE05QBAAAAAAAAwCNgxowZat26tdzjLJQTHBys5557TmvWrNHChQvVt2/fNPX94osvqmzZslq3bp2aNGmS6Iiay5cv66WXXpIkrVy5Up06dZIkmc1mvfXWW5o2bZqef/75BNNHpdTgwYP1xRdfqFChQtbXzGazxo8frzFjxmj06NGaPXt2gv327t2rhg0b6vz588qTJ0+qjvn1119r1apVat++vaTY69moUSNt3bpVzZo1U6NGjTR79mw5OztLkj744AONHz9ekydPjrdeSYECBbRhwwa1bNky3oijCxcu6IknntC4cePUv39/+fn5paq+5Fy8eFF9+vSRs7Oz1qxZo1atWlm3rV27Vp06dVKfPn109uxZubi4qGjRomrRooU2btyoHTt2qGnTpvH6W7p0qSIjI/XCCy9YzzcqKkbduvXQ5cuXNWrUJL377lvy8ood43H27Fm1bt1a7733ntq2bauqVava7NxGjBihrVu3auXKlSpVqpTq168vSdqzZ48iIiI0efJk9e/f39o+KCjIOprJ19c30T59fX21f/9+Xbp0SdWrV7dZrSnFyBgAAAAAAAAAeAR07tw5XhAjSblz57aOJlm5cmWGHv+HH35QeHi4nn76aWsQI0kODg6aOHGiihYtqv3792vnzp1p6r9FixbxghhL3x9++KGKFSuWYOqsuL788stUBzGS9Nxzz1mDGCn2eg4aNEiSdOXKFX355ZfWYEKKDQlMJpO2bt0arx8fHx+1atUqXhAjSaVKldL7778vs9ms1atXp7q+5EybNk2hoaGaMGFCvCBGktq2bauhQ4fq8uXL1um7JKlPnz6SpIULFyboz/Ja7969ZRhSWJj088+rdfz4UbVv310vv/y2HBz+ixTKli2rzz//XDExMZo5c6ZNz83Dw0O//fab+vTpo+vXr2vlypVauXKlrl+/rlq1aqlJkybx2oeEhFh/zpUrV5J9SrGBmz0wMgYAAAAAAAAAHhFnzpzR77//rrNnzyo0NFRms9k65dSZM2cy9Njbt2+XFPtl/YNcXV3Vs2dPffHFF9q+fbsaN26cpmPcuXNHq1at0rFjx3Tv3j3r1GvR0dG6c+eOAgIClC9fvnj7FClSRHXr1k3T8SzTYcVVunRpSVLdunUTTHfm7e2tfPny6dq1a4n2t2PHDm3ZskVXr15VRESEDMOwtrX1+7N+/XpJirdYfVxNmzbVl19+qb1796pr167WtkOHDtXSpUv11VdfWYOmq1evauvWrfLz89NjjzXW3btSRIS0eXPsMTp2TPoYUuzoJFv6999/1b59e127dk3z589X27ZtJUl//PGHhg8frubNm2v9+vUJRvdkZYQxAAAAAAAAAJDFGYahESNGaOrUqYmu9yFl/F/8W9ZsSWqqLcvrV69eTVP/P//8s1566aV4oxweFBwcnCCMKVGiRJqOJ0nFihVL8Jqnp2eS2yzb79y5E++1wMBAdevWTZs2bUryWLZ+fy5evCgp6Totbt++bf3Zy8tLHTt21C+//KK1a9eqY8eOkmKvvdlsVvfuzykgwCTDkFxcpKtXY48xZEhvSQlDuMSOYQv9+vXTsWPHtHz5cnXp0sX6et++feXp6anu3btrxIgR2rNnj6T/3jNJCgsLk5eXV4I+Q0NDJcWOfrIHwhgAAAAAAAAAyOIWL16sKVOmqHjx4po6daoaNmyoAgUKyNnZWVFRUXJ1dU0ypEmM2Wy2eY0PTtGVGpcuXbKuATJt2jS1b99exYoVs07L1qhRI+3atSvRc3Rzc0vzceNOu5WabQ969913tWnTJjVr1kwfffSRqlatqjx58sjR0VHr169XmzZtUvX+pITlPezXr1+y7SzrrVj06dNHv/zyi3766Sd17NhRhiH9+GPsFGXt2/eWo6Pk6Bj/GC1btlW+fIXk7CzFmbXNKn/+/Ok8m/9cvnxZW7ZskaurqzUsiqtz585ycXHRvn37FBERITc3N3l5ecnb21uBgYG6cuWKKleunGC/K1euSJJKlixps1pTgzAGAAAAAAAAALK45cuXS5K+/fbbeGucSNL58+cTtHdxcZGkREeZxMTE6Pr166muoWjRojp16pQuXbqkKlWqJNie0pEaifn9998VFRWlESNG6PXXX0+wPbFzzEqWL18uR0dHrVq1KsGojIyq3dfXV+fOndPnn38uHx+fFO/Xrl075cuXT6tWrVJAQIhOnvxXR44cVtWqtVS9evwQo0gRX0lSnz4v6sknu8vbW4ozCCVDWEITDw8POVpSoTgcHR3l4eGhu3fv6t69eypcuLAkqUaNGtq2bZsOHjyYIIyJjo7WsWPH5ObmpvLly2fsCSQh5dEeAAAAAAAAAMAu7t69Kyn2C/gHLVmyJMFrRYoUkSSdPn06wbbNmzcrOjo6weuWAOf+/fuJ1mBZn+Pnn39OsC0qKkq//PJLvHapkdz5bdu2TTdu3Eh1n5np7t278vLySnR6rMTeH1t48sknJf0X1KWUs7OzevToqbCwMP300wotWhQ7KqZnz4TTkDVvHnuM335L3THSwxKuBAQE6MKFCwm2nzt3Tnfv3pWHh0e8ETmWkHLp0qUJ9lmzZo0iIiLUqlWrdI2kSg/CGAAAAAAAAADI4ix/zf/999/Hm+5q+/bt+uyzzxK0f/zxxyVJCxYssI5YkaQLFy7otddeS/QYRYsWlSSdOnUq0e0DBw6Uu7u7Fi1apN9++836utls1qhRo3T16lXVqVNHjRs3Tt3J6b/zW7BggXVtDyl2/ZkhQ4akur/MVr58ed29e1eLFy+O9/rUqVO1efPmDDnmW2+9JXd3d40YMUK//vprgu2RkZFaunSpdaSJJBmGFB4udejQR5K0fPlCrVr1sxwcHNSt27MJ+ujYsbsqVKispUsXatq0cYqMjIy33TAM7dy5Uzt37rTZeZUqVUrVq1eXJA0ePFiBgYHWbffu3dPgwYMlSV26dJGT03+Tf7344ovy8vLSypUr412Pmzdv6p133pEUe83shWnKAAAAAAAAADySwsPtXUHybFnfa6+9prlz5+qbb77Rli1bVL16dV29elU7duzQW2+9pcmTJ8drX6ZMGfXt21fz589XzZo19fjjjyssLEy7d+/WU089pbCwMF26dCnePn5+fqpevbr279+vevXqqUqVKnJ0dFSnTp3UqVMnlShRQjNmzFD//v3VsWNHNW7cWMWLF9fBgwd16tQpFSpUSAsWLEjT+XXq1ElVqlTR/v37VbZsWTVu3FgRERHavHmzatasqUaNGumvv/5K8/XLaCNHjlSfPn3Uq1cvTZ8+Xb6+vvr777918uRJDR8+XFOnTrX5McuWLauff/5Zzz33nLp3766yZcuqUqVK8vDw0NWrV3Xw4EGFhobq0KFD8vX11f37UkiIFBYm1anTWMWLl9SmTWslSY8/3lJFihRNcAwnJyfNn79CTz/dRpMnf6h5875WjRrVVbBgQd2+fVuHDx/WzZs3NXXq1DSFcEn5/vvv1apVK23YsEFly5a1rnuze/du3blzR35+fpo0aVK8ffLly6fZs2fr6aefVo8ePdS8eXP5+Pjozz//1L179/Tmm2+qefPmNqsxtRgZAwAAAAAAAOCR4uoq5csXG3YEBGTdR3h4bJ2uruk/5/Lly2v//v3q2LGjbt++rVWrVikkJEQzZsxIdGSMJM2cOVPvvfeevLy8tG7dOl28eFEjR45MdJoxi2XLlqlLly46f/685s+fr1mzZungwYPW7c8//7y2b9+uDh066J9//tHSpUsVHh6uoUOH6sCBA6pYsWKazs/FxUXbt2/X0KFD5ebmpjVr1uiff/7Rq6++qg0bNsg5sVXjs5DevXvrt99+U4MGDXT48GH98ccfKlq0qDZt2qROnTpl2HE7d+6sI0eO6OWXX5bJZNKGDRv022+/6ebNm+rYsaOWLFmiihUrKzRUunNHCg2VnJ0lV1eTund/ztpPjx4JpyizKFOmnDZvPqR33hmvYsV8tXv3bv366686ffq0atWqpenTp6tPnz42Pa/69evr8OHDGjRokLy8vPTnn39q06ZNKlSokN577z0dOHDAOpIrru7du2vbtm1q06aNDh06pN9//11ly5bV3Llz9fnnn9u0xtQyGXHHtCFZQUFB8vb2VmBgYKJz/wEAAAAAAABIv4iICF24cEGlSpVKcn2HsDDpgRmTsiRXVylXLntXgZwqKkoKDpYiIiRHx9ggJq0iIiRvb8nT03b1ZVUpuQdZpDQ3YJoyAAAAAAAAAI+cXLkIOYCkxMTEBpahoZLZHBsKmkz2ripny/LTlE2cOFEmk0lvvPGGJCkgIECvvvqqKlSoIHd3d5UoUUKvvfZavEV8EtO/f3+ZTKZ4j7Zt22bCGQAAAAAAAAAAkPEMI3YES0CAFBQkOTgQxGQVWXpkzL59+zRjxgxVr17d+pq/v7/8/f01efJkVa5cWZcuXdKQIUPk7++vpUuXJttf27ZtNWfOHOtzV1tM1ggAAAAAAAAAiGfFihVasWJFitq++OKLatKkScYWlIVMnDhRJ0+eTFHbyZMnK3/+/ClqGxMjhYTEjoYxmaSHzK6VaXbs2KEffvghRW27dOmiLl26ZGxBdpJlw5iQkBD17t1bM2fO1Pjx462vV61aVcuWLbM+L1OmjD755BP16dNH9+/fl5NT0qfk6uqqwoULp7iGyMhIRcaZeDIoKCiVZwEAAAAAAAAAOc/hw4c1b968FLVt3rx5jgpj1q5dq61bt6ao7dixYx8axlhGwwQHS9HRsevCODraolLbOHv2bIo/C35+ftk2jMmy05QNGzZM7du3V6tWrR7a1rIwTnJBjCRt2bJFBQsWVIUKFTR06FDduXMn2fYTJkyQt7e39VG8ePFUnQMAAAAAAAAA5ERjx46VYRgpevTv39/e5WaqLVu2pPja+Pn5JdtXTIwUGCjdvRv7s5tb1gpipNglRFJ6vmPHjrV3uRkmS46MWbRokQ4ePKh9+/Y9tO3t27c1btw4vfTSS8m2a9u2rbp166ZSpUrp3LlzGjVqlNq1a6ddu3bJMYlP58iRI/Xmm29anwcFBRHIAAAAAAAAAADs6sHRMC4usevDIOvKcmHM5cuX9frrr2vDhg1ye8ikdkFBQWrfvr0qV6780MSsV69e1p+rVaum6tWrq0yZMtqyZYtatmyZ6D6urq6sKwMAAAAAAAAAyDKy6towSF6Wy8oOHDigmzdvqnbt2nJycpKTk5O2bt2qL7/8Uk5OToqJiZEkBQcHq23btsqdO7eWL18uZ2fnVB2ndOnSyp8/v86ePZsRpwEAAAAAAAAAgE1FREgBAbFhjLNz7IgYPBqy3MiYli1b6ujRo/Fee+GFF1SxYkW9++67cnR0VFBQkNq0aSNXV1etWrXqoSNoEnPlyhXduXNHRYoUsVXpAAAAAAAAAGzIMAx7lwBkCWbzf6NhDIPRMBktI+49WW5kTO7cuVW1atV4Dw8PD/n4+Khq1aoKCgpS69atFRoaqlmzZikoKEjXr1/X9evXraNmJKlixYpavny5JCkkJERvv/22du/erYsXL2rjxo3q3LmzypYtqzZt2tjrVAEAAAAAAAAkwrLGc3R0tJ0rAewvOlq6ezd2fRhHR4mVNTKe5d6T1HrzaZHlRsY8zMGDB7Vnzx5JUtmyZeNtu3Dhgvz8/CRJp06dUmBgoKTYC3bkyBHNmzdP9+7dU9GiRdW6dWuNGzeONWEAAAAAAACALMbZ2Vmurq4KDAxU7ty5ZTKZ7F0SkOkMQwoPjw1h7t+PDWH4Vch4hmEoMDBQrq6uqV4eJTkmg7F+KRYUFCRvb28FBgbKy8vL3uUAAAAAAAAA2VZQUJCuXr0qT09PeXt7y9nZmVAGOUZMjBQWFhvGmEySUxYYVhEVJXl6Sh4e9q4kYxiGoejoaAUGBiokJETFihVLUQ6Q0twgC7yFAAAAAAAAABCf5UvN27dv6+rVq3auBsg89+9LERGx/3V0lByyyGIj0dGSu3v2nybN1dU1xUFMahDGAAAAAAAAAMiSvLy85OXlpejo6HjrRQPZkdks/fOPtHevFBoqFS8eG8ZkFefPS02bShUr2ruSjOPo6GjTqcniIowBAAAAAAAAkKU5Oztn2BekQFYQHCxt2SIdOCB5eUnFisWGM2azvSv7T0hI7Do2bm72ruTRRBgDAAAAAAAAAICdXLokbdwYO/KkZMnYdVmQ/RDGAAAAAAAAAACQyWJipP37pW3bpPBwqUIFyYlv7LMt3loAAAAAAAAAADJRYKC0ebN06JCUN6/k62vvipDRCGMAAAAAAAAAAMgk587FTkv2779SqVJSrlz2rgiZgTAGAAAAAAAAAIAMFhMj7dkTOy1ZdLRUqZLk4GDvqpBZCGMAAAAAAAAAAMhAoaHSpk3Svn1S/vxSgQL2rgiZjTAGAAAAAAAAAIAMcuOGtG6ddPq05OcneXjYuyLYA2EMAAAAAAAAAAAZ4PTp2CDm1i2pfHnJ2dneFcFeCGMAAAAAAAAAALAhs1navz92ajLDkCpWlEwme1cFeyKMAQAAAAAAAADARiIjpa1bpb/+kvLkkQoVsndFyAoIYwAAAAAAAAAAsIGgIGn9eunwYal4ccnLy94VIasgjAEAAAAAAAAAIJ1u3pR+/106e1YqU0Zyc7N3RchKCGMAAAAAAAAAAEiHixdjg5jr16UKFSQnvnnHA/hIAAAAAAAAAACQRsePS2vXSuHhsUGMg4O9K0JWRBgDAAAAAAAAAEAqGYZ04IC0YYPk6Bg7NRmQFMIYAAAAAAAAAABSISZG2rlT2rJF8vKSChWyd0XI6ghjAAAAAAAAAABIoehoafPm2DCmYEEpXz57V4RHAWEMAAAAAAAAAAApEB0t/fmn9Ndfkq9v7KgYICUIYwAAAAAAAAAAeIi4QUzx4lLu3PauCI8SwhgAAAAAAAAAAJIRN4gpUULy9LR3RXjUEMYAAAAAAAAAAJCE6GhpwwZp1y6CGKQdYQwAAAAAAAAAAImIiooNYnbvJohB+hDGAAAAAAAAAADwgKgoaf16ac8eghikn4O9CwAAAAAAAAAAICshiIGtEcYAAAAAAAAAAPD/LEEMU5PBlpimDAAAAAAAAAAAxQYx69ZJe/dKfn6Sh4e9K0J2wcgYAAAAAAAAAECOFzeIKVmSIAa2RRgDAAAAAAAAAMjRoqP/WyOGIAYZgTAGAAAAAAAAAJBjmc3S1q2xQQxTkyGjEMYAAAAAAAAAAHKs3bulHTukYsUIYpBxCGMAAAAAAAAAADnS339LmzZJPj6St7e9q0F2RhgDAAAAAAAAAMhxzp6NXSfG3V3Kn9/e1SC7I4wBAAAAAAAAAOQoN25Ia9dKUVFS0aL2rgY5AWEMAAAAAAAAACDHCA6W/vhDunlTKlXK3tUgpyCMAQAAAAAAAADkCFFR0oYN0pkzUrlykslk74qQUxDGAAAAAAAAAACyPcOQtm+XDh2SypSRnJzsXRFyEsIYAAAAAAAAAEC2d+yY9NdfsWvEuLvbuxrkNIQxAAAAAAAAAIBs7fp1aeNGyc1NypPH3tUgJyKMAQAAAAAAAABkW2FhsevE3L0rFStm72qQUxHGAAAAAAAAAACyJbNZ2rpVOnUqdp0Yk8neFSGnIowBAAAAAAAAAGRLR45Ie/ZIxYtLzs72rgY5GWEMAAAAAAAAACDbuXVL2rJF8vCQvLzsXQ1yOsIYAAAAAAAAAEC2cv9+bBATEMA6McgaCGMAAAAAAAAAANnKoUOxU5T5+bFODLIGwhgAAAAAAAAAQLZx7Zq0bZuUN6/k7m7vaoBYhDEAAAAAAAAAgGwhOlravFkKCpKKFLF3NcB/CGMAAAAAAAAAANnCvXvSpUuSr6+9KwHiI4wBAAAAAAAAAGQrDnzzjSyGjyQAAAAAAAAAAEAGIowBAAAAAAAAAADIQIQxAAAAAAAAAAAAGYgwBgAAAAAAAAAAIAMRxgAAAAAAAAAAAGSgLB/GTJw4USaTSW+88Yb1tYiICA0bNkw+Pj7y9PRU9+7ddePGjWT7MQxDH374oYoUKSJ3d3e1atVKZ86cyeDqAQAAAAAAAABATpelw5h9+/ZpxowZql69erzXhw8frtWrV+uXX37R1q1b5e/vr27duiXb16RJk/Tll1/qu+++0549e+Th4aE2bdooIiIiI08BAAAAAAAAAADkcFk2jAkJCVHv3r01c+ZM5c2b1/p6YGCgZs2apSlTpuiJJ55QnTp1NGfOHP3111/avXt3on0ZhqFp06Zp9OjR6ty5s6pXr6758+fL399fK1asyKQzAgAAAAAAAAAAOVGWDWOGDRum9u3bq1WrVvFeP3DggKKjo+O9XrFiRZUoUUK7du1KtK8LFy7o+vXr8fbx9vZW/fr1k9xHkiIjIxUUFBTvAQAAAAAAAAAAkBpO9i4gMYsWLdLBgwe1b9++BNuuX78uFxcX5cmTJ97rhQoV0vXr1xPtz/J6oUKFUryPJE2YMEEfffRRKqsHAAAAAAAAAAD4T5YbGXP58mW9/vrrWrhwodzc3Oxay8iRIxUYGGh9XL582a71AAAAAAAAAACAR0+WC2MOHDigmzdvqnbt2nJycpKTk5O2bt2qL7/8Uk5OTipUqJCioqJ07969ePvduHFDhQsXTrRPy+s3btxI8T6S5OrqKi8vr3gPAAAAAAAAAACA1MhyYUzLli119OhRHT582PqoW7euevfubf3Z2dlZGzdutO5z6tQp/fvvv2rYsGGifZYqVUqFCxeOt09QUJD27NmT5D4AAAAAAAAAAAC2kOXWjMmdO7eqVq0a7zUPDw/5+PhYXx84cKDefPNN5cuXT15eXnr11VfVsGFDNWjQwLpPxYoVNWHCBHXt2lUmk0lvvPGGxo8fr3LlyqlUqVL64IMPVLRoUXXp0iUzTw8AAAAAAAAAAOQwWS6MSYmpU6fKwcFB3bt3V2RkpNq0aaNvvvkmXptTp04pMDDQ+vydd95RaGioXnrpJd27d09NmjTR2rVr7b4uDQAAAAAAAAAAyN5MhmEY9i7iUREUFCRvb28FBgayfgwAAAAAAAAApNHdu3d15MgRPf744zKZTDbr99Yt6YcfpIIFpVy5bNYtJP3zj9S2rdSokb0ryVpSmhtkuTVjAAAAAAAAAADZ27Bhw9S8eXONHz/e3qUAmYIwBgAAAAAAAACQaQzD0Pr16yVJY8eO1fbt2+1cEZDxCGMAAAAAAAAAAJnm9OnTunPnjiTJbDbrueeesz4HsivCGAAAAAAAAABAptm5c6ckqW7duipfvryuXLmiAQMGiOXNkZ0RxgAAAAAAAAAAMs1ff/0lSWrVqpUWL14sFxcXrVq1Sl9//bWdKwMyDmEMAAAAAAAAACDTWEbGNG7cWDVr1tTnn38uSRo1apSioqLsWRqQYQhjAAAAAAAAAACZ4s6dOzp58qQkqWHDhpKkYcOGKV++fAoJCdGRI0fsWR6QYQhjAAAAAAAAAACZYteuXZKkihUrysfHR5JkMplUv359SdLu3bvtVhuQkQhjAAAAAAAAAACZIu4UZXFZwpg9e/Zkek1AZiCMAQAAAAAAAABkCksY06hRo3ivN2jQQBIjY5B9EcYAAAAAAAAAADJcVFSU9u3bJynhyJh69epJks6ePavbt29nem1ARiOMAQAAAAAAAABkuEOHDikiIkI+Pj4qX758vG158+ZVxYoVJTFVGbInwhgAAAAAAAAAQIaLO0WZyWRKsJ2pypCdEcYAAAAAAAAAyNaOHTumwYMHy9/f396l5Gh//fWXpIRTlFkQxiA7c7J3AQAAAAAAAACQkT799FMtWLBAoaGhWrBggb3LyZEMw4g3MiYxljBmz549iomJkaOjY6bVB2Q0RsYAAAAAAAAAyNbOnTsnSVq0aJEuX75s52pyposXL+r69etydnZW3bp1E21TpUoVeXh4KDg4WCdPnszkCoGMRRgDAAAAAAAAIFu7cOGCJCkmJkZffPGFnavJmaZOnSpJeuyxx+Tu7p5oGycnJ2tQs2fPnkyrDcgMhDEAAAAAAAAAsq3w8HBdv37d+vz7779XYGCgHSvKebZv366vvvpKkvTRRx8l25Z1Y5BdEcYAAAAAAAAAyLYuXrwoSfLy8lLlypUVHBysH374wb5F5SDh4eEaOHCgJGngwIFq1apVsu0JY5BdEcYAAAAAAAAAyLYsU5SVKlVKb731liRp2rRpio6OTrT9/v379fnnnys0NDTTaszOxo4dqzNnzqho0aKaPHnyQ9vXr19fknTs2DEFBwdndHlApiGMAQAAAAAAAJBtWcIYPz8/9e7dW4UKFdKVK1e0ZMmSRNv3799fI0aMUIMGDXT69OnMLDXb2bdvnzWAmTFjhvLkyfPQfYoUKaKSJUvKMAzt27cvgysEMg9hDAAAAAAAAIBsyzJNWalSpeTq6qpXX31VkjR58mQZhhGv7Y0bN3T8+HFJsSMz6tatq19//TVT680uoqKiNGDAAJnNZj333HPq0KFDive1TFW2Y8eOjCoPyHRO9i4AAAAAAAAAwKPNbDZr0qRJ8vf3l7u7u3LlypXsf729vVW1alU5OGT834rHnaZMkoYMGaJx48bp8OHDOnHihKpUqWJtu337dklSuXLlVLhwYW3fvl3du3fXW2+9pQkTJsjZ2TnD680u/ve//+nYsWMqUKCAvvjii1Tt27p1ay1evFhfffWVXnvttRSNqJEkw5BCQ6WoKCkyUsqEj1eOEhUVe42RNoQxAAAAAAAAANJl8+bNGjlyZKr2eeedd/Tpp59mUEX/eTCM8fHxUbNmzbR+/XqtW7cuXhizdetWSVKbNm00ZcoUjRo1SpMnT9bnn3+uvXv3avHixSpSpEiG15xSK1asUP/+/TV//nx16tTJ3uVYHTlyRJ988okk6euvv1b+/PlTtf/zzz+vzz//XCdOnNDHH3+sKVOmpGi/sDDp/99mZJAnn7R3BY8uskEAAAAAAAAA6XL06FFJUpUqVTR8+HANGTJEffv2Vc+ePdW+fXs98cQTatCggapXr24NRWbOnKmIiIgMr+3BMEaKDVskad26dfHabtu2TZL0+OOPy9nZWZ999pmWLVum3Llza/v27apVq5Y1sElMZGSkrctP1qJFixQYGKixY8dm6nGTc//+fQ0YMED3799X165d1bNnz1T34ezsrGnTpkmSvvrqK508edLGVQKZz2Q8ODEikhQUFCRvb28FBgbKy8vL3uUAAAAAAAAAWcKQIUM0Y8YMvf/++xo/fnyybWNiYuTn56crV65o8eLFevrppzOsrsDAQOsUV8HBwfL09JQkHT9+XFWrVpWbm5sCAgLk7u6ugIAA5c+fX4Zh6Pr16ypUqJC1n9OnT6t79+46duyYHB0dNWHCBI0YMUImk0mSZBiGPv30U3344YcaNWpUpoUjNWrU0JEjRyRJ+/fvV506dTLluMn59NNP9d577ylv3rw6fvx4ukYSde7cWatWrVLbtm31+++/W693UgxDunRJmjdPKlBAcndP86GRiFOnpI4dpcaN7V1J1pLS3ICRMQAAAAAAAADSxTJyoWLFig9t6+joqH79+kmS5syZk6F1WUbF5M+f3xrESFLlypXl6+uriIgI62iYHTt2yDAMVahQIV4QI0nly5fX7t271adPH8XExOidd95Rt27dFBgYqJCQED3zzDMaOXKkoqOjtWTJkgw9J4v79+/r1KlT1ufff/99phw3OadOndKYMWMkSVOnTk33lG6ff/65nJ2dtXbtWv3+++8PbW8ySR4ekouL5OoqubnxsOXDxSX2GiNtCGMAAAAAAAAApMs///wjKWVhjCT1799fkrR+/XpduXIlo8rSxYsXJcWfokySTCZTgqnKLNOPNWvWLNG+PDw8NH/+fH377bdycXHRihUrVLduXTVs2FC//PKLnJ2dJcUGU4GBgRlxOvFcuHAh3rRoP/30k0JCQjL8uEmJiYnRgAEDFBkZqbZt26pv377p7rNs2bIaPny4JGn48OGKiopKd5+AvRDGAAAAAAAAAEizgIAA3bx5U1LKw5iyZcuqadOmMpvNmj9/fobVlth6MRYPhjFx14tJislk0pAhQ7Rjxw6VKFFCZ8+e1bFjx1S4cGFt3rxZpUqVkmEY2r9/v61PJQFLAFazZk2VK1dOISEhWrRoUYYfNynTp0/XX3/9JU9PT82YMeOhU4ql1Pvvv69ChQrpzJkz+uqrr2zSJ2APhDEAAAAAAAAA0swyVZavr2+8qcAeZsCAAZJipypLybLWe/fuVY8ePdS3b19FRESk6BjJhTGtWrWSg4ODTpw4oRMnTujgwYOSkg9jLB577DEdPHhQzz77rDp27KgDBw6ocePGqlevniRpz549KaovPU6cOCEpdsq1l156SZL9pio7f/68Ro4cKUn67LPPVKJECZv17eXlpQkTJkiSPv74Y924ccNmfQOZiTAGAAAAAAAAQJqldooyix49esjDw0Nnz57Vzp07k2x34MABdejQQfXr19eyZcv0448/ql+/fjKbzQ89RnJhTN68eVW/fn1J0pgxY2Q2m1WqVCkVL148RfX7+Pjop59+0qpVq1S0aFFJsva3d+/eFPWRHnHDmH79+snZ2Vn79u3T4cOHM/zYcRmGoUGDBiksLEzNmze3BkO21K9fP9WpU0dBQUEaPXq0zfsHMgNhDAAAAAAAAIA0O3nypKTUhzGenp56+umnJcWOjnnQoUOH1KlTJ9WtW1e//fabHBwc1L17dzk7O2vJkiV65513HnqM5MIY6b+pypYuXSopZaNikhN3ZExKRvukhyUEq1y5sgoUKKCuXbtKkmbOnJmhx33QDz/8oE2bNsnd3V0//PCDHBxs/5Wzg4ODvvzyS0nSrFmzrKOYgEcJYQwAAAAAAACANLOEMZUqVUr1vi+88IIkafHixXr33Xc1c+ZMrVq1Sl27dlXt2rW1evVqOTg46Pnnn9fJkye1dOlSa3Dz+eef64svvkiyb8MwrGGMn59fom0sYYxFs2bNUn0OcdWuXVtOTk66fv26rly5kq6+kmM2m61hjOW6W0akLFiwQOHh4Rl27LiuXLmit956S5L0ySefqEyZMhl2rEaNGum5556TYRh6/fXXMzzsAmyNMAYAAAAAAABAmqV1mjJJatKkiSpWrKjQ0FBNmjRJL730kjp37qwVK1bIZDLpueee04kTJzR//nyVK1dOktS7d2/rGiLDhw/XsmXLEu379u3bCgsLk8lkUsmSJRNt89hjjylv3rzW5+kdGePu7q7q1atLyth1Yy5fvqzQ0FA5OztbA5AWLVqoSJEiCgoK0r59+2x+vPv378d7zTAMDR48WMHBwWrQoIFee+01mx4zMZ9++qly5cqlHTt2aMmSJRl+PMCWCGMAAAAAAAAApElkZKTOnz8vKW1hjMlk0h9//KGpU6fqlVdeUbt27VS5cmU999xzOnbsmBYuXKgKFSok2O/dd9/Vyy+/LMMw1Lt370TXnLGMiilatKhcXV0TPb6jo6OefPJJa7vSpUun+hweFHeqsoxiCcDKly8vZ2dnSbFTeVnWrLFVGHPgwAG1b99eJUqUUJMmTXTv3j3rtoULF+r333+Xi4uLZs2aJUdHR5scMzm+vr567733JElvv/22wsLCMvyYgK0QxgAAAAAAAABIk7Nnz8psNsvLy0tFihRJUx9+fn5644039NVXX+n333/X8ePHtXDhQlWuXDnJfUwmk7788kt17txZkZGR6tSpk3W6NIuHrRdj8cwzz0iSunTpIpPJlKZziMsSiOzduzfdfSXlxIkTkhJODffYY49JSn8Yc/r0aXXr1k1169bV77//Lik2XGrdurXu3bunGzdu6PXXX5ckjRkzJtn3ytZGjBihEiVK6PLly/rss88y7bhAehHGAAAAAAAAAEgTSwBSsWJFmwQZqeHo6KiffvpJ9evXV0BAgNq1a6fr169bt6c0jOnWrZsOHz6syZMn26Quy8iY/fv3J5jay1YsYcyDIYgtwpiIiAi1aNFCy5cvl8lkUp8+fbRy5Ur5+Pho3759evLJJzVo0CAFBASoZs2aevvtt9N+Imng7u5ufa8+/fRT/fvvv5l6fCCtCGMAAAAAAAAApEl61ouxhVy5cmn16tUqW7asLl68qPbt2yskJERSysMYSapRo4bc3d1tUlPFihWVO3duhYWFWUMTW7Nc9wdHxtStW1eSdP78ed25cydNfS9ZskT+/v4qVqyYjh07ph9//FGdOnXSpk2blD9/fu3fv1+rV6+Wk5OT5syZY50mLTP16NFDjz/+uMLDw/Xuu+9m+vGBtCCMAQAAAAAAAB5xa9asUfv27eXv75+px7WMjHkwFMhMBQoU0Nq1a1WgQAEdPHhQPXv2VHR0tDWM8fPzy9R6HBwcrCNUMmLdGMMwkhwZkzdvXpUtW1ZS7MictPT91VdfSZKGDRsWr//q1atbAxkpdt2emjVrpuUU0s1kMumLL76QyWTSokWLtH37drvUAaQGYQwAAAAAAADwCDObzXr11Vf1+++/W79IzyxxpymzpzJlymjNmjVyd3fX2rVrNXTo0FSNjLE1y7oxGRHGXL9+Xffu3ZODg4PKly+fYLslCEpLGLN3717t379frq6uevHFFxNsr1atmg4cOKAlS5boo48+Sn3xNlSzZk0NGjRIkvT6668rJibGrvUAD0MYAwAAAAAAADzCtmzZoosXL0qSVq5cmWnHNZvNWSaMkWLXalm8eLEcHBw0a9YsnT17VpJ9whjLujF79+61ed+WKcpKly4tNze3BNvTs26MJczr1auXChQokGibEiVKqGfPnnJ0dEx1/7Y2fvx4eXt769ChQ5o7d669ywGSRRgDAAAAAAAAPMJmzZpl/fmff/7RmTNnMuW4V69eVWhoqJycnFSmTJlMOebDdOzYUd988431uZOTk3x9fTO9DsvImOPHj1vXsEkrs9msgIAA6/OkpiizsKwbk9ow5vr161qyZIkk6dVXX01LqZmuQIECGjNmjCRp1KhRCgwMtHNFQNIIYwAAAAAAAIBH1N27d7Vs2TJJUvHixSVJq1atypRjW0bFlC1b1i6LuCdl8ODBGjVqlCSpQoUKdhnBUaRIERUvXlxms1nfffddmvsxDENdunRRgQIFNG3aNEkPD2Nq164tBwcH+fv7p2oNoZkzZyo6OloNGzZUnTp10lxzZhs2bJgqVKigmzdvaty4cfYuB0gSYQwAAAAAAACQyUJCQvTTTz+pU6dOqlixotavX5+mfhYtWqTIyEhVq1ZNb7/9tqTMm6rMMl1WVpii7EHjx4/X4sWL9fPPP9uthiFDhkiS3n77bY0fP16GYaS6j++//16rV6+W2WzW8OHDNWrUKGsYU6lSpUT38fDwsAY1KR0dEx0dbQ2NXnnllVTXaU8uLi6aOnWqJOmLL77Q2bOnrNsOHfpd48c/oR07FtqrPMCKMAYAAAAAAADIJKGhoXrhhRdUsGBB9e7dW6tXr9apU6f01FNPpWkEhWWKsgEDBqhTp06SpJ07d+r27ds2rTsxWWm9mAeZTCY9/fTTqlatmt1qGDlypHUKrQ8++EAjRoxIVSBz6dIljRgxQpLUsmVLSdKECRO0detWSUmPjJFSt25MUFCQRo4cKX9/fxUqVEg9evRIcY1ZRbt27fTUU0/p/v37GjPmLZnNMVq5cow++6y9TpzYrG++6aOFC9+W2Rxj71KRgxHGAAAAAAAAAJlk0qRJmjt3rsLDw1W2bFmNHj1avXv3VkxMjIYOHarhw4crJiZlXxj//fffOnDggJydndWnTx+VLFlSNWvWlNls1po1a2xe+7lz59S6dWs1adJELVu21NKlSyUlPUIjpzOZTBo7dqx1erEpU6borbfeStG+hmHoxRdfVEhIiJo0aaL169dr5syZcnD47+vc5EKwlIQxd+/e1UcffaSSJUvq888/lyQNHz5cLi4uKaoxq5kyZYqcnJy0YcNv+uabelq9+mNJUuXKzSVJv/02WZ991kGhoffsVyRyNMIYAAAAAAAAIBPcvXvX+sX87Nmzdfr0aY0bN04//vijxo8fL0maNm2aunTpouDg4If2N2fOHElS586dlT9/fkmyjo7JiHVjZs6cqQ0bNmjnzp3atGmTbt26JUmP1Poi9vD6669bRzB9/fXXCgkJeeg+P/zwg/7880+5ublp9uzZcnBw0Isvvqhly5bJzc1N9evXl6enZ5L7W8KY/fv3JxiNc+fOHY0ePVp+fn4aO3as7t27p0qVKmnBggV655130nGm9lWhQgW99tprkiR//4NycXHXyy//qNGjN+vVVxfJxcVdf/+9VmPHNlZUVISdq0VOZDLSMllhDhUUFCRvb28FBgbKy8vL3uUAAAAAAADgEfLhhx9q3Lhxqlatmg4fPhxvlIMkLVmyRP369VNERISqV6+uNWvWqHjx4on2FRkZqaJFiyogIEC///672rVrJ0k6ePCg6tSpo1y5cun27dtyd3e3Wf2tW7fWhg0b9Morr6hRo0aKiopSiRIl1KJFC5sdI7syDENly5bV+fPntWrVKnXs2DHJtv/++6+qVq2q4OBgTZkyRcOHD4+3PSAgQB4eHnJ1dU2yj6ioKOXOnVtRUVE6e/asypQpo5s3b+rzzz/X9OnTFRoaKkmqVq2aRo8ere7du8vR0dE2J2tHgYGBatCgse7dc9DgwQtUoUJ167YLFw5qwoQnFRISoJEjN6hatVZ2rPTR9M8/Utu2UqNG9q4ka0lpbsDIGAAAAAAAACCDBQQE6IsvvpAkjRkzJkEQI0lPP/20tmzZokKFCunIkSOqV69ektNMffrppwoICFCxYsXUunVr6+u1atVS8eLFFRYWpo0bN9qsfsMwdOjQIUlS37599eyzz6pfv34EMSlkMpnUpk0bSdK6deuSbGeZniw4OFiNGjWyjvSIK1++fMkGMVLsovY1atSQFDtK6s0335Sfn58mTZqk0NBQ1apVS7/++qsOHz6sp59+OlsEMZLk7e2tLVuO6LXXjqh48erxtpUqVVvlyzeWJF2/fsYe5SGHI4wBAAAAAAAAMti0adMUFBSk6tWrq2vXrkm2q1+/vvbs2aNq1arp+vXratasmZYtWxavzYYNGzR27FhJ0v/+9794X6SbTCbrVGUrV660Wf1Xr17V7du35ejoqGrVqtms35ykbdu2kqS1a9cm2WbWrFnasGGD3NzcNGfOnHSFJJapyt58801NnTpV4eHhqlevnlavXq0DBw6oa9euiYaCj7rkzqlw4XKSpOvXT2dWOYBV9vttAwAAAAAAALKQgIAA61oxSY2KiatkyZLasWOHnnrqKYWHh6tHjx6aMGGCDMPQ5cuX9dxzz8kwDA0aNEh9+/ZNsH/nzp0lSatXr06wXkhaWUbFVKpUSW5ubjbpM6dp0aKFnJycdO7cOZ07dy7B9n///VdvvvmmJOmTTz5R+fLl03W8Jk2aWH9u1KiR1q5dq927d6tDhw4ymUzp6vtRVaRI7DVlZAzsgTAGAAAAAAAAyEBTp05VcHCwqlevri5duqRoHy8vL61cudI6TdWoUaM0YMAAPf3007p9+7Zq166tL7/8MtF9H3/8cTk6OurGjRvy9/e3yTlYwphatWrZpL+cKHfu3NaA5MGpygzD0EsvvaTg4GA1bNhQr7/+erqP98wzz2jGjBnatGmTduzYoTZt2uTYEMbCMjLm2jVGxiDzEcYAAAAAAAAAGSQ0NFRfffWVJOnDDz9M1bRQTk5O+uKLLzR9+nQ5Ojpq7ty52r17t/LkyaOlS5cmOULF1dVVZcqUkST9888/6T8JEcbYimXdmAenKpszZ47WrVsnV1fXdE9PZuHg4KCXXnpJLVq0yPEhjEXhwrEjY27duqD796PtXA1yGsIYAAAAAAAAIIP89NNPCgwMVJkyZZJdKyY5L7/8sn777Td5eXnJZDJp/vz5KlWqVLL7VK5cWZJ04sSJNB3zQYQxtmEJYzZt2qSoqChJ0uXLlzV8+HBJ0vjx41WhQgW71Zfd5c1bVC4u7oqJua9bty7auxzkMIQxAAAAAAAAQAYwDEPffPONJGno0KHpWiy9TZs2OnXqlI4ePaqOHTs+tL0tw5iAgABdunRJklSzZs1095eT1ahRQ4UKFVJoaKh27txpnZ4sKChIDRo0sIYyyBgODg7WqcpYNwaZjTAGAAAAAAAAyAB79uzR4cOH5ebmpv79+6e7v8KFC6tKlSopamvLMObvv/+WJJUqVUp58uRJd385mYODg1q3bi0pdt2YuXPnau3atXJ1ddXs2bNtMj0ZkvdfGMO6MchcWTKM+fbbb1W9enV5eXnJy8tLDRs21B9//CFJunjxokwmU6KPX375Jck++/fvn6B927ZtM+uUAAAAAAAAkMNYRsX06tVLPj4+mXrsSpUqSbLNmjFMUWZblu8kf/nlF+tImI8//tj6niFjWdaNYWQMMpuTvQtIjK+vryZOnKhy5crJMAzNmzdPnTt31qFDh1SxYkVdu3YtXvvvv/9en332mdq1a5dsv23bttWcOXOsz11dXTOkfgAAAAAAAORst2/f1uLFiyXFrvmS2SpWrCiTyaTbt2/r1q1bKlCgQJr7IoyxrSeffFImk0nnz5+XJNWvX19vvfWWnavKOSwjY65dY2QMMleWHBnTsWNHPfXUUypXrpzKly+vTz75RJ6entq9e7ccHR1VuHDheI/ly5fr6aeflqenZ7L9urq6xtsvb968mXRGAAAAAAAAyElmz56tqKgo1alTR4899limHz9Xrlzy8/OTlLqpyr777jtVqVJFhw8ftr5GGGNbBQoUUO3atSXFfl85Z84cpifLREWKMDIG9pElw5i4YmJitGjRIoWGhqphw4YJth84cECHDx/WwIEDH9rXli1bVLBgQVWoUEFDhw7VnTt3km0fGRmpoKCgeA8AAAAAAAAgOWazWd99950k+4yKsUjtujGbNm3SsGHDdOLECQ0cOFAxMTEKDw/XyZMnJRHG2FLfvn0lSRMnTmR6skxmGRlz586/ioqKsHM1yEmybBhz9OhReXp6ytXVVUOGDNHy5cut/wOJa9asWapUqZIaNWqUbH9t27bV/PnztXHjRn366afaunWr2rVrp5iYmCT3mTBhgry9va2P4sWLp/u8AAAAAAAAkH0FBARo7NixunDhgvLkyaNevXrZrZbUrBtz5coV9erVS2azWZJ08OBBfffddzp69KhiYmJUsGBBFSlSJEPrzUleffVV3bp1S2+88Ya9S8lxvLwKKFcubxmGoZs3z9m7nEdGSEiAwsMDZBiGvUt5ZGXZMKZChQo6fPiw9uzZo6FDh6pfv34JUvzw8HD99NNPKRoV06tXL3Xq1EnVqlVTly5dtGbNGu3bt09btmxJcp+RI0cqMDDQ+rh8+XJ6TwsAAAAAAADZ0IEDB/T888+raNGiGjdunCRp6NChypUrl91qSunImKioKPXs2VO3bt1SzZo1NXnyZEnS+++/r3Xr1kmKHRVjMpkytuAcxGQyKX/+/PYuI0cymUysG5MGy5eP0+TJPvr22/fsXcojy8neBSTFxcVFZcuWlSTVqVNH+/bt0xdffKEZM2ZY2yxdulRhYWHWYX2pUbp0aeXPn19nz55Vy5YtE23j6uoqV1fXtJ0AAAAAAAAAcoQtW7aoVatW1hlYatasqcGDB2vQoEF2rSulYcybb76p3bt3K0+ePFq2bJlKliypn376SQcPHtTHH38siSnKkL0ULlxe58/vZ92YVAgMvCFJypu3oJ0reXRl2ZExDzKbzYqMjIz32qxZs9SpUycVKFAg1f1duXJFd+7cYXglAAAAAAAA0uzq1at65plnFBMTozZt2mjPnj06ePCghgwZYvdF2S3TlF27dk337t1LtM3ChQs1ffp0SdKCBQtUunRpOTo66ptvvpHJZNL9+/clEcYge2FkTOoFBt6UJOXLV8jOlTy6smQYM3LkSG3btk0XL17U0aNHNXLkSG3ZskW9e/e2tjl79qy2bdumF198MdE+KlasqOXLl0uSQkJC9Pbbb2v37t26ePGiNm7cqM6dO6ts2bJq06ZNppwTAAAAAAAAspeoqCg9/fTTunnzpqpXr65ff/1V9erVyzLTeXl5ealYsWKSEl835siRI9bROx988IHat29v3Va/fv1437sRxiA7KVKkvCTZbGRMdHRkllpL5datS4qICLFpn0FBsWFMnjyMjEmrLBnG3Lx5U3379lWFChXUsmVL7du3T+vWrdOTTz5pbTN79mz5+vqqdevWifZx6tQpBQYGSpIcHR115MgRderUSeXLl9fAgQNVp04dbd++nWnIAAAAAAAAICl2feKff/5Z4eHhKWr/9ttv66+//pK3t7eWLVtm1/VhkpLUVGX37t1T9+7dFR4ertatW2vMmDEJ9p0wYYJKlSqlqlWrqkyZMplSL5AZLCNjrl9P/8iY27f/1ZAhBTVjxoB092ULt25d1JtvltMHH9RTRESozfplmrL0y5JrxsyaNeuhbf73v//pf//7X5Lb4yaR7u7u1sXGAAAAAAAAgMR88MEH+vzzzzV8+HBNmTIl2bY///yzvvzyS0nS/PnzrWsfZzWVK1fWhg0b4oUxZrNZ/fv319mzZ63rwyQ2pZqPj49OnDghZ2dnOThkyb/pBpKU3EAVSxhz7951hYcHy909d5qP8/fffyg8PEjHjm1Icx+2dO7cPsXEROvq1X+0ePFI9ev3Zbr7NJtjFBx8WxLTlKUHd1EAAAAAAADkeDExMfrpp58kxa6dEh0dnWTb48ePW6fwGjVqlDp16pQpNaZFYiNjJk2apJUrV8rFxUVLly6Vj49Pkvu7ubnZfe0bIDW8vSVfX+nKlaTbeHjkkZdX7Drk6Z2q7MyZ3ZJigx2zOSZdfdlC3PNZt+4rHT++Kd19BgffkWGYJUl58uRPd385FWEMAAAAAAAAcrydO3fq2rVrkqRbt25pw4bE/8o9KChI3bp1U1hYmFq1aqWPP/44M8tMNUsYY1kzZuPGjXr//fclSV9//bXq1q1rt9qAjODiIjVvLnl4SDdvJt2ucGHbrBtz9mxsGGM2x1in8rKnGzfOSpJy5cojSZox4wWFhQWlq0/LejHu7j5ycnJOV185GWEMAAAAAAAAcrzFixdLknUUyIIFCxK0MQxDL7zwgk6fPq3ixYsnOb1XVlKpUiVJ0qVLl3Ty5Ek9++yzMpvNeuGFF6yje4DspnhxqXFj6fZtKTIy8Tb/rRuT9jAmJOSu/P1PWp/fveuf5r5sxXI+zz03SQUKlNLt2/9qwYI309WnJYzx8GC9mPQgjAEAAAAAAECOdv/+fS1dulSSrAvZr1ixQsHBwfHaTZ48Wb/++qt1eq8CBQpkeq2p5ePjo4IFY79AffLJJ3Xr1i3VqlVL06dPl8lksnN1QMapW1eqXFk6fz7x9WOKFIkdGXPt2uk0H+Pcub3xngcEXE1zX7ZiCWP8/Gpp6NB5MplM2rJllg4d+i3NfVpG/BDGpA9hDAAAAAAAAHK0bdu26ebNm8qXL5/ee+89lS9fXuHh4fr111+tbTZv3qz33ntPkvTFF1+oXr169io31SxTlV25ckV58+bVsmXL5O7ubueqgIzl4iK1aCHlySNdv55wuyWM8ff/J83HsExRZnH3rn3DmPDwYGtwUqhQWVWs2FRPPRU7KmbmzBcVHHwnTf0GBlpGxhSyTaE5FGEMAAAAAAAAcjTLFGXdunWTs7Oz+vTpI+m/qcquXr2qXr16yWw2q2/fvho8eLDdak0LSxhjMpm0YMEClSpVys4VAZmjcGGpWTMpMFAKCYm/rXTpxyRJFy8eSvOaKpYwxsUlNty09zRllvVicufOLw+PPJKknj3Hq1ixSrp377rmzn0lTf0yTZltEMYAAAAAAAAgx4qOjtayZcskSc8884wkqXfv3pJiF7u/ePGievbsqZs3b6p69er69ttvH7npvbp06aJcuXJp4sSJeuqpp+xdDpCpataMnbLs4kXp/v3/Xs+fv4QKFy4nszlG//yzJdX9ms1mnT27R5JUo0Y7Sfafpuz69dgwxrIejiS5uLhpyJB5cnBw1K5di7R795JU92sZbZMrF2FMehDGAAAAAAAAIMfavHmz7ty5owIFCqh58+aSpNKlS6tx48YyDENPPPGEdu3aJW9vb/3666/KlSuXfQtOgyeffFLBwcF655137F0KkOkcHWOnKytbNnb9mLiqVm0lSTp27M9U93v9+hmFht6Vs7ObataMDWPsPU2ZZb2YQoXKxnu9TJnH1LnzKEnS7Nkv6969ROZtS4ZlZIynJ9OUpQdhDAAAAAAAAHIsyxRlPXr0kJOTk/V1y1RlFy5ckBQ7ZVmZMmUyv0AbcXDga0DkXJ6e0pNPxv7XP85MYukJYyxTlJUuXVf585eUlHWmKYs7Msaia9fRKlmypkJC7mjWrMEyDCPF/TIyxja4CwMAAAAAACBHioiI0PLlyyVJTz/9dLxtPXv2lLOzsyTp/fffV4cOHTK9PgC24+srPfGEFBwsBf3/EjGVK7eQyeSgq1f/SfUUY2fOxIYxZcs2UN68xSRlhWnKYkfGFC5cNsE2JycXDR06X46OzjpwYJW2b5+f4n4ZGWMbhDEAAAAAAADIkcaPH6+7d+/K19dXTZs2jbfNx8dH8+bN07hx4/TRRx/ZqUIAtlSjhtSokXTlihQRIXl65lXp0nUlSceObUxVX5aRMeXKNVC+fLFhTFjYPUVGhtm26FT4L4xJODJGkkqUqKaePT+WJM2b95ru3Lmcon4tYQwjY9KHMAYAAAAAAAA5zqFDhzRx4kRJ0rRp0+To6JigzbPPPqvRo0cnug3Ao8fBQWrWTKpWTTp3ToqJSdtUZRERofr33yOSYkfGuLt7ydU1dj0pe01VFh4ebJ1O7ME1Y+Lq0OFtlS3bQOHhQZoxY4DMZnOy/UZEhFgDJg8Pwpj0IIwBAAAAAABAjhIdHa0BAwYoJiZGPXr0UPfu3e1dEoBM4uoqtWkjlSolnT0rVanyXxiT0nVULlzYL8MwK18+X+XLV0wmk8k6Vdndu/aZqsyyXkzu3Pnl4ZEnyXYODo4aOnSeXFzcdezYn9q48btk+w0MjB0V4+LiLhcXT5vVmxM5PbwJAAAAAAAAkH1MmjRJhw8fVr58+fT111/buxwAmczbW2rbVvrlFykkpKFcXNx17941Xb36j3x9Kz90/7jrxVjky1dM16+fsdu6Mdevx4YxSU1RFleRIuX17LOfat681/TTT28rOjpSkZGhCg29K09PH3Xq9K4cHGJHBFqmKPPyKiiTyZRxJ5ADEMYAAAAAAAAgxzhx4oQ+/jh2zYQvvvhChQqxIDWQExUrFjtCZsUKN5Uq1VSnTq3X0aMbkg1jzOYYnT+/X/v3L5cklSvX0LotT56ikuw3TZllvZjkpiiL68knh2nfvuU6cWKzFix4M962kiVrqlatpyTJOvWZlxdTlKUXYQwAAAAAAACyPcMwtGjRIr3++uuKiopS+/bt1bt3b3uXBcCOKlWSQkOlzZtb6dSp9Tp27E+1a/e6dbthGLpx45yOHt2gY8c26PjxzQoLu2fdXqFCE+vP+fJljWnKUjIyRpIcHBz08ss/asGCt2Q2x8jDI4/Ont2ty5eP6fLlI9YwxjIyxtub4Dq9CGMAAAAAAACQrV28eFFDhw7V2rVrJUlVqlTR999/z5Q7AFSnjvT00620YYN04sQW3b17TadObdfRoxt09OgG3b59KV77XLnyqEqVJ/TYY11Vpsxj1tcta8Y8OE2Zv/8pmUwOKlIkZSFJWllGxhQunLKRMVJsgPTaa4usz1es+J8uX35fV64ct75mGRnj7c3ImPQijAEAAAAAAEC2tWfPHrVs2VKhoaFycXHRBx98oHfeeUcuLi72Lg1AFmAySf361dBbb+VXUNBtDRtWNN52R0dnlS/fSNWqPamqVVupdOm61vVU4sqbN+E0ZeHhwRozpqEcHBz19ddX5OzsmmHn8V8Yk/bQp1ix2Cnarlw5YX3tvzVjGBmTXoQxAAAAAAAAyJZCQ0PVp08fhYaGqlGjRpo9e7YqVKhg77IAZDHOzg7q1q2T5s6dLUkqVqyqqld/UtWqPamKFZvKzc3zoX0kNk3ZuXN7FRp6V5J069ZFFS2aMfef8PBg6wiWlK4ZkxjLejn+/v/IbDbLwcFBgYGWacoYGZNehDEAAAAAAADIlt59912dPXtWxYoV02+//aY8efLYuyQAWdQXX0xV1649detWDV24UESlS0vu7inf3zJN2d27/jIMQyaTSadP/2XdnpFhjGW9mNy588vDI0+a+ylYsLScnV0VFRWuW7cuqlCh0goKig15vLwIY9LLwd4FAAAAAAAAALb2559/avr06ZKk2bNnE8QASJaXl5c6dWqr3r2LqE4d6fx5KSws5fvnzVtEknT/fpSCg+9Ikk6f3mndfuvWBZvWG9f167FhTHqmKJMkR0cnFSkSGxhdvRo7Vdl/I2OYpiy9CGMAAAAAAACQrdy7d08vvPCCJGno0KFq3bq1nSsC8Khwc5PatZPq1pUuXEh5IOPk5CIvrwKSYqcqM5vNOnNml3V7xoYxsevFpGeKMgtf3yqS/gtj/lszhpEx6cU0ZQAAAAAAAMhWXn/9dV25ckVlypTRZ599Zu9yADxi3Nyktm1jf96/XypVSsqV6+H75c1bTEFBtxQQcFUODo4KDw+ybrt5M+PCGMs0ZekdGSNJxYrFrhtz5cpxxcTcV0hI7Cgfb++CCg1Nd/c5GiNjAAAAAAAAkG2sWLFC8+fPl8lk0rx58+Th4WHvkgA8giyBTGpGyOTNW1SSdO+ev3W9GEdHZ0kZOzLm2rXTkqTChdM/MsYSxly9ekLBwbet69/kzp0/3X3ndIQxAAAAAAAAyBZu3bqlwYMHS5LefvttNW7c2M4VAXiUWaYse+yxlAUyefMWkyQFBFzVmTOxYUzNmu0kZdzImJCQuzp3bq8kyc+vdrr7iztN2b171yVJuXPnl4ODY7r7zukIYwAAAAAAAPDIMwxDQ4YM0c2bN1W1alV9/PHH9i4JQDbg6ho7QqZevYcHMvny/RfGWEbGNGr0nCQpJOSOwsODbV7fwYOrFBMTreLFq6po0Qrp7q9QoTJydHRWZGSYzp/fJ0ny9i6U7n5BGAMAAAAAAIBsYOHChfr111/l5OSk+fPny9XV1d4lAcgmXF2lNm0eHshYpim7dOmwrl8/I0mqVq21PD3zScqYqcr27PlFklS/fk+b9Ofo6GQNdU6c2CJJ8vIqaJO+czrCGAAAAAAAADzSrly5oldeeUWSNGbMGNWqVcvOFQHIblISyFimKbOMKPH1rSJPz7wqUKCUJOnWrYs2rSk09J6OHFkvSapXr4fN+i1WLHaqshMnNksijLEVwhgAAAAAAAA8sgzD0MCBAxUYGKjHHntM7733nr1LApBNWQKZ+vVjA5mQkPjbLdOUWZQr10iSrGGMrdeNOXhwtWJiolWsWGX5+la2Wb+Wvu7duyaJacpsxcneBQAAAAAAAABpNWPGDK1fv15ubm6aP3++nJz4ugtAxrEEMiaTtGePVKyY5O0du80yMsaifPnYMKZgQcvIGNuGMbaeoszC17dKvOeMjLEN/u8EAAAAAACAR9K5c+c0YsQISdKECRNUsWJFO1cEICdwcZHatpXc3KQdO6ToaCl/fil3bh85Obno/v0oSf+FMf9NU2a7MCYsLFBHjqyTJDVoYNswplix+KNsGBljG0xTBgAAAAAAgEdOTEyM+vXrp9DQUDVv3lyvvfaavUsCkIM4OUlPPCG1bi3duyf5+0smk0l58xaVJOXOnV+FC5eT9N/IGFtOU3bw4Grdvx+lokUrJghP0qtQobJydHS2PmdkjG0QxgAAAAAAAOCRM2XKFO3cuVO5c+fWnDlz5ODA11wAMpeDg9SwodShgxQTE7uOjGWqsvLlG8lkMkmKPzLGMAybHHvPnqWSYqcosxzHVpycnFWkSHnrc29vwhhbYJoyAAAAAAAAPFLu3Lmj0aNHS5KmTp0qPz8/+xYEIMcymaRatSRPT2ntWsnV1U/STpUv38TaJn/+kpKkiIgQBQffkZdX/lQdIyQkQAsXjtC1a6dVsmRNlSxZQ0eOrJVk+/ViLIoVq6wrV45LYpoyWyGMAQAAAAAAwCPl3LlzioqKUrFixTRgwAB7lwMAKldO8vCQgoPHKl++Cnr88SHWbS4ubsqbt6ju3vXXrVsXUhXG/PvvUU2Z0kU3b56XJJ0+vdO6rUiRCipevKrtTiIOX98q2rPnF0lS7twFMuQYOQ1hDAAAAAAAAB4pAQEBkqSCBQvafHoeAEirokWloUPLys/vA504ITk6xo6YkaQCBfysYUyZMo+lqL/du5doxowXFBkZpgIF/NS58yhdu3ZKFy8e0o0b59S16+gMuwda1qFxdfWQm5tHhhwjpyGMAQAAAAAAwCPFEsbky5fPzpUAQHx580pdu0peXtLevVLBgpKPT+y6MadP/6Vbty4+tA+zOUaLF7+v1as/lSRVrdpKr766SLlz+2Rw9f8pV66BHB2dVbJkjUw7ZnZHGAMAAAAAAIBHCmEMgKzM3V1q107Kk0faulUKC4sNYyTp5s0Lye4bEhKgr756VkePrpckdejwtp555n9ydMzcr/J9fIprypQz8vTkPmsrhDEAAAAAAAB4pBDGAMjqHB2lJk1iR8qsXy+ZzbFhzK1bSYcx//57RFOmdNXNm+fl4uKul16arUaNemVWyQkUKFDSbsfOjghjAAAAAAAA8EghjAHwqKhSRfL2lq5diw1jbtxIPIx5cH2YN99cwRRh2YyDvQsAAAAAAAAAUoMwBsCjxNdXev752DDm9u2LunPHbN1mNsfo55/f1ZdfPqPIyDBVrdpK48fvJ4jJhhgZAwAAAAAAgEcKYQyAR02lSr5ydHRUTEyULl++puDgYsqXL0DTp9t/fRhkDt5VAAAAAAAAPFIIYwA8apycnFSiRAlduHBB1atf0PHjdzRpUlfdu5c11odBxiOMAQAAAAAAwCOFMAbAo8jPz08XLlzQ6tVfa/Xq1QoLC1OePKU0ePBy1ajBtGTZHWvGAAAAAAAA4JFCGAPgUVSqVOy6MYsXL1ZYWJhatnxSP/64T97eNXT+vGQ2P6QDPNIIYwAAAAAAAPDIMAyDMAbAI6l06dLWn99++22tXfu7OnTwUc+eko+P9M8/UliYHQtEhmKaMgAAAAAAADwygoODFRMTI0nKmzevnasBgJTr06ePDh48qGeffVY9evSwvl62rJQ/v7Rpk3T4cGwwU7Cg/epExiCMAQAAAAAAwCPDMirG3d1d7u7udq4GAFKuZMmSWrZsWaLb8uSROnWSihaVtm2TzpyRSpeWHB0zt0ZkHKYpAwAAAAAAwCODKcoAZFdOTlKDBtLTT8eGMv/8I4WE2Lsq2AphDAAAAAAAAB4ZhDEAsjs/P+mZZ2KDmStXpGvX7F0RbIEwBgAAAAAAAI8MwhgAOUHu3NJTT8VOXWYY0unT0v379q4K6UEYAwAAAAAAgEcGYQyAnMLRUapTJ3aUTMmS0qlTUmCgvatCWhHGAAAAAAAA4JFBGAMgp/H1lXr2lJo0kW7dki5elMxme1eF1CKMAQAAAAAAwCODMAZATuThIbVpI/XoIeXJI/3zjxQSYu+qkBpO9i4AAAAAAAAASCnCGAA5lckkVaokFSkibdsmHTwo5colFSsmOTDsIsvjLQIAAAAAAMAjgzAGQE6XJ4/UoYPUvXtsGHPypBQaau+q8DBZMoz59ttvVb16dXl5ecnLy0sNGzbUH3/8Yd3evHlzmUymeI8hQ4Yk26dhGPrwww9VpEgRubu7q1WrVjpz5kxGnwoAAAAAAABsiDAGAGJHwlSrJj33nFSnjnTlinT5MmvJZGVZMozx9fXVxIkTdeDAAe3fv19PPPGEOnfurOPHj1vbDBo0SNeuXbM+Jk2alGyfkyZN0pdffqnvvvtOe/bskYeHh9q0aaOIiIiMPh0AAAAAAADYCGEMAPwnXz6pUyepW7fYUTKsJZN1Zck1Yzp27Bjv+SeffKJvv/1Wu3fvVpUqVSRJuXLlUuHChVPUn2EYmjZtmkaPHq3OnTtLkubPn69ChQppxYoV6tWrl21PAAAAAAAAAA9lGIZu3bqlggULpngfwhgAiM/BQapeXSpeXNq+XTp8WHJxiX3OWjJZR5Z/K2JiYrRo0SKFhoaqYcOG1tcXLlyo/Pnzq2rVqho5cqTCwsKS7OPChQu6fv26WrVqZX3N29tb9evX165du5LcLzIyUkFBQfEeAAAAACBJQUFBevHFF7Vo0SJ7lwIAj6R9+/apcePGKlSokBYuXJiifQzDIIwBgCTkzfvfWjLe3tKJE9K9e/auChZZNow5evSoPD095erqqiFDhmj58uWqXLmyJOm5557TggULtHnzZo0cOVI//vij+vTpk2Rf169flyQVKlQo3uuFChWybkvMhAkT5O3tbX0UL17cBmcGAAAAIDtYv369Zs2apWeffVbvvvuuzEzQDQAp4u/vr/79+6tevXrWP5Ldtm1bivYNDw9XZGSkJMIYAEiMg4NUpYrUu7fUrFlsGHPmjBQVZe/KkCWnKZOkChUq6PDhwwoMDNTSpUvVr18/bd26VZUrV9ZLL71kbVetWjUVKVJELVu21Llz51SmTBmb1TBy5Ei9+eab1udBQUEEMgAAAAAkSaGhodafJ02apFOnTmnBggXy9PS0Y1UAkHVFRERoypQp+t///me9h1arVk1Hjx7V+fPnU9SHZVSMs7OzPDw8MqxWAHjUeXlJrVtLZcvGTl129qzk4yMVLCiZTPauLmfKsiNjXFxcVLZsWdWpU0cTJkxQjRo19MUXXyTatn79+pKks2fPJrrdsrbMjRs34r1+48aNZNedcXV1lZeXV7wHAAAAAEixXypKkq+vr1xcXLRy5Uo1bdpUV65csXNlAJC1GIahpUuXqlKlSnr//fetU9Hv2bNH06dPlySdO3cuRX3FnaLMxLeJAPBQpUtLvXpJTz0lxcRIJ09Kyaz4gQyUZcOYB5nNZusw1AcdPnxYklSkSJFEt5cqVUqFCxfWxo0bra8FBQVpz5498dahAQAAAICUsvz7pFGjRtq8ebMKFCigw4cPq169etq3b5+dqwOArOHw4cNq8X/s3Wd0VOX39vHvpJEQSEILHaT3XqQpSA0gTUBQaQooiBQVFPiLWFAQsYAo0ouKShEQRUCE0KRJr6H3EpoJJEDK3M8LnpkfkQApM5mU67NWljDnnPvsgSSYuWbv+6mn6NixIydPnqRAgQL88MMPbNy4kZo1a1K0aFEATp8+TXR09CPX034xIiKJlykT1K4NXbpApUpw5gycPg2aspuyUmUYM2zYMNatW8fJkyfZu3cvw4YNIzg4mBdeeIFjx47x4Ycfsn37dk6ePMmvv/5Kt27dePLJJ6lYsaJ9jdKlS7No0SIALBYLgwYNYtSoUfz666/s3buXbt26kS9fPtq2beuiZykiIiIiImmZLYzJlCkTderUYevWrZQrV44LFy7w5JNPMn/+fBdXKCLiOqGhobz88stUrVqVtWvX4u3tzbvvvsuhQ4d4/vnn7V0tefPmxdvbm9jYWE6fPv3IdRXGiIgkXe7c0LYttG8PWbPCwYMQFubqqjKOVBnGhIaG0q1bN0qVKkWjRo3Ytm0bK1asoEmTJnh5ebFq1SqaNm1K6dKlefPNN2nfvj1Lly6Ns0ZISAhh93wmvfXWW/Tv35+XX36ZGjVqcPPmTZYvX463t3dKPz0REREREUkH7g1jAB577DH+/vtvWrRowe3bt3n22WcZNWoUxpg414WFhdG6dWuyZMlCkSJFqF27Ns888wxr165N8ecgIuIMS5YsoUSJEkydOhVjDJ07dyYkJIT333//vn1e3NzcKFKkCECC9o1RGCMikjzu7lChwt0umXr14OpVOHIEoqJcXVn65+HqAuIzffr0Bx4rWLBggn5I+e8PPBaLhQ8++IAPPvgg2fWJiIiIiIj8N4wB8PPz49dff2Xw4MF8+eWXjBgxgkOHDjFt2jS8vb0JDQ0lKCiInTt3AhAREcHJkycBOH/+PJs3b07x55FR3bhxg3379lGrVq149504deoU+fLlw9PT0wXViaRNxhi+/PJL3nzzTYwxVKtWjfHjx1O3bt2HXlesWDEOHjzIsWPHaNKkyUPPVRgjIuIYfn7QrBmUKAEbN94NZLJlgzx5QFtyOUeq7IwRERERERFJ7eILYwDc3d354osv+Pbbb3F3d+eHH36gUaNGbN++nSeeeIKdO3cSGBjI6tWr+fvvv/nqq68AOHr0aIo/h4zIarUyZ84cSpYsSZ06dRg3btx95/z444889thj/N///Z8LKhRJm2JiYujXrx9vvPEGxhj69OnD5s2bHxnEAPZ9Y9QZIyKSsiwWKFYMOnWC1q3BzQ0OHIDwcFdXlj4pjBEREREREUmCB4UxNq+88grLly8nICCAv//+m+rVq3P48GEKFSrEhg0beOqpp6hduzbdu3cH4OrVq3FGLYvjbd++nXr16tG9e3cuXrwIwJgxYwi/5xWHqKgohg8fDsCKFStcUqdIWtSlSxcmTZqExWLhs88+45tvvsHDI2EDWYoVKwYojBERcZVMmaBGDejWDWrVgitXNLrMGRTGiIiIiIiIJMGjwhiAxo0bs3nzZooXLw5AqVKl2LBhAyVKlLCfkzVrVnLlygXAiRMnnFhxxnX58mX7/qGbNm3C19eX0aNHU6pUKa5du8bEiRPt586ePds+Ou7AgQP2v2cRebCLFy/y888/Y7FY+OWXX3jjjTfiHf/3ILbOmGPHjj3yXIUxIiLOkz07tGwJnTtDkSJw7BicPQtWq6srSx8UxoiIiIiIiCRBQsIYuBvAbN26ldmzZ/P3339TsGDB+85JzIiejOzy5cuMGjWKNm3aJOhF25iYGL766itKlixp30j8+eefJyQkhKFDhzJixAgAPvvsM27cuEFUVBSjRo2Kc/3Bgwed9nxE0outW7cCULZsWdq2bZvo6+/tjPnvHsD/pTBGRMS57h1d1rYt+PjcHV32/7/9SjIkrF80EX788Ud+//13rly5Qv78+enQoQPNmzd39G1ERERERERcKqFhDEC2bNno1q3bA48XLVqULVu2KIx5gH379vHFF1/www8/2P/c79y5w/Llyx94TXBwMAMGDGDv3r0AVK5cma+++op69erZz+ncuTMffPABhw8fZuLEiWTLlo3Tp0+TN29eChUqxJYtW9i9ezeVK1d26vMTSetsYUzNmjWTdP1jjz0GQHh4OFevXiVnzpwPPFdhjIhIyvDygqpV7wYz27bB9u0QEeHqqtK2RHXGLFu2jLJly8Z5p5CNMYb27dvTpUsXfvzxR1auXMnMmTN5+umnee211xxWsIiIiIiISGqQmDDmURIzoiejOXLkCFWqVGHGjBncuXOH6tWr4+npyYoVK/jrr7/uO//MmTN06tSJp556ir1795I9e3YmTZrEP//8EyeIAXB3d+fdd98FYNy4cXz00UcADBs2jMcffxyA3bt3O/kZiqR9tjDG9nWTWD4+PuTPnx94dIegwhgRkZTl7w+NG0OXLlC3LmTN6uqK0q5EhTHLly8nJCSEoKCg+45NmTKFRYsWYYyhbdu2fPXVVwwYMAAvLy8mTZrEypUrHVa0iIiIiIiIqzkjjFFnzP327dtHTEwMhQoV4u+//2bbtm306dMHgLfffhvrPUPMf/vtN0qXLs28efNwc3Pj1Vdf5fDhw/Tp0wd3d/d41+/cuTMlS5bk2rVrnD17lvz589O7d28qVaoEKIwReRSr1ZrszhhIeCitMEZExDXy54dnnoEKFVxdSdqVqDBm06ZN5M2bl+rVq993bOLEiVgsFnr27MnChQvp168fX375JfPnz8cYw4wZMxxWtIiIiIiIiKspjEkZ58+fB6Bq1arUrl0bgHfeeYcsWbKwfft25s2bB9wdS9ahQwciIyOpW7cu27dv5+uvvyZHjhwPXd/d3d2+dwzA8OHD8fb2jhPGPGoPC5GM7MiRI4SFheHj40P58uWTvM69+8Y8yJ07d4j4/zNyFMaIiEhak6gw5ty5c1SrVi3ex/fv3w/A4MGD4xx7+umnKV26NFu2bElGmSIiIiIiIqmLM8KYkydPEhsbm+z10pMLFy4AkC9fPvtjgYGBvPXWWwD83//9Hxs3bqRVq1bcuXOHNm3aEBwcnKh9Xjp37kzz5s1p0KABPXv2BKBcuXK4u7tz9epVzp0757gnJJLO2F7vqVq1Kp6enkleJyGdMbauGDc3N/z8/JJ8LxEREVdIVBhz9epVsmTJct/j//zzDwCFChWiVKlS9x0vVaoUly5dSmKJIiIiIiIiqY8jw5j8+fPj6elJTEwMZ8+eTfZ66YktjMmbN2+cx19//XVy587N8ePHqV+/Pjdv3qRRo0b89NNPeHh4JOoeHh4eLFu2jDVr1tj/Pr29vSldujSgUWUiD5Pc/WJsEtIZYwtjsmXLhptbol7SEhERcblE/cvl4+NjbxG/l+1dEDVq1HjgdQ+azysiIiIiIpIWOTKMcXd357HHHgM0quy/bD+D/jeMyZIlC++99x4AsbGxPP744yxevBhvb2+H3Vv7xog8mu01oeTsFwMJG9eo/WJERCQtS1QYU7ZsWTZt2sTFixfjPL548WIsFgtPPvlkvNedPn36vv9xFhERERERScscGcZAwt4VnhHFN6bMpmfPnjRr1oz69euzbNmyeCc5JIfCGJGHu337tv3rI7mdMbYw5uzZs/bvr/+lMEZERNKyRIUxHTt2JCoqiqeffpo1a9awd+9e+vXrx6FDh/D09KRDhw73XRMZGcmOHTvsP1iIiIiIiIikB44OYxLyrvCM6EFjygA8PT1Zvnw5wcHBTnlxVmGMyMPt2rWL6OhocuXKReHChZO1Vq5cuciSJQvGGE6ePBnvOQpjREQkLUvUIN1+/frx/fffs2PHDho3bhzn2Ouvv06ePHnuu2bevHncuXOHRo0aJa9SERERERGRVERhjPNFR0dz+fJlIP4wxtlsYczhw4eJiIjA19c3xWsQSc3u3S/GYrEkay2LxULRokXZs2cPx44di3dPYoUxIiKSliWqM8bLy4vVq1fz2muvkTdvXjw8PChSpAijR4/m448/jveaKVOm4OfnR1BQkEMKFhERERERSQ1sYYyj9ihRGHO/S5cuYYzB3d2dXLlypfj98+TJQ+7cuTHGsG/fvhS/v0hqEhsbywcffMD8+fPtjzlqvxibR41rVBgjIiJpWaLCGAB/f38mTJhgn+F59OhR3n777Qe+A+Lvv//m+vXrlC9fPtnFioiIiIiIpBbqjHE+24iyPHny4OaW6B9fHUKjykTuWrx4MSNHjuTZZ59lypQpQNzOGEewfR88duxYvMcVxoiISFqWqDFlIiIiIiIicpejw5giRYoAcOXKFcLDw/Hz83PIumnZw/aLSSmVKlVi5cqVCmMkw7MFMAB9+vTh5s2bHD16FIAaNWo45B7qjBERkfTMNW8tEhERERERScOsVivR0dGA48IYPz8/cubMCcCJEyccsmZal1rCGFBnjGRsJ06cYOXKlQB07twZYwxvvvkmACVLliRbtmwOuY86Y0REJD1LVGdMw4YNH3mOxWLB19eXAgUK8MQTT9C+fXu8vLySXKCIiIiIiEhqExUVZf+1o8IYuPtC5JUrVzh+/Lg9BMjIzp8/D0C+fPlcVoPt72HPnj1YrVaXjUsTcaVp06YB0KRJE+bOnYufn5+9U8ZRI8ogbmeMMSbOSHyr1WoPRQsVKuSwe4qIiKSURIUxwcHBiVp88uTJDBs2jHnz5jlsMzcRERERERFXs40oA8eHMVu3bn3gu8IzmtTQGVOqVCm8vLy4ceMGJ06csL9YLJJRREdHM2PGDABefvllLBYLkyZNwmq1Mm3aNNq0aeOwexUqVAh3d3du3brFyZMn7eMbAbZv387ly5fJmjUrtWvXdtg9RUREUkqiwpiZM2c+8hxjDJGRkRw/fpzff/+dkJAQWrRowe7du8mfP3+SCxUREREREUkt7g1jHDkJwDai50H7JWQ0qSGM8fT0pHTp0uzZs4eQkBCFMZLh/Pbbb1y8eJHAwEBat24NgJubG1OnTmXcuHH4+/s77F5eXl7UrVuXdevWsWTJEgYNGmQ/tmzZMuBud46np6fD7ikiIpJSEhXGdO/ePVGLf/rppwwfPpxPPvmEzz//nM8++yxR14uIiIiIiKRGtjDGy8srzhid5FIYE1dqGFMGkDt3bgCuXLni0jpEXME2juzFF1+8L3x2ZBBj06FDB9atW8eCBQvihDF//PEHAC1atHD4PUVERFKCU4fdWiwWPvroIwoVKsTy5cudeSsREREREZEUYwtjHDmiDBTG/Fdq6IwByJEjBwBXr151aR0iKe3kyZOsWLECgF69eqXIPZ955hkANm7caA9kL1++zNatWwEICgpKkTpEREQczek7D7q5uVGzZk1OnTrl7FuJiIiIiIikCGeFMbYRWCdPniQ2Ntaha6c1sbGxXLp0CXB9GJMzZ05AYYxkPBMnTsQYQ+PGjSlevHiK3DN//vz2PWEWLVoEwIoVKzDGUKlSJY3AFxGRNMvpYQyAu7s7Vqs1JW4lIiIiIiLidM4KY/Lnz4+npyfR0dGcO3fOoWunNZcvX8ZqteLm5kZgYKBLa7F1xmhMmWQUFy5coH379vZx86+88kqK3r9Dhw4ALFiwAPjfiLLmzZunaB0iIiKOlCJhzK5du/TOBRERERERSTecFca4u7vz2GOPARpVZhtPFBgYiIdHorY7dTiNKZOMwhjDtGnTKFOmDL/88gseHh6MGDGC9u3bp2gdtvutW7eOCxcu2Effa78YERFJy5wexkyZMoWQkBCeeuopZ99KREREREQkRTgrjAHtG2OTWvaLAY0pk4zh6NGjNGrUiN69exMWFkb16tX5559/+OCDD7BYLClaS+HChalRowZWq5Xhw4dz7do1/P397ePLRERE0qJEvb1o3bp1CTovMjKS48eP8/vvv7N8+XK8vb154403klSgiIiIiIhIaqMwJn6nT5/m9OnT1KtXL9lrpaYwRmPKJD2LiYnhiy++YOTIkdy6dQsfHx8+/PBDBg4c6NKutPbt27Nt2zZmzZoFQNOmTV3eJSciIpIcifpXrEGDBol6N4QxhixZsjBnzhxKly6d6OJERERERERSI4Ux8WvVqhV79uzhm2++oW/fvslayzamLF++fI4oLVnUGSPp1a5du+jVqxfbt28HoGHDhkyZMoVixYq5uLK7YczQoUPtv9eIMhERSesSFcYUKlTokWGMxWIhc+bMFChQgCeffJKXXnqJPHnyJKtIERERERGR1ERhzP1CQ0PZs2cPAP3796dIkSIEBQUleb3U2hljjEnxkU1yv4iICO7cuUP27NldXYrLXL58mePHj+Ph4YGHhwc+Pj4UL14cN7dHT6S/ffs2H3zwAWPHjiU2NpaAgAA+++wzXnzxxVTz+V28eHEqV67Mrl27AJL1/URERCQ1SFQYc/LkSSeVISIiIiIiknYojLnf5s2b7b+OjY3l2WefZePGjVSoUCFJ66XGMCYqKoqIiAiyZMni4ooyNqvVylNPPUVISAgHDx5MFd1TKe3atWuUL1+e0NDQOI9Xr16dSZMmUb169Qdeu3fvXjp06MDhw4cB6NChA1999VWqfCNthw4d2LVrF1WrVk2V9YmIiCTGo98uISIiIiIiInE4M4wpUqQIcPdd7zdu3HD4+s6yadMmALp27UqDBg24ceMGLVu2tIcqiWW7LjW80O7r62v/u9aoMtdbvnw527ZtIzw8nBUrVri6HJf46KOPCA0NJUuWLBQoUIA8efKQKVMm/vnnH2rWrEm/fv24fv36fdfduXOHjh07cvjwYfLmzcsvv/zC/PnzU23Q0b9/f/r168dXX33l6lJERESSzSlhzOHDh/nqq68YN24cwcHBzriFiIiIiIiIyzgzjPH397d3YqSl7hhbGFO/fn0WLlxIyZIlOXPmDF26dEnSerY9Y1JDZ4zFYokzqkxca/z48fZfZ8TXHE6cOMHEiRMBWLBgAWfOnOHChQucPHmSLl26YIzhm2++oWzZsuzfvz/OtaNHjyYkJITcuXOzZ88e2rVr54qnkGB+fn5MnDiROnXquLoUERGRZEtSGPPXX38xYMAAVq9efd+xSZMmUa5cOQYNGsRbb71Fo0aNePrpp4mOjk52sSIiIiIiIqmBM8MYSHujyqKjo9m6dSsAderUIXv27Pz22294enqyevVqe1CTUFarlYsXLwKpI4wByJkzJ6DOGFfbv38/K1eutP8+ODgYY4wLK0p577zzDlFRUTRu3JimTZvaH8+TJw/fffcdq1evplSpUly8eJGgoCDOnj0LwMGDB/n4448BmDBhgv1zWkRERFJGksKYqVOn8vXXX1OoUKE4jx87doyBAwdisVgYNGgQX3zxBWXLluWPP/5g3LhxDilYRERERETE1RTGxLVnzx5u3bpFQEAApUqVAqBEiRJ07doVgE8++SRR6129epWYmBiAVDM+ydmdMVarlRMnTmS4YCGxJkyYAECzZs3w8PDg9OnTGWp/23/++Ye5c+disVgYO3YsFovlvnOeeuopNm7cSOnSpTl79izNmzfn2rVrvPzyy0RHR9OyZUs6duzogupFREQytiSFMTt27KBy5coUL148zuPTp08nJiaGjz76iM8++4yBAweybt06smfPzty5cx1SsIiIiIiIiKs5O4wpVqwYkHbCGFvnS61atXBz+9+PmUOGDMFisbBkyRIOHjyY4PVsI8py5cqFp6enY4tNIlsY46zOmE8//ZSiRYsyc+ZMp6yfHly9epU5c+YAMHz4cGrWrAlknFFlxhiGDBkCQJcuXahSpcoDz82RIwfLly8nX7587Nu3j4oVK7JhwwZ8fX35+uuv4w1xRERExLmSFMZcuHDB/m6ne61evRo3Nzd69uxpfyxbtmy0bt2aY8eOJb1KERERERGRVESdMXHZwpjatWvHebx06dK0adMGuBs2JNSFCxeA1DOiDJw7pswYw9SpUwHs/5X7TZ06ldu3b1OlShWeeOIJGjRoACQujJk5cyYVKlTg999/d06RTrRs2TKCg4PJlCkTH3744SPPL1y4MH/88Qd+fn6cO3cOgFGjRlG4cGFnlyoiIiLxSFQY4+bmhru7O5GRkfz888+4u7vH+di2bRtWq5VcuXLFeXzmzJncvn0bd3d3+xoiIiIiIiJp1e3btwGFMTYPCmMA3n77bQC+//57+94Vj5IawxhnjinbvXu3/Q2MmzdvTvCfU0YSHR1t37R+0KBBWCyWOGHMo8a7RURE0KNHD1566SX27duX5kKvmJgY3nrrLQAGDhyY4EClYsWKLF68GD8/Pxo2bEj//v2dWaaIiIg8RKLCGKvVSmxsLJkyZSIoKIjY2Fj7x/r16zHG0KtXrziPx8bGMmLECPz8/IiNjbWvISIiIiIiklalVGfMyZMnU/3PTxcvXuTEiRNYLBYef/zx+47XqlWL+vXrEx0dzRdffJGgNW1hTL58+Rxaa3I4szNm4cKFcX7/yy+/OPweadHOnTsZO3Ys3bt3p3r16pw7d47cuXPTqVMnAOrUqZOgfWP2799PjRo1mD17tv2xHTt2OLt8h5o9ezYHDhwge/bsDBs2LFHXPvXUU1y4cIE///xTb44VERFxoSSNKStdujSrV69m//799scmTZqExWKhWbNm951/9OhRChQokPQqRUREREREUhFnhzEFChTAw8ODqKgo+/4pjnTixAn69+/PoUOHEnXdlStX6NixIy+99BLR0dHA/7piypcvj5+fX7zX2bpjpkyZwh9//MGpU6ewWq0PvI/tOafGzhhHhzHGGObPnw9gD7P+G85kRJcvX6ZWrVq8/fbbzJkzhz179gB39yGyfd35+vpSo0YN4MGjymbPnk3NmjU5ePAgefPmZenSpQCcOXPGKV1OzhAREcGIESMAeOeddwgICEj0GpkzZ46zn5OIiIikPI+kXPTKK6/Qt29f6tSpQ+PGjTl79izbtm2jWLFi9nnANlFRUfz55580btzYIQWLiIiIiIi4mrPDGHd3dx577DGOHj3K8ePHKViwoMPWvnr1Kk2bNuXo0aMcPXqUP/74I0HXnTx5kqCgIEJCQoC7+4N+9tlnDx1RZhMUFETFihXZs2cPLVq0AMDHx4dSpUpRunRpSpcubf91yZIlM9SYsgMHDhASEoKXlxfTpk2jQoUKrF+/nkuXLpE7d26H3ist2bRpE1FRUeTOnZt+/fpRtmxZKlasSIkSJeKc16BBAzZt2sTatWt58cUX7Y9HRkby2muvMXPmTACaNGnC999/T2BgICVKlODIkSPs3LmTJk2apOjzSoovvviCCxcuUKRIEV599VVXlyMiIiJJlKS3Rbzyyiu89dZb3L59m0WLFrFt2zaKFy/OL7/8godH3HxnxowZXLlyhXbt2jmkYBEREREREVdzdhgDztk35s6dO7Rr146jR48CsHLlSvvG3g+ze/duateuTUhICLly5QLg888/Z968eQkKYywWCzNmzKB9+/aULVsWT09Pbt26xa5du/jpp5947733eO6556hSpQq+vr4sWbIEyBhjymxdME2bNqV8+fJUr14dYwyLFy926H3Smi1btgDQsmVLRowYQfv27e8LYoA4+8bYHDx4kJo1azJz5kzc3Nz48MMPWb58OYGBgQBUrVoVSBujykJDQ/nkk08A+Pjjj536PUdEREScK8k9qmPGjCE0NJS///6bPXv2EBISQoUKFe47r1atWqxZs4a2bdsmp04REREREZFUIy2GMbY9PtevX4+fnx9ly5bFarUyZ86ch163evVqnnzySS5evEiFChXYuXOnfezYSy+9xLZt24CHhzEA1apVY8GCBezfv5/IyEgOHz7Mr7/+yqeffkrPnj2pW7cu2bNnByA2NhaLxUL58uUd8Mwdw1mdMQsWLACgQ4cOALRv3x7QqLKtW7cCxLsP0b1s+8acOnWKkydP8t1331G9enX2799Pnjx5WLVqFe+8806cEV1pKYx5//33uXnzJtWrV+fZZ591dTkiIiKSDEkaU2bj7+9PrVq1HnpO5cqVk3MLERERERGRVCethTHGGN59912+//573N3dWbBgAWfPnuWll15ixowZDB06FIvFct91P//8M127diU6Opr69euzePFiAgICGDVqFNu2bWP16tUAZM+enZIlSya4Hg8PD0qUKEGJEiVo1apVnGNXrlwhJCQEX1/fRK3pbLbOmMjISG7fvo23t3ey1zx8+DB79+7Fw8OD1q1bA3fDmGHDhrFmzRquXbtmD6gyEqvVag9jatas+dBzs2TJQo0aNdi0aRNt27Zl9+7dADRq1Igffvgh3lFvaSWMCQkJYfLkyQB8+umn2vNFREQkjdO/5CIiIiIiIomUlsKYM2fO0LJlS0aNGgXApEmTaNKkCR07dsTX15ejR4+ycePG+6778ssv6dy5M9HR0XTo0IHly5fbNw738PDgxx9/JH/+/MDdrpj4wpykyJkzJ3Xr1k11b+zz8/Ozj+V21KgyW/dLo0aNyJYtGwAlSpSgQoUKxMTE8OuvvzrkPmnN4cOHCQ8Px8fHJ0HdUbZRZbt378ZisfD++++zYsWKB+65U6VKFQCOHj1KWFiYw+p2tGHDhhEbG8vTTz9tf44iIiKSdimMERERERERSaSUDGOOHTuWpOuNMUydOpVy5crxxx9/kClTJr744gt69+4N3O0osI09sm1yDne7Et566y1ef/11APr3789PP/10XydIYGAgv/76K02bNmXw4MFJqjEtsVgs9i4VR40qs40os40ms7GNLMuoo8ps+8VUq1btvn1p49OyZUsAcufOzapVq3j33Xdxd3d/4Pk5cuSgUKFCAOzatSv5BTvBxo0bWbRoEW5ubvY9Y0RERCRtUxgjIiIiIiKSSCkZxly+fJkbN24k6tq//vqLOnXq8PLLL3Pjxg1q1arFzp07GTRoUJzzXnzxReDuOLKbN28SFRVFt27d+PTTT4G7e4WOHz/+gS9sV61alRUrVmSYd+3bRpU5ojPm7Nmz7NixAzc3t/v2WLWFMytXruTs2bPJvldaYwtjHrVfjE3dunX5559/OHjwIA0bNkzQNal5VJkxhiFDhgDQs2dPypYt6+KKRERExBEUxoiIiIiIiCRSSoQx/v7+9k3jT5w4kaBrNm7cSMOGDWncuDGbN2/Gx8eHzz77jA0bNlCmTJn7zq9Xrx7FixcnIiKCmTNn8vTTT/PDDz/g4eHB7Nmzefvttx02fiw9sP19OCKMOX36NACFCxcmV65ccY6VLVuW2rVrExUVRa9evTDGJPt+aUlC94u5V7Vq1eyj3hIiNYcxv/zyC5s2bSJz5sy89957ri5HREREHERhjIiIiIiISCLZwhhHbOL+MAnZNyY6Opqff/6Z2rVrU69ePdasWYOXlxcDBgzg+PHjvPHGGw/sbLFYLPbumAEDBvDnn3/i6+vL0qVL6datm+OfUBpn64xxxJgy214l/v7+9x2zWCxMnz6dTJkysWLFCqZOnZrs+6UVt27dYvfu3UDCO2OSwhbG7Ny502n3SIro6GiGDh0KwJtvvkm+fPlcXJGIiIg4isIYERERERGRREqJzhh4eBhz9epVRo8eTZEiRejcuTObN2/G09OTXr16ceTIEcaPH0+ePHkeeY9u3brh5nb3R8NcuXIRHBxMUFCQY59IOuHIzpjw8HAg/jAGoEyZMnz88cfA3RflE9odldbt2rWLmJgYAgMD7fu6OIMtjDl48CCRkZFOu09iTZkyhaNHjxIYGGgfVSYiIiLpg8IYERERERGRRHJlGLNv3z569+5NgQIFGD58OOfOnSMwMJCRI0dy+vRppk6dmqgXsQsUKMBbb71FvXr1+Pvvv6levbrDn0d6YQtjHNkZ4+fn98BzBg4cyBNPPMHNmzd58cUXsVqtyb5vanfvfjHOHJGXN29e8uTJg9VqZc+ePU67T2J99dVXAIwcOZKsWbO6uBoRERFxJIUxIiIiIiIiiZTSYczRo0dZunQpjRs3pkKFCkybNo3bt29TpUoVZs+ezenTp3nvvfcS1AkTn9GjR7N+/XqKFy/uyPLTHduYspTojAFwd3dn5syZZM6cmbVr19pfqE9LjDEcPHiQqKioBJ1v2y/GmSPKbFLbvjE3b97k8OHDAHTo0MHF1YiIiIijKYwRERERERFJpJQOY1asWEHr1q3566+/cHNzo3379qxbt47t27fTrVs3p9chdzlyTNnD9oy5V7FixRg3bhwAQ4cOJSQkJNn3Tknvv/8+ZcuWJU+ePLzyyiusXbv2oR0+ts6YmjVrOr22KlWqAPGHMcYYvv/+e1566SVCQ0OdXgvA/v37McaQO3duAgMDU+SeIiIiknIUxoiIiIiIiCSCMSbFwphSpUrZ93MJCAhgyJAhHD9+nAULFvDEE084dYyT3M/WGeOIMWW2zpiHjSmz6dOnD02aNOH27dv06NGDmJiYZN8/JezZs4ePPvoIgOvXrzNlyhQaNGhAoUKFGDJkCDt37sQYYz//ypUr9pF8NWrUcHp9D+qM2bt3L/Xr16dr167MnDmTOXPmOL0W230BKlasmCL3ExERkZTl4eoCRERERERE0pKYmBj7C8jODmPy58/PvHnzCAsLo1OnTvj6+jr1fvJwruiMAbBYLEyfPp3y5cuzefNmxo0bx9ChQ5NdgzPFxsbSu3dvYmJiaNeuHa+99hpz585lwYIFnDt3jnHjxjFu3DhKly7Nc889x3PPPceRI0eAuyFkQECA02u0hTH79u1j7Nix3Lp1i9OnTzN79mxiY2Pt5+3atcvptQD2vWsUxoiIiKRP6owRERERERFJBFtXDDg/jAFo3749L730koKYVMAWxqR0ZwxAwYIFGT9+PHB3c3dbF0Vq9fXXX7N161b8/PyYOHEiDRs2ZNq0aVy6dIlFixbRsWNHvL29OXToECNHjqRkyZJ069YNSJn9YgAKFy5Mjhw5iI6O5u233+a9995jxowZxMbG8swzzzBx4kTgfyGJsymMERERSd9SZRgzadIkKlasiJ+fH35+ftSuXZs//vgDgGvXrtG/f39KlSqFj48PhQoVYsCAAfZ3FT1Ijx49sFgscT6CgoJS4umIiIiIiEg6ktJhjKQetjFl4eHhREdHJ2utxHTG2HTv3p1WrVoRFRVF9+7diYqKSlYNznL69GmGDx8OwCeffEK+fPnsxzJlykTbtm2ZN28ely5dYvbs2TRr1gw3Nzd7x1GtWrVSpE6LxcKUKVPo1KkT3bp145VXXuH1119n5cqVLFy4kFatWgFw8ODBOF/3zmCMsYcxFSpUcOq9RERExDVS5ZiyAgUKMGbMGEqUKIExhtmzZ9OmTRv7PNnz588zbtw4ypYty6lTp+jTpw/nz59nwYIFD103KCiImTNn2n+vH5xERERERCSxbC/Kuru74+7u7uJqJCUFBARgsVgwxnDt2jVy586d5LUS2xkD/wsPypUrx86dO/noo494//33k1yDMxhjePXVV4mIiKBevXq8/PLLDzzXz8+Pbt260a1bNy5dusS8efM4d+4cPXr0SLF6n3nmGZ555pl4jxUsWJCAgAD+/fdfDh06RKVKlZxWx7lz57h+/Tru7u6UKVPGafcRERER10mVYYzt3Sc2H330EZMmTWLz5s307NmThQsX2o8VK1aMjz76iC5duhATE4OHx4OfUqZMmciTJ4/T6hYRERERkfTPFsbozV0Zj7u7O9mzZ+fq1atcuXIlWWFMUjpjAPLkycM333xD586d+eijj2jVqhXVq1dPch2OtmPHDn7//Xc8PT2ZMmUKbm4JG8iRO3du+vfv7+TqEsdisVCxYkXWrVvH7t27nRrG2LpiSpUqhbe3t9PuIyIiIq6TKseU3Ss2NpaffvqJiIgIateuHe85YWFh+Pn5PTSIAQgODiYwMJBSpUrRt2/fR266eOfOHcLDw+N8iIiIiIhIxqYwJmOz7RvzqJ8nHyUpnTE2nTp14tlnnyU2Npbu3btz+/btZNXiSHPnzgWgXbt26aLDw7Z/i7P3jbHtAaT9YkRERNKvVBvG7N27lyxZspApUyb69OnDokWLKFu27H3nXblyhQ8//PChrc9wd0TZnDlz+Ouvv/jkk09Yu3YtzZs3JzY29oHXjB49Gn9/f/tHwYIFk/28REREREQkbVMYk7E5KoxJameMzddff03u3Lk5cOAA7777brJqcRTbmykBnn/+eRdX4xi2bpjdu3c79T62sEdhjIiISPqVasOYUqVKsWvXLrZs2ULfvn3p3r07Bw4ciHNOeHg4LVu2pGzZsrz33nsPXa9z5860bt2aChUq0LZtW3777Te2bdtGcHDwA68ZNmwYYWFh9o8zZ8444JmJiIiIiEhapjAmY8uZMydw942BSRUVFWXvZklKZ4ytjilTpgAwbtw4Nm7cmOR6HGX9+vWcP3+egIAAgoKCXF2OQ6RUZ4zCGBERkfQv1YYxXl5eFC9enGrVqjF69GgqVarE+PHj7cdv3LhBUFAQWbNmZdGiRXh6eiZq/aJFi5IzZ06OHj36wHMyZcqEn59fnA8REREREcnYFMZkbI7ojLl3BHZyfs5s3bo13bt3xxjDa6+9luR1HMU2oqx9+/bp5uujfPnyuLm5ERoaysWLF51yjzt37nDo0CFAYYyIiEh6lmrDmP+yWq32H3rCw8Np2rQpXl5e/Prrr0na3O7s2bNcvXqVvHnzOrpUERERERFJxxTGZGyO6IyxjSjz9fV95N6nj/Lxxx8Dd8douXLvmKioKBYsWACknxFlAJkzZ6ZEiRKA87pjDh06RExMDP7+/hQoUMAp9xARERHXS5VhzLBhw1i3bh0nT55k7969DBs2jODgYF544QV7EBMREcH06dMJDw/n4sWLXLx4Mc7+L6VLl2bRokUA3Lx5kyFDhrB582ZOnjzJX3/9RZs2bShevDjNmjVz1dMUEREREZE0SGFMxubIzhhHTF/Imzcv/v7+GGM4duxYstdLqhUrVnD9+nXy5s1L/fr1XVaHM9i6VZy1b8zevXvt97FYLE65h4iIiLheqgxjQkND6datG6VKlaJRo0Zs27aNFStW0KRJE3bs2MGWLVvYu3cvxYsXJ2/evPaPe/d0CQkJsb/byN3dnT179tC6dWtKlixJz549qVatGuvXr9cPUCIiIiIikigKYzI2R4Qxtp9V/f39k12PxWKxd24cPnz4vuOXL19Okf1Pf/zxRwA6deqEu7u70++XkipVqgQ4L4zRfjEiIiIZQ/L6oZ1k+vTpDzzWoEEDjDGPXOPec3x8fFixYoVDahMRERERkYxNYUzG5ogxZY7sjAEoWbIk//zzz31hjDGGJ598kvPnz3Pw4EHy5cvnkPv9V0REBEuWLAHS14gyG1tI4qwxZQpjREREMoZUGcaIiIiIiIikVgpjMjZbZ0xoaGiS13BkZwzcDWMAjhw5Eufx8+fP2zeGX7BgAQMGDHDI/UJCQujUqRPe3t5UqFABgMjISIoXL0716tUdco/UxNYZc/DgQe7cuePwr32FMSIiIhmDwhgREREREZFEUBiTsZUuXRqLxcKxY8c4e/ZskjZcd3RnzIPGlNn2IgGYN2+eQ8KYO3fu0KlTJ/vIri1bttiPPffcc+lyz5OCBQsSEBDAv//+y6FDh+zhjCNcvnyZCxcuAFC+fHmHrSsiIiKpT6rcM0ZERERERCS1UhiTsQUGBlK7dm0Afv311ySt4azOmIeFMRs3bnTI3jHDhg1j9+7d5MyZk++++4533nmHNm3aEBQURL9+/ZK9fmpksVjsXSuO3jfG9ndUrFgxsmTJ4tC1RUREJHVRGCMiIiIiIpIICmOkbdu2APZ9UhLLWZ0xly5dsq8NsG/fvjjnLViwIFn3Wb58OV988QUAs2bNokuXLnz44YcsXryYP/74g9y5cydr/dTMWfvG/Pzzz3HWFxERkfRLYYyIiIiIiEgiKIyRNm3aALBmzRp7l0tiOLozxt/fn8DAQCDuvjG2rovGjRsDd0eVJdWlS5fo3r07AP3796dly5ZJXistso0mc2RnzMKFC5kyZQoAffv2ddi6IiIikjopjBEREREREUkEhTFSsmRJypQpQ3R0NH/88Ueir3d0Z4ytJvjfqLKYmBgOHDgAwMiRI7FYLGzevJlTp04lem2r1UqPHj0IDQ2lfPnyjB071mF1pxX3jikzxiR7vRMnTtCzZ08A3nrrLZo0aZLsNUVERCR1UxgjIiIiIiKSCApjBP7XHbN48eJEX+vozhj4Xxhj64w5evQod+7cwdfXlzp16vDkk08CSRtV9tVXX7F8+XK8vb356aef8Pb2dljdaUWFChXImjUrly9fTtLf+b2ioqLo3LkzYWFh1KpVi1GjRjmmSBEREUnVFMaIiIiIiIgkgsIYgf/tG7Ns2TL750RC2TpjHBnG2PaNsXXG2PaLKVeuHG5ubjz77LPA//YoSahdu3bx1ltvAfD5559Trlw5R5Wcpvj4+DBw4EDgbqeR1WpN8lrvvPMOW7duJSAggJ9++glPT09HlSkiIiKpmMIYERERERGRRFAYIwA1atQgb9683Lhxg+Dg4ERda+uMceaYMtt+MRUqVACgffv2uLm5sW3bNk6cOJGgNSMjI3nuueeIioqidevW9OnTx2H1pkVvvPEGfn5+7N27l19++SVJa/z77798/vnnAMyYMYPChQs7skQRERFJxRTGiIiIiIiIJILCGAFwc3OjdevWACxZsiRR1zqjM+beMWXGGHsYU758eQBy585NgwYNAJg/f36C1nzjjTc4dOgQefPmZfr06VgsFofVmxZly5aN119/HbjbHRMbG5voNfbv309sbCwFCxakXbt2ji5RREREUjGFMSIiIiIiIomgMEZsbPvGLFmyJFFjq5zRGVOsWDHgbufFlStX7uuMAeyjyubNm/fI9RYtWsTkyZOxWCx899135MyZ02G1pmWDBg0iICCAAwcOJDjUupdtfJwtJBMREZGMQ2GMiIiIiIhIIiiMEZuGDRuSJUsWzp8/z/bt2xN0jTHGKZ0xPj4+FCpUCLi7z8uxY8eAuGHMM888g7u7O9u3b7cfj8/Zs2fp1asXAG+99RaNGjVyWJ1pXUBAAG+++SYA77//fqK7Y+7dy0dEREQyFoUxIiIiIiIiiXD79m1AYYzc/Rxo3rw5AIsXL07QNZGRkfYX8B3ZGQP/G1W2ZMkSjDHkypWLwMBA+/FcuXLRsGFD4MHdMbGxsXTt2pVr165RvXp1PvjgA4fWmB4MGDCA7Nmzc+jQIX766adEXbt//35AnTEiIiIZkcIYERERERGRRFBnjNyrbdu2QMLDGNuIMjc3N3x9fR1aS4kSJYC7I8YgbleMzaNGlY0dO5bg4GB8fX2ZO3cuXl5eDq0xPfDz82Pw4MHA3e6YmJiYBF+rzhgREZGMS2GMiIiIiIhIIiiMkXu1aNECDw8PDhw4wJEjRx55vm1EmZ+fHxaLxaG12Dpjzp8/D8QfxrRr1w53d3d27drF4cOH4xzbsmULI0aMAGDixIn2cEfu99prr5EzZ06OHDnC3LlzE3RNaGgoly9fxmKxUKZMGSdXKCIiIqmNwhgREREREZFEsIUx3t7eLq5EUoOAgAAaNGgA3B0P9ii2zhhH7hdj89/wJL4wJkeOHDRu3Bggzgb04eHhPP/888TGxtK5c2e6d+/u8PrSk6xZszJkyBAAPvjgA6Kjox95jW1EWdGiRR3eFSUiIiKpn8IYERERERGRRFBnjPyXbVRZQsKYeztjHM3WGWPzoH1J4htV9tprr3H8+HEKFy7MpEmTHN61kx7169ePXLlycezYMb777rtHnq8RZSIiIhmbwhgREREREZFEUBgj/9W6dWsANm7cSGho6EPPdWZnzGOPPYaHh4f99w960b9du3Z4enqyZ88eDh06xA8//MB3332Hm5sbP/zwAwEBAQ6vLT3y9fVl6NChAHz44YeP7I6xdcY8KCQTERGR9E1hjIiIiIiISCIojJH/KliwINWqVcMYw2+//fbQc22dMc4IYzw9PSlSpAhwdxRWlixZ4j0vW7ZsNGnSBIBPPvmEvn37AjBy5Ejq1q3r8LrSsz59+pA7d25OnjzJrFmzHnqurTNGYYyIiEjGpDBGREREREQkERTGSHzatGkDwOLFix96nq0zxhljyuB/o8ri2y/mXrZRZbNmzeLGjRvUq1eP4cOHO6Wm9Cxz5swMGzYMgFGjRtm/P/yXMUZjykRERDI4hTEiIiIiIiKJoDBG4mPbN+bPP/8kIiLigec5szMGoGrVqgA8/vjjDz2vTZs2eHp62mv5/vvv44w4k4R7+eWXyZcvH6dPn2bGjBnxnnP+/HnCwsJwd3enVKlSKVyhiIiIpAYKY0RERERERBJBYYzEp3z58hQpUoTbt2+zcuXKB57n7M6Yt956i0WLFjFo0KCHnhcQEECnTp1wc3Nj6tSpFC5c2Cn1ZAQ+Pj727piPP/6Y27dv33eOrSumRIkS+t4hIiKSQSmMERERERERSQSFMRIfi8Vi745ZsmTJA89zdmdMlixZaNu2LT4+Po88d/r06Zw5c4aOHTs6pZaMpFevXhQoUICzZ88ybdq0+45rvxgRERFRGCMiIiIiIpJAVquVmJgYQGGM3M8WxixdutT+efJfzu6MSQwvLy/y5cvn6jLSBW9vb/7v//4PuNsdc+vWrTjH9+/fDyiMERERycgUxoiIiIiIiCTQvZtzK4yR/6pTpw45cuTg2rVrbNiwId5znN0ZI67z0ksvUahQIS5cuMCUKVPiHLN1xpQrV84VpYmIiEgqoDBGREREREQkgRTGyMN4eHjQqlUr4MGjylJTZ4w4lpeXF++88w4Ao0ePJjIyErjbUafOGBEREVEYIyIiIiIikkD3hjGenp4urERSqzZt2gCwePFijDH3HbeFMeqMSZ969OhBkSJFuHTpEp9//jkAp06dIjIyEi8vL4oXL+7iCkVERMRVFMaIiIiIiIgkkC2MyZQpExaLxcXVSGrUtGlTfHx8OHnyJHv37r3vuG1MmTpj0idPT09GjBgBwIgRI2jXrh3Lli0DoHTp0nh4eLiyPBEREXEhhTEiIiIiIiIJdG8YIxKfzJkz06RJE+Bud8x/qTMm/evevTtDhw7Fw8ODxYsX89prrwEaUSYiIpLRKYwRERERERFJIIUxkhBt27YF7t83JjY2loiICEBhTHrm5ubG6NGj2blzJ0888YT9cYUxIiIiGZvCGBERERERkQRSGCMJ8fTTT+Pm5saOHTs4ffq0/XHbiDLQmLKMoHz58qxdu5bZs2fzwgsv0L17d1eXJCIiIi6kMEZERERERCSBFMZIQuTKlYu6desC8Ouvv9oft4Ux3t7eeHl5uaQ2SVkWi4Vu3brx/fffky9fPleXIyIiIi6kMEZERERERCSBFMZIQrVp0waIu2+Mbb8YdcWIiIiIZDwKY0RERERERBJIYYwklC2MWbt2LdevXwf+1xmj/WJEREREMh6FMSIiIiIiIgmkMEYSqnjx4pQvX56YmBiWLVsGqDNGREREJCNTGCMiIiIiIpJACmMkMWzdMT/++COgzhgRERGRjExhjIiIiIiISAIpjJHE6NatGxaLhd9//529e/eqM0ZEREQkA1MYIyIiIiIikkAKYyQxSpYsSYcOHQAYM2aMOmNEREREMjCFMSIiIiIiIgmkMEYSa9iwYQD89NNP7NixA1BnjIiIiEhGpDBGREREREQkgRTGSGJVqVKF5s2bY7VamTdvHqDOGBEREZGMSGGMiIiIiIhIAimMkaQYPnw4AMYYQGGMiIiISEakMEZERERERCSBFMZIUtSrV4969erZf68xZSIiIiIZj8IYERERERGRBFIYI0ll644BdcaIiIiIZEQKY0RERERERBJIYYwkVVBQEDVr1gSgZMmSLq5GRERERFKah6sLEBERERERSSsUxkhSWSwWVqxYwdGjR6lSpYqryxERERGRFKbOGBERERERkQRSGCPJERAQQPXq1V1dhoiIiIi4gMIYERERERGRBFIYIyIiIiIiSaEwRkREREREJIEUxoiIiIiISFIojBEREREREUkghTEiIiIiIpIUCmNEREREREQSyBbGeHt7u7gSERERERFJSxTGiIiIiIiIJJA6Y0REREREJCkUxoiIiIiIiCSQwhgREREREUkKhTEiIiIiIiIJpDBGRERERESSQmGMiIiIiIhIAimMERERERGRpFAYIyIiIiIikkC3b98GFMaIiIiIiEjiKIwRERERERFJIHXGiIiIiIhIUqTKMGbSpElUrFgRPz8//Pz8qF27Nn/88Yf9+O3bt+nXrx85cuQgS5YstG/fnkuXLj10TWMM7777Lnnz5sXHx4fGjRtz5MgRZz8VERERERFJRxTGiIiIiIhIUqTKMKZAgQKMGTOG7du3888//9CwYUPatGnD/v37AXj99ddZunQp8+fPZ+3atZw/f55nnnnmoWuOHTuWCRMm8O2337JlyxZ8fX1p1qyZfcyAiIiIiIjIwxhjiIyMBBTGiIiIiIhI4liMMcbVRSRE9uzZ+fTTT+nQoQO5cuVi7ty5dOjQAYBDhw5RpkwZNm3aRK1ate671hhDvnz5ePPNNxk8eDAAYWFh5M6dm1mzZtG5c+cE1RAeHo6/vz9hYWH4+fk57smJiIiIiEiqd/nyZQIDAwGIjIzEx8fHxRWJiIiIiIirJTQ3SJWdMfeKjY3lp59+IiIigtq1a7N9+3aio6Np3Lix/ZzSpUtTqFAhNm3aFO8aJ06c4OLFi3Gu8ff35/HHH3/gNXB3BEF4eHicDxERERERyZhCQkIAKFSokIIYERERERFJlFQbxuzdu5csWbKQKVMm+vTpw6JFiyhbtiwXL17Ey8uLgICAOOfnzp2bixcvxruW7fHcuXMn+BqA0aNH4+/vb/8oWLBg8p6UiIiIiIikWYcPHwagVKlSLq5ERERERETSmlQbxpQqVYpdu3axZcsW+vbtS/fu3Tlw4ECK1jBs2DDCwsLsH2fOnEnR+4uIiIiISOph64xRGCMiIiIiIonl4eoCHsTLy4vixYsDUK1aNbZt28b48ePp1KkTUVFR/Pvvv3G6Yy5dukSePHniXcv2+KVLl8ibN2+caypXrvzAGjJlyqSNOUVEREREBPhfZ0zJkiVdXImIiIiIiKQ1qbYz5r+sVit37tyhWrVqeHp68tdff9mPhYSEcPr0aWrXrh3vtUWKFCFPnjxxrgkPD2fLli0PvEZERERERORe6owREREREZGkSpWdMcOGDaN58+YUKlSIGzduMHfuXIKDg1mxYgX+/v707NmTN954g+zZs+Pn50f//v2pXbs2tWrVsq9RunRpRo8eTbt27bBYLAwaNIhRo0ZRokQJihQpwogRI8iXLx9t27Z13RMVEREREZE0ITY2lqNHjwLqjBERERERkcRLlWFMaGgo3bp148KFC/j7+1OxYkVWrFhBkyZNAPjiiy9wc3Ojffv23Llzh2bNmvHNN9/EWSMkJISwsDD779966y0iIiJ4+eWX+ffff6lXrx7Lly/H29s7RZ+biIiIiIikPSdPniQ6Ohpvb28KFSrk6nJERERERCSNsRhjjKuLSCvCw8Px9/cnLCwMPz8/V5cjIiIiIiIpZNmyZbRs2ZIKFSqwZ88eV5cjIiIiIiKpREJzgzSzZ4yIiIiIiIirHD58GNCIMhERERERSRqFMSIiIiIiIo8QEhICQKlSpVxciYiIiIiIpEUKY0RERERERB5BnTEiIiIiIpIcCmNEREREREQeQZ0xIiIiIiKSHApjREREREREHuLmzZucO3cOUGeMiIiIiIgkjcIYERERERGRhzhy5AgAOXPmJHv27C6uRkRERERE0iKFMSIiIiIiIg+hEWUiIiIiIpJcCmNEREREREQe4vDhw4BGlImIiIiISNIpjBEREREREXkIdcaIiIiIiEhyKYwRERERERF5CFtnjMIYERERERFJKoUxIiIiIiIiD2CMsXfGaEyZiIiIiIgklcIYERERERGRB7h06RI3btzAzc2NYsWKubocERERERFJoxTGiIiIiIiIPICtK+axxx4jU6ZMLq5GRERERETSKg9XFyAiIiIiIpJaWK1WBgwYwJEjR8iVKxeXL18GtF+MiIiIiIgkj8IYERERERGR/2/Lli18/fXX9z1etmxZF1QjIiIiIiLphcIYERERERGR/2/Lli0AVK9enU6dOnH58mViYmJ4/fXXXVyZiIiIiIikZQpjRETSsJiYGDw89K1cRETEUTZv3gxA27ZtGTx4sIurERERERGR9MLN1QWIiEjSbNmyhdy5c9O9e3dXlyIiIpJu2DpjHn/8cRdXIiIiIiIi6YnCGBGRNOj48eO0atWKa9eu8cMPP3D16lVXlyQiIpLmhYaGcvLkSSwWCzVq1HB1OSIiIiIiko4ojBERSWOuXbtGixYtuHz5MgCxsbH8/vvvLq5KREQk7bN1xZQpUwZ/f38XVyMiIiIiIumJwhgRkVQsPDycMWPGMG3aNDZv3syVK1do27YtISEhFCxYkN69ewOwaNEiF1cqIiKS9mlEmYiIiIiIOIt2fRYRSaWMMXTr1o0lS5bcd8zPz4/ff/+dmJgYpk6dyooVK4iMjCRz5swuqFRERCR92Lx5M6AwRkREREREHE+dMSIiqdTUqVNZsmQJXl5eNGnShHz58gHg5eXFggULqFChApUrV6Zw4cLcunWLlStXurhi1/vuu+94+umnuX79uqtLERGRNMZqtbJt2zZAYYyIiIiIiDiewhgRkVTo0KFDDBo0CIDRo0ezcuVKzp07x9WrVzl79ixNmjQBwGKx0LZtW0CjygDeffddfv/9d3744QdXlyIiImnMoUOHCA8PJ3PmzJQvX97V5YiIiIiISDqjMEZEJJWJiorihRde4NatWzRu3NgeygBkz56dXLlyxTm/Xbt2ACxdupSYmJiULDVVuXjxIidPngRgzZo1ri1GRETSHNt+MdWrV8fDQ9OcRURERETEsRTGiIikMiNGjGDHjh1kz56d2bNn4+b28G/VdevWJUeOHFy/fp1169alUJWpj23OP0BwcDBWq9WF1YiISFpjC2M0okxERERERJxBYYyISCry999/8+mnnwIwbdo0+z4xD+Ph4UHr1q0BWLx48SPPj4mJ4ejRo5w5cyZZtaY294Yx165dY+/evS6sRkRE0hrbvyMKY0RERERExBkUxoiIpBK3b9+mZ8+eGGPo3r27ffxYQtj2jVm8eDHGGIwxXLhwgeDgYKZMmcLgwYNp3bo1pUuXxsfHhxIlSlC0aFEOHz7spGeT8jZt2gRgHy2zevVqV5YjSTB58mS+/vprjDGuLkVEMpiIiAh7iK8wRkREREREnEHDkEVEUolRo0Zx6NAhcufOzeeff56oa5s0aYKvry9nzpyhYsWKnDx5kps3bz70mpiYGIKDgylZsmRyyk4VYmJi2LZtGwBdunRh1qxZrFmzhtdff93FlUlChYaG0qdPHwB27drFt99+i7u7u4urEpGMYvv27VitVvLly0eBAgVcXY6IiIiIiKRD6owREUkFdu3axSeffALAN998Q/bs2RN1vY+PD08//TQA+/bt4+bNm7i5uVGsWDGaN2/OwIED+frrr/nzzz85deoUgwcPtt83PdizZw+3bt0iICCAvn37ArBu3TpiY2NdXJkk1L59++y/njZtGs8//zxRUVEurEhEMhLbfjG1atVycSUiIiIiIpJeqTNGRMTFYmJi6NmzJzExMbRv355nnnkmSetMmDCBp556irx581KyZEmKFi2Kl5dXvOdWrVoVgJ07dya57tTk3jn/1apVw9/fn7CwMHbu3En16tVdXJ0kxIEDBwAoUqQIZ8+eZd68eYSHh7Nw4UIyZ87s4upEJL3bsGEDoBFlIiIiIiLiPOqMERFxsc8++4wdO3aQLVs2Jk6cmOR1AgMDeeWVV+x7wzwoiAGoUqUKcLejJD10j9j2i6lVqxbu7u48+eSTAKxZs8aVZUki2MKYTp068dtvv5E5c2aWL19Os2bNCAsLc3F1IpKeHTx4kN9++w2AZs2aubgaERERERFJrxTGiIi4UEhICCNHjgTgiy++IE+ePCly3xIlSuDj40NkZCRHjx5NkXs6k60zpnbt2gA89dRTgMKYtGT//v0AlCtXjqZNm/Lnn38SEBDAhg0beOqppwgNDXVxhSKSXo0cORKr1Uq7du2oVKmSq8sREREREZF0SmGMiIiLWK1WevXqxZ07d2jWrBndunVLsXu7u7tTsWJFIO3vG3P58mV7oFSzZk3gf2HMunXriI6OdlltknC2zpiyZcsCUKdOHYKDgwkMDGTnzp08+eSTnDlzxpUlikg6tGvXLubPn4/FYuH99993dTkiIiIiIpKOKYwREXGRSZMmsWHDBnx9fZk8eTIWiyVF71+5cmUg7Ycxtk2Xy5QpQ7Zs2QCoWLEi2bNnJyIign/++ceV5UkChIaGcuXKFSwWC6VLl7Y/XqlSJdavX0/BggUJCQmhXr16HD582IWVikh68+677wLQuXNnKlSo4OJqREREREQkPVMYIyLiAqdOnWLo0KEAjBkzhsKFC6d4DekljLl3vxgbNzc36tevD2hUWVpg64opUqQImTNnjnOsZMmSbNy4kVKlSnH69GmeeOIJdu/e7YoyRSSd2bJlC0uXLsXNzY333nvP1eWIiIiIiEg6pzBGRCSFGWPo06cPN2/epG7durz66qsuqaNKlSoA7Ny50yX3d5T/7hdj07BhQwD++uuvFK9JEue/I8r+q2DBgqxbt47KlSsTGhpK/fr1+fvvv1OyRBFJh0aMGAFA9+7dKVmypIurERERERGR9E5hjIhICvvuu+9Yvnw5mTJlYvr06bi5ueZbcYUKFXBzc+PSpUtcvHgxRe8dGRnJqFGj2LNnT7LWiY2NZevWrUDczhiAJk2aALB69WrmzZuXrPuIc+3fvx+AcuXKPfCcwMBA1qxZQ926dQkLC6NJkyb89ttvKVWiiKQza9eu5c8//8TT09M+qkxERERERMSZFMaIiKSgS5cuMWjQIADee+89SpUq5bJaMmfObH8ncEqPKvv6668ZMWIEderUeWDnijGGiIgILl68yJEjR9i+fTvBwcEsXbqUH374gW+//ZZ33nmHmzdvkjVr1vu6KkqVKsWQIUMAePHFF5Md/IjzPKozxiYgIICVK1cSFBREZGQkrVq1omfPnvz7778pUKWIpBfGGHtXTK9evXjsscdcW5CIiIiIiGQICmNERFLQa6+9xvXr16lSpQpvvvmmq8tx2b4xv/zyCwARERG0aNGCRYsWAXDr1i2+/vprSpYsiYeHB1myZCFv3ryULFmS6tWr89RTT9G6dWu6dOlC3759GTNmDHB3RJm7u/t99/n4449p3LgxkZGRtGvXjmvXrqXck5QES0hnjE3mzJlZsmQJ/fv3x2KxMGPGDMqWLcvixYudXKWIpBd//vkn69evJ1OmTPzf//2fq8sREREREZEMwsPVBYiIZBS//PILCxYswN3dnenTp+Pp6enqkqhSpQo//fRTioYxFy5csO/zEhQUxPLly+nQoQM9e/ZkyZIlhIaGxjnfYrGQNWtW+4efn1+c/wYEBPDKK6/Eey8PDw9++uknqlevzvHjx3nuuedYtmxZvMGNuMbly5e5fPkyAKVLl07QNV5eXkyYMIFnn32WXr16ERISQrt27RgzZgxvv/22M8sVkTTOGMM777wDwKuvvkr+/PldXJGIiIiIiGQU6owREUkB169fp1+/fgC8/fbbVKlSxcUV3WXrjNm5c2eK3fPXX38F4PHHH2fp0qX07NkTq9XK1KlTCQ0N5bHHHuObb77h7Nmz3Lhxg5iYGMLCwjh79iwHDx5ky5YtrFq1il9++YXZs2czfvz4h463ypEjB4sXL8bHx4eVK1cyYcKElHqqkgC2EWWPPfYYvr6+ibq2Xr167Nq1y95lNnToUGbNmuXoEkXShdGjR9OsWTOuXLni6lJcaunSpWzbto3MmTMzdOhQV5cjIiIiIiIZiMIYEREnM8YwcOBALl68SOnSpe1z6lODSpUqAXDkyBFu3ryZIve0jSRr27YtHh4eTJ06lffff5969eoxZ84cDh8+TN++fcmfPz9ZsmTBzS35/1RVqlSJzz77DIAvvviCmJiYZK8pjmELYxIyoiw+3t7ejBs3jrfeegu4u//DsmXLHFafSHrw66+/Mnz4cFauXMnIkSNdXY7LWK1W+7/BAwcOJDAw0MUViYiIiIhIRqIwRkTEySZPnsx3332HxWJh2rRpeHt7u7oku9y5c5M3b16MMezdu9eha1+8eJHp06cTERFhfywsLIzVq1cDd8MYuDuG7N1332X9+vV07drVaePbXnzxRQIDAzlz5gwLFy50yj0k8WxhzMO6mxJizJgxdOvWjdjYWDp27MiWLVscUZ5Imnfu3DlefPFF++8nT57MoUOHXFiR6yxYsIA9e/bg5+fH4MGDXV2OiIiIiIhkMApjREScaP369fTv3x+4u5l83bp1XVzR/Wwj0+LbN8YYQ69evejRowdWqzXBa164cIG6devSq1evOC8C/vHHH0RHR1O6dOkE7w/iKN7e3vTt2xe42x2TELGxsfzyyy88/fTTfPDBB84sL8Pav38/kPTOGBtb2BkUFERkZCRt27YlKirKESWKpFmxsbF07dqVa9euUbVqVVq0aEFsbKy9kywjiYmJ4d133wXgzTffJHv27C6uSEREREREMhqFMSIiTnL27Fk6dOhATEwMHTt2TLUbi9v2jYkvjFm6dCnTp09n9uzZbN68OUHrXbt2jaZNm3L8+HEA5s+fz/z584G4I8pcoW/fvnh5ebFly5aHPp+bN2/y1VdfUbJkSdq3b8/vv//OyJEjOXz4cApWmzE4qjMGwNPTk/nz55MtWzYuXrzInj17kr2mSFr2ySefsGbNGnx9ffnxxx/57LPPcHd3Z+nSpaxZs8bV5aWouXPnEhISQvbs2Rk0aJCryxERERERkQxIYYyIiBPcvn2bZ555htDQUCpUqMDMmTOxWCyuLitetjBmw4YNcbpfjDFxukEWLFjwyLVu3rxJixYt2LdvH3nz5qV3794A9OvXj7Nnz9r38mjXrp0Dn0HC5c6dmxdeeAGIvzvm7NmzvP322xQoUIABAwZw/PhxsmfPTokSJQCYNGlSitab3l29epVLly4BUKZMGYesmSVLFmrWrAnA1q1bHbKmSFpy6dIl5s6dy4svvmjvBJk4cSIlS5akdOnS9OnTB7jbHZKYjse0LDo6mvfeew+At99+Gz8/P9cWJCIiIiIiGZLCGBERJ5g8eTLbtm0je/bsLF68GF9fX1eX9ED169cnS5YsHDhwgG+//db++PLly9m+fbv99wsWLMAY88B17ty5Q9u2bdmyZQvZs2fnzz//ZOLEiVSoUIHLly/TsGFDbt68Sb58+ahevbpTn9PDDBw4EICFCxdy+vRpALZv384LL7xAkSJFGDt2LGFhYZQsWZJJkyZx5swZxo8fD8DMmTPj7IEjyWPriilcuDBZsmRx2LqPP/44gPaNkQwhMjKSFStWMHjwYCpVqkSePHl44YUXmDVrFrGxsXTp0oXu3bvbzx85ciR+fn7s3LmTL774IkN8T5s5cyYnTpwgd+7c9OvXz9XliIiIiIhIBqUwRkTECYKDg4G778AtWrSoa4t5hMDAQEaPHg3crff06dNxumL69euHr68vZ86cYdu2bfGuERMTw3PPPcdff/2Fr68vy5Yto1y5cnh5eTFr1izc3d05cuQIAG3atMHNzXX//FSqVImGDRsSGxvLgAEDaNCgAdWrV2fu3LnExMTQoEEDfv31Vw4ePEifPn3InDkzzZo1o1ixYoSFhfHjjz+6rPb0xpEjyu6lzhhJz2JjY9m2bRujR4+mYcOGZMuWjaCgID777DP7aL7KlSszZMgQVqxYwezZs+N0ZubKlYvhw4cDMHjwYPz9/alevTpvv/02N27ccMlzcqbbt2/z4YcfAjB8+PBU/eYIERERERFJ3xTGiIg4mDGGv//+G4B69eq5uJqEefXVV6lbty43b96kb9++rFq1is2bN+Pt7c0777zD008/DcQ/qsxqtdK7d28WLVqEl5cXS5YssXcmAFStWtX+wh+4bkTZvV5//XUAlixZwtq1a/Hw8KBLly5s376dNWvW0KpVqziBkZubG3379gXg66+/fmiHkCTc/v37AShXrpxD17WFMYcOHeLff/916NoZgT6/U6eIiAjGjRtH/vz5qVmzJsOHD2fNmjVERUVRsGBBXnrpJX788UcuXbrEzp07GTt2LE2bNo03/B40aBADBgygYMGCxMbGsn37dsaOHZsu91KZMmUKZ8+epUCBArz88suuLkdERERERDIwi9FP3AkWHh6Ov78/YWFhmjUtIg90/PhxihUrhqenJ+Hh4Xh7e7u6pAQ5dOgQlSpVIioqisDAQEJDQxkwYADjx49nwYIFdOzYkSJFinDs2DH7u6yNMbz55pt88cUXuLm5sXDhQtq2bXvf2lFRUbRq1Yo7d+6wcuVKvLy8UvjZxWW1WmnYsCF79+7l5Zdf5rXXXiN//vwPvebatWvkz5+f27dv8/fff1O7du0Uqjb9KleuHAcOHGDu3Lk899xzDl27aNGinDhxgj///JPGjRs7dO30bOjQocyZM4cJEybQoUMHV5cj3A1hvvnmGz799FMuX74MgJ+fHw0bNqRx48Y0adKEEiVKJHlfsjNnzrBs2TL69OmDxWJh9+7dVKhQwZFPwWUiIiIoVqwYly5dYvLkyQpjRERERETEKRKaG6gzRkTEwWxdMdWqVUszQQxA6dKlGTFiBAChoaF4eXnx1ltvAdC8eXN8fHw4ceIEO3futF8zatQovvjiCwBmzJgRbxAD4OXlxYoVKwgODnZ5EAN3O12Cg4O5evUqo0ePfmQQA5A9e3Z7YPD11187u8R079ChQxw4cABPT0+aN2/u8PW1b0ziBQcH88knn3DhwgU6duzIJ598oi6ZVOC5557jrbfe4vLlyxQtWpSZM2dy5coVFi1aRL9+/ShZsmSSgxiAggUL8sorr9ChQweMMbz99tsOrN61pk6dyqVLlyhatCgvvviiq8sREREREZEMTmGMiIiDbdq0CSBNdk689dZb9ndE9+zZ0x5S+Pr60rJlS+B/o8omTpzIu+++C8CXX34ZZ4Po9OrVV18FYP78+YSGhrq4mrRt4cKFADRu3JiAgACHr699YxLnzp07vPLKKwAUL14cuNsl07t3b6KiolxZ2iMZY/j999+pU6cOFSpU4Ny5c64uyWGuXLnCb7/9BsC0adM4dOgQPXr0wNPT0+H3Gj16NJ6envzxxx/89ddfDl/fFZYvXw7AgAEDnPJnJiIiIiIikhgKY0REHMzWGVOnTh0XV5J4Xl5eLF26lFGjRjFmzJg4x2wji+bPn8/3339P//79AXj33XcZOHBgitfqCtWrV6dmzZpERUXx7bffurqcNM0WxrRv394p69/bGaPujkcbM2YMhw8fJnfu3Gzbto2vvvoKNzc3pk+fTuPGjTl58mSy1p80aRL9+/dn8ODBDB8+nLFjx3L16tVkrWkLYWrWrMnTTz/Npk2b2LdvH8899xwxMTHJWju1+PPPPzHGULFiRXr27OnUQKF48eL2vbGGDBmC1Wp12r1SQmxsrP3f4wYNGri2GBEREREREbRnTKJozxgReZSbN2/i7++P1Wrl7NmzCRp/lVbcuHGDwMBAbt++jcViwRhD//79GT9+fLJG5KQ1P/30E8899xz+/v6cOHGCbNmyubqkNMe2r5K7uzsXL14kZ86cDr/HrVu38PPzIyYmhlOnTlGoUCGH3yO9OHz4MBUqVCAqKoqffvqJTp06AbBs2TI6derEzZs3yZIlC1988QU9e/ZM9Nf7pk2b4g2n+/Xrx8SJE5Nc95dffsnrr78OQObMmenZsyezZs3ixo0bvPPOO3z44YdJXju16N69O3PmzGHIkCGMHTvW6fe7cuUKxYoVIzw8nO+++44uXbo4/Z7OsmvXLqpUqYKfnx/Xrl3D3d3d1SWJiIiIiEg6pT1jRERcYOvWrVitVgoVKpSughiArFmzEhQUBNx9R3rXrl358ssvM1QQA/Dss89Svnx5wsLC+Pzzz11dTpr0yy+/AFC/fn2nBDEAPj4+VKxYEUjcqLLo6Gh27Nhh3+z7s88+c0p9qYUxhj59+hAVFUVQUBDPPvus/ViLFi3YtWsX9erV4+bNm/Tu3Zunn36a8+fPJ+oeM2bMAO6ObhwyZAidO3cGYPHixUnuWrp06ZJ9TGLfvn05ceIEEyZMYMqUKQB89NFHrFq1KklrP8z169dZsWJFinTeWK1WVqxYAWD/3utsOXPmZNiwYQAMHz6cmzdvpsh9nWHDhg3A3S5VBTEiIiIiIpIapMowZvTo0dSoUYOsWbMSGBhI27ZtCQkJsR8/efIkFosl3o/58+c/cN0ePXrcd35K/XArIhlDWt4vJiF69+4NQLt27Zg+fTpubqnynxGncnNz44MPPgDuvjP/8uXLLq4o7XH2iDIb274xW7Zsife41WolJCSE77//noEDB1K7dm38/PyoVq0affr0YerUqQwePJgjR444tU5XGj9+PGvWrMHHx4dvvvnmvnC1WLFiBAcH8+mnn+Ll5cWyZcsoX748P/30U4LWj4iI4Oeffwbu/v/d2LFjmTlzJpkzZ+bcuXPs3LkzSXWPGDGCGzduUL16dSZOnEhgYCAAnTt35uWXX8YYwwsvvMCFCxeStP6DDBw4kKCgIDp37uz0vXT27NnDpUuX8PX1pW7duk69170GDhxI4cKFOXPmDEOGDEmx+zqaLYypV6+eiysRERERERG5K1W+irZ27Vr69evH5s2b+fPPP4mOjqZp06ZEREQAULBgQS5cuBDn4/333ydLliw0b978oWsHBQXFue7HH39MiackIklgjGHJkiUcOnTI1aUkWFreLyYhWrRowaVLl1i4cGGG3gy5bdu2VK1alZs3b6bI6KD05OzZs2zevBm4++foTLZ9Y+7tjDl48CDDhw+ncePGZM+endKlS9O1a1cmTJjA5s2buX37NgEBATRp0oSiRYsCsHLlSqfW6SrLly/nzTffBO4GJUWKFIn3PHd3dwYPHsyOHTuoWrUq169f57nnnqNTp05cuXLlofdYuHAhN27coGjRojz55JMAeHt707RpUwCWLl2a6Lr37NnD9OnTAfj888/vC4W//PJLKlasSGhoKC1atODUqVOJvkd8rFYry5YtA+4+r3bt2nHr1i2HrB0f2+bzDRs2JFOmTE67z3/5+PjYu5m+/fZbe3dOWmKMYf369YDCGBERERERST1SZRizfPlyevToQbly5ahUqRKzZs3i9OnTbN++Hbj7okCePHnifCxatIhnn32WLFmyPHTtTJkyxblOs/5FUq8pU6bQtm1bypcvz2uvvZbszZ6dzWq12l9kTq+dMQCBgYEZbjTZf1ksFkaNGgXAxIkTHf7u+/Rs0aJFwN3AMl++fE69l60z5p9//iEmJobff/+d6tWrM3r0aP766y/CwsLw9vamTp06DBw4kO+//57Dhw9z7do1Vq5cae8ES49hzMGDB+nUqRNWq5UXX3yRAQMGPPKacuXKsXnzZkaOHIm7uzvz5s2jfPny/Pbbbw+8ZubMmQC8+OKLcb5vtG7dGoBff/01UXUbY3jzzTexWq106NCBJ5544r5zfHx8mDdvHjlz5mTXrl1Ur16dtWvXJuo+8dm7dy9Xr17F29sbHx8fli1bRsuWLZ02yssWxriii7thw4b0798fgJ49e3L9+vUUryE5Tp06xfnz5/H09KRGjRquLkdERERERARIpWHMf4WFhQGQPXv2eI9v376dXbt20bNnz0euFRwcTGBgIKVKlaJv374PfXH3zp07hIeHx/kQkZRx69Yt+yio2NhYvv76a4oXL8748eOJjo52cXXxs72I6+PjQ+XKlV1djjhZUFAQderU4fbt23z88ceuLifNSKkRZQClS5cma9asREZG8vbbb9OmTRsiIyN58sknmTx5Mjt37iQ8PJyNGzfy5Zdf8sILL1CiRAl7aGDr3li9enWq/b6TFFevXqVVq1aEh4dTr149Jk2alOCA1dPTk/fee4/NmzdTpkwZLl26RKtWrejZs+d9/590/PhxgoODsVgsdO/ePc6xli1bYrFY2LFjB2fPnk1w7b///jurVq3Cy8uLTz755IHnlSpVin/++YcqVapw5coVGjduzMSJE5O8Rw3c/TwAeOqpp1ixYgVZs2ZlzZo1NG3alH///TfJ68bnxo0bbNy4EYBmzZo5dO2EGjNmDCVLluTcuXMJCutSyv79+6lSpYq9Oyo+thFl1apVI3PmzClVmoiIiIiIyEOl+jDGarUyaNAg6tatS/ny5eM9Z/r06ZQpU+aRY4GCgoKYM2cOf/31F5988glr166lefPmxMbGxnv+6NGj8ff3t38ULFgw2c9HRBLmm2++4fz58xQqVIgVK1ZQqVIl/v33XwYNGkSFChXso2JSE9t+MdWrV8/QI7wyinu7YyZPnsylS5ecdq/w8HBWr16d6HfgG2OIjY3lzp07REZGcuPGDe7cueOkKh8tNDTUPjromWeecfr93Nzc7O+K//zzz4mNjaVbt26sWrWKl19+mcqVKz/0a7Vy5crkzJmTmzdv2rveEuL69etMmzaNo0ePJvs53Ovdd9+lR48eyerEiIiI4JlnnuHYsWM89thj/PLLL0kagVW9enW2b9/OG2+8gcViYcaMGVSoUIE1a9bYz5k1axYAjRs3vu//oQIDA6lVqxbAQztrbIwxBAcH20OBQYMG2cfIPUjhwoXZsGEDzz33HDExMfTv358uXbpw48aNxDxVO1sY07BhQ5544glWrVpFtmzZ2LRpEw0bNnzkyLbE3ismJobixYtTrFgxh62bGJkzZ2b27Nm4ubnx/fff88svv7ikjntZrVZ69uzJrl27+OCDDx4Yrmm/GBERERERSZVMKtenTx9TuHBhc+bMmXiPR0ZGGn9/fzNu3LhEr33s2DEDmFWrVsV7/Pbt2yYsLMz+cebMGQOYsLCwRN9LRBIuLCzM5MiRwwBmxowZxhhjYmJizJQpU0yuXLkMYAATFBRk9u/f7+Jq/6d3794GMG+//barS5EUVL16dQOYSZMmOWX9mJgYU7t2bQMYLy8v07x5czNp0iTz/fffmzfeeMM89dRTJnfu3Mbf399kyZLFeHt7G09PT+Pm5mb/Wrn3w8fHx2zdutUptT7KJ598YgBTo0aNFLvnsGHD7M99+PDhxmq1Jur65557zgDmnXfeSfQ1gHnqqafMDz/8YG7dupXY0uM4f/68fc369eubmzdvJnqN8PBw88QTTxjAZM2a1ezZsydZNdmsXbvWFClSxF7fgAEDzI0bN0zBggUNYH788cd4r/v4448NYFq0aPHAta1Wq1m2bJmpU6eOff18+fKZf//9N8H1Wa1WM27cOOPu7m4AU7JkSbN79+5EPcfo6GiTNWtWA5jt27fbH9+9e7cJDAw0gClbtqw5f/58otZ9kD59+hjAvPbaaw5ZLzlsX0O5cuUyly5dcmktU6ZMifP9bNu2bfGeV65cOQOYxYsXp3CFIiIiIiKSEYWFhSUoN0jVYUy/fv1MgQIFzPHjxx94zpw5c4ynp6cJDQ1N0j1y5sxpvv322wSdm9A/VBFJnvfff98AplSpUiY6OjrOsX///dcMGTLEeHp6GsC4u7ub/v37m6tXr7qo2v/Riz8Z09ixYw1gGjVq5JT1x4wZE2+okpyPvn37OqXWh4mOjjaFChWKE7KmhJCQEPPEE0+YqVOnJun6mTNnGsDUrFkzQedfuXLFeHl5GcBYLBb7n3n27NnNwIEDzb59+5JUx6xZs+L8HT711FMmIiIiwdf/+++/9lDP39/fbNq0KUl1PEh4eLh5+eWX7fXlzp3bACYgIOCBQdS+ffsMYDJlynRfuBQbG2sWLlxoqlatal8zU6ZM5tVXX33gG3QeZcOGDSZ//vwGMN7e3mbq1KkJDuc2b95sAJMtWzYTExMT59jBgwft6xYrVswcOnQoSfXZWK1W89hjjxnA/Pbbb8layxFu375tKlasaADTtm3bRAeaNlar1axbt85s3749UZ+7NqGhoSZbtmz2vwfADB069L7zrl69av+cSerPByIiIiIiIomRpsMYq9Vq+vXrZ/Lly2cOHz780HPr169v2rdvn6T7nDlzxlgsFrNkyZIEna8wRsT5rl69avz8/Axgfv755weed+TIEdOmTRv7Cy7ZsmUzEyZMMFFRUY+8h9VqNTdu3HBk2ebq1av2F15d/c5hSVm2Lkt3d3eHv/C3d+9e+wv7M2bMMPv37zejR482TzzxhKlbt6557bXXzIwZM8w///xjQkJCzNGjR82JEyfMmTNnzPnz582lS5fMlStXzPXr182NGzfM4sWLDWAKFCiQ5BdUk2rhwoUGMDlz5kx2l0hKOnv2rD1YSUjo++WXXxrAVK5c2Zw6dcqMHDnS3iFi+6hdu7aZMWNGorpbbN02rVq1sndoNGzY0Gzfvt1s3rzZrF+/3vz111/mjz/+MEuWLDELFiwwc+fONbNmzTJTpkyxd3Bly5bN/PPPP8n5I3moZcuWmbx58yYo+LNaraZo0aIGMIsWLTLG3A3tvv/+e1O2bFn7Gr6+vubNN990SNfJ5cuXTfPmze1rv/DCCwn692D06NEGMO3atYv3+PHjx+3dQR4eHmbAgAHm8uXLSaoxJCTE3gmXlA4oZ9i1a5f9TRBz5sxJ0hqfffZZnK+DwoULm6CgIPP666+bKVOmmA0bNjz0a+yll14ygKlYsaL5/vvvDWBKlChx3/eypUuXGsCULl06SXWKiIiIiIgkVpoOY/r27Wv8/f1NcHCwuXDhgv0jMjIyznlHjhwxFovF/PHHH/GuU6pUKfPLL78YY4y5ceOGGTx4sNm0aZM5ceKEWbVqlalataopUaKEuX37doLqUhgj4nxvv/22AUylSpVMbGzsI89ftWqVqVChgv3FnTJlyjzwe4JN7969jYeHh1mxYoWjyjZz5syxv0gkGY/t3ftJ7b6IT1RUlH3dp59+2iHhya1bt0zmzJkNYHbs2OGAKhOuQYMG9lFhaU358uUNYObNm/fQ86xWq/370VdffWV/PCYmxvz++++mXbt29lFZtlFhr7zyyiNHZsXGxpqcOXMawKxbt85s3LjRZMmSJdEdUTlz5jS7du1yyJ/Jw1y9etV0797dlCtXzhw5cuSh5w4cONAeikybNs0UK1bMXq+fn5955513khxqPEhsbKwZM2aM/e+iVKlSjxzZ1qRJEwOYCRMmPPCcs2fPmhYtWtjr9/f3N2PHjk1U+Lhz505TpUoVp3bbJdVHH31kf16nT59O1LUHDx40mTJlsl//sM/TwMBAU79+fdOnTx8zfvx4s3LlSrNkyRL78Y0bN5rw8HD7env37o1zL9v/R/Tq1cuRT19EREREROSB0nQY86AfzmbOnBnnvGHDhpmCBQs+8AXbe6+JjIw0TZs2Nbly5TKenp6mcOHCpnfv3ubixYsJrkthjIhzRUZG2l9gXLp0aYKvi46ONt9++639xUq4uwfBwYMH7zt37ty59nOKFy9u7ty545Da27dvbwAzYsQIh6wnaYtt74tmzZo5bE3buL5s2bI5bB8KY4xp27atAcx7773nsDUfZffu3fbuoaSOmHKlN954wwCmZ8+eDz1v69at9nFa165di/ecCxcumNGjR8cJHTw9PR+6j8+2bdvs4Y2t+2/Dhg2mYsWKJn/+/KZIkSKmZMmSpnz58qZKlSrm8ccfN0888YRp2LChCQoKMq1btzY9evQwBw4cSPofgpP89ddf9/3/Xo4cOcxHH32UqH1hkmL9+vVxxpZNmzYt3tDz9u3bxsfHxwAJGjO3atUqU7ly5TgdIHPnzn3oGwwiIyPN0KFD7QFRtmzZzPr165P1/BwtOjraPP744wYwTZo0SXBAfO++V82aNTNWq9WEhoaatWvXmm+//dYMHDjQNG3a9L4Osvg+7v0abNWqlQHM+++/H+d+devWNYCZNWuWQ5+/iIiIiIjIg6TpMCa1Uhgj4ly2MUaFChVKUhfA9evXzRtvvGE8PDzsL4jeO+rs1KlT9nfk2jY3//TTT5Nd961bt4yvr68BnDr+R1Iv21ghDw+PZO1fdOfOHfPbb7+Zrl272j+Pf/jhBwdWasyMGTMMYKpVq+bQdW3+/fdf8/nnn8d5t3qvXr0MYJ599lmn3NPZli9fbgBTsGDBh35veuWVVwxgnn/++UeuGRsba1avXm3fmL5Hjx4PPNfWkdC2bdsk1Z+aRUVF2YP0vHnzms8++yxFR3NdvnzZBAUF2V/s79Kly31jy9auXWvv2Ejov02xsbFm1qxZ9rAHMDVq1DBr166979wtW7aYkiVL2s/r2LGjuXDhgkOen6MdOnTIHkx98803CbrGNp4sa9asj+youXHjhtm2bZuZM2eOGTZsmGnbtq0pVaqUcXd3N4UKFYrTJWXbz+nejtSIiAj7aMejR48m7UmKiIiIiIgkksIYJ1AYI+JcnTt3NoAZPHhwstYJCQkxTZs2tb+w9emnn5qYmBhTv359A5jHH3/cTJ061T4GJzEdcvH57bffDLhmHw5JPWwbXP+3i/NhrFarOXjwoPn222/Nc889ZwICAuK8C7xLly4O/5y6dOmSfX+js2fPPvC827dvmytXriRqbavValq3bm3vghk0aJA5fvy48fb2NoDZsGFDcst3iYiICPtIpPg67mzn2Pa7+uuvvxK89oYNGwzc3RclPDw83nOefPJJA5hJkyYlqf7UbufOnebnn3922V5CsbGxZvTo0faulNKlS8cJE0eOHGkA07lz50SvHRERYUaNGhVnrFzbtm1NSEiIiYqKMu+99579vvny5TOLFy925FNzivHjxxvAZM6c+ZFj6EJCQuxf/8kZ43jnzh0TExMT57GrV6/a/+yOHDliYmJizDPPPJOsN3WIiIiIiIgkhcIYJ1AYI+I8ERER9u6Sh43rSaiYmBgzYMAA+4tfts2rfX19zZEjR0xsbKz9seTOle/du7cBzKuvvprsuiXt+uCDDwxgWrZs+cBzYmNjzd69e83EiRNNx44dTe7cue8bw5MnTx7Tv39/s379+gTtm5QUtpFB3377bbzHjx49aooVK2YyZ8780MDmv2zvVLeFPbYRXICpUqVKmn5xtHHjxgYw48ePj/f47NmzDWCKFCmSqL83q9Vq74qYPn36fcfDwsLsXVLHjh1Lcv3yaOvWrbN3snh5eZk+ffqYkydPmieeeMIAZsqUKUle++LFi6ZPnz728MDDwyNON0znzp0fONoutYmNjTVPPfWUAUzdunXvC0ls7ty5Y/9ek5ixZolh+7ocM2aMvTPNy8srUYGoiIiIiIhIcimMcQKFMeIIMTExafoFSWdZsGCBAcxjjz3msD8fq9VqH49i+7j3xc6NGzfaXzhO6mbmsbGxJk+ePAYwK1ascEjdkjYdOHDAHj5cv37dGHP382Pnzp3myy+/NO3atTM5cuS4L3zJlCmTqV+/vnn33XfN2rVrH/jCpiPZ9rhp0aLFfcd2795t/5wGzLRp0xK05qlTp+ydIWPGjDErVqyI82JzYjqGUqOxY8faX1SOj6175cMPP0z02qNHjzaAqVev3n3HFi9ebODuHlfifKGhoaZly5b2z1sPDw97gOKIsVf79+83Tz/9tH39gIAAM3fuXAdUnrJOnDhhsmbNagAzduzY+45brVbz8ssv28eTnTp1yil1fPPNN/YuHdu/5/Pnz3fKvURERERERB5EYYwTKIyR5Lp165apXr26yZ8/v5k3b55CmXs8++yzBjBDhgxx+No///yzCQgIML169brvz/z55583gKlZs2aSNovetGmTfdzZnTt3HFWypFFly5a1jxdr1arVfWPHAOPj42MaNWpkPvjgA7N27VqXjGbat2+fPQi6d3+ODRs22PdVsnVjdO3a9ZHrxcbGmoYNGxrA1KlTxx4o3b5923z++edm2LBh9o3n06ojR47Yx6/du2+FMcYcPnzYwN29qM6cOZPotc+dO2ffxyokJCTOsb59+xrA9OvXL1n1S+IEBwebRo0a2b9uH7VfUGKtXr3aDB8+/JF7qKRm06ZNs3ei3DvWzRhjvvrqK3s48ttvvzmthvPnz8fpxEuvo/xERERERCR1UxjjBApjJLlsmzDbPtq1a2fOnz/v6rJcLiIiwv6u1m3btjnlHg/qNjhz5oz93b3lypUzx48fT9S6Q4cOTfJeApL+vPvuu/eFL1myZDFBQUFm9OjRZuPGjakitLNaraZo0aIGMIsWLTLXr183H3/8sX1j7rp165r58+cn+EXoCRMmJHgPibSsatWqBjCTJ0+O87jt+0Dz5s2TvHaLFi0MYIYOHRrncdvf06+//prktSXpNm7caF5++WXz+++/u7qUVMdqtdq7iHLlymXef/99c/nyZbNq1Sp7N9Enn3zi9Dpso8ree+89p99LREREREQkPgnNDSzGGIMkSHh4OP7+/oSFheHn5+fqciSNOX/+PCVLliQiIoI2bdrw+++/ExMTQ0BAAJ9//jk9evTAYrG4ukynM8bw4Ycfki1bNvr164ebmxsLFiygY8eOFClShGPHjqX4n8P27dtp3bo158+fJ1euXCxevJg6deok6NoyZcpw6NAhfvzxRzp37uzkSiW1u3DhAp06dcLf35/69etTv359qlSpgoeHh6tLu8+gQYMYP348pUqV4vz589y4cQOAFi1aMH/+fIwxZMuWjejoaI4dO0bRokXjXefw4cNUrlyZW7du8fXXX/Pqq6+m5NNIUZ988glDhw6lUaNGrFq1CoCYmBgKFSrEhQsXWLBgAe3bt0/S2gsXLqRDhw7ky5ePU6dO4eHhwbFjxyhevDgeHh5cu3aNrFmzOvLpiCTbhQsXqF+/PkeOHAHA29sbT09Pbty4QdeuXZk9e7bT/02/evUqhw8fplatWhni/6NERERERCT1SWhu4JaCNYlkaMOGDSMiIoLatWuzaNEitm/fTrVq1fj333956aWXaNq0KSdOnHB1mU63Y8cORo4cyYABA2jfvj03btxg3rx5ADz77LMueSGlWrVqbN26lapVq3L58mUaNmzI3LlzH3nd4cOHOXToEJ6enjRv3jwFKpXULm/evKxbt46lS5cyePBgatSokSqDGIDWrVsDEBISwo0bNyhXrhxz5szh119/JXPmzPj6+lKzZk0AgoOD410jJiaGbt26cevWLZo0aULfvn1TqnyX6NixIwBr1qwhNDQUgD/++IMLFy6QK1cuWrVqleS1W7VqRY4cOTh//jxLly5l1apVDB06FIC6desqiJFUKW/evOzfv5+5c+dSrVo1bt++zY0bN3j88ceZMmVKivybniNHDmrXrq0gRkREREREUj2FMSIpYMuWLcyZMweA8ePHY7FYqFixIps3b2bs2LF4e3uzatUqypcvz/jx44mNjXVxxc5z74u6ixcvplatWvz222/A/17odIX8+fOzbt062rVrx507d3jhhRd49913sVqtD7xmyZIlADRo0AB/f/+UKlXEIZ588knat29Po0aNWLp0KXv27KFr1664u7vbz6lfvz4Aa9eujXeNsWPHsmXLFvz9/ZkxY0a6fzG0aNGi1KhRA6vVysKFCwGYNm0aAF27dsXLyyvJa3t5edGlSxcAnnnmGZo0acKCBQsAaNOmTTIrF3EeT09PnnvuObZt20ZwcDAffPABS5cuxdvb29WliYiIiIiIpCoaU5YIGlMmSWG1WqlTpw5btmyhR48ezJw5875zjhw5Qq9evVi3bh0AtWvXZtq0aZQtWzZJ94yNjeX69evkzJkzWbU7Q+vWrVm6dCldunRh9erVnD9/Hrj7IufRo0dd/mKu1Wpl+PDhfPLJJ8Ddbp1Zs2bh4+MT57zz589TqVIlrly5wjfffJPuOwIkY/rzzz9p2rQphQoV4uTJk3G+Pnfv3k2NGjWIjo5mzpw5dO3a1YWVppxx48YxZMgQGjRowNy5cylYsCCxsbHs378/yd+zbfbt20eVKlWIiYkhb968BAUF0aJFC9q1axcnJBMRERERERGR1ENjykRSiZ9++oktW7aQJUsWPv7443jPKVGiBGvWrOHbb78la9asbNq0iSpVqjBq1Ciio6MTfc+ePXsSGBjI66+/TmRkZHKfgsPExsbaA6cBAwawfft2+94s3bp1c3kQA+Dm5saYMWOYMWMGnp6ezJs3jwYNGnDx4kX7ObGxsXTt2pUrV65QqVIlXnzxRRdWLOI8derUwcPDg9OnT3Py5En743fu3KFr165ER0fTtm1be0dHRmDr4Fu7di1jx44lNjaWWrVqJTuIAShfvjw7duxg165dnDt3jhkzZtChQwcFMSIiIiIiIiLpgMIYESdbtGgRAK+//jp58+Z94Hlubm688sorHDhwgJYtWxIVFcWIESOS9CLnxo0bMcbw5ZdfUqlSJdavX5/k+h1pz549hIWFkTVrVqpUqUKePHkIDg5m48aNDBs2zNXlxfHiiy/y559/kj17drZu3UrNmjXZs2cPAGPGjGH16tX4+vry888/axSLpFsP2jfmvffeY+/eveTKlYvJkyeniiA1pRQuXJhatWrZv8cC9OrVy2HrV6hQgUqVKmWoP1MRERERERGRjEBhjIiTHTx4ELg7eiwhChQowNKlS+3jzBYtWpSo7hhjDGfPngXubmp79OhR6tevz8CBA4mIiEhk9Y5l23eiXr169k3NPT09qVOnTrL2WnCW+vXrs3nzZkqWLMmZM2eoW7cuo0aNYuTIkQB88803lCpVysVVijjXf/eN2bRpE2PHjgVg8uTJBAYGuqw2V3n22Wftv/b19Y3zexERERERERGR+CiMEXGimJgYDh8+DECZMmUSfJ3FYqF79+5kyZKF6Oho+xoJce3aNW7fvg3AgQMH6NWrF8YYJkyYQMWKFR+4EXdKsN3b9uJuWlCiRAk2b95Mw4YNuXnzJiNGjCA2NpZu3brRrVs3V5cn4nQNGjQA7nbGRERE0L17d6xWK127dqVdu3auLc5FOnToYP91p06dyJo1qwurEREREREREZG0QGGMiBOdOHGC6OhofHx8KFSoUKKutVgslC9fHri7qXNCnTlzBoDAwEACAwOZOnUqK1asoGDBghw/fpwGDRrQv39/bt68mah6/ssYw6FDh4iNjU3Q+Var1b5fTFoKYwCyZcvG8uXL6d27NwAlS5bk66+/dnFVIinDtm/MqVOn6NKlC0eOHKFAgQJMmDDB1aW5TMGCBWnZsiWenp7069fP1eWIiIiIiIiISBqgMEbEiWwjykqVKoWbW+K/3JISxthGlBUoUMD+WNOmTdm3bx8vv/wyABMnTqRixYqsWbMmzrUXL15k9uzZnDt37pH3GTJkCGXKlKFq1ar3rROf/fv3c+3aNXx9falWrVqCn09q4enpyeTJk9m6dStbt24lS5Ysri5JJEVkyZKFGjVqALB48WIAZsyYQUBAgOuKSgXmzZvHyZMnqVq1qqtLEREREREREZE0QGGMiBPZwpjEjCi7V4UKFYDkhzEAfn5+TJ48mZUrV1KoUCFOnDhBw4YN6devH3/++SedOnWiYMGC9OjRg+eff/6h91i/fj2ff/45AHv27KFhw4a0a9eOo0ePPvAa2+bfderUwdPTM8HPJzWxWCzUqFEDf39/V5cikqJso8oAXn31VZo0aeK6YlKJzJkzky9fPleXISIiIiIiIiJphMIYESc6dOgQkPQwxtYZs3fv3gRfYwtjChYsGO/xJk2asHfvXl555RXg7ib0TZs2Zd68ecTExACwbt26BwYrkZGRvPTSSxhjeP755+nfvz/u7u4sXryYsmXLMmTIEMLCwu67zrZfzL0v6opI2tCiRQsAihcvztixY11cjYiIiIiIiIhI2qMwRsSJbJ0xpUuXTtL1tjDm+PHjREREJOga254x/+2MuZefnx/ffvstf/75J4899hi+vr688sor7Ny5k2bNmgEwZ86ceK995513OHr0KAUKFOCbb75hwoQJ7Nmzh6CgIKKjoxk3bhwlSpRgypQp9v1kjDFpdr8YEYF69eqxdu1a/v77b3x9fV1djoiIiIiIiIhImqMwRsRJjDHJHlMWGBhIrly54qz1KA8aUxafxo0bc+TIEa5du8a3335L5cqV6d69OwCzZ8/GarXGOf/vv//myy+/BGDKlCn2cV1ly5bljz/+YNmyZZQuXZrLly/zyiuvULVqVVavXs3Bgwe5fPkyPj4+9r0nRCRtefLJJ8mVK5eryxARERERERERSZMUxkiynTx5kt69e3Pnzh1Xl5KqXLx4kfDwcNzc3ChRokSS10nsvjGJCWMAPDw88PLysv++bdu2+Pn5cfr0aftoMbg7nuzFF1/EGMOLL75I8+bN71urefPm7NmzhwkTJpAtWzb27NlDo0aNaNOmDQC1a9eOcy8RERERERERERGRjEBhjCRLbGwsTZs2Zdq0abz22msYY1xdUqph62QpWrQomTJlSvI6idk3xhhjH1P2oD1jHsXHx4fOnTsDMGvWLPvjb775JocPHyZfvnx8/vnnD7ze09OT/v37c+TIEft+Mrb9ZzSiTERERERERERERDIihTGSLO7u7nz11VdYLBamTZvGlClTXF1SqpHcEWU2tjAmIZ0x169f59atWwDkz58/yfe0jSpbuHAhN2/eZPHixXz77bdYLBbmzJlDQEDAI9fIkSMHEyZMYO/evbRs2ZIiRYrwwgsvJLkmERERERERERERkbRKYYwkW7Nmzfj4448B6N+/P3///beLK0odDh06BKRsGGMbUZYzZ068vb2TfM/atWtTokQJIiIiGD9+PD179gRgyJAhNGrUKFFrlSlTht9++43jx49TrFixJNckIiIiIiIiIiIiklYpjBGHePvtt+nQoQPR0dG0b9+e8+fPu7okl7N1xpQuXTpZ65QrVw6A8+fPc+3atYeem9j9Yh7EYrHYu2Peeecdrl27RrVq1fjwww+Tta6IiIiIiIiIiIhIRqQwRhzCYrEwc+ZMypUrx8WLF+nQoQN37txxdVku5agxZX5+fhQuXBh4dHdMcveLuVfXrl2xWCwA+Pr6MnfuXLy8vJK9roiIiIiIiIiIiEhGozBGHCZLliwsXryYgIAANm3axMCBA11dksuEh4fbu4OS2xkDCR9V5qjOGIBChQrx9NNPAzBx4kRKliyZ7DVFREREREREREREMiKFMeJQxYsXZ+7cuVgsFiZPnszUqVNdXZJL2PaLyZMnT4I2u3+UChUqACkbxgDMnTuX/fv306NHD4esJyIiIiIiIiIiIpIRKYwRh2vevDmjRo0CoF+/fmzatMnFFaU8R40os7F1xuzdu/eh5zlyTBnc7XYqW7asQ9YSERERERERERERyagUxohTDBs2jGeeeYbo6Gjat2/PhQsXXF1SirJ1xjhiRBnEHVNmjHngeY7ujBERERERERERERGR5FMYI05hsViYNWsWZcuW5cKFC3Ts2JGoqChXl5ViHN0ZU7p0adzd3fn333/te9H8lzFGYYyIiIiIiIiIiIhIKqQwRpwma9asLF68GH9/fzZu3MigQYNcXVKKcXQYkylTJkqWLAk8eN+YsLAwIiIiAMifP79D7isiIiIiIiIiIiIiyacwRpyqRIkS/PDDD1gsFiZNmsT06dNdXZLTRUVFcezYMcBxY8rgf6PKdu3aFe9x234xOXLkIHPmzA67r4iIiIiIiIiIiIgkj8IYcbqWLVvy/vvvA/Dqq6+yZcsWF1fkXEeOHCE2NpasWbM6tEOlbt26ACxfvjze4xpRJiIiIiIiIiIiIpI6KYyRFPF///d/tG3blqioKNq3b8+NGzdcXZLTbNiwAYAqVapgsVgctm7r1q0BWL9+PdeuXbvvuMIYERERERERERERkdRJYYykCDc3N2bPnk3BggU5d+4ca9ascXVJTmN7bg0bNnToukWKFKFChQrExsaybNmy+47bxpQpjBERERERERERERFJXRTGSIrx8/OjcePGAGzbts3F1TiHMYbVq1cDjg9j4H/dMb/++ut9x2ydMQULFnT4fUVEREREREREREQk6RTGSIqqWbMmAFv/X3t3HlVlncdx/HORRWRTBEECizTFJTUR1DSXUpwylTZtO5prFk6i2aKn4kzL2KS2zdHSQmzRM0oTaaaZhYNiLo1Kk42iIIYpYG4gqEjwzB8e7njdEuG593J9v865x8vv+T2/3/d5jl+P3O/9/Z4tWxwciTl+/vln/fbbb/L29la3bt3qfPzqYsyqVatUXl5uc4xtygAAAAAAAADAOVGMgV3FxMRIOrsyxjAMB0dT96pXxdx2223y9PSs8/G7du2q5s2bq7S0VP/6179sjlGMAQAAAAAAAADnRDEGdnXzzTfLy8tLx44dU25urqPDvRIg9QAAHPNJREFUqXNmPS+mmpubmwYPHixJWrZsmbXdMAzrM2PYpgwAAAAAAAAAnAvFGNiVp6enbrnlFkmut1VZZWWldbVKv379TJtn6NChks4+N6Z6dVFJSYlKS0slSdddd51pcwMAAAAAAAAAao5iDOzu3K3KXElWVpaOHz8uf39/denSxbR5br/9dvn4+OjAgQPatm2bpP9vUdakSRP5+PiYNjcAAAAAAAAAoOYoxsDuYmNjJbneypjq58X06dNH7u7ups3TsGFDDRw4UNLZ1TGVlZXKzMyUxPNiAAAAAAAAAMAZUYyB3VUXY7Zt26aKigoHR1N3zH5ezLmGDBkiSZozZ45CQkI0YcIESVJkZKTpcwMAAAAAAAAAaoZiDOyuVatWCggI0OnTp/Xzzz87Opw6UVFRoXXr1kmyTzFm0KBBatCggY4cOaIjR46ocePGevDBBzVjxgzT5wYAAAAAAAAA1Ix5eykBl+Dm5qaYmBh9++232rJlizp37uzokGrthx9+UFlZmZo2baoOHTqYPl9QUJAWLVqkH3/8UQMHDtStt94qDw8P0+cFAAAAAAAAANQcxRg4xLnFmPHjxzs6HElSaWmpRowYoYKCAvn5+cnX11e+vr7W9xdrq37/z3/+U5LUr18/ubnZZ8HZ8OHDNXz4cLvMBQAAAAAAAAC4ehRj4BDVz4354YcfHBzJ/y1YsEBpaWm1GsMeW5QBAAAAAAAAAOoXijFwiOpizI4dO1RWViYfHx+HxmMYhubOnStJSkxMVHR0tEpLS3XixAmbPy/3Pjg4WPfee69DrwMAAAAAAAAA4HwoxsAhwsLCFBYWpoMHD2rbtm267bbbHBpPenq6srOz5evrq5dffll+fn4OjQcAAAAAAAAA4Drs83AL4CKcaauyOXPmSJJGjBhBIQYAAAAAAAAAUKcoxsBhqosx33zzjSoqKmyOff3114qKilLv3r31+++/mxrHr7/+qmXLlkmSnnzySVPnAgAAAAAAAABceyjGwGH69esnSVq9erXat2+vtLQ0HTt2TKNGjdKdd96p7OxsrV+/XmlpaXU2Z35+vsaMGWMz5rx581RVVaW+ffuqffv2dTYXAAAAAAAAAACSZDEMw3B0EPVFSUmJAgICVFxcLH9/f0eH4xJSUlL0/PPP69ChQ5Kkhg0b6vTp07JYLOrYsaN+/PFH9ezZU5mZmbWeq7rgsn79eknSmDFjNHPmTLVt21ZFRUVaunSpHnjggVrPAwAAAAAAAAC4Nlxp3YCVMXCoUaNGKScnRy+88IK8vb11+vRptWnTRpmZmVq5cqU8PDy0YcMGbd26tdZzzZs3T+vXr5eXl5csFouSk5PVsmVLFRUVKSwsTPHx8bW/IAAAAAAAAAAAzkMxBg7n5+enV155RXv27FFqaqqysrJ06623KiwsTMOGDZMkvfPOO7WaIz8/X88++6wkaebMmUpPT9d1112nY8eOSZLGjx8vDw+P2l0IAAAAAAAAAAAXwTZlNcA2Zfb3ww8/KDY2Vh4eHsrPz1doaGiNxzAMQ4MGDdKqVavUs2dPrVu3Tm5ubjp69KgmT56svXv36vPPP1dwcLAJVwAAAAAAAAAAcFX1epuyGTNmKCYmRn5+fmrWrJni4+OVnZ1t06dv376yWCw2rwkTJlx2XMMw9NJLL6l58+by9vZW//79tWfPHjMvBbUUExOjHj16qKKiQu+///5VjbFo0SKtWrVKnp6e+vDDD+XmdvavfWBgoD766COtX7+eQgwAAAAAAAAAwDROWYzJyMhQQkKCNm3apDVr1qiiokJxcXEqKyuz6Tdu3DgVFBRYX2+88cZlx33jjTf07rvv6v3339fmzZvl4+OjgQMH6vTp02ZeDmpp0qRJkqT33ntP5eXlNTr36NGjSkxMlCQlJSUpKiqqrsMDAAAAAAAAAOCy6sU2Zb/99puaNWumjIwM9e7dW9LZlTGdO3fW22+/fUVjGIahsLAwPf3005o6daokqbi4WCEhIVq4cKEefPDBPxyDbcoco6KiQpGRkTpw4ICef/55vfDCC/Lx8bmicxMTE/XOO++oQ4cO2rZtG8+FAQAAAAAAAADUmXq9Tdn5iouLJZ3dVupcixYtUlBQkDp06KBp06bp5MmTlxwjLy9PhYWF6t+/v7UtICBA3bp108aNGy96Tnl5uUpKSmxesD8PDw9NnjxZkvT666+rRYsWevHFF1VUVHTZ87KzszVnzhxJ0ptvvkkhBgAAAAAAAADgEE5fjKmqqlJiYqJ69uypDh06WNsffvhhffrpp1q7dq2mTZumTz75RI8++uglxyksLJQkhYSE2LSHhIRYj51vxowZCggIsL4iIiLq4IpwNSZPnqy5c+eqZcuWOnr0qF599VVdf/31Gj9+vHbt2nXRc6ZOnarff/9dd999twYMGGDniAEAAAAAAAAAOMvptyl74okntGrVKmVmZio8PPyS/dLT03XHHXcoJydHLVu2vOD4999/r549e+rgwYNq3ry5tX3YsGGyWCxasmTJBeeUl5fbPKOkpKREERERbFPmQJWVlfriiy80c+ZMbd682do+ePBgPfPMM+rVq5csFou+/fZbDRgwQO7u7tqxY4fatGnjwKgBAAAAAAAAAK7IJbYpmzhxolasWKG1a9dethAjSd26dZMk5eTkXPR4aGioJF2wtVVRUZH12Pm8vLzk7+9v84JjNWjQQPfdd582btyo9evXa+jQobJYLPryyy/Vu3dvde/eXUuXLtWUKVMkSU8++SSFGAAAAAAAAACAQzllMcYwDE2cOFFpaWlKT09XZGTkH56TlZUlSTarXs4VGRmp0NBQfffdd9a2kpISbd68WT169KiTuGE/FotFvXr10hdffKGdO3dq/Pjx8vLy0pYtWzR8+HD99NNPatKkiZKSkhwdKgAAAAAAAADgGueUxZiEhAR9+umnWrx4sfz8/FRYWKjCwkKdOnVKkpSbm6tXXnlFW7du1b59+7R8+XKNGDFCvXv3VseOHa3jREVFKS0tTdLZD+8TExP16quvavny5frpp580YsQIhYWFKT4+3hGXiTrSpk0bzZs3T/n5+XrppZfUtGlTSdJrr72mwMBAB0cHAAAAAAAAALjWOeUzYywWy0XbU1JS9Nhjj2n//v169NFHtWPHDpWVlSkiIkL33HOPXnjhBZutxCwWi/Uc6eyKm6SkJM2fP1/Hjx9Xr169NHfuXLVu3fqK4rrSvd/gWCdPnlReXp7at2/v6FAAAAAAAAAAAC7sSusGTlmMcVYUYwAAAAAAAAAAQLUrrRs45TZlAAAAAAAAAAAAroJiDAAAAAAAAAAAgIkoxgAAAAAAAAAAAJiIYgwAAAAAAAAAAICJKMYAAAAAAAAAAACYiGIMAAAAAAAAAACAiSjGAAAAAAAAAAAAmIhiDAAAAAAAAAAAgIkoxgAAAAAAAAAAAJiIYgwAAAAAAAAAAICJKMYAAAAAAAAAAACYiGIMAAAAAAAAAACAiSjGAAAAAAAAAAAAmIhiDAAAAAAAAAAAgIkoxgAAAAAAAAAAAJiIYgwAAAAAAAAAAICJKMYAAAAAAAAAAACYiGIMAAAAAAAAAACAiSjGAAAAAAAAAAAAmIhiDAAAAAAAAAAAgIkoxgAAAAAAAAAAAJjI3dEB1CeGYUiSSkpKHBwJAAAAAAAAAABwtOp6QXX94FIoxtTAiRMnJEkREREOjgQAAAAAAAAAADiLEydOKCAg4JLHLcYflWtgVVVVpYMHD8rPz08Wi8XR4ThUSUmJIiIitH//fvn7+zs6HKBeI58A+yHfAPsh3wDHIPcA+yHfAPsh3+DMDMPQiRMnFBYWJje3Sz8ZhpUxNeDm5qbw8HBHh+FU/P39+QcQqCPkE2A/5BtgP+Qb4BjkHmA/5BtgP+QbnNXlVsRUu3SZBgAAAAAAAAAAALVGMQYAAAAAAAAAAMBEFGNwVby8vJSUlCQvLy9HhwLUe+QTYD/kG2A/5BvgGOQeYD/kG2A/5BtcgcUwDMPRQQAAAAAAAAAAALgqVsYAAAAAAAAAAACYiGIMAAAAAAAAAACAiSjGAAAAAAAAAAAAmIhiDAAAAAAAAAAAgIkoxriQGTNmKCYmRn5+fmrWrJni4+OVnZ1t0+f06dNKSEhQ06ZN5evrq/vuu09FRUU2fZ566ilFR0fLy8tLnTt3vmCeffv2yWKxXPDatGnTH8Y4Z84c3XDDDWrYsKG6deumLVu22ByfP3+++vbtK39/f1ksFh0/frzG9wGoK66QU48//rhatmwpb29vBQcHa+jQodq1a1fNbwZgIlfItb59+14w7oQJE2p+MwCT1fd8u9S4FotFqampV3dTADup7/knSbm5ubrnnnsUHBwsf39/DRs27IL4AEdz9lxbt26dBg8erLCwMFksFn3xxRcX9Pn8888VFxenpk2bymKxKCsrq6a3AbALe+WbJBmGoVmzZql169by8vLSddddp9dee+0PY0xNTVVUVJQaNmyom2++WStXrrQ5Tr7BnijGuJCMjAwlJCRo06ZNWrNmjSoqKhQXF6eysjJrn8mTJ+vLL79UamqqMjIydPDgQd17770XjDV69GgNHz78svN9++23KigosL6io6Mv23/JkiWaMmWKkpKStG3bNnXq1EkDBw7UoUOHrH1OnjypP/3pT5o+fXoNrx6oe66QU9HR0UpJSdHOnTu1evVqGYahuLg4VVZW1vBuAOZxhVyTpHHjxtmM+8Ybb9TgLgD2Ud/zLSIiwma8goIC/eUvf5Gvr6/uvPPOq7gjgP3U9/wrKytTXFycLBaL0tPTtWHDBp05c0aDBw9WVVXVVdwRwBzOnmtlZWXq1KmT5syZc9k+vXr10t/+9rc/uFrAseyZb5MmTdKHH36oWbNmadeuXVq+fLliY2MvG9/333+vhx56SGPGjNH27dsVHx+v+Ph47dixw9qHfINdGXBZhw4dMiQZGRkZhmEYxvHjxw0PDw8jNTXV2mfnzp2GJGPjxo0XnJ+UlGR06tTpgva8vDxDkrF9+/YaxRMbG2skJCRYf66srDTCwsKMGTNmXNB37dq1hiTj2LFjNZoDMFN9zqlqP/74oyHJyMnJqdFcgD3Vx1zr06ePMWnSpBqNCziD+phv5+vcubMxevToGs0DOIP6ln+rV6823NzcjOLiYmuf48ePGxaLxVizZk2N5gLsydly7VySjLS0tEser4s5AHsyK9/++9//Gu7u7sauXbtqFM+wYcOMQYMG2bR169bNePzxxy/oS77BHlgZ48KKi4slSYGBgZKkrVu3qqKiQv3797f2iYqKUosWLbRx48Yajz9kyBA1a9ZMvXr10vLlyy/b98yZM9q6davN3G5uburfv/9VzQ04Qn3PqbKyMqWkpCgyMlIRERE1jg+wl/qaa4sWLVJQUJA6dOigadOm6eTJkzWODbC3+ppv1bZu3aqsrCyNGTOmxrEBjlbf8q+8vFwWi0VeXl7WPg0bNpSbm5syMzNrHB9gL86Ua4CrMyvfvvzyS914441asWKFIiMjdcMNN2js2LE6evToZc/buHGjzdySNHDgQD6LhMNQjHFRVVVVSkxMVM+ePdWhQwdJUmFhoTw9PdW4cWObviEhISosLLzisX19fTV79mylpqbqq6++Uq9evRQfH3/Z/3QcPnxYlZWVCgkJqdXcgKPU55yaO3eufH195evrq1WrVmnNmjXy9PS84vgAe6qvufbwww/r008/1dq1azVt2jR98sknevTRR684NsAR6mu+nSs5OVlt27bVrbfeesWxAc6gPuZf9+7d5ePjo+eee04nT55UWVmZpk6dqsrKShUUFFxxfIA9OVuuAa7MzHzbu3evfvnlF6Wmpurjjz/WwoULtXXrVt1///2XPa+wsJDPIuFU3B0dAMyRkJCgHTt2mPINpaCgIE2ZMsX6c0xMjA4ePKiZM2dqyJAhWr9+vc2e3fPmzVO/fv3qPA7AnupzTj3yyCMaMGCACgoKNGvWLA0bNkwbNmxQw4YN6/Q6gLpQX3Nt/Pjx1vc333yzmjdvrjvuuEO5ublq2bJl3V0EUIfqa75VO3XqlBYvXqwXX3yxzuIG7KU+5l9wcLBSU1P1xBNP6N1335Wbm5seeughdenSRW5ufM8TzsnZcu2RRx6p8zgAZ2FmvlVVVam8vFwff/yxWrduLensl3Kio6OVnZ0tb29vtWvXztp/+vTpPI8aTolijAuaOHGiVqxYoXXr1ik8PNzaHhoaqjNnzuj48eM2FemioiKFhobWas5u3bppzZo1kqSuXbsqKyvLeiwkJEReXl5q0KCBioqKbM6ri7kBs9X3nAoICFBAQIBuuukmde/eXU2aNFFaWpoeeuihWsUI1LX6nmvnjytJOTk5FGPglFwh3z777DOdPHlSI0aMqFVcgL3V5/yLi4tTbm6uDh8+LHd3dzVu3FihoaG68cYbaxUfYAZnzDXAVZmdb82bN5e7u7u1ECNJbdu2lSTl5+erX79+NvlWvU1aaGgon0XCqfD1FRdiGIYmTpyotLQ0paenKzIy0uZ4dHS0PDw89N1331nbsrOzlZ+frx49etRq7qysLDVv3lyS5O3trVatWllffn5+8vT0VHR0tM3cVVVV+u6772o9N2AWV8wpwzBkGIbKy8trFR9Ql1wx16p/EageG3AWrpRvycnJGjJkiIKDg2sVF2AvrpR/QUFBaty4sdLT03Xo0CENGTKkVvEBdcmZcw1wNfbKt549e+r3339Xbm6utW337t2SpOuvv17u7u42+VZdjOnRo4fN3JK0Zs0aPouEw7AyxoUkJCRo8eLFWrZsmfz8/Kz7HwYEBMjb21sBAQEaM2aMpkyZosDAQPn7++vPf/6zevTooe7du1vHycnJUWlpqQoLC3Xq1CnrB0rt2rWTp6enPvroI3l6euqWW26RJH3++edasGCBPvzww8vGN2XKFI0cOVJdu3ZVbGys3n77bZWVlWnUqFHWPoWFhSosLFROTo4k6aeffpKfn59atGhh/YcUsJf6nlN79+7VkiVLFBcXp+DgYP366696/fXX5e3trbvuusuEOwZcnfqea7m5uVq8eLHuuusuNW3aVP/5z380efJk9e7dWx07djThjgFXr77n27nzr1u3TitXrqzDuwOYyxXyLyUlRW3btlVwcLA2btyoSZMmafLkyWrTpk0d3y3g6jl7rpWWllo/85CkvLw8ZWVlKTAwUC1atJAkHT16VPn5+Tp48KCksx9eS2e/5c83+uFM7JVv/fv3V5cuXTR69Gi9/fbbqqqqUkJCggYMGGCzWuZ8kyZNUp8+fTR79mwNGjRI//jHP/Tvf/9b8+fPt/Yh32BXBlyGpIu+UlJSrH1OnTplPPnkk0aTJk2MRo0aGffcc49RUFBgM06fPn0uOk5eXp5hGIaxcOFCo23btkajRo0Mf39/IzY21khNTb2iGP/+978bLVq0MDw9PY3Y2Fhj06ZNNseTkpL+8BoAe6nvOXXgwAHjzjvvNJo1a2Z4eHgY4eHhxsMPP2zs2rWr1vcGqEv1Pdfy8/ON3r17G4GBgYaXl5fRqlUr45lnnjGKi4trfW+Aulbf863atGnTjIiICKOysvKq7wVgb66Qf88995wREhJieHh4GDfddJMxe/Zso6qqqlb3Bahrzp5ra9euvei4I0eOtPZJSUm5aJ+kpKQ6uENA3bFXvhnG2c847r33XsPX19cICQkxHnvsMePIkSN/GOPSpUuN1q1bG56enkb79u2Nr776yuY4+QZ7shiGYfxhxQYAAAAAAAAAAABXhWfGAAAAAAAAAAAAmIhiDAAAAAAAAAAAgIkoxgAAAAAAAAAAAJiIYgwAAAAAAAAAAICJKMYAAAAAAAAAAACYiGIMAAAAAAAAAACAiSjGAAAAAAAAAAAAmIhiDAAAAAAAAAAAgIncHR0AAAAAANiLxWKx+dnd3V0BAQFq3ry5oqOjNXjwYA0dOlTu7vyqBAAAAKDuWAzDMBwdBAAAAADYQ3UxZuTIkZKkqqoqFRcXa/fu3crOzpZhGGrVqpUWLVqk2NhYR4YKAAAAwIVQjAEAAABwzaguxlzs16Dc3FxNnz5dS5cuVaNGjbRhwwZ17tzZzhECAAAAcEUUYwAAAABcMy5XjKk2duxYJScn65ZbbtG2bdvsFRoAAAAAF+bm6AAAAAAAwJnMnj1bPj4+2r59uzIzM22OffXVVxo9erTatm0rf39/+fj4qFOnTvrrX/+q8vJym76zZs2SxWLR9OnTLzlXXFycLBaL1q5da8q1AAAAAHAOrIwBAAAAcM24kpUxkvTAAw/os88+08svv6wXX3zR2h4aGqpTp06pQ4cOCg8PV3FxsbZs2aJjx47p9ttv1zfffKMGDRpIkg4fPqzw8HA1adJE+/fvl7u7u80ceXl5atmypVq1aqXdu3fX8ZUCAAAAcCasjAEAAACA81Q/K2bnzp027fPmzVNhYaE2bNigJUuW6Ouvv9Yvv/yiu+++W+np6Vq0aJG1b1BQkO677z4VFhZqxYoVF8yRnJwswzA0duxYU68FAAAAgONRjAEAAACA8wQFBUmSjh07ZtM+dOhQeXt727T5+fnprbfekiQtW7bM5tiECRMkSR988IFNe2VlpRYuXCgPDw899thjdRk6AAAAACfk/sddAAAAAODaUr2NWfW2Zufas2ePVq5cqZycHJWVlamqqsraf8+ePTZ9b7vtNrVv315ff/219u/fr4iICEnSypUrdeDAAd1///1q1qyZyVcDAAAAwNEoxgAAAADAeQ4fPixJCgwMtLYZhqGpU6fqrbfeuuQzZ06cOHFB2+OPP66nnnpKCxYsUFJSkqT/r5QZN25cXYcOAAAAwAmxTRkAAAAAnGf79u2SpHbt2lnblixZojfffFPh4eH67LPPdODAAZ05c0aGYai8vFySLlqkGTFihBo1aqQFCxaoqqpKBw8e1MqVK3XDDTdowIAB9rkgAAAAAA7FyhgAAAAAOEdxcbFWr14tSerXr5+1PS0tTZL03nvvadCgQTbn7N2795LjBQQE6MEHH9SCBQu0evVqbdu2TZWVlRo7duxFt0EDAAAA4HpYGQMAAAAA53j66adVVlammJgY9ejRw9p+7NgxSVJ4ePgF5yxduvSyY06YMEGSNG/ePCUnJ6tBgwYaNWpUHUYNAAAAwJlRjAEAAAAAnV3dMnz4cCUnJ8vHx0fJyck2x1u3bi1Jmj9/vs12ZOvXr9fMmTMvO3ZMTIy6dOmiZcuWKS8vT4MGDVJYWFjdXwQAAAAAp2QxLvXkSQAAAABwMdXbgo0cOVKSVFVVpZKSEu3evVu7du2SYRi66aabtHjxYnXt2tXm3N27d6tLly4qKytTu3bt1LFjRx04cECZmZl6+umnNWvWLF1//fXat2/fRef+4IMPNH78eEnSihUrLtjqDAAAAIDrohgDAAAA4Jpx/jNa3N3d5e/vr7CwMEVHR2vo0KEaMmSIGjRocNHzd+3apWeffVabN29WaWmp2rRpoyeeeELjxo2TxWK5bDEmNzdXrVq1Unh4uPbt23fJOQAAAAC4HndHBwAAAAAA9lLb76JFRUVp+fLlVzV29XNlxowZQyEGAAAAuMawMgYAAAAATFZSUqKoqCgdOXJEeXl5PC8GAAAAuMawMgYAAAAATJKSkqKMjAytW7dOBQUFSkxMpBADAAAAXIPcHB0AAAAAALiqjIwMffTRRyotLVVCQoJef/11R4cEAAAAwAHYpgwAAAAAAAAAAMBErIwBAAAAAAAAAAAwEcUYAAAAAAAAAAAAE1GMAQAAAAAAAAAAMBHFGAAAAAAAAAAAABNRjAEAAAAAAAAAADARxRgAAAAAAAAAAAATUYwBAAAAAAAAAAAwEcUYAAAAAAAAAAAAE/0POR798H+YYH0AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 2000x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 1, figsize = (20, 7))\n",
    "\n",
    "plt.plot(df_plot_2['y'], 'k-', df_plot_2['AutoARIMA'], 'b-')\n",
    "\n",
    "# Specify graph features:\n",
    "ax.fill_between(df_plot_2.index, \n",
    "                df_plot_2['AutoARIMA-lo-80'], \n",
    "                df_plot_2['AutoARIMA-hi-80'],\n",
    "                alpha=.35,\n",
    "                color='blue',\n",
    "                label='auto_arima_level_80')\n",
    "ax.set_title('Google daily closing stock price', fontsize=20)\n",
    "ax.set_ylabel('$US', fontsize=15)\n",
    "ax.set_xlabel('Day', fontsize=15)\n",
    "ax.legend(prop={'size': 15})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The lower limit of this prediction interval gives the 10th percentile (or 0.1 quantile) of the forecast distribution, so we would expect the actual value to lie below the lower limit about 10% of the time, and to lie above the lower limit about 90% of the time. When we compare the actual value to this percentile, we need to allow for the fact that it is more likely to be above than below.\n",
    "\n",
    "More generally, suppose we are interested in the quantile forecast with probability $p$ at future time $t$, and let this be denoted by ${f}_{p,t}$. That is, we expect the observation ${y}_{t}$ to be less than  with probability ${f}_{p,t}$. For example, the 10th percentile would be ${f}_{0.10,t}$. If ${y}_{t}$ denotes the observation at time $t$, then the Quantile Score is\n",
    "\n",
    "$$\n",
    "{Q}_{p,t}=\n",
    "\\begin{cases}\n",
    "2 - (1-p) ({f}_{p,t} - {y}_{t}) & \\quad \\text{if ${y}_{t}<{f}_{p,t}$}\\\\ \n",
    "2p - ({y}_{t} - {f}_{p,t}) & \\quad \\text{if ${y}_{t} \\ge {f}_{p,t}$}\n",
    "\\end{cases}\n",
    "$$\n",
    "\n",
    "This is sometimes called the “pinball loss function” because a graph of it resembles the trajectory of a ball on a pinball table. The multiplier of 2 is often omitted, but including it makes the interpretation a little easier. A low value of ${Q}_{p,t}$ indicates a better estimate of the quantile.\n",
    "\n",
    "The quantile score can be interpreted like an absolute error. In fact, when $p = 0.5$, the quantile score ${Q}_{0.5,t}$ is the same as the absolute error. For other values of $p$, the “error” $({y}_{t} - {f}_{p,t})$ is weighted to take account of how likely it is to be positive or negative. If $p>0.5$, ${Q}_{p,t}$ gives a heavier penalty when the observation is greater than the estimated quantile than when the observation is less than the estimated quantile. The reverse is true for $p<0.5$.\n",
    "\n",
    "In Figure 23, the one-step-ahead 10% quantile forecast (for 4 January 2016) is ${f}_{0.1,t} = 37.22$ and the observed value is 741.84. Then\n",
    "\n",
    "\\begin{gather*} \n",
    "{Q}_{0.5,t} = 2(1−0.1)(37.22−37.09) = 0.234\n",
    "\\end{gather*}\n",
    "\n",
    "This is easily computed using the `quantile_score()` function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "quantile_score() function: Return quantile score.\n",
    "\n",
    "* prob: Percentile analized.\n",
    "* f: Int. number, interval value predicted.\n",
    "* y: Int. number, value observed.'''\n",
    "\n",
    "def quantile_score(prob, f, y):\n",
    "\n",
    "    if y >= f:\n",
    "        q_score = 2 * prob * (y-f)\n",
    "    else:\n",
    "        q_score = 2 * (1-prob) * (f-y)  \n",
    "\n",
    "    return q_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.14467084350585965"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "quantile_score(0.1, df_plot_1['AutoARIMA-lo-80'][0], df_plot_1['y'][0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Winkler Score\n",
    "\n",
    "It is often of interest to evaluate a prediction interval, rather than a few quantiles, and the Winkler score proposed by Winkler (1972) is designed for this purpose. If the $100(1−α)%$ prediction interval at time $t$ is given by $[{ℓ}_{α,t},{u}_{α,t}]$, then the Winkler score is defined as the length of the interval plus a penalty if the observation is outside the interval:\n",
    "\n",
    "$$\n",
    "{W}_{α,t}=\n",
    "\\begin{cases}\n",
    "({u}_{α,t} - {ℓ}_{α,t}) + \\frac{2}{α} ({ℓ}_{α,t} - {y}_{t}) & \\quad \\text{if ${y}_{t} < {ℓ}_{α,t}$}\\\\ \n",
    "({u}_{α,t} - {ℓ}_{α,t}) & \\quad \\text{if ${ℓ}_{α,t} \\le {y}_{t} \\le {u}_{α,t}$}\\\\ \n",
    "({u}_{α,t} - {ℓ}_{α,t}) + \\frac{2}{α} ({y}_{t} - {u}_{α,t}) & \\quad \\text{if ${y}_{t} > {u}_{α,t}$}\\\\ \n",
    "\\end{cases}\n",
    "$$\n",
    "\n",
    "For observations that fall within the interval, the Winkler score is simply the length of the interval. So low scores are associated with narrow intervals. However, if the observation falls outside the interval, the penalty applies, with the penalty proportional to how far the observation is outside the interval.\n",
    "\n",
    "Prediction intervals are usually constructed from quantiles by setting ${ℓ}_{α,t} = {f}_{α/2,t}$ and ${u}_{α,t} = {f}_{1-α/2,t}$. If we add the corresponding quantile scores and divide by $α$, we get the Winkler score:\n",
    "\n",
    "\\begin{gather*} \n",
    "{W}_{α,t} = ({Q}_{α/2,t} + {Q}_{1-α/2,t})/α\n",
    "\\end{gather*}\n",
    "\n",
    "The one-step-ahead 80% interval shown in Figure 23 for 4 January 2016 is [37.22, 38.66], and the actual value was 37.09, so the Winkler score is\n",
    "\n",
    "\\begin{gather*} \n",
    "{W}_{α,t} = (38.66−37.22) + \\frac{2}{0.2} (37.22−37.09) = 2.74\n",
    "\\end{gather*}\n",
    "\n",
    "This is easily computed using the `winkler_score()` function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "winkler_score() function: Return winkler score.\n",
    "\n",
    "* a: Percentile analized.\n",
    "* l: Int. number, low interval value predicted.\n",
    "* u: Int. number, high interval value predicted.\n",
    "* y: Int. number, value observed.'''\n",
    "\n",
    "def winkler_score(a, l, u, y):\n",
    "\n",
    "    if y < l:\n",
    "        winkler_score = (u-l) + 2/a * (l-y)\n",
    "    if l < y < u:\n",
    "        winkler_score = (u-l)\n",
    "    if y > u:\n",
    "        winkler_score = (u-l) + 2/a * (y-u)\n",
    "\n",
    "    return winkler_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.1593513"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "winkler_score(0.1, df_plot_1['AutoARIMA-lo-80'][0], df_plot_1['AutoARIMA-hi-80'][0], df_plot_1['y'][0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Continuous Ranked Probability Score\n",
    "\n",
    "Often we are interested in the whole forecast distribution, rather than particular quantiles or prediction intervals. In that case, we can average the quantile scores over all values of $p$ to obtain the **Continuous Ranked Probability Score** or CRPS (Gneiting & Katzfuss, 2014).\n",
    "\n",
    "In the Google stock price example, we can compute the average CRPS value for all days in the test set. A CRPS value is a little like a weighted absolute error computed from the entire forecast distribution, where the weighting takes account of the probabilities."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "create_levels() function: Return a dataframe with AutoARIMA [10 to 90] prediction intervals.\n",
    "\n",
    "* train_dataframe: Dataframe with train values and time variable (ds).\n",
    "* test_dataframe: Dataframe with test values and time variable (ds).\n",
    "* season_length: int, number of observations per unit of time. Ex: 24 Hourly data.\n",
    "* freq: str, forecast frequency ('H', 'D', 'W', 'M', 'Q', 'Y').\n",
    "* horizon: int, forecast length. '''\n",
    "\n",
    "def create_levels(train_dataframe, test_dataframe, season_length, freq, horizon):\n",
    "\n",
    "    models = [AutoARIMA(season_length=season_length, approximation=True)]\n",
    "\n",
    "    fcst = StatsForecast(df=train_dataframe, \n",
    "                     models=models,\n",
    "                     freq=freq,\n",
    "                     n_jobs=-1)\n",
    "\n",
    "    levels = [90, 80, 70, 60, 50, 40, 30, 20, 10]\n",
    "    forecasts = fcst.forecast(h=horizon, level=levels)        \n",
    "    forecasts = forecasts.reset_index()   \n",
    "    df_fore = test_dataframe.merge(forecasts, left_on='ds', right_on='ds')\n",
    "\n",
    "    return df_fore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ds</th>\n",
       "      <th>y</th>\n",
       "      <th>unique_id_x</th>\n",
       "      <th>unique_id_y</th>\n",
       "      <th>AutoARIMA</th>\n",
       "      <th>AutoARIMA-lo-90</th>\n",
       "      <th>AutoARIMA-lo-80</th>\n",
       "      <th>AutoARIMA-lo-70</th>\n",
       "      <th>AutoARIMA-lo-60</th>\n",
       "      <th>AutoARIMA-lo-50</th>\n",
       "      <th>...</th>\n",
       "      <th>AutoARIMA-lo-10</th>\n",
       "      <th>AutoARIMA-hi-10</th>\n",
       "      <th>AutoARIMA-hi-20</th>\n",
       "      <th>AutoARIMA-hi-30</th>\n",
       "      <th>AutoARIMA-hi-40</th>\n",
       "      <th>AutoARIMA-hi-50</th>\n",
       "      <th>AutoARIMA-hi-60</th>\n",
       "      <th>AutoARIMA-hi-70</th>\n",
       "      <th>AutoARIMA-hi-80</th>\n",
       "      <th>AutoARIMA-hi-90</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016-01-04</td>\n",
       "      <td>37.972000</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>36.800831</td>\n",
       "      <td>37.248646</td>\n",
       "      <td>37.550785</td>\n",
       "      <td>37.790916</td>\n",
       "      <td>37.996929</td>\n",
       "      <td>...</td>\n",
       "      <td>38.673428</td>\n",
       "      <td>38.983215</td>\n",
       "      <td>39.140602</td>\n",
       "      <td>39.303276</td>\n",
       "      <td>39.474712</td>\n",
       "      <td>39.659714</td>\n",
       "      <td>39.865726</td>\n",
       "      <td>40.105858</td>\n",
       "      <td>40.407997</td>\n",
       "      <td>40.855812</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016-01-05</td>\n",
       "      <td>38.076500</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>36.549133</td>\n",
       "      <td>37.052544</td>\n",
       "      <td>37.392189</td>\n",
       "      <td>37.662132</td>\n",
       "      <td>37.893715</td>\n",
       "      <td>...</td>\n",
       "      <td>38.654198</td>\n",
       "      <td>39.002445</td>\n",
       "      <td>39.179371</td>\n",
       "      <td>39.362240</td>\n",
       "      <td>39.554955</td>\n",
       "      <td>39.762924</td>\n",
       "      <td>39.994511</td>\n",
       "      <td>40.264454</td>\n",
       "      <td>40.604099</td>\n",
       "      <td>41.107510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016-01-06</td>\n",
       "      <td>37.966499</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>36.322594</td>\n",
       "      <td>36.876038</td>\n",
       "      <td>37.249447</td>\n",
       "      <td>37.546219</td>\n",
       "      <td>37.800823</td>\n",
       "      <td>...</td>\n",
       "      <td>38.636890</td>\n",
       "      <td>39.019749</td>\n",
       "      <td>39.214264</td>\n",
       "      <td>39.415310</td>\n",
       "      <td>39.627178</td>\n",
       "      <td>39.855820</td>\n",
       "      <td>40.110424</td>\n",
       "      <td>40.407196</td>\n",
       "      <td>40.780602</td>\n",
       "      <td>41.334049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016-01-07</td>\n",
       "      <td>37.049999</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>36.114902</td>\n",
       "      <td>36.714222</td>\n",
       "      <td>37.118576</td>\n",
       "      <td>37.439949</td>\n",
       "      <td>37.715656</td>\n",
       "      <td>...</td>\n",
       "      <td>38.621025</td>\n",
       "      <td>39.035618</td>\n",
       "      <td>39.246254</td>\n",
       "      <td>39.463963</td>\n",
       "      <td>39.693394</td>\n",
       "      <td>39.940987</td>\n",
       "      <td>40.216694</td>\n",
       "      <td>40.538063</td>\n",
       "      <td>40.942421</td>\n",
       "      <td>41.541740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016-01-08</td>\n",
       "      <td>36.545502</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>35.922016</td>\n",
       "      <td>36.563938</td>\n",
       "      <td>36.997040</td>\n",
       "      <td>37.341255</td>\n",
       "      <td>37.636559</td>\n",
       "      <td>...</td>\n",
       "      <td>38.606289</td>\n",
       "      <td>39.050354</td>\n",
       "      <td>39.275963</td>\n",
       "      <td>39.509148</td>\n",
       "      <td>39.754890</td>\n",
       "      <td>40.020084</td>\n",
       "      <td>40.315388</td>\n",
       "      <td>40.659603</td>\n",
       "      <td>41.092705</td>\n",
       "      <td>41.734627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2016-01-11</td>\n",
       "      <td>36.653500</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>35.408009</td>\n",
       "      <td>36.163460</td>\n",
       "      <td>36.673161</td>\n",
       "      <td>37.078251</td>\n",
       "      <td>37.425785</td>\n",
       "      <td>...</td>\n",
       "      <td>38.567020</td>\n",
       "      <td>39.089622</td>\n",
       "      <td>39.355133</td>\n",
       "      <td>39.629559</td>\n",
       "      <td>39.918762</td>\n",
       "      <td>40.230858</td>\n",
       "      <td>40.578392</td>\n",
       "      <td>40.983482</td>\n",
       "      <td>41.493183</td>\n",
       "      <td>42.248634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2016-01-12</td>\n",
       "      <td>37.266998</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>35.253056</td>\n",
       "      <td>36.042732</td>\n",
       "      <td>36.575523</td>\n",
       "      <td>36.998966</td>\n",
       "      <td>37.362247</td>\n",
       "      <td>...</td>\n",
       "      <td>38.555183</td>\n",
       "      <td>39.101460</td>\n",
       "      <td>39.378998</td>\n",
       "      <td>39.665855</td>\n",
       "      <td>39.968163</td>\n",
       "      <td>40.294395</td>\n",
       "      <td>40.657673</td>\n",
       "      <td>41.081120</td>\n",
       "      <td>41.613911</td>\n",
       "      <td>42.403587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2016-01-13</td>\n",
       "      <td>35.978500</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>35.104546</td>\n",
       "      <td>35.927025</td>\n",
       "      <td>36.481945</td>\n",
       "      <td>36.922981</td>\n",
       "      <td>37.301350</td>\n",
       "      <td>...</td>\n",
       "      <td>38.543839</td>\n",
       "      <td>39.112804</td>\n",
       "      <td>39.401871</td>\n",
       "      <td>39.700645</td>\n",
       "      <td>40.015507</td>\n",
       "      <td>40.355293</td>\n",
       "      <td>40.733662</td>\n",
       "      <td>41.174698</td>\n",
       "      <td>41.729618</td>\n",
       "      <td>42.552094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2016-01-14</td>\n",
       "      <td>36.569500</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.961739</td>\n",
       "      <td>35.815758</td>\n",
       "      <td>36.391960</td>\n",
       "      <td>36.849911</td>\n",
       "      <td>37.242786</td>\n",
       "      <td>...</td>\n",
       "      <td>38.532928</td>\n",
       "      <td>39.123714</td>\n",
       "      <td>39.423870</td>\n",
       "      <td>39.734100</td>\n",
       "      <td>40.061039</td>\n",
       "      <td>40.413857</td>\n",
       "      <td>40.806732</td>\n",
       "      <td>41.264683</td>\n",
       "      <td>41.840885</td>\n",
       "      <td>42.694904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2016-01-15</td>\n",
       "      <td>35.524502</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.824017</td>\n",
       "      <td>35.708458</td>\n",
       "      <td>36.305183</td>\n",
       "      <td>36.779442</td>\n",
       "      <td>37.186314</td>\n",
       "      <td>...</td>\n",
       "      <td>38.522408</td>\n",
       "      <td>39.134235</td>\n",
       "      <td>39.445080</td>\n",
       "      <td>39.766361</td>\n",
       "      <td>40.104946</td>\n",
       "      <td>40.470329</td>\n",
       "      <td>40.877201</td>\n",
       "      <td>41.351460</td>\n",
       "      <td>41.948185</td>\n",
       "      <td>42.832623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2016-01-19</td>\n",
       "      <td>35.953999</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.314968</td>\n",
       "      <td>35.311844</td>\n",
       "      <td>35.984428</td>\n",
       "      <td>36.518978</td>\n",
       "      <td>36.977573</td>\n",
       "      <td>...</td>\n",
       "      <td>38.483517</td>\n",
       "      <td>39.173126</td>\n",
       "      <td>39.523487</td>\n",
       "      <td>39.885612</td>\n",
       "      <td>40.267235</td>\n",
       "      <td>40.679070</td>\n",
       "      <td>41.137665</td>\n",
       "      <td>41.672215</td>\n",
       "      <td>42.344799</td>\n",
       "      <td>43.341675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2016-01-20</td>\n",
       "      <td>35.928001</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.196442</td>\n",
       "      <td>35.219494</td>\n",
       "      <td>35.909740</td>\n",
       "      <td>36.458328</td>\n",
       "      <td>36.928970</td>\n",
       "      <td>...</td>\n",
       "      <td>38.474461</td>\n",
       "      <td>39.182182</td>\n",
       "      <td>39.541744</td>\n",
       "      <td>39.913376</td>\n",
       "      <td>40.305023</td>\n",
       "      <td>40.727673</td>\n",
       "      <td>41.198315</td>\n",
       "      <td>41.746902</td>\n",
       "      <td>42.437149</td>\n",
       "      <td>43.460201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2016-01-21</td>\n",
       "      <td>36.333500</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>34.080872</td>\n",
       "      <td>35.129448</td>\n",
       "      <td>35.836922</td>\n",
       "      <td>36.399197</td>\n",
       "      <td>36.881577</td>\n",
       "      <td>...</td>\n",
       "      <td>38.465633</td>\n",
       "      <td>39.191010</td>\n",
       "      <td>39.559544</td>\n",
       "      <td>39.940453</td>\n",
       "      <td>40.341869</td>\n",
       "      <td>40.775063</td>\n",
       "      <td>41.257446</td>\n",
       "      <td>41.819721</td>\n",
       "      <td>42.527195</td>\n",
       "      <td>43.575771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>2016-01-22</td>\n",
       "      <td>37.272999</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>33.968048</td>\n",
       "      <td>35.041546</td>\n",
       "      <td>35.765831</td>\n",
       "      <td>36.341469</td>\n",
       "      <td>36.835316</td>\n",
       "      <td>...</td>\n",
       "      <td>38.457012</td>\n",
       "      <td>39.199631</td>\n",
       "      <td>39.576920</td>\n",
       "      <td>39.966881</td>\n",
       "      <td>40.377838</td>\n",
       "      <td>40.821327</td>\n",
       "      <td>41.315174</td>\n",
       "      <td>41.890812</td>\n",
       "      <td>42.615097</td>\n",
       "      <td>43.688595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2016-01-25</td>\n",
       "      <td>36.681000</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>33.644287</td>\n",
       "      <td>34.789295</td>\n",
       "      <td>35.561829</td>\n",
       "      <td>36.175812</td>\n",
       "      <td>36.702553</td>\n",
       "      <td>...</td>\n",
       "      <td>38.432278</td>\n",
       "      <td>39.224365</td>\n",
       "      <td>39.626789</td>\n",
       "      <td>40.042725</td>\n",
       "      <td>40.481056</td>\n",
       "      <td>40.954090</td>\n",
       "      <td>41.480831</td>\n",
       "      <td>42.094814</td>\n",
       "      <td>42.867348</td>\n",
       "      <td>44.012352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>2016-01-26</td>\n",
       "      <td>36.689499</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>33.540771</td>\n",
       "      <td>34.708645</td>\n",
       "      <td>35.496601</td>\n",
       "      <td>36.122845</td>\n",
       "      <td>36.660107</td>\n",
       "      <td>...</td>\n",
       "      <td>38.424370</td>\n",
       "      <td>39.232273</td>\n",
       "      <td>39.642731</td>\n",
       "      <td>40.066975</td>\n",
       "      <td>40.514061</td>\n",
       "      <td>40.996536</td>\n",
       "      <td>41.533798</td>\n",
       "      <td>42.160042</td>\n",
       "      <td>42.947998</td>\n",
       "      <td>44.115871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>2016-01-27</td>\n",
       "      <td>35.879002</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>33.439243</td>\n",
       "      <td>34.629539</td>\n",
       "      <td>35.432629</td>\n",
       "      <td>36.070896</td>\n",
       "      <td>36.618473</td>\n",
       "      <td>...</td>\n",
       "      <td>38.416615</td>\n",
       "      <td>39.240028</td>\n",
       "      <td>39.658367</td>\n",
       "      <td>40.090755</td>\n",
       "      <td>40.546429</td>\n",
       "      <td>41.038170</td>\n",
       "      <td>41.585747</td>\n",
       "      <td>42.224014</td>\n",
       "      <td>43.027103</td>\n",
       "      <td>44.217396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>2016-01-28</td>\n",
       "      <td>37.415001</td>\n",
       "      <td>Close price</td>\n",
       "      <td>Close price</td>\n",
       "      <td>38.82832</td>\n",
       "      <td>33.339596</td>\n",
       "      <td>34.551899</td>\n",
       "      <td>35.369839</td>\n",
       "      <td>36.019909</td>\n",
       "      <td>36.577610</td>\n",
       "      <td>...</td>\n",
       "      <td>38.409000</td>\n",
       "      <td>39.247643</td>\n",
       "      <td>39.673717</td>\n",
       "      <td>40.114101</td>\n",
       "      <td>40.578197</td>\n",
       "      <td>41.079033</td>\n",
       "      <td>41.636734</td>\n",
       "      <td>42.286804</td>\n",
       "      <td>43.104740</td>\n",
       "      <td>44.317047</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>18 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           ds          y  unique_id_x  unique_id_y  AutoARIMA  \\\n",
       "0  2016-01-04  37.972000  Close price  Close price   38.82832   \n",
       "1  2016-01-05  38.076500  Close price  Close price   38.82832   \n",
       "2  2016-01-06  37.966499  Close price  Close price   38.82832   \n",
       "3  2016-01-07  37.049999  Close price  Close price   38.82832   \n",
       "4  2016-01-08  36.545502  Close price  Close price   38.82832   \n",
       "5  2016-01-11  36.653500  Close price  Close price   38.82832   \n",
       "6  2016-01-12  37.266998  Close price  Close price   38.82832   \n",
       "7  2016-01-13  35.978500  Close price  Close price   38.82832   \n",
       "8  2016-01-14  36.569500  Close price  Close price   38.82832   \n",
       "9  2016-01-15  35.524502  Close price  Close price   38.82832   \n",
       "10 2016-01-19  35.953999  Close price  Close price   38.82832   \n",
       "11 2016-01-20  35.928001  Close price  Close price   38.82832   \n",
       "12 2016-01-21  36.333500  Close price  Close price   38.82832   \n",
       "13 2016-01-22  37.272999  Close price  Close price   38.82832   \n",
       "14 2016-01-25  36.681000  Close price  Close price   38.82832   \n",
       "15 2016-01-26  36.689499  Close price  Close price   38.82832   \n",
       "16 2016-01-27  35.879002  Close price  Close price   38.82832   \n",
       "17 2016-01-28  37.415001  Close price  Close price   38.82832   \n",
       "\n",
       "    AutoARIMA-lo-90  AutoARIMA-lo-80  AutoARIMA-lo-70  AutoARIMA-lo-60  \\\n",
       "0         36.800831        37.248646        37.550785        37.790916   \n",
       "1         36.549133        37.052544        37.392189        37.662132   \n",
       "2         36.322594        36.876038        37.249447        37.546219   \n",
       "3         36.114902        36.714222        37.118576        37.439949   \n",
       "4         35.922016        36.563938        36.997040        37.341255   \n",
       "5         35.408009        36.163460        36.673161        37.078251   \n",
       "6         35.253056        36.042732        36.575523        36.998966   \n",
       "7         35.104546        35.927025        36.481945        36.922981   \n",
       "8         34.961739        35.815758        36.391960        36.849911   \n",
       "9         34.824017        35.708458        36.305183        36.779442   \n",
       "10        34.314968        35.311844        35.984428        36.518978   \n",
       "11        34.196442        35.219494        35.909740        36.458328   \n",
       "12        34.080872        35.129448        35.836922        36.399197   \n",
       "13        33.968048        35.041546        35.765831        36.341469   \n",
       "14        33.644287        34.789295        35.561829        36.175812   \n",
       "15        33.540771        34.708645        35.496601        36.122845   \n",
       "16        33.439243        34.629539        35.432629        36.070896   \n",
       "17        33.339596        34.551899        35.369839        36.019909   \n",
       "\n",
       "    AutoARIMA-lo-50  ...  AutoARIMA-lo-10  AutoARIMA-hi-10  AutoARIMA-hi-20  \\\n",
       "0         37.996929  ...        38.673428        38.983215        39.140602   \n",
       "1         37.893715  ...        38.654198        39.002445        39.179371   \n",
       "2         37.800823  ...        38.636890        39.019749        39.214264   \n",
       "3         37.715656  ...        38.621025        39.035618        39.246254   \n",
       "4         37.636559  ...        38.606289        39.050354        39.275963   \n",
       "5         37.425785  ...        38.567020        39.089622        39.355133   \n",
       "6         37.362247  ...        38.555183        39.101460        39.378998   \n",
       "7         37.301350  ...        38.543839        39.112804        39.401871   \n",
       "8         37.242786  ...        38.532928        39.123714        39.423870   \n",
       "9         37.186314  ...        38.522408        39.134235        39.445080   \n",
       "10        36.977573  ...        38.483517        39.173126        39.523487   \n",
       "11        36.928970  ...        38.474461        39.182182        39.541744   \n",
       "12        36.881577  ...        38.465633        39.191010        39.559544   \n",
       "13        36.835316  ...        38.457012        39.199631        39.576920   \n",
       "14        36.702553  ...        38.432278        39.224365        39.626789   \n",
       "15        36.660107  ...        38.424370        39.232273        39.642731   \n",
       "16        36.618473  ...        38.416615        39.240028        39.658367   \n",
       "17        36.577610  ...        38.409000        39.247643        39.673717   \n",
       "\n",
       "    AutoARIMA-hi-30  AutoARIMA-hi-40  AutoARIMA-hi-50  AutoARIMA-hi-60  \\\n",
       "0         39.303276        39.474712        39.659714        39.865726   \n",
       "1         39.362240        39.554955        39.762924        39.994511   \n",
       "2         39.415310        39.627178        39.855820        40.110424   \n",
       "3         39.463963        39.693394        39.940987        40.216694   \n",
       "4         39.509148        39.754890        40.020084        40.315388   \n",
       "5         39.629559        39.918762        40.230858        40.578392   \n",
       "6         39.665855        39.968163        40.294395        40.657673   \n",
       "7         39.700645        40.015507        40.355293        40.733662   \n",
       "8         39.734100        40.061039        40.413857        40.806732   \n",
       "9         39.766361        40.104946        40.470329        40.877201   \n",
       "10        39.885612        40.267235        40.679070        41.137665   \n",
       "11        39.913376        40.305023        40.727673        41.198315   \n",
       "12        39.940453        40.341869        40.775063        41.257446   \n",
       "13        39.966881        40.377838        40.821327        41.315174   \n",
       "14        40.042725        40.481056        40.954090        41.480831   \n",
       "15        40.066975        40.514061        40.996536        41.533798   \n",
       "16        40.090755        40.546429        41.038170        41.585747   \n",
       "17        40.114101        40.578197        41.079033        41.636734   \n",
       "\n",
       "    AutoARIMA-hi-70  AutoARIMA-hi-80  AutoARIMA-hi-90  \n",
       "0         40.105858        40.407997        40.855812  \n",
       "1         40.264454        40.604099        41.107510  \n",
       "2         40.407196        40.780602        41.334049  \n",
       "3         40.538063        40.942421        41.541740  \n",
       "4         40.659603        41.092705        41.734627  \n",
       "5         40.983482        41.493183        42.248634  \n",
       "6         41.081120        41.613911        42.403587  \n",
       "7         41.174698        41.729618        42.552094  \n",
       "8         41.264683        41.840885        42.694904  \n",
       "9         41.351460        41.948185        42.832623  \n",
       "10        41.672215        42.344799        43.341675  \n",
       "11        41.746902        42.437149        43.460201  \n",
       "12        41.819721        42.527195        43.575771  \n",
       "13        41.890812        42.615097        43.688595  \n",
       "14        42.094814        42.867348        44.012352  \n",
       "15        42.160042        42.947998        44.115871  \n",
       "16        42.224014        43.027103        44.217396  \n",
       "17        42.286804        43.104740        44.317047  \n",
       "\n",
       "[18 rows x 23 columns]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_fore = create_levels(google_train, google_test, 5, 'D', 30)\n",
    "df_fore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "crps() function: Return Continuous Ranked Probability Score.\n",
    "\n",
    "* df: Dataframe with test values and forecast intervals values.\n",
    "* model: Model analized name.\n",
    "* levels_list: list, levels analized. '''\n",
    "\n",
    "def crps(df, model, levels_list):\n",
    "\n",
    "    import re\n",
    "\n",
    "    obs_values = df.y\n",
    "    q_scores = []\n",
    "    \n",
    "    lvl_lo, lvl_hi = model+'-lo-', model+'-hi-'\n",
    "    lvl_low = [lvl_lo+str(i) for i in levels_list]\n",
    "    lvl_high = [lvl_hi+str(i) for i in levels_list]\n",
    "    lvl_list = lvl_low + lvl_high\n",
    "\n",
    "    for level in lvl_list:\n",
    "        s = -[float(s) for s in re.findall(r'-?\\d+\\.?\\d*', level)][0]\n",
    "        prob = (1- s/100) / 2\n",
    "        res = obs_values-df[level]\n",
    "\n",
    "        for e in res:\n",
    "\n",
    "            if e > 0:\n",
    "                q_score = 2 * prob * (obs_values-df[level])\n",
    "                \n",
    "            else:\n",
    "                q_score = 2 * (1-prob) * (df[level]-obs_values)  \n",
    "            \n",
    "            q_scores.append(q_score)\n",
    "\n",
    "    crps = np.average(q_scores)\n",
    "\n",
    "    return crps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.375661729982259"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "crps(df_fore, 'AutoARIMA', [90, 80, 70, 60])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, the naïve method is giving better distributional forecasts than the drift or mean methods.\n",
    "\n",
    "## Scale-free comparisons using skill scores\n",
    "\n",
    "As with point forecasts, it is useful to be able to compare the distributional forecast accuracy of several methods across series on different scales. For point forecasts, we used scaled errors for that purpose. Another approach is to use skill scores. These can be used for both point forecast accuracy and distributional forecast accuracy.\n",
    "\n",
    "With skill scores, we compute a forecast accuracy measure relative to some benchmark method. For example, if we use the naïve method as a benchmark, and also compute forecasts using the drift method, we can compute the CRPS skill score of the drift method relative to the naïve method as\n",
    "\n",
    "\\begin{gather*} \n",
    "\\frac{{CRPS}_{Naive} - {CRPS}_{Drift}}{{CRPS}_{Naive}}\n",
    "\\end{gather*}\n",
    "\n",
    "This gives the proportion that the drift method improves over the naïve method based on CRPS. It is easy to compute using the accuracy() function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'skill_score() example with drift and naive crps'"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''skill_score() example with drift and naive crps''' "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Of course, the skill score for the naïve method is 0 because it can’t improve on itself. The other two methods have larger CRPS values than naïve, so the skills scores are negative; the drift method is 26.6% worse than the Auto ARIMA method.\n",
    "\n",
    "The `skill_score()` function will always compute the CRPS for the appropriate benchmark forecasts, even if these are not included in the fable object. When the data are seasonal, the benchmark used is the seasonal naïve method rather than the naïve method. To ensure that the same training data are used for the benchmark forecasts, it is important that the data provided to the accuracy() function starts at the same time as the training data.\n",
    "\n",
    "The `skill_score()` function can be used with any accuracy measure. For example, skill_score(MSE) provides a way of comparing MSE values across diverse series. However, it is important that the test set is large enough to allow reliable calculation of the error measure, especially in the denominator. For that reason, MASE or RMSSE are often preferable scale-free measures for point forecast accuracy.\n",
    "\n",
    "***\n",
    "\n",
    "## Bibliography\n",
    "+ Gneiting, T., & Katzfuss, M. (2014). Probabilistic forecasting. Annual Review of Statistics and Its Application, 1(1), 125–151.\n",
    "+ Winkler, R. L. (1972). A decision-theoretic approach to interval estimation. Journal of the American Statistical Association, 67(337), 187–191."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.7 64-bit",
   "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": "3.10.7"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "b1896fd924129e7fd2d4d004268bb9f63a53b23ae7875be394e789ef8193cc30"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}