{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Algoritmos supervisados: un proyecto sobre predicción temporal\n",
    "\n",
    "En esta sesión seguimos profundizando en técnicas de *Machine Learning* aplicadas a **predicción** mediante varios proyectos. Como habéis podido comprobar con en el notebook anterior, la aplicación de modelos no conlleva un trabajo excesivo, lo importante es la coherencia del proceso con el objetivo pretendido. Por tanto, lo complejo es determinar el modelo.\n",
    "\n",
    "En muchos problemas de predicción aparece el **tiempo**. Por ello, en este proyecto trabajaremos con datos de cotización de varias compañías para predecir **valores de cierre futuros**. Nuestro objetivo será cómo utilizar el **tiempo** dentro del modelo. \n",
    "\n",
    "Para series temporales, los modelos realmente específicos (porque modelan explícitamente la dependencia temporal) son, por ejemplo:\n",
    "- Modelos autoregresivos y de medias móviles: AR, MA, ARMA.\n",
    "- ARIMA y su extensión estacional SARIMA.\n",
    "- Modelos de suavizado exponencial (Holt, Holt‑Winters).\n",
    "- Redes neuronales con memoria temporal: RNN, LSTM, GRU.\n",
    "- Arquitecturas secuencia‑a‑secuencia y modelos con atención aplicados a series.\n",
    "\n",
    "En ML, modelos como regresión lineal, SVM, árboles, Random Forest, etc., no son específicos: solo sirven para series si \"gestionamos\" el tiempo en features (lags, ventanas, etc.).\n",
    "\n",
    "¿Entonces vamos a perder el tiempo? NO! Gracias a este proyecto veremos las limitaciones de los modelos de ML en la gestión temporal y también replantearemos el tiempo para que pueda ser consumido por el modelo como cualquier otra característica.\n",
    "\n",
    "A nivel general, en problemas de predicción (clasificación o regresión) existen múltiples algoritmos como:\n",
    "- Linear regressión / Regresión Lineal: https://es.wikipedia.org/wiki/Regresi%C3%B3n_lineal \n",
    "- Regresión logística (para predicción de clases) : https://es.wikipedia.org/wiki/Regresi%C3%B3n_log%C3%ADstica\n",
    "- Máquinas de vectores de soporte / SVM: https://es.wikipedia.org/wiki/M%C3%A1quinas_de_vectores_de_soporte\n",
    "- CART / Classification And Regression Trees https://www.nature.com/articles/nmeth.4370\n",
    "- Gradient Boosting / Potenciación del gradiente: https://es.wikipedia.org/wiki/Gradient_boosting\n",
    "- Random Forest / Bosques aleatorios: https://es.wikipedia.org/wiki/Random_forest\n",
    "- Artificial Neural Networks / Redes neuronales artificiales:  https://en.wikipedia.org/wiki/Artificial_neural_network\n",
    "- k-vecinos más próximos  / K-nearest neighbors (k-NN): https://es.wikipedia.org/wiki/K_vecinos_m%C3%A1s_pr%C3%B3ximos\n",
    "- Análisis discriminante lineal (LDA): https://es.wikipedia.org/wiki/An%C3%A1lisis_discriminante_lineal\n",
    "- etc.\n",
    "\n",
    "\n",
    "La gran mayoría están implementados en la librería de python [scikit-learn/supervised_learning](https://scikit-learn.org/stable/supervised_learning.html)\n",
    "\n",
    "<img src=\"images/ModelSelection_ref_book_5.png\" width=80%/>\n",
    "\n",
    "> Fuente de la imagen: [5] Machine Learning & Data Science Blueprints for Finance\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preparemos los datos\n",
    "\n",
    "En este caso, usaremos los datos obtenidos en [yahoo finance](https://es.finance.yahoo.com/?guccounter=1) sobre la compañia de *Google*. Los datos se encuentran en: ```data/GOOGL.csv'''\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</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>2017-01-03</th>\n",
       "      <td>2017-01-03</td>\n",
       "      <td>40.030998</td>\n",
       "      <td>40.571999</td>\n",
       "      <td>39.844501</td>\n",
       "      <td>40.400501</td>\n",
       "      <td>40.400501</td>\n",
       "      <td>39180000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-04</th>\n",
       "      <td>2017-01-04</td>\n",
       "      <td>40.494499</td>\n",
       "      <td>40.671501</td>\n",
       "      <td>40.205502</td>\n",
       "      <td>40.388500</td>\n",
       "      <td>40.388500</td>\n",
       "      <td>30306000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-05</th>\n",
       "      <td>2017-01-05</td>\n",
       "      <td>40.375000</td>\n",
       "      <td>40.687000</td>\n",
       "      <td>40.296001</td>\n",
       "      <td>40.651001</td>\n",
       "      <td>40.651001</td>\n",
       "      <td>26810000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-06</th>\n",
       "      <td>2017-01-06</td>\n",
       "      <td>40.749500</td>\n",
       "      <td>41.448002</td>\n",
       "      <td>40.575001</td>\n",
       "      <td>41.260502</td>\n",
       "      <td>41.260502</td>\n",
       "      <td>40342000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-09</th>\n",
       "      <td>2017-01-09</td>\n",
       "      <td>41.318501</td>\n",
       "      <td>41.521500</td>\n",
       "      <td>41.081001</td>\n",
       "      <td>41.359001</td>\n",
       "      <td>41.359001</td>\n",
       "      <td>28178000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Date       Open       High        Low      Close  Adj Close  \\\n",
       "Date                                                                            \n",
       "2017-01-03  2017-01-03  40.030998  40.571999  39.844501  40.400501  40.400501   \n",
       "2017-01-04  2017-01-04  40.494499  40.671501  40.205502  40.388500  40.388500   \n",
       "2017-01-05  2017-01-05  40.375000  40.687000  40.296001  40.651001  40.651001   \n",
       "2017-01-06  2017-01-06  40.749500  41.448002  40.575001  41.260502  41.260502   \n",
       "2017-01-09  2017-01-09  41.318501  41.521500  41.081001  41.359001  41.359001   \n",
       "\n",
       "              Volume  \n",
       "Date                  \n",
       "2017-01-03  39180000  \n",
       "2017-01-04  30306000  \n",
       "2017-01-05  26810000  \n",
       "2017-01-06  40342000  \n",
       "2017-01-09  28178000  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "data = pd.read_csv(\"data/GOOGL.csv\")\n",
    "data.index = pd.DatetimeIndex(data[\"Date\"])\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2017-01-03 00:00:00\n",
      "2020-12-30 00:00:00\n"
     ]
    }
   ],
   "source": [
    "print(data.index.min())\n",
    "print(data.index.max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['Date', 'Open', 'High', 'Low', 'Close', 'Adj Close', 'Volume'], dtype='object')\n"
     ]
    }
   ],
   "source": [
    "print(data.columns) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date\n",
       "2017-01-03    40.400501\n",
       "2017-01-04    40.388500\n",
       "2017-01-05    40.651001\n",
       "2017-01-06    41.260502\n",
       "2017-01-09    41.359001\n",
       "2017-01-10    41.300499\n",
       "2017-01-11    41.493000\n",
       "2017-01-12    41.476501\n",
       "2017-01-13    41.547001\n",
       "2017-01-17    41.373001\n",
       "Name: Close, dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Nuestra variable objetivo será el valor de cierre de la compañia.\n",
    "data[\"Close\"][:10] "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Análisis de una serie temporal\n",
    "\n",
    "Antes de construir un modelo de predicción conviene entender la serie temporal: ver si presenta tendencia, patrones estacionales o mucho ruido.\n",
    "\n",
    "Una herramienta clásica es la descomposición de la serie temporal, que separa la serie observada en varios componentes. En el caso aditivo se escribe como:\n",
    "$Y[t] = T[t] + S[t] + I[t]$ \n",
    "\n",
    "- T_{t}, componente de tendencia, que recoge la evolución a largo plazo (creciente, decreciente, etc.).\n",
    "- S_{t}, componente estacional, que recoge patrones que se repiten con un periodo fijo (por ejemplo, días de la semana, meses, trimestres) [1]\n",
    "- I_t, componente irregular o ruido, que recoge las variaciones aleatorias no explicadas por los otros componentes.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[2mResolved \u001b[1m99 packages\u001b[0m \u001b[2min 10ms\u001b[0m\u001b[0m\n",
      "\u001b[2mAudited \u001b[1m94 packages\u001b[0m \u001b[2min 0.34ms\u001b[0m\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "!uv add statsmodels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "\n",
    "result = seasonal_decompose(data['Close'], model=\"additive\", period=30)\n",
    "trend = result.trend\n",
    "seasonal = result.seasonal\n",
    "resid = result.resid\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/QAAAJ6CAYAAACG6Al+AAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjcsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvTLEjVAAAAAlwSFlzAAAPYQAAD2EBqD+naQABAABJREFUeJzsnQV4FFcXhk/cjXhIAiFIcHeXAoV6S4221N2dv0K91CltKXWnpS1VtLi7a0gIcXf3/M+5u7OZ3WwgZGc2u5vvfdgnu7PL2J07c88953zHrqGhoYEAAAAAAAAAAABgVdi39Q4AAAAAAAAAAADgwoFBDwAAAAAAAAAAWCEw6AEAAAAAAAAAACsEBj0AAAAAAAAAAGCFwKAHAAAAAAAAAACsEBj0AAAAAAAAAACAFQKDHgAAAAAAAAAAsEJg0AMAAAAAAAAAAFaIY1vvgKVTX19P6enp5OXlRXZ2dm29OwAAAAAAAAAAbJyGhgYqKSmhsLAwsrdv3g8Pg/48sDEfERGhdPsAAAAAAAAAAADnJCUlhcLDw5v9Hgb9eWDPvHQivb29z/dzAAAAAAAAAADAJIqLi4VjWbJHmwMG/XmQwuzZmIdBDwAAAAAAAADAXJwv7RuieAAAAAAAAAAAgBUCDz0AAAAAAAAAAItn0aZ4+npbInXyd6eLegXTrMHh5O/pQu0ZeOgBAAAAAAAAAFg83+1IpNzSKtqfVEDzV52i8e9sov1J+dSegUEPAAAAAAAAAMCiKSqvoaziKvH+rrFR1CXQg0qraunNlaeoPQODHgAAAAAAAACARXMio1j8Dfdzo+dm9qJPbhwkPp/MKBY129sryKEHAAAAAAAAAGCRLN2bTF9uPUvDu3QQn3sEa8q4dQ3yJEd7OyqrrqOMokoK83Wj9gg89AAAAAAAAAAA2oSs4kr6YVcSvbsmllYfy9T7rqSyhl5fcZLiskvpx13JYlknfw/x18nBnqICNO/v+G4fpRdWnHM7ibll9MWWBKqurSdbAh56AAAAAAAAAACqw6Hx+5IKyN/DmSpq6iiygzs9tOQg7UnUCNvZ2xFtenIiRfq7U2VNHd32zV4qrqzVWwcr3EsM6ewnjH0Ou7/927307W3DyN/TWRj7cmrr6unyT7ZTUUWN+HzXuC4209ow6AEAAAAAAAAAqM6SPcn03J/HdJ/Zw55WoPGsR3Rwo5T8Chr3zkZ69Yo+wrhn459hwz85v1z3XuL+CV3p5z0p4v2pzBIa8eZ66ujrRkvvGUHhfo2/+2Vvis6Y/3lvsk0Z9Ai5BwAAAAAAAACgOsv2p+p9PptbRtV19WRnR/Tk1B665S/8dUx43RkvV0f68Y7hIl+eYXV7iYgO7nTwhYt0AnlMWmGFyLmXYCX8BetO6z4n5JRRXqlGLd8WgEEPAAAAAAAAAEBVOIT+QHKheL/tmYnk4+ak+25ghC+N7hqg9/utcbni74uX9BIh+J/fMpjevqafLodews/DmWb2C6XnZ/YUXn1m+ZF0MVmw80wefb75DOWWVotoAGkyYFeC7dSuh0EPAAAAAAAAAEBViis1Ie/sjQ/zcaM7xkTpvrtvQlcK8HShS/qF6pYl5WlC7LsEeoq/k2KC6dohEc2u/86xXSj2tYupg4ezMOAnvruJbvhiFy3cEC++f2Z6DxrXLVC835mgmSywBZBDDwAAAAAAAAA2Bqu+/3kwjW4cFim82G1NqVbcztPZkezt7eie8V1ocCc/Gh7VgRy1InYf3ziInphaRs8sO0J7zuaTk4MdddUa9C3BycFeTAp8vzNJb/m03sE0rXcI+bo7i9D9Kb2CyVaAQQ8AAAAAAAAANsYtX++h+OxSOppaRItvHtzWuyNy2RlPV40J6uLo0CTMnuHQ+KV3j6CdCXlkb2dHPu6Nofkt4clpPSjY25U6+3vQU78fFvXpP7huANnZ2dGILv7iZUvAoAcAAAAAAAAAG4ONeWZDbDZZApKHnkXuzgcb36Oimxr7LcHb1YkemNhVvB/epQO5OTmQu7Ptmr22e2QAAAAAAAAA0M5xMajJ3laUSB56F/OZoAGeLmTrWEbrAgAAAAAAAABQHGdHyzD5dDn0rhcWQg/OjWW0LgAAAAAAAAAAxWGhOEtAyqH3MqOHvj1gGa0LAAAAAAAAAEARyqs1xrNFeejbIOS+PWAZrQsAAAAAAAAAQBHySqt17+vqGyzirJboQu5h0CsJziYAAAAAAAAAWDk74nPpsy0JdOXAjuRgb6dbXlJZQ5aAtB/w0CsLDHoAAAAAAAAAsHLu+n4flVXX0ebTOWRnp68uX1/fQPYyI19N/j6UJrzxN43opLc8NrNE/O3o62aW/WgvWG3IfV1dHb3wwgsUFRVFbm5uFB0dTa+++io1NDSGlPD7F198kUJDQ8VvpkyZQnFxcW263wAAAAAAAACgNGzMS8hMIvFeKhlnjjz5R345RM//dYyOpRXplvP7w6mF4v2ILv5m2Zf2gtUa9G+99RZ9+umn9PHHH9PJkyfF57fffps++ugj3W/488KFC2nx4sW0e/du8vDwoGnTplFlZWWb7jsAAAAAAAAAKEVtXT1JDvhBkb70+EXdaUrPIN332cXmsX+OpjYa8SuPZoi/L/1znC75aBvV1DVQj2AviugAD72SWG3I/Y4dO+jyyy+nmTNnis+dO3emn3/+mfbs2aPzzi9YsICef/558Tvm+++/p+DgYPrrr7/o+uuvb9P9BwAAAAAAAAAlSCusINa+c3G0p9/vHaULr5++YAudyiyh9KJK6hbspfrJlrzwzIqjGfTUtB60ZE+y+HxJv1B6ZnoM2cnzAUD79dCPGjWK1q9fT6dPnxafDx8+TNu2baOLL75YfD579ixlZmaKMHsJHx8fGj58OO3cubPZ9VZVVVFxcbHeCwAAAAAAAAAslWNpGpslOtBTL1c+1MdV/M0orDDLfuw9m697n5RXTlFzV1J1bb34/O6s/hTRwd0s+9GeMJuHnkPfW8rDDz983t88++yzwtiOiYkhBwcHkVP/+uuv0+zZs8X3bMwz7JGXw5+l74zx5ptv0ssvv9zifQUAAAAAAACAtoIjk9edzDKanx6qFaBLN4NBX1NXT7sS8nTCdxw1IOHl6kiuTg6q70N7xGwG/QcffKD3OScnh8rLy8nX11d8LiwsJHd3dwoKCmqRQf/rr7/STz/9REuWLKHevXvToUOH6NFHH6WwsDCaM2dOq/dz7ty59Pjjj+s+86RBREREq9cHAAAAAAAAAGrw+ZYz9MeBNBFWz0zoEaj3fZjWQ5+cX656A7CKPQvz+bg5iVD7R5ce0n0X6OWi+vbbK2Yz6DkEXoKN8EWLFtFXX31FPXr0EMtiY2PprrvuonvuuadF63vqqaeEl17Khe/bty8lJSUJDzsb9CEhIWJ5VlaWULmX4M8DBgxodr0uLi7iBQAAAAAAAABqGb9JeWXk5epEI6Nbp/oen11Cb6w8Jd67OtnTQ5O60dhuAXq/GRTpJ/6uP5VNlTV1innJOSrg2x2JFJddKkLqO3Vwp0h/TTg9C99d1j+MXvjrmE5dP9AT9pVNieJxubnff/9dZ8wz/J69+Ndcc40ubP5csHff3l5fAoBD7+vrNTkaXM6OjXrOs5cMePa2s9r9fffdp/gxAQAAAAAAAMD5SC0op2kLtug+731uygV7sNmgflNrzDPrn5hgtL47h+BL4e9rjmfS5QM6KtJAfx5Mo5f/PaG3TJpMiA7yEHn8D07qSm+u0uwjPPQ2JoqXkZFBtbVNayFyHjx70FvCpZdeKnLmV6xYQYmJifTnn3/S+++/T1deeaX4ntUTOQT/tddeo3/++YeOHj1Kt9xyiwjJv+KKKxQ/JgAAAAAAAAA4H/8cTtf7nJhXdsEnbeeZPOF1Z2YNDjdqzDNsWF8zOFy8/2JrAn20Po7yy6pNbiSpJJ2crXG5OmE+xtvNSfddADz0tmXQT548WYTWHzhwQLds//79wnMuV6U/F1xvnr35999/P/Xs2ZOefPJJsc5XX31V95unn36aHnroIbr77rtp6NChVFpaSqtXryZXV00uCQAAAAAAAACoxa/7UujNVSeFYJzEP4f0DXpJsK6qto7eX3uaDiQXnHe9B1May8NFnkc5XjLoWQn/vbWn6f6f9uu2m1F04WJ5FdV1OuN94Q0DaUxX/TD/nqHe4q+3a6NB39yEA7BSg/7rr78W4fBDhgzR5awPGzZMKNB/+eWXLVqHl5eXqDPPefMVFRV05swZ4Y13dnbW/Ya99K+88opQta+srKR169ZR9+7dVTwyAAAAAAAArJ+s4kqq48LmoNWw4fv070fos80J9PmWBF3uPAvYOTnY0UStgF16YaXwnPd4fjUtXB9Hs7/Yfd51ZxdX6t7fMrLzOX/LpeJGd23M09+VkC/2bcr7m2nSu5tFbv2FsC0+l6pq64WRfmm/UPrxzuF6IfWDO2ny9r3dGrO7O/rBoLepHPrAwEBauXKlqCF/6pQmr4LLz8HYBgAAAAAAwLywcff9zkQ6k1NKDvZ2QhF9e3yeMAK/v324WAYunN1nNSXcmHfWxNLyIxk6T/347kHULdiTNsbm0FurT+m3xzkM7KOpReTv6UyJeRrV+vlX9SUf90ZPeHNcOTBctKk8cqC8WrOdwymFNNyg3N25WHdCkyJ9Ua9g4UCV9uOO7/bRlJ7BOuE9uYc+DB562zLoJTp37iwEHaKjo8nRsU13BQAAAAAAgHZHcWUN3fndPtpzNr/Jd2wAbjiVLQw3cOGs1Rq+EiczinXvLx8QRkUVNc3+39q6enJ0sBfec2cHe5ELv+NMLs3+cjc1NBA5aidZOvl7tGhfLukXSr/uTaE9iZp2nvfPcd13B5JbbtDX1zfocvfZeJeY3DOYVjw8hsL9GsP/3Z0bFfURcm9jIfesUH/HHXeIuvNcQz45OVks53z3+fPnt8UuAQAAAAAA0O548tfDesb8raM60/9mxOhEzE6kNxqhoGWwMb548xn6abfGxnnxkl7k7KgxuyI6uNFNIyLp4j4hNLVXMHm5GHdq5pRWCTX8Qa+upcs/2U7Lj6TTE78eFsa82EZ9g4igGBbVoUX7xF7zX+8dSTuenUQxIV563+1PajqZ0xyxWSWUW1pFHs4OTbbdO8xH1KCXCPJu1C0L8GxMiwbK0iZu8blz59Lhw4dp06ZNNH36dN1yFsR76aWXRH15AAAAAAAAgHrsTsij/7Re5I9vHEjTe4cIr7AE1ziPyy5BE1wgC9bF0ccb48V7TxdHMUlyxcCO5ObkQG4yrzUbvJ/dMph2J+RTn44+dPcP+3QGOwvWncwoEWHxR9OK6MElB/W28dS0HnTv+OgLTofg0PdVj4wVKvk/7UmmFUcyaH9SgYialsLn5ZRU1oiSdzEhGqG7Q1oxvn7hvrpJiuZg4/6/x8aRq6OD0XUDKzbo//rrL1q6dCmNGDFCr3HZW8/idgAAAAAAAICWw6HZv+9PpRMZxRSfXUoJOWUUFeBOP905Qs/wOp5eRIs3J9DT03rQK8s1dcRnD4+kS/qF6a2va5Cm9FhcVima4QLYFperM+aZO8ZEiXD5Dh7GPdSjogPEi0l4YwbNWryT9iUViND6yppGZXw5Z96YYZKuAdtfo7oG0JDOHUQ+fEF5DZ3NLaMu2nJzcu75YT/tOJNH783qT+tOZgkPPTMg0rdF2+oerB8NAGzEoM/JyaGgoKAmy8vKyjB7AwAAAAAAwAXy464kem3FSb1lHBrN4fRjujWWFZu5cJv4+6+2FjqHTj86pWkVKMkjG59TKry0XjKBM2CcnJIqenTpIfF+bLcAUc5tzqhzK9AbGtr3jI+mzH+PU2pBYzm5L28ZQsfTi+mDdacpKsBDMZFCnuhhg5sjAOKyS5sY9FxGj4155onfDut9NyCiZQY9sNEcei5Xt2LFCt1nyUvPJetGjhzZFrsEAAAAAACA1SKFQk+OCaIPruuvqwX+34lM3W/ySqua/L/nZvbSKzkmD83u5O8uStft1QqptbeyfTd/tZv+PpRm9Pt/DqfTI78cFBUCJL7YmiAmUboHe9LnNw8Rxrmk+N5SWIBw69MT6bkZPcXn/uE+NK57ID04qSu9ekUf+vrWoaQkXQI1onpc4SApr4xOZRbrjPmpH2xp9v8NhEHfvj30b7zxBl188cV04sQJqq2tpQ8//FC837FjB23evLktdgkAAAAAAACrhUPtmVtGdabx3QPJz92Zbv1mLy3bn0oRfu704fo4Kq2q1fs/7J2/fmhEs+scFe1PSXmaEnaTYtqX0j2fO1al3xqXS5cP6CgM3O92JIowdRaVe/hnTU47f755RCfx/mBygfh7z7hovVz5C4WdnXeN60JjuwdQdKAnOWl1DaTtKEmXAI1XntM0xr+zSbw//OJUenPVSdH2xvB2ddQTvAPt0EM/ZswYOnTokDDm+/btS//9958Iwd+5cycNHjy4LXYJAAAAAAAAq6O8upb2JeaLHGimZ6gmZ5mNejY8y6rr6PWVJ5sY80z3EC+R390cUm63FHZty+xKyBNed1Z8n/P1Hr0Sc8wzvx8RIoGP/nKIvtuRpFv+y55koWrP5dw4LJ7pG+6jyD5x2oNkzKuFpJXw3/HGEnuvrjhBv+xNIQ6ifn5mT6HE37ejjyifx7DAH7Ac2qz4O9ee/+KLL9pq8wAAAAAAAFg1LGg298+jInebGda5AwV5ueq8vA9M7EoPaT3JxhjbtTG33hgjtLXJ2bjNL6tuVtjNWmFD/NNNZ+hwapEQfDMGK9Nz2sHfWs2B5Pxyemv1Kd33bMT/fSid+oX7CEV6Vyd76hLQstrwlsDEmECKDvSgMzmaCSGGxRWZZ6fH0J1ju9D1wyJF3XtWxj+YUkgPTIxuwz0GFuGh5/J03377LRUXo64lAAAAAAAArRHBu/P7fTpjnnlyWg+938zoG9rk/7FDfv/zU2j5Q2PosYuaiuHJ4dz6HlqVcvZg2xKcM/72mlh6b+3pJsY8RzasfHiseF9RU0eHUwt15eQk2HCXOJBcQLvPanQGBkb46ZX+s3TcnR3ps5sHi/QLOZy3f/e4LrrSe6wFMDEmiB6/qDu5OLY+nQAoT5tcbVyejmvRh4SE0KxZs+jvv/+mmpqattgVAAAAAAAArIqiihp6a5XGS3z1oHAhhPfM9BgaFtVB73eshv73A6PphmERtO/5KfTsxTG06pFx5O/pIuqet6Q2+KiuGi/9jjO5ZAse+R92JtIlH22lye9tps+3JIjlM/uG0htX9tX9bmjnDiJ1QTLab/lqj/grlf/jv3weOZec4dMoGfTDu+i3gTXQNciLPp49SG8ZlzJE7XjroE1C7lkE74MPPqB169bRkiVL6JZbbiEHBwe65ppraPbs2TR+/Pi22C0AAAAAAAAsnuVH0qmkqlbkP79zTb9z5sH3j/AVL+be8RceKj0o0o++2Z5IJzM09cetAVaaP5RcSJN7BumM0t0JeXTd57t0v3FysCNHe3vx97mZPYWq///+PKpLNeD/J9WBZ/0Bd2cH+vSmweRkb0ceLo6ifByX+3tl+QlRx511DJjhUZoJEGtjYo8gumZwuC7cnj8D66DNcujt7e1p6tSp4rV48WL6999/6fXXX6evvvqK6uoayz8AAAAAAAAANHy/M5Fe/Pu4eH/tkPBzGvNK0C1YI5p2OquEGhoaLN5ryykIVy7aLuq4fzVnCE3uqVHn//NgY/m5p6b1oBuGRZKHi4MIpZdKy61+dCwdTS2iGX1D9NYZ4u1Kqx4ZS34GGgK+7k7iL/+frOIqIRo3MNJ667PfPjqKVh/LpEcmd9NFIwDLp81bKjMzUxj0b731Fh05coSGDlW2tiIAAAAAAAC2Uhv91eUnxHsO92axMrVhTzRTUllLj/xyiGrq6unLrQmiXjkb+JZEZU0d3fX9PmHMM3u0XvPUgnKh2i5FKbBYIAv8cS64vE48q8rPGhKhm7RYeMNAmtAjkP5+cHQTY15u0LNQHsPG/IXWnbckeoV509GXpoqSecB6aBMPPYvhLVu2TITbb9q0ibp06SJC7ZcuXSrU7wEAAAAAAAD6fLsjkWrqNEb093cMJ29XjUGpJnIBNC7rNiDCl15bcVJ8ntkvlD65UT/3ui35attZOpRSqPt8LK1I7PP//tCE0jOX9GsqFNgcl/UPE6/m8HHTN/JvHdWZrB1Lj8AAFmLQBwcHk5+fH1133XX05ptv0pAhQ9piNwAAAAAAALBIJO83G1gs5vbiP8dpye5ksezzmwcLw9pcfHTDQFH+jpXQ/z7UGLq+4kgGfXIjWcz5kvK/54zsRN/tTKLt8XniJSc6UJNCoASSh16zXg+a1ls/VB8AmzToubMtXLhQeOTd3d3NvXkAAAAAAAAsgoKyanr+72Pi/UU9g2lIZz8K93OnhJxSuvrTHUJtffFNg2ndyWydMR/k5UJTtHnh5mJktEborayaS7gV6X3HIfhOFlCm7WxumXhx7jeX4/t5TwpV12lE7a4a2JH8PZ0pxMeN3AzKs5mCr1ujQX/PuGjV9QwAsBiD/oEHHqAJEyZQt27dzL15AAAAAAAA2lyFfdHGM/Tb/hSRmy55uxlWro/PLhXv/zuRJQz+rKJK3f99/pJeZjcc5YarsUmJIG9XagtYWZ695F0CPOlmbWm5/uE+5OvuLAx7yaC/d0I0dQ/2Unz7nIfP9dqra+voykEdFV8/ABZp0LO6PRvyeXl5MOgBAAAAAIDVciGq7xtPZYsc+MOphVRYXtPs7yRjXkLyzDNcU14qQWdOHB3sycvFUZTKMySvjQz6n/ck09w/jlKglwt9cO0ASivUCOEN7qSpA3/1oI4i7J4NfDWMeYbb/vvbh6mybgAsOod+/vz59NRTT9Gnn35Kffr0aYtdAAAAAAAAoFXU1zfQ7C93U3Fljcgv73KevOyyKlaIP0jFWm884+/hTBf1ChYl1HzcnEQ4e1JeGR1OKRSG/62jo6iDuzM9uvSgEMJjBfW2MOYljBnzTH5ZNbUF/xxK15Wp+/ew5r1Uyo95eHI3ig7ypCsHwnMObJs2MehvueUWKi8vp/79+5OzszO5ubnpfZ+frykxAQAAAAAAgKWRkFtKOxM0YmtXfLKdFt88mEZFB5xTfZ2N+U7+7vTxDYMo3M9NhIrLvfs+bvbUL9xXvG4e2aiWHh3kQTvi84Txbyn8du9ImrV4p85D3xZU1tbp3rOSPfP8zJ66yRV/Txe6RXYeAbBV2sSgX7BgQVtsFgAAAAAAgFZ55Dedzha52YMi/Wh/UoHuOzbU7/lhP618eCxFdHCn1ccy6NNNZ4Sg3eNTu5O9nR19sjFe/PaxKd2pb7jPBW2ba6Pzy1Lo6Osmjm1G3xBaeTST8kqr2mQ/OOpBoqKmjlhWYHTX5idVALBV2sSgnzNnTltsFgAAAACgTY1CdsiizrP18faaWFq8+Yx4f/3QCDqaplF6v2NMFB1MLqADyYX04JIDdM2QCHpt+Qmqqq0XavCbT+fQ/ROjxWfO9b58QPM1zS2dFy/pRe+vPU2LZmvqzkcFeIi/8rrv5qSsSuOhD/B0phuGRYpz2zVInVx5ACyZNjHomTNnztA333wj/n744YcUFBREq1atosjISOrdu3db7RYAAAAAgOKkF1bQTV/tpmAvV1py13AY9RYscnc8vZi6BXuSi6MD1dU30LL9qTpjnvllb4ruPednsyF5+SfbhQEvlXTjnHgu5RaXXUqPLT0slg3t7GfV7X77mCi6dVRnncL++O5B9MnGM7TldI44Tw5mVt4vq9Z46H+5ewQMedCuaZOikZs3b6a+ffvS7t276Y8//qDSUo2a5+HDh2nevHltsUsAAABsnAXrTotc16JzqEsDoAYsnMb5xgk5ZSLvmr21wDL5YO1puuSjbfTxhngRUfHUb4fp6WVHxHfB3i7CeGTD3NXJnmb2C6U+HX1EzvvXc4aSt6ujCPt+ZHI32v/8FPr3odGiBJ3EhB5BZO3Iy+UNivQlNycHKiivoYQcfWV+peAa9ywUeK6Qew+XNvNPAmARtEkPePbZZ+m1116jxx9/nLy8GkNjJk2aRB9//HFb7BIAAAAb5lhaES1YFyfeb4jNoisHalSQAVCT6tp6USebBdGkklpMaVUtuTo54ORbGEdSC+mTTRpP/Pb4XCqqqKE/DqYJz/OsweF0zeBwGtK5A/1276gm5eomxgTR+icmUEV1HUX6u4tloT5u9Ns9I+mV5SeEGB6vw5bgUnZ9OnrT3sQCkZJQWVNH713bn4K8lCth9+bKU/T19rP08KSu9PjUHiJSgl+sY8DK/4y7Mwx60L5pkx5w9OhRWrJkSZPlHHafm5vbFrsEAADAhll2IFX3nr2khvDg/I8DaSIX9Imp3UW4LNNceCz/nj14bs4OdOfYLiruObD2qJBFWgPR0d6Oaus1BkhJZS0FeLq08d4BwyiKJ387LELHGc6J5xffAt6/tj9dPkC/9JmxewPnyBvi5+FMH1w3wGZPNkcnsEG/9kSW+PzWqlhh1JtKdkklvfDXMVpzXLPehRviKSG3jJYfyRCfN5zK1v3WwxmTY6B90yYGva+vL2VkZFBUVJTe8oMHD1LHjqgVCQAAQFnWndQMCpnYzJIm33+0IV6IPUneU/Y07UsqoL8fGE1hvm5NBpqXfbSdMosrxWeuC82KzwAYhgNLxjwzZ1RnWnEkQ1w3pbJa5KBt4PsAh4lP7xNCpzJL6NrPdmonWpyF2BqrpjPXDApvYsyDRvoZKPZnae+LprJsf5rOmJeQjPmeod50NreUKmvqdZECALRn2qQHXH/99fTMM89QZmammOGsr6+n7du305NPPilq1AMAAABKGlYp+Y3hzpzDzAJlVbV1IjeTjXe54NWfB9No1bFMyimpopVHNQNICf7tbd/s1RnzzM97ktFYQEd5dS39sieZJr67Sbdsaq9gempaD/Jy1fhRSiqtX8fht30p9PxfR8Wx5pupDnlz541V5h/46QDtTcxv0Xo4EufSj7bRfT8doKs+3UEXf7hVGPPMV3OG0twZMRQT4kWRHdzp3gnRih6DrdE/3Ffvs1Kaf/JSeFwer7M2jYGZe3EM9QiGmj0Abeqhf+ONN+iBBx6giIgIqquro169eom/N954Iz3//PNtsUsAAABslIwijTHv7uxA3YK96HBKIT3yy0FRG3r32XzycnGk8uo6IXh197hooWh9IqNY/B/pr8Szy44IBWw5e862zIgA1sHuhDz661A6PTezJ3leoNgWG7aXf7JNN4HEBuGbV/XV1cb2lAx6Wf1sa9UGmPvHUV0KwdJ9KfTn/aNV2x6nuDz31zFasjuZRnTpIDzm3E8Zzmt/9o+j4v2Koxl0eN5UXcqMMXhS7uYvd1N1nca7ezC5seTaN7cOFRE3/LplZGfVjseWYG0AF0d7ndAj31eVIFdr0M8Z2YnmXdqbbv56NyXmlYtlER3cjaY3ANBeaRMPvbOzM33xxReUkJBAy5cvpx9//JFOnTpFP/zwAzk4tDwPJi0tjW666Sby9/cnNzc3oZy/b98+vQfAiy++SKGhoeL7KVOmUFycRhQJAABA+yCtsFJnXH143QCRb8k5n2zMy42rKwZ0FDWlVz4ylr6+dYhYti0uV4Tgsyd/f1KBMPRY5Pnb24bSV3OGiPepBRX018G0NjxCoBTc1td9vktEXSzZnXRBk0bf70ykm77cLYx5NnCem9GT1j4+TmfMM16uGkPT2kPu+TxJxryTg50wio9qy7WpAUfVsDHP7ErIF5MJnBLDL8mYl1i4Pk5E4DTH8fQiXZ+/bXRnenhyN7q0fxhdNySCxnZrbCvQMjjSVp6WxKr0SpBdojHoB0b6CWV9N6fGybVQH1dRXQAAoKFNZSHZQy956Vkor6CggPz8/Fr0f/m3o0ePpokTJ4r69YGBgcJYl///t99+mxYuXEjfffedyNd/4YUXaNq0aXTixAlydVVOgRMAAIBlkVlUScn55TQw0pcytIN7HnR2DvCg+Vf3o0eXHqL6hgZ6+bLeuvz37rIQzpFdAqijr5tQJh/z1gYqlJW6YwNNKj81e3gn+mFXEj326yFhJNw8olOr95knodnLBfXztmO9TGvhVEZTrQXm38Pp9PaaU/TKZX2Esjmnblz5yQ5dGgZ79bm0mTGDg6NBbCHkXioXxqXbpvYKoX+05+S724bplTUzBS4v6e7iIGq5/671xrPBzaHWnBaTpw3zZ4dwg2ZuQcAVBb7dkSjaYHAk133XGJ3smf9meyKtPZEpfndRr2Dh+QWmE+brSmdzNWKjuxLyRAoEG+KmIOXiB3m76O6PEnyPvHd8tLjH8/UHQHunTQz6Rx99VHjT77jjDmHMjx8/nnbs2EHu7u7CYz9hwoTzruOtt94SkwHffPONbplcZI87/oIFC0QI/+WXXy6Wff/99xQcHEx//fWXyOMHAABgO9TW1dO2+FwRjnnNpztEbWQO15Q8PezVYdgbx4Y+C1/1CDGeh8nq9e9c049u/HK3njHPnlceSErwhAAbDN/vTBKKzB3cnUVt6pbAhuA7q2Mpo6hShGrHZpWI/OsVD4+l6MDG2tXAfOxLLNC9P5xaKIxA+QTLifRieujng+L9bd/uFXm9fJ1xebMOHs50y8hOIhw8KsDD6PqlHHr2cFsz0v7z5MVjF3Wn1ccyaWtcrohSuHV0lCjdtmRPMl0xIIz8W6HmzxNprEHg7+FMlw0Io1VHNUb4o1O60+BOfvT8Jb10v2VhuxkLt4oJuRl9Q0Uoflx2qUiP4b7FfbZ3mI/oW6xaLwFvvHLwed8enyfec+DGlYt20O7/TaZg7wtznm08lS2updTCcl14vVQCj/uXHO6X719ru9UDALB4g/73338XofLMv//+K0LvpZD75557TgjknY9//vlHeNtnzZpFmzdvFur4999/P911113i+7NnzwrRPQ6zl/Dx8aHhw4fTzp07mzXoq6qqxEuiuFg/VxIAAIBl8uCSg7T6uGbgL/Hdzsaw6ck9NV51JtyvUWCpOUZ1DaAvbxlC+eXVNCLKXwwy+a/cA8nv2ajnvFH2Cn60IU4IODVX7k4+6cz1lfn/GLLjTB4MejPBooccXn/1oHC6YmBH2n1WY5QwZ3LKaOoHW2jhDQNpQISvyOm9/du9ev9fMjqY+8ZH013jzl3CUMrJlwTYsosrydXZgby1ofjW5qH3cHEUkxcs+Pf6ypP0854UYdC/vzaWvth6VgjnsaecS0JeOzRCXPf3/rifAj1dRCk3eT9h1flQX1dxLnadyRN5+myQf7Y5QXzfJcCDBkXqC7CJ5YGetPHJCeLcckrDsKgOdNWiHaL9GNbH4Ik+qXQgT0D0CvWGQa8gNwyNFGkOn2xsFBflygEXYtDX1zeISCf5BGpEBzfxYh6f2p1OZhbTTcNbHwUFgK3SJgY915oPCdGEyKxcuZKuvfZa6t69O91+++304YcftmgdPAnw6aef0uOPP07/+9//aO/evfTwww+L/Pw5c+YIY55hj7wc/ix9Z4w333yTXn75ZZOODwDQ/mBhNC6N5uHsSKO7+gsv0vmMOmAcrgPNJaPY68lh7y2BQ5gNjXkJHuhzXeRJMfrPg5YwpVfj/4mUqSzL4XZ+bEp3Wro3RQxiWUHb0d6e5q8+SXMv7qkLvU7JL6ebvtpN1w6JoHA/N50xzyHE90+MpnfWxIp8/DPZpRe8n+DCYY/tY0sPiTQH9i5zpYPTWaXk7GBP86/uS2+vjhUhvXd+t49WPzqWHv3lkAir7xLoQX89MFp473/clUxrjmXSHWOjaNbg8PNuU/JWf7YlgVYeyxD59kFeLrRBa5BKsHHE3nwp595SPfQezpp95muar1+OMjmWVqQrL8b94YYvdtPJjGLx4lrvkidXCA/O6CkmQThEm7267JH//JYh4rcSt4+OEukxVw3q2Ow9NdTHTU91nVMBpJJmU3oG68pW8sTDPbIIG6AMPLHJkUtyg57v3xfC6ewSnTH/xpV9qXeYt4igcnF00LXx8ofGoskAsBSDno1qzmNnsbrVq1cLw5wpLy9vsSgel7obMmSIUMxnBg4cSMeOHaPFixcLg761zJ07V0wSyD30HNoPAADNGQUfrosTA3SJD9Zp/r5+ZR+RYw1aDgsqzf5iN+3Rlp/67vZhNL574Hn/HxsOUlg9e1QPJBWIMFA2rgd18mvxxEBr8XF3omm9g4WR8tI/x+lIWpHI673ko210/dAIevny3sLgScorF3+5JJaUU/37fSOF4cYG4jPLjtKZHBj05oAFDyVlbvk1xBEWVw0Kp8k9g+nqT3dQfHYpDXlN06ndnBxo8U2DhReZX49f1F28WkpfWV69pITPKSErj2QID7bkqb70423it8vuG0WWCKerMNIkBF//U3sHC0P+l73J5OjQaHhLxvlv2jx4OZwLzwb9fyc0Bjfnxd/81W7hVWfmX9WXrh8WeUH75mBvR7V1jfnWL1zSkw6lFIiJmptH4n6sFh7ayR0Jjj65kHSp6Qu2ivfDozrQjcMvrM0BaO+0icr9bbfdJrzyffr0EbOtUlj87t27KSYmpkXr4MkALncnp2fPnpScrFFBlSIAsrIaBW6kz9J3xnBxcSFvb2+9FwAANMery0/ojHn2tHprc2SZvw+mK37ieJDEA35b5eMN8Tpjnpnz9R5KloU1NyeA99GGePG+Z6i3ELljLxzn0nO+vNrGvMTtY6KE4vfhVI0xL/HL3hThCZaXt2Pjkb2I654Yr/PCckk95lByIWWXtHwwDFoOi6uxJ54njr7celbTbqOjaPNTE0TVgg+u608vX95HLOfSZ4bGOpegk4snXij9IvSF8u4aq9H+eXtNLH23I1Hs1xdbE0S4OVdVOJVZbNEh9yxaJ3GD1vDmqAVpsoKNa0Mu6x+me5+YV0b/+/Mofbqp0bMrGfPM0CiNYOWFIu0Ll7jr5O9B/z02XlSvcDcwOoFyGIohsg5CS5HSI5irWxDpAgCwAIP+pZdeoi+//JLuvvtukS/PRjTD3vlnn322RetghfvY2Fi9ZadPn6ZOnTrpBPLYcF+/fr2et50nDUaOHKno8QAA2idxWSUiZ1Tip7uGi1zOcVqPcm5Zox6HqSTmlglv4bA31tO0BVto55nGXF9bCbN/5vcj9OF6TWlRyYPNPLPsiJ7CsRwWUJr6wWbacjpHfB4jKxFmbvqF+9IPdwwX+diGub4rj2bqVNCZyTFBtPTukXo5phwqzBMSrJZ/+cfbRT10oBysmv3U74dp/qpTNP7tjWLiiCdV7hwbJYw+9shfOTBcr4b59N4hos45G6acVsF59qbAXv1uQRrBQ84NZpE3Ftbj/Px5/xynz7ckiKgSCRZN5NBlTtewJMqqG3PoJUZ28RelISW4H/BEibwvsxjlh9cPoNOvXSxKPrLxLpWj4/Ow97kpNEWrdcGida0Vh+SJmJcu7UVfzhmqE1TzddcXVQPqYqx0IKdjcNWQPw+mijKH1y7eSUdSC3VRHJyqxukbAIALw66huVGShcM586NGjRL57uzt37NnjxDE+/zzz2n27Nk6Jfz58+frla07cuTIBZWt40kAFtMrKiqCtx4A0HhvqKwRRheX6hnWuQN9fstg3YAxPruEpry/RXjrj7w0TZGz9tbqU3peLOaFS3qJuunWBKtfP/LLQeG9eWhSV5reJ1SIIb3w9zH6STuwn9k3lD6ZPYj+OJBKj/96WCz75rahNFFbKk5i7Yksuuv7feJ9J393mjOyM80Z1dmoV7Ct0jFqahtoc1wOPfzzQbFfP981QoRSs4q+Mfjauev7/eK6YgGvfx8aI4x8oNFJ4EoCLFZ4IYN+niziMHD2znO9dDkPT+pKj0/t0aK2VMq7y9EXbKhz9AhP/uWUVNETvx0Wk1J8z+A0AHkqgMTXtw5plQ6EKWXjNsZm0x8H08SE4oOTugqdAN7f+386IGrAc+32t67pp/s/7/8XSwu10TL7np9CAVrNgHt/2E8ZxZW0aPYgXcTM2Lc3CE++s6M9PTM9hm4cFin6BUcpcPnAIZ076P4/sA46P7tC957vc3z/knPpR9voaFqR3jL+3aiu/kL8kMt+vnqFJkIGAEAttkPbLPaIPecffPABnTx5Uhcuz+Xs5Kr052Lo0KH0559/ipz3V155RRjsXKZOMuaZp59+msrKykQkQGFhIY0ZM0bk7KMGPQDAFNhzyiWr2LsU5uNKn940SM/7Iw1CiytrRegsD1hbWsaMDZaTGSXC2y9Xvj6l9WCwOvMJ7XsO9+d8Q2P1ri0RHqg/sOQAbTiVLT7f++MB+uGOYVqlcU2kA58ryUDgPGYOTWev5fyVp2hct0Dh1eNyWNvjc4WAmeZ3HWn+Vf1afJ7NhTAAnTUhxhF+bqKe9vnaqmuQFy1/aIxQAmehtk82xtPHNw4y2z5bMi//e0JoFPDrQgx6FkR77s9jTZYHerm0WCBNyVBtLsP1zqz+evvxza1DafqCLaLcmgRHBHyw7rTuM4vy3a1V0WexxpgQdSZ6+J7FEwzLj6TrpY48/fsRKq2sFVUBpP2Ue+gZVrjfEJtNwzr76xnji28e3GQ7k2OC6dd9KbTw+oF64pPcT3iiD9ieh54nwo1db/zMYzB5CUDraBODftGiRfTII4/QNddcI/4yu3btohkzZggj/4EHHmjRei655BLxag7Oz2djn18AAGAKG05lCQOLBylrjmu0OVhk6bObhzSps8yGOHtXa+sbKK+sSk+B+VyeazbiNmtDx1mwi0XdJCTjdfaISD3j5N8j6VZj0H+w9rQw5lmoms8RhxLf/NUevd+sfHiMntr3AxO6CvV4Vs9msTkeEP59qFGbgM/zs9NjLM6YN2RgpF+Lf8tGEpdA4+uN654zLJjHBl3fcB+6f0J0u6qgwIGErEvAHna5l5sNYw5F5xB11kvg0nLsSecKA/LrgSd/GA7l/t+MnsJgZCN/VHRAE4O0rRAh/Rd1F55vychnj/jBlALan1gg0jB4gvDd/zQGPnvNNzwxQbHtc8j/GytPanRA3Jzo38OaPtY1yJOGdPITOhDMK8tP6P0/D1kOvRTa3lIl8pcu601zZ8ToVMyBbcECh3zf4nrx4nNplRAFNcTT1VEXch8T2np9CgDaM23yJGNlejbcH3zwQd0yLjnHefH8XUsNegAAMFeJJvYmsyfBUCCLDSxj4kA8sGX16rzS6hYZ9AvWndYZ88zhlEKdQc+ib5LA0CV9w4TBy0bMN9sT6Zc9KfTQpG56RrClInnm37q6n8iVvPKT7cJI4dxZDnu+pG9oE2ElVs+ed2kvEXr/w67GmvISfI6CLqDWsbUQqg1LTsgtEznGLo72oiwfv7ge98WyyR5bhsXXxr+zSRiccniioy6kgSa/t5mq6+p1FQ44HJzLC3532zCdirqUynHN4AhRs5y5bbTlpapwvj57KNm4CfN1E0b+t9rjOJ2liVRhNXBWkk/IKaMvtybQnWPPXfdegisn3PrNHooK8KQnp3YXeg/yyZFZi3c2MbYenNiVnpymSUd4eHI3mvDOJnGu2ej3dXcSJRYn9Dh/BYpzAWPettmVkCeqlPAEJEd4GIPz6jnFhOco+doCAFw4bTIC5PD36dOnN1k+depUeuaZZ9pilwAAQA9WmF5xJEMI9rB3mI15rhc9a0i4CL+9dVTnc3r32MPGBj0b4ufzoLOwm6SUzwPkTbE5Ij9VYtkBjWdyaGc/YeBePqAjXdovTOTcsjrwj7uSRA1gS4bzmNk4lcSz2KPKtbe5zne/jj7k6NC8h/3KgR11ufTMfROidXoCnDNviwR7N0Z9sAo4RyJIbIvPFefriy0JoiTfsxc3Xx2GDWGu7X0hHn2+5jmfdUtcjgj7ZvV+tT3wbCgaM+4OJBfojHk+DvbCrz+VTTsT8qigvFpnzDMZRRrRQTZ2x769UW89UQEeQmTNkuHJrOdn9qTbvtlL4wz2lZX139WG6bO44q6EfHptxUlxL+B7zfn4aH2cyFfnF983vr99mE68kytLsDHP6+EJEQl5WgNPMPx670gRScTK8e0pQgS0nlu/2SsEQj+/ZYjot3L4llbfwKlmmj7cqYO7xUTMAGBttEmM4mWXXSby3w35+++/zxlCDwAA5iC/rJpu+GIXfb39rDCsSyo1is4sYvXUtBh6YGLX8w48+kdoPGA7tOG+xuBwRM4fZ8E0hsfIz8/sJf7yhAILo7Gx89u+lCYDbB783zehq3jP5bd4oG1O2Oib8eFW8eozbw09vvTQOX+fWlAuJkXY08zGAcM5toMi/c5pzDNsPLAxw6rgy+4bKc4/hwJz3XdDNXlbgSeNvGQlEDl9Q4I9ziwGyCrtXILtm+2a8muGsGgg109ng62lcDuxt3bF0Qxx3fNkk9q8vuIk9Z33H605rtkWG/C/7k0R1/SxtMaybSyMeOUgjcr8t9sTxYu5pF8oXT80gv43I0YIRUqwh5sF2J6b0VOIc1mDsTC6awAdnjeVnjiHUJ/8GP/RhsafayKNSz+y7gAjKfhzfrzknZfC6Vl9fsF1A8Q5e/3KPiLSQQ5PpoyM9ocxD5qFJ50NOZBcKLzzXNJTDqe/yEH+PACtx2xPt4ULF+rec/34119/nTZt2qQrIcc59FzC7oknnjDXLgEAgFE+3RQvjE/26nEuc1x2CTnY29O941sW3sqwIjuHSm+MzaGXGhqaDILZCOOccg45l5h/VV9hqHLpNc6fXnsiU4TsJ+aVk4ezA83spx9mffmAMLEOjgKY9dkO+uDaAbpa5hcKh/LmllZTiE/LwtcXbTyjE+dj/jyURi9e2qvZ0lCSBgCf09ao0HPpMH5JrHt8PNk6XNKupFJz3iRNBgnWb5C80ywWx4Ygn1tOxegX7iO83WzsM++tPU13jeuiy2VtDq428NbqWOEx45DqwvIayippLLWnBvuT8unLbZoJiSd/PUzdH/IS6Sesk7DyWAZ5aMXoOAqBw8S5jFmIt6vwUnNevZR2IaWnaOrLJ4hJubWPjW9ilFoDzVVAkOgd5iM8+eyh3xqXY7TSBWsLsEYFp/FIqTzXDgkXKvms1XE8vVicq//9cVTc67icHEfO8H3K1NJ8oP3y6U2D6a+DabQ3MV+nNSOlW3FJQ47IYuaM7CSiq/galvD3RFlBACzeoOeceTl+fn6ifBy/JHx9fenrr7+m559/3ly7BQBoRxxMLqCvtyfS7OGRNKKLf7P5fF9s1RgYbMxfO7R1NXFHRfsLo4sHMBx2yHngUv4uGzFshDFsnLDX8c4xUTpxPa5Hzgb9GytP6dbHuayGStss7vXalX1EiC57Mr/ZkUhvXNn3gveVDbln/zgqRMc+v3kwTe0dcs7fc2TBptPZuvD3VUczxKTDfT8eEGrWy/aniugGTlFgI4zL0+1LytcdG2gZgZ4uFK9VE//z/tF0PL1IeGTt7ezos5sHU+95a/SE3zbH5tDrK08Kg37ZfaPIXWYYcpj1+dqV810lMbS7xnahd9bEitrtHCWiRIg1H8t7/8XSxJggumZQuMjB5muXYRE7Fn6b+O4m3e859USiT5gmbYW97H89MFqkoXB+PWtVTJMdF/cJPlcVNXVWacy3FPaUM/sSC4QXXpok41z55Ycz9NTxmYk9Aun1K/sKPQ6G8/Q5yoPPMUfNaCKDEEYPTIOjrljXgXVn5Aa9l4ujmECSjPmXL9eUpvv1npF07Wc7xfuhnTvg9ANg6Qb92bNNQwJzczWhqAEBlp3XBgCwTHjQwANTVmE+32D0lz3JotZ5TV2D8HxzaOnUXiFNRNj+O6EZhLCXnPPlWwsbHsO7dBCGOXvI5nyzh7Y+PUl8Jw10eJD95ZyhTTzW3UMavexssPBkAOfKNhcJ8NS0HsL4SsrT5KhfKLd+u1cYfMzdP+wXnjw26Jrz9sdmllBlTb3IaX56Wg/h4ftq21mR19z/5f90v2ODjcMtOa2A9QSYYVEYtLWU5y/pKXLZ+Tpi8UV+XT8s0uhv15/M1oVTH0ktEp/ZSyvBofNs0HOfYbXpTv4eTdYhlSJjYSr2+vI1xe3c44XV9Ns9I8VEAZcZyy6uoltGdhZ6Di3h3TWxIiKlTJsWsupYphC14/3lCBWOPll2/yia/cVuoYxtjN5hjSXaOIqE0y6ao6VRJtYMl6zjlAxOi+BzydcGV8a478f9YmJEgvPiebKQjSy+z4T7uVGXQA+hM8DGPEd+cNlNFqkEQCnkmh8MR/1Ua/v2HWMaI934ebDu8XF0MLlQlPcEAFhJDj0L4rGKPRvxwcHB4sXvWfGevwMAgJbA5eMu+3ibyPdduP7cOcIchsweaDbm2YPARgqr1vd5aY2o/fzM70eooKxaGN4crisJsZnqsZrQI0j3nsWoJNaf1Bj0Vw8ONxp+3j1Y48lnltw1vFljXkLybEjhjBfqnZeMeYlf96XSm6saowMMkUoMcc4jn6MnpnYXecqGsOp4dKCHzpjn0znGwoXJLAkOrV54w0CaPbzTeX/LkylH0zQh6MxrK06IqAmJvw+ni4mYe3/YL1TjXzUoP8YkaieExvcI1AvP5wkbNuS3x+fRM8uOihD+z7dqwvlbwl+H0nTGvMSSPclCJ0JSUGcD9ZPZg0TECqti//fYOJF+IuHngXBcOXzfkPr97rN5IuqHVezZmGetCY6OefXy3rTtmYl0z/ho3X2G++tHNwwU90Fe9N61/UUYPgBKwqkx8muV04Mk8TvDicCuQV40a0gEIkQAMAGzKsTk5+eLnPm0tDSaPXs29eypGQBy2P23335L69evpx07dohwfABA+0IK62WDmo2Hj28cJJSdjYV7s9jX/FWndKXcPtkUT/eM188R5vVJZePYe8zM6BtC780aIHJ02aBg79apzBLxWqoVnmNYDZsF8EyFPfCvLm/8zDmtP+1OEsr0PMgZ2814ySf2kN41Noo6eLgIQ+d8dNKGFqcXVorc4Y2nskW+/flyphnpHDLsKWHPLhvzxiYHuAb8V1vP0ofr48TnGG0kAacCcI42C2yx0ffK5b3Fb8d0DaTa+np65d8T4vw+PqW7yAsHysNh1/I24+gIhg3k/hE+Iirk9m/36tqb+wRHYUjebFY35zKIDOe6StEhfD0x7NF9Z03jJA+381PTjGsxcFQGe/P5+uNrUdoXnpwYGOErokB4UojV+plu2gksToPZOXeSbmD/za1D6cGfD9IVA+C5MwZ7Nzk3mfOQeXKNNRb4nsNpL+cqB8cTRVueniA0EiSBSgCUhNNgOHKM+/vHG+Npx5k83XfeMrFPAIAymLVXvfLKK+Ts7ExnzpwRnnnD77hsHf81zLcHANgu5dW1IrT35z3J5GBnp/PkPbb0EM29uKfwNrEByd9xnufUBZspq1jj8WURMFaCZw8iGwgDIzWTgfsS8+mhnw/qylhJcO1pFpyaO6OnUJFmRW8OM37pn+O637LQ3DvX9Bc5vabCOfMs5sWTD4w8HF2UoNOGSBvCBs1zMxuVrFuSa+3qZC8iDy5ZuJXSiypFaHVLSrpJXl2eRGBPCZGdMOg5fJ89qKwiz/vDEyQ3fblbhHNLTIppjEBgjO2zM9nT/Kv7CcE8Qw0AYBpcOvHbHYniOuLJIomVD4+le37cJ6JCbhoRKUTOOLxaPnnDLN2bInKxY0K96InfDjeZIHr76n707n+xYsKLIwAMrxuujc7blae8PLL0kCj3yErpHMYtV2HnOuvcr96d1Y/u/G6frs/JvXnyqBgubfj3A6NxmTTDcFn6ihQ1c8uozi2q7c59Ef0RqAWnsklpMb3CvOn6z3eJ+whPMEKrAQDlsWvgUZqZ6Ny5M3322Wc0bZqRaX3O8Vu9mu69915KTNR4CSyB4uJi8vHxoaKiIvL2Pr+nDABwftjLvudsPtU3NIi8dnk4+rkY3dVfeAalMN3bR3cW9cnZS8Veda6tzKF9Y9/aqAvzlvL52FvAoafGYOEvrukd4OVCd4/t0iSv3lSu/nSHLrzYz92JLuoVLPIIe8hy5U3l8k+20+GUxrQlDrnlkP5HJncTnlI+51wGjEOXObWAa9fzZw7X5VSEWYPD6Z1Z/cVEAJehk2A1bf6/7/6nEdliob93ZvWjCd2DWpxDDdSBRaZOZZSIXOoJWjE5zkc//sp0kSfP7XxJvzBhRH+97Sy9svyEEEC7cXikzhtv6IlnT+9fD4zSGYUllTU04o31YqKNjfSXLustomh2n9WIHDKPX9Rd9EfuR8PfWN/s/ibOn6l7z6XpXvz7mBD44xBwDPJb1/7dnlul+8z3lm3PTLKK8nygfcGT7iyUyhPHLIoJAFDWDjXrXT8jI4N69+7d7Pd9+vShzEz1a94CANqWt1fHioe7HBbE8XBxEB6/64dG0i97k4XYkzwPWDLmn57eg+7X1mCXwk5ZfI7DxTnUm415FoNa99h4cnSwE7l7bLQ0B4eBPy+r7aw0nEsuwSJ4aghQdQ/y1DPo+Rx8uukMnckuFZMZN365m4oraoT6OdelljOhRyA9PT1GvPd0cRQvNuwZeVkhpk9H7/Pm9APzwIruLIbGSG0mlRzjiglXDWoUdbx9TJQYSLNxz2KGXFqqoFzj1ZeM+SendqcHJ3XT24aXqxP9du8o0S+5r3GqSFSAO015f4vuN++vPU3XDY0QOfoMb4PFEnNKq4SonzE4h3vR7MGKn5P21v4cUcQl/rjkHKe6wJgHlgjfE+5tZkIdAGA6ZjXoWfyOve/h4eHNKuF36AAFZABsLbeXPfE8+JQ4mqYxPDmc3t3JkW4d3ZluMFDvHtddk1/OXr8/DqRRYUU1FZbVkJOjnVDYlrhtdGfaeSZPeCMf//WQrq47hyNLHmQPTTW4NkM+mcAicWrAYnO/7U9tspxV+yXlfuap34/off/FLUNoSs8gPQ8ph3BLBr0hKC1kmfxy9wj6+1AaPTKle7O/4RQVieuGRurq1DOc8373OOMDbg6ZlcOpGazxIJV3FNvfk6Irk8fXE6uqM8UVtSKdRl5CDyjHW1f3E2k9oT7IhQcAgPaKWQ16DrV/7rnnaO3atSKXXk5VVRW98MILNH36dHPuEgAthj1UnKfMJZu49A9CRI3DpbI2nc6hY6lFQiSNBbI4T569cRxqrvmNJsT+oxsG0YCIc9clZ+851zpvDg4NZrGtW77eo/NQuzk5NJkgaEuk+ruMr7s6at0cWs0eeB7Y3/n9PrGMjahyA3VxKdeWWXLncBrVtanqPE+SLN6cIEQEeUJFKrP34MSu4jtgefTp6CNeLYUrE/Tt6CO0HFhXgj1oF6Ib8dS0GHEtcxTNiqMZdCS1UKdE3yO4cQLgfzNiREoAe5KB8nA6DYx5AABo35g1hz41NZWGDBlCLi4uonRdTEyMEFo6efIkLVq0SBj1+/bto4iICLIUkEPf/uAyXpnFlUIxmg0xVgH+cF2cnriTlDOqplG8JS5H5DjfPjpK8ZxupeF+zDm2ro72NPbtjU3E6CQ49JvrW8e8sFp83v/8FBEarASc583idhx++uZVfXWhx5bAsv2pOtExeR6xGpRV1VJvbQ785JggmtIrmBaujxOq/Zz77OHsSE9O60E3jehktGSesbZlhWKuOMBpDADI2ZWQJwSveJLTz91ZiOUtvmkQTe8TihMFAAAAmMEONatBL4XV33///fTff/+JgaLYCTs7uuiii+jjjz+mrl01ebGWAgz69gXXB2fDi8v5nE+c7ac7RzRrWLKatFy5uSVwf2A1dy4nxeXUJD67ebAoAWOp8H6zYvX6U9l6y+8e10WE+PKkBItxGcLe4+MvT1M80oFD/FtiqJp7kujnvck0rHMH6makFJ/S/Lo3RYTZc2SDPF8/PruE/D1cUNMbKAbn3w96da14zx5+Fr/a+OQEvfB+AAAAANiIKB4TFRVFq1atooKCAoqL09QyZiMeufPAEmBxJ2PGfPdgT7pzTBf6cXeSKNsVn13a5DdVtXXiu1mLd+pyuF+4pNd5jUv+f7sS8ikhp5QWrNP0Cf4v3m5OYl+OphZZtEG/6lhmE2Oe60b/b0ZP3efrh0UIITwusSUxUFsOTWkszZhnOMJi9vBOZtvetUMjxMsQTVk6AJTVh+gS4EEJ2vKRnJYk1bEHAAAAgPq0WW0TPz8/GjZsWFttHgCdd1kyKllQ6ni6Jr9473NThBL0gnWnqbCiRuQOB3m70vS+IdTvpf9EHfTdCXnCC3pxnxCRgzpz4Ta9s8rGK+eQf3LjoGYN18Lyarr/pwMipFniyoEd6eXLewsV6hf/Pq4rd2ZJ5+yJXw/T3qR86tTBg7bF5+omMDhPdvWxTBHOLYfrHXP5M1abzyyqoOuHRdK4bhrROwCAdfPLPSPo590ptOJoOk2KCbbISTUAAADAVjF7yL21gZB724DzirneNofB+7o5U0pBOX28IZ4OpxbS7/eNonUnsuiNlSeFwcnGO+cYN8fo+RtESP25YDGx/45nUW19gxBne+PKPkaNehZy23I6R7wP83Gl8A7u9NlNg0VI9IHkArpq0Q7xHeedc96+JZQk4jxsw1JmDJdDU6McGwAAAAAAAO2NYkvNobc2YNBbL3xpv77iJH2/M4mqZSrj54LLLX1285Bzepj2JebTq8tP0OHUIr3lQV4uQoDsvvHRFNHBnaYv2EKntHWZv71tKE3oEaS3b+zdv+eH/eLz21f3axIizb955JdDOjG+zv7u9P51A8jLxZG6BHqKUnAOdnYXLJj3ycZ4+mb7WTHRwIJ7POnAJd+4DjzPOfx7OJ26BnlSv3CN+nxOSZWIJOBleWXVNPyN9SJPfVCkrwi3XXcym64a2JHendXf4sX7AAAAAAAAsAZg0Jv5RALLQ9Rk/uWQ7nOApwvll1VRfQORp4ujXp1tP3cnemJqD7p+aAQ5yuqlNwcb2xwK7+XqRP8cTqOC8hp6ZnqMqN8t8fDPB3XG+EOTuor1M5lFlfT0siM6z/zUXsH0+S1DjG4nr7RKqMYblh6TcHKwo6m9QuiNK/uSi5O9qPfMBvm946Npau8QOpZWRN/vTKTL+ncUdcp526PmrxfnwLCsGZd6C/V1pYScMnF+Nj01Qaj8X/T+FnGuxncPpGFRHeidNbHUJdCD1j8+HqX7AAAAAAAAUAEY9GY+kcBykMpsPbb0EGWXVJGHswO9ekUf4YHmkHqpdi/XTb5y0Q5ytLejbc9MohAfV0X3I7WgnMa8tVG87xnqTf8+OFqE+F/9qUY0z9nBnm4fEyWM/XOF0rMAHytJL958hjYYiM9J8Lr8PJxEbj/T0ddNKE2PeHO9+L/8/WtX9hHG/tY4Tc57r1BvOiGrSW5IJ393SsorN/rdnJGd6OXL+1zA2QAAAAAAAAC0FBj0CgGD3rpgT/Jd3+2jnQkakTlWW17x8BjhSTfGhlNZFBXgqVqJpeziSpr47iZRo71fuI8Q3eNwdYb3q3eYzwWtjxXxVx3NpKziSrp2SARtjM2mx3/V1Dc3pDmD3cXRnpY/NEaE0HOoPXvxHeztqYEaxEQARzI89PNBMRHAcBT9/Kv60erjmWJCgSdIfrt3FPUKwwQXAAAAAAAAagCD3swnElgGrLB+74/7hUf6xuGR9Mjkbm1ec3v1sQx68rcjeiH+N4/oJKIGTKW2rp5Gv7VBeOYfndJN1B3n/P4fdyXrfsMq/Gyo700qoJziShH6f/Xg8HOut6i8hlYdy6BNsTlCF+Aa7e9PZRZToKcL+Xu6mLzvAAAAAAAAAOPAoFcIGPTWxUfr4+i9taeFSBsLyFkKnLvOueeVtXVCAE9JtfozOaVCuG5EF3/dspT8clp3MotOZhTTAxO7Uid/dSIQAAAAAAAAAG1nh7Z9DSwAFMydP5KmUZ7vFuxlUeeV8/Pfu7a/KuvmUnz8ksMq+7eNjlJlewAAAAAAAADLAAY9sAk49PzGL3fTnrP54nO3IH0DFwAAAAAAAABsjfPX5wJAQcqra+loahFVNFOGrbV8tzNJZ8yP6NKBRkY3hp8DAAAAAAAAgC0CDz1QPQz+TE6ZKBHHBveKIxlUUlUrROsmxQTRhzcMIBdHB5O2seNMLr258qR4//zMnnTn2C4K7T0AAAAAAAAAWC4w6NsprLj+vz+OkquTPT03sxf5uBkv63Y+Y51LsDk6aAI91p3IorO5ZaK2uoO9HaUXVtDTvx+hbfGauucS7s4OVF5dJ8qgbTyVQ9P7hDRZ76pjmXQ8vYjuGNOFOhhRqeeybaxozy+pRN3lA8LojjHIGwcAAAAAAAC0D2DQt0PYYH5t+Qn653C6+OzsaE+vXdG3xd5wroOeWlAuapyzunpMiDdll1RRbmmV+M1X285SpL87HUwuoJq6BuGN5xrsfcN96KKewUKN/aV/j9P3O5NEibktT00Uv//jQCp9tCFe1D8vqqgR6/p1Xyq9eWVfGtXVX3jyuSb6sgNp9PI/x4WnX6J/hC+9dXU/srOzU+WcAQAAAAAAAIClYdfA1h1oN2Xr6usb6OV/j4uccwk2uJfdN0oY3Iawl33tiSzam5hPW07nUHFloxHdEoZFdaD5V/WlLgYq7JxHf+nH28T7cD83Gh7lTyuPZlBFTZ3Oix/o5UJJeeW6/8PGfL3sag3wdKEZfUPool7BNKRTB3JzNi10HwAAAAAAAAAsgXZXh37+/Pk0d+5ceuSRR2jBggViWWVlJT3xxBP0yy+/UFVVFU2bNo0WLVpEwcHBNmvQn0gvFp7zxNwyEbKeUVRJTg721DXIk2rrG2jDySwqq64jdmS/cWVf2nAqWxjsbFQvf2gM+bprwtv3J+XTxxviRbg8e9nlcGj7yC7+FNnBnTxdHSk2s0QY4rztaX1CqK6ugdKLKkTptAndA5v1mm+Pz6XZX+7WW8brfHJaD7F+L1dHendNLH2/K4mqa+t1v3FysKNHp3Sne8Z10YX7AwAAAAAAAICt0K4M+r1799K1114rDnTixIk6g/6+++6jFStW0LfffitOxoMPPkj29va0fft2mzTo9ycV0NWf7jjv79jT/fY1/emaweEitP3Sj7ZRcn658HhzaHxeaRUdTtXUc2d6BHvRzH6hwmse6uMm3ivFtrhc+mVvMoX7uQt1+lHRASIFQA7n6VfV1lF6YSVV1tSJyQdp4gEAAAAAAAAAbI12Y9CXlpbSoEGDhOf9tddeowEDBgiDng88MDCQlixZQtdcc4347alTp6hnz560c+dOGjFihE0Z9Is2xdOve1MoMa9ceONjQrzIz91ZlG+zt7OjMzmlIv+9sLyGHpvSnab0aoxSYM/6jV/uEt/J4fU8N6MnjekWILz8AAAAAAAAAADUp6V2qNWL4j3wwAM0c+ZMmjJlijDoJfbv3081NTViuURMTAxFRkae06Dn0Hx+yU+kpcNzMu+siSWemnFxtKfPbx7cJGedeWBiV6P/v1eYN616ZCw9u+woebg40KX9wijEx5X6hfsKtXoAAAAAAAAAAJaHVRv0nBt/4MABEXJvSGZmJjk7O5Ovr6/ecs6f5++a480336SXX36ZrAkOSb9+aATV1xNdNyzCqDF/PjiU/rvbh6myfwAAAAAAAAAAlMdqDfqUlBQhgLd27VpydXVVbL0srPf444/reegjIiLIkmFhuDev6tfWuwEAAAAAAAAAwIxYbWI0h9RnZ2eL/HlHR0fx2rx5My1cuFC8Z098dXU1FRYW6v2/rKwsCgkJaXa9Li4uIkdB/gIAAAAAAAAAACwNq/XQT548mY4ePaq37LbbbhN58s8884zwqjs5OdH69evp6quvFt/HxsZScnIyjRw5so32GgAAAAAAAAAAaOcGvZeXF/Xp00dvmYeHB/n7++uW33HHHSJ8vkOHDsLT/tBDDwljvqUK94xUBMAaxPEAAAAAAAAAAFg/kv15vqJ0VmvQt4QPPvhA1J1nDz0r10+bNk2Ut7sQSkpKxF9Lz6MHAAAAAAAAAGBbsD3K5etstg692tTX11N6erqICLCzs8wSbpJwHwsFKpHzP3ToUKOVA4DltNGFgPa0nrZqCWhPy2+jC8GW29OW2smW27M9tpM1tifayTraE+1kPe1ZbAX3PjbT2ZgPCwsTTup26aFXAj554eHhZA0oJeLn4OBgsRe2tdMWQotoz9ZhqaKYaE/Lb6MLoT20py20U3toz/bUTtbcnmgn62hPtJP1tKe3hd/7zuWZt3qVe6AeDzzwAE6vDYH2tC3QnrYF2tO2QHvaFmhP2wLtaVugPRtByL0NwCEjPHtTVFRk0TNM7Rm0kfWAtrJ80EbWAdrJOkA7WQdoJ+sA7WQ9FNuQ/QQPvQ3g4uJC8+bNE3+BZYI2sh7QVpYP2sg6QDtZB2gn6wDtZB2gnawHFxuyn+ChBwAAAAAAAAAArBB46AEAAAAAAAAAACsEBj0AAAAAAAAAAGCFwKAHAAAAAAAAAACsEBj0AAAAAAAAAACAFQKDHgAAAAAAAAAAsEJg0AMAAAAAAAAAAFYIDHoAAAAAAAAAAMAKgUEPAAAAAAAAAABYITDoAQAAAAAAAAAAKwQGPQAAAAAAAAAAYIXAoAcAAAAAAAAAAKwQGPQAAAAAAAAAAIAVAoMeAAAAAAAAAACwQmDQAwAAAAAAAAAAVggMegAAAAAAAAAAwAqBQQ8AAAAAAAAAAFghjm29A5ZOfX09paenk5eXF9nZ2bX17gAAAAAAAAAAsHEaGhqopKSEwsLCyN6+eT88DPrzwMZ8RESE0u0DAAAAAAAAAACck5SUFAoPD2/2exj054E989KJ9Pb2Pt/PAQAAAAAAAAAAkyguLhaOZckebQ4Y9OdBCrNnYx4GPQAAAAAAAAAAc3G+tG+I4gEAAAAAAAAAsEjq6xvaehcsGnjoAQAAAAAAAABYDLV19bQxNofmrzpJZ3LKqIOHMwV6ulCglwuF+LiSi6M9VdfWk4uTPfUJ86Hrh0VSewUGPQAAAAAAAAAAi+CHXUn07ppYKqqo0S3LL6sWr9iskia/79sRBj0AAAAAAAAAANCmbD6dQy/8dUy893V3oisGdKQ7xkRRWXUt5ZRUUXZxFWUUVVBtfQM5O9pTSWWt8Ny3Z+ChBwAAAAAAAADQphSV19DcZUfE+2uHhNObV/UjB/tGQbiYkDbcOQsGongAAAAAAAAAANoM9ro/uvQgpRdVUid/d5p3aW89Yx40Dzz0AAAAAAAAAADahI/Wx9GH6+NEGL2jvR19dMNA8nCBmdpScKYAAAAAAAAAAJidDaey6L21p8X74VEd6NEp3alfuC9a4gKAQQ8AAAAAAAAAwKxUVNfRi38fF+9vHdWZXrqsN1qgFSCHHgAAAAAAAACAWflkYzylFlRQmI8rPTWtB85+K4GHHgAAAAAAAACAWUjMLaN3/oullUczxOcXL+2NnHkTgEEPAAAAAAAAAEB1ckur6JrFO8Vf5qYRkTStdzDOvAnAoAcAAAAAAAAAoCr19Q301G+HhTHfNchTqNn3DPXGWTcRGPQAAAAAAAAAAFSjoaGBFqyPo42xOeTsaE8f3ziQYkJgzCsBDHoAAAAAAAAAAKqw/mQWvbMmlk5llojPr1/RB8a8gsCgBwAAAAAAAACgOGdySuneH/dTTV0DuTk50IOTutKsIRE40woCgx4AAAAAAAAAgOK8vfqUMObHdgugj28YRD7uTjjLCmO1dejr6urohRdeoKioKHJzc6Po6Gh69dVXRX6GBL9/8cUXKTQ0VPxmypQpFBcX16b7DQAAAAAAAAC2zuGUQlpzPIvs7YjmXdoLxrxKWK1B/9Zbb9Gnn35KH3/8MZ08eVJ8fvvtt+mjjz7S/YY/L1y4kBYvXky7d+8mDw8PmjZtGlVWVrbpvgMAAAAAAACALfPuf7Hi75UDw6lrkFdb747NYrUh9zt27KDLL7+cZs6cKT537tyZfv75Z9qzZ0+jkuKCBfT888+L3zHff/89BQcH019//UXXX399m+4/AAAAAAAAANgaReU19MryE7Q1LpecHOzo0Snd2nqXbBqr9dCPGjWK1q9fT6dPnxafDx8+TNu2baOLL75YfD579ixlZmaKMHsJHx8fGj58OO3cubPZ9VZVVVFxcbHeCwAAAAAAAADAuamsqaOrPt1Oyw6kkp0d0VPTelBEB3ecNhWxWg/9s88+K4ztmJgYcnBwEDn1r7/+Os2ePVt8z8Y8wx55OfxZ+s4Yb775Jr388ssq7z0AAAAAAAAA2BY/7EyiMzllFODpQotvGkRDOndo612yeazWQ//rr7/STz/9REuWLKEDBw7Qd999R++++674awpz586loqIi3SslJUWxfQYAAAAAAAAAWw21/3hjvHj/9LQeMObNhNV66J966inhpZdy4fv27UtJSUnCwz5nzhwKCQkRy7OysoTKvQR/HjBgQLPrdXFxES8AAAAAAAAAAC1j0aZ4Kqqooe7BnnT14HCcNjNhtR768vJysrfX330Ova+vrxfvuZwdG/WcZy/BIfqsdj9y5Eiz7y8AAAAAAAAA2CK7E/Lomx2J4v2zF8eQA9eqA2bBaj30l156qciZj4yMpN69e9PBgwfp/fffp9tvv118b2dnR48++ii99tpr1K1bN2Hgc936sLAwuuKKK9p69wEAAAAAAADggqirb6CjaUWUWVRJeWVVlFdaTbmlmr85pVWUX1ZNfcK8qU9HH4oK8KCx3QLJ2VFdH+77a0/TRxviqKGBaEzXAJrYI0jV7QF97Bq4vpsVUlJSIgz0P//8k7Kzs4WhfsMNN9CLL75Izs7O4jd8aPPmzaPPP/+cCgsLacyYMbRo0SLq3r17i7fDXn1Wx+d8em9vbxWPCAAAAAAAAACMk1VcSQ8tOUh7EvNbfIqCvV3ottFRNKSTH/l5OFOEn7uiBv6B5AK6atEO8f7aIeH0wiW9yMvVSbH1t2eKW2iHWq1Bby5g0AMAAAAAAADagtq6etqbWCAM52+2n6Xc0mpyd3agHiFe5O/hQoFezuJvgKcz+Xu6kKeLozD4U/LLaffZfMopqdJbH//u3vHRNGtwBPm4m2543//Tflp5NJOuGtSR3r+2eZ0ycOHAoFcIGPQAAAAAAAAAc5NeWEH3/3SADqUU6pbFhHjRpzcNFuH056O6tp7+OpRGv+9PFd59Nu7Lq+t03/cM9aYXL+lFI6P9W7V/8dmlNPWDzVTfQLTm0XFikgEoBwx6M59IAAAAAAAAAGgtHDgdm1VC+xIL6GxuGS3dm0KlVbXk5eJIY7oF0Khof5o1JIJcnRxatf6aunr640Aqfbn1LMVll4pljvZ29NbV/S5Ylb64sobu/HafiAaY0jOYvpwzpFX7BJoHBr1CwKAHAAAAAAAAqCFwt/ZEFu0+myfKvbEhn5xfrvebARG+tPD6gRTp767otllI7+V/T9C/h9PF52sGh9OwqA40vU8IeZ8jBz6jqII+3XSGftuXShU1dSL8f8XDY1sUMQAuDBj0CgGDHgAAAAAAAKCkIc9l3l769zidztJ4yiVYsG5EF3/qEuBBAyN96dJ+YWSvUgk4jghgo/5bbbk5xsvVkZ6eHkM3DY8UVcP4N9vj82j7mVzaGpdDx9KKdb+NDvSgly7rLZT0gfLAoDfziQQAAAAAAAAAOfX1DcLrziH0Z3JKhRd+x5lcKq6sFd/7uDnRlQM7UrC3K0UFuAvj2MPFfJXF2WDffDqHdibk0boTWXQmp0ws7xLoIRTx2ZN/PL3RiOe5hV5h3vTUtBga1y1AGP1AHWDQm/lEAgAAAAAAAIBkyC87kEofbYhvEkYvecIv6RdKT0+LEeXkLGWfv9yWQG+tjhVRBBJODnZ0af8wGta5A13UK1io6QPLsUPNN/0DAAAAAAAAaDOqauvI3s6OnByUq0MOGikqr6GVxzLoeHoR7Tmbrwund3G0Fznm7PXuGeJNo7sFUL+OPuRoYe3Aof13j4umy/p3pNNZJZRZVElOjnY0rlsgjHgLBgY9AAAAAAAANkx2SSU99+cxWn8yS5QYYyGzyT2D6elpPSiig7Jia+2RwvJqWn4kg95ZEyvE7SRYnf7BSV3plpGdyc25dcr0bUGIj6t4AesABj0AAAAAAAA2SG1dPW2LzxXGfFphhW451yJndfP/jmfSJf3CaFiUn/DKWpPRaQkk5pbR838dE4JxDdoI9W5BnjSpZ5DwxI/tFgDPNlAduwZWQgDNghx6AAAAAABgbXDI9wNLDlBOSZX43NnfnT6ZPYjC/dyFQNv8VSdpV0K+7ve9w7zpqzlD4Zk9D1y2TRK249Jttdpc865BnnT90Ai6dVRniwulB9YJRPHMfCIBAAAAAACwBE5lFtOsxTuppLKW/D2caUbfUHrsou7UQSa+xj69HWfyaOeZPPp5TzLllVVTTIgX/Xn/aHjqjVBTV09fbj1LH6w9TdV19brlE3oE0guX9KLoQE/zNC5oNxS30A41i4d+4MCBLS5pcODAAbIkYNADAAAAAABLp6K6jn7bn0Lb43NpR3welVTV0tDOfvTDHcPJ1encofQp+eV05aIdokTZkE5+dMOwSLqodzB5uzpRe6a6tp5WHcsQ0Q6cupCUp1Gr7xXqTf0jfOnyAWGiZjwANq9yf8UVV+jeV1ZW0qJFi6hXr140cuRIsWzXrl10/Phxuv/++82xOwAAAAAAANgMbIjf+d0+OpRSqFvWp6M3fXnL0PMa8wwL4y28YQDN+XoP7UsqEC+/FU70270jqWuQF7UXUgvK6e9D6ZScV04F5dW0P6lARC5IcLTDMxfH0KzB4ai/DtpvDv2dd95JoaGh9Oqrr+otnzdvHqWkpNDXX39NlgQ89AAAAAAAwNLgIfzBlELalZBHX209KwxPX3cnumtsFxoZ7d+qsmicW//3oTQRgp9VXCW80H/dP8pmjVf2wG84lUWxmaV0MKWANp/O0YnbSQR5uQhPfJ+OPjQpJoi82nnUAminIfdyeKf27dtH3bp101seFxdHQ4YMETtsScCgBwAAAAAAlgR7kJ/944jIgZfoHuxJi2YPFuJsppJdXEnj39lEFTV19PnNg2lq7xCyJUoqa+hgciG9/O9xOpNTpvfd6K7+NCLKn3w9nKlTB3caFe0PkTvQJlhUyL0cNzc32r59exODnpe5uqLeIQAAAAAAAMZy5NeezKIDSQW0dG+KMLZdnexpfPdAGtM1gK4dGkEujsqUnQvydqXbRnemRZvO0Lx/jpOnqyONig6welE7jjz4bkeinhEf4OksPO+dAzxoeu8Q6gJxO2BlmN2gf/TRR+m+++4T4nfDhg0Ty3bv3i1C7V944QVz7w4AAAAAAAAWS119gzBEP9+SQMn5GlE2ZkSXDjT/qn7CEFWDe8ZF04qjGUII7sYvdtNrV/Shm0Z0Imuktq6e7vlhP204la1bFurjSuO6BdKzF8eQn0z9HwBro03q0P/666/04Ycf0smTJ8Xnnj170iOPPELXXnstWRoIuQcAAAAAAOamsqaO/jmUTkv3pQhxNibQy4Vm9g0VxvzUXiFkb2+nemj6a8tPin1gT/b2ZycpFgVgDtjM+e9EFn2xJUEI/XFEw9yLe9Kl/cP0SvgBYIlYbA69tQGDHgAAAAAAmJP47BJ64rcjdFirWu/kYEdPT4uhG4dHkoeLo9lD1ce+tZEyiytpwXUD6IqBHckaYBPnleUn6JvtieIzz30svGEgXdIvrK13DQDrzqGXqK6upuzsbKqvr9dbHhkZ2Va7BAAAAAAAQJuRU1JFS/cm0wfr4kSovZeLI909rgtdPqAjRfq7t8k+OTnY0+zhkfTe2tP03c5EqzHov9p2VmfM3zU2im4c3omiVEpPAKAtMbtBz2r2t99+O+3YsaPJLBqXxKirqzP3LgEAAAAAANCm4fUv/XNchLZLsbMs1PbczJ4UbQEibdcPi6SFG+KEMvyR1ELqF+5Llpwvv/JYJr256pT4/PzMnnTn2C5tvVsA2I5Bf+utt5KjoyMtX75c1KO31bqWAAAAAAAAtKQE3f1L9tOxtGLxuW9HH7pqUEe6dVRnixknS7n7fx1Kp+92JNF711qmQb89PpceXXpIRDowVw7sSHeMiWrr3QLAtgz6Q4cO0f79+ykmJsbcmwYAAAAAAKDNYbG5H3Yl0e/7UylBW0KNRdoWXj+QxnSzzPJwc0Z1Fgb9sgOp1KejN9022rIMZY4cuOv7fVReXUd+7k40a0gEPTalu8VMigBgMwZ9r169KDc319ybBQAAAAAAwOxwLvzuhDz690gGnUgvooLyGiEwV13bqCM1PKoDvX/dAOro62axLTQw0k94uzk3/eV/T4jUgNstxPt9NreMbv92rzDmR0X701dzhpKbs/Wo8QNgVQb9W2+9RU8//TS98cYb1LdvX3JyctL7/lwKfgAAAAAAAFgy6YUV9MeBVNqbWCDqxmcVVwpD05CuQZ5CrI3Lz1lLHXTOR/d0caQP18fRxxvjRV16Z0f7Ntuf/LJqemfNKVq2P42q6+qpV6g3fXbzYBjzoF1h9rJ19vaaTm8Y/mKpongoWwcAAAAAAM5FakE5vfffadp8OkcYmYb4uDnRjL6hNL57gMhH9/dwoU7+7lYZDs6ic6Pmb6DskipaNHuQOK62iny44YtdtOdsvvg8tLMffXLjIArydm2T/QGg3ZSt27hxo7k3CQAAAAAAgKLEZZUIL/zRtEJadiBNF0LPNvqQTn50Wf8w6hrkRUHeLhTh596mnmwlcXSwp1lDwumTjWfol70pbWbQf7IxXhjzHs4O9NWtQ2lEF/822Q8A2hqzG/Tjx4839yYBAAAAAAAwifr6Bsovr6a4rFL6evtZWnsiS+/7EV060OMX9RAq9baev33tkAhh0G+NyxHRCeF+7mbd/rG0Ilq4Pk68f/WKPjDmQbvG7AY9s3XrVvrss88oISGBfvvtN+rYsSP98MMPFBUVRWPGjGmLXQIAAAAAAKAJGUUV9M32RKFILw+nZ088C7B1D/aii3oG08hof6sMoW8Nnfw9aHRXf9oen0e/7Uulxy7qbpbtcoruX4fS6PUVp6i2voEu7hMiStMB0J4xu0G/bNkyuvnmm2n27Nl04MABqqrS1Ink3AAWylu5cqW5dwkAAAAAAAAdaYUVtHBdHO1JzKfEvDKh6M6wvc7575NiAunucdFC2K69ct3QSGHQ/7wnma4bGkFhZlDof3tNLH266Yx4z+eevfPtZRIFAIsRxRs4cCA99thjdMstt5CXlxcdPnyYunTpQgcPHqSLL76YMjMzyZKAKB4AAAAAQPsJq//7cBq9+NdxKqmq1Ssrd/e4LjS2W6DN5MKbSmVNHU1+b7OY/AjycqFvbhtKvcN8VK0zf8Un26m+gUR9+XsndCEXR9tObQDtm2JLFcWLjY2lcePGNVnOO1tYWGju3QEAAAAAAO2cvNIqen3lSdp4KlvUiWcGd/KjByd1FTnxAZ4ubb2LFoerkwMtvWcE3fHtPorNKqFHfzlE/z02ThWPeU1dPT2z7Kgw5lls8JEp3RTfBgDWitmnGENCQig+Pr7J8m3btglPPQAAAAAAAOaAA1W3nM6hSz/aRn8cSBPGPKumP35Rd1p69wia2CMIxvw5YDE8NurdnBwoLruUDiQXqNJGC9adppMZxeTn7kQvXtpL8W0AYM2Y3UN/11130SOPPEJff/21mMFLT0+nnTt30pNPPkkvvPCCuXcHAAAAEKGjOxPyaMPJbBE+yuWnOD/znvFdKNRH/bxQAID54P69+ngm/bYvhY6mFVGh1iPfJcCD3riqr/DMOzkgrL6l+Lo708x+oUI08Jc9KTS4UwfF2io+u4Se/O0IHUrRRPHOu7Q3JlgAaGuD/tlnn6X6+nqaPHkylZeXi/B7FxcXYdA/9NBD5t4dAAAA7Yzy6lpRdmrDqWzam5gvDPiMokpdDWmJbfG59MveZLp+aCT1j/ChPmE+wsiHABMA1gF7dosqaig5v5z2JxXQifRiOplZTKezSvX6u5ODHc0e3omemNqdvFyd2nSfrZUbhkUIg375kQzhQVfiPFbV1tE9P+ynMzll5O7sQI9M7kaXDwhTZH8BsCXMatDX1dXR9u3b6YEHHqCnnnpKhN6XlpZSr169yNPzwlVC09LS6JlnnqFVq1aJyYGuXbvSN998Q0OGDNHdyOfNm0dffPGFyM8fPXo0ffrpp9StG/JuAACgPcHPg9/2p9LHG+LF4N4YId6uNKVXkDDc7e3thPdub2IBfbsjUfebXqHedGn/MIoJ8aIhnf0w+G+nsDHIE0H5ZVVUU9dA/cN9bb7uuKVTVF5DW+NzKCmvnFILKuhUZrEI0a6s0Z+okwj2dhGTdRf1ChYTdZwPDlrPoEg/cR7js0vpn8PpYoLEVL7eliiM+QBPZ1r+0FgK8XFFEwHQ1ga9g4MDTZ06lU6ePEm+vr7CkG8tBQUFwkCfOHGiMOgDAwMpLi6O/Pz8dL95++23aeHChfTdd9+JGvcc0j9t2jQ6ceIEubripgAAAO3BkGexpo82xNOKIxm65d6ujqJm9IQeQSLMlgeKkR3c9bzvswaH0+pjmbT9TC6dziylQ6mFdCKjWLwYHzcnoXp92+jO5O5s9oA30EapGT/uShJls/Jk9chZ9bxniBd5uzmRh7Mjebg4kpcr/3WgbkFeIg/bxx2eXyWoq2+grOJKYbBzpA174E9lltDWuBwxuWIMzrseEOFLfcN9qVeoF/UM9aYIP3cxcQeUge+d1w2JEMKCS/em0A1DI006v5lFlfTRhjjx/n8zesKYB8CSytax9/ytt94SIfemhu6zt3/r1q1Gv+fDCgsLoyeeeEKE8zMs+R8cHEzffvstXX/99S3aDsrWAQCAdXIwuYAe+vmg8NYxjvZ29PjU7jR7WKdWGVcFZdX016E02pdUQIdTCvXWyyWbOvq5iQmCnqFewsDnEFFWxg71cUWYvpUbkMsOpNKm2Gzaczafcks1hjyLgAV4OVNNbQNlFleecx0O9nbUO8ybOvl7UP9wH7p8QEcK9IJq+rlKx+04k0frT2UJ4z27uIoKK2rEhAoberUsdW6E7sGe1KejjxBqiw70EO87+rrB+24mckuraOSb68XECpf5e+/a/qItLpTYzBJ6bcUJ2hqXK/QMfr93JO6hoF1S3MKydWY36FevXk1z586lV199lQYPHkweHh56359rZ+Wwd5+97ampqbR582bq2LEj3X///UJ0j0lISKDo6GhR337AgAG6/zd+/Hjx+cMPPzS63qqqKvGSn8iIiIjznkgAAACWAT/WOOTz2WVHqaKmjlyd7Gl4lD89dlF34aVTysj793C6UF5OzDMewi/BnnwO1R/fI1B481E32fKprasXqRn7Egvosy1nRNivBBuID0/uSlcPCidHB3txvfH3CTmlVFZdS6WVtVRaVUelVTVUXFFLu8/miZxtOTwJNKZbgCiHNio6QESLtGfYgOPJsuPpxWLiLKOoQjdx0twECUfWxIR6k7+HMwV5u9BUETrvZdb9Bk3562Aa/e/Po1ReXUfhfm604uGx4h7YEnJKqui+H/eLSVOGHfz/PDhGTMwA0B4ptjSD/pVXXhHeci+vxputPLSRd4M/c559S5BC5h9//HGaNWsW7d27V6jnL168mObMmUM7duwQIfmsoh8aGqr7f9dee63YztKlS42u96WXXqKXX365yXIY9AAAYNmeoX8OpdPKoxkixL6kslYsH9c9kBbNHkSeLo6qeRKzSiqF15DDflloj9+z8B4PaHmAKvcmju7qT1/eMhT51hacG//znmQR6is3KDlF47bRUTSii7/wGHKI/YWQkl9Ox9KKKCm/nNYcz6SDyRrFbomrBnWkVy/vI0L1bY3iyho6lVFC609m0dncMmHc+Xk4i7riibllYpmxSTFOWZjZN5R6hHhRkJcr+Xk4ickwTo8J9nIRkynAMknKK6Obv9ojJsVYo4DvweerGsB2AAvg/XciS0x4Te4ZRLePjqLhXdr3ZBdo3xRbmkHP+fMZGRkif/5csAe9JTg7O4vwfTbcJR5++GFh2HMZvNYa9PDQAwCAdZBaUC5E7o6kFgkjnr3mEmxw3T8hmh6c2LVNB/6s0sx5vvsS8+ntNbHCyB8U6UtPTutBQzt3QGksC4CHQayVwCHem05nU0p+hS6kPjrIg2b0DaWbR3RSVACRS3FtOZ0rSqb9fSiN+NLtF+5Dn9w4iCI6XHiIsiVENLB3PTGvjEqraim3pJr2JxfQoeQCKtZOrp0LyYBjrQFORWCDn6NaIFRnvXBa0jWLd4jwe54E44gWNu6NpZrwvfuHnYn00r8nxLXw94OjqXcYvPIAFLfQoDfbVLA0b9BSg/18sJFuKKrXs2dPWrZsmXgfEhIi/mZlZekZ9PxZHoJvCJfQ4xcAAADL9cZ/uC6Oft2XQlWy0lOcm3zVoHDhRe3k724RxgB7FDlclF99w31oztd76UById34xW6hsv3M9Bi6YkBHiHO1ESys9tI/x2lXQr5uGesePDKlG10/NEK1CRcODZfCw28cHkl3fb9PTEyNf2cjXTkwnOZf3deiJ3t4THcyo4SOpRdpwuUPpumJBBrCKuWjuwbQwAhfKquuE2H1LJjG/TTK30OI1LERD2yH/hG+9PnNQ+iBJQdEyUB+vb7iBD0+tQfNGhJOzg72VN/QQO//d5p+2ZsiJoKY+yd2hTEPwAViNg+9vb29MKZZjV4JbrzxRkpJSdETxXvsscdo9+7dwjsvieKxIB6H+kuzHEFBQRDFAwAAK4OFsVg5mb2Z8nxm9vzcOSZKGMutEV8yN1wH+8utCbTpdA7law0gLvU0ONKPBkb60pWDOiLHXkXKqmqFwB2nRnDkRHpRBfEoyMXRXhjwQzp3oEkxQWYPfefr4s1VJ4UIGHPfhGgx2WNpsKI8iwPyhFpCbmM/lNISeoV5k7erkwir51B51gbo7O9hk6kEoGVwWgVrmqw6linKCDaHl4sjXT4wjF6+rI/QSAAAkOWF3LNBzzskz5s3Rn5+4yz5ueDQ+lGjRol8dw6j37NnjxDE+/zzz2n27NniN6ymP3/+fL2ydUeOHLmgsnVQuQcAgLaBPTpcA57Fxjg/XR5S3y3Ik168tBeN6RpglerHHIr/1bazovyZlO/PRHRwo+dm9BKhqRjUKgPnarMHeffZfNp4KruJJ/niPiGiLJYlhLovP5JODy45SHxJ/3D7cCGcZyk50c8sOyLOoTRqZLFJrj3eJdCDxnQNpCk9g5DXDs6pN/Lz3mRauD6Osoqr9CJi3riyD03uiXseAFZh0C9YsEDs1LlgQbuWsnz5cqGYz/Xn2WBngTxJ5Z7hQ5s3b54w8gsLC2nMmDG0aNEi6t69e4u3AYMeAADMCwvKvfffafp6+1md8cAM7ewnwpPHdw+iDjYSnltUXiPytuOzS0UKgTTQZW/V2O4B9PzMXhTm69bWu2mVcIkzNuA/XB8nJoQkOvu7i+uIjdGoAA/y97SsNDtWCF+yO1nkGn9z69A2Ufjm8dP2+Dzak5gv8v03x+aIUHkmsoO7EPG7c2wX1cQmge3C1xZXH+EJWp6j9XB2wEQQANZk0GdmZoqQd2sCBj0AAKhPdnEl/bY/VRdSL3njWUiJPajdg70o0r/tPahqT2R8sjGevt+RRCXafFIOXX7psl50Wf+O8Ni3AB7S8MQIRz7IldNZMX328E40LMqPxnYLtOj8dJ6IuOzjbbpSdxN6BNKH1w9scekvUymprBElH1cczdBbzhNq780aYPP9EAAALAWLVbmHQQ8AAO0XfuTklFZRUl65UEHm8l1J+WUih1gWUS/qF796RR+het3e4MkMVj9/8e9jQihNMkjHdQuk52b2hMfeSKm5LadzhMDdnsQC8V6CIzlmD4+kO8ZEka+79UR1pBVW0PxVp0QZRr4eOJJg3qW9xGSEWqkYnJqwPT5XbJcjGlht/LIBYUJtnhXHR3TpYJXpLQAAYK1YnEEPDz0AwBph5d13Vp+ivw6li3rKEoGeLsJrzINrHuPyMJcHu5qxtp0wXAvKq8nR3p5cnOxpeFQHmt4nRAhEtcf6yXmlVfTGylP03/FMnffZkCGd/OjaoRHCcGUF+PZuPLCBtXjTGfp8a4Iuz97X3Yneuaa/yLFv7+cmo7CS1p7Moi+2JFBmcaXuO+6TT0ztTtcNiRAGvTVfR0dSC+n2b/eJyg4MK4OzMvzEGE2Nbq7Jbup55HJzrDHAwmWSUCPnNX85ZwgNiPBV5DgAAADYgEFvrSDkHoD2yfH0IjHAXX44Q3jLlIKNDRZ0m9k3lHp39KZgb1eRK+vv4WJzIdWcH84G19HUQlp5LJNySjRGCdtXYT5u1C3YU5SYiw70pJgQL4sQJbPUGt9H0opEeTXJY8/q/pf0C6WbRnSy6PBxpeChSnFFLe06m0fL9qfSxthsUd9aIsjLRQgkRgd5ihB1W6phzROJC9bG0ZI9SVRZ01imkeEwfNYEmN4nVOS18/2kJedy9bFM+nhjvCg5VysLjeHychf3CaV7J0RTR2g3AABAmwKD3swnEgBgO/y6N4Xm/nlUl8cd5uNKr1/Vl3qHeZMde9+pgc7mlFFqQQU1aAfIur8NpH3P4b5OIoycvV5rjmcKhWgODzaGs6O9CG31c3ciN2cHKquqE9EBnE/LBht7yib3DKKBkX5WIUR1NLVI1NaWe065PNsbV/al/hE+KM3WSmX8d1bH0lcyscBpvYPpoxsGievHVuB+xPnjx9KKROoBe6m55jkLacnhY+bJoFtGdhLGrIujA9kyfD9KL6wQ5+Tb7YlCsE4Ol97jFBWu/90/3If6hPuIEnLSOeX7FYfSs8bA2hNZuv/n7uwgFMavHtRRTIq0xwgiAACwRGDQm/lEAgCsm0MphUIR+2BKoS4Hd2KPQLpiYEcx2FXCiOayPdklVcK7uC0+lxJyyoTXOq+sSk/N/Xx0CfAQCtPXDgm3uME3exNXHsmgl/49LryJrIjN4eH9wn1oaq8QMVkBTCOzqFKUN3t7dSxV19WL6/Tp6THUI9iL7K00yoMNTp74OpZeTO+vPS30FYzBk2uX9g+jKwd1pO5B1nu8StWEzyquFOUdf9mbYvSccR68owOnAPGEUONkopODHd07PpquHxZJod6u7fo8AgCApQKD3swnEgBAVmtILN6cQG+vOaVnVD88qSs9dlF3s+Tfckh1cn658J6xV76iuk54zbxcnUT+PedP7zyTSxtP5TTxeHMe7chofyGa1dY58vP+OU6rjmXqIhvY0PzwhoE6LyFQlk2x2XT3D/t1UR8cds6ieZf2C7MaA42F7N5adYq2n8nTi15hbzN7mjkqhieD+nb0oXA/d3J1woRQc/exvYkFdCC5QEQ0HE4papIqxEZ81yAvcU7vGtuFeoR4qd6+AAAAWg8MeoWAQQ+A7SHlkO44k0cHUwroWFqxWD6lZzCN6epPw6L8qVeYZU7gFZZX0x8H0kT+qyRgxczsF0pzRnam7sGeIq/WHBMRfB4ziippz9l8em3FSZ1wF08uXDWwI90/savN6QJYGmy8vfffadp9Nk+XX8051eO6BwrRRtZr4Drr7Kn1dHUkfwsRiePIlP9OZNKry0/o5YWzGNuMviH00KRuQlsCmObB58lBFr6rb2gQ1RHag94CAADYCjDozXwiAQCWDeein80to/jsUvphV5IwQiXY6Hzxkl40Z1RnshbYsP9meyLtSsijvYn5eiXfOLeYvbUxId6i/BuHvXOOsSllu9gw2Hw6m05klFBOSSVlFVfR6awSkZcrwZMJXKe6b7jtCJJZ0/W9aNMZ+nb7WSrWKuIbg43kgRG+IvqD/w+/grxdRRoHixLytcKeWzUmYngCKDarRKQKbDiVrVs+tlsAvXBJL1EBwpa0AAAAAABTgEGvEDDoAbBiVezKWjqVUUxL96WIes5yTyCH9F4/NIIGd+4gFMOtWdGZFfk/35IgakjnljZ67eXw8X160yDqF96yMlRs6HFINwtw8STI1rhcKq/WFyVj2PPbOcBDiLOxVxUh0W1LWVUtrTuZRScyiul0ZgnFZZdScUWNmPApq65tkVYDe/UfnNRVCM6F+riSk6M9Odnbi5Bt1mxIyCkVonU8scOe9sKKGjFZ5Kz9nsnVLi8sr6GkPI2ApKGoHYtAcmTJPeO6WJwWBAAAANDWwKA384kEAJgHFpbjnPLskkohMMd/U/MrRA46h3znl9dQQVm1WG6sxFOXQA/qH+5Ld4/rIkJQbQ02xPk8sAEVl1VC6UWVtOJIhjg/XMP6tjGdaXy3QBrexb+JF5bP2cHkQo3I1p7kJp5enhQY1z2AQrzdRJ34YB9XGh7VgdydLV91H2iiLCTVeC75xjoN7BFPK6igpPxycY3EZ5VQmZGJGwkvV0fR/1oLR/tP7x0iRPzaWvcBAAAAsGRg0Jv5RAIAlDdMWdSJy8NticsRqvB5pdVUUqnxNrYULgPHKuus5syhxpaQP9wWubRP/nZYr1QVh1jfNa6LMNJZiO9AUgF9vzNJqKZL8HfjewRSlL8HDY3qIEphtcfz157gVI6PNsQLcbXkvHLKk+k0SLCnnr3rXPM8yNtFhO9zv6ytaxATBZyvzfXM/TycydfNmUJ8XIS338PFkTycHVHpAAAAAGgBMOgVAgY9AOqHxrPXcHNsDiXmlVNKQTml5JcLNfdzhQezt53zxNmgCPNxo07+7hTk5Uq+7k7CkODv2OBACHjjef73SIYozcdl8zgU2hicB9+no4/worJIoLWopQP1ImJq6uuFsa6J/qimMF9XYcQDAAAAQD1g0Jv5RAIALqzm+/qTWZRdXCXUudmQN4aHs4MQ6hoQ4StqwUcFuJO3m5Mw5l0cUb6qtbBH/qddSbTyWCZV1dSJsnKB3i5CmX5STBC88AAAAAAAbQwMejOfSABA87Bnj5XYWUjrn8PpdDilUO97Vyd7mhwTLErFSUrb/OJweYR4AwAAAACA9kZxC+1QKBkBAFRlzfFMeuGvY0LAToLF2ab1CaHuQZ4UFehBE3sEifxaAAAAAAAAQMvBCBoAoDhZxRpldRaz2xSbI5ZxTvvASF8aEOFHs4aEU4CnC848AAAAAAAAJgCDHgCgqIDWT7uT6K3VsSJPm2FNtbvHRdNjF3VD3jsAAAAAAAAKAoMegPMYqFzyi+t6czkvLsnkYGcnajHzy8/dud2rgLN6+u6z+SK0fkd8HsVmlYhz1y/ch6b3CRG58T1CvHCdAQAAAAAAoDAw6EG7pq6+QYSFn0gvpvTCCsosqhS53gXl1VRcUUMlVbXnLZ02tHMHGtTJV4i4cd3ujn5uFODhYrOGfm1dPW2MzaGdZ/LoTE4pHU8vFhMecmX6p6fH0M0jOtnsOQAAAAAAAMASgEEP2i1c6/yRXw7SgWR9xXVjcG1zF0d7cnKwF/WYSyprqKy6Tnjv153MEi85dnZEro4OQr2d66Dz/2ugBqqvJ6pvaCA3JwfycdeUXwv0dBHq7lxHPczXTby4jJh80oHN4rY0juOzS+lgcgGdzS2jvw6mUXpRpd737s4ONLNvKI3pFkCjogMo0Av58QAAAAAAAKgNDHrQ7lh7IkuUTttwMksY5V4ujnRRr2AK93OjUF83Id7m5+EsjGpvN0fxl41yQ2rq6oV3endCHp3IKKa0ggpKK6wQgnD1DUQVNXXiRVRz/p3ar/+RDX02/LncG4f5Mxzi7+/hLIzlXqHe1D/CV7yi/D0UN/Y5/52N+NNZJfT7/lTaczZf73suJ3dZ/zCKCfWmrkGeIrwedeEBAAAAAAAwL3YNnAALmgV16G0rH/6t1afosy0JumVDOvnRB9cNELXPlYIN/YKyaqqsqafK2jqtUV4v6qnbixdRuda7X1ReIyYBeEKAQ/4ziiopv6z6grbn7epIAyL9qHeYN0UFeIjQ/xBvV3JzdhCefYEd/7MTkQO8jPeFJwzY+19dWy/2gcPn2Yg/klpEh1IKxXcSDvZ2NLSzH3X296BhUR1oRt9Qo5McAAAAAAAAANNBHXoAZOSVVtGry0/QX4fSxec5IzvRJf3DaHCkn+LebQ6vD/J2Nck7zsa9o72dMMrZ883Gd2F5jTD2M4oqhNF9OKWQjqYVUXFlLW05nSNeShLs7SImBziE/vphERTq46bo+gEAAAAAAACmgZB7YNOwAfzUb4dpY2y2CINnI3n+1f3omsHhZKl4ujhS9+CmqvDyuu2XD+go/rLnPzazhPYnFQjvekJuqQj9zyrWqPIzHITDvnZjsTjseecXG+9dAz0pOtCTuod40ahofwr3Uy5qAQAAAAAAAKA8MOiBzcJh5Ld+vYfiskvFZw5Jf/biGBrbLZBsBY4G6NPRR7xaijDwGzTCfRx6DwAAAAAAALBOYNADm+NUZjFtjs2hr7adFSXoOJ/8i1uGUN/wlhu9tgwb8bDjAQAAAAAAsH5g0AObgXPP31h5kpbsTtYt6xHsRd/ePhT53wAAAAAAAACbAwY9sHq4Pvr2+Fz6ZW8KpRZUiGWTYoJoTNcAmjUknLxkNd0BAAAAAAAAwFaAQQ+slvLqWpr393H6bX+qbhnXkn/nmv40Mtq/TfcNAAAAAAAAANQGBj2wKiRV9+PpRfTl1rNC8I6rzk3rHSKM+KsGhQuVeAAAAAAAAACwdWD5AKthx5lcenbZUUrOL9ctC/JyoQ+vHwiPPAAAAAAAAKDdAYMeqAaXR+Pa7/Xir6ZUWl295j0v5+9dnRzIxdFefK6qraPq2nqqrKmnwopqUUP+VEYJbYnLoYScMp0h7+HsQP0jfMXrjjFRevXZAQAAAAAAAKC9AIMemBT+vuNMHh1OKaSckirNq1TzN7e0isqr61q0Hi6hxsZ+S3530/BO9NT0HuQNoTsAAAAAAABAOwcGPdDBHvKU/HLKLq6i4soayi2tFp5xNtIrquuosqaOKvilfc+Ge0lVrcln0NCYd7C3Ix83J/J1d6JOHdxpeBd/GhTpR9GBHuQPbzwAAAAAAAAACGDQt1PYcP/7UBodSikS7/PKqiivrLpFnnI5AZ7ONK57IIX7ulGAlwsFerpQIP/1chHidGyc29nZCeE6e/HXjuztNe/tiHQTBE4O9uTiZE/ODvbk6GCv1mEDAAAAAAAAgM1gMwb9/Pnzae7cufTII4/QggULxLLKykp64okn6JdffqGqqiqaNm0aLVq0iIKDg6m9w0b8u/+dbrKc89NDfd00HnI3J4oK8BCf3Z0dyM3JQeS8u2nfs8HeI8RLGO2txcvBHnXiAQAAAAAAAKC9GvR79+6lzz77jPr166e3/LHHHqMVK1bQb7/9Rj4+PvTggw/SVVddRdu3b6f2Tr8IX7q0fxgNiPClbkGeQliOverscWePOgAAAAAAAAAAy8bqDfrS0lKaPXs2ffHFF/Taa6/plhcVFdFXX31FS5YsoUmTJoll33zzDfXs2ZN27dpFI0aMoPYMe9c/umFgW+8GAAAAAAAAAIBWYvXJyg888ADNnDmTpkyZord8//79VFNTo7c8JiaGIiMjaefOnc2uj0Pzi4uL9V4AAAAAAAAAAIClYdUees6NP3DggAi5NyQzM5OcnZ3J19dXbznnz/N3zfHmm2/Syy+/rMr+AgAAAAAAAAAA1N4N+pSUFCGAt3btWnJ1dVVsvSys9/jjj+uF7rNXH556AAAAAAAAAADmQLI/G85ThsxqDXoOqc/OzqZBgwbpltXV1dGWLVvo448/pjVr1lB1dTUVFhbqeemzsrIoJCSk2fW6uLiIl+GJjIiIUO1YAAAAAAAAAAAAQ0pKSoTAu80Z9JMnT6ajR4/qLbvttttEnvwzzzwjDHAnJydav349XX311eL72NhYSk5OppEjR7Z4O2FhYSIawMvLy2LV33nSgY+X99Pb29vk9Q0dOtRoGgOwnDa6ENCe1tNWLQHtafltdCHYcnvaUjvZcnu2x3ayxvZEO1lHe6KdrKc9i63g3seeeTbm2R49F1Zr0LOB3adPH71lHh4e5O/vr1t+xx13iPD5Dh06iIZ66KGHhDF/IQr39vb2FB4eTtYAH6MSF6SDg4PFXtjWjlJtdCGgPa2nrVoC2tPy2+hCaA/taQvt1B7asz21kzW3J9rJOtoT7WQ97elt4fe+c3nmrd6gbwkffPCBMMjZQ8/q9dOmTaNFixa19W5ZReUAYDugPW0LtKdtgfa0LdCetgXa07ZAe9oWaM9G7BrOl2UPLB4OGeHZGxbws+QZpvYM2sh6QFtZPmgj6wDtZB2gnawDtJN1gHayHoptyH6y+jr0QCPkN2/ePD0xP2BZoI2sB7SV5YM2sg7QTtYB2sk6QDtZB2gn68HFhuwneOgBAAAAAAAAAAArBB56AAAAAAAAAADACoFBDwAAAAAAAAAAWCEw6AEAAAAAAAAAACsEBj0AAAAAAAAAAGCFwKAHAAAAAAAAAACsEBj0AAAAAAAAAACAFQKDHgAAAAAAAAAAsEJg0AMAAAAAAAAAAFYIDHoAAAAAAAAAAMAKgUEPAAAAAAAAAABYITDoAQAAAAAAAAAAKwQGPQAAAAAAAAAAYIXAoAcAAAAAAAAAAKwQGPQAAAAAAAAAAIAVAoMeAAAAAAAAAACwQmDQAwAAAAAAAAAAVohjW++ApVNfX0/p6enk5eVFdnZ2bb07AAAAAAAAAABsnIaGBiopKaGwsDCyt2/eDw+D/jywMR8REaF0+wAAAAAAAAAAAOckJSWFwsPDm/0eBv15YM+8dCK9vb3P93MAAAAAAAAAAMAkiouLhWNZskdtxqD/5JNP6J133qHMzEzq378/ffTRRzRs2DCjv/3222/ptttu01vm4uJClZWVLd6eFGbPxjwMegAAAAAAAAAA5uJ8ad9WJYq3dOlSevzxx2nevHl04MABYdBPmzaNsrOzm/0/bIRnZGToXklJSWbdZ2uhorpO5GmoSWVNnarrxzYu7DzV1avb3kxtXT1V1arb7tW19WI7amOO65fhfqj2turr1d+GOc4XX8NqX1+2cN/i9at9f5fgvsh90lb6Iz8b1cYWrjFb2Ya5+kpNXb142cI90lzHYo7nljnOlznukeiL7Q+rMujff/99uuuuu4TXvVevXrR48WJyd3enr7/++pwzGiEhIbpXcHCwWffZkjmaWkTPLjtCm0/nUL+X19Cry08qvo1DKYX04t/HaFtcLsW8sJre+y+W1CA+u5TGvb1RbGPx5jOqbONYWhENfnWt2MYPOxNV2caB5ALqO2+N2Mbv+1NVuck/tvSQWP/NX+0mNSiprKG5fxyh7fG5dNEHW+jiD7cqPnlQVF5Dz/15lPYm5tOo+Rvo8k+2k1rsOZtP0f9bKc7Zv4fTVRuovLnyJP11MI3u+n4fDX1tHeWWVim+jbdXn6K1J7Loxi930bDX14m2UhJe9ztrTtGv+1LE+fr7UBqpwc4zefT8X0dpwrsb6aL3t6gymNx4Kps6P7tCHAffv9Rg9bFM3Tb2J+Urvv7MokqatXiHWP9L/xwnteDr9uMNcWKgOvn9zTRj4VZxvakBP1P6v/yfOKZf96ao1ufn/nFUXM+9562mD9fFqXINv/LvCfrveKY4li+3JpBaz62Br2jO1897klXZxr7EfOr+/CpV+/yO+FxdX+FzpjRFFTV053f7xPofWHKA1IDv6c/8foR2nMmlsW9tpFmLd5JacF98d02seHZNfGeTKvdI7uN8DXd7bhVNX7BFlT7P+833rjXHM+m6z3fSmLc2UFlVraLbYOP6teUnxH1+5sKtYiyphlHPx8LPx67PraKpH2xWxamy4VSWrp/wPUYNVh7N0G3jcEqh4utPK6ygyz/eJtb/2grl7RJbxa7BXNP2JlJdXS2M999//52uuOIK3fI5c+ZQYWEh/f3330ZD7u+8807q2LGjUKsfNGgQvfHGG9S7d+8Lyl3w8fGhoqIimwu55w5pSOL8mYpuo8vcFWR4z1J6G3wJX/bxdjqaVqTaNoydL2vcxvtrT9PC9Y2D07NvzlC8esMLfx2jH3bpR8IcnjeVfNycFNvGQz8fbGJcJ7wxg+zt7ayy3dlwYENezuKbBtP0PiGKbWPV0Qy67yf9gerimwbR9D6hVnVP4cHc2Lc3Un5ZtW7ZrrmTKcTHVdHtyI8l3M+Ntj0zSdH1G26jf7gP/f3gGEXXf+3inbQnMV/Va5cNBx6gMs/P7KkbgB15aSp5uyrX59vyPqzGdsyxDW4bnlQ9m1um2jbaok0c7e0o/o0Ziq7/6d8P06/7UlU9Br7H871ejhrbYfg5z897ifVPjKfoQE9Ft7H8SDo9uOSg7vP+56eQv6eLotvgsQSPKeT8++AY6hvuo9g2vtuRSPMMJjxXPjyWeoUpO+ZfsjuZ/vfnUd3nPc9NpiAv5Z5beaVVYjKiTBtV1C3Ik9Y+Pp7U7ItjugbQj3cOV3T9bMwfTlV3PG9NtNQOtRoPfW5uLtXV1TXxsPNnzqc3Ro8ePYT3no39H3/8URj1o0aNotTU5j2fVVVV4uTJX6D1mCGqm9Ycz9Iz5kHzGHoBq1QI+9qZ0HRWWMmZe57AMeYpr1XhYjOc73RxVOeWGZvZ9D6j5AQIk5hX3mSZt4LbkBvYavLtjkTVt2XonVG6LSSPoBwXJwfFtyE35tVCfl39caDRO6uWh15tzOHjMEeKkNQecmPeHOdLjeq+hteSqwp9ZYdK3kw5HA1piBr9hKMv5RP3aj27TmboP7cqVRhPHDHiAXZ0UPYiS8xr2kc8XBwUj478aIN+m9gr3Fk4OlUy5tV6bhk+e9XYhtyYBy3Hagz61jBy5Ei65ZZbaMCAATR+/Hj6448/KDAwkD777LNm/8+bb74pZkKkl62WrCuvbhqy5KbwQ1LpkGFjcMjS+2vVCeM3dz5SYbn6BlFeafU5jQpltlGlqrHd3M1ejfC17JIq1R9exrajxvEk5JQ2WabkwPh0VkmTZUqfL75ePzOSUlNbr+xA8kx2mertHmdwvtSaLFIbebufkA3w1ZhgK1Y4RaSlfbF7sLKezaT8ppNrakxKfWhg1KlBrsEzRY2+klKgf75cnexVfzaqMbFjLG+6TuHtiLS6Xw816X8K3yIFuSX650zpUHgmz8jkrdLPxrSCCtW3wd75jKJK1baRVVxJ3+9MUr0vxmaWqOYUMCcn0otFhMkpI84Ua8VqRhABAQHk4OBAWVn64Ur8mXPjW4KTkxMNHDiQ4uPjm/3N3LlzRViD9OJydbZIQk7TGclO/u6KbuNMdlMDQml2J+TR6azSNjlfSnPGiMGlNAUGkwbFKhj0BeVN16nkg2udQciiWkYdE2dwball0Bvzoil9PPFGri8lvUOsY2GI0ueLc7WLK2tVH3zFZZeo3u6G58vFUdkJVbWFnc7V7mpNsJnjmWLseDp4OKu+DaVZfzJb5KKqOWlvrK+okWahdl9hKgwm7StrlL3/NjdBoHQ/4agMPl+BXi6qPx+T8stUN+izS5pWpVJ6stDYs1HpdvnOiO6SksexdG9Kk4hLdZ5bhga9ssXS1LiGjPH34TSRLvKbLM3G2rEag97Z2ZkGDx5M69ev1y3jEHr+zJ74lsAh+0ePHqXQ0OZzRrmsnVSizpZL1RnzdNQrPFNsbOCtNDlab7AaM/Zy0g0GRmrcKIsr9M8X5wkqTU1dg+oeLycj4XBKDibytZMSnQ0moNQwIMzR7s1NrCh9PMa2oeSAosRIf1f6fOWUGO/vSg/w0gsrVW93w/Ol9D2srKqpQa+G17E5YUU1PPTGrjFr7IvGjkPp272xvqLGddwWfcVF4b5irF+UGYliNIXm0tuU7ifsqWWm9Q5W/floOGYpV6EiRKnRCdx6VY9DjXaR+qOcOoPxmBLrl99H1PCepxn0d6Un8DKKmkZLqEGVdsJObdvBnFjVkXDJui+++IK+++47OnnyJN13331UVlamqzXP4fXsYZd45ZVX6L///qOEhARR5u6mm24SZetYKK+9YyyEXOkbmOGMt5olhTr6uumWqdFBM7QPSTUHLYbnS408QcMSTMYeZKbAHl/DSQOlBxOV2mOYGBOkugFhGCKnlkFvrK8ofTzGPE5KtouxY1D6fEnbkPd3NQar0qBYwtNFWS+EsfOl9ISqsfZQ2uvY3HaUHqw2ty1nB3ur7Itt1VfUuH9ltcWzUWEPvbFnVrmRCTE1yh8q3U+ksZ3hOVLj+Wg4jlTDu2q0P6p0ztR6pvCEkfH7inL3Y2n97s6Nzyp3Z+XHkJkGBrfSmhmGYy61qGymn1gzVmXQX3fddfTuu+/Siy++KPLiDx06RKtXr9YJ5SUnJ4ta8xIFBQWizF3Pnj1pxowZQuBux44douRde0ftwb28w3QJ9FCt80vb6OjnruogL8vgJuPlqvwA3/ChonROrahvbhCGq7SH3hxeCOkYogI8qEuAh6oeiExDw06Fdjd7f5SdMyXbpcrI+pXuJ9IxGBrYSg/wDL0EasikNR0M16k+SFXa6yi/duXtrlaIr9r3SHP1RWN9RWlvmnSu5KHXSofHGusramzDsN2V7o+Gz0U1+oqxbajRT3SGioFDQI3no2G7qOGhb6uxqpLPxuq6ejI2X6uoo0N7DG4yI15pfQZjYyKlJ6S41Ko5qGymn1gzVmXQMw8++KDwsrMa/e7du2n48MZyCZs2bRKl6iQ++OAD3W9ZCX/FihUihx4046GvU2fQ0j3IS6/8k5JIqqoBno05jpEKawEYu4mp4YUwfHApfTM29lBRWhRPfl3JQ++VfXBpQ6UcHeiqQR1V9UAYep/UEsA2R8SMtI3rh0WoEroorX9CjyDVVHylth/TLUCvzys9wMssrjLDYLj+vEKlSl9TSnsd5duZ0TdUz3hU45xJYZLW/myUtjG4k59qefrS9dUjuPH5K/feKUVmkX5fsSM71ftKhTn6iuLbaDwG+YSk8sZpYyjxu7P6q+uhN5jAV2fCUN1nI1eckNZ347BIlZ6Njeuad2kvdRwd2m3cPKKTqlFSWQbPRrVSRtSOiqyU+okKUQxthdUZ9EDZi3lij0D658HRqj9Yfrl7hHjvpLD3XD7L9uUtQ1QbUEg3mbHdAlQJj5UfC9ejVmPwLX+ozOgboooonuSFYGP+6EvTdAN8JQfE0nniPMoHJ3UjB23SmBoPL2m2mGutim2oVIpLOqYVD4/RXWNK5wlKg6/L+nekXqHeKrSLZv3+ns708mW9NetX/Bg05ynA00XUhZfCiZXejhSRM7qrvyrrlx9Ln46atihVqb9HdHAjb22kRKkKIbHSdjr6udHuuZMb+7wqBoTmHPXT3iPZgFBaF0DaxmX9w+jHO4ar+mz0cHGkxTcNUnTdjduo023jnWv6qfbcMnw2qhmZIbW70n1Fnk8brfXSKt8fNevz93Cmfc9PIWdtdIniE7e1jWOiawaHU6iPqyrPE/kx9e2o7Y8K31+4b0tRf9ufnaRrf0UdBLJJidnDOzU+UxR8NkrOLZ7fvnVUZ/Jzd1L8OCQRVHaa3TkmSpVri9sj0+DZqPxkumb9E3oEqpY2oNdPrLS6jDFs50hA60JXXZ1kDxb1Qr8kQ155T0ejt1atB6Q8r6ezv4cqxyG/wQR7u+o86sbK3CjxUJEeWkoLF8rbg9tdEvZTI7RMCpXy0N7wa1QYsEgPFw7vV2uwKh9U+Lo76yYolLzGaurqdW3Ag1bJQ6RGKgSHQUs6FmpF/XDba+4ryl9fvC5JbFNqdzXDVcN83FTxCErnipXN/bTeX6W3IbajMyDsyd7eTtfnVblHao8pRHuP5GZpLs2n9duo14WOS9eX0vcW+aSk9NxSPKxfZthJ90q+D6h1j5Sejar0lVr9dlcrmoX7ihQpUa6wcSp/bsmfjWpFf0htLhlExnQCTE7h025LmjRQOm1I3rfZUys9G5W8juURAPzsUuXZKBsX2dnZ6SJllN1GY7u7a49BeWHdWl2uvvRsVPreKE0YRGjTaJW+biUQcg9sBvnslO7BouJMsRqGnX7eUONxqDETLXns2AulnsBMo4dTQsmBi/yhIoUxKe6h1w1UNYMInXGqQviaNGBxdFBvQJyvrYErtbsag1UWEpQmbuT9UY1JELENJwdZu6gzoHCwV8n7ZKBMq8Zx5JZWiXPP65YMCDWM0yqD/q54Dr3s/uuhHUCWqZLjqi8upEafb9xWfZPwdKW9grpJI0cHclRhwkj/+SvrK2pNdjupc0+RDCvuL2o/Gw37CudqK1lyU/5McVeprzR/71JnQkr3fNReX8q3fYOYUNNvF/VSIdR+NvLEGk9IqvJsrNXPb2+8ryg/LuJJKbXG9NLkna+7k4j8USWHXruNxjGXSk6UGv1+YgvAQ99OkV/M0oBCebXVep0XQq2HV5XewEidmxgPGEu0g0ZJA0Cd/FBtiKSzo85ro+SgonGAb68TYFIrh14atKjyADYIlVLDo81ka3PFuC04xFut2WK5F0I9Y7tezwuh5oDCHBN4kvGoxmBVisYJ8nJRzXsqPxadR1ClnF0+Vx4uDqp4HY0bEOq0vWZbjQNjqaa60kJcRiemVDO2zdBXVLqnMNklVUJXhCMZuL+o3e7+Hi6qVNHRm/yS+opKonhq9xPDZ7BabS8X+ZPaRa1JED5XPHGv5rNRGk+o82xUf8wij/rRpSEq7T3XGts80a3apIFWkyNcxQlCY/3EFrCdIwGtDF1V0UNvDk+HfGCkkqdWuolxKJbk2VazBIzGo6b8AFy+fql2qNIq9815z5VVU9ffhpNKAyP5w0uN0O7mvOdqXMe6h72jvQj5UyPcU/6AlPq7Wjn0jdeX8vcuKeSPU1+kCQM188E7aAfDynsdGwd4ankdDbejpgGh2VZ9E8NLcTVyM3i29fqKap7axuevlPKm+D1SN/nF90h1Jj/kx8KeQanOtpLtLr8/ekh9RfE8fcN7l1o59I0TeZrtqGPcVcpS+Lhd1E5TYNS4jg23oY6xbThmUeE4ZM9GabyiuGCdkWejksfAkYp5ZdqIH21aqFq6RVVSP4GHHlg78puYWYxttScN5AMjlcpohPi4qjbzaTiQVGMALm8PqbSQ0nXo5VEAjDly6B1UMh4zjLS7GmHEkqeJH8K8HTX6ijyfltFF5aggLiQm8FQzhBqjfhg10myy5F4IVSdytGHEsvBxRb2ORoxfNXLom3oeVYxqkEVk6e6RKqYqqObd1JuIVr+vqBXFpOsrPo0eOzX7CqdyeWjbXUnRWP2+ol2/ihNFjFrny3DiQK22r9KLAFIpTcFgPOFghmejmmOWpqmI6uTpq/F8l4+JWDNBjXFKdkmliPjh0tNqiqsyFdprFXXogY2F3KvtTbNXMay/8UbZaGyrZNCbyWOnNwCvUscLocuhV7oOfbP5tOp4gtU0IKTZaHl4mZoeesNzpk44vP5gUo069Orm0Ks/WNVFZqg9kaMdUPi4Owkvl9L54NLkgDwvWE2VezX7vG5b0iBMTHpqPfQqeQX5maKWLotcSFC1vmIGDRv5s9Eck54csuyufTaWqtJXGq8rpftK81oT6obcq9X2xs6Z8n1RO5GjS7FS/9loljGLCpN4xqN+1IlaZA+95EhRYzI9yNtFGPWa9asUcl+LkHtgI0g3Y5FPqx1QcL9RNORTN/hSz0MvH7SqFeardxMzg2HHA0l1PPRNQ+6VzqGXtwfTGDVRr5iyruE21BoY6Rt2Kk7kyLxP8nOmZL5+c1ENSl7H8gGe2h56wxx6JdtFN5Ej90KoWNWCc8E91OzvIoxYmiCsU9XIVjPEt+mkp8qeVEd1cnYNvUPqqZ2rP2mv56FXKcy3aTqa1O7qPBt115VKZeuaeoKVbRNzPR+NtYnSHnpdP1HR2G7chnoRRo0TUvptooZav15/V7iqhbGIHGXT3aqaTBByOyhdmpSByj2wGYx1fqZOwY4jH7CqZQjLB19qTRo03sRcVC7J1HgsUukUJWe8dSHRMlE8VrlX8map9my3XFm30ahTN4deky+m4kROk7BCNfIEDQ1htVXu1QrzVH+Ap0u18FZ3IqfKWD64gv1dHjHRGBKrfsi9Wm1veH/RHZOKhpda0T/mCOtvjJaSi2CqE4Kr11dUmfySqdCroJ0gjyzTeZsVF8XTTxdSq580jcZS6RqWVwZQSXTTHOk8zd6/VNH9USdqgtcjOQDMF5GjfFvoxlxiwqBR4k3p42gQJRf1+6MtYDtHAlr9gJQ6v2pGhDmEf1RU69cftJgh9NpJnUGFMQ89H4YaXsFGb7OyN325sq7aIlxGw0lVFHxqEtWgYsif2mH9qnnomxHFUyPsT/WJHFmbeKjhdZRN4KnldTQcSDLmOWcy4VCVDC/xTFHJ6ywfTKrlqTVHGUn9Abh67S6fyHNXNYdenb6o2YZ5tCYaU+vM46GXCwkqfc7MoQfQfJWOequZNNAX1lUvssiYk0NZD31F45hLe+9VehvGKgvZCjDo2ynywaTcQ6+GESHPb1c6rF+uVKm+h95N3bBC2bGoMQCXeyG43SUlVCVr0audTytX1uWBBKPWgFguhqjuYNV4VIMaAwrDsH5V+rteCcl6VYxHN5UGeDxzLw1aWPhH3ZrqMq+gCl5HtScI5dvQbMccKveN15hayv3GKsBYY0k56XnCJf7UOo4sc/cV2USOan1FhWgZc2lN8P3LUL3bHBNG5qg4obrIrmFJORWFfBV3dMjvwypVlWLdj/yy6ib9XUkHWqa2VLBcdE9sQ/HopXrde4jiAatHGrTwwFiv46iRtyvyBGXhM4qGeKsf1q8viqdefqi+F0KNQUvjw5FLl0nCeErm0TfWiVYnn1aurMvHoNasPU86seKqOcTRmtZWV08Bt9EQVjaaRT6QFAaEiqI/anqDS6pqdV4m1ZW75WHEOsV2dYwUDxXWL9+GMV0DNdXONVEH6oriGT5T1EhN0jx/1fXUalSv1TEejT8b1UxHk6WPKJqOpn56SvPPRuUdHJrtqOyhNxIBpGZfNJuxrcqzS3+SRXFHh7YtnB3tyd7eTpX0l2ytsc3b4DKFaoxT5GXx1HI0yvs7b0NybNkC8NC3U6q1HZ3reup3HOUMFsmb5uxop7tJKn2jlERFnHgbKhldJZWahxTfxNTMqZWEU9z0BJ/qlF+/9kGvq0WvoEGvaw+tQqnSxna1bv2N15MaA2IeGEnXr6+bk6rlyyQPo7u2zXWCdQo+jA3bRemHsVzAT3NPUb6fyAWEpPZXemAk9QWO/lDTe8r1dqVrSSOKp7xoXYWRHFelvdlSu/P1xANJ8d4M4mh8ztQqWycdE1/HUr1o1Z5b3FdUOl/Gno2K1oyuq9cZkFypQa266vrPLnXCuyuqZdFxKoT0n/PZqHCbGN4j1ajdLhfDlPdF7jt8b1PtuaVCipW8v4ttqGAMS8fhrN1/pR1DNZJBb/B8V7LNpYpIPB5iZ4oakYS6bfBYWyr9osK1W609X9xHJMeQLQCDvp0idUK+Qep3zgYqKq+hiz/cSp9sjDdxG9Jgr2lYf05JFU37YAt9ve2sSduoNbINKaw/Jb+cLnp/M/28J1mZ43BoOjCKzy6lye9tor8OppGpSANTfjh6yDxqx9KKaNJ7m2j1sQzT1q/1OEgPXy9d6TrN8vTCCnG+ftyV1OptSOfF0HCUlj/z+xGa/eWuVqddGK7f2MDooZ8P0p3f7TPJoyafFOJ2N2agvrb8BF21aLuuBFVrKdE+xLxcHY0a259tPkPTF2zRhbu1BmldhoawdD4XrDtNMxdubXWpJvkDl7dhbEDx5qqTdMUn2/W8uq05Bvn+G07mvPDXMbr2s52tVvdtcv0aTORsOZ1Dk97dRPsS88kU5HnfbGzLvYJrjmeKbRxNLTJtG9q25MkCuSI8X28zPtxKH66LI1OR+ol8wtZwsJqQU0pT3t9My/anmrw96R7Gx+MpCXFV19KJ9GJxj1xxxLR7pOExGeZxskea75Hf7Ug0aRuN11nTvpKUVybO16/7UhR/Nkr94nRWiXhu/Xs43eRjEMdh3zQ94UhqoWiTdSeyyFTKZc8ueUj8nrP5oq9sjctRZP3cV+TRcWyIzVq8g+b9fczkY2j+2VgvnlX3/bif7v1hv0nPLfkksDQZZTip/tI/x8UxmaqwrhtPuDjqzpl0Lp/67TDd9OVukw0x3TXczCTIu2ti6dKPtpmkoyHdp6Rno+F1/OryE3T1pztMmqg433HM/eMI3fD5rlafr8b1G050a/Z546ls0U8OJBe0+hiaezZK2151NEP095MZxSZvg+9ZPEEsmQ3ScfxxIFXct/iZYgp10rHIJmxtAds6GtCqh73hDWZ/cr7olH8fMs1IlTohr9swrH/h+jiKzSqhV5afMGkb0gOM128Y1v/aihMUl11Kc/84qthxGN6In1l2hM7klNGjSw+RqVToBquyOvTVdbQlLocScspoxdFME9evMaQkb6C31oCUvJI7zuSJ87X8SOsHedJ5kc6T/HyxEf/r/hTaHp9HGdrcywtev7a95RNEcqOLj5EHqetOZlG6NnyrNcgfrDxTbMxA/eNgGh1ILqTTmaY9XIoravXaw1A99o8DaXQqs4QOmvIw1g7eGg1h/et4wbo4Op5eTD/vTjZ5AsRYP2E+25xAh1IKaVUrJ6YaH/aNs+qG2/lhV5IY5O9KaJ3BLa1HurwMQxdv+XoPJeSW0U1f7SZTkLyL7FHhAZKHzOvIBj1vY8OpbEW2wQNu3fqr6sREwYmMYlp2wHQDWz4AkzBsk//9eVRMfD7x22GTt6c7Jmf9HPpt8Zp7pCkGqrFjMszj3JmQa/I9srlnihTW/+Lfx8X5evr3I61eP99rJbtQ/myUju2xpYfEc4snP1t/DI19m1dv2O53fb9PtMmd3+8jU2BDSvKiesgmu/laWH8qS/QV7jOK9RWZfk1ibhntTSygX/eZ3lfO9WzktLdVxzJp9fFMk6JopG3w7VEXMWPw7Pp2R6I4JtMnQRrHE3wP41Bshvef7y3b4nMprUAjctZa5Pd8zV/96/i3/Sl0NK3IJCPyXO3CfLXtLO1PKqD1J7OUOw6DieKf96TQzoQ8OpRSoNB50l//bd/uFf3k1q/3KHie9Nvivp8OiP5uyj1FWldzE/bLj2SI+9bWuNxWb0PvWGwo3J6BQd9Oaew41OQGIIWYmxpyJO+chmH9UgidqchvMvLOydtWIiSPB0bSuMWYYadUzhgP5MplIfHSYJU9plI7SJ5cUyMA3LTrNsyhl9ZvSrvr2tyu6Q25tLpWN8g01YMqv57kD2C5PoNpHvqG8xqokofD1HaR/r+UAmFobEvfm1KXvsnDuJnQRXnIptITIBKtPQ6dsS2fzGkm7K+1Oh2N0TjnDl2UciJbi+RRktJf5F5Hqf+VVtUo49U0EMWr0R5La6MxzjUBYmwQplRotLhH6gx6NrwaS2VJ11RBeeujWJr2Ff37TK3s2WhKXzS8jxlOGiih2q9vbDc+G2sVfG7JvcHyyQ/p2Eq158pUpIlo3bPRSF+R2sX0CAB90T2pnXm8YqpH+1zPRvn+m1I73HAb5wrtN3V8J4+aYKTzxu0ubUoKoTb9/mLc2FaiPzYxIpt5NpqSGiGda+nZ1VyOe2ubXpogbO48SZhyz29qbBvfhin3lmYnV6R+qL0XmDrmqjPST2wBGPTtlEZPp76Hnh9akse2xsTcmMYoAP2wfl6ulL5QcwMjpfLS5IaBPHRRDZEkaXUe8sFqNQ9a6hUZIFXUNEYAMLpa9Nqbo+Qplgb8pnqe5H+5PeS5+q036vTXr/dgqas3aTDU3HXF164xw043mDRxcCwNSqSQe8PcNCklwhRtiOYexqYaJuebADF1ENxcNI58W4bbNwXDCBC1xBDl3i3N30avo7Qt040UyeuoXzVD6iM8KDL1HmY4AcI0tr1y90dpsknaHh9To4e+0bBTwqCvkz0b5QM+Pm/SPczU60FvIlqN55ZsPRoPvfL52nqTBnbywbfCfUX73OKQaPYCe8j7ikLPRr3ID21f4cOTG0CmToI0Z6xwP5EL05o2cWvk+eigUrtIOfTSpKS2XeTHYuqkYXOeZz4WfrZI+2DK5MS52kXxFFcjHnQl8sPPF8mg6rNRyWd8c+0tTQ5rnV6mjrlqjfQTW0DTA1vAwoULW7zShx9+uLX7A8zEuUJ0dAaEgl4I6S8vU3JQIRncvP/yzqmUoJie51E8GBsHrkoeh1xR11DwSTKwTR3g6zz0ToaieLUGHnpTDEdq5rqq1/dCtHJAXC9rb2NGndzDbIqtYuzaFduQrV8pw0uaUPEy4qEXkQ3ah5cSgxbHZsLllJ8AUWFAcY7BKvd3JbZleH2ppdzdGC2j76HntpYGkooZ9CJMWT8vmOHtsKiZKXV4jUXMqFUJRC5SxlEHjSr3jZMgBeWmC3zK+4qUx8mLlIpe0w+HZ8+28sJ7zaXASBPRSmxF3leM9fkGpQ1H7XUqTU5xX2ExPqX7Cl9bEnLjlLfh6+6siG6RGs9GzbrO/XxU0glhmMLnqW0Xw3OmiBFpRARVPoljivOp3gzGsDRONeboUGLSW3cfVkks1lzPxibjLoNoCUmI0dTrqt5gnNruDPoPPvigRb/jGzsMesun6Q3GXrWwQukGyZ2nWjv4VmJIwQ+n5j30ynpqjXV+zeBI2YcjD1p4EKlTvZZ56E0NM9I9gLUPXm83bQ69dr1KDFalG6+xB4t+WGGDcjn0sgewfL2mPCjPN+PN116j4WVqyH2tXnvIB99yD4cpxyMNWqR+rrTHrrkHsZoDCjUGRufLp1SKxmgZKVy1UbROOg5Tw1X1hL5kXkdpwlbahhIG/bmiJpS6R0reGfbSckSA3EMvTeQVlFWLvmmKcrHhs5H7Pq9fRBnpno0m3FtkJ0QtD7388Sf30Cv53GoujUfxdtdO5DT2labPRqX6Cnvo+fnLf9nIL5RFfJi6jWYnicWzsXHdiqS8Nfd8VNQJoZ/CJ01Kys+ZyaHR5zAilRhPiP9r8GxUYwKkeQea5r6i3PrVK4lojmejobif4ThFStU1eaKoXn+c2u4M+rNnTVMjB5aF4Q1GUvjU3CgVDitsRnlTWWNbU1tdigJQztOhP/hq7jtTaVSg1/dCCA+9Qh47aRuGZeukGXVpwGLKbLehJ1gKxeVJHP2Qe9NyteVeLbnaqvzBblLOWzN5ddJy+bpNbRedyr2Lk56KrGGaghLH01SUp16d+4nB7L18YGSn0MDLcLCqjKdDPzRSrZJPhtEy8v6uy0E2MaywccDtoOd1lF9T7OUK8iLF+olh1ISSSKr97gYeQU0aQWO/5HBM6d6mxLUsjq1Oil6Tno2mG13NGdtqeOjlKRGKTeLVmSeapanehPLPRnnIveavo1impLe56ZjLeA69IhPRzVSBkT8fTS3XpRPxNUgbUvacGYi5yp+NepMgSk7eNz4b5U4tU06XYYqrfDwhlZwzhSbjlWbG2iZNdJrh2dg0NUFfa6DRoFdmcs3RxlTuW2zQA9tCugE2inRo/j71+xHdQ02pkHvDGyWr90oPZ1MwVNllJIM+Ob/cZCVMY2JfdrL+z6V/WKlfCeT5rowUTppWWCFUaRkWlWNPq1wUrHUhcvqieNvjc4USdWJeuWIeAkmcRWrz99aepqsGdjR5UGlMGE26drm8jKuTvUIDI+Oq8NLxLdp4RvdbUw2v4mZy6LkEV4BHY4inEkaE4fGwevPIaH8T9l5/36ScY3mUwYt/H6Onp8eYvg0jOfTyGXxFyrA1ES9qXP/uhDxSCsNoGWlQvPl0jiKDYb6/S2WWPJwd9byOr604qcg2jPUT+Xvufx9viBMq1MrqDmg9gtpzlllcKZSoJQrLalpt0BtGfcmvMa5oIhkrSngEpW3Iw/p5G4UKpA00pwHCsII+Px9NRSfEZaSvcKlVpYRvDdtd6itcqYFfStyD5SH3YhsuDpRbSnp9xdQ8/eY8wVzVYEbfkCa/M2Ub0vPX0EBduKHxHmmndNqQ9q/eOVNKd8DgubJkd7LeJLEauj9fbD1LvcK8SR0RanvdPVLeJlIUmmKCi3UNouqLuvoyDbQpNluZbRhcv9I2uMw1l/bjv0r2RXvbctC33qBPTU2lf/75h5KTk6m6Wl+I5v3331di34CKNGdsy8t/mGIQGRsYSQ+yfUmtL791Lg+9PKz/msU7FdmGdLOXysDYye63Vy3aQUohhRW6O0mD1aZdk+/17GWXcq1NjQKQRPF4MMMPSCVmu5vmatvplXlTzkPfdMDCPPdnY71gZXPOGx9eXCf6g3Wndb9VTOVe2x7y4/luZ5Ki7WLYF5nHljaWE2vtBL5hqoV85vv7nUkU5usm24adchUOtNvjmtdcdlGpgZExr+N1n+8ipShrxusox5TrSgpP19uG1uuovw2FclyNhPj+vDdFNwBT45xJkyGG5JdXU6S/u2LPFOkak19fJuU5y+5LhmH9axWo2a7ZP+ODb+aSj7Ypso3m+oqS2ziX+JocNhxNSbWQh9w3t40SE6tOnOt8rZSVpFUyVUze9iczSkQpOQlT/arSPcZDNwnS9JyZnKZgkEMvPy4u9aamh77Js9GEKZDmJnN4Ql1+j2x1KmIz/Z2jtK79rHEsbIr92txEJy+/9Zu9im7DMHXg4V8O6j2rlIuWsSdq7wb9+vXr6bLLLqMuXbrQqVOnqE+fPpSYmChuqIMGDVJ+L4HinOuBL/8N19vdcDJbGBuX9AulIZ07tGj98gnTxlnD5rv6ltM5ogbrrMHhNDDSr1W5iM0dh8SGU1mitvP1QyOpT0efVt3s5WH9xlh9LEPUWb95ZCfqHux14QMKyUNvZEAh3cj4GLgu6q2jOlOXQM9WhxVKdc8N4dnupXuTaffZfOHFv3pQeIvPly5X26DslyHH04rpzwNp5ORoT/3Dfei6oZGtN+qauSnzoJsjHBZtjKcBEb40a0hEi7ZhzFCR3/i5Jryhh509dwvWnaaOvm50x5ioCxpYNuehNySzqJKe+/OoMFY7+XvQbaM7k4ujg0kqu81di++uOU2BXi5059goXWjdBU2AGOSmJWmjP+TG6vtrNefr9tFRLYo6MRrerX1/PL24Sfvll1XTwvVxFBXgQbeM7NSiNmnMnzbIp2xmwPjDzkQR2XLPuC4U5O1KrVe5b9qO0qDly60JlFVcSfeMj6YAT5cLigDg/XfR1oeWvI7629Ao3X+4Po5SCyrI39OZbh7RicL93FuV4yptk2nOmOea4VwD+6YRnSgmxLvVIlzSfcwQzqNnA+/dNbHi/3T0c6Pbx0TpQvQvNOrLWF/h6+vXfSm060yeEGa7kHuk3nNL7olqxqHN54ufj3y+eoZ6t8pjd77yTBwNtCshj+aM6kRdW5iD0VxfaY4/D6bSoeRC0RZ8/7rQiRxdX9E+I/X2RVumlscrbLjyfavF13B9g64Mpft5+uOJ9GL6aXcSTegRRBf1CiYlon8M4Wf7N9sTRbTZ0M4d6PIBHVu+jXPcIw1rtfNxc777gnVx1MnfXYwnLuS5ZZgC497MOcsuqaRPNsRTt2AvcQ0r5SCQw/dg9uCyORkd6NHiZ4peVKGRyW45rPtUVM7PrViK9OdttPx8NacFYHiP5HTH3NIq+nhDPEUHedJNwyNbtA2dsJ808aH9K09/kJNRVEGfbjpDfcJ86NqhESbZDFXNROJ8tvmMeAbfOz6a/GSRhq1J0zU04KXnFp8nnli6b0L0BUVlGRtHtluDfu7cufTkk0/Syy+/TF5eXrRs2TIKCgqi2bNn0/Tp05XfS6A4hjNUxi5s/s3cZUd1JSI4rGbTUxMvPIdP2zmbE8Ljjnnvj/vFA3lHfG6Lt6FfB/f8hsojPx8Sx7LzTB6tf2JCqz1Q5zLo7/vpgPCkH08voj/uH02tNbabS0ngG/RjSw+JCZPYzBJaes/IVtVwlnuEDeFwpv/9eUx3jIdTClt8LOcz7CTY0JKuqyW7SQxcWjI5Ybh+w/dyOP/to/Vx9MveFPppdzKN7RZIIT6uresfsuM4m1v2f/auAlyu6uqevHhe3N1D3IgRCAmQ4K6FIsW9uLb9cSlWirVYC8WKa0txh0AggoQkJCEhTtxd/m+dufvOnjPn3rkzc8+8N+/t9X3vy8vLy1w5tmXttVN+F4fNKxPmaSMM6NWqvtq1a9NI19m4ZZtPjfZV7gPe2TPj5qQcbB0a11H7920VWwCP8NzXc9U/P09QmEE5HLVTsxwCIFWsBgfNxdcmLfDf18D2DdWgDo1zzD6VWA0XOOAvfDPXL1cZ2qlxJGcoTWvA+3wuJEeAMfZ/r03W3+O3/3RQL5VrD2dbRhDrFXONU1j/eGC0a1CLLdTOk0FozzpuVZPmrtRGPXeIbz+qf5aGEdc1CDeSznt6gn7P381bpV4/f4TKRRMg6HkAGJFvfr/QH3ugYZ3q6qThHXPK0Nv2ejiZf3j5e3++TPhlhXot4rMke0YnnYggejqytWc9Od53yHLdhzmt3wTW4+//PUH/GyjsL52za7RrpNXQBwf+8ByU7fx19Sb14ImDVLaBnLC1AixctVFd+dL3+vtl6zar+44bGO3z2bunLLOdMbNVn1tIPOA8mXz9voEskUiZ4IB9/o63pvlnI66zW9emkQN5ppI+X5t4J+Y8fHH8PH+d4AyOGpRKLRMMztDjnf37q7k+0wyB9WyuYe4vQXvLY5/NSmllhnNr1y5N89KXsd3LY1/M8p9lcIdGqn+7htGeI4CunnYv25BMSZ5bwzs3jhRgy+Z8B54c+4tmzgEjujVNYdBFzp57fy5YtTHtd2cuWatu/d9U/T0Cqb8f3S3j56deI9yex7xCKRdKOSkAd/5e0a6Rcp0KJoqXE99gypQp6qSTTtLfV6tWTW3YsEHVrVtX3XDDDeq2226L+x4FDmAq+S5da+/fyzdJLNyoqp82dXgc5DagDp0Oh2womrSJER0+rG3Rpq3b/Gcxs4XRlLUzG6yIdtPrQTY1G5jONpScbUDdPr3aiXNWRqbnmT2ceQ29zcDh4wdDKV9algmzj2jUa9gz9PZrwMnjOgqotc1d6Tp5DTPTgWjx9MXJ1Cen5mYCOeiYw/U8gygoQ2BGqW0HafTWPMEHJTQVCItWbchpnZhjAqYEvwZfHwtW5j72gcGc7TtS9C2irsc0wyvkwIfRQliYxdyyBfBsWUdg4pwVOc0rU5NDX8MSJMSehYwgRzbr3RZgCzMmQcGl/zN/RbS5lS7CFc5kQS961IlzRJ5jFtaX7UxBRpdn87NZi9nQPREoIASd0WHX4PM3qMwYwSr6t58M9lGka0RwIsD+ICADmU+pWBDT4vv5K/3vwTaIaqtQphl7MLFZ6lrWI86TnxnFxVw3ubYuCzsb8V+ysSXIHLCJVNqcR9gTuax7W5lCqS1gqM/G5DWQTMkGZmabmH8Z7YmI6z2beYxAJz8bkeWOiq2Rz98tKZ8b3S6KxsCj++BzKlfbK2y9T+fzKuLZGFamawJ6Ukho5XI2BjFZKm2GvrS01K+bb9WqlZo5c6bq3bu3/vvSpfkLkQncIrUPbhU/q5EJyCQiS5ItdTHTotnvr5/634M+g0UdZaHZMnZBOOrvrI6oSn4LP+jeBtzwrv/95ixroczDMQgXPjuJXWO7Fh+KUqJg9nDmFO9MQGQ/an2ieXBFrTszswfBn5+egQiqZ6XMFmH5uuiGZNjh9ca3C1J+F4EVfBG+mpWNQ59wFup6wmXZ9EbN5nmCavhMgIaX27gY78uYK/NWJINoN/znR9WteZKNEWXvyTaY8+BHM33BrHyeI2wsjmU19cuzcLZSNDMy1INf8nyyhhPPA6ciKBAXFiDkP+O45c1EFgXo3qKeDoIsy8ZxtDCYuCCXiYFsj+TOcy4iXEH452ez0hzsqGvFxvqKgmza5dlYX0Hgug0IVERFNmfjif/4KsUxijrHslkre975UU7PkZqhT4w7Fz7l4PXOSAyAho1ym2xE92j8algcx4c/+Tnl7xAbGxmBvRR0NhLNPxOi7o+Ja6Q6dmHzDM/D98hszpOoQcmPpiVFPgGU8Z0xsnPO+0v0s3Fz7M72dW/8mPL3qGeKTTw0qBwA+31nNmejPkc2Gfpb/zdFLWWfm+35m4nJgL337KcmJD8/4nmSTZkujg4IeBMmzFmh/ZOgRFimAEilztDvsssu6rPPEqInBxxwgLr00kvVzTffrE499VT9b4LyjdQ6wcwTGpkdOkSjLs5Uwyj6NMNC5X1Mo1EXMz8DV1oGFRtU51wNIzIwwgCjhdOMczHAowAHZC49nIFM9dd0sFAgJ5eDa0GGKPZOLer6BnFW9W5szKMyLrJxVNIUXdn1Mg3rt3NXRZ5fZv08kGlJ0jtbvm5Lziq7UUXjIq93w6AoCcnQA5zRsDyPYE7QgcwN1awcuiypi8nPz9Kh32KoakdY9xiyb2YvzzlAmCmjtGvXJlk7XLaAJ2fCQCOBgzN/4Dhm0/KIqNGlGd4Vd+azXSs21lcYyGHkPepd1W+C2RKVkWUqd4cBpQ8cJrsh6tkYtWY5iEWXKZBD1O6odctRu1KYYovAkggsgmwCt7azEeKqYUjO3ewDOTaV+8x7ZHbnySavVIz2MGI3hOHr2cuzsovM/SXT/KdAcVbOti+855VXBdSdm4hqs9ieI4yJ+jMr6Ytqs6QzI9PbFhIe+eRnNZnZw1HPxmxp/bme7/qzvTM+avINwbHv5iWTKpmvlb5OKq1DDxX7YcOG6e9RRz969Gj13HPPqY4dO6p//OMfcd+jIAMgAHTcw1+q3te8pcb/sjwnOnwY4Mw1KU3UcJ355Dfqe8MAyBg0yHLN3PzfKTpzmWmBZpOFMHHTf39Uy9Zu0vWj2QjZ6J9FOJBwb5e9+K2ud4/yvkxaYVT8+X9TtVAaavazEbCJAux1FPG88N8T1dRFqQZAFEd4tlFvboKEA699fbJ6f8qvGQ9sG8V3dgbDiICILhxtUCUzOdw2Y5UuGSQmyJkTg258Vz355S8Zlcr9HvRM0CWoNMV8Z/8eN0c9//Vcv146mwz93OXR6IKPItu5coNmgoQF2jJlHXk/XxMQZMM1YOyFqSLb6rWjqt2+M/lX9bt/jtMidmHvK0i8KBNgQGL+ItuJcckU9DPXY1DW0cRp//pG3fbW1IzOgJk9i+JIUc0pKJh3vTNNzY3Q2szmnP7MShGCdDqSWZZv1UXPToxkjNG4ZdPylNbKe1N+VY99PitjsDjbMwVGKJUy/P7fE9PKceKme0J4E8B+H2Yoh9H6baUXHGAgnfr411rUKux95RqYwH1f9dJ3+ivK+9qwJakHkQ2uevl7df0bk1No5VHEFqNStr+evUKv9w+nLY6ksp7UHEiMyayl4eurq+ecXvTcJC3AuHj1RjWD0det17DUBkdtHYc1cvJj49S/vpid8f9QwJDvMVGcaATxdrvtA127HyUxYmZSMwXvd2qZWO8PfjxTvTZpfqTAurnnz4qYILjznZ/0XMbelemcN9fKz0ui2Sxgs8H5R3CKv/NMfe75muxttN7DrfCxenXiAj2PoTkT9r6CWvlmAkrTzntmgmYGzAl5tza/xAw2huGoB8equ9/9SdsSmSA19AxQt+/Xr59Pv3/wwQfVd999p8XxOnTITsVSkD9wEIFyAsf7yL+PDXQgoZR77tPjU6KDUQ9jMrSh8H3w/Z8FHpIfTl2sLn3+W7XMi/pRH9xsgPZmI2//UB1y/+dakM0GGP9nPjE+8Bn26B5OhXvqyzlq0E3vqcMe+DylTpUD9VIIYARR2CiCHvgcE+arVybO1++L1xSZyvsQifrVyyplMrZsgHDOgfd+lkJrJiBogc30rR8SbXFKs2AAIDNCgm3vT12sSyNskWXUY2FeoT4uWUudeEeZhMg4vQzOCqjSNkCxGqrovoo+M1T7tw0XpoEQD6HvdW+rI//+hXro41T6JAfe1TlPp88tKjW57pBEeZENUAsGsBb/79Uf1BlPJOaPCRgbcGSonrR+7eS4NM6gCAtqNOGKl77TBp8NL42fp5V/4QiZjkomcUD+znb98we6DdUfXkkITnE8O26Ouu71yb4hkCuFDdc4+sGx6rTHk+1vTCcD8xjgTD8o8UcBWmWixztE7M73PscEDBoE4PQ1vOeoxZyIXToHC/dhv0OJBzQbrn75e91P3IanvvxFjwnVfGabdaSyiFF3fGTNdGFfPvvJ8X7tNWf8DM3QoYT/+30fzFC73/5hYB0ysjwIJtoovrTmEaQgXYiwPfLVSQv0Xh/k3MFhwh5JtZ5BegOZ1sr1b/yoTvzHuMA1D4eczrmomRsIiBJ7Cefr/vd8qjsSBCnJw4mla+SyVh74cKZubYf9/sJnJ1p/B/OczkYbOQ7ioGGAIBc6qUDUipeVcLwz2b5HEtC5JAwQKsVX2PvCmsczUpbSJlSXCRDe3OfuT6xq32AN4dzC+wJqs7WSqYytRf3EvoP1fspjX6t/fJYQETUBUda/fTTDUFNP/NvO7cPPrWZMCO+KF79TQ295Xz9LUHYX8wvrxDwfowrqIdANijyC6+c8lVquRsA+jz3yWa9lHIaeMvMmG8dER+9sxDq+7IVv9ZcNaKF75Yvf6SCJGfzKdA2+3lGeeK0nWmory7nmtR/0Hmp2hMhUpsHPxp1vfFfvXbe8mRQuNbWbLnlukl/jTcGc7l7gIcoeMOTm93TpDZ1NtjP4Uu9d0udx6nnPDJ1Exv68TM9jJDx4218OBHlIAJbGgpeJDesUdjZu1nMTNtfIOz60/g72fm4r0TUgMAw0DznneTD8nvena1vCBvgSOBuRZLIlByp1H3oAdfSLFy9W240a1vbto7WgEsQDGINtG9VWM72o3ze/LFd9LQfqSf8clyauQxvAf34/Qp379AQ1JyArY2bCoHhva8sGowjR3U+mL0k77F85d1f12fSlvjJlGCibhOinTUkUvTV9HQBW7/bcmbuoiXNXakfRrN8KAmrhbAf48Y8mawv5czx12jA1Y8laLZzD6/YyUf7RtsXEqY+nOnzcqHjtvN20+r9NtASHKFHeCIj6mq16UCeHzfS/aqH3+anG8AtnD9eOiC3bY6tHQmDCdKKQzUMfXXxBsZxvlJfv212/O1JUNWHW0aK/u6mICof0/g8TRtF5e3ZJOwivPbi3VsiHkWm2lAMwF4iJQZnihz6ZqS4cY1dFxTsn8Ov89dgBavHqTerwgW1Saps59u3dMqXW8sufl4dqISDLY2bo9+/TUr+3O96eZv2/3YxAUlDv6ns/mK4zGmg7ZxpGRw1qq7Nv42av0I6IiQP7tkpjr/BeyTwLBiCgyD8feOSkwbplERy/qJl0eh+2fYXAs454DlAk0dopyFgw8WHAvsBr8qoxh/6B3+7sGx1B44lpvJXN5de/XaDutahs/+nVVIOJZx3/8bvB6pynJmiGhwkY/xOYTgOwceu2tBIdBFN5eRHP0GP+woh9b8pi6zOg/ZoJOEIN66QGmGAE3+wZsLt2aZI27pi7cCKwTpAhjQoYXLYAIBwmDv7Mr5+/m35nZkkHoW3jVAeAvxvbmicWAF/3L52T2CNtonS2PRLzpIWlhSEFpCDYZr6zZ84YpucWVNRNNK1bI+XaCBgBQWc1mCi2tfL4KUN01wQEJqHUHgW2/RQ4k+mT8GwwrrF4zSbNQPk2YnYNz2F7X3zNm3MZ5zzuweaowxbiInwAAqemLsBt/5uasqfxYPofD+ypxxZsGxtwznImFc43tJXkwL2RrbNPrxZprImbDu+jalavmqbJEja38BGY67ZAJs2vxDWSY3LEzm00PfyQAa11IDAKuJgsx3++W6iTFPgCSpnuwGED2+ixDzq39u3TMiWQ/t/vF6oHLL9HgeNBHRulBaJ/O6y9XqNBNmT7xqn2z3PfzFW3HZVIPpqZbwAdXMyz8cyRnbVz//nMpdo2tCWLzLPx3+PmqluPSL8ObC8kqAh0jYvH7KTnI94Z15cIw9uT7ec8ncH88/HZdxzVT++V3zGxSKB29aqBXTVemjBP3XVMeocTBHkINBbY4+86ur/uIIKga9TyT7AwzVa4CNz99GuS2UVr5M6j+6vPZixRA9o10ja/DUM7NbHaMCaQMMI+gFKT8/fqqn8mNfTIAvz0k9p9991V7dq1dUa+U6dO+guUe/wpKDxqsHroJhkisnyTog0ArUQu27e79fcfOnFQGm1nZQB9k6haFEXmCwZOM5w1W8QYBg0O4vTnCugJygU02DWGdW6i+17a/l+rgKxklCwff47hXZroXs1Re4ADUUSGzOwTAhkX772T9ffQ9sd8Hlu/cJNCZTIA0KrmAm9zM3H7Uf3SssU4sE1wep55AMMZgIFvw5ieLVSrBqljzoMzBGIJAKRsytWbMX6X7L2T6mtph3P8sPa6n6uJ4Z0Tjkgm8Lm1V48W6tih4X1hYTxlA3JEeA093sF5e3YNzEZ0apq5vR8fexibn3kBNjK+4aiizcvIbumtfXBA24JoYbRwosbxdYI+zccP66BaWoz1qJn1TGOCLAH2lPZe9icu8Mj9gf1a6S/b+uKIIsBlgqvQj+7ZQp2ym721mq2NnE3kzsx2cucXrYmu2K+H9fMv28e+z5iOEfVKJiQzT8kxwZrHvtWxaamqm0VvYDNwEATu2PVr21BdGnDvwEhLNjqMHv3FzKVpcwwtFS8cY7/G3ccMSMseZVKjh1ib+c5Q7nDxmG5WMbbHTh6aUp9MrJ4olGX+HOidfspundLORhj4eIagcyDzNUpSrnHM4HYp9kgmRGVD8HHHOf/7gPv9628GpBnq/AwhmOU9PNiNc++KgHPr4P6t01p52hhPyM4SEDgwVe6b16ulrtzPfg3sA2ZwnhCFRs7HHXsA9sgOTbLfn0yYCWX+zqp751YQorQ/5UAyxywVw7mFZ7EJM+NMicpGIExesDqt1R8+G8H+XSxZ535tG1gTWUF7vyl8SPMSwVPskbb/1zpia10b+Lw/enA7fW6Ze8qxQ9tlZNBEXe9HDmqrz60aVaOvd1swwUxa0WOg/eCZI7uEsrJuPaJv2s+2WPZ42jcRQKyofehz4huccsopqqSkRP3nP/9R48ePVxMmTNBfEydO1H+6xAMPPKADB7Vq1dJ1/OPG2Sl0hBdeeEH16NFD/37fvn3Vm2++qSoiaFMKOrxswFzmzonNmIARj6zj/V6WisCjaRx92qQ6/rYF07i0ehpNGQbNP343JO13IxktFoPANFrQM3Ts1aM1EwERWI4owky257C9LzivyKxfuvdOKbXWZjY9CDCuOGxCM8jW7dm9uXrmjF0yHvTtjIi1rf7UZnzBOd6lcxP15GlDU35uo/XzQ4kEtbjmQFBQ5pYj+ugMK4+q294p1yygeWcbcxtV/Q8H9FS/27VjmtEdRQch2w3/wRN2Vj1a1tfBhWxR3+L42N7bE6cO1XS944a2yzj2fF4T48U0npvUTX9nWIdDOjZS+/VumfLzhrXTf9ccBpswls3YBctl5i0H6HeVjVJ0Nr1jsf5m3Ly/umhMNx3pzwY23902Hgh2PnrSYPWHA3qoe44d4P88KoPezLDbrnHM4La6nhaZ4kx7o2kgmpoZtvWFOU49fDG/Mq13234ZtE5shjfGHJl1/InsM8HGTLAh7ZkC9hdkjxDcxthwhLVpIgPcDCzWtLw30Ht37dpU/SvCOwPMAJ25VnAWNzLOxt27NdVsOzCpTARl2ThsGSjzTNm3dws17o9jtINx9f491P2/jda/nWBbjrYxuemwPjqzbgZ4owqIRlkre3ZvpgZ3bKz+c8GIjMEvc62URvh84IZDeuukAR9PouAHrZM3v1/Iaugzn41X7d9DB0YQtDDHMMq4285HGxDgp30Y6yUTOJNM//+IZRBI2CCIjgB7VIBNkiyxS/0323sDwwmMob16NE/5eVj7QlwjSP+lkcWeuPs3AzTbYlCHRhn3VdMut13DhpfP3U1Nu2k/dcHobmpoCJUdMMuubOeveW8Ierx2/gg19cb99Py6+fA+/r9F6WBl21Ns44G1DnYb1j3HRsta5KWXtjLdIMFF2PJYh0+fntB0C1sjPAHmB4pEFE+pSZMmqYceekjtv//+asCAAap///4pX64A4b1LLrlEXXvttTpwgGvtu+++mvZvwxdffKGOO+44ddppp+lgw2GHHaa/fvjBXidSzECG3ezXa8LMzpqCObZFw+uK+EbJe4ty0OYY9HegkZGJoWvAUcFBybEhQmsX25o0n4U2OrwnOHnciYhiUNjEhWybGBxHZDcRRSaRKf0cEVXizQPSHmSp6o8Jr12yGS3mAVJqqaG3PQe9r96tEZFOZoRtYmrcwKZ/TzFaAg47MBzwLH9mEVZkNcxDiqufEs3UdqjYDmDMAxxS9xn05yCjyIzsZuPQgz4JnB+SpQiCrYVg2NijzCBjJtVbe3zdm+/NXIv0znCg/vnIvhnfmXkg2sbFlqHHNfBuYbTceXQ6VTHMIIs6JsjYwDG7aMxO6pD+rVU2yGa9j+nVQmcRkJHmP48Cc08OG3NkijONh0k55QyAoGeoxX5mBlfC5lU0hz49WIh3g8w6xh5BwzBDz7a/pzl2AfsLvTeMTaZnyvQ8tvdGv4MyAc5gCxKcNDtL2N6ZuR4pAIf3RUE8ckSzzdQGPQsFL7DmQRs/qJ+7tYLMOrK4vA458tmYxVpBYJVqtoOuYQZRowaKalYv0f/GHRUbW4/bPihdIDYWH5OaAZlNPBuC75fu012N7pnqoEZ5X1GDntgjaR8+dEDmcTfZLWYCIggYF8wv2F6cCRaGWUvWpWXow8YeYwC7hdsTNsYMP0900MDQ/QlLEODcwnr5i0FLD7InTCZrlLML18CzIMjCn8XG8qRuKdk422QbY0wwv/bv08r/t+aWwJQJzo4kVLf8rHn9WpohcMIuHTK+K84esQWMgxgAdHbt1jWVaWhbI52blaZonwQ9S6XL0Pfq1atM+s1DXf+MM87QDAHcA8T46tSpo/75z39af/+ee+5R++23n7r88stVz5491Y033qh23nlndf/996uKhj+yjTJoczHr1MzNxUYp5b+DSB4i4ORY2RaNmfm03YuZFeRRxZOGd1C/GdxO7ea1UIoSjd4aYQMwN7p7WTbNdo1tESKftveFTYxw/aFJpyvoOcwDsXYEo4IHK3jm0aaCagZU7Bn69Gvw9/XgCYN8CtS8leuzNvCDKOr0HChhOHW3Tv5nmbR+W1DIzvxIP4DJWEVpwVmjOvuU+CBj2KzJzKblIj0P5go9T6MMRktQ5iPT2OMwvu7gXv7P51qygmR8gQYb9N5s74yo9aBAX3twL58Gbs5hGEbm2qsaMUMPo5gA6inGJnHN6pFpi5nAaf14lzT2ttKeXBwhc73DQRjtBT2jZDps69FmYPD1/up5u4UG8NKc3+rVMu5ZFIgiPH9WMhtsU7q3UdaD1kndmunjyY3Ga9gcDtojTWE90/GqHuR4sZ/f/Zv+oWvFHO9qWZ6NyESRkxGUoTf3SVvi0Dwb+TVOG9FZn42glNPnZeoKYlsr5rPYfucKjwpuM9RzZa/xtXLvsckAa5QzPpez8dHfDQ69Bg8U2z4/aJ+hZwOFnK5ns4WCxoY729Wr2a/Bz8zz9+yma8eJ/Whb95zen03QE3tv8r5K1LFD2gUGYYEtxn4ftWsOvSckLKARExRMNttPBnWd2KF2BF4D84yzQMz1yM8sKKIHBQ3sDn1V3wHFNYhxEDSHzc4iUc4ufjZCG4gCLbb7Icc0m3PLTHbhc0nw1RagMQPqtmeokuH8femcXf3vbe8q03sJWif87Po3Y6za1gg/k6n7ktTQQ0zkttvUFVdcoT766CO1bNkytXr16pQvF4AAH+j9Y8aM8X8G2j/+PnasXSwBP+e/DyCjH/T7wKZNmwryPHED1ELaXGyT2VRstk3msCwEgHrnf548RDspWONoR2EiCn09KENPdZ4QMUFmOMjxMn9mjegZz2JudHBW/M148/aMjp1NUdr2vrgyLQIoh3mbcZADaY5BaQTaH88GgFJPB6TdaNkR+vn6GjZjlY0JDhXaLG3ZLdv7j3Rwec8G4wWGfTtPwCrsAPY/32JoNg4xEOBkX71/T03xD8tymG2assrQs3EhJzhoLZrgKvehzBN22J+8WyetQxA0LvTeeP2Z+d7CjJbEc3RSv/fo2Cjl4WNtH/f0uWQTvDIzWicM6xC4TtZs2pJTzW0zZqwCVGoTjZFjcawt64RTxgFkuoLml436WWqsR9uz8X0bmU1qZ2UN4JnrPUKG3pzjoHkSDTPqeg/M0FuYJ83q1kpxKI7c2VuTQXuk8d7NDL1tXMy1cvjAtuq4oe0DnykoMx6FxQSAnfHvM4eFMgBM+q3tPYbdB8YdZyMpZNvmcto5ZStHM96XLVh97JD2ftAg0/5u24ttbD9u4EPzgpIDUR36NPaajW3C1krX5vV8rZRoayX184MC0TQf8d5uObyvNVMaZgvxszWIXcKBkgtcp23DOoEMTHSkyTfoCZy+eydfcDPKHI7aeYDv91RnH9aGjUBzw9xfbCWMfOxxDZSN2dYjHxckD5KUe5WZvcb2FFzj9N29MyXIntiQvT1hno1neNewnSlrjbMxWoY+3WmHSF9UG9J2RplrBYEBzqhBiQIF0qMkAtOeIZDlmfz58C5N/Plsm1t83Gd5bV+lhh5iVmPGqC+//FL3n2/evLlq1KiR/mrYsKH+0wXACNi2bZtq0SJhwBLw90WL7Iqt+Hk2vw/ceuutqkGDBv5Xu3aptarlGRQtDXJWzAPGzAZFMfawcEmt3da6LlPGAGhiOBE2g4I2Z9uzmM627Zrms9g2IYpG2jYx07Ezo/lRjBZ9jQxjYj57lJpXfqjwa9icxzSKnCWibnsO0y+jWnxEtM15xEWyCFE2SvMQIIPFzKDZMoLWDL2lHjybMbdR5LKJ4PL3SOMIwyNKkMuWod9kKTcxD2M6JG1ZwaRDnzS6ImUhzPnFDCXuQNgO4ZIsKPcp1/DeF7LxplNiZiGi0klN2rs/9lmKiQVlCUCBNMeDxt1m4Nvel7kebdPNfFdh+3wmRo5trdvWVztPkMsuihc9gGej3KfvkcFZTtveHZkabYxNUMDQdv/VopyNVezvbOGqDdYzyRwb2++YZ6PNjk2UxCjrPrbOMGRNB8z2LLY60qA1b1vntrPV9r7M4FdyzUfM0BsJCTvtOo+1kmWfe/1//M/fGom5Zu4tNhHYIPjPYnlfUTK1NphrMczusj0T79KR7dm4PkLQkxhC5vq0zRlz7Gk9mvaEue5ISNQ8G236MmnnVrb2RA6MF3+O2YLdaeNuseXMenSLuK2/FiOsE9teb14DLCybkx/0HDb7OtoeH329b2HPUVEz9Dm1rfvww2jtgYoRV199ta7TJyBDXyxOPbU8C8oKmosmLUNvOVxshzXEgKCUbxPGs9GjM9U62860sMVPwmsEm8NkUgVtB1yYQWFSpay0/rSgQbqD5B+QETP0tSPQSbPaxCJQ5DLRSSnLinvFpogDEEwKwjbLmGdDVecG99iflZpn1KBaM/SWzw/L0Ec1Is1xzwZcBZ5nEXGtTOJBthp6W5YkzWjxAi3mO4NhROuCU83M+Yb7xH1zWrvpCPFrbmDPYhsXWxbaSrkPMIwoCMLXAe+ikE3v2DTHka1F3GdYtwLbfmHuhTal/rAAXloZD+vhTKhiIS+a4xG2N5rXSKOnZ+giYQaKYAyb72pbVjX0qYEqfIxpKNPzRGUxhTn02H+pBaf5bkk13FwrtrMxrYY+wh4JsSncC97nolUb0wRJzfVizdCnnY32bBjeGc75jQa7zFwrtvM4E3stbM3zMdkcoA1gm2e24Fcm59FElFILs3wkLNhtvv8wJe0gZPP5+ZyNma5lsphydehpLWIPRmDVDNCm7S8R31lNy9mYyekCqPzOfB6zDMu8RmqwOzhDz68RRV/G3AfofWFt29qxmbZqLn3Pw/Z7c73bnW2VMaibjbNtKwszj1LOVI1ie2XyG/iexe2VbM7Grew5kvMqt7VYoRz6UaNGqUKjadOmqmrVqurXX1N7MeLvLVumqjET8PNsfh+oWbOm/ipG0GQOdujDqTM1LHUqtoOBRNLQkzzTNWwwnV57hqAkJHu+JWujJYzKajtY0jL0lmuYmzcMurToagaH3jwwSyMIPtnaDgVdw8zQ2BxLK53UeF94Ljjx0E7AAckdenumVmUN3+A2DmDbu7eNZ1YZ+oA1Yh7AUdXwzcOFH5pYj5kcet4RITxDH2S0BJcp8Gtb31udGrpeMegamKN0iPL3Zsvs2rKONofezExxuiTmcYpDH7GHvQnTqKjlfSZeDRTV+XiZrIAoDr1N+I4bxDAc+eeY76aU9XCOWmaRybA3r2HS023PZVOXp/WNayDIxffsbBg5ppOEuWbumxn3SFMXwFhL/PMw5r5DH2jgB7NZgsvRMs8HrJO2DWurn5eu04GQdId+e9YZ+qDyEnLozXdmrhW7vkzmszFozdvGO0rJW9hayZVybyu1yCaDar5/SohkA8pQ24Pp9qxjrkJcYYF7c9wzJDwzMgopoGzuH+aeH72GvqrVCUbwi8+VwDKFCAEKc14H2RNRmBN0JoE2vo69b3O/ruWxi2iOmXvbKlMULwdVdXpfVPLG7zMKMyNdMT59zMIST9Ey9CpjsDtpb1vYSxnsrOrMoES3nUVbNubAVt2R9rMsCDJFgZwfZ+XKlequu+5Sp59+uv66++671apViV7ELlCjRg01aNAg9f777/s/2759u/778OHpLV0A/Jz/PvDuu+8G/n6xI0nJshvAmRamTUnSdtgT5X6a1aHfnrVDb6PphlGAItHhjWfJlgVgBg2iGC3WjF2G7JOZ0YxCkQ1mZkQQxbNQ5KKUWnDKqimUZTNMc8lC8KxgyudbxtdGXYN4ViYRJzNLa8I8gKPQ5W1jhTmdDcXb1rbOFDrS1zDGLygLwcedG8JRugOERfD5PLYZRrb3lXDiwscF843mobke1xhZCJueRTbZJ8DMbJpzLNcMPTeAzeeIIlBp22+zohHnIFplG0cYX9SKKy1Y5P0+v9UgO9VknlgNvQxlSWnPZFKv2fvhgntplHvPwF+0emMaK8Ecm3RV7fT3aJsjbUMYM+Z7zjVDH2aAm2slmr6M9RKhZ1eVLIMGzWJw6M11YKOrB62VKGejqaIfBbTeozoQ+dB8wxI2acyMHPfIWmzNRMmgl0YMgvBx4fueeY3AIEiEd2bOh7ZUYmPYK0E2atTOOeY8p/9ms1XTM/Q5OPQ8yJJhvUc5t6zne0jJW/rZWJIxaJDtes9Uqst9BC6cm8/ZmA9bprwip6f55ptvVJcuXbQTv3z5cv0FBXr8zGUfelDhH3nkEfWvf/1LTZkyRZ1zzjlq3bp1WvUeOOmkkzRlnnDhhReqt956Swcepk6dqq677jp97+eff76qiMhUr52JOmNVO7cY4jt5Dj2ciHURKH6ZHHrbRlorC2fbdnalteaxLNyw2qQomdpIRkuNLGvSItSHbtlq/z/2jF1mWqGNBWAbE6pxz0Rhy/Xg8unjOX6+7uOcgXZvZmkzj3vEVIfFgUjWU2c2sm019FZaYQCNeNm6zSlGKz+EuaFqyw6Z69FGRbcdxvQc3FkPqqvnaspBCAqApGcdo42JuR6RPaF5ba55c0xsCuQlEWiFyGrS6zOdiEwiXLZr2GmFIU5EmkOfEwkvnDHjvf+67LODAl8m5T4Xxy4TzZfvX/zzzbWCGm78DB8HSnyYQZmtYGwmFoDt9djWShp7LWAfDTpXoqyVdIfebgZGLe3IOdgdErS3BVvNfcn2atLqdn16d+asYNRsc+rnBzMJg/apXIW4/GeJQL22aSfYYO5h2Kt95f4IJT22oKQNfC1ibgTtw8FBkBxK+Gj/WrkhxUkNuoZtXEzGTFD5i+1ZcM20rjk5sDPM8pew9W6bWmmUe0tSxyx5CxvzKDX0Vsp9FqViYUh16LM4G7flpvVU4R36iy++WB1yyCFq9uzZ6uWXX9Zfs2bNUgcddJC66KKLlCv85je/UXfeeae65ppr1IABA9SkSZO0w07Cd3PmzFELFy70f3/XXXdVzzzzjHr44Yd1z/oXX3xRvfrqq6pPn2T/0IqEMMfOnqE3shAhfehNo4MEbmYsXpt/hj5Lwboodc5plHsbCyAkA2FuxFEMI7uBn52xalK2bGNiOqI07rbnMA92G60wCuU+NINuraHPfqOkzzeF9+wMAPvn20TeomZpgZXmuEcIUBGC6F+m0W2bS1Fr6G21qUTX547Xtiwy9JneWVAWlcYlE/02iHYfdT2mG6vRxiQbmq85h21jlFYnaPl8buCZ88v8TLsYZZWM6zNMHCtXSqwN/no3M1zeu+KOddA6MQOIYYZeEIvJXP9mAJK/R/75ZlcXjE3QHpYmGJunQz83jeYbjWYapRwtq7ViK0erGi1DH8ac4PduM5TTgt1Zno1R1rjt1ZjOSBhr0VyPuQS/kgGD4Ax9muBijk5EWPbRFEfbms8eGWJPmNna0ojvLD3QYmdOxBkEwZmD/wY2zpK1mzK+G1vQIFOGPiwxhDExL5XLc6QEDcwAXgRmhrlObOeOWfIWzl7KTLm39bPPRngvDKDcB9lcdcLOxhy7MVWKDP2VV16pqlVjhmK1arqVHf7NJZBd/+WXX3R7ua+++koNG5ZoFQOgjd7jjz+e8vtHH320mjZtmv79H374QR1wwAGqoiJsMtuMiihGi83ABLo1tyvdRxLFM7KotqhlWGbbpNzbYD6bLTAaVkOfJrxneS5zY7RuYr7RYj+oMlGNrDWvhtESRsUzRfFstEIbHdpOuaeMXWYKWy4HV4t6tfS9YOMFLdb//CxEuDJl6MOytPZyDvt8ttY6R6R/2Z7H5njZssQ22MaFDC/cJw9i2N5bpncW5AjTu+H1bUHrP5JDH7B/mUZLUA0kB57T9lxBTorplNifo0rG9R7GzEgLqNrEhSyfFyR0tyESjTj3DH2Q0j2td54dCVon9cwMfYihF3RumZ9tOgc8yMWDVzbD1bZWkA02r2GuE9seaTMGk+8sc+vNoPKUKBm9oMCUmbGzdSSIonIfdo00hz6CE5Ht2Zg+5rY7zHw2hiU5zLWSS/CLyj8QZE/bQ7x10qphrUg2VSYEOcFRxRBN4DbQ7jjtOmG1zhHKhqIgaFwCM/Q5ZLaxL6DVsrkebQEowEYCiCK0G2RHmjZkXuUWNXIPdptCqzYbIKzkzbRR7eyZLGroIwSKwhCNcr817f9lY0dWKoe+fv36OhtuYu7cuapevWSPVEFhQRHmIBGptCxEmmFkWegBh32X5qX6z9leP8cgw8gGEhuJkqHfGIFyHwVhdfpRWAC2DcfMloYqe0asD43y/s0sBBmxtnGP0rYuavsxEspaaNBVbf8/l40S16QDeOFKlm22fX7AvMy3dd2KdZmZGYDt6qajQevRrHOzHSxhiuuZQOOyYGV6EATONo++295bJlph0Hok9gen8Qe9L1vruqiGUXr2KfPhDwZRGCsnE+XeuibSKPe1stqH05xG25hXieKkeJ9vzKvENTKv96hIrne7PgM3soPGJEqGnsbdVmISaY9k98GdMptDb10rEfYvewlG9D3SFhC2oVFEUbygwFT6XpP53ApyMIPOYN5BI3GN3Az8sKB9lLViu+3AtbLRTYaes1SCnOompTVjcepKvfsze87ra+ewRzYuTRfxzSaQk2tnAP3/AvawIIc+aB1kQpuI6z2wc04WZ6P5vky2Xz7PEXQN82y0BbvNSwaW8QScv2ltim3XMCwi29kYF+W+AXPozX25ToDNpe87psRThXPoQX0/7bTT1HPPPaedeHw9++yzWhzvuOOOi/8uBZGA1mIAspu2KJoZrTcXOmhy5mETdNiTiBePsGazKHu3aRC6qMKiebbIpw0QSfOvkSWtP4o4mnnfzSx1wmHXMA0jG0CpNmFGTEm86tc1qUak9QC2GC106HHY3hcd3ulR9cyU+IP6tVJRQMEJfg0biyFIKbhP6+S8CkIYlTQq5f7ag3sHUhbNcVnM2Ab6MyNGo6/ev4e1f7MJakvHn4fmFZwu7tjYMh192FoMgi34RQc7N4SC3teAdg0zXyOgs8VaoyVT0DXofQU5EGFOirk32oy+QR0apfw96Br+ely9ybjv7Rn31r16NE+hboeud+PzoxrcPVvVV1EQuN5pbrH7D86slaQEikNr6ANLxcLXSynb07iyvE00zV8rFi2IlPtO29vT79t2btnWYjYZKDjbPVrWi2B8B6yVCNTrmlEz9AEBb/MzbfuZ+b7CDHx7G6vMIpUkzsvHKquzMUKGntZjEMBSoYxh2npnGiO8i4n5LLt1baKigGyBlRs25yyK96cDe2bcv8IVz012YLo9ceHobt6/BTv7xNigHvDJ+97ul6CFtVs9a1TnSKLBtIeldmfZ4Z/d/HNtcyybs9Gcx9YMveX8vWyfnSK3xM1EubdpJwzv0iRy5wzr2RhBzHOf3omy5/AAnv0ZbNewAe2ygaMHtQssP23h21zpZyPdN7d3xaFXStexH3HEEVqErmPHjvrr5JNPVkcddZS67bbbMg6MwA0oCwYDzBahMjfj1YbzgizhxGv2VpOu2VsN75zYBE7cpYP1WrasSjYtvp44dai/+I4a1DY7yr0l8mnDuD+O8Z36g/u3Tr+GX+uamyge8MP1+6q+3qZP7fwiH44R3heM0u+u20d9e+0+vlN0pPG+WnpZ7V9XbUpTKI1itOAeMeZ4ls7NEsyLQwe0DhnzzDQ5c6O859iB6our9lK3H9lP/31ox8ahBxe/Rja1T2eP6qxO3a2T/n50gDEW3t0g2rj/bteO6vOr9lIvnbOrCoLPNojAaLDhzJGd1WdX7qmePn0X/fcgpXjbO+P17bzOzCaAtWeP5uq+4waGZiRsLWd8FkAGUTwK6Nx8eEK7pGOT1HZeGWmFhpMSFATD+7rjqMT86t7C7rRGp9xbWorVranG/2mMNuDBMGrDnG7bejQz26bRYnOkYGx8/ccx6r1LRvlraLCxVqh8wZxXtmtw1WrCa+ftpsb9YbQ6Z48u+u/HDW1vfY6g9U5zKyWQEzKnJ12zj+rv7V3dmbMapWYX+1mm5YL39D32yGv2UTu3Tw26pD1TteC1EjbHaI/EdWifP2xgm0hrka+VKAbky+fu6jsq+/dplWUNfeZANM55PAdhj+72fTKozZS539uDBlX1mdW5aal+ZmL0WT/fqnqduW4XDt/E/9tbX4c6sOzXp6V1rZgiiPoaEXqqP3LSYDX26r3UjYcl9q5ROzULtLvS13tynfB91cwEP37KUPXl1aPVFft1TzxDb3tbZQoc2DSE0lhMAQ7S6bt3Vn85pr/+vrvFXsmWOWELGF40pps+G58/a3hgUoLORnNckkGQkpSgpjn+V+3XQ1/jX6cO1X8nWzJwD2N6NBRcw7nI2+/abArYQdcd3Et/36153azWiZkUsj0HcN6eXdWleyec+p3bN8yuHG1j5mA3BGlxpmQqfwsO4JmlJOnX6NCkVI8H9j8w/ppYkhDZBPBseP33u6lxfxyturJx4N+nno2pa5G/G7JxK2INfbVcW8jdc8896tZbb1UzZ87UP4PCfZ06dkNNUBhgQWLzhKAbNkreCstmGC23RBBJbfuxU4aoaYvWqH5t7RFKMlw2ZTCMgoBN8MPL9lAzF69TfdrUD1n8FuGyiBl6vI8vrt5L/bJsvTXSGqZQa9L6g+qZYVQ8ddowNW/lej9jYN6DvkYeQiA0js+cMUxN/3Vt2pgg0gzfANFKjGlTRmtNoxUG9ENv6NWKvX7+CDV76TrVu3XwmJj90aO0A6E+9kcPbquNeuqUYIIckI2MOkmfj72XLhWkzAxjFVmIQwa0Tsl0RT1Y0rQTQrwJivS+e/FI//1ZGTMZFLWDkBDxSuypb180MpANQLXM3GjhhhGvdQ46wBDw6tKsbqCTaq+hjy6Kh2f57dD2qn/bhqpDkEOfh/APXQPBQcwvPIv9GiURs472a8Cp/9+Fu+t9iWeQOFp6xqU57pmE18zsBoywJWs2BRotiyxGi5mhCRLZa16/ljYi9+7VIpDVEkSFj9rdgLNuEMCFuF6PlsH7Sq5BT7NLxPuXjgocG1J4TgkYWvZhW1kXrfFXzt1N/bzEfm4lHYjtmoFFpTS0VrD+MrGyQBv9+k9j9PvqHTA2gSr3EdcK3hcCFAgKBTE2AgUkTQM/4ByDLfLC2cPV0rWbfQcuyjNEKQ80SxT+e8Huao7lnKfr4h7QBpTvhdsiUO6xRvEZxw9trwP3tjOlVYNaauqiNYHrHVlZOPSzlyXquM0liT0aa/rskV3Url2aqp6t6oU+K+wfPresLKaQOXb4wDZaA6kTc2yiO16ZEwS4L5yN+MJeifcTdDaaQclkEKSKZtvgveq/G8Fsfo3/XjAihZmT6Vn8cSmBQ19LzfFEP4M65yB4P6hD44zvy1wnZEs0qlPdD8LYaP24xvl7dVWjujfztamiPEc26x1nChIqGL+gvTGo5M38zKAOChiLdy4apbUIzLKelM/PkXKPtdu8XuIzEJBevXFrWveclgE2F59bHZuUqk+nLw21I4sVuSvmYDHXqaP69u2rVq9erd555x3VvXt31bNnktIjKDywecKhx0bJnSbbQg+qWaTFR1kVG8LaWAHIMJi13iZwgPYNCBjUNsRmOH0yiso9N1yCaFPhrfGiBQ2opqdBnQbZH45ZCIHQ+7KNCTZPOPEw/rGRpTj05gEcQoMDsNlnel8YE4w1HYD0HLjuUk9NNogSj8MrdF5ZNAcoegvDar5XWx8WWYUTE0bxDsrSYozMNRFljGyBnMT92o0W8/CyiVGasGU20w3j9Ow53hOnJYZlCHtZgjj+NWyOEAsaRHlfGPswCmPQesxG5R7X6Nc2u31L33cEWiHB5pxYM/QGnTRKr3sOOAE2xkSr+rX9fRDjwTUSslELxp4altHORMFMDeSErxM4dw1y2Iez3SOBoGBOkBCb7WxEG8iwAEXQuUXPgvmD91SjWpW09ZhOBrUHcYOc+fC1El7/agYobIHITOOfzVpBAMwmvBb2+bagZyYhObwv294CZwr7K2rrQcPljp/JAgijiIedKUlGjpFtZmPO13GQXkqmc4sE2nDf6zZvS3HMsmlbh+sHzd9s12Mm3YGgYBGdjYFBEDj0XjCb/h6EsHVCATx+NvKgQaIV77LE30NslrD3FezQb/HtIvo+aM/PdG4FjYnJzAhbi5yNkFWwOwIjh9A+IFifKYCXjSgegIB08/rhrMgdRtCL3k3HphU3Q59TeOKYY45R999/v/5+w4YNavDgwfpn/fr1Uy+99FLc9yjIAkHZm2zq26PAlungB2Sm2qZM4GJOPFuLRWqjMuV0DZaB4lR1HP44MOO8hr0eLb4xCXIe03rt5ihiw1kZQRkunkEOyqhEvYYt24xoepD4UTYIOrhWWg7dfMbIX4umY2d8Zr5rJclqSM+e48DiDl+u2nu2w9jWlimf9xWo3J2mcp/72JMBGkQjj+MamYzVfOv36teu5r+rIMpqHAiijyczjyVZ1UAGIRuKb76wlRHQWuGMg+XrNuW/R1r2MFtdfy4IVrnPTLmPfI2IbTdzCbpky8zI1fCGQR81sJrregxa73zMo4irRRkPOitWGAGnNO2EAq3HXNti+tTogLOxmkG5L4nRnuD7F5Vq5GOzBAkfr/KSQjzBkuscC+o8EOfZGFT6FK0DTPmwhZt77DgkE02WFb0blAHl23GivCKn0+WTTz5Ru+++u/7+lVde0U7WypUr1b333qtuuummuO9RkAWCDq9sI2CZYK+hTxpGNspNNuBODj9YoA9gCmHEETTgbALKzuchPO6jFqP1m0KFcY5JkmoU3v/Y7A+cDXg9ri1Ty+vYcj24fOfUEjBowPqPhmXQMiETRa4hq/nLx6FIrsUNKWMfpR95NrDX0CeNFv75uRr4tSJS7vN5X0GGUdQa+igIohWan5nPNYLqdjO1DY3DSYlSixgVnD6ecg2LA5yX40ilPFtTA6txBz0DNTos4o65Pg72Vzo3MgUN8kFUCm4+CMoKZpOxC0NYyVuU0pGoCKpxj2utBDmnfMyhKB8HqBWn6aykdTfIaz0Gj0uUrjnZBUHsY4JseYrAZY7UaCuzjAWi+TVytlkyZehZoiPftnX5sNcyXiMi5T7n9R7KVo1nn69VvarftScouYUxpz1aMvSIPK1apRo3Tgj1vPXWW+rII4/U9PsDDzxQTZ8+PZaBEeSGlh4dMy1azA7hIHXTbEDiQikGPqNi5tI31DRabRsACeJxwykOB3WD5Ro2MZdsQc8Af84sQYiTNRE1C5FPezQYVUQPT6lx9w5hqovLZ6MkOnxQoIhgZijiiESTkcTbouRj9NH7wLOgDCb5mQblPs/gFz0P1zYIUrmPM7qeFMXjn58PcyKa0ZJf0CCohj6eLARfi1BwDmvvlWtGKJX9YQTwYtxTggy8JOU+3nE3M2n5fm7YtWwByTiMu5Rzy1ICE5eqclAW1aTg5nWNqA59jmslKLgWVUAy3wx6PqwS++dvCFwnUVqDRgEJ45n6R+kZ+vzXY5Qa+tIcWv3x/Wvxmk0pgc4Uyj3PnudoT9oSTz7lvmpSoyaOoEFQ2zrepSbfDD0fEyQJzHHPZ/8PDOClnY25zS2/1DFHUbx8z8at3rvB3tnKs81E5R4Rjnbt1NixY9W6deu0Q7/PPgnF1BUrVqhatTL3Gxa4QzD1OrFgcC52YpSTOJ0iv6ZWq7rmHzSwRSXJkcPBxqlMuTqo5OzYrkHR8HzA6c6ZahF5a5tsQXV8YYGcOOAHcjanjzuvLc41o0JBFpvjGBddNciQJGYGH3cutpXLdYhqydej6ZDmG2Czv7Mk5T6O90aOcIq2gfcc/FDM931FqRPMtfdxqNCXqTWRxzXQax2vBO9nmacpYQuu5RMwDFK6jzNISN0R4GRzhgllT6uzcc+lh3cmNlacFN9QZhkz8G0tPHO+ho1yX1KiW3LFdY1Mwa98goW2NW8z8HN9nmyYGflQ1oNq3ONaK8HB9OQ6aRGgLJ57hn5z6LjXzWOOhbV1NddjmO5AGJqWJlrGYVuB9o/pkGKdpDrbeVLuraJ4qNPPP2gQ1I6NxogYIvoaOZcOpI8J1k3a/pjjeGRzNnKbNqfPz0MgOgpaZWCvYT3S3JIMPdpSXHSROv7441Xbtm1V69at1R577OFT8SGSJyg7BNE9/WxaSYm686j+qler+uqvvxkQby0iM4zuPXaAvsaDJwzK+Rq2DYCyqDjYHjt5iBZdefyUIblfw3J4UWQVQYMXzx6ur/HsmYnWYdkC2UvKLKdvlEnDCOq5T5w2TLkstUA7GSeBHG9u7dyhoRrTs0VgG6wosLetS27Etx/VT6tLX5DHs2SKqENIiebW30/YWeUDm+oqj0Zjjdx1TO7rMMXx4pR7ZhgBJw3voHsqU4vFeDL0SeYE2gfhfaHNU66wiftBK4GYLRh7vK+/5PG+glpVcuMeHR5uOCTRpioXIIBC6rt8PfL1jjl85f49HGQdE9dAW727jk60p8p3j4fRzcuckuUcVXS7Q7yvW47I/dxH8M/vsxwS9MQe+dTpue+RfK1sCChPefR3g/V18pnH1qAnWyvoioK1gj9zhU3cjzt2Nx3WR18DnVHyv0ZwkCWfNZ+qkxMcNMA17jl2QPxrhV2D92fPNWCAwCN3rPk6QSs69Js/a2Sif3quaFRaPY2hhr0LpYgA2sJij7zXa0Mav7ZB8p39dlj7nAP3+H82pXu+TiD6d9zQdmpMz+aBHV5yyWzzsxHBdLSmQ1vVIPX3XK7BS/gGd2yk9u3dQh07pF3O78uW3KJAN2IE2OtxpvzpoJ6xl7zxsxEdli4YnZvtFda5ivYUJLXyWetBPhDv8oV9/rhh7fS5ha4SFQk5zeBzzz1XDRs2TM2ZM0ftvffeqsQzGjt37iw19GUMXrfLkcxyVtFKlG9emNBAyBXJWud0tWBQzKD8nfc1LMbXyg2bfVo0VG3RFiUfYJNZqRJq0f41WC01ekDnew0Yxlu2bQ2s24UD+dZFI/O6RpAAG13jnycPVnv1aKHyBRn5aAFkGkbVPYM4js9PpcglM8HHDG6nv/JBphp61OqjNzu+4liPPy5cbc3Qd2lWmvcaCQqu+dkhL5h0w6F9YrnGhgDDCP2ZbT2a8zWM1jED+YiBbfIe+yDaHznCOOTRAitfYD1iLWLc+7dLHfehHRur588enufnhytrYz8JauWUExV+y3a/5VeS/VGi2x3iK1/gWrhGmlihN49h6OW7RwavlWTGDs5jvtexXYPvYejykfe5ZTG+OQUXgdUTdunghAVAawVOVj7PkVLytnlbCsuD1grYhPm+q8Aad29MXjpnuG5LlivgCNarWU3XscOJoDaTyTKLEl2q9vTpuSUFOKgrAe/0w4MIhw5srY4Z4uZ85PP45sP7qOOHdcj7bETXmpRgt8H6uvWIfnldo2aGxBNKZO45NvfgR9g6oeQTxuyhEwfHPiY07nVrVFNHDmqrv+K+Bl/vu3Zpop45I/c5HKQDwANFb188MmMXmUxoZTkbeRASc+vwgW31V0VDznysQYMGqcMPP1zVrZtsEYMa+t122y2uexPkcXihR+P6zVtDBazygU/H22qrqY25TpDT4f0MffV4rlE97Brx1L1lojLFQYnmWQhOj/WvEVO/TZ/ezbKotr7UuSK0b2xcCtEBkeiVMc+toECLTYQrH9iydbbWYnldw8JkoaBBvnoZUYwW/FsspQMBmU3u1MUBXkcfVqKQ8+fXT/98/hxxXAMBOvoca815TOMepuIc+9q3BQxjnsc2McHkHulO5Z5TcPOhXPvXCKBex7VWgkreuBMRy1qhtRjYIq0ktn2er0czqBoHGlso97RHImhAQbd8EKQxkzKPY3xn1rMx7vI6R5oZwR1gtvsZ9DjsSBuzjIJ3caz1sMBEXGdKKOU+xnOrpeVs5MySONdjeUPOM2HevHnq9ddf11n6zZtT63n+8pe/xHFvghyAGlZEjLHJw7nr7PXktfWMzgd0cPDFGbdhZNvEqBVIQ6Z4ng9stUk2cbR8ENgOxFKLmiuIvobNePWGrf6982h0HEiK1oWrRMerQh2vYRQcUU/qM8QFm2CSKwdik0PdAapdtirpx2QY2Wo318RstGQSeovrfdlq3ONci0E19LHPrWoluoWnXUQunncVhTkR2xwLUb2Ow0lJXCOdcu9qzQdRcPOppc1EJY5zreBsRCAisDViHAa+L8C2Ua9Bum8ujhbHNaYvXptaYuNgndDZlJKhpz0yR8p4dmKF8b0z29norhtEukClyxaSXAQ3H22ksP1xzabENUpjGndbyVuQ+G2+exbvEY+1TjmoOPbgVpazkesAxLkeyxtymgnvv/++OuSQQzTFfurUqapPnz5q9uzZepB23jm/mlOBiuVwmbF4bYpD78owwmFMizPOiDe/hi173tCrJXPBArCJo7nIPvlRyRgOLxJgW75us1q4eoPv0Mc+Jr7K/TYn2RQbRY4Mo9iyzQFZWn9uxTTuQdTobbFnIUJ0Bxw6EIUwjCj7BEprIYJrcWfoU4zVmAwj/vlL127SOgPIzsVtGNH7Sjj0FhG5GHv4ZlJUr16AtRLX/mIrS+L0a1dZQU7BjaO/cvCYxOfUUcmbKyeCC7DhvF2ydpNPyY0zMGVf7/EGcbhNQuVhwFrPsYvLoY/SLzyWTKqNGh3zOinE2WgLfJEtAWc+XmaZJUMf87gHlT3lu9+bbaJpn+TaMnEGuxdxm4tR7iuaEB5HTjPt6quvVpdddpn6/vvvtar9Sy+9pObOnatGjRqljj766PjvUpB3hCpuSqmth3vcQQM7Hd6Ns73Reo3qBennG5fxney3u9Eda8LSVi5OOqktoh47XTWAShp3ICdIkClu59F/ZymtBGN2ICxzeJur50hx6LcUlFYY1zqxGqsxBqYQvIOCORx4ZB4Tn58c/ziChClMrIJR7u0CbHEH8/C5dF65CkxZNS1crpW4KbgBrIk4bYlMdbtxjHsmAbZ4nIjg9R6nA+GL4jGH3mcxxe7Qp4uXxXkWh52Nca8T3jI49qBBWFIopnaFtv3RD3bHzV5zdDZSMii9dC/e7HlLb17h/azZuCVlrWNaxRHsLK/I6e1NmTJFnXTSSfr7atWqqQ0bNuha+htuuEHddtttcd+jIEv4ytqshiRuSqltccZuGFkcr1WURY2hR3xmyr3bGvq4I/jWAzJuOqnFwI+ThmcVX3MUKArqQx8n5d4WLY6dqm6Zw7EbRhY6Xtx7iq1HvDNjNYhyH9M68deipY4vjswQGFHm3OKK8HHTx60U9ZjGPZQ5EXfJkKWNqDM2C6fcM1FaV2cKUXDjXyupjl2cVPJMAba4xsTuPMZ3NoadvXHtjymieOvSRfFiK0sK0JiJM1sbVDYU9zopyNkYYkPGZaf6mlU2Uby4Si1quD0bsQ6ojaaN6Zf4nfzHpE6Nan5LWFqP/p4V41osj8jp6UpLS/26+VatWqmZM2f6/7Z06dL47k4QO+Ukro2SL851nvhe3CwAqttdtyk9ex4XLZquwZVi4xZHowjqUtZv1ewKEAeo122Yamxc72s9e19xjrv/+UzQ0RW7xOzdu8KFQ+8F16CATIrtrhgHKIOgz447i1qnerU01fmkQFJchlH6NeI2WsjoBVWdi0fGrdPA2xXSdeIWQzQDt1zJN65xp/fOa0Lj3lOA+rUSaw6q165EuACIsNFtU6uvOGnEfD2iVIEQd+kTLxWjz447Q09jv3rDFqN8IL61UjfgbIy7/jz0bIwjQ29JojjJ0NtE8TbGW5ZEaxHPwlusxm1HcvEy2iNdrROyUZ2cjQWwU+ls5DZL3MFumj84G10FVfn5a2sfGddaaWmsR7/MsQJn54GcVs0uu+yiPvvsM/39AQccoC699FJ18803q1NPPVX/m6BsATomsJxt+i7qHsnxIQc4zkxt4vPT68Xidrbpc+hzeWu8uKjX6N0JjP9lRcrP484MNSlNf19xOyq2tjlxRj9toj9xZ4JtRhEMilUxjzsd9qTkDH2DuOmkAEWjYROR4xW380iaFXAgKEtAB3Fcz0F0UqxFMvDidlJ6tqyv18LiNZvUvBVcjClep472YFA9KRsRd6kFXYN6UnPjO65rdGhSqv+ctXSd09rg4V2a6D/f+fHXlJ/HrZ8BZgPtYbTnx5mpTT1TkvtL/Gdj+pqPO/jVtlFtPcc2b9uufpi/ysla6dcmcTZ+Y5yNcXfQCD0bYyqBMfvDu6mhr+4HiyjIEneGvmerenp+YV59PXuFMyYTvTPMJwS8E58f7zujsxzOr1liE1cihesa0LkVu53KzkZC3OOONtAAxB353hVnYKovrXc2r7iKPgnlxeYDrfNsrpjP3vKKnGY0VOzRhx64/vrr1ejRo9Vzzz2nOnbsqP7xj3/EfY+CGA4XF4qr5qKJ2zBqbHPoPeMlrshnIyP4gQ2ZnElylPIFetkD42YvD2gpF+/hRePhgvbXmGr41rk3jJKZzXgPeXoG/p5gINF4xJmhxwEVuE5ieh5k+klJ1wwaxPXOEL2neUrrMe65RfMX7wdtN12J4pHhMm7Wcv/nyTkWX1cACNUBy9bSmMSbGaK9a5kxr+I0jKinNkRWCXHv88A+vVpodfZv5640aMvxt/4iI3vZuk1unAh/vaezGuI6f7Hm6xlrPu6aWsyhwR0a6e+5Yxfnuh/SKXE2fjM7uRYLcTZu375DEaEljuewJlEcrBNkz+mVkHMXd6YW72NMzxb6+7cnL0r5tzjPYtDhKbvtByVjXiew4Wgr9AN4Me8pjRatYBMAALO0SURBVCzBW0oKxWWnclsYc9cFM6NZvZqqc9PSNIc7zvU+1FvvX7P1HndJJZBmczkoEyuPyOnpoG7fr18/n37/4IMPqu+++06L43Xo0CHuexTkOZkBKCED1Vnte9yKq8lruImywcED/S9OZ5si98s9oxubMj1HXG3rBrRrqA+PJWs2qV+Wrfd/7r+vmDaZJnXTx32To8OLDBccjmQYxfEc3KmjqD0yRACVeOSLxqU19Z9wGukwoXmF90Q1cXHBNPbiHnd+DX8tbov3GjDuk45K6nPUiLHesdQw8MhJias1DzC0Y7pRQYFCKITH9b7MrGDs690I3CY/Pz7DqEuzdId+k4P527x+LbVz+4Tz+M6PSSfCxVpp4q1/qkN2Ni7MufPXSkxno+06/lqJaQ6nGOAs+EXOCq3VfDDEW4vICvIgMc5KoH5c57xxNtL+GNd6of0XYq5ULuZiPULQy2QuUolSXA49sG/vlvrPdyYvSklCxH8WpwYl414nCG5SHTu9r6RNFM8zlLLgLc0vtA6Oc/7S2budMXLWOTgbh1jPxs2xzS/++SabIc752zjwbBSH3oqVK1eqRx99VCveL1+eGPwJEyao+fPnxzYogvgytXTI1K0Zn7PS2DgkqU6pNCaDwnToYUhSBLc0pucwHVRev1sa03PAUenXtqGfpU8fk2pOxh3OKm1kcV3DD4D4Y56sG4vjYOFOHQVZqDYtroOLR+1NowjXiCu7mTaP1xrrxMEBRtng9d47i/OQTB/7+J8jaD26MFr4WpztUco7NK0T23VoPZKxGvt6N4xhGo84x5wy9DOXrPUNMHqOuPZgwr69E1nBt35IOvS0v8Q7x6qnzDGqpY97H6ZxcXE26uuY69HhWgElnjKD/lrxyjHy3bdojnEn4scFq/WfPVvVVy7ORn7OQ0QrTkaOy/0xpSzNC0i5uM7u3ZrqZ1qwaqP6npVbxH0Wm46Xi3USdDbG9QyahRcwv+Kyt1MYOXQ2Ohh3YsxYz8YmdWIpP0UwaOnazX4Z1+xl8X1+prMx7jOrvCGncAWy8TvttJNWtL/zzju1cw+8/PLL2sEXlC14ts6vRfU2mDgOL/86vuKqG4OCFiV9PhlecT6H+Qx0Dag7xyn65EcmWaZjrXew1In7cCRDdVP876tRwJgjC0EGjSunjuh5sUbtfaPIO+RjXB9B4+Kvk5iex3aN5HqP7xpBRrGLwEQy+EHjEt9zDO6YyAT/vGSdL84zy2POdIrBSTGzgkm2gbfeY5pjZoY2uU7iG4+OTetoii9ovZQ1dTHuPCv41azlljPFwVpZ52at2ErenKyVtPW4LfZ31bt1ff1ekBX8afEa/bNZS7214tFzXWQFJy9IOJF9WifKY+LfgxPvCmysOM557tT5Z4qjdWJqwNCzxLnXI7C+R/dmabT75NnlJgnh72EusrUOz9+04NpmB/a2GfxwsN6Jvfb9vFW+2v3sGM9GzKv+7RqkrHcKGHSMaT8JOxvjXovlDTlZ4Jdccok6+eST1fTp03UfegIE8j755JM470+QxyaJOjQyVlxk7NIdr3g3GD8jaCzKOJ3tIJZBnO8JGNqJahFZhj7mKG4yAJIQFaNnQUQ0LmfbpI8ns3XuDq64M5v8GjTuNBYuIrjpRouDrKNxjUKMiwtnO21+OTDwQFnt3qJeSu2uC6MieExiZhf5DkT88wp96Ns3rpNCu3dxHcr49mhZT4skvT91sfOMnau1Ynfo4ze+g54jTicCdbNUCkHBaMqoxbVW6Gwc59Xtrt64xXciEFCIdy26y2oH20NxO/SpwrEuzkceYHt7ckKoEuuSasTjmsdBSYg4maRp56+DsTcdSEoMuVjvaVnnGNd7u8a1VYv6NTUTduLcFbqkcu7y9bGud58hN2tFaoAwxmB6owAmaamDhE15Qk5W/tdff63OOuustJ+3adNGLVqUKqIhKDwg/ER1wBQtdpGxa2xQv3zDKK7oLatLQ7TQhbNtio0ko93xLvxBHRprmjcMFXOTiZttgDo3fLZv4MV5OHpjgronHPC+0RInlTTtAI5/TNKi9g6uEXStYnW2k1TlLe6yKWkMkPgDkcAQz4mYMGelvhbVJXaM0agIzAzF5TgGMaRiHHNeRz/TC3okDcn4g1/7eE7Eh9MWO8usmNovcTvbNO7QAKGSp/UFzNjFvlY8AxxrBXv+HM/ZJgGtuD5/8vxVWrl9ike3b9Owtn/exM1adFE2YgrGrne0Tswaet+2i3nc9+zRXDPvEMiDU8d7h8fGwgwItBRbZtsMrrrQs0hjkzoIGIFpQutx4pyVuhMMnHsk0agVXFy0/olzVjgJEAImW8ZlwqboHfqaNWuq1asTmy7HTz/9pJo1S9B0BGULLECAWpu4iOKCPsOvETeltBbLKuMaLpztel7PVZQHrt+Ca8Qf+KC6bRLcSlLl4j3wacyBTVu2+eNRWiP+MacafReU2OTc3e5M9Mecuy6eI+15ttDzxO9s1/SfJ/WdxXnY16qW+s5csH6oTpCUm5MGXrzj0qJeLf86szyDAgYL9S2OAzWDxj2m92Xu8S7qzbkA6QZvLFzSFzt5GgYkUunC+PbXvj8u8ToRfI8kEbF1xbxW6idEBNds3KIWrNygnwmsr9YNa8f0+Ym1CMdh4+bt6gfPoe8VU3aerxUEJPDl5Gw09kf/GjGvE3PckxoQ8Y47FPVJQBIBT1onYEdSK9bY9zAH6yR51m93ltkmW9U8G+Mce3PcXQjG8vXIz0YEuiHIGOvZ690/sePiKuGxzSsXJcflETmtykMOOUTdcMMNasuWLX5UZ86cOerKK69URx55ZNz3KMgB1GKC2r+4yECa14g76s1p9biGC2cbC58ug893RV/jThcYB/p6MWdqsQ7pncE4chGV5K1FtEPvcF5Re5lkJjj+50iuj/gP+eS1vHXiiUr5hneMc4xaBlKvcxcGRXJcTF2O+MaF7pc+25XzWNVTnsYcS9Lt4xPl4e2qSMgz7sy2OR6uglLUospfKw6cIQL1hqbuEy50AUh13F8rMYvi8T3S38M2O1wrjgMt/EyhbBqouXGVvaW8r+3bY6+f19dg6tbOzkZjncTNWCTQPKXxdums8LHn9ldcwrFpe5iDdeKf9d56d7FOyuJsdMUAofeVcjbGyFyjdYLPR3kNlRDEmaGvluabuAmuVQiH/q677lJr165VzZs3Vxs2bFCjRo1SXbp0UXXr1lU333xz/HcpyHlRIhrtythLv0b86qH8Gi6cbVyjLtsofRaAg4WfxppwcK2UA9iFs82MLz4mLuYVniFBj4zfgSDHgeauCwfYPMC2mc52nM/jH5LuDSP/nTkoHajrZSFov3JV+0bONp7FRYYgdS26cU75OnFFV7WtlWRAKn42S/W0ORY/m4WeJ/29xR+I3upwfzEdO1c129TqCXuLi7ViBqInz18da/18etDA1dmYXCco36O5G/c6McfdxXliBr9wdrk4h809zMU6obNxmxGQdHE2+s/hwFYNOhvjtlV9h5udjbE62yXpn9+0bs1Y31U10zdxWCZWnpDTG2zQoIF699131WeffaYV7+HcDxo0SI0ePTr+OxTktSgXrtqgurWo6+SwT0akEwbr1EWr498oq1bRCx9G8RxPnCPuDQwbCXqSY6P8xctAxE1f49Qv1KKBuuhik4FBjHjnwpUbfGp/nM62yZqY7QmalDowjMjwirtVYUqWzpu7Py9d64xyT8+DetpFqzYyqlyc416SEgSZsjCxFus6Wu8wWEm4qjTGsfeN1c1b9WFMyupxjwvN441bt6nPZy6LPQvB5xgYOYtXb4w/E8wyzXo/cVCaYhpgWC9E+XTBYqJx2bxtR+oeGSebhQW/+FqJOxCN94U57OIapmOHa9D5G7fRmhyT7eqT6UudrBVcA+9q5uK1atqvCTX93m3cOPTzVqz3Szpc7I8oTZi/coP/87jXCc/U4p0tWr3R2dlFY499mBiGLlgNKBH8dTU/G6s5ORvxvmhsXIw9Ah/YI0mwME5blUo2sSdCU8pvSewogItrQBiPl0LFHfj61NtP4vx882zEeLvQ+ymPyOrpxo4dq5YtW6YOOugg/fcRI0aomTNnqttvv12tX79eHXbYYeq+++7TNfaCsgUtmjOfHK/aNqrti+S5yti9PGGeFtAASmN37rbrGq5b3pyqf1aH1SjGATo85i7foO79YEbiGjXcZejveHuaFhwhxHktOoCPenCsk8/nxuqydZvU3e/9FH+giM2rs578xul7wnMgiPPvcXP130udUO4T13pvymL1/tT3lddJMt4sBHueBz6coTUhXLEzMC53vjPN/3mcBl6pd79wGq9/Y7L/87jXIzl1b36fFHGNMwvBDaMXx89Tr0yc72cL4srY0ZgjsLbrnz/wfx53DTWnSF7+4nfOrsOzwd/OXZnyTG7Wynb1/Ddz1a+rN8X+PLgG1iKCX49/MduvF43T2aZ3grXyt49mOlnzfK2MYy1X414rFIj+7aNf+arhcYlwmYHo/f76qf993GMO3PnOT/oLwI/iqjcncEbhzf+d4v/cBbuM1uPZT01wch06U16eOF+9Omm+P4fjZeQk1/uf/5d8X3E6d0ma+g71x1d+8H/uar2f+/R4/+dxM0CIJfXcNwmbyBXlHgFu2MJxfz4fD/S7342fIxU8Q5/VToO6+cmTk0bW999/r8444wy19957q6uuukq98cYb6tZbb3Vxn4K8ItIb1HSv5VC8tJZk5POv7013anyRsi6weE0iIh0X6H4nedFIYK2XhYoTJJrDnfm4x4QOYFefzzfk7+Ymah0BYgPE+fk4HD+fkcieAnHVbJrO6Zc/J6+h4ruE9b7JmQdKHdQFI0NAxmTcY8+pynAgXKx3+ixE1J8Y+4uzOUzPwhE35d4sTyHEV6ttP77jNuz5WkFgguCyZteEq4Dhjf9JGviuKJ/Xv/Gj27WyeatvGBfrWjHHHazCuOq0zfI9DheZYPO54nwOc4/85+ez0n7uej3GmxRKvjO2RTpb7498mnxfcWa2+dnIHeE490jOyPlw2pJQmy+OEgVX6922Dl0F001IDT3DpEmTUmj1zz77rBo6dKh65JFHdG/6e++9Vz3//PNpL1FQeAQZRrFmORmthde7xeh3+Y4KKZQCP3r0xbg3St6WJe5rcNVrE3FGWG3jTvTCuA9hapsT9/tKUnyTYx43uGgKtc4BZvy61sG1gpyU+OuCSQSGEKdqu1mmQCh1YrQk16LZwcGVUUH91l2Pe1wGcdDn254tv+vY51acATb/WkEOvQMnAs8D9pqTgCHVnW/fkXI2cgX8uFSvsVaoD3bca75QBjgXrXMF2/hSuWAsn29Zj+aaiTtDz0EsTNdjT0xMp2ejQ/HmQp6NNWJkZ5QGnI2FGPNm9Wo6XYfcBnN5jtSMmS1T3pDV061YsUK1aNHC//vHH3+s9t9/f//vQ4YMUXPnJqNTcWL58uXq+OOPV/Xr11cNGzZUp512mq7dD8Mee+yho6P86+yzz1aVAXyjbNWgljZcMMldRNqQSYWoBaFbi3qxL36qpQT+cEBP5eKApL7nwIWjuylXGXpzg+kQoyNh28hGdY+3lSSNycJVSabEpXt3j/HzE9sS1eoCPVrGN6dSo/bbU4JFJwzvEOt1EtcqsbYCa+61b4k7C8Gpqg3rJFqOxU3rT3VSSmIX/oGxin2LEHeWy2ZIxulsBa1F9NYujVkUz0TftvGpg5trxUXdfCbHDi3SOsRIyfRFsrbvSMnYdG5a18nZyJ3fOGnkvJa6Z6v6TtZ80FppFeNz2ObyxWN2ivXzbdcAhno9sV19/sid4m/jzIOePAjpJMBmWY+H9G8d2+fb7hn7Pp0DLgTSXDjb/GxsFPP6s52NLmHaK/Ad4jx/bXNqj5ht1GoBgSK+T1ZEZLVq4MzPmjVLtWvXTm3evFlNmDBBXX/99f6/r1mzRlWv7mYyw5lfuHChFuNDu7xTTjlFnXnmmeqZZ54J/X8oCUCpAKFOnXizMOUVfFGir+S/z9hFrdywWbVkRnJc18AmRsrtJw3voBqzbEFc1yD1U+DEXeJ1usgwQr0NHZhH7NxGxQ3TYbjjqH5q1E7NVBMWDIlzs9yrR3N17cG9VLtGMbfj8jZLCLwBwzs3UUcPbuvs84GXztlVxQneYogc+o5N6qg9uzeP9Tr8WoSHTxykBndsHGuGgAvWoRUMvbM46Xi8NQ9Fum87sm+shz2JUcJogfODoNHfj99ZxQ3TkHzqtGGxX8M0XJ49cxdtUMTVz9c0vE4f0Un9bteOql3MTAO/Vn87Arc19Ng8dsqQWK9hXouA+bVH9+axZoiIGs3Xyv2/HagaxGiIc1EmytohSBznmifKMMSxqN/9nw7sGT8F1xiTO4/uH9sctjnDfds0UMM6N4n18/U19HtJ2BGHD2yjLhrTLVZWjrkev7hqr1htIZPZBep149LE/d973EDlAnzse7Wqrx46cZBq3TDJaom7TOGZ04ep3m0axMyWoZrtpCP84WV7qDhBaw5Bg0alNbQo3qMnDY71GubZCDHE587cRbm2Vx48YVC8n2+M7TGD28Y6p2xr8ZK9d9JrPu6zsagd+gMOOEDXyt92223q1Vdf1c7x7rvv7v87FO/Rvi5uTJkyRb311lvq66+/VoMHJxYJxPdwP3feeadq3To4Yoh7bNmypaps4IsGxjcMido1Yl40LNNBipuuaKuUoYdBHHe2jjbKFV6GHtnguK9ho9zjsG/uMNOBzGmcma20DL2nrossR5zvy/x8RIjjrwsuSZu7O8XILEm9Vuq7gXMSt6FH7wzdGkhN3RWrAQE8ciAQLHRVH0pBwjidOYLp9LgwvE2jFJknMDNczStQ+V0YLL6xui25VpqWuhG+pWAeoWGdGrHPMRoX3kUhzp7nKcFu9s44vT8O8JIdOrviNoxta8XFeuTBr7jr8wt1NnJHCKwSF2ORQrnfvFV36QCaOxgTcz3CfnEVLCS0jHmP5OO+yBO/xDPFyYo0BUr9PTLmMeFn43ZPjMfFuWU6wwhQxPr5xh7fNuaEk+1sxLur6M48kFUo98Ybb1TVqlXTfedRN4+vGjWSg/3Pf/5T7bPPPrHfJNT1QbMnZx4YM2aMKikpUV99lVBFDcLTTz+tmjZtqvr06aOuvvpqrcYfhk2bNqnVq1enfBUjUg4XR3UjvNaZspxx16jQRklOiotnKTUo90G17nFn6GtaKPhxjruLz+cb/q9eBj3ug4uMO//zY2Qw2Chy5Di6Wiem0eLiOtWNd4Z6yriDIElHaLvatIXWu5uOExgXKrlw8b7MMXGx5k2jIu53hSwpj6NRWym3a2W70z0ybVwcjH2yvVgy+BX3HmZ/Z3F3Hyjxy138s6sQa8XFmJQUwF5xfDaaSRRXIOo1fLpVXns0V9fj4mIu16LLPYyegc7GJqU1nTFMeAkfgjpxotRj5CBDT0ED1/M4cY249WtSP8/JnlXVvc1VHpGVxQfH+JNPPlGrVq1SdevWVVWrpk6mF154Qf88bixatEg1b55KhUVgoXHjxvrfgvDb3/5WdejQQWfwwR648sor1bRp09TLL78c+H+g0s/LCIoVfNG4duyQhUg69HG30ChJqRtysfjpgFzmGUVxb8QE897dOCrJz3T1HHQIU//bZnXdCJrQ57tw6HkrLldzNyjD5eI6Vc13Vi/+yL0v/OPQsStl9eW0Hl28LzMT7GKtmJR7J4GckhKfLeFsvRcgcJu8lvt3Zu4vcIrjbmfE2Wsu3xmytRu3bE6eXY7mmPMxSXG23c8tl/PK1ecTEKzFpVAS7nLc05/JhfNYUoB92LBXXDBMWFnSJk9cOe6zkUQwsZ9sKtBadHGNROeHZMefQuxZNSuJQ5/TUzZo0CDNmQfgYPOMfSaAvm+K1plfU6cmeo/nAtTY77vvvqpv3766Bv+JJ55Qr7zyipo5M9lyyQSy+AhY0JcrkT/XSDkgHWVTknWCySxn3NfyM/QuHXojk+nKsUvP0LvI1BZg3C0UcpcHissDWM/dLcWfdTQd1GZOWA08gOexGhzU65oq8G6yjiVlkKF3mz11vd4LwWZJy9gVwInA/hK76KIviocgiHt2WSGChYVaj84SEI6DBjyA59KBwFwtNYQ13SVtXL+zwjOloAMSN6qysiRXATwb464iMHKKdc8qj3ArWZsBl156qTr55JNDf6dz5866Bn7x4sUpP9+6datWvs+mPn7YsITw0YwZMwJr/WvWrKm/ih2FoH+lGC0+BbfEiaOyzqNGulj8pebh6Ipyb9y7i+ukGPgFyqLFnUFPCxg4OICTgo5Jg9vV+zKdbTfj7nZMgOqWTG3c7eTIcCEqdLHS4e3XcGQQe10pndFuPWMVNE9q/VQIB6JQToQTBhAXjHV0NgKlaY5dAZwu13T1AgRW3QfX3KwPvkeuYWrnzs5610GQgqx39xoQtrMxdrZq1RIdFCS6vbuz0T1rAmsleZYU5/leHlGmDn2zZs30VyYMHz5crVy5Uo0fP14NGpRQXPzggw/U9u3bfSc9CiZNmqT/bNWqlaroSD283NdW+rRPZzX07jL06VkOVzX07qnXriOftuhn3Aax6Zy6zNBzSqy7Gnpj3C3spvLOmkhb7w7LFMCYIbEyZzTMAjuOeHcuem0XYr2TsbqBB1mcORCpn+siYJQeMHRY0uPQwAfM9l7FyjQxBeVcwPVaKVQNvT/uTN7JHTvHLXMiTV+mAHu9S02eDR7d3pU9gbNx+dbNBSlRoHnsQiAalPiNyh0zsqSkil+WUpky9EXxlD179lT77befbkE3btw49fnnn6vzzz9fHXvssb7C/fz581WPHj30vwOg1UPED0GA2bNnq9dff12ddNJJauTIkapfv36qoqMQ9C+qDdYUXEciWXSgoOdq4vOFch/lfbl6V2bGGfTouMXXzIy2kwOYta1z6ZxaDeKC0Apd6g64rQvm6t2usl3pWUfXNGL3wSJ31yCWVAEyggXIrKQFJF0Gv1L2l+I8u9JErBzvXwVxTqsXL+W+UKUWhRgXri/jynk017ubLg3eHsnaK7s+G6sVJEjsivVVOPFmQBz6cgao1cNhHz16tG5XN2LECPXwww/7/47e9BC8IxV71PK/9957WnUf/w/0/iOPPFK98cYbqjKgkBS2RJbTjcNt9hB18SxpWY6CZegdR1cLQF10mQl2eQ0STXE5dwNFZhxH1d2xGhL3jfeF9+YyQ8/hhpJXkpJJcZOFcG8YFUIzg4xutEtyyTagz3avN2EwgOoWZn9xVZ5SMSj3FaCGvgBsGbPVbvJ6FaBMoUAlby5LbHivezcBvOoFZUm5Kh0phLNdrQBnY3lDmVLuswEE95555pnAf+/YsaPaQbKJSql27dqpjz/+WFVWpCwYx2rnLmmFfnbIz9C7cCBSP9MV9dq8d+dKu87GvcRxrbb7enBOH6eeroXIOmJM4m6Zk7iG+3dGz5GShXARYGNOCi5pZm1jNyQL4JwWoiViDQelHPwaLllShVRUL2SGXqteOy1PKeyZUpBAdEEU2x2XDbjO0BvaCYU46920W3Wrol/oEj7aI10Fifl6L0hngwKUvxTibKwpNfSCipOhdxtlQ7KOaofiNvD97FBBa+irFm0NfQoF19m4V3GrGFsAg9tvwbZtu98+pZij0emsBgdCgsZadHXg8/WINeLCMCpMFxD3WccUyqpjlXsad5fOStVCZIPTOkK4WCuWvtQFody7D4C4EskqpJhrhaihdxwkLowaeQFKBAtSjubeTrWdjS5QiM4phWFmlDi/RnlD0WToBeUw4s2usX6TuzZWKZ9fCIfe0SbGHUb4KCYVLHYKbgEoWS7r0ajnbtw9olNasG3foZ36Ys/UptOIa8V+DXMtYq65MCT5enSXhSgAxbcQtYiFcIS8uUXj7tI4MrPBbjLOhcvYQVsmWZ5SnMFos7zCNcOoEC0RXZfxuKb48kCO0/VYyDKFArWNdbneXe+RKQ69sxKrAgitFuAaVQtgd5U3VI6nrIQoZNs6gFTu495k6Br+5xdp6x+zbV0hBGAK4ai4rEfTn1+vhpsMbUqfaLd96AsRXEtnNbho9WeuxapFa6wW3Nl2nD0vxHp3Pe78WkHzuhi6dCSuQRk7LpLlVm/CVZCYj4mL8hfzcwuSeXTM/HBN8S2EY1foPvSF0E5Awqm+oZkU59xyfjYyW7UiCK0W+9lY3iAOfQVF6qIsTPsyF9cyswOuVI+h1u7yGiblvqJkBV2L4rkQrErVf3CrQq2vVeAyCDAa6hhBqmJRhk9z6AsQZCkEa6JYhb7Ma7i8jisHvixEsuga61mvcNfsMldBYrP9ogsUpATGsRNRCLaMPejpMMDmWCCtMIknrvnjKEFQqLORBSOK2YYsBB2+Cvu+slDuK8dTVkIURP3YoC46UbkvQMsvm2HkAnwDLoxCqXtKlktj2NXnp3doKFzbOmcZLv7OHARZrO2rirhOsBBBlkIY+KmaGYVhf7jMCLowtjM9T9xtN/k11no1tdjTXDjDdQu8Vpw59AUsHXF1jUK2ySoY5d51hr7A2WAXCQjb2egqSFwIOzU1uOZmTylh+7yrfWt7UiO90qjcV46nrIRIVT92M8yopTPtr7g3srSaSkfPwg9IV5sxz9AXs0IpdyJcKtDrz3d0AFOEGH2iqdVMIWo33RnEbsfE2n7PWYaeKfkWQNfAlXZCqoKzo+dgQ+JOkdgwVh1dp1Dggeh6Dui3+hreO1qzcWvRn1t8y6rqKOBSCJqvc5X7AtbsFkJnpNBt6wrBlCrU2eiOvVYxVO6Zr+3sObYyj77Yz6yoqBxPWQnBo9EuMhA24wibpSnOFfdG2aphbeUCpWyjNJWD40Kt6u6vwd+Xq3HnqrGN6lSP//PZHHLx+dyQWL1xi5q9bL3+vmPT0grRc9XVOzODa60auFqL1ZyvE+5s13Igumgaq67W4sYt251fo1B7sAkX9eCmqFv9Wm73l+/nrdJ/dmhSWrRrhbMmCsH6cjePmU3kuCTJpc1lOnZO7TvH74x/vqv5y/fhhnXi15Ypq7PR2Tph9orZFjMubPK6YgEuRI/Na5iluxUVonJfQcE3GFdZCH+z9NZNz1b13Hw+g4trmIeJq/fFHfpCbMaujFVu4NWvXd3pmLt6Blof035do9vWtWlYW3+5AH+e2o4OL77eXdTP29di/aJdi/xZ6hSgXaELISaA2oWa+4tLY7WXo3E3gQ4Xzq/heD3+sCDh0A/t1Nh5INrlOe/6ffH1WM/Rnr+d+pM6Olf4enf1DDb7wbl95/A6rj/f3L/q15azMbsxcTOP129272yvZ2djZUHlCFtUQvBoscvDhV/HhYHPD8kmpTWciaRxJ8KFk5pGuS9AxNDVAblx6zan16hegLlLhwjZeMMcGdzpgnXus09c4DFOmOybHgUIrrlai9UL8L4KsQdTuYhLmJT7Hi3djLsJV4Gp1Gu4HXvX+0u9mtWdBz856lR3MybU2s9p8Is5EVxczMV6d/UM1j3SpX2X4gy7YOO5D0QXYh8uk7OxAM/iKgBSiHNrGy+iryQQh76CgotOuFqU5obvInue+vn1nYkmFSLizcU/qjuiXnOjxdVzcOVmN10H3B8oZtbRVQbNpPgWIkNfiCyay0xtIdZioVkTpY6MVZ7pcAWTcu+KmVEoZ7sQLAAzCDKkABl6/r0ruForfB6XOlsrW51qmaSyIt0GVwrBYioE0zOFKVWjeJlSZXE2FuJZXM1jziwTxAdx6CsoNrHaSpcRXL74e7R0kKFn0UKXmSEu9uO6rtJlhn4dM1qcUZkcOxHVCki5L4RDz1GnEPXajgxis6tF1+Z1izYLkVqi4CpL654FwAN4rpBaAlNNtWpQSxUCrpxHDndORPKddWlW6kyMixv4ZhChmN4Xd7Z5ALSY1krKOnGYRElz7ByxmIBtLMnp4jp8zhaCKeUsq51Sd17NWfleIdhrZgtcF9jCJ5YgNohDX0GxfgvPorob5qVrN/vfd2lW13mG3hV4fV0hMkM1qhWn0VKI6GoqRc49DQ/GdidHgngmnNEKC5GhZ++sXePazmq2ubHq6lk406d2EY8JV/J1BW6sumRJmSjEPuxMb4KtlaGdmihX4Gd7IYalThEzTVxfI4WqXMAa+loO+9BvYIGWuo6FBCvK2YjEk6ugVCHYa3xMXJ2NAjcQh76Cgjt2hTLAXLSfSGEAOKpLArhdXIj35SpDXxEMI9d1eyatH/WthVojhXBQC5F17Nbc3Vrk1GFXbf4KzZoohHPqCtxYLRTdvlDGZCHKLVzqc/B1z8vsii9DX/znViHEXG2ZWpdbJH9nLpzUQjClCkPrL4ydyse9LtPPKDbWRCFRo5Io3AOV50krGQpxQHK0b1zHyecuWbvJOcXXzNAXAq56bxZC2dM1C4AbpoVQvi0U3d6lonrKNQri0Ltbi1yXwVXf60K/r2I2jHibSlddRireWikp+P5SiOCXK82BisAs43DdcYCPtatscCHsyFSmlHsx10IE8FyUntocelPbxE2GvnjPrYr0DFEhDn0FRSEOSA5XBv6MX9c6FWArK4feVEV1QZFzBdcKpZu2utd/4IfWkI6NK9Th4iqzyY1Il8E1jkIQJ1w5KYWg9RcCfNx3alFAh74ga8WtgY9a2taO6mnLZK24ytCzEsFiPbd4wKAQLQQLwcwopB3pah9OzdC7bxfsco/knZJcjXtKINpxULWYGXjlEeLQV1AM69zYKv4VN+jzDxnQ2snn79o1UX/YoYkbBgBhRNdmBaXn7NLZTV3l6J4t/BpnV9i/byv9Z982DZx8fuPSGv73ruq0G9au4bdC7F6gNlzAgHYNnV/DlXgkd1ALlXV0Ncc4erV2TyPv7Eijgcbh8IFtVCGcuEI69CO6NXV+DVd0+IbeHrabd34VArt2aVq059Y+vVr6AoKucFC/hI0yqEMjJ5/fvF5N54K0NvRv626PHNTRzbuyoZujvYXbwK7E6moyR9ulPcHP4M6O1gq/RntHdjfZQUft3Fa5PnMP7u/GNymPqLJjR4FTk0WG1atXqwYNGqhVq1ap+vULVz+YL7Zu267e+G6Bzj62beTOGZ65ZK36Yf4qdUj/1k7qkBEh/s93C9Qe3ZurZuzAdNGz8vVv56vBHRqrdo7KB4BZS9ep7+atdPa+Nm7B+1qoRu7UVDWvV8tZpuO/3y3UwQPufMeJj6YtVg1qV1cD27szKN798VfVsn4t1dehQUSYumi1mr10vdqvT8JwdYHv561Si9ds9IM6LjD+l+Vq3aZtauROiQCYK0xbtEbvLQd4wSMXwL61cNVGtXcvd+9r0tyVasX6zWrP7s2dfP6KdZvVe1N+1e+JCybFjc9nLNWsokIEcmYvXae+dbhHAjMWr9Vr8sC+rZydWzh/9+rR3JnCPWHOsvVq/Jzl6tD+bZzRr7EWJy9YrQ7u18rpubVH92bO3tfaTVvV/75fqMb0bKEaOTq3Ppy6WJ+J/QsQuJ3+6xr1069r1YH9Wjm1i974doEOgriyi76du1ItX7dZ7dnDzR5J59aGzdudBgm/mLFUrz9XQa/CnvMr1LpNW52d88vWblIfTF2s564r1sTiNRvVJz8tVQf1a+UsMVTe/FBx6GN6kQKBQCAQCAQCgUAgEBTSDxXKvUAgEAgEAoFAIBAIBEUIcegFAoFAIBAIBAKBQCAoQhSv9G6BQBIDoDwIBAKBQCAQCAQCgUDgGuR/ZpK8E4c+A9asWaP/bNeuXVxjIxAIBAKBQCAQCAQCQSR/FLX0QRBRvAzYvn27WrBggapXr54zxd04ojcIOMydOzcW4b4hQ4aor7/+OpZ7E7gZo2wg41k8YxUFMp7lf4yyQUUez4o0ThV5PCvjOBXjeMo4Fcd4yjgVz3iuLoK9D5l5OPOtW7dWJSXBlfKSoc8AvLy2bd31SowTmIxxTMiqVauW24ld7IhrjLKBjGfxjFUUyHiW/zHKBpVhPCvCOFWG8axM41TM4ynjVBzjKeNUPONZv5zvfWGZeYKI4gnScN5558lbqUCQ8axYkPGsWJDxrFiQ8axYkPGsWJDxrFiQ8UxCKPeVqEehoOwgY1Q8kLEq/5AxKg7IOBUHZJyKAzJOxQEZp+LB6grkP0mGvgKgZs2a6tprr9V/CsonZIyKBzJW5R8yRsUBGafigIxTcUDGqTgg41Q8qFmB/CfJ0AsEAoFAIBAIBAKBQFCEkAy9QCAQCAQCgUAgEAgERQhx6AUCgUAgEAgEAoFAIChCiEMvEAgEAoFAIBAIBAJBEUIceoFAIBAIBAKBQCAQCIoQ4tALBAKBQCAQCAQCgUBQhBCHXiAQCAQCgUAgEAgEgiKEOPQCgUAgEAgEAoFAIBAUIcShFwgEAoFAIBAIBAKBoAghDr1AIBAIBAKBQCAQCARFCHHoBQKBQCAQCAQCgUAgKEKIQy8QCAQCgUAgEAgEAkERQhx6gUAgEAgEAoFAIBAIihDi0AsEAoFAIBAIBAKBQFCEEIdeIBAIBAKBQCAQCASCIoQ49AKBQCAQCAQCgUAgEBQhqpX1DZR3bN++XS1YsEDVq1dPValSpaxvRyAQCAQCgUAgEAgEFRw7duxQa9asUa1bt1YlJcF5eHHoMwDOfLt27eIeH4FAIBAIBAKBQCAQCEIxd+5c1bZt28B/F4c+A5CZpxdZv379TL8uEAgEAoFAIBAIBAJBXli9erVOLJM/GgRx6DOAaPZw5sWhFwgEAoFAIBAIBAJBoZCp7FtE8QQCgUAgEAgEAoFAIChCiEMvEAgEAoFAIBAIBBVQVG3Rqo1lfRsCxxCHXiAQCAQCgUAgEAgqGO54e5ra5db31ZNf/lLWtyJwCHHoBQKBQCAQCAQCgaCC4W8fzdR/Xvf65LK+FYFDiEMvEAgEAoFAIBAIBBUUJeGaaoIihzj0AoFAIBAIBAKBQFBJVdIFxQ1x6AUCgUAgEAgEAoGggkIy9BUb4tALBAKBQCAQCAQCQQVFiWToKzTEoRcIBAKBQCAQCASCCgpx6Cs2xKEXCAQCgUAgEAgEggoKqaCv2BCHXiAQCAQCgUAgEAgqKIRxX7EhDr1AIBAIBAKBQCAQVFCUiCpehYY49AKBQCAQCAQCgUBQQSE19BUb4tALBAKBQCAQCAQCQQWFJOgrNsShFwgEAoFAIBAIBIIKiipSRF+hIQ69QCAQCAQCgUAgEFRQiMp9xYY49AKBQCAQCAQCgUBQQSEJ+ooNcegFAoFAIBAIBAKBoAJhx44d/vciilexIQ69QCAQCAQCgUAgSHEG//f9QjV25jJ5K0WKrdvFoa8sEIdeIBAIBAKBQCAQ+Phi5jJ1ztMT1HGPfClOfZFi89bt/vdCua/YEIdeIBAIBAKBQCAQ+Ji9bJ3//dzl6+XNFLlDL5T7ig1x6AUCgUAgEAgEAoGPFes2+9+v2bRV3kwRYvO2pEO/ndXTCyoeKrRDf9111+m+i/yrR48eZX1bAoFAIBAIBAJBucXydVv879dsTH4vKM4M/dZt4tBXZFRTFRy9e/dW7733nv/3atUq/CMLBAKBQCAQCAQ5Y8X6ZIZ+7UbJ0Bd7hn4L+15Q8VDhvVs48C1btizr2xAIBAKBQCAQCIoCyxnlfq1Q7osSG7ds878Xh75io0JT7oHp06er1q1bq86dO6vjjz9ezZkzJ/T3N23apFavXp3yJRAIBAKBQCAQVMYMvdTQFyfWbdpmbWEnqHio0A79sGHD1OOPP67eeust9fe//13NmjVL7b777mrNmjWB/+fWW29VDRo08L/atWtX0HsWCAQCgUAgEAjKClu3bVcLV230/75GKPdFiXWbk6USUkNfsVGhHfr9999fHX300apfv35q3333VW+++aZauXKlev755wP/z9VXX61WrVrlf82dO7eg9ywQCAQCgUAgEJQV/v7RTLVkzSb/72tFFK8osZ5l6LdsLz819Nu371CfTl8ipRwxokI79CYaNmyodtppJzVjxozA36lZs6aqX79+ypdAIBAIBAJBvtixY4c2ZgWC8orJC1ape96frr8/elBb/afU0Bcn1jHtA3St21ZO9p4HPpyhTvzHOPXn/00p61upMKhUDv3atWvVzJkzVatWrcr6VgQCgUAgEFSyFlL73/OpOu6RL8v6VgQCKyCcdunz3+p66/16t1Qn7NJB/1xU7oufcl+ehPHuevcn/edTX4brmgmio0Kr3F922WXq4IMPVh06dFALFixQ1157rapatao67rjjyvrWBAKBQCAQVCKAYjp1UULDB1n6kpIqZX1LAkEKvp27Us/R+rWqqZsO7+M78is3SB/6Ys/Qk0Nfq3pVVdaBTQ7ZC4vEoX/99dcj/+4hhxwS67XnzZunnfdly5apZs2aqREjRqgvv/xSfy8QCAQCgUBQKHw3b1VKPWvNkrI1rAUCEySE16NVfdW0bk3f+Vu/eZtav3mrqlOjQucBK1Rpz9K1m9VaVkNfXoTxNhssgVnL1qkuzeqW2f1UFDhfmYcddlik36tSpYrati114uWLZ599NtbPEwgEAoFAIMgF389POvTlpZZVIOAgIbxm9WrqP0trVFW1qpeojVu2q6VrNqv2TcShLwa8+f0idd4zEzI602XVQYHjkU9+Vn8+sl+Z3U9FgfMa+u3bt0f6ituZFwgEAoFAICiPDr30hBaURyz2HPrmnkOPZFuT0sT3S9YmVe8F5Rtjf15q/fmmLeXAoTeCmc9/M1dN/zW4nbggGiqVKJ5AIBAIBAJBofHr6o0pbcDKA/VVIMiUoQeaet8vFYe+KPDyhHmBYnObtpZ98pT2vmolVdQ+vVoo+Pd3vD2trG+r6FFw7sy6devUxx9/rObMmaM2b96c8m8XXHBBoW9HIBAIBAKBwCkmzlmZ8vet5agndK5YtWGLps82qZt0/gTFjcVrEjX0zevV8n/WrG4N/eeytak2u6B84pLnvw38t02GIF1ZgPa+alWrqAtGd1Pv/Pir+mT6krK+raJHQR36iRMnqgMOOECtX79eO/aNGzdWS5cuVXXq1FHNmzcXh14gEAgiAEq11asKwUogKBZMnLsi5e/FnqGHMvVB932q5q/YoL68erRqXj/pAAqKEwjOTPO6MLRqkBxPiOMBkqEvPvRoWc/vrFH+MvQlqqU3z6DRIGr3+aGgFuHFF1+s28itWLFC1a5dWyvO//LLL2rQoEHqzjvvLOStCAQCQeyqsnOWrVef/LREzVux3tnb/b9Xf1B9rn1bzVgsNWcCQXnHT7+uURc/N0m9NH5eys+LXRQP4lpzl2/QdNm/vj+9rG9HEAM+m7FU19A3Lq2hBndslNGhx5l397s/qXd//FXefzlC1+ZJxfhXz9tNffWH0ap7i3rlroYeGfrarIVeeWAPFDMK6tBPmjRJXXrppaqkpET3g9+0aZNq166duv3229Uf/vCHQt6KQCAQxIoLn52kRt7xoTrpn+PUUX8fq6PNceOz6UvVk1/+og++N75dqJZJTWNWQNslBF0EmbFo1UZ16uNfqw+mirGeD859eoJ6ZeJ83UKKt50Hy6aYwYWtnvlqjmRvKwAmL1it/xzZramqWS3paDXxKPemQ//htMXqnvenqzOe+KbAdyoIQ81qCdfuxsP66LaDLerXUjWrl5Qbp5mCmaihp7aIwIYtZc8eKGYU1KGvXr26duYBUOxRRw80aNBAzZ07t5C3IhAIBLHWkv73+4X6+ypVlFq0eqPqde1bav7KDbFdA1S5a177wf87DKl9//qJ2rBZDsGoOOJvX+igC7KmgnD89pEv1QdTF6uzn0xvfSSIjhmL1/rf79m9uWpUp3qFyNBvM0oG0HpKUNwgZln7JqUpP/cz9Gs2W3vWC/I71+Omwa/3bALKynMnvzxQ7imYWbWkiv6q4d2bOPRF5NAPHDhQff311/r7UaNGqWuuuUY9/fTT6qKLLlJ9+vQp5K0IBAJBbPh8xlJtoIPqVrdmNb8m7M//mxrbNV6dOF/9vHRdys+Q9Ztl/KwiAC1s3p68KNbPBD2Uagnfn7I41s+uaNi4ZZs/18pD3+JiBrJQhH16t1DVPO2LLUVeQ2+K+iFbKyhuoIQCaN+4TsrPgyj3VRSjnAhyYoyNuv0jdcxDX+rzKc7PBerUSGa/yWkuXxn6xD0R7T6b5ATsnje+XRDreyt2FNShv+WWW1SrVq309zfffLNq1KiROuecc9SSJUvUww8/XMhbEQgEgtjw7byEgvUunRurNRsThykQJyWejC0Ti1bHxwIoL9j77k/UWU+OV+N/WR7bZy5g2aSmHoVUYAeviR3QrqG8pjxQmxnVe3Rv7jv4RZ+hN+6ft+QTFCfmLE9k6Ns1qp3y82b17JR7QX6AACHYfN/OXRlrYJ4y9NyhpxKK8lFDn1S55w49AslRgTKP3/97ohp918dqzcYtju60uFBQh37w4MFqzz339Cn3b731llq9erUaP3686t+/fyFvRSAQCGLr+frQxz/7FLdbj+jr/9vmGKPh67yoe/1aqc1JFqysWLRHHqU3W33lm/UnlIcshQ2gQy5cVfYBmue/mVthar3Lei5TgO/lc3fVtaxkxG7ZXnFq6IEV67fIXClirNu01S8R6xBAuV+9cWsgZduFZkxFBw+CPfLprFg+ExlrOkPr1KhWLin3vA89D3pmQ7mnUiYwyU57/JusggEVFdL3SCAQCPLAH1753v++W4t66jeD26k/HtAzdroyDC6gTaNUOuSCGOv0ywNmLknWHK9YH1/fY54BIUpiecPp//pGDb/1A/WjJ05VFli+brNWuy7vwY9iwBczE++xXq1qaqDHdCCaaUXJ0MNRQB0sIH3KixcInmJM2zSs7bcSIzSoXd13vmiM4TiO/XmZ/zvlsTTn0U9/Vnvc8aGa6zEPyhuQnSf8e9ycWFguGAcKtnF2UNKhL0cq995eSMJ4ueoBjZu9XF36/LeVnn5fUIe+U6dOqnPnzoFfAoFAUGwogQoe6/laUlJF9WxVP/4M/abEYdemYaqxFafwXnlz6Gcujo+GOG9F8j2t9d5lHGKIcQUHkDn5dHrCAXznx3j1A7IBAkS8LLE8ZHSKEVj7N/13iv7+2CHtVBVvnyDHqNj70JNRXqNqiV/CIrT74sXXsxPlTUM7NU77N8xdU+n+hW/m6RpmQnmgcnMg4ID1N3vZ+hTGUXmCKSqI8yRfcKfYSrkvTw69T7nPXhSPxEX/76Beeg/67/cL1Vez4ivRK0akcjcdA+J3HFu2bFETJ07U1PvLL7+8kLciEAgEeWP1Rjh0iUPosVOGqIZ1EkZPde+gijVD7zmOpZ7ons0BzgeLV29UH/+0RB2xc1s/4xbFaSGxnbgwk6mCfz5zqVq5frP/XvPBfObQE9shH4DiN/quj1T1qiXq8yv30oGcfPCXd3/yv29SWnY1/kuMOtnyZqgXCx77fJZmhYCufMHobv7PaW2ZonLFhm3e/VetWkU1q1dT/bp6k1qyFg5Kg7K+NUEOIApz79aJYLQJzGOMMTn0z36d6FKVGvhLOFlxAgFTTh2PChJAJYZBeW0NyhEHbZzsETi5OJsIftu6ckBNp73DpNxn8/zrvOfcr09LNXn+KvXyxPnqw6mL1S6dm6jKioI69BdeeKH15w888ID65hvpYykQmJHWxWs2ptWzCcoPfvKMhpb1a+mWVARycuPN0Cec0D26N1M/zF+l1fS/nbdKTf91raZKRnXCg3DI/Z9rCiBqfk8d0Snj77/346/qnKfHq5sP76uOGdxOxYUZLECBe3lvymJ11KC2eX8uZzLE4dCjxRO6DACzlq1TXZrVzfmzfl29MaXt15oY7i9XEKUW1Fu8s/KQ0Sk2IDh27/vT9fdX7d9D1auVdCjIyDZr0IsNSdpsFdXMq7FetEpE04oV1NWiU1O7vWG2riPGGMHFPjFu1nJ1zENj1e/36qou3ad7Vv8XwenyXA5gK5eLx6Hfmka3d0m5n710nTrukS/VmSM7q1N2y2w38A4fZLOQKB6YFB9NW6KuO6R3aBBm67btvm1VWqOq2rNHc+3Qo83q1V65Y2VEuaih33///dVLL71U1rchEJQrHPvIl2rUHR+pqYvKrp5WEC0L0KNVst+rK4eeaOKNS2uq9y4ZpV4+dzdVq3qJPqBJnThXQNCI6vm4IRSGc5+ZoA/mK178TrnIFCHrByCoFQe4Q782BoeZ0yOhUpwPXp4wX3H/bvWGsnPoKQPXxlO6Fsp99nhxwjydQerfrqE6YmCblH+rWlEo98wo7+g5gT/HxBYS5AcECLPRaAA9/ZdlCYeexjLQoV+3ybqHunDob/jPZP3nfR/MyPr/fsLOsfV5lFhh/7vnvenq+3mrVL7iuXe+PS2lzpvX0FOrWxcK9y4p9//8fJYuHbj+jR8jzzm/bV3V1Br6z2csU69MnK/+9mH4eK9ngQ8ELkZ2a6aO2LmNunBMt0pdR18uHPoXX3xRNW6cXrcjEFRWwPkgJ+GLGUnhGUH5azsDdG9ZzxoNj1MhnCLvdWtW1TWNMKS7Na+Xch+54lMmgtbQq03LhDiDFTzyTuJ1w7xaTsoI5QMc8hB7izNDv3xd0qH/Yf7qvO7tBa/Gs5UnRlWWbXio1WLbhrV9I7MiG0m3vTVVnf6vr1OyY3BWsP/m+tzE3NmnV4u0UgwqxyHaabGC95Lu2rxuGrumsgBzpDx1gvjf9wvVsFveV/v+9RP12qT5kZysxWs2aUcQZ0o7Q3SV0JRa11GG3tAOcRH4Iyc0W2B/J02AfAO4f/9oprr7vZ/Uwfd/ltccueT5b9X9H85QY2cu839GNfRUYhVHhn7tRntpniuVe1D7CeN/WRHp/9B68Sn3nkNPePyL2WnlCBwUoMH/x/Ub1Kmu/nLMAHVQv9a+VkllREEd+oEDB6qdd97Z/8Lf0Zf+D3/4g/4SCOIG1E3f/H5h0Rmkr09Kis1Q9FJQ/kCONMTwOIhW64Jyz+sJd2qRuO5PrCVbLrib1W8vXp09bXbinBWx9CgG0wBZfxzw/dsmVMHj+FwzK2HSRXPBi+PnBtadZwMYQaC74pmPHdLebw9VVqAygtaeQ1+eKKswBJHl+nDa4lg+D+cCDHaUdfzri9n+z3/3z3Hq0Ac+Vx9FZKuYmLkkEZQiR9eWoSfaabGCC1tRuQmxayoLwGw66L7P1AH3fFpu2mZ9N3+VPxYXPjtJ/faRL3WgNAwURG3bqHagJgqVVdB+bAZFXQR4wUCLkvne76+faHo+4cufl6Wsr3wCuF/9nL/QGt/Pf/HYdGjzSO+sQ5M6WYvCBQGfy0Xj0mroYx4nHij/5pflOWXozXvC3+/7IFGyZAMFk8BCqMwOfJk69Icddpg69NBD/a8jjjhCXXvtteqHH35QZ555ZiFvRVBJgCj1uU9PUG9P/lUVC2BkvjJxXqzZxGICapOf/uqXcpX1CBonKofo3qK+nXKfwzOgjuyql75LM8Ioy4DaeUL3lglDelqeDj2no0fpg45n5+fo4X/7Qp32+NcqX3zn0Ro7Nyv1KffLPIpnPjCN7Xwp9yhL4HvK0hzbDa1Yt1md8URCP+aAvq1Uq4Zln6Gf7VFv23tGJjDq9o/UlIWJuf725EVqUp4lBrkCdExkuf785tRYPo+MX4A6DGDdUabpze8W5uTkkVClTVchats6aFSUZweZ9icEKChwgX0k19ZTxSqKOnnBajV98Vr1OlN8L0tARBTo37aBPoeg/P3s13MjOfQdQ/R6qKaZSo3MgJQLyn2tCBn6Rz+dpUvfUGvP25hxmGyCbMc4X/BsM+wbfs6ilIE0NuIICi33xr+RISRLGfvVMSjpm+wO3vowlz70e/ZolvY76KIQZAPSHpOLWGJFRkHfBpx3gaBQgMFE9USI2EINMxMgNvbQJz+r8/fsmkajLhTQ2/WnX9fGWu9bTLjqpe91H2zQVq8/tI8qr0D9GyLvMGi7NC+10tBg9MDAj6qAjgP92tcm60j9oQPaqOFdmvhzmerrOJXOz9AvWpOX4jw/5GGUwwAxexFzwKgzSS8wqhKOfuZnfeuHRTpQdSQTu4Nze9Fzk3zD0m+TFAPl3qxNRI2p7V5hKIDt0K9tg9DnMB29XIMOE+as8J3Kc/bo7DtwcRtdUYHnp1rR4Z2b6KANxhlz/eb/TlE3HNpbnfXkeP3vU2/cr+Dsoec9x2TlhvznhKnPQJnHH73ABc+cZQO8K5w7MFZt/59aNYUFLL+bt1Kd7gV6Zv/5QFW+KfdVNGUYpTor129RPy9dq3q3LozSPWr2MQc5m6SQ4FoXyBTHKQ6aK1Z4pUAQEsU+fec7P+kA5Am7dAgVNgsTxDMdepxpJpw49Gx/GXrze+qJ04aqHi1Tg+e8bSjt6VMWJgLcO7dvqCbMWZlXm9I4HPoFLEhOe/zClRv9MitiImyM4R2uXGd36Fs3SKyRBd514wLOUoKp9zT91zX67O3btoGV3UNspcMHtlU3/WeKWubdO84dJEOwn1Bg38pWrCns1YJm6FevXh35SyCIEzyLRPVDmfDgxzN1b1Vk9ikjVWgK9+n/Su34UJky9DAU4MwD/xr7S1bCPmUliNe5aWlarV91Nt+2ZFEriywh0e64s8EztqXsEKOgEzJEO/3pfynMjqhATR0ZY8i0IQhx439+jGzIPn/WcO9ztkeiDMIgPP+ZCerSF75NKRXgbYZgAJAIUz50dgJlPlC/DAcEhsMCS43eqY9/ranWlK0NwuSFq6zK8NmCnMhROzVTXZvXU/W9TA3U/eMA9rBD7/9MfRSRov79/FXa2ELXBk2/ZfWRWIs/e1TybMQTowAZ7b99NCOUtgtD+Bsvcx5HyQSANlwEv2OB59zkSoElgx3OPG8bRaCsVNjextcC1gsEy/49bo7VkYoKODsIpHEDPB8kjfIS7UR1LTDtHlTfve76WO365w/0syEQdclzkyIxjOICd/a+/Hl5ytwpK/gZ2tIaateuTfX3YJyElR3OiuDQ1/ccegQbV3jXcE255zFVZIJtAqx0XwCdY2S7DenYOC8bCutt7vINsWbofYfeW4cInFMN+aY4KfdG61Nif8W9PniGfhVjPOHd7X33J1p7gFgjBGrZSXoiwE2HJZI3xw5p5wcjOJ2fgxJ1pZKhL6xD37BhQ9WoUaNIXwJBXMBmcuzDSQrWW5MXqeG3vq9enTg/9P9xtfCj/v6Fen/KrwUXaMJm1a15Xb2x5UsXKzb8tDiVOl4WQZV8BfEA7ghlY+i8M3mR//0SdlCSGi7q4njwAI4XxxNjf1HZgjuPf/3NAAV/47/fLwx12Ig1guj5kI6N/OflFOYgjJ251HcG/vd98nl5rfzxw9prhxIRfBzqRFPMFRs9ISBkmWi8vjNo4whsgB0DvDg+ODCC+wHNNuVn6zfnFHyiMW7uZSHIOI3yHjmrKIgCf/XL3+vWhic/Fq0cggwoKNzDSeNZt+b1a6ZkmpBFzoWmDSq5Scu+9c2p6va3pqX1tuYBLV7jjj0xDl0UtJcjLF+3SY8hOfZcYCobEN3eVj/PKfdhbev4e/9xwWp14L2f6bF8Zpz9/UQBdAfOfmq82ufuT7L+v1jvpoI9z9ADvjBegRx6zHsCgn5wHNC6ihgk+WDG4jW69hzMvjCYTBqUSxUaWBtPjJ2t1wcCd+Q8wSnq07qBHh+saxJgC6XcR8jQI4jB14hLUTyTgm4LmPCgGWwnBMBoXx3YvlFeDv0jnyZbieaDCUws7pdl0InZrhZ5e2kiQ5840+MoV6Fgi1lD38rL0ONsiassBuPDu71gn6CgI/+5yQpIdshIjh0YtO9cPFLdcGgf/94zOfRma77KDucO/Ycffqg++OAD/fXPf/5TNW/eXF1xxRXqlVde0V/4vkWLFvrfBIXFnGXrUw7FigQY3LzGC5soDrQwQ93c+NF2CLRHCH5lG42NmhEjWtKlz3+r//xiZiIzeM+xA306dbbZunyyOGUN9FTnoGx9XMBBBsOHlLzzwVQv2GAK4uXq0GtnhwWQeOQ7Sc9LpZbC6eI0e1JJzwZklNarWU31adNAnTS8o/77U18GBwcoyIR6ftwDKeODNp8Jn7Ds9/9+SFLXyQg7sG8rXVOIrwHtEsJ4mTLmUSn3CIaATs/Fo2ziRwgmBOHzGUs1DR3j/s2fxvi0dFvWKhPIMG7qOfREHUZwI4qBvM0T5Trsgc9TsiM2Omo2BrSpOgxs2rJdzVuRdOizyZLjcyFQeu8HM/SeeuVLqZm2ab8m1tI3s1dYHZbht36gnmTzEe+bjLp8wNfYdm8MeWApF6ouObS2+nmgqpeVChMqW8ICDVDrpgAab8WVLWgNwdBGkCAbnPDoVzobzgOsJm2W9kGyKfAeb39rqk/njgMIeI6640PtbCPAQTj+ka/SdDjywQvj56kvZi7LaC+YdGzoSxQa17w2WX9d+/pkHbijcj3syTgbWnhBX7NFGrcXSKgNbLMgULARe9an05dY94e4Ya5xmy20nQX2sN/R+sV5RnRtnhRBIBC2LwUEXxo/T3cDMIF/v/V/+Wt1YN/DfOLrBkwnCrDgTCeHngLPuDa6FZh96rNz6FMz9PVrVfP1d3hgNh+YArrYEuhd8zI0s2yUgoHVWSki7AjYvJizjT12QdCZStdAD3pBAR36UaNG+V9PPPGE+stf/qJuvfVWdcghh+gvfH/nnXeqxx57zPWtCBhAK9/jzg/VH1/9Ia9MBw6DbMTLcC0c+HAW/puD4FDUDfRPryYOe7P+ZvKCVYHPy9uIIFK4e7em2nCE4Bc2fVBxQV89/tHwyD3oxDhYqd4zSh/PlybMU/v89RPteLSoX1P1bFXP33yziS6P/2W56nvd2zpi7xJ4j8i0xfl5Fz47UX00bUma4E2YQBgcc4xNJhVfwq3/m6INH7SQyQc4kMZ7gR6zpg9AzTzRyaKqWX89O1lPbWbo6QC2Oew3Htrb/z4XJ4eMJDLYBnVIZDXC3jsd0ET/z3QA8zXGnRJQiynrQoYYX7NDvdZ1mB/5gBxV1Cr289TzkWHG/kXZig+mLo5UNwmHHhjRtakuCyCWhJm1jwIqJyAFaWQmqJ6SgjhBgMN/5N+/8P9uM9Ka10vOlyj7PFHMbbXxcASxt+ai7YG1vfvtH6p7308oF3MBMYwNBQpsTIPZS9enXItouHGUIpndITAH+bpbu2lL7Bl6MmLDMvQ80PDwJz+nrFWwCh74cIb6xhD+ygQebMsmk4w1QuPyGuu+Qm33KEPfzwu+oWwDc+261yerv300Ux3NxMryAQKj+EwE5+FscyFPlBxxRD0PgjDT+zxzjuEdojSEHHcqPaIgIfayQgpaYq5CKBIwZVpoTwazxmSjcGDf0PorVUtCtQgoQw/c9N8pVlYJ5hUU5/k+kQ9s55kZ6OTMKOwllNVtXLcGs6GS/we258g7PlR3vzdd/ee7Bbr0C/YA7fmwZ5/5ak6ayCEP0mfD9MC+p++ntIY/T1DS+fKE+f6ZTgr0FHhGIOmcpyfod5kNsO7me3up2X4WDjPZD5nOlqigssD2jVFeVCXFnuAsDjPTTmWIFAw0kZFy79fQiyhemancjx07Vg0ePDjt5/jZuHHjCnkrlR4Q2wJVCb12UfuVC7CRQqG55/+9pS5+bpKOxoMyHmQ4Inp6wbOTdEbpT6/+oM57ZkIajS9fgG4Gxxs0UxxAj5w0WP12WHt19f49tOEBh4kbAhxrNm31DxBk6Pbs3tz/N2z6eFf43M9nLAt11qnO8x7PeM0EMmbpte3cvpHefEn8LJtM2GUvfKeZBYjYuwIOvEPu/1xn2r7yAhugZ1/2wre69pocnqiAQ3X0g2O1sYjABnDOHl10PR8MbKhaBwH/D2MThWoOh4R+D/ebT+boH5/9rGvroM0woH3CkDWRbes6YnUgs2DW0PsZeq8OjuM3Q9qrvx+/c85ODhky9WpVM/rVBt83XafUq2HzM/QZqOIwxjHfcfhTRg9ZP7wjGP9AU08MD6CeyFGFfFBrPOK2D9Sjn/6sLvj3RHXu0+O1Q5J06Kuqvm28DP28Veri5yep/je8o1k4n7Cs0yqmEWCCWCO7dUvUp+7VI7FPIKOSLUgdnzL0WPdtPKM6U3YGrS25A2wzfklYEIDAUCZsCKEyIliDoBMhm7lm6zJC6wJOEO19KHky2TM8u4Z9odSbc9jncsHZT45Xxzw4VjsG2NNNByk1Q58L5X5deIY+AuWeO/T8HlASMvSW99Udb0/TZ2g2QGtEApzAqIraPFDEa5rNDH2vVvX19zDkoU1AbcR4gCTf4AvpXhzcv3Xo75K4Vq4gloU5x25+c4ouDQGtH2VAtHcis926QS09j3MJ7BGySa5g/Cix0K5xbXXKbp2sThEFHIP6epMuBj4jyMGKkg3dvHWbrnFHkPaGDBosQXjhm7namaY9wEYNN2n3RN8GlqzZrG0CcqAp4MzX0P95thGCi8TygC1LZyzYQ3945Xvd8g+g8yJKCz0Tt7BuHCirIp0JDjjZxIjCHn7Fi9+qyz2tgKjtS7GXHnL/Z6rT1W/q4BbuldhtKdeKeLZkq0GCJBQp9VNQnzvjpmO+jVTuA4IkdG4FOfS0LutIS+eyc+jbtWunHnnkkbSfP/roo/rfBIUDMkugtgKZ6sRsgJG8110fqfenLtYHOwwEGOTo6QuH1wYYIGAG8Ghn0CGTK9A2A5sANpj/XjBCb2q3HN5XnTWqi+rZKpFJpZZEJl7zIt1wTtAOI0zlfmnARsONJAQOohiE5uYKwwiwHUaZEJeYVhiQHaGo+PPfzNNGCNqsIar8j89mqTOf+CYr2j8CAaYzAgPp/w7qqb9/6OOfdUbK9q7JSH3nx8xUx2eN+tNXLTS7qAENyppdd0hvX7zNRLJ1XTTDmQJNlJXmhjCn59lAwZ9cxp+yTCTI5t93iENPtcWUAfEz9BkMaaJqggVA6wvZHAh+ERoyqmBrL4BBWYdM+NuHM3XAABkkZFje/H6RNrAp8wGHHtfFM+JdIYiE5wRrgxuKpogPAb9LAbgBXqaf9lHodGTDVoJDAJV7UwuBsmRPj5sTmoWl9nIEmzI+9w84XT5Thr62xXiFoZ6rs2tjlvzgsS4oo01AkAKUcOwLMOx5HfvFY3by98Vcgle4Z4wT2lohQzZ72Xpf0R/4xAj0ZUu55/W7XYIy9F4mK0xzgb/nIGA8grKuQcE0fp/vRmRYkVaIuQ6TvaSr+GuLxnn+yvV+gDAu0Hvt3bq+uv6Q3pohk+l3w4BgNMo4oC1gniukpUOZQJxnoPvz9YjgMK05sJv6ew4U74eeDbMO9hSSHVHX1Wn/+lr9/t8T9fdtG9bxA4HAdQf38lk2RLn/NeCdgFEBkH0UhEzdS3jwIxeWAp4bjiyc6UE3vadtw/Vb0t8F7wBkriOukYTOC3Q+Yd+mvZkzDfh5icAVfocy54QTh3dIY9rxwMtbPyxUFz07Ma28Cb/DA64Ikpp7AoLXOAtprNBhAzZVtsAeyktNLhqzk2pu6OsACDrRs8YBSjqACUZBCeh9YL3wwKzJ3Nti6G9EzdA/OXa2GnLze/575R1/BAV26O+++2513333qb59+6rTTz9df/Xr10//DP8mKCy6tqibJgQXFV/OWuYbCLt1TRhDhKDP+8yr4cNhQ20s4mgJYmbZAWR323rZPcIunROO0ic/pWeQcQDd9e5P+vtTvUg36FE4FGCYvHfJyJTf3xiQHaIsIyETLR2bvpl97NU6cbDSYZSN4VyI3u2UlafDDMrJcDhpc8bBPjeDiBnGHf8Pdb/k5OGwoewwDj4wJFD2QN0HzOwFD0Tx3uxB7+VxT1RrV68V3F/fS0Tosy05mboQjs1mbRigNVAQKGgVtZ0PZVB38hzdlBp67wAmB9dEXc94ziWjSMZX/dqUofcUd8Mceu86dF06gDNpEyAYBOzerZmffcd+wSnA3LCkWvaoGQVbOzNk1ImmicwFmBNk2BBgDPFpwAV9OLihSiUKCMBgn8D4jfWeLwogCAgjEQKYaK9E6O5pZ6Ak6agHxwYqZ5u0SRhNuD90OqD75CrtUVrrmTX0/7twd38/JHT0WrFl41Dbltin3j48c3Hq88FYQ09pBAjRxnCNR3uHA4dADBlxfK7jvqPQrLmjd5/HoIIa/TFDEuv4kU9n+U6+vkaW5xOxNOBcBe1JlAUN26ttWgaHD2yjx+KxU4b4P0O2PkrwlItXUXbbJjgKJ5+3noJThWCXLYiU7CWdruGB84Crj+Nzx/zlYx1MhwMGlgSvgY8CGjuU5CCA+NTpw9TJuyb0PrINiODd/+bhL9X/vfqDOuWxr9UXjFWGZ6RXSnMMfdzPfHJ8ytxAEJICrQiGYk8DournoI77vKcnqMP/9rk65qEvdaYcwceHPk61IYLAEyfIrvPuJ6SDkuLQG8kTvAOUMhFLxZbRzQZcDDGXKk4zCIC69vVeQG10j+b+ng2tIZtiOoBWZ8nvd6T0KUewGeuT7+1gw1EC5a53EswLDmTVKXBE10EAetgt76ubPBbC2U9NUK9OWpCmOTNxbiK5xN83zgkCbJ0PL9tDZ6lreXaPyRiKCh6YAJPitBGpe7YZLI6Lck8ZelMwNaFFEpKhz0C59wUYjXP4uW/m6n2AgpF1pIa+7Bz6Aw44QP3000/q4IMPVsuXL9df+B4/w7+5wgMPPKA6duyoatWqpYYNGyb0fg8dGpf6NNVsgAj0Bf9O0JEOG9BaPX36Lin/bvs8RC/JQUHPSQh0hBnOuSKZOUytHwLowIXBayo0v/HtQm2Qd25Wqs7bs6v+GShE710ySn1+1V66pdSfDkxkjIMCETBUqEaUgBqtMOCaZHQ/eMLO6vw9u6o9PKo/6pIAOMdRaYtm1ge1fhCB404rfob2XPSZcHbMWtIwcAcDzjtqvYDebRr46qSj7vhIZ49gMCDzYIrO3PHWNK24DLozaGVwtD6+Yk/1zBm7qBsP66MGtmuoMwIomaADi1NGzZrnTEEpjDeML9zfA79N0NPJKHsuotbBh1MXqz//b6qm2wPIyNjaUhEo0x21hp4OPXLoMDfIESXDsWV9e4aeaPq5ZC2/9dYCZXiiZOiJdk7OFRnytlZwHCTMhPpimt9PfzXHp6libKiGnxsgCNRF2SvM7A1lPXzKvRes4PXNj508xM+aUkAJ7x6ZWmT5+dohCiTeNxkjMMig0Au8mYF2j89CSQocDmJkjO7ZIoV6eOk+3dVtR/b1x4McLzhuKCVAIEzXShpBDswfZA0vfu5btccdH6WUGphZkqDsMFFca3mGEoIrF4xO7IfAEQPbaBXibEuByNB+5vRh6or9uuvvx81epvedu99LBFJpPtz3wQzfkYKQm8kG4doieA/XvzFZ9fi/t7TDmCmgyTPaNFf7t22o9u3dUgd8yYGmIHU2z4hg1p/fSlBsgxzNqG3raM8YuVMyCz24YyN1zcG9UkrB9DNFOBvIwUWQkQIyJpsH+xtK6IiyjGDrCf/4KoXZgfGiwImpcg+09BhECARwUx2fC4cPf4KVApYE9l6T9o/xpPWGM3bPOz/SGVuuN0EdIYCD+rXSZzYFfpPPG84UMplE701ZbHVMiTlmatKAlYX3R2JnyLSO6t7MD0hF2YehnYNuImAV8rkQ1LGCw7Q/kLzYr08rPbYn7NJea7jYgiwcf3j5e51Nfcdzjnq3Tu0VbsNl++yU8nc+9lxgOReH3uwoAdYn7fnXHtxbnbZ7Z/29aavwd8cDOViPOMsosI7gF9kqBJRjDvSCqT/MX51iU1C3DzoTsCYxNzEfseYe/WxWyu+a59N/vk09C3Ce7tmjudY6wHz5+PI9fZp6vmrtnFH22vm7BdolyXM69ewAs8EmDJgJtJ8iQ8/HHOuP70umQ0/BQN62joPeCy85wN7+06LU810c+jJ06AFQ62+55Rb18ssv66+bb77ZKd3+ueeeU5dccom69tpr1YQJE1T//v3VvvvuqxYvjq5CXlGBzES2Gfrnvp6jFWaxcaIG9uoDkk4uARsuid/Rgf2VR0ND5K1BneosAhcvRZxEjChgwIFDH/2esf8j+4WDm8S2aCM/fECblKghenkSFRlRz3+fsYs1cojNZp+7P/b/ftzQdn6tdlit4pueyjcykTiQL9u3u399UKbAEsBGCWMrCrhBiyAKIs7IsPBoPn6G56UMCZxUtDJCOQTGDNnwsLYm5ND391gWhH5tGqRExHGAoqYfmQfUo3EVblNxF04cDiH8eeIuHXx6H6hoZMCPvutj35jE/HqfGWGYw2GZdjJKQbfGmHKDG3X/mUR88Nmoyca7QTTe9vwmamRZQ0+HcqJ/dRX/YEyINUbP0GfDOIAx9I5X37x3r5ZGDb19DiCDROKFNDZUCpCpx+0a5hCbLZIu37e7OrBfgr5OwHyiIFGmLD2e29beDu3IOOUe+MMBPdWe3ZupV8/bTRtZfzt+kF6Dp+/eyTdA9rjzIz3mnJrMKbYcRLtHsCzMqcScPf7Rr9TB933mP485pjDuoIswzCu9oPUGVX6UEiAQ9p/vFvqBGAKCEESPRg0xAhm8lGWZ5+T8UVNa37WWOyUp91VTSiBgyF80ppu64+j+/rNnxRzy1gBqOKERQlkiLox6xM5trP/XD6J4c7zU22Nw/SkL16jHPk84W8ieZnJuySnkwB6LefbsmcPVjzfsq6beuJ+69fB+fjvCqLXm2EcxBgiCnLJbiENfNXOgj/YMCkIDQaU9UQLyPLtN79EcP3JasUaxjq9+5Tt9RqIc5NMr9tRrFuuIgmZmDX2Kw7ByozUAh/VD3UGAI1lrWKyHfte/4+u//DBvlZ77qKlGcI0/A2Fwx8bqg0v38IPgUTP0Zo09dxJTHfqt1j2cB7mAbi3q6QAcssh4LebatAEZXAICXc+euUtaaUQQzAwrssywqZDxvemwvin/RkFR04njyuum3kYQTO2CR383WH18+R5ppTNceT5TYAJ185grxK7kJSWYYziPIHC3k8coNTvhcC0KHuQkxgFnLpg4dmj7FCFA7AX/d1Av/+8Na1dX1VkXGaxZsmXNNUROKJ2rZjLn2CHtNVvisyv3Uh9dvqfuP08wxZuzxUpvre3RvZlOPgWBnhUB9Ic/manXFJ4BTCjYaNmWStB+ixJXc/3xszjNoff3DrsLSnsUvx+sS87AALi9KSiAQ//dd9+p7R69At+HfbkAVPXPOOMMdcopp6hevXqpBx98UNWpU0fa5DGHHk5DUM0oBw55CIpgMR7Sv7V6+dxdfToXj5DjwEdmC/Vg17/xo647A60NICOejMIggToc4lCT5/S/KPCzORaHHk7iYQMTB9JDn/ysMzq4RzixJOQG4z4IXCXUFCtBOQGJIQGo24dThgMgSNgDeMGrlzIFbQhE96Ja2zCAgs5p0jyT/v7UdOo/WqQ9/vks3yBGPd41r/2gDdNznh4f6DSRwXHFfj1S/g3OMhewMmmdXKmcjOQzdu+kbji0t7rtyIQBbUM7L3NHWQYYgHhvfO7A0AwTkKHfpQPt2oN7qcnX76t7qINlgHrdMOoq3isZHDg0Md9/MyQ8EBkl081BgnKgk5LqOQ5cOAnkkHIjgIMy5ZhvUSn+1BILxi2MwWFeSUrNkPtGVo238iEHncT6MlH5aH3C+OE0czgrZ45MZGBMIEsSpY4ehgk5SdzJgCNL843UhDs3q6seO2Wob/Tt3auFNrTO37Obdly4oUY1pjYBQQLR7ldkoN1/9NNifx5n0kXo0CSVQcUz7NgfaU4Q8O+8hSHumwfm6P+DEYE5BbXuKA49cP5e3XRdJt5rXaphz6Il3ibPEONttGAM8uDTKbt2smZsyNmiPZ2LhZplBMvXJsoOgoIqZpslk2qMPR5BHwR34MhiDVCpGMf381apkx8b548N9o7XPH0Y7GdBYk+pGfrgdUprmN8b11lAmQaBWo6FATokJL5Ijgc3lrHP8NaQT479RetP4FbhtGEPRktLnon1Ve7ZmNH5uGj1BqtDj/HnjiwcC9DeKRiG/QR17XAIuIMHyrvv0FsCGyZz1zbOHOaZjHaxFODiDj3mGALRnB0GEcwjd27r75N8PHb22EXIukPw9y/vTEtzknAW3fPedDXFYyXBCd+1a1NdJkjB6SCdHwJ3zo8f1l6N7tk8sM69Ndubw864KNlOM/MLR5QUzvlHR3XoL3kuIQB35YvfpWTo8ZmX7r2TuumwPtrOROCY2viiJIIH2XiGnp8/CBCbddac8n7HUf30O+et+nbp3ETt27uF/3eci1zdHvsKPxd5kJk/Ms5V7G9ILKG16eOnDFEnebX4rS3lOLQn5gKMKdH9EYAIA9k/mP8Q7INoLGeThNmqNvzqJWZw/9C1IGD98sCUWUPvt60LzNCnawLZWm2GBWsqI5w79AMGDFBLly71vx84cKD+0/zCz+PG5s2b1fjx49WYMWP8n5WUlOi/Q3Hfhk2bNqnVq1enfFVU4GAnpx50ozDgcEIvYGxmQzs2VvccOyAlOnbn0f31IUcH1p9e+cF3AE78R7KDwZFeFoYy9KhrNunvMJZAa0JWeb+/fqo6XvVf9djnqfSmIJAREFS/iIwH/zdswjBwYciCygfBnTDw7BSyxdjQEInkLU5Qo42DtUHtGqHq0siA0ia1T6/kIcJBbU6i0PDM2jtOgQNTgK7Jcd0bqWq0E+YkroMMrO3wh8NCRv+Qjo3VS+cM9/8NNGZ+KFOGkcApf5Q5HNShsa73M/UOOOB8Ez76aYk67V/fqCu8PtYwrohuHla/TQd9a895oi4CmLdwXhB1nzg32IgiRw720j9/N0Q9edqw0HsGyLmK0k8c9cL0XpERbUZOz+qNvvEGmh7Vt5soZWsxm8wp6NvAmJ4tfGMtqIYeztctb6a2K6I5Qu8V9xrGECDjFgc2nB6o8w/u0Eg9c8awQJog/+ww0DqDoX3YgGS2FwFLcn5t/dXN7Ph5e6Vm34ihkyIgaBhOeBa8Q64TYEPj0qQzQhktm2Ac0LFpYn5R3S7Vk5LauQkELrgDD10AbviaxpqNleLXbocY90mHOto8w3XJCIZxTJRpzFPKmJ81qrNmbpm9k/n7JEeUzg7oJZh7K4T2UG5wxN+S7fw4bHoENqoxKMt+GYXHouI4+P7P9B5J7S9hwOIZ4VhmqkUmBzgsQ08BCQQXwAq79Yi+vuga8M+Th6TUYmd6/6C4AwtXbvCNZR4ARZCUZzqpG8iIbs18R55KIsiIt2XZeIbedu5hjZtnGWUXeYAKrBh+lrw4fq6f8WvGWjESzADK01/94mf+g94JBx4F1zAdeuzJFEyH8zf26r3U347fWa+B3zGWF2VYqVwK2eW97vpY3fvBDPXsuNSSrnOfnqDLTJBtRFa/g/deeSkBWt+GgRxJ1JbffHjfUME6OFuYl7je0hAdjdII2c40h75uTX1tc91G1cRFUgHA/KRzC6ryn1yxp/r96G7qhF06+OsT7we2Gz4bNibORpw1fN7SGQpnnsRVub0HhiaBdJxIswbAePBMe+0a1VLKCswgN2d28P2QGJX79G6pg9VgkIQF+XiwLlsgSEyCeFxQ1gbzrEHXEm4vZCpVCcrQY2ywHvb39kwEx3jChdhh5v4WVEPvBx1Z21Bb9wiMj6CADv2sWbNUs2bN/O9//vln/af5hZ/HDQQStm3bplq0SHWW8PdFi+yq2Lfeeqtq0KCB/1XR1ffpsOZZKJszA5oop0aaBwgOjbuO6e/TAnl0HZssMppTbtjPF2vhRvLjXoYYNTyomYOxZAKZ/n8aNUvZZugB3N8Hl47StfG0uZMyKmoTMym58swcVGPRIgsHN/WCRT0WZZuplRfKFGy49rXJ+pBFRpbEv0z09ZS0YSCgfjYIyIZQXTNF+kHLJYD2jqgxF+D7wwE9dIY6qFONSYvnkVY40XBYB7RL1jv3bFnfz6YCMET/fERf//PRXo4chrD2WCYQNaeaVmQmCaBigzpNVEFOo8RBj/70mDOgvpEhb/bZRRaUauh46QkybygtOf1fX+t6S4oUwzjg9Ylh8B2PCO3CwBAgIKrf1MskwPCkYERQdp4ORmorFNXRQhCBhALpIA5jFmiWgvceENBDzfG5e3ZJuTcY5UFRfkTlSQmZ1tH+fVupF8/Z1c++5JOhp7kJ4/K6Q3r5mheg3BN11tZf3QTo0hCZItBehjl1+Yvfpu1fhD5t6qeJNpnBDZ6VRVALU4kzUDjISKY9LVNGHOuVKy1/P39lGuWeZ7NsrQDp98MCH6lskPRgFZ4Zhi6MNgR8dr7xXf/f4NDj/5ORTS2z6nqGGXcMjvUYMJSZpcAdOU8J9krqXEPbVOwDJjsBwH7w1FcJR/U3gxOfDec7aA86wCujgLApXwu8LSc5mfM84xWGeZiuBheRCyvNoOshOIUWs8cNbZ/y75gzlIHk7S1t4IEbPJMvtsocZmJSUctIMvJ59hK0Z77PJkXx0mvoUfpha8unO5MwJhtAWhGLWMAObDOe2YajgZaxQWVHYAxiXhzQt6Vm22CvgohdEMOP9ijse3cd3V9//+9xc/UcwVnLQeONuQI2De0hl+3TXWvewMEnu4HagnEtjUksYaHXhvf5fzygp653pvMEn0FnEYLfYYFRP0Ad0jeegPlIGWBa8zb9hjoRsp2cAcS7m9CfBIxdNtlevD6ac0H080QQPnGPaI+Hkj7YOLZn4fsIn4cjGJOU1gEFVGjf4wEAzG2crWQWmtoFXMSOO8a0Z1GZQCaY748jk+glL0HIZE9h7prX4vZCJlHbIKFNUtQnOwD7Lx+XoAx9kMo9lcyiNTDdEzE8OTMmUyvFygbnDn2HDsl6WHwf9lUecPXVV6tVq1b5X3PnRhPMKlb08aKfPJsLwAlCnSdocIfd/3lKdpVviiYGtEvPdsDBhLgZ32y4oUnON7L55HjBGYbTTU4xgN6mmVrC0KZKxp8N2HyQTSZHkKK6YXR7fjgSNe3fX83RRi0ZX3DKXz5nV99AJ/rTv8b+otvScMAQJGXvsEACavLImEP9LATu4GBiM0W2FIJzEIWCeispZp/qKZySg0+4/IVvfYoofu/MkV3UC2fvqqbcuJ+afvP+KRTooGwWOacNSxPPhsPuk8v3VO9ePFJn2O4/LiFq9sLZw7Whghq1c/fomtYFgBz7qKImpkAYRPReO2+EN4410w4j1CojgIA5c/4zE/13wR01Ah1CfI6jxSIygxBLOurBL7SyfZAjFwQKboWpi8PwMeu+iT0AwAFeuDqcmm2uI4zRjMUJHYsw8AAZX9N0YMIQ4ocyP/gP7tda1xwTLRyGAo2RKV5I4A4Ez4JkAn1uUHmOWbIAjQR8PgUP8Qgk/ET1+GEAQ+Hag5L0QXIakVmkfcu2XCkoMc1z6OEIQqxtrzs/8vdXTkOGgNXDJw5OaaOUeh+pDA9qoUUwdUKQOeX7Kpwqfj0EILn2h62ePEqgrbRGNf/5bbRqtAKEJgdUoGcYLenIIaBsJO0xNN9v8QKAKMWhDBqBgkBk8CdKxVKvz//OA5IIsiAYgyUB/Yw/H9lXvXLururhEwcFPif2MVwLmWxQsrnmhuk4JPUQMjtYtOfxDgQmqFY0LDhAhnkmx4mvm/P36mrNfk3xnEwzcEBZeU5X9h36kBp6ejb8nTs1tnIgzFnsVR8wdfiEQ5+Y76bNz9uzEbCGxl61lxbVBOsHY4e962uLvaA7y3jBA7xDBDkwt/CecGYQ6Lk+9xgi5GzzuQzNGwr8AKTEzvdNLmiLscK7wfo5adcOaboIpNFjq/PnoPunUqdM8OvovblgagzAseLU8iCYv0NZZ9NJRPnhHnd8aNXiwXtGa1veLUeXOXnjHdaKjNhjVJKARIqtuwXf5/mz8qA4rQM8A5JUCMbt07tFynxOOPNV/HVovjfeMYk79MQ6aeixNDMhLJFk1o2b4O8eDJxMMANiUTP06N5wxYvf+naAL7RZrcQ/i2g+E3OB9myscc4Wo3MqKMDO7YOTH/taz9tvvDFHtw+C1NCXoSjev/71L/Xf//7X//sVV1yhGjZsqHbddVf1yy+pLR/iQNOmTVXVqlXVr7+mUq/w95Ytkxkpjpo1a6r69eunfFVkgN7EewIT4AQhygYBNRjoOKgSCse9Q6nGdxzVX7elg/osarvu/+1A9dyZu6QtXL6pkCHOHci/HNNfO2tvXrC7+uzKPX1KOjesbPAzqRF64PLNFgeV2X4vCGSMotbPFIzhGzO/BzPq/9SXc7SjAQGl248Krh/H54FueZ6XCYXA3cAb39WOAnqhI9uBGvirPIE7GBcH9W2lDyfy58BEwIYL6u79HyTqZrsxIwuHJA6s+367s08bpjILGD+3vjlF/RUUwa3bfeeGv7v2TepoUSAARvhL5+yq6fgE7hxQ9jJKJjDIsASr4Os/jtHX5YYmP4yIYsppjDAuke034dd9eg49HCMeDIFxcp5HgcymtzJlu4IOSTAu+l73jhpx24f+z0A9BygTASeODmmz1ZoJyixc/PwkNeYvn6T10w3K3CGTzdcnz8LwzCQpfsMhsbEUoDYNzGSUVQ7KbsBRNTM9cTj0NDfJmMM1yFnG3oK5drSXmc2EMb1a6Oxb4nO3pF3fDICShgQJMGpByqfGawcG+ye0OoBVHmUffc/fuXiUvk4QanpjQlTk9YYDaAYgQX/kWXwE+LjT+OOCVSllC3AuzN7JtnZpJmDkUstBqA6blO/bPaV3BDK/mZ1axkLjTqwLomzSfIcj9s2f9lZX7tdD9WuT6kBRmQOtqyVrNvqGsw1coBHBGOy3CFpec1AvLxvayNqrmT/nfr0TdgJ0RSijzvdyMmqJPcIZSkEgh4VKKNJZHMlAWtg68VtFZnDoyYnD+QSDm4xvOitxfRKqQ5kBDzBRSR5AAXAwXlJq6NlegM/nf0eQC8J1NruDHH3MQexVyMbxdlhEsd3HE+vk17ABjhnGFX/2bJVYi2Z3GGTg0XKMNAVwduB8MsXeIM5JgRcKIkdp62ZjUXFGAomztqiHFq3pa4wHRsPEDmlMbcGNMIcegZJLnp+UopcA4FkzsRPN+cgdX1uGGYEwW3ce1MsjwIq2gdyB81uhhjj0CORzoGyJAkuwpfAcOF94MJAH+bjdwq/zl2MGqC+u2iuNsk6vhAIZS9akrjUK3gJIflFwgTLSSHJEBcSAbc+eSROHO8pcRDMIZmKAd/IIy9DDsX7+m3na7jSTZzR3SN+C5jzmJ9XJ88AjBYKDgtnczoIfgvIZ7Iko9wUbhyAq92Xo0EPdvnbtxGRCDfv999+vbr/9du14X3zxxbFfr0aNGmrQoEHq/fff938GgT78ffjwZO1vZQbRRCFgEdQWCo71G78fof7ymwEp/U1t0Ariu3VS9/92Z13bdVC/1laj6dABbfx6dVDtEMmlnpYT/29vtVePFv5BhAACOYiZ2qsla+gzb6R8M9m3T8vIWUN6B+b74rRlgGfLqrM6Qzzrsx4N/5K9E0JTmQB6H9rRcCB7ZFK6YIhgDJ4/a7hP40dfexJwoygnatdNYPOFANI5e3TxszZQkYaAIHq2o82On6HP4qDq6GVxuRFJjkYUCjRwzOB2WlQGGTU8Gx8rCD3xOi0YqKQZAFEdwgWju1kzXkmF9o0+hQ02dpdmpZolwpGNQ0/MgaUBytv/8+rXAZyHEAHatUsiU15ag2XoSTwtg/Fm0pgf+DBd9IyArC+tlSM87QsCp7QhKk9GCh3gQRmULs3qpqkdc/gGQBbv0JZdytSKitMtqfwEgOheNuJD9H9hnIHuhyAnAbWdJmAMEjMIY2ZSQeEQ0p6BjFCmoEYyQ+859EYLta7e+ybgeuRcm58NBx0OPu8MAXDhItBHidacKdBGwQu0NRt5x4cpFGOiZpMDwUF7HQwzDj6n4BzAKdupZd0U0SRyGlIp96kihXBWk+fKxjQ2QqemdSOXzAAnekJWcOIhjmdS/BGsS3RLWZ221wWBghcIvsAYP+rBseqov3/hzxdOxQ+bI0kH2+7Q3/zfH9Wlz3/rB+7I+aO9ExnRRLZ6o3a+4IjjrOfOWQ+vRzfXf6B91pahx/d8jWGeIABrAvOfdACwx6F9m1lPTPXV+Aw+H6OMX7O6tdJYKHBWrnr5Ox1Ewn0iuHOol+3jzjpKGe4/bmDKPoK12KNl/aycpd/v1VVfB0EfSlaQQx9UYsdbaj7w4Uw9F54dNydNdyCToKYJSqBAGwHBE17ilSkrzpEy1iwAwMXmODZaSnLQhtcE1m+U88EMgmD/IjsLtsvEa/ZW3123T2DSCYEeJJpgC5h2l21ekcNI+5CN1cRtm2fGJew66uiTSaSOA+16cf/mu8ykwUO2FOjnh7HsdRDMxAB1hooSHOT7KgUC+NxpWq9Gyt4AG5vWUTYOvemo0z6PpAHXPBFRvDJ06EFf79o1kfV49dVX1VFHHaXOPPNMXbf+6aefOrkmWtY98sgjmh0wZcoUdc4556h169Zp1XtBwgilwwUGADKRxz38pR+RRIQe/dHJOYkLMFSQ7afsBuq7/Q3AcjiQAQkF6aANDhvGDFJEjnBAcaeUK5tGMfLg7JlA6xgOrtDJa4igFYANCu8dom5RgAjo1fv3TKG8nT2qi7pwdLIvLO6JaL/IXoP+j+wr2u2ZDgiEyDKVYUyevypFKO7T6Uv858gkvsKBukawO4DpixPjQ1S8qBFWBIXQ/xoiM2n6DZ5IEtgbidZlG7Sxj0N45E7NdBDgxkN7+zW5JpJCThtSqLowlni/9Gyp4pRJshkBvJYSLJTpN+2vRYAIdbz5O23Rap9eGCSeRjBZKbzXetChjIyHSUOHM0XGDvr2IqADUPY3aG2RAwwjCyUm5z09QRuOFBDwW9Zl8Q7NbG5YxwAyGPi6Rp9hAOsGwmvZgD4Hc+nMJ8anUOWpvZ2JFoztYc5tCGBmMmQ4KNgFhwC14WZ7UVs2kITOTEFKUsHmjrfp0PN/s+3BHGhZynH3u4le8qahb7ZuIuzOxKmCHAoY71TSAZChS6wbONOrNmz2qeJoIwm2EzJ1ZgnNUi+zRgZnVGA/RXtDAOwGel+4BwpKnPiPr3xRVNS7ZwLRREFd/cdns/T6RqD1q1nLdNeRlEBwgAp0CuXewgCCkNsjn85SL02Yp/799dwUJXbaJ2B0g/1B2XkE5PDOsV/T2c8zwPT+ESA4/V/fqNe99p1mkJTPSwQ4hnduas3Ql3rvAWuaAj9gox0zuG2KUwAH708HJfQwzH7zQaCgD9bcGU98o0549Cs16Kb3dFALTgE6nLx87m5+EJJ3DcBz44zhZzOYNFFYRcj2I7APWwABRAjkAmgJC3w3P+GYmy07OeBsIoCA1rL9rntHs+8Oe+DzlJpqmttBbUxNZMrk55LpTM3Q2+1DG+W+xMIEwBhznZog8GCzCQSkMH9Np5/2/YvH7KTHFYkmJFLCgBadmCenjUj8X59yn6EtJsqNAGIOZdpHTeA6bQ1NlUxdcjZsTvw7glNRkkNmYoDvk2EBCwJ1iiHqfYpDb/gJYFXRPsVt4NUZzkHTxuPJBLAeoJMBfY9MwsSVDQWVCKxbt65atmyZat++vXrnnXe0sw3UqlVLbdiQufYjF/zmN79RS5YsUddcc40WwoOi/ltvvZUmlFeZgcMVThDoNCbevmhk5CxqtsBhCoMFGYLfPTbOaohy8T44H6Dng2IOZ5YD2Q20oSPRrTBnhsAjwVGi73zTRa9SUJA4TBVTHnhAP2MA0Xaix8MIjbIBE7CZff2nMar/9e/4egU8CchVkMkJPn33zv5Gi3dCCr6knm8DZbiQoecCUJPmrFT9PZG+KLXIfHO+/tDe6uWJ8zULA5lUqguLSrkPA7I9D3w0QyvVIztMHRsSc6aqDgKEAS3MgOmL1+qMCG/FQm3qSPU5F8q9qfBKhygy5JQVM+cOib186PV7x3VB0w6DaQiFBQBOefxrX0HeRrWEA7xhe2L+3vv+dK3zAA2HMOOP3hfaUh3592QXkX17J4SquMJ9NoAjAUMOmWo4ylRqwYGMyIMejZZnSdF5A0bl4I6Nsq63I2PMLJe5av8egd0G8M6xxjC+GA/unEEVPZMhE2S8/t3TnuAIE0nEPkq0Wrw/OCcQkDKfhdN6iT6K9W867EEBVpvwFPZzQpCeE84cvAMKcAQZ8aBw0p5FhiHNM/xfqDQDEGiEQCYXeIUKtkmNt7U8i8oQ+XXVRt+4hrGPeu1jHhrrtyvFXmbWWdtQ6s1D/L8pC5Msmt8+khCeJc0MIKyumRxs7CM4a/icfJDNF9wzmFxEK8fegn0HYwYjmzqt9PBo6uft2VVrhfBaVX09bz/DGUvZc8A8w/i8bNmgpv77gycMUmc/NT5lzzU1dfA8ECuEcw91eALmBrRYUOaRaV6aDn2CcZJaHoO9yLRpujB7oYnnnOJ8hq4PHNBzDHsjDH8/IanLAHow1iECPsjYQ1fDVFs3gbITzK0zn/wmpWQGNg7e9cfTl+gxxbYdlXFEnUKCEDVDz8FLKxp7mjombDoRNouniqoSiXJPziQFkHmZZpDA2uX7dNeaLxToiwK06MQXgRx6UxOErkv7H/ZOBFySpYnZBa+Bu4/pr89aBHSiUO6DWo0GwbQLePlYFFE82pN8h56tY9Ohxx6/zqL14Qe2Q+xI2NikV0JBB5qn0ORCUiRKmUhlQkEz9Hvvvbc6/fTT9ddPP/2kDjjgAP3zyZMnq44dw6nc+eD888/XNfpoSffVV1+pYcMSdaqCBIIOBRhQrpx5yvJSqztQnNHL+abD+wTeC0TcKFMMxxjZP2QicNDBkeMbRpBSapAYCLJu2QBtSP558mAtYgORpf9eMCLtd246NPkslEUhZ54O7myBDRIRfBgbA9s1SqkR79A4nO5577EDtSP/2MlDQjdCZKQR/YTRgF7EBBhyVD+V7UGFrCwdJLyjQhSV+0yACCFlWJBBo0wmtRDKBKjcgiqN5wXdnhx6GKPAoQOSNVthytQm6HAzhXSA2Z4hAufUlvmhDD05P//9/e6h9b62MpMwMR2iZgeJeHHDCQbFAfd+6iv6Bhl/zesnqdAc33sqz8kMfXbGI+YqZZhsdE0c7HAW6HmpcwcAAxjtdGytyTLBnD/oCnHvcQN1gCLTXooyIrMkByKYVGudaSwzZaMyBWyGseAP7ikoc0wt8QBijFx3SO+MhpIZAKV1gTVE4//M6cO0o20DxoVrlgQFiXjNLLGCsJcQQwrvGNnQ3bo2TaeGs/OA1mBQ/XUYiAIO5g61IINWCVhrT5yWtCVQ2hQU6LGpiePebQ7Pf75d6BvOYeMAp5ucaV4n/LePZvgikIRbDu/r7zP4THLO8V4o8EPnEQJfZ43qkjZH8R5+N7yDZj1xmI5UK/b/aD1QaR+tfwSEMb95MADCnKA9Q9uABzKI0YPrRlk3PJhqg40ejneJ7D/OPgrKIGj17bX7aBo031OyAUr58N4RlIKtQsGfEWy+2gBGgNn5A/MP+/YpXiIBZ3+mjgqE1g4y9CVRMvQ2h94ypRGQWhelhp6tL1MPh7pHmECwHOOXTfLEBDFl/ut1DuIB2aM9RgnNIwTYKJBptjeNAgS7UI5Ce9WmLYluIdDcoRaK2eqecJjsSm7jBen98Dp92kcoecbtAdpXCJxxS/Y5zusoooFgllIJG9l0dVknBnHmy9ihf+CBB3TtOjLmL730kmrSJLEg0Sv+uOOOK+StCBiovtqkjpOwjEtce3BvbSRDKR210c0tPWYJ6LkKoD89HGO03UK/4Yc+mal71xOiiNcAEICDMYJsVliP0CCgzh+GMgxgm8Owa9emWiCQNjO+KQJRsw0mQBcDNREHKn9fvHbVBhicr58/IqOaPz4X9VwcVFdPkWjqk54NiDXxnefg4WDP5LREBZWNgGlCB0dYKxjzeXt7BifEIYmOTob8zu0b+dH9qHMrVeV+c1rrGWJsNA0whHjUG5kyW1baBD/szL7OHDCcaBz/euwA6++EZQaDDK6gwOA4r5zGz9BH0LcImjskUMXx4bTFKX3ZaSzzhRm4OGP3zjrjFmZIkJN9/4cz/MwKRD3xXyCqh6ADsjhRAogkiheEsOwcUYlJdAzlI7wFGTF05ixf5wsWUmZ9p+aZ96WORiCKWpjxVmMIKqATQlBwdUTXZhnnFHeQuUEOp4+M2Pt+OzDl98hhwz4A4/FPr37v97LPxaGnMgpoXoDFgCADNGAABJqgyI9bg85JFJQaTBFzDyTxxTC6Pe1bxJQiFhBa2N3z3nT9PQLk+Gzs3aYwHb0HOJrfeB1YwMLJzLTqo544dWhK+YrpKPHyPNpDOTsGTjM+C1912BwnQS8EeXkAIAqbxUTYOAc5n3iuDy/bIyWJge/zSWrAwRvjlbs88kmiqwjWTRQNGjN4AWeeixdnk1XPRLkvzaGfdyZRPGCjhXJvc6yxV2ZqOWwGmqmcwf/cDOslH/DA53FD26UIGR8+sK3fIUjrpHjBtdp5zh3aF3BmvD91sc7aw94lIMCODjnZtAAGOoXofARl6Hl3FArgJSn3yetiH+bdVzD/aT8mjRsEAkgvJNPaJpuWdEBE1b4cUe6haA8hPBPXX399IW9DYADOytir99L0QWz8D50wWP2yfF1kBdV8gE2Iq1aGAZRAHIbIAmGzpOgvKMFHDUpESbHXQFk/atR63B/HZJ0xzAZUr6Sj60YNbDZ16EHgm3i7CArLUYGMF6jsENDBPICx+s7kRX6GYY8QymAQujWvp7NBd76TqLfFGMYVZaVaKhxwVKsV1aGnjOOXPy/X7f/wJ0D1u7jHV8/bVX0xY1mkGlkCXR+HF7JxvJ4u2WItSBSGC81Ec0JMQyhIa4KylXBMgsSMwupFgwzJIDrzhDkrdTCLRPhyWW9gKSDryNkdAOrzb30zoapOWfRsWgtmAgJnaL/28EmDI83Vnb0sJwVT4OBjbg5q38gXpIRyepTPqhUyBqgzDtMi4Bk3aqOJDCe1FDx6UFv13bxVmm598XOT/N/B/UZRZkYAFPsN7QdbPcVzEkrC2ibDPYgKy+uhg+ZU0DyEHsaEX1aoS/bZKSV4wdcdnHh0aeHZaq7aHhXklBLD5MC+rVMMUWibgDUVVRvCzIYiMENnA95ZFIV7/qzIqlEQ89FPZ2nnCIFHdJfBR9kcKHJ4UWOP3wEbheZAFPB3bo4vjwXSWKTMR1b3CrYC7Qsj2XxAdhr7BoBymWxh7jEYd9I/CNrzsCZd+IRHDGyrs6Bv/5jIhDYtrRFp/ZuBHgTfQE23afRkQv3a4XtuLo4nn1dmZjZMFI8/AwE2XVLQOJoonlneErTPxI1bj+inOwAREHwa2qmJnq/QaFjp6XpkU5YYFrzYtGWbZnwRXp4wT5eloEQWe9MQr0w16hi293SxYGNAnJln6GGXIChhMj+oQ42+Hy9QTZo6pUYwCPYKabk0qF1Nbfe6eKCM5oyRnf1/Q9LA7FpgAvsH1i2xraJoY1VmFPztQPzuoYceUj///LN64YUXVJs2bdSTTz6pOnXqpEaMSKctCwoDZHCQJaeDpl+d6JnIQgGbOdrZwfg5bkh7te9fP9FCWdjIPpy6xK+tyUYoIxunLxdQzTlUkL9g9egwtuLCHUf10z2nw+rycgHaO2EvRkYfh/c9xw7UQR9QM7Mx/gi8VV7coGAGIunEhMhmbKmGkmrWQYnkNF7MvUzMBhPUNg3OPMpD/nBgT98Qpmh10D3yqHcYfTSMcm/L0CNjyenHQYalqdCeep2qgc+L5zFVt2GsQd8gm5aSJvp4GWWzXdy7P/6qtQ+Q8fr48j1zyuaFAaUtZm/uMCDQBRVlaE+gVvf4YQkxSnxPDr2pdREEEic0x+KeYwf4GWKui0GAXQtn4OnTh+nae4hJApfus5M2ruDsgAr8f69N1gYcejkTTJpvGOCIkkNPWgG2TghBVFfsIWeN7Kz/T1BG9dQRnXT9MbGzCMg4f3DZHtb/w9cUd+ZRk5nLvmUq1yNDZ2bKsxF6NIMX2GfB6LhgdFdd0kQBxagOPbBs3SadXXvS0/q4cHS3UAeVWBNgu3HRxKjg88RUB+/nnXkA7S+YjwgWoRb2wjFM/FM7A5s0W41npFHHj2z0qbt1ysnZNN8x2Cnk0Ls+802M6t4sZV/MxKQjXLlfd73fITuKrCaCmaTeD2TDbssUQMj2bDPFL4PeKQm2pd5L+u/BSdziBQXDzgcuqAemF/YNOs/yodRHBTGwQIu//pDe+tzBHK7hLTQENilY3yDPhA2x5PBu+F5wyfPJDgVIFG3wHOts9Ij269PKt3PubzlDB/ZofcA2MdkhvHyM7CubKB6AMaFOOwgO87mH/Y0YkEhmZJqX/do0UBO9wJ7tWoJUFPTtgGZ/4oknquOPP15NmDBB17QDq1at0i3t3nzzzULejsBAMdSkoA0eAQZ8r2vf0vWD+EK0jzth5QEQBNJ902evUPd5Qj+gUPMa13wRtbd2tgCD4O7fJCnZqEMbe9VoVatGbjR5riQMmCUI+YD6zqLNGh1s2SjM8hZgYCSgDCQOAwHOOA5DUObGz1mhJl2zj6HIXiNjhj6qkJdpCCGLgJo7lAwA17z2g3aAz9+ra0Zaalj2pzSEngmFZ4i/mT3jn/nqF9+JyFblHiC6OEoquBgOObP79GoRuzOfC0hF2VYTe+v/pmYUpDQBo52cZbw/UFi5OOJTpw1Tr387Xwcxz316gj+ncB/YC/l+iPdOzj3eIbKmXLTPJnYXBq7fgc+BgZfshFA1Uubsak/ILgjINKM/dBRNFFu2EMwO/P3GQ/vk5MzbWozlon0SlqHHfHj1vN3099/OXZWVQ08CbthPIHoI1hoyhhDyDIO59rN16NFRANoiuKY5n1HbfN9xA1OEaTEfXzh7uA7ukQgpfxdmXT7G6j+/313lCmQYYQ9QUBPXpGBtoR163AtYiI9/MTtlzDKha/N6auzVo9VbPyxUZz81QY3/ZaUWwiXcfmSS9p0rbjuyrw6YUIAwG/DzMUhTx1ZDb6fcJ2nYVDdtA/Ytzk7kQQ0eYHAFPqehzUKgkk1kt/MRxLOVXGn2Qsh5vMQrt6mdg12GYBzYRfgafNO7mu2DL+y3xO5BiSQXuUUS7c//m6pbudpax3F7pWX92mrL1iQzlWxgDNX5eyUDe0GAVsm/vCCl7VqCMqyhv+mmm9SDDz6o28hVr56c7Lvttpt28AWCbIANnWdPUIPnUsQvV5DSPNGGeC1rsQER1yjCTzaYnQcoexkH4LRCZAuUN6J0Nc4iQo4MEYwDfD104qDYjD5uOHPhKjr0g+6R17BHrfs1PwtsBfS3Rpbnm9nLtVI/MmQk7BP2uWFq3WFlB3BmUTMO4x2lPDd7IpcIJPy0KOF88xq7qCDlbGQrOAOA1lQ2Dl9ZAMyMPx/RVzOMsim14fsZnCSIdPHsCd4LxEJ5xpQ72kGAg8VbwuWSoTezehDDyyZDHxUojYoq/gXwtYv2go+fMjRnZ97MQMNJzDfwbWZWuXDWMCYiGOWZ/dZ16zb77TYHd2ic8R55uzNQgwe0yy5Igc8HYwvt36AlYwKK+qaqOIJO3JkHiLGEQGDc4IyllPZ7ETPkcYLEf3O5PgVk0cIUoq3kiEOjJxug0wD2dXRoIBwzuF1OzjxQjdE/4NDadAFsQXtrhn7LduveYWIbc+ihMcF1Jmx95OPCFft112fMzYelB2sT90IOfaK8DoiikxBln/h1zcaUGnazFHaeV65Tu3p++Vnah0CvP/bhL9Xed3+sAxTE+iFAS+TBj2f6YprmePH9F8HQVpbWin8/fpA60WilbAOYWPw8E8p9OcrQT5s2TY0cOTLt5w0aNFArVyZpFQJBVKD2j1qXkPhMeQOyh8hoJFroKNXJ0sO+MgCODN4FVOSfO2t4rMEXGJioZT3ovs/8n2XjlCOb/+LZu2qDvVsWTk0mBDnNy9eHH/pc/Cdq5hmtyc7fs6s2BO7y+oIjyv725EUphhUJy5EqvQ13HdNfffrTUv07J/5jnJ9p/e66fTIK02AshnRsrP53YSLDhgAC6plJHC+XGnoEkXCwIzuwYOVG9dmMpeqlCfMzCl2VJ6D1Vj7OH4yZIKOVtwqLSkuEsJ0pMpiNUCeyjsjMX+117kAPYxsNM0h92hUwV1rUr6lbZMbF2EInE2RYkenPF6azzfdBct5s3SJs4ArSVKvaKELA6Iid22qhKQT94HznGnTJdw+/4+h+WuQ1rC97rkDgkOjYA9o31B0XsJeWhbAWWBNgqKE8KFthRgTw4BiBnURlOyiRzBb79Wmpv055bJxPic4nOGVmxHHe8qB1Nn3oqX4+E+WeVx/h3rMJ9OWDc/foqlsXBr0vCiwkMvR0tucXOAKzc9ys5er2t6b5rKDz9uyivp+/OoUBR0HtfDsGEXMOAQmUyAEo5eSCszaYTvYOlRykhIZM+lztz5gmYcC+hGAfZemFch+Ogu5sLVu2VDNmzEhrUffZZ5+pzp2TqqkCQVSAIvrW5EU66mvLFJQH4BC4bN/uvmPUuam7WvLyDoiLuQJKAlDOgJpg0PayzdoSbT9OBNW/Z6qhR6YU6tQ4EKNmHnCgY55BMIfj/SmLU7IiVONntv3hQPeEIz2hSer3i3vNxRg+aXgHX2U8V8o9ZWrh0EOV+/o3Ev1ps9UZKDZwhz7MmOFGFaelhsHmRJksmkz7GvQFINKEkqLFq5MOPb+fQtS2mnjrwpGa7puLqr0NV+3fQwuv5tq+zIRNG8F0kKOInqUq+of3BOfA+PzxwEQHlrIEgi8unHnTMUSQAx0XygpYK+hr/n+v/aADr9kCDh0c+igtKzMBXQbigHkuYS5SoCA8Qx++H5SGnDEUtCIUyqHPdN90H+j2QaK8+WboQYPHuYkacupJD8FX3gGHI5saehuo1IEHEhes3KDnHYL51xzcS13z2uSMDj3X7qFShN8Mbqee+ybR8hPIhuXJafelZRCMKyYUNHR+xhlnqAsvvFD3gsfiWLBggXr66afVpZdeqs4555xC3oqggoBUpZERLM+0W4iPQODnsAGts6pTFWQHULxBLfzPBSPKRflFkEMB6jsQ1qbxyv16qJOGpwY/o8B87h8XrtaijBw4oKOKKJIgWa416mgPyQ3QXGlzCCwApjNfFnWxhQKvow4y5EyE6BkG0oBJdT6XNUP77pI1G9VaT+WeBx+ovVShVKgpc52p93Y2wHuJy5nnyvn5guY9MmkLPcp9HN1TKgK481ceAn4H9mulJvzf3qGB1CCYug2ckZMtEFgAU+7hEwepfGCuZxszxFZDz/+bWX6FPS4sAGgGK6vH1PY2X1D5AW9bl3cNfbWqmprOzwDQ4m2BQLyzfII8nDmHoDmBsvPQyQg6Y81xB+PHZHuZLQWj6INw2v3Qjo21mGx5tvHLAwoa7rjqqqvU9u3b1ejRo9X69es1/b5mzZrq8ssvV6effnohb0VQQYBWNziYuLJueQQCWIhwCtwCGWTQCssLeE9moiAi0E9UUBsdLV9kagVDAbCoDvphA9vo+vuo6uwmEKU/cXgHTR0Ecm0TCbohV2TniCo0VcyISjdsE3FOIRv//FnDdQkM6uEz9T0PAokgUccRUxTv0n0S9adQ+hckg1OmcCTB1ikiU4ae2hFGpdxXBixatSlvVlB5AS/FgJBgPs+DYFccTDlTCd2mDWA69Os3b01RLcc6oFLEKB1QTGeW1OXLGtW9siI49HFl6ClwAxuBmA84O02WAspOkcjIRyeEPpsCs4SPf1ri6+bUCsiqm21OYZO/cu6uKUxUM/iTTYcGBCueO2uXohDtLmsUNLyFAfnjH/+oli9frn744Qf15ZdfqiVLlugaerStEwhymVP79G6ZV8RaIHAFMzO0fP1m35BHNiKOQ99E0MGbq6o1MpOfX7WXL3CXC04f0VlTlndu3zBNLCsqoPp8xu72c6IshK4Kga3MgM1EqXz0pME6y/6nA8OV4zmGdmqsMypgTeQqdkl13Ki99GvoGTUSpSBnjeqSJohWmXHDob21GN3l+3ZP+7dsmAC2Fmgu9pRiBAVNKwJ4ttPsTFFooDUw9AhuMc4DWwbXpNxf8lyy5RpgCnNmYm+ZyelsMr0uUb0ata3boRat3pSRfZcNuAgeAjn8TID4LNiIyGLnCxKR5JR7CiQM7xzM3jIDiLDJB7ZvlOLoc9YF/PJs2VrizJejDD3a01133XXq3Xff9TPyhx12mHrsscfU4YcfrqpWraouvvjiQtyKQCAQFAxNDYoYaufPfmq87wi5OKi4oB6yrlDetdHgs0G+VDcYXnce3V/li0v27q4zET/9ukY99eUc39GN2tqv2MBZFJm0FNAaD1+FRqnnvHNHQ8SLwoHgBujXtvUPBfMLn52kTmFtsYJg65JRUctPsgUJ0SKDWewAywnBnzvenqbbBZYlwLSxsW1swqSmKB70jsLYRJkc+mGdGuuOKWVRQx8GEv5Ei7l5K9bHyr7jjLYuzUq1ThBawyIwDqZdXLBR7smGQMkHtakzgcBkJnAHHh18xEEvYof+mmuuUQ899JAaM2aM+uKLL9TRRx+tTjnlFJ2hv+uuu/Tf4dQLBAJBRQJRYjmFjcSNcs1UZ5Oh79a8nq6h5wdr7zYNnAlRuQayvdAVePiTmf7PurWo67RlUVnisn26q0c//blclZGYIOcdbaeoxlXaC2VGkFELBXOUQmTDjuAQhz6Be48bqB77fLY6Y2TFEFw+b8+umlnVrlE8onauzzpTIM0G6FygA8AP81dHotyj9zvOgF27NC1XDj3dB0plSMwyaulTJvAyA7R+RMYe2f9eMdsP5NCjBMtsLYl3bqvdx7ELob5MqMo6nWRDtxeUQ4f+hRdeUE888YQ65JBDNNW+X79+auvWrerbb7+VSI1AIKiwMEXx/v5R0hG9dJ+dnFyTt6/ZvG27prlPmLNSHTO4rTpzZOeiaPGWCbyGFEGLigq0XYur9ZorlHr18qiPpfJOydAXBjaHpjyIgZYHgM593SG9VUVCj5ZugsBxoHfr+ppOzUu8baJ4HG0a1lIP/HZnNeqOjyIFAjHfjx/WofxR7r1a/tnL1vnBjbjaIyKoi4DHRWO6+YFA1LTHjWQN/SarDQMB6j26N9NdEqCpk/g/1SMF01My9DmWdgkyoyCrYd68eWrQoISiZp8+fTTtHhR7oV0IBIKKDDg2L56dzLYhiwm8d8lIZ8YZN+h/WbZOPX7qUHXNQb3UHw/opevQK0IGjxt+O7WQ2uyyBBmuULhP9qEXo60sAPVygaAs0K1FPfXVH0ZnbFvHMaBdoxTHtzRLJxj07fIUWKPytjjFbvFeoWFz9OB2qhBBcjMRTywgPOPjpwxVfzigZ9bZdl5DLxl6dyjIati2bZuqUSNpRFarVk3VrStGmEAgqPigtisEtIuDY+0Sj508RGcN/nRgL02JO3VEpzQ12mIGp2aCci8oO5DzDmd+3eb0PvQCt6DWXyjh+euxA+R1C8oMphCcmaHnDvhbF+2ug8s8+Eft36Ii184cccO8b9DUi1l4kaOxYTdwhzxTiYQtQy8OvTsU5NRFXd3JJ5+sM/PAxo0b1dlnn61KS1PrOF9++eVC3I5AIBAUFFyo7hCvT6tLoA3ZD9fvm7NyeXkHV3yvyJT7YkApieJt2uq3VBLKfeHwv4tGqu/nrVT79m4prEdBuYIpilenZlW1eX2irp7OJr6XV1HZOejlhXJvMgW6FmGQGUKdaCM4i7XABBpaVOwJHSK2yuN96MvLmFVEFMSh/93vfpfy9xNOOKEQlxUIBIJyAe5YR+3/Huc1KxrQi9nW1kdQtqJ420QUr+DA/Jc1ICiPMCn3pTWqqZXrt6Rk17mDmG3TFwSx0O3EbA9bFl0IOHYq0iDzXj2aq398NivlZ2ElembbwSBIhr4COfRoTycQCASVFTxDjyyFID/0bdNAiwR1bV5xFe6LBUlRvG1+/aVk6AUCgalyzwVbeV11rg797t2aqVfO3VV1jOhYuoLZV71Yy8BGWxx6WycNBBDnr9ygW+hl09YPkAy9O0ihm0AgEDgGbyUXl/ptZQayOheNcdMlQJAdyHnfun2HL/ooongCgQBdVrZu2+5nsDmzqk719HMwF6Hsge0blfmL5k5qreolur1cser9mGhsUO6B18/fTS1ctVH1adMga40BcejdQSxLgUAgKGSGnhk1AkGxo46lTZqI4gkEAmDj1u2qrkFJP2GX9laR1k5lnGmPI0MPQTwb+6AYYHO2WzZIFToEmtStqb+iIlXlXuwfVxB1AoFAIChghp6LAAkExQ5k35CVIsB2kzkuEAiAT39aonb78wfqncmL/LZu+/RqmfJynjptmDp1t07qpF2TPeaLCdS2DtipRXHWz/PuAxxx6HOk9KEvJ60GKyLkzQoEAoFj8PY6Ul8sqGgY0zPZ/xx19LlQZwUCQcXDOU9P0PXWZz45Xm3Ztt3a5m1Et6bqmoN7FW32ljv00HUpZvRoWV8dOqC1VfcgV1RlNfScrSiIF/JmBQKBwDGonRcglHtBRcOdR/dX/dsm6inbR2xlJBAIKhdQT286wBUtYF/sDj2wdO2mWD9PMvSFgdTQCwQCgWN4doxGTenDKqhgqFW9qnrh7F3V41/MUj1b1S/r2xEIBOUQRLk3VeGLHVzFvXPT4tQB4Nita1P1+Yxlqnm96HXyUWvoRRTPHcShFwgEggJm6IWOLKiIgKF25sguZX0bAoGgnGLr9oqZoa9Xq5rq0TJRO9+5WfFn6E8b0Uk1r1dLjejaNJbPS+1DX5xlFcUAcegFAoHAMbZRg26BQCAQCCohtlKG3qihL3aUlFRR//n9CFVSpYr+vtgBp/uoQW1j+zyeoZeSQ3eoWGEyAx07dtTZMP715z//uaxvSyAQVOIMvUAgEAgElQ0kilfRMvTU7aMiOPMuwAM4cYjsCSpphv6GG25QZ5xxhv/3evWKu6WEQCAoPohDLxAIBILKjK0eU606qzkXVHxwlXvorQjcoMI79HDgW7ZM7XkpEAgEhcRhA9qof4+b69fZCQQCgUBQWQDadUWl3Aui19DXFofeGSp8mAwU+yZNmqiBAweqO+64Q23dujX09zdt2qRWr16d8iUQCAT5YFjnJur9S0epV87dTV6kQCAQCCoV0NptiyeKJw595QKvoa9do8K7nWWGCp2hv+CCC9TOO++sGjdurL744gt19dVXq4ULF6q//OUvgf/n1ltvVddff31B71MgEFR8dKkA6rcCgUAgEOQiDEtSMkK5r1yQDH1hUHShkquuuipN6M78mjp1qv7dSy65RO2xxx6qX79+6uyzz1Z33XWXuu+++3QWPghw+letWuV/zZ07t4BPJxAIBAKBQCAQVLwe9IBk6Ctzhr5C55HLFEX3Zi+99FJ18sknh/5O586drT8fNmyYptzPnj1bde/e3fo7NWvW1F8CgUAgEAgEAoEgPlRElXtBMKoxUTypoXeHonPomzVrpr9ywaRJk1RJSYlq3rx57PclEAgEAoFAIBAIgiEOfSVuWyeieM5QdA59VIwdO1Z99dVXas8999RK9/j7xRdfrE444QTVqFGjsr49gUAgEAgEAoGg0qBKlVQKtqCS1dCLKJ4zVFjeC2jzzz77rBo1apTq3bu3uvnmm7VD//DDD5f1rQkEAoFAIBAIBBUKj50yRLWsX0sN7pBMnDUpreF/L4J4lQ88gCN96N2hwmbooW7/5ZdflvVtCAQCgUAgEAgEFR57dm+uvvzDaHX3uz+pb35ZoX926IA26p+fz0r8giTnKx1KQMvwIJR7d6iwGXqBQCAQCAQCgUBQdk7chaO7+d9v3proRS+oPNi6PTnmtWtULdN7qcgQh14gEAgEAoFAIBDE41ywTHyDOtXV7/fqKm+2kmJ7smOhqlVNHHpXqLCUe4FAIBAIBAKBQFBYsAS9n6WvUbVEdWtRV4aikqFb87pqTM/mqmndmqpEBBGdQRx6gUAgEAgEAoFAEAuGd2mS6mxULVG/Z9R7QeVBlSpV1KO/G1LWt1HhIQ69QCAQCAQCgUAgiAWDOjRWz565i+rQpI68UYGgABCHXiAQCAQCgUAgEMSGXTqnZukFAoE7iCieQCAQCAQCgUAgEAgERQhx6AUCgUAgEAgEAoFAIChCCOU+A3bsSPRbWL16dSHGQyAQCAQCgUAgEAgElRyrPf+T/NEgiEOfAWvWrNF/tmvXLq6xEQgEAoFAIBAIBAKBIJI/2qBBg8B/r7Ijk8tfybF9+3a1YMECVa9ePd16obxGbxBwmDt3rqpfv37enzdkyBD19ddfx3JvAjdjlA1kPItnrKJAxrP8j1E2qMjjWZHGqSKPZ2Ucp2IcTxmn4hhPGafiGc/VRbD3wU2HM9+6dWtVUhJcKS8Z+gzAy2vbtq0qBmAyxjEhq1atWm4ndrEjrjHKBjKexTNWUSDjWf7HKBtUhvGsCONUGcazMo1TMY+njFNxjKeMU/GMZ/1yvveFZeYJIoonSMN5550nb6UCQcazYkHGs2JBxrNiQcazYkHGs2JBxrNiQcYzCaHcVwCAMoLozapVq8p1hKkyQ8aoeCBjVf4hY1QckHEqDsg4FQdknIoDMk7Fg9UVyH+SDH0FQM2aNdW1116r/xSUT8gYFQ9krMo/ZIyKAzJOxQEZp+KAjFNxQMapeFCzAvlPkqEXCAQCgUAgEAgEAoGgCCEZeoFAIBAIBAKBQCAQCIoQ4tALBAKBQCAQCAQCgUBQhBCHXiAQCAQCgUAgEAgEgiKEOPQCgUAgEAgEAoFAIBAUIcShFwhiwpIlS9SOHTvkfQoEeWLu3Llq27Zt8h4Fghgg60kgiBdr166VVyooVxCHvhxj69at+s/t27eX9a0IQjB79mx1wAEHqLPPPltVqVJFxqucYtOmTWV9C4IMmDVrljr44IPVcccdp/vCSoCs/ELWU/mHrKfiAdl5Esgs3/jll1/Uvvvuq6688kr9d7HPyye2VkL/SRz6cooLL7xQHXjggfr7khIZpvIIOBtnnXWW6tatm/ruu+/Up59+qo1cGa/yh4svvljttdde6tdffy3rWxEErCUExLCWZs6cqb755hv9cwTIBOUPsp7KN2Q9FRcuueQSdcIJJ+jvq1atWta3Iwix97p27aq+/PJL9fHHH2tnUey98ocLK6n/VHmetEgwZcoUPRFfe+019e6776qnn3660kWZigF33XWXatiwoZo0aZL6+uuv1YMPPqiaNWumfvjhh7K+NQEDnMPDDjtMvfXWW2rs2LHq8ccfl/dTznDHHXf4a2ncuHHq2WefVR07dlSff/55Wd+awICsp/IPWU/Fg4kTJ6q9995bPfXUU+q5555Tb7/9tv65ZOnLF/7yl7/4Z9SECRPULbfcoqpXry4JgnKGKZXcfxKHvhxOyFatWqnHHntMR5kuu+wytWXLlkoVZSrvWLdund4s/vrXv6qvvvpKDRgwQLVv31799NNPPkW4smwg5R3z5s1TrVu3Vo8++qg2dHEQz5gxo6xvS8AAx/3uu+/WWY+dd95Z1a1bVy1YsMBfQ7KWyg9kPZV/yHoqHiAZ0KZNGx1o/u1vf6vtPcrSS7lR+cD06dO1g3jPPfdoe69v377669tvv/UDLzJW5QNTKrn/VGWHzMQyhUnZWbZsmVq8eLHq2bOnrs3ebbfd1EknnaRuvfVWofeUo3HCsiE6MP5t5cqVeqxOPvlkv7ZKUDZ1U9WqVfP/jjrspUuXqi5duugx69Wrlxo2bJhk6svRGPG1BAMJxuygQYPU7rvvroNmgrKDrKfyD1lPxQuUgMHeg4P40Ucfaaf+8ssv1yUttBcKyhabN2/W2Xg6o3Beff/99+qII45Q1157rTrxxBNliMoI4j+lonKELcopbrjhBnXaaaepG2+8UTvyQJMmTbQzD7Rr105dffXVmu4zZ84c7VBK/KV8jBM2d4rOkqNfp04dUT4tQ1xzzTXqmGOOUb///e91pBaGboMGDbQzT2N2++23a3rjJ598Upa3WmlhjhHWEF9LMGDXr1+v2rZtq1asWCHCa+VorGQ9lT/IeioeICkDR/2hhx7STiLQokUL7cwDYPr97ne/U7fddptas2aN3guFnVT241SjRo0UsWN8j/JK6CWRMKjY5YWH+E/pEIe+jFrIIAP14osvqtLSUvW3v/1N7bfffvrvfHPAhn7ssceqfv36afoIICJR5WecyJHHRt+4cWPthKC+io+hoDDtAkeMGKFeffVV1b9/f/XOO+9olfT77rsvbSygoL7PPvuoP/7xj2rjxo0yPGU8Rvfee2/KWsJYITDWsmVLXcJSs2ZNWUsFhqyn8g9ZT8WDadOmqd69e6t///vfauHChTpJA5V00Lf5+YQa7d/85jfaWSTqvaD8jBO390Drhs7LZ599JkNUYIj/FAJQ7gWFxeOPP75jwIABO1auXKn/vnbt2h2HHHLIjhEjRuyYNGmS/tmWLVv833/jjTd2VKlSZcfHH3+s//7222/vmDZtmgxbORinrVu3+r9/ww036N9fsmSJjE0B8frrr+/o2bPnjjlz5ui/b9y4ccdFF120o1OnTjs+//zztPX0ww8/7KhevfqOJ554YsfmzZv1+vrss89kzMp4jLCWtm/frr9/6qmndrRs2XLHvHnzZFwKDFlP5R+ynooHd911147hw4f7Z9DChQt39O/ff8cxxxyzY8aMGfpn9G/YF++///4d9erV2zF58mT9s48++mjH8uXLy/AJKgeijNO2bdv0n5s2bdpx6qmn7jjggAN2rFmzpkzvu7JB/KdgSIa+DIDaeNTkIOsL4M9LL71UZ6NAtwJQY0qR29GjR+vILehYu+yyi1btRs22oOzHiYvX1KtXT23YsEHThyVDXzigBnHt2rWavghgfNACrU+fPn6mg9dsIwp//vnn67EcMmSIOvroozXNW1C2Y8TrRTFeyNTj/wkKC1lP5R+ynooDKFOZPHmyat68ub+/gX0EhhjKKP/xj3+k2HvYFw844ADNZjr++OP1n/i77IPlY5yQpUeGHjT8pk2b6kw+RFzF3iscxH8Khjj0ZQBQfbGB80165MiRav/999e1iu+9957+GW0S8+fP17Xbv/zyi663gpDK0KFDy+LWKxWijhPV/4KOD5owxkdKIwoH1LnBUYTqLKF79+7qlFNO0Wvn+eef1z+jGji03sJaglgeBPIwvmgdJCj7MaK1NGbMGDVr1iwJXJYBZD2Vf8h6Kg7AfkCdNQL9OH9of0MQGeV8oHOjdR239+BcLl++XO+VPXr0UIsWLdJ7paB8jBPZEUi0YYxgT4i9VziI/xQMcegLCNoIkGlHiyb0XOaAEYsI7fjx4xODU1Ki63qgfIo2TlDWfOSRR3QmWFB+xomyv2BNnHHGGTrKKxFb96B3jL6jP//8s/riiy90ixICDmIIDb3//vv/3969ANlc/38cf/v5ueQaco/IpXJZ11YtfihGU4gkph23IaHcUqaaGrbkNm7ptskQcsldJQblnjak5BZGKHJtsmVXos9/3p//fM+cXaylXef7Od/nY2ZnncueOee8fM73fL6fz+f9Eb2vtic9o96nTx97Nl7bkxa+oT35JyOvLelofv/+/aVy5cq0pWzK5N9mRXvyf0a0p8jxOoU9e/a0J//1eKOjv9ph9zqLOvrrbaOq7Wnr1q3SqlUr27ncuXOn3W6V45O/cvLalBYu1BPSWvuA73vZj/5TJmQwHR83YP/+/aZx48Z2fa7y1oSmX8fboUMHU6dOncvWWzdo0MD069cvdDk5OTm0Xhv+zMlbV4Wsp+vYjh49alJSUi6rWRCe0zPPPGPuuOMOs3379jR//9hjj5lOnTqFLusaRc0e/syItpS99HgS/ll3tc892lN0ZER7yl7eZ96VeFmlpqaaJk2amObNm1+WZ6VKlWztHc/p06ep5+LznLzjW/jtyFpXem/pP10bI/RZOAVO94vXKVJa+VJHAFX4lkx6Zk/vp2f6xo4dK3v37pUJEybYvbKVnhHUkd8iRYqEHlfPzmpVaPg3p/D96ZE1dNTp6aeflvvvv99WptdlDjpqoWfOvREpzUmnX+lUuDfffNPm9/bbb9vp9OH0DLpHc9NRX/gzI9pS9tA8tGaBrsd9/PHHZcaMGaHPPW8kivYUfRnRnrIvK53ppXuR6/cJncnnjdJ6W9JpVpqPfm9ISEiQdevWSWJiYuh+ui2n1uXRHXKUXq/bFjds2DCbnnXwZEdO3hp7ptlnbU76XXvx4sWXvbf0nzKPnkgWGDVqlO3c6UFVO4H65VbXPXn/Gb0PAN2iSe+3aNEiKV++vD0g69pRLXj3ySefyJAhQ2T//v12yhWyHjm5QddUa60CbQuzZ8+2WzbqViUvvPCCvV0LFXrtSZc36H20jU2cONFOl9P2o0VsBg4caPeb1y/HIKOg0qnZWvxRT0zqMaZw4cL2s1BPxoRPIaU9kRGuTb/bae2VHTt22O96+ltPxIwZM8bergXTvPakhT1XrFghTZo0kaFDh9ofbXcbNmyQ119/3U7b1rXYig6iGzkhay1fvtwOWuqxaeHChXZ5cfrtu72c6D9dQyZG8ZGBKVOmmJiYGDNv3rzQdQkJCXaKTvg03969e5sSJUqYmTNnppkGp1tm6dYXul1G/fr1zddff837nQ3IyR1z5syx28XoVG5Ply5dzCuvvBK6PHjwYFO0aFG7vVl4e/r+++9NfHy8admypW1TmzdvvunPPwjIyB26DVbTpk3NuXPnQtMZ33vvPbsV6sKFC237efHFF02RIkVoT2SEa1iwYIGpXr16aEtN3dZ22LBhJm/evHZLVNWxY0dTpkwZM3369DTThydNmmSX+tWsWdMe45KSkni/swk5+Z9uBd2zZ0/Tv39/M3LkSNsH0mNTON0ikP5T5tChv0FeJ+LMmTOXrfcYO3asqVGjRmjvSr1937595uzZs5f9vef48eM3+lRATlFFP9Dz5csXunzs2DFTu3ZtM378eLN+/Xp73cmTJ+1aU0/6Nhje1kBGQTZw4EDTqFGjNO3k3XfftR16rQ+ixzBtT+FthvZERkjL+86mxyftrIfTk88PPvig+d///mcv68DM1b7v6b8PHjzI25tNyMkdepzZtGmT2bt3r73cvn1707p1azswE34f+k+Zw5T76+RVPPemg+i6Gm+qlHedTvPZvXu35M2b117W26tUqSKFChUKPU76tW3e/szIGuTkVk5eBVOla7J1Dai2I50ur8tT9PKyZctsdWdd56aXw6v/pp+uGN7WQEZBbk/aTvRY9Pnnn4fayaZNm2w70uPUp59+KsWLF7frRD20JzKCyIIFC2zlc90dxfvOplOAdY9ynY7t0csvvfSSbN68WVauXGmPXbo/+ZW+7+m/K1asyNubhcjJvZy840xcXFxoW0Zd5vDLL7/YtfRef4r+03XIZMc/8BYvXmzPyuo0359++inDCrI6Mq/VZ6dNmxb49+1mIyd3cwqvYqrXrVixwlSrVi20E4GaPXu2Hb3/+eefI/K8g4SM3M5Kpyqq3bt3m3bt2pnChQvbqcAFChQwsbGxdmcCrYbeqlWrCD/7YCAjd+gxR5dIajspXry4adiwoZ3Crb799lt7XBo1alSojXmzLNu0aWM6d+4cwWceLOTkbk6LFi0K9aPCZ4T17dvX7jawevVqe5ndBDKPEfpMmDVrlowYMcIW6qpWrZotKJRRBVkdDdGCG6mpqddzbgX/Ejm5nZNXnEtVqFBBfvvtNzsa0rlz59CIo47ea0VULXADMsLV25Meg3SU45577rEFWHWnjttuu00++ugjSUpKkjJlytiK6drWEJnPPDLyF91NQNvKyJEjbV46Cr9kyRKpVKmSLbSq3+nq1KkjjRo1ssWNv/rqqzSzLLVgq1fEC+QUdBm1pw8++MDuiqP9KB2F977j9evXzx6Xli5dKufOnbPHsH379qWpeI8ro0OfAe8/j25zpRUuR48eLW3atJG1a9fan/D7ePQ/X9myZe2Hu26RkX76I7IeOUVnTtqW9MP+xIkToZNnOm24bt26EhsbG6FXEd3IKDqzKleunHTv3t1ucfboo4+GqkDr7hH65QpkBLEdiFOnTknXrl1te9ETLjolWE/CJCcnh7Y60+UqemJ58uTJdlcWj3b4w7cdRvYgp+jIyduOU+l3PP3Op1tKt2vXTrZu3Wp3GNAdWuLj49PsGIaruI7R/MDQAgzpp3l404G1iqlOq9LK9J7099XLAwYMMHFxcbaKI8gpyK63PXlLWVatWmWnXmmBycTERNO9e3c7pXjChAk3+RVEPzIKzvHp0KFDtkK37gahRfEOHz58k555cJCRu1lt377dXLx4Mc2xaNasWbYwa/gU+/nz59uK9bq8cty4cXaqvU4r3rBhQwReRfQjp+jOKfz2LVu2mFy5ctnCrb169brsfrgyRujD6J7wWqxE96y87777ZOrUqaHbvDND1atXl7Zt28qhQ4dk2rRp3kmRNCdJdPqITiXRvRW9PbORdcgpunPyZrQ0b97cTtPSx9AiKToFX6c46v7yIKOgyYrPPR1BnDJlisTExMiRI0dk/vz5tugkyCjo7Umn1KvatWvb9qTHIW9mmBZk1et1hNEbpdeCrXPmzJGWLVvaqcRnzpyR9evX2+n4IKegudH2lH6UPjEx0c7AbNasmRw4cEDef/99ez9kwlU6+oGzcuVKU6FCBfPOO+/YYlzPPfecPUM0efJkk5KSkmYUREc3evToYe69917zxx9/2OsuXLhgf3tnorzLIKcg+rft6fz582nO2upevyCjoMqq45P67rvvzLp16yL0SqIXGUVHVqmpqfY+OsqoP3o5JibGzJw586qP5/0NyCmIsrI96ZZ1H3/88U1+BdEh8B16b2pIQkKCqVevXpovPlptsX79+qFqjOE+++wze9vQoUPtf0CtFHzkyJGbGl6QkJMbyMn/yMgdZOV/ZBTdWeluENpZ0anESn8PGjToJj/zYCEnN5CTvwR+yr23367ux6vFgXSKvBY7UcOHD7cV67XaohYQCi8ypNNBdFrIa6+9JvXq1bN/U6JEicxMisANICc3kJP/kZE7yMr/yCh6s1K6b7YWlSxdurQMGDDAFvQ6fPiw/bv0yy1BTkFCe/IZE8CpIf369bOFtZKSkkLX69SQggULXjZlXq+vWrWqWbt2bei+WuhO/z5nzpymadOmZseOHRF4JdGNnNxATv5HRu4gK/8jo+jPas2aNaERyA4dOpgiRYqYYsWKmerVq9uCXSCnIKI9+VtgOvTHjh2z0+K1CqlW961Zs6YpXLhw6EP+xx9/NGXLljWvvvqqvRxeVbFUqVJpKmvv2rXLNGjQwMyYMSMCryS6kZMbyMn/yMgdZOV/ZBS8rM6dO2cf5/bbbzdz586N0KuJXuTkBnJyQyA69Pqh3LVrV9OxY0dz8ODB0PWxsbGmW7du9t/Jyclm+PDh5pZbbgmthffWh+jWWT179ozQsw8OcnIDOfkfGbmDrPyPjIKb1datW2/6awgCcnIDObkjEGvo8+XLJ3ny5JFu3brZbRW8bRIefvhh2bNnj10HVbBgQXnyySelbt268sQTT9g1Uro+RLf2OXnypN0KCOQE2pML+MxzB1n5HxkFNyutkQRyCirakztyaK9eAkALmHh7wnv7IcbHx0v+/Pll8uTJofsdPXpUmjZtag8C9evXt/te33333TJ79mwpWbJkBF9BMJCTG8jJ/8jIHWTlf2TkDrJyAzm5gZzcEJgO/ZU0atRInnrqKenatavt5Cvt6B84cEC2bdsmSUlJUqtWLXs7yAm0J9fxmecOsvI/MnIHWbmBnNxATv4T2A79wYMHJS4uTpYtWxaaUnXhwgXJnTt3pJ8awpCTG8jJ/8jIHWTlf2TkDrJyAzm5gZz8KRBr6MN55y82btwoBQoUCHXmExIS7B6jun4KkUdObiAn/yMjd5CV/5GRO8jKDeTkBnLyt/9KwGjhE/XNN99I+/btZdWqVdKrVy9JSUmRmTNnSokSJSL9FEFOzqA9+R8ZuYOs/I+M3EFWbiAnN5CTz5kASk1NNZUrVzY5cuQwefLkMaNGjYr0U8IVkJMbyMn/yMgdZOV/ZOQOsnIDObmBnPwrsGvoW7RoIVWqVJHx48dL3rx5I/10cBXk5AZy8j8ycgdZ+R8ZuYOs3EBObiAnfwpsh/7SpUuSM2fOSD8NXAM5uYGc/I+M3EFW/kdG7iArN5CTG8jJnwLboQcAAAAAwGWBq3IPAAAAAEA0oEMPAAAAAICD6NADAAAAAOAgOvQAAAAAADiIDj0AAAAAAA6iQw8AAAAAgIPo0AMAAAAA4CA69AAA4Kq6desmOXLksD+5cuWSkiVLSosWLWTq1Knyzz//ZPqd+/DDD+XWW2/lnQYAIAvRoQcAABl66KGH5Ndff5VDhw7J8uXLpVmzZjJgwABp1aqVXLx4kXcPAIAIoUMPAAAylCdPHilVqpSULVtW6tatKy+//LIsXbrUdu515F2NHz9eatasKfnz55dy5cpJ37595c8//7S3rV27Vrp37y5nz54NjfYPGzbM3vbXX3/J888/bx9b/7ZBgwb2/gAA4Nro0AMAgOv2wAMPSK1atWTRokX//4XiP/+RSZMmya5du2T69Ony5ZdfypAhQ+xtcXFxMnHiRClUqJAd6dcf7cSrZ599VjZv3ixz586VHTt2SIcOHeyMgP3795MKAADXkMMYY651JwAAENw19L///rssWbLksts6depkO+G7d+++7LYFCxZI79695fTp0/ayjuQPHDjQPpbnyJEjcuedd9rfZcqUCV3fvHlziY2NlREjRmTb6wIAIBr8N9JPAAAAuEnHBHT6vFq9erWMHDlS9u7dK8nJyXZt/fnz5yUlJUXy5ct3xb//4Ycf5NKlS1K1atU01+s0/GLFit2U1wAAgMvo0AMAgBuyZ88eqVixoi2WpwXy+vTpI2+88YYULVpUNm7cKD169JALFy5ctUOva+xz5swp27Zts7/DFShQgFQAALgGOvQAAOC66Rp5HWEfNGiQ7ZDrFnbjxo2za+nVvHnz0tw/d+7cdjQ+XJ06dex1J0+elMaNG5MCAADXiQ49AADIkE6BP378uO18nzhxQlasWGGn1+uofJcuXWTnzp3y999/y1tvvSWtW7eWTZs2SWJiYprHqFChgh2R/+KLL2wxPR2116n28fHx9jH0ZIB28E+dOmXvExMTI4888gjJAACQAarcAwCADGkHvnTp0rZTrhXo16xZYyva69Z1OlVeO+i6bd3o0aOlRo0aMmvWLNvhD6eV7rVIXseOHaV48eIyZswYe/20adNsh37w4MFy1113Sdu2bWXLli1Svnx5UgEA4Bqocg8AAAAAgIMYoQcAAAAAwEF06AEAAAAAcBAdegAAAAAAHESHHgAAAAAAB9GhBwAAAADAQXToAQAAAABwEB16AAAAAAAcRIceAAAAAAAH0aEHAAAAAMBBdOgBAAAAAHAQHXoAAAAAABxEhx4AAAAAAHHP/wH/OoOgIjq4WQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1200x800 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Paso 1\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, axes = plt.subplots(ncols=1, nrows=4, sharex=True, figsize=(12,8))\n",
    "result.observed.plot(ax=axes[0], legend=False)\n",
    "axes[0].set_ylabel('Observed')\n",
    "trend.plot(ax=axes[1], legend=False)\n",
    "axes[1].set_ylabel('Trend')\n",
    "seasonal.plot(ax=axes[2], legend=False)\n",
    "axes[2].set_ylabel('Seasonal')\n",
    "resid.plot(ax=axes[3], legend=False)\n",
    "axes[3].set_ylabel('Residual')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predicción con ARIMA\n",
    "\n",
    ">En estadística y econometría, en particular en series temporales, un modelo autorregresivo integrado de promedio móvil o ARIMA (acrónimo del inglés autoregressive integrated moving average) es un modelo estadístico que utiliza variaciones y regresiones de datos estadísticos con el fin de encontrar patrones para una predicción hacia el futuro. Se trata de un modelo dinámico de series temporales, es decir, las estimaciones futuras vienen explicadas por los datos del pasado y no por variables independientes. Fue desarrollado a finales de los sesenta del siglo XX. Box y Jenkins (1976) lo sistematizaron.\n",
    "\n",
    "\n",
    "Fuente [Wikipedia](https://es.wikipedia.org/wiki/Modelo_autorregresivo_integrado_de_media_m%C3%B3vil)\n",
    "\n",
    "- https://www.statsmodels.org/dev/generated/statsmodels.tsa.arima.model.ARIMA.html\n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DatetimeIndex(['2017-01-03', '2017-01-04', '2017-01-05', '2017-01-06',\n",
      "               '2017-01-09', '2017-01-10', '2017-01-11', '2017-01-12',\n",
      "               '2017-01-13', '2017-01-17', '2017-01-18', '2017-01-19',\n",
      "               '2017-01-20', '2017-01-23', '2017-01-24', '2017-01-25',\n",
      "               '2017-01-26', '2017-01-27', '2017-01-30', '2017-01-31',\n",
      "               '2017-02-01', '2017-02-02', '2017-02-03', '2017-02-06',\n",
      "               '2017-02-07', '2017-02-08', '2017-02-09', '2017-02-10',\n",
      "               '2017-02-13', '2017-02-14', '2017-02-15', '2017-02-16',\n",
      "               '2017-02-17', '2017-02-21', '2017-02-22', '2017-02-23',\n",
      "               '2017-02-24', '2017-02-27', '2017-02-28', '2017-03-01',\n",
      "               '2017-03-02', '2017-03-03', '2017-03-06', '2017-03-07',\n",
      "               '2017-03-08'],\n",
      "              dtype='datetime64[ns]', name='Date', freq=None)\n"
     ]
    }
   ],
   "source": [
    "# Podemos trabajar con diferentes periodos para la predicción!!!\n",
    "import pandas as pd\n",
    "\n",
    "data = pd.read_csv(\"data/GOOGL.csv\", parse_dates=[\"Date\"], index_col=\"Date\") ##VAMOS A HACERLO BIEN YA!\n",
    "# ¿Probamos meses? \n",
    "print(data.index[:45])\n",
    "# tenemos muchas muestras de cada més: 30,31,28,29, etc.\n",
    "# ¿Que deberíamos de hacer?¿Qué muestra mensual puede ser la mejor?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Date\n",
      "2017-01-31    41.009499\n",
      "2017-02-28    42.246498\n",
      "2017-03-31    42.389999\n",
      "2017-04-30    46.226002\n",
      "2017-05-31    49.354500\n",
      "2017-06-30    46.484001\n",
      "2017-07-31    47.275002\n",
      "2017-08-31    47.762001\n",
      "2017-09-30    48.686001\n",
      "2017-10-31    51.652000\n",
      "2017-11-30    51.808498\n",
      "2017-12-31    52.669998\n",
      "2018-01-31    59.111000\n",
      "2018-02-28    55.195999\n",
      "Freq: ME, Name: Close, dtype: float64\n",
      "48\n"
     ]
    }
   ],
   "source": [
    "# RESAMPLE ! Vamos a coger la última muestra de cada mes:\n",
    "close_monthly = data[\"Close\"].resample('ME').last()\n",
    "print(close_monthly[:14])\n",
    "print(len(close_monthly))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                               SARIMAX Results                                \n",
      "==============================================================================\n",
      "Dep. Variable:                  Close   No. Observations:                   48\n",
      "Model:                 ARIMA(3, 1, 0)   Log Likelihood                -130.621\n",
      "Date:                Tue, 09 Dec 2025   AIC                            269.241\n",
      "Time:                        10:57:51   BIC                            276.642\n",
      "Sample:                    01-31-2017   HQIC                           272.026\n",
      "                         - 12-31-2020                                         \n",
      "Covariance Type:                  opg                                         \n",
      "==============================================================================\n",
      "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "ar.L1         -0.0901      0.128     -0.704      0.481      -0.341       0.161\n",
      "ar.L2         -0.2209      0.137     -1.609      0.108      -0.490       0.048\n",
      "ar.L3          0.1170      0.165      0.709      0.478      -0.206       0.441\n",
      "sigma2        15.1343      4.051      3.736      0.000       7.195      23.073\n",
      "===================================================================================\n",
      "Ljung-Box (L1) (Q):                   0.48   Jarque-Bera (JB):                 2.17\n",
      "Prob(Q):                              0.49   Prob(JB):                         0.34\n",
      "Heteroskedasticity (H):               3.24   Skew:                            -0.52\n",
      "Prob(H) (two-sided):                  0.02   Kurtosis:                         2.88\n",
      "===================================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n"
     ]
    }
   ],
   "source": [
    "# Paso 2.a  Cargamos/Fit/Entrenamos ARIMA \n",
    "from statsmodels.tsa.arima.model import ARIMA \n",
    "\n",
    "order= (3, 1, 0) # order = p,d,q !!OJO!!\n",
    "model = ARIMA(close_monthly, order=order)\n",
    "\n",
    "model_fit = model.fit()\n",
    "print(model_fit.summary())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               0\n",
      "count  48.000000\n",
      "mean    2.001236\n",
      "std     6.843555\n",
      "min    -8.461429\n",
      "25%    -0.455059\n",
      "50%     1.450549\n",
      "75%     3.769838\n",
      "max    41.009499\n"
     ]
    }
   ],
   "source": [
    "residuals = pd.DataFrame(model_fit.resid)\n",
    "residuals.plot()\n",
    "plt.show()\n",
    "print(residuals.describe()) # más o menos valores cercanos al 0 ok"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El cálculo del **orden en ARIMA** ha de hacerse seguiendo un proceso más metodológico:\n",
    "\n",
    "- Elegir d (diferencias)\n",
    "  - Comprobar estacionariedad (gráfica, test ADF, KPSS, etc.).\n",
    "  - Aplicar las diferencias mínimas necesarias.\n",
    "- Elegir p y qq\n",
    "  -  Mirar ACF y PACF de la serie ya diferenciada:\n",
    "     -  PACF que “corta” en el rezago 3 sugiere p≈3.\n",
    "     -  ACF que “corta” en cierto rezago sugiere el q.\n",
    "  - Probar varios modelos cercanos y comparar AIC/BIC: quedarse con el que tenga menor AIC/BIC y residuos ''limpios''.\n",
    "\n",
    "NOTA: Esto queda fuera del alcance de esta asignatura!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2021-01-31    86.247888\n",
      "2021-02-28    87.308516\n",
      "2021-03-31    87.231475\n",
      "2021-04-30    86.938066\n",
      "2021-05-31    87.105654\n",
      "2021-06-30    87.146347\n",
      "2021-07-31    87.071324\n",
      "2021-08-31    87.088709\n",
      "2021-09-30    87.108477\n",
      "2021-10-31    87.094075\n",
      "2021-11-30    87.093041\n",
      "2021-12-31    87.098629\n",
      "Freq: ME, Name: predicted_mean, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# Retomamos el modelo entrenado con ARIMA para predecir futuros valores\n",
    "forecast = model_fit.forecast(steps=12)\n",
    "print(forecast)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(close_monthly, label=\"Histórico\")\n",
    "plt.plot(forecast, label=\"Predicción\", linestyle=\"--\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Y ¿cómo obtenemos métricas de la calidad del modelo? **Debemos a reestructurar** el dataset en dos secciones: train y test."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(38,) (10,)\n"
     ]
    }
   ],
   "source": [
    "n = len(close_monthly)\n",
    "train_size = int(n * 0.8)\n",
    "\n",
    "train = close_monthly[:train_size]\n",
    "test = close_monthly[train_size:]\n",
    "print(train.shape,test.shape)\n",
    "\n",
    "# ¿Esta separación mantiene la coherencia temporal de la serie?\n",
    "# ¿Podríamos usar otra estrategía como un muestreo aleatorio?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Entrenamos\n",
    "model = ARIMA(train, order=order)\n",
    "model_fit = model.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predecimos\n",
    "forecast = model_fit.forecast(steps=len(test))\n",
    "forecast.index = test.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE: 11.929347532992498\n"
     ]
    }
   ],
   "source": [
    "# Evaluamos\n",
    "from sklearn.metrics import mean_squared_error\n",
    "import numpy as np\n",
    "\n",
    "rmse = np.sqrt(mean_squared_error(test, forecast))\n",
    "print(\"RMSE:\", rmse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualizamos\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.plot(train, label=\"Train\")\n",
    "plt.plot(test, label=\"Test\")\n",
    "plt.plot(forecast, label=\"Predicción\", linestyle=\"--\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reflexiones\n",
    "- ¿Cómo mantenemos este modelo mes a més, con nuevos datos?\n",
    "- ¿Qué ocurre si cambiamos de modelo ARIMA(1,0,0) o ARIMA(5,1,2)?\n",
    "- ¿Qué ocurre si reducimos el threshold de entreno?\n",
    "- ¿Qué ocurre si el periodo de la serie no es mensual?\n",
    "- ..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicción con un algoritmo supervisado\n",
    "\n",
    "Vamos a complicar un poco más el estudio. Ahora el objetivo de nuestro analisis será predecir el valor del cierre de MSFT en función del valor del cierre de IBM y GOOGLE y el uso de dos indices Dow Jones y S&P500\n",
    "\n",
    "Disponemos los datos en dos ficheros:\n",
    "- \"data/Stickers.csv\" contiene los valores de las compañias\n",
    "- \"data/stock_data.csv\" , los valores según los dos índices\n",
    "\n",
    "Es importante que comprendáis el formato del fichero antes de intentar abrirlo con pandas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            Adj Close                             Close              \\\n",
      "                GOOGL         IBM       MSFT      GOOGL         IBM   \n",
      "Date                                                                  \n",
      "2017-01-03  40.400501  115.346306  57.138737  40.400501  159.837479   \n",
      "2017-01-04  40.388500  116.774376  56.883072  40.388500  161.816437   \n",
      "2017-01-05  40.651001  116.388054  56.883072  40.651001  161.281067   \n",
      "2017-01-06  41.260502  116.960693  57.376114  41.260502  162.074570   \n",
      "2017-01-09  41.359001  115.663658  57.193520  41.359001  160.277252   \n",
      "\n",
      "                            High                               Low  \\\n",
      "                 MSFT      GOOGL         IBM       MSFT      GOOGL   \n",
      "Date                                                                 \n",
      "2017-01-03  62.580002  40.571999  160.487579  62.840000  39.844501   \n",
      "2017-01-04  62.299999  40.671501  162.399612  62.750000  40.205502   \n",
      "2017-01-05  62.299999  40.687000  161.940720  62.660000  40.296001   \n",
      "2017-01-06  62.840000  41.448002  162.447418  63.150002  40.575001   \n",
      "2017-01-09  62.639999  41.521500  162.332703  63.080002  41.081001   \n",
      "\n",
      "                                        Open                           Volume  \\\n",
      "                   IBM       MSFT      GOOGL         IBM       MSFT     GOOGL   \n",
      "Date                                                                            \n",
      "2017-01-03  158.709366  62.130001  40.030998  159.655838  62.790001  39180000   \n",
      "2017-01-04  160.000000  62.119999  40.494499  160.391968  62.480000  30306000   \n",
      "2017-01-05  159.904404  62.029999  40.375000  161.806885  62.189999  26810000   \n",
      "2017-01-06  160.152969  62.040001  40.749500  161.271515  62.299999  40342000   \n",
      "2017-01-09  160.248566  62.540001  41.318501  162.017212  62.759998  28178000   \n",
      "\n",
      "                               \n",
      "                IBM      MSFT  \n",
      "Date                           \n",
      "2017-01-03  3069278  20694100  \n",
      "2017-01-04  3536944  21340000  \n",
      "2017-01-05  2805686  24876000  \n",
      "2017-01-06  3080993  19922900  \n",
      "2017-01-09  3336635  20382700  \n"
     ]
    }
   ],
   "source": [
    "data = pd.read_csv(\n",
    "    \"data/Stickers.csv\",\n",
    "    header=[0, 1],      # dos filas de encabezado!!\n",
    "    index_col=0,        # la primera columna es Date\n",
    "    parse_dates=True    # convierte automáticamente la columna Date en datetime\n",
    ")\n",
    "\n",
    "print(data.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               Adj Close                      Close               \\\n",
      "                    ^DJI        ^GSPC          ^DJI        ^GSPC   \n",
      "Date                                                               \n",
      "2017-01-03  19881.759766  2257.830078  19881.759766  2257.830078   \n",
      "2017-01-04  19942.160156  2270.750000  19942.160156  2270.750000   \n",
      "2017-01-05  19899.289062  2269.000000  19899.289062  2269.000000   \n",
      "2017-01-06  19963.800781  2276.979980  19963.800781  2276.979980   \n",
      "2017-01-09  19887.380859  2268.899902  19887.380859  2268.899902   \n",
      "\n",
      "                    High                        Low               \\\n",
      "                    ^DJI        ^GSPC          ^DJI        ^GSPC   \n",
      "Date                                                               \n",
      "2017-01-03  19938.529297  2263.879883  19775.929688  2245.129883   \n",
      "2017-01-04  19956.140625  2272.820068  19878.830078  2261.600098   \n",
      "2017-01-05  19948.599609  2271.500000  19811.119141  2260.449951   \n",
      "2017-01-06  19999.630859  2282.100098  19834.080078  2264.060059   \n",
      "2017-01-09  19943.779297  2275.489990  19887.380859  2268.899902   \n",
      "\n",
      "                    Open                  Volume              \n",
      "                    ^DJI        ^GSPC       ^DJI       ^GSPC  \n",
      "Date                                                          \n",
      "2017-01-03  19872.859375  2251.570068  339180000  3773010000  \n",
      "2017-01-04  19890.939453  2261.600098  280010000  3768890000  \n",
      "2017-01-05  19924.560547  2268.179932  269920000  3785080000  \n",
      "2017-01-06  19906.960938  2271.139893  277700000  3342080000  \n",
      "2017-01-09  19931.410156  2273.590088  287510000  3219730000  \n"
     ]
    }
   ],
   "source": [
    "stock_data = pd.read_csv(\n",
    "    \"data/stock_data.csv\",\n",
    "    header=[0, 1],      # dos filas de encabezado!!\n",
    "    index_col=0,        # la primera columna es Date\n",
    "    parse_dates=True    # convierte automáticamente la columna Date en datetime\n",
    ")\n",
    "\n",
    "print(stock_data.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            Adj Close                             Close              \\\n",
      "                GOOGL         IBM       MSFT      GOOGL         IBM   \n",
      "Date                                                                  \n",
      "2017-01-03  40.400501  115.346306  57.138737  40.400501  159.837479   \n",
      "2017-01-04  40.388500  116.774376  56.883072  40.388500  161.816437   \n",
      "2017-01-05  40.651001  116.388054  56.883072  40.651001  161.281067   \n",
      "2017-01-06  41.260502  116.960693  57.376114  41.260502  162.074570   \n",
      "2017-01-09  41.359001  115.663658  57.193520  41.359001  160.277252   \n",
      "\n",
      "                            High                               Low  ...  \\\n",
      "                 MSFT      GOOGL         IBM       MSFT      GOOGL  ...   \n",
      "Date                                                                ...   \n",
      "2017-01-03  62.580002  40.571999  160.487579  62.840000  39.844501  ...   \n",
      "2017-01-04  62.299999  40.671501  162.399612  62.750000  40.205502  ...   \n",
      "2017-01-05  62.299999  40.687000  161.940720  62.660000  40.296001  ...   \n",
      "2017-01-06  62.840000  41.448002  162.447418  63.150002  40.575001  ...   \n",
      "2017-01-09  62.639999  41.521500  162.332703  63.080002  41.081001  ...   \n",
      "\n",
      "                   Close                       High               \\\n",
      "                    ^DJI        ^GSPC          ^DJI        ^GSPC   \n",
      "Date                                                               \n",
      "2017-01-03  19881.759766  2257.830078  19938.529297  2263.879883   \n",
      "2017-01-04  19942.160156  2270.750000  19956.140625  2272.820068   \n",
      "2017-01-05  19899.289062  2269.000000  19948.599609  2271.500000   \n",
      "2017-01-06  19963.800781  2276.979980  19999.630859  2282.100098   \n",
      "2017-01-09  19887.380859  2268.899902  19943.779297  2275.489990   \n",
      "\n",
      "                     Low                       Open                  Volume  \\\n",
      "                    ^DJI        ^GSPC          ^DJI        ^GSPC       ^DJI   \n",
      "Date                                                                          \n",
      "2017-01-03  19775.929688  2245.129883  19872.859375  2251.570068  339180000   \n",
      "2017-01-04  19878.830078  2261.600098  19890.939453  2261.600098  280010000   \n",
      "2017-01-05  19811.119141  2260.449951  19924.560547  2268.179932  269920000   \n",
      "2017-01-06  19834.080078  2264.060059  19906.960938  2271.139893  277700000   \n",
      "2017-01-09  19887.380859  2268.899902  19931.410156  2273.590088  287510000   \n",
      "\n",
      "                        \n",
      "                 ^GSPC  \n",
      "Date                    \n",
      "2017-01-03  3773010000  \n",
      "2017-01-04  3768890000  \n",
      "2017-01-05  3785080000  \n",
      "2017-01-06  3342080000  \n",
      "2017-01-09  3219730000  \n",
      "\n",
      "[5 rows x 30 columns]\n"
     ]
    }
   ],
   "source": [
    "### IMPORTANTE! Tenemos dos ficheros, necesitamos unirlos para tener un único dataset. Usamos el indice como elemento en común entre ambos\n",
    "dataset = data.merge(stock_data, how=\"inner\",left_index=True,right_index=True)\n",
    "print(dataset[:5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gestión del tiempo.\n",
    "\n",
    "Para gestionar el tiempo, necesitamos transformar la serie en algo más \"masticable\" para el modelo. \n",
    "\n",
    "En este caso, podemos reconvertir la serie en una frecuencia de 5 días laborales."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">Adj Close</th>\n",
       "      <th colspan=\"3\" halign=\"left\">Close</th>\n",
       "      <th colspan=\"3\" halign=\"left\">High</th>\n",
       "      <th>Low</th>\n",
       "      <th>...</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Close</th>\n",
       "      <th colspan=\"2\" halign=\"left\">High</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Low</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Open</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>...</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GSPC</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GSPC</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GSPC</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GSPC</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GSPC</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></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>2017-01-09</th>\n",
       "      <td>40.811901</td>\n",
       "      <td>116.226617</td>\n",
       "      <td>57.094903</td>\n",
       "      <td>40.811901</td>\n",
       "      <td>161.057361</td>\n",
       "      <td>62.532000</td>\n",
       "      <td>40.980000</td>\n",
       "      <td>161.921606</td>\n",
       "      <td>62.896001</td>\n",
       "      <td>40.400401</td>\n",
       "      <td>...</td>\n",
       "      <td>19914.878125</td>\n",
       "      <td>2268.691992</td>\n",
       "      <td>19957.335938</td>\n",
       "      <td>2273.158008</td>\n",
       "      <td>19837.467969</td>\n",
       "      <td>2260.027979</td>\n",
       "      <td>19905.346094</td>\n",
       "      <td>2265.216016</td>\n",
       "      <td>290864000.0</td>\n",
       "      <td>3.577758e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-16</th>\n",
       "      <td>41.454250</td>\n",
       "      <td>115.311779</td>\n",
       "      <td>57.321336</td>\n",
       "      <td>41.454250</td>\n",
       "      <td>159.789677</td>\n",
       "      <td>62.780000</td>\n",
       "      <td>41.554250</td>\n",
       "      <td>160.693115</td>\n",
       "      <td>63.142500</td>\n",
       "      <td>41.189251</td>\n",
       "      <td>...</td>\n",
       "      <td>19896.634766</td>\n",
       "      <td>2272.324951</td>\n",
       "      <td>19952.964355</td>\n",
       "      <td>2276.262512</td>\n",
       "      <td>19822.260254</td>\n",
       "      <td>2262.965027</td>\n",
       "      <td>19900.620117</td>\n",
       "      <td>2270.549988</td>\n",
       "      <td>301407500.0</td>\n",
       "      <td>3.455635e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-23</th>\n",
       "      <td>41.534500</td>\n",
       "      <td>116.330086</td>\n",
       "      <td>57.162468</td>\n",
       "      <td>41.534500</td>\n",
       "      <td>161.200766</td>\n",
       "      <td>62.606000</td>\n",
       "      <td>41.677701</td>\n",
       "      <td>161.780115</td>\n",
       "      <td>62.864000</td>\n",
       "      <td>41.245400</td>\n",
       "      <td>...</td>\n",
       "      <td>19798.198047</td>\n",
       "      <td>2267.995947</td>\n",
       "      <td>19842.650000</td>\n",
       "      <td>2273.432031</td>\n",
       "      <td>19736.757812</td>\n",
       "      <td>2261.320020</td>\n",
       "      <td>19814.990234</td>\n",
       "      <td>2269.583936</td>\n",
       "      <td>337072000.0</td>\n",
       "      <td>3.353020e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-30</th>\n",
       "      <td>42.338200</td>\n",
       "      <td>122.245384</td>\n",
       "      <td>58.869874</td>\n",
       "      <td>42.338200</td>\n",
       "      <td>169.397702</td>\n",
       "      <td>64.475999</td>\n",
       "      <td>42.755399</td>\n",
       "      <td>170.248566</td>\n",
       "      <td>64.816001</td>\n",
       "      <td>42.054700</td>\n",
       "      <td>...</td>\n",
       "      <td>20029.408203</td>\n",
       "      <td>2290.141992</td>\n",
       "      <td>20060.282031</td>\n",
       "      <td>2294.039990</td>\n",
       "      <td>19958.350391</td>\n",
       "      <td>2281.860010</td>\n",
       "      <td>19999.477734</td>\n",
       "      <td>2288.083936</td>\n",
       "      <td>352768000.0</td>\n",
       "      <td>3.602052e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-06</th>\n",
       "      <td>40.954400</td>\n",
       "      <td>120.744125</td>\n",
       "      <td>58.201519</td>\n",
       "      <td>40.954400</td>\n",
       "      <td>167.317398</td>\n",
       "      <td>63.744000</td>\n",
       "      <td>41.201500</td>\n",
       "      <td>167.986615</td>\n",
       "      <td>64.106001</td>\n",
       "      <td>40.713400</td>\n",
       "      <td>...</td>\n",
       "      <td>19952.764062</td>\n",
       "      <td>2285.850049</td>\n",
       "      <td>19997.016016</td>\n",
       "      <td>2289.337988</td>\n",
       "      <td>19885.774219</td>\n",
       "      <td>2277.549951</td>\n",
       "      <td>19937.026172</td>\n",
       "      <td>2283.824023</td>\n",
       "      <td>354716000.0</td>\n",
       "      <td>3.707408e+09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            Adj Close                             Close              \\\n",
       "                GOOGL         IBM       MSFT      GOOGL         IBM   \n",
       "Date                                                                  \n",
       "2017-01-09  40.811901  116.226617  57.094903  40.811901  161.057361   \n",
       "2017-01-16  41.454250  115.311779  57.321336  41.454250  159.789677   \n",
       "2017-01-23  41.534500  116.330086  57.162468  41.534500  161.200766   \n",
       "2017-01-30  42.338200  122.245384  58.869874  42.338200  169.397702   \n",
       "2017-02-06  40.954400  120.744125  58.201519  40.954400  167.317398   \n",
       "\n",
       "                            High                               Low  ...  \\\n",
       "                 MSFT      GOOGL         IBM       MSFT      GOOGL  ...   \n",
       "Date                                                                ...   \n",
       "2017-01-09  62.532000  40.980000  161.921606  62.896001  40.400401  ...   \n",
       "2017-01-16  62.780000  41.554250  160.693115  63.142500  41.189251  ...   \n",
       "2017-01-23  62.606000  41.677701  161.780115  62.864000  41.245400  ...   \n",
       "2017-01-30  64.475999  42.755399  170.248566  64.816001  42.054700  ...   \n",
       "2017-02-06  63.744000  41.201500  167.986615  64.106001  40.713400  ...   \n",
       "\n",
       "                   Close                       High               \\\n",
       "                    ^DJI        ^GSPC          ^DJI        ^GSPC   \n",
       "Date                                                               \n",
       "2017-01-09  19914.878125  2268.691992  19957.335938  2273.158008   \n",
       "2017-01-16  19896.634766  2272.324951  19952.964355  2276.262512   \n",
       "2017-01-23  19798.198047  2267.995947  19842.650000  2273.432031   \n",
       "2017-01-30  20029.408203  2290.141992  20060.282031  2294.039990   \n",
       "2017-02-06  19952.764062  2285.850049  19997.016016  2289.337988   \n",
       "\n",
       "                     Low                       Open                    Volume  \\\n",
       "                    ^DJI        ^GSPC          ^DJI        ^GSPC         ^DJI   \n",
       "Date                                                                            \n",
       "2017-01-09  19837.467969  2260.027979  19905.346094  2265.216016  290864000.0   \n",
       "2017-01-16  19822.260254  2262.965027  19900.620117  2270.549988  301407500.0   \n",
       "2017-01-23  19736.757812  2261.320020  19814.990234  2269.583936  337072000.0   \n",
       "2017-01-30  19958.350391  2281.860010  19999.477734  2288.083936  352768000.0   \n",
       "2017-02-06  19885.774219  2277.549951  19937.026172  2283.824023  354716000.0   \n",
       "\n",
       "                          \n",
       "                   ^GSPC  \n",
       "Date                      \n",
       "2017-01-09  3.577758e+09  \n",
       "2017-01-16  3.455635e+09  \n",
       "2017-01-23  3.353020e+09  \n",
       "2017-01-30  3.602052e+09  \n",
       "2017-02-06  3.707408e+09  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Reconvertimos la frecuencia agrupando los valores. La semana empieza en el lunes !\n",
    "# https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.resample.html\n",
    "data5days = dataset.resample(\"1W-MON\").mean(numeric_only=True)  \n",
    "data5days.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 ('Adj Close', 'GOOGL')\n",
      "1 ('Adj Close', 'IBM')\n",
      "2 ('Adj Close', 'MSFT')\n",
      "3 ('Close', 'GOOGL')\n",
      "4 ('Close', 'IBM')\n",
      "5 ('Close', 'MSFT')\n",
      "6 ('High', 'GOOGL')\n",
      "7 ('High', 'IBM')\n",
      "8 ('High', 'MSFT')\n",
      "9 ('Low', 'GOOGL')\n",
      "10 ('Low', 'IBM')\n",
      "11 ('Low', 'MSFT')\n",
      "12 ('Open', 'GOOGL')\n",
      "13 ('Open', 'IBM')\n",
      "14 ('Open', 'MSFT')\n",
      "15 ('Volume', 'GOOGL')\n",
      "16 ('Volume', 'IBM')\n",
      "17 ('Volume', 'MSFT')\n",
      "18 ('Adj Close', '^DJI')\n",
      "19 ('Adj Close', '^GSPC')\n",
      "20 ('Close', '^DJI')\n",
      "21 ('Close', '^GSPC')\n",
      "22 ('High', '^DJI')\n",
      "23 ('High', '^GSPC')\n",
      "24 ('Low', '^DJI')\n",
      "25 ('Low', '^GSPC')\n",
      "26 ('Open', '^DJI')\n",
      "27 ('Open', '^GSPC')\n",
      "28 ('Volume', '^DJI')\n",
      "29 ('Volume', '^GSPC')\n"
     ]
    }
   ],
   "source": [
    "# ¿Usaremos todas las features? \n",
    "for ix,col in enumerate(data5days.columns):\n",
    "    print(ix,col)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0,\n",
       " 1,\n",
       " 2,\n",
       " 6,\n",
       " 7,\n",
       " 8,\n",
       " 9,\n",
       " 10,\n",
       " 11,\n",
       " 12,\n",
       " 13,\n",
       " 14,\n",
       " 18,\n",
       " 19,\n",
       " 22,\n",
       " 23,\n",
       " 24,\n",
       " 25,\n",
       " 26,\n",
       " 27,\n",
       " 28,\n",
       " 29]"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Vamos a utilizar: el valor de cierre y el volumen de las acciones de cada compañia; más los dos indices de cierre.\n",
    "# Limpiamos series innecesarias\n",
    "columns_a_conservar= [3,4,5,15,16,17,20,21]\n",
    "columns_a_eliminar = set(range(len(data5days.columns)))-set(columns_a_conservar)\n",
    "columns_a_eliminar = list(columns_a_eliminar)\n",
    "columns_a_eliminar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">Close</th>\n",
       "      <th colspan=\"3\" halign=\"left\">Volume</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Close</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GSPC</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></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>2017-01-09</th>\n",
       "      <td>40.811901</td>\n",
       "      <td>161.057361</td>\n",
       "      <td>62.532000</td>\n",
       "      <td>32963200.0</td>\n",
       "      <td>3165907.2</td>\n",
       "      <td>21443140.0</td>\n",
       "      <td>19914.878125</td>\n",
       "      <td>2268.691992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-16</th>\n",
       "      <td>41.454250</td>\n",
       "      <td>159.789677</td>\n",
       "      <td>62.780000</td>\n",
       "      <td>25812500.0</td>\n",
       "      <td>3535767.5</td>\n",
       "      <td>20125200.0</td>\n",
       "      <td>19896.634766</td>\n",
       "      <td>2272.324951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-23</th>\n",
       "      <td>41.534500</td>\n",
       "      <td>161.200766</td>\n",
       "      <td>62.606000</td>\n",
       "      <td>29210800.0</td>\n",
       "      <td>6789606.8</td>\n",
       "      <td>22419380.0</td>\n",
       "      <td>19798.198047</td>\n",
       "      <td>2267.995947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-30</th>\n",
       "      <td>42.338200</td>\n",
       "      <td>169.397702</td>\n",
       "      <td>64.475999</td>\n",
       "      <td>56426000.0</td>\n",
       "      <td>4919526.2</td>\n",
       "      <td>33673920.0</td>\n",
       "      <td>20029.408203</td>\n",
       "      <td>2290.141992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-06</th>\n",
       "      <td>40.954400</td>\n",
       "      <td>167.317398</td>\n",
       "      <td>63.744000</td>\n",
       "      <td>35357600.0</td>\n",
       "      <td>3245800.8</td>\n",
       "      <td>32173440.0</td>\n",
       "      <td>19952.764062</td>\n",
       "      <td>2285.850049</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Close                             Volume             \\\n",
       "                GOOGL         IBM       MSFT       GOOGL        IBM   \n",
       "Date                                                                  \n",
       "2017-01-09  40.811901  161.057361  62.532000  32963200.0  3165907.2   \n",
       "2017-01-16  41.454250  159.789677  62.780000  25812500.0  3535767.5   \n",
       "2017-01-23  41.534500  161.200766  62.606000  29210800.0  6789606.8   \n",
       "2017-01-30  42.338200  169.397702  64.475999  56426000.0  4919526.2   \n",
       "2017-02-06  40.954400  167.317398  63.744000  35357600.0  3245800.8   \n",
       "\n",
       "                               Close               \n",
       "                  MSFT          ^DJI        ^GSPC  \n",
       "Date                                               \n",
       "2017-01-09  21443140.0  19914.878125  2268.691992  \n",
       "2017-01-16  20125200.0  19896.634766  2272.324951  \n",
       "2017-01-23  22419380.0  19798.198047  2267.995947  \n",
       "2017-01-30  33673920.0  20029.408203  2290.141992  \n",
       "2017-02-06  32173440.0  19952.764062  2285.850049  "
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data5days_clean = data5days.drop(data5days.columns[columns_a_eliminar],axis=1)\n",
    "data5days_clean.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['GOOGL', 'IBM', 'MSFT', 'GOOGL', 'IBM', 'MSFT', '^DJI', '^GSPC'], dtype='object')"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# En cualquier momento, podemos simplicar el uso de multiindex eliminando un nivel superior \n",
    "data5days_clean.columns = data5days_clean.columns.droplevel(0)\n",
    "data5days_clean.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "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>CloseGOOGL</th>\n",
       "      <th>CloseIBM</th>\n",
       "      <th>CloseMSFT</th>\n",
       "      <th>VolGOOGL</th>\n",
       "      <th>VolIBM</th>\n",
       "      <th>VolMSFT</th>\n",
       "      <th>DJI</th>\n",
       "      <th>SP500</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></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>2017-01-09</th>\n",
       "      <td>40.811901</td>\n",
       "      <td>161.057361</td>\n",
       "      <td>62.532000</td>\n",
       "      <td>32963200.0</td>\n",
       "      <td>3165907.2</td>\n",
       "      <td>21443140.0</td>\n",
       "      <td>19914.878125</td>\n",
       "      <td>2268.691992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-16</th>\n",
       "      <td>41.454250</td>\n",
       "      <td>159.789677</td>\n",
       "      <td>62.780000</td>\n",
       "      <td>25812500.0</td>\n",
       "      <td>3535767.5</td>\n",
       "      <td>20125200.0</td>\n",
       "      <td>19896.634766</td>\n",
       "      <td>2272.324951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-23</th>\n",
       "      <td>41.534500</td>\n",
       "      <td>161.200766</td>\n",
       "      <td>62.606000</td>\n",
       "      <td>29210800.0</td>\n",
       "      <td>6789606.8</td>\n",
       "      <td>22419380.0</td>\n",
       "      <td>19798.198047</td>\n",
       "      <td>2267.995947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-30</th>\n",
       "      <td>42.338200</td>\n",
       "      <td>169.397702</td>\n",
       "      <td>64.475999</td>\n",
       "      <td>56426000.0</td>\n",
       "      <td>4919526.2</td>\n",
       "      <td>33673920.0</td>\n",
       "      <td>20029.408203</td>\n",
       "      <td>2290.141992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-06</th>\n",
       "      <td>40.954400</td>\n",
       "      <td>167.317398</td>\n",
       "      <td>63.744000</td>\n",
       "      <td>35357600.0</td>\n",
       "      <td>3245800.8</td>\n",
       "      <td>32173440.0</td>\n",
       "      <td>19952.764062</td>\n",
       "      <td>2285.850049</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            CloseGOOGL    CloseIBM  CloseMSFT    VolGOOGL     VolIBM  \\\n",
       "Date                                                                   \n",
       "2017-01-09   40.811901  161.057361  62.532000  32963200.0  3165907.2   \n",
       "2017-01-16   41.454250  159.789677  62.780000  25812500.0  3535767.5   \n",
       "2017-01-23   41.534500  161.200766  62.606000  29210800.0  6789606.8   \n",
       "2017-01-30   42.338200  169.397702  64.475999  56426000.0  4919526.2   \n",
       "2017-02-06   40.954400  167.317398  63.744000  35357600.0  3245800.8   \n",
       "\n",
       "               VolMSFT           DJI        SP500  \n",
       "Date                                               \n",
       "2017-01-09  21443140.0  19914.878125  2268.691992  \n",
       "2017-01-16  20125200.0  19896.634766  2272.324951  \n",
       "2017-01-23  22419380.0  19798.198047  2267.995947  \n",
       "2017-01-30  33673920.0  20029.408203  2290.141992  \n",
       "2017-02-06  32173440.0  19952.764062  2285.850049  "
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Y renombramos columnas: MUY IMPORTANTE EL ORDEN!!\n",
    "columns_names = [\"CloseGOOGL\",\"CloseIBM\",\"CloseMSFT\",\"VolGOOGL\",\"VolIBM\",\"VolMSFT\",\"DJI\",\"SP500\"]\n",
    "data5days_clean.columns = columns_names\n",
    "\n",
    "data5days_clean.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Con los datos preparados, podemos analizar los valores de estas series y sus posibles relaciones"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Su desomposición\n",
    "result = seasonal_decompose(data5days_clean.CloseGOOGL, model=\"additive\", period=30)\n",
    "fig = result.plot()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "o incluso podemos mirar la matriz de correlación para detectar algún patrón común entre los datos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x151f2dfd0>"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 480x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#con matplotlib \n",
    "import matplotlib.pyplot as plt\n",
    "correlation = data5days_clean.corr()\n",
    "plt.matshow(correlation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#con seaborn \n",
    "import seaborn as sns\n",
    "sns.heatmap(correlation, vmax=1, square=True,annot=True,cmap='cubehelix')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Con los datos ya preparados aplicamos un modelo supervisados de predicción. PERO esta vez, aplicamos *multiples modelos**.\n",
    "Nuestra variable objetivo, por ejemplo, será el valor de cierre de Google.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Aplicación de modelos de supervisado\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# Vamos aplicar diferentes modelos\n",
    "models = [LinearRegression,SVR,RandomForestRegressor]\n",
    "\n",
    "# Preparamos las variables: features y target\n",
    "X = data5days_clean.drop(labels=[\"CloseGOOGL\"],axis=1)\n",
    "Y = data5days_clean.CloseGOOGL\n",
    "\n",
    "# Particionamos los datos de entreno y test al 80%\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.20, random_state=33)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: LinearRegression\n",
      "Model: SVR\n",
      "Model: RandomForestRegressor\n"
     ]
    }
   ],
   "source": [
    "# Variables de salida\n",
    "model_names = []\n",
    "train_results = []\n",
    "test_results = []\n",
    "\n",
    "# Entrenamos y testeamos\n",
    "for model in models:\n",
    "    print(\"Model: %s\"%model.__name__)\n",
    "    model_names.append(model.__name__)\n",
    "    res = model().fit(X=X_train,y=y_train)\n",
    "    train_result = mean_squared_error(res.predict(X_train), y_train) \n",
    "    train_results.append(train_result)\n",
    "\n",
    "    test_result = mean_squared_error(res.predict(X_test), y_test)\n",
    "    test_results.append(test_result)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "ind = np.arange(len(models)) \n",
    "width = 0.35 # the width of the bars\n",
    "\n",
    "fig,ax = plt.subplots()\n",
    "plt.bar(ind - width/2, train_results, width=width, label='Train Error')\n",
    "plt.bar(ind + width/2, test_results, width=width, label='Test Error')\n",
    "plt.legend()\n",
    "ax.set_xticks(ind)\n",
    "ax.set_xticklabels(model_names)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Más allá de la regresión lineal simple\n",
    "\n",
    "En la sesión de hoy seguiremos trabajando con las técnicas de regresión, profundizaremos\n",
    "en las técnicas de regresión lineal y veremos otras técnicas de regresión no lineal.\n",
    "Por otra parte revisaremos las métricas más usadas para este problema. Finalmente\n",
    "trabajaremos en la automatización de los procesos de aprendizaje automático.\n",
    "\n",
    "1. Uso de la regularización\n",
    "2. Regresión no lineal: un ejemplo\n",
    "3. Métricas de regresión\n",
    "4. Buenas prácticas I: Conjuntos de entrenamiento y test.\n",
    "\n",
    "\n",
    "### 1. Uso de la regularización\n",
    "\n",
    "Una forma de encontrar una buena relación entre el sesgo y la varianza es ajustar la complejidad del modelo a través de\n",
    "**la regularización**. La regularización es un método muy útil para manejar\n",
    "colinealidad (alta correlación entre características), filtrar el ruido de los datos y\n",
    "eventualmente evitará el _overfitting_. El concepto detrás de la regularización es introducir\n",
    "información adicional (sesgo) para penalizar los valores extremas de los parámetros.\n",
    "\n",
    "Si tenemos que la expresión para una regresión lineal es: $$y = w_0x_0+w_1x_1+...+w_nx_n=\\sum^m_{i=0} w_i X_{i}$$\n",
    "\n",
    "Y la función que queremos minimizar es: $$J(w) = \\sum^n_{i=1}(y^{(i)}-\\hat{y}^{(i)})^2 $$\n",
    "\n",
    "\n",
    "### Ridge Regression\n",
    "La _Ridge Regression_ es un modelo que usa una penalización aplicando la norma L2 donde simplemente agregamos\n",
    "la suma al cuadrado de los pesos de nuestra función de coste:\n",
    "\n",
    "$$ J(w)_{Ridge} = \\sum^n_{i=1}(y^{(i)}-\\hat{y}^{(i)})^2 + \\lambda\\| w \\|_2$$\n",
    "\n",
    "El uso de esta regresión sería el siguiente:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Ridge\n",
    "\n",
    "ridge = Ridge(alpha=1.0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Al aumentar el valor del hiperparámetro $\\lambda$, aumentamos la fuerza de la regularización\n",
    "y encogemos los pesos de nuestro modelo. Hay que tener en cuenta que no regularizamos\n",
    "el término del _intercepto_ ($w_0$).\n",
    "\n",
    "### Lasso Regression (L1)\n",
    "Un enfoque alternativo que puede conducir a modelos dispersos es la regresión LASSO.\n",
    "Según sea el valor del término de regularización, ciertos pesos pueden coger el valor cero,\n",
    "lo que hace que este tipo de regresión LASSO también sea útil como técnica de selección de características\n",
    "supervisada.\n",
    "\n",
    "La función a minimizar en este caso es:\n",
    "$$ J(w)_{Lasso} = \\sum^n_{i=1}(y^{(i)}-\\hat{y}^{(i)})^2 + \\lambda\\| w \\|_1$$\n",
    "\n",
    "El uso de esta regresión sería el siguiente:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Lasso\n",
    "lasso = Lasso(alpha=1.0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Elastic Net Regression\n",
    "\n",
    "Sin embargo, una limitación de LASSO es que selecciona como máximo $n$ variables si $ m > n $.\n",
    "_Elastic Net_ ofrece un buen compromiso entre la _Ridge_ _regression_  y LASSO, que tiene una\n",
    "Penalización L1 para generar una selección de características y penalización L2 para superar algunas de las limitaciones de la regresión LASSO, como es el número de variables seleccionadas.\n",
    "\n",
    "La función a minimizar en este caso es:\n",
    "\n",
    "$$ J(w)_{Elasticnet} = \\sum^n_{i=1}(y^{(i)}-\\hat{y}^{(i)})^2 + \\lambda_1 \\sum^n_{i=1} w^2_j + \\lambda_2 \\sum^n_{i=1} w_j $$\n",
    "\n",
    "El uso de esta regresión sería el siguiente:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import ElasticNet\n",
    "elastic = ElasticNet(alpha=1.0, l1_ratio=0.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regresión polinómica\n",
    "La regresión polinomial es un caso especial de regresión lineal en el que se construyen nuevas características en\n",
    "función del grado del polinomio que se quiera construir.\n",
    "En las secciones anteriores, hemos asumido una relación lineal entre las variables del modelo y\n",
    "la variable objetivo. Una forma de no tener en cuenta el supuesto de linealidad es\n",
    "usar un modelo de regresión polinomial, es decir agregando términos polinomiales.\n",
    "\n",
    "$$ y = w_0 + w_1x + w_2x^2 + ... + w_dx^d $$\n",
    "\n",
    "donde $d$ es el grado del polinomio. Aunque podemos usar un polinomio\n",
    "para modelar una relación no lineal, todavía se considera una\n",
    "modelo de regresión lineal debido a los coeficientes $w$.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAy0AAAE/CAYAAABLkSy/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAByaUlEQVR4nO3dd3gU1dvG8e9k0ysEQg0Qeu+9CKGIgBSpEkG6gIj6s2HvvWMFERCVLiAWxAaE3nvvoUNCICEJ6TvvHwsILzV1d5P7c125JNndmWcNezPPzJlzDNM0ERERERERcVQu9i5ARERERETkVtS0iIiIiIiIQ1PTIiIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yIiIiIiIg5NTYsTMwwjwjCMRMMw4gzDiDEMY5VhGCMNw3C636thGO6GYcy59J5MwzBC7V2TSH6Sl/LkaoZhvHIpU9rZuxaRvCov5cftjkcMw3jNMIxUwzDir/oqZ59q8xen+8sk1+limqYfUAZ4D3gWmJQTOzIMw5IT273KCqA/cDqH9yMiN5aX8gTDMMoDvYFTOb0vEclT+XG745FZpmn6XvV1KIfrEdS05BmmacaapvkrcD8w0DCMGgCGYXgYhvGRYRhHDcM4YxjGeMMwvC6/zjCMMYZhnDIM46RhGMMunVWocOmxKYZhjDMM4w/DMBKA1oZhlDAMY65hGFGGYRw2DOOxq7blYhjGc4ZhHDQMI9owjNmGYQTeYf0ppmmONU1zBZCenf9vRCRjnD1PrvIVtgOnlKz+PxGRO+Ps+aHjEcelpiWPMU1zHXAcuOvSj94DKgF1gApASeAVAMMwOgBPAu0uPRZ6g00+ALwN+AGrgN+ArZe20xb4n2EY91x67qPAfUAroARwHttBA5f2t80wjAey432KSM5z5jwxDKM3kGya5h8ZetMiki2cOT/uQBfDMM4ZhrHTMIyHs7AdyQjTNPXlpF9ABNDuBj9fA7wIGEACUP6qx5oChy/9eTLw7lWPVQBMoMKl76cAP1z1eGPg6P/b1/PAd5f+vBtoe9VjxYFUwDWD7+s4EGrv/7/60ld++spLeYLtoGY/EHKr96Yvfekre77yUn78v21edzwCVMPWCFmAZtiGn4bZ+3eQH75ckbyoJHAOCAK8gY2GYVx+zMD2QQPbh27DVa87doNtXf2zMkAJwzBirvqZBVh+1eM/G4ZhverxdKAocCLD70JEHIEz5slrwI+maUbc5nkikrOcMT9uyTTNXVd9u8owjM+AXsCMrGxXbk9NSx5jGEZDbCGxAjgLJALVTdO80Yf0FBB81felbvAc86o/H8N2VqTiTXZ/DBhimubKDBcuIg7HifOkLRBsGMaoS98HAbMNw3jfNM33M7E9EckgJ86PjDKxNWCSw3RPSx5hGIa/YRidgZnAVNM0t5umaQW+BT41DKPIpeeVvGrM52xgsGEYVQ3D8AZevs1u1gFxhmE8axiGl2EYFsMwalwKJoDxwNuGYZS5tK8gwzC6ZeA9eBiG4XnpW3fDMDyNq07JiEjuyAN50haogW3sfB3gJDCCq8a0i0jOyAP5ccvjEcMwuhmGUdCwaQQ8Bvxyp9uWzFPT4vx+MwwjDttZhReBT4DBVz3+LHAAWGMYxgXgX6AygGmaC4HPgSWXn3PpNck32pFpmulAZ2wHAYexnTmZCARcespnwK/A35dqWoNt3CkAl25Y63eL97IX25mYksBfl/5c5nb/A0Qk2+SJPDFNM9o0zdOXv7ANCzlvmmb8nf+vEJEMyhP5ccmtjkf6XqoxDvgBeN80ze9vsS3JJsalm4pEMAyjKrAD8DBNM83e9YiI81KeiEhmKT/kRnSlJZ8zDKP7pcugBYH3gd8UECKSGcoTEcks5YfcjpoWGQFEAgexDaHQfOMiklnKExHJLOWH3JKGh4mIiIiIiEPTlRYREREREXFoalpERERERMSh5crikoULFzZDQkJyY1cicgc2btx41jTNIHvXkVHKEhHHozwRkexyqzzJlaYlJCSEDRs25MauROQOGIZxxN41ZIayRMTxKE9EJLvcKk80PExERERERByamhYREREREXFoalpERERERMSh5co9LZJ/pKamcvz4cZKSkuxdigCenp4EBwfj5uZm71JEMkx54liUJ+LMlCeOJTN5oqZFstXx48fx8/MjJCQEwzDsXU6+Zpom0dHRHD9+nLJly9q7HJEMU544DuWJODvliePIbJ5oeJhkq6SkJAoVKqRAcACGYVCoUCGdVRKnpTxxHMoTcXbKE8eR2TxR0yLZToHgOPS7EGenv8OOQ78LcXb6O+w4MvO7UNMieY5hGPTv3//K92lpaQQFBdG5c+cMbSckJISzZ89m+Tki4ryUJyKSXZQnWaOmRfIcHx8fduzYQWJiIgD//PMPJUuWtHNVIuKMlCcikl2UJ1mjpkXypE6dOrFgwQIAZsyYQVhY2JXHzp07x3333UetWrVo0qQJ27ZtAyA6Opr27dtTvXp1hg0bhmmaV14zdepUGjVqRJ06dRgxYgTp6em5+4ZExG6UJyKSXZQnmafZwyTHvP7bTnadvJCt26xWwp9Xu1S/7fP69u3LG2+8QefOndm2bRtDhgxh+fLlALz66qvUrVuX+fPns3jxYgYMGMCWLVt4/fXXadGiBa+88goLFixg0qRJAOzevZtZs2axcuVK3NzcGDVqFNOmTWPAgAHZ+t5E5OaUJyKSXZQnzklNi+RJtWrVIiIighkzZtCpU6drHluxYgVz584FoE2bNkRHR3PhwgWWLVvGvHnzALj33nspWLAgAIsWLWLjxo00bNgQgMTERIoUKZKL70ZE7El5IiLZRXmSeWpaJMfcyRmHnNS1a1eefvppwsPDiY6OzvR2TNNk4MCBvPvuu9lYnYhkhPJERLKL8sQ56Z4WybOGDBnCq6++Ss2aNa/5+V133cW0adMACA8Pp3Dhwvj7+9OyZUumT58OwMKFCzl//jwAbdu2Zc6cOURGRgK2MadHjhzJxXciIvamPBGR7KI8yRxdaZE8Kzg4mMcee+y6n7/22msMGTKEWrVq4e3tzffffw/YxpKGhYVRvXp1mjVrRunSpQGoVq0ab731Fu3bt8dqteLm5sZXX31FmTJlcvX9iIj9KE9EJLsoTzLHuHoGgpzSoEEDc8OGDTm+H7G/3bt3U7VqVXuXIVe50e/EMIyNpmk2sFNJmaYsyV+UJ45HeSLOSnnieDKaJxoeJiIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yIiIiIiIg5NTYuIiIiIiDg0NS2Sp0RHR1OnTh3q1KlDsWLFKFmy5JXvU1JSsmUfy5cvp3r16tSpU4fExMRs2SbAO++8c833zZo1y7Zti0jGKU9EJLsoT7JO67RItnKkedBfe+01fH19efrpp6/8LC0tDVfXrK2pOnLkSFq0aEH//v2zWuI1fH19iY+Pz9ZtgtZVEOelPMk85cntKU/yF+VJ5jlKnuhKi+R5gwYNYuTIkTRu3JgxY8bw2muv8dFHH115vEaNGkRERAAwdepUGjVqRJ06dRgxYgTp6enXbGvixInMnj2bl19+mX79+hEeHk7nzp2vPD569GimTJkCQEhICK+++ir16tWjZs2a7NmzB4D4+HgGDx5MzZo1qVWrFnPnzuW5554jMTGROnXq0K9fP8AWEgCmafLMM89Qo0YNatasyaxZswAIDw8nNDSUXr16UaVKFfr160dunIQQyc+UJyKSXZQnGZO1lk7kVhY+B6e3Z+82i9WEju9l+GXHjx9n1apVWCwWXnvttRs+Z/fu3cyaNYuVK1fi5ubGqFGjmDZtGgMGDLjynGHDhrFixQo6d+5Mr169CA8Pv+V+CxcuzKZNm/j666/56KOPmDhxIm+++SYBAQFs3277f3P+/Hl69uzJl19+yZYtW67bxrx589iyZQtbt27l7NmzNGzYkJYtWwKwefNmdu7cSYkSJWjevDkrV66kRYsWGf7/I+LwlCfKE5HsojxxyjxR0yL5Qu/evbFYLLd8zqJFi9i4cSMNGzYEIDExkSJFimRpvz169ACgfv36zJs3D4B///2XmTNnXnlOwYIFb7mNFStWEBYWhsVioWjRorRq1Yr169fj7+9Po0aNCA4OBqBOnTpEREToIEMkhylPRCS7KE/unJoWyTmZOOOQU3x8fK782dXVFavVeuX7pKQkwHaZc+DAgbz77rt3vN2bbesyDw8PACwWC2lpaZmq/VYubz8n9yHiEJQnyhOR7KI8cco80T0tku+EhISwadMmADZt2sThw4cBaNu2LXPmzCEyMhKAc+fOceTIkVtuq0yZMuzatYvk5GRiYmJYtGjRbfd/991389VXX135/vz58wC4ubmRmpp63fPvuusuZs2aRXp6OlFRUSxbtoxGjRrd2ZsVkRylPBGR7KI8uTU1LZLv9OzZk3PnzlG9enW+/PJLKlWqBEC1atV46623aN++PbVq1eLuu+/m1KlTt9xWqVKl6NOnDzVq1KBPnz7UrVv3tvt/6aWXOH/+PDVq1KB27dosWbIEgOHDh1OrVq0rN7pd1r17d2rVqkXt2rVp06YNH3zwAcWKFcvkuxeR7KQ8EZHsojy5NU15LNnKkaYUFBtNUSrOSnnieJQn4qyUJ45HUx6LiIiIiEieoqZFREREREQcmpoWERERERFxaGpaRERERETEoalpERERERERh6amRUREREREHJqaFsmT3n77bapXr06tWrWoU6cOa9euzbF9hYaGomkzRfIu5YmIZAdlSda42rsAkey2evVqfv/9dzZt2oSHhwdnz54lJSXF3mWJiBNSnohIdlCWZJ2utIjdRUXB+vW2/2aHU6dOUbhwYTw8PAAoXLgwJUqU4I033qBhw4bUqFGD4cOHc3lh1dDQUJ544gkaNGhA1apVWb9+PT169KBixYq89NJLAERERFClShX69etH1apV6dWrFxcvXrxu33///TdNmzalXr169O7dm/j4eACee+45qlWrRq1atXj66aez542KyHWUJyKSXbIzT5Ql2cA0zRz/ql+/vin5w65duzL0/OnTraa3f5pZqEy86e2fZk6fbs1yDXFxcWbt2rXNihUrmg8//LAZHh5umqZpRkdHX3lO//79zV9//dU0TdNs1aqVOWbMGNM0TXPs2LFm8eLFzZMnT5pJSUlmyZIlzbNnz5qHDx82AXPFihWmaZrm4MGDzQ8//PDK69evX29GRUWZd911lxkfH2+apmm+99575uuvv26ePXvWrFSpkmm12t7b+fPns/weM+JGvxNgg5kLn/3s/lKW5C/KE+VJTn4pT/IXe+eJsuR6Gc0TXWkRu4mKgmEjrQT0XIlv33ACeq5k2Ehrls9o+Pr6snHjRiZMmEBQUBD3338/U6ZMYcmSJTRu3JiaNWuyePFidu7ceeU1Xbt2BaBmzZpUr16d4sWL4+HhQbly5Th27BgApUqVonnz5gD079+fFStWXLPfNWvWsGvXLpo3b06dOnX4/vvvOXLkCAEBAXh6ejJ06FDmzZuHt7d31t6giFxHeSIi2SUn8kRZknW6p0XsJiICvAom4V4kDgD3InF4FUgiIsKHoKCsbdtisRAaGkpoaCg1a9bkm2++Ydu2bWzYsIFSpUrx2muvkZSUdOX5ly/Xuri4XPnz5e/T0tIAMAzjmn38/+9N0+Tuu+9mxowZ19Wzbt06Fi1axJw5c/jyyy9ZvHhx1t6giFxDeSIi2SWn8kRZkjW60iJ2ExICiec9SYn0AyAl0o/EGE9CQrK23b1797J///4r32/ZsoXKlSsDtjGk8fHxzJkzJ8PbPXr0KKtXrwZg+vTptGjR4prHmzRpwsqVKzlw4AAACQkJ7Nu3j/j4eGJjY+nUqROffvopW7duzexbE5GbUJ6ISHbJiTxRlmSdrrSI3QQFwcTxLgwb2RyvAkkkxngycbxLls+KxsfH8+ijjxITE4OrqysVKlRgwoQJFChQgBo1alCsWDEaNmyY4e1WrlyZr776iiFDhlCtWjUefvjh//d+gpgyZQphYWEkJycD8NZbb+Hn50e3bt1ISkrCNE0++eSTrL1BEbmO8kREsktO5ImyJOsM2z0vOatBgwZmXpsrWm5s9+7dVK1aNUOviYqyXYoNCSHLBxg5JSIigs6dO7Njxw57l5JhN/qdGIax0TTNBnYqKdOUJfmL8sTxKE/EWeXFPHHmLIGM54mutIjdBQU5ZhiIiPNRnohIdlGeOBbd0yJyB0JCQpz2TIaIOBbliYhkh/yWJWpaRERERETEoalpkWyXG/dJyZ3R70Kcnf4OOw79LsTZ6e+w48jM70JNi2QrT09PoqOjFQwOwDRNoqOj8fT0tHcpIpmiPHEcyhNxdsoTx5HZPNGN+JKtgoODOX78OFFZXYZasoWnpyfBwcH2LkMkU5QnjkV5Is5MeeJYMpMnalokW7m5uVG2bFl7lyEieYDyRESyi/LE+Tnd8LCoKFi/3vZfEZHMUpaISHZRnojkPKdqWmbMMAmpkE7H3gmEVEhnxgyNSxSRjFOWiEh2UZ6I5A6nGR4WFQXDRloJ6LkS9yJxuEf6MWxkc9q1s2jhHxG5Y8oSEckul/OkQFg4bgWSlCciOchprrRERIBXwSTci8QB4F4kDq8CSURE2LUsEXEyyhIRyS6X88StQBKgPBHJSbnetGR23GdICCSe9yQl0g+AlEg/EmM8CQnJ9hJFxElkJk+UJSJyI5nPEw9Mq+371Bgv5YlIDsnVpiUr4z6DgmDieBdi5zYnfkYosXObM3G8iy6/iuRTmc0TZYmI/H8zZpiEVE6hY+/4DOfJB58lY1w6mkqNKKo8EckhuXZPS3aMIw8LM2jXzkJEhA8hISgURPKprOaJskRELouKgoceT6LQ4HBc3KwZzpOQehdgD/i5u1G9QwxhYUbOFy2SD+XalZbsGkceFAQNG+ogQyQ/y448UZaICNjyxNM7HRc32xivjObJthMxuFtceKBJKfZExhKfnJZjtYrkZ7nWtGgcuYhkF+WJiGQXW554kRLlC4BphRQz/Y7zZNuxWKoW96N5hcKkW002HTmfY7WK5Ge51rRoHLmIZJfLeZK8qzTJW8sSH16Tb8cZyhMRybArxydzWpDwa2PMZDcqD96Eh1/qbV9rtZrsOBFLzeAA6pcpiMXFYO3h6FyoWiT/ydV1WjSOXESyS1iYwbKUGBbuPoFH7cN8ccSTHT8XoU3lIjSvUBgvd4u9SxQRJ/Hf8Ulhoi31GT1nLY/P2MzEgQ2xuNz8HpWI6ATiktOoVbIAPh6u1CgZwLrD53KxcpH8I3ealuQ4iFgJnv4E+RUnqH4guDjNEjEi4iiSLsDBJeDqCT6FGde3PJEpVQjfE8XiPZH8svkE09cexcPVhablC9GmShFaVy5CqUBve1cuIo4m6QIcWgpu3uBbhKCCxQgK8gAK8VpKdV6av4MP/trD8x2r3nQT20/EAlAzOACAxmUDmbIygqTUdDzddOJEJDvlTtMSfQCmdPrve4sHFK4EQZWhZH0o3RiK1QKLW66UIyJO6txB+PG+a35UxLMAfYKq0KdIVVK7NWCrUYU/jnuxeG8kr/yyE9hJpaK+tKlSlDZVitCgTEFcbnHmVETyiXMH4Yeu1/7MPxiCKtO/WA2oWoyPl16gajF/7qtb8oab2HY8Fg9XFyoWsd0P07hsIBOWHWLz0Riali+U0+9AJF/JnaalcCUYMA4SYyD+DMQchai9cHQ17Jhje45HAFRoC5U7QuVO4OGbK6WJiBMJqgyDJ0LqRbh4DuJOwfkIiNwDO+bhtvE7GgANCpThler3cLJ4WxbGlWfR3mgmLj/E+KUHqV7Cn5c7V6NJOR1QiORrhSvDoAmQkgDxkRB3GqL3Q+RuWP01/a2p9PeErfPLc+pYN4o37WvLoKtsPx5L9RL+uFpso0cahARiGLDu8Dk1LSLZLHeaFncfKBd648cunISja+DgYtj3F+ycZ7tUW6Uz1B8EZZqBobOiIoItG8o0u/FjViuc3QtHVsL+f2HTj5RIm8BQvxIMrdWbuC4P8tcJLz79Zx99J6zhnupFeb5jVUIK++TuexARx+DuDSEtbvxYahKc3MzF/eFYVs6j6KZPYdMnULwO1HkAaoeR7u7HjpOx9GlQ6srLArzcqFrM/9LN+BVz5W2I5Be5eiP+DfmXgBo9bF9WKxxbC9tm2ZqX7bOhaA1o8jDUul/Dx0Tk5lxcoEhV21fDYZCaCHsXwrbZsOpL/FZ+Tq9K99Cl92gmHi3NV0sOsGTPMsZ0qMzQFmUxdHJERC5z84QyTfEu0xTXGo8Q+vUfDPDbyBBzLS4Lx8CiN4ir1JPA1HrUCq5xzUsblQ1k5vqjpKRZcXfV/bsi2cWxPk0uLlCmKXQZC0/ugS6f237+yyPweT3YMBnSbz8FoYgIbl62kyEPzIQndkDLZ+D4Bjx+7MwjEY+xso8rLSsF8daC3QyZsp7o+GR7VywiDqhKMX9evD+Ut6JDebrg55jDFkPVLvjvnMYS9ydpt+8N2zDVS5qUCyQp1cr2EzF2q1kkL3KspuVq7t5QfyCMXAEPzAbfIvD7E/B1E9j9G5imvSsUEWfhXwLavAj/2w4d3odzhyk4pyffur7PZ63dWXkwmo6fLWfVgbP2rlREHNA91Yvx1N2VmLf5BN8eKgDdx/NptZ+YSXv8DvwCXzaEP1+Ai+doGBIIwFpNfSySrRy3abkk6qzB+th7iOr2L4TNBMMCs/rDD93g7H57lyciTiQq1pv1lpFEPbAZ7n4T49hauq3pw+o6/1DUI4V+k9bywZ97SE232rtUEXEwo9tU4N6axXl34R7mr43kz8MezAt6HOOxzbYh7GvHwRf1KbR3JpWCvFl7SE2LSHZy6KZlxgyTkArpdOydQEhFKzM2dYCHV0Gnj+DkFvi6KSx5B9JS7F2qiDi4a/Kkihszjj4Kj22B+oMptGMyv/A/Xq94mK/DD9Lnm9UcO3fR3iWLiAMxDIMPe9eiuKc/j83czL6oWFb+7s+MBcWh25e2kSFBVeC3x/g2/SXOH9lOmk6AiGQbh21aoqJg2EgrAT1X4ts3nICeKxk20krUOVdo9BA8ugGqd4el78O3beDUNnuXLCIO6qZ5khAInT+Bhxbh4luEAUdfZEWFaUSdOUWnz5cTvjfS3qWLiANJiHVl+4S6GG7pGBYT97KnbVkSBRStDoP/gPvGUSztOD/xLFF/fgDWdHuXLZInOGzTEhEBXgWTcC8SB4B7kTi8CiQREXHpCb5FoOe30HcGJETaGpeVn9tmIBMRucpt86RkfXhoCbR+keCTf7PU9wXu9d3PsO83MG/TcXuVLSIOJiICPNwMDBfbfbWuAUnXZolhQJ0HuDB4OUusdSi+/l2Y0hlijtmrZJE8w2GblpAQSDzvSUqkHwApkX4kxngSEvL/nlilE4xaA5U7wD8vw7SeEB+V2+WKiAO7ozyxuEGrMfDQYixeAbwb/xIfB87nmdmbGL/0IKYm/xDJ9+702CSoeGne83uBiYWfg9PbYHxz2yRCIpJpDtu0BAXBxPEuxM5tTvyMUGLnNmfieBeCgm7wZO9A6PMjdP4UjqyCb1rCsfW5XrOIOKYM5UnxWjA8HKNuf7rFz2Jh4Cd8u3Atb/y+C6tVjYtIfpaRLGlUrhBfRNfHOnwZBJazTSL0zyuQnpb7hYvkAQ7btACEhRlEHLCwcI4PEQcshIXdYvE3w4AGQ2DoP7Yzpt91tK3rgm08+/r1tv+KSP6UoTxx97HdWNvtayqm7GaJ3ytsXLWI4VM2s2pNurJEJB+70yxpXLYQsYmp7E0NgiF/2Y5RVn4GU7vDxXM6NhHJIIduWsB2VqNhQ258RvRGiteCEUuhXCv4/Qn2jn2G8hWTbTMGVUhnxgydKRXJrzKcJ3X7YQz9Bz9vT+Z4vIXHgZ/p880aQiqnKEtE8rE7yZJGZW3rtaw7fA5cPWyjQbp9DUfXEvdpG+5ptEvHJiIZ4PBNS6Z4FYQHZnOx9iNUjpnALyM6ULTvwv9mDNJZDRG5U8VrEd1jCRtO1OEr9895osRkAh9czLCRuuIiIjdXKtCbkgW8WHs4+r8f1u3H+ft+JzEmnvAH23LPA1/o2ETkDuXNpgXAxcLO4Hf438oPuctzC7Pc36RkkaPXzvIhInIHDkcG0WvRTOam38VTbnN43288BervV5aIyC01KhvIusPnrpnI40BSIzru+I4TFGKK2wf0KbZQxyYidyDvNi3YZvn4dt1QBka9TFnjFPNcX6ckx6+fgUxE5BZCQiA2OoBHTzzD52n38YDrEn5o9jAJLiftXZqIOLDGZQM5G5/CwagEAEzTZOGJvUTXvkCvuDdYb63MWPevebjaVzo2EbmNPN20XJ7lY/60x+n223w8U6ysHXYvQWlaiFJE7tx/Mwa14I0fv+bxpe/T2rKFgN97cvasxnSIyI1dfV9LSpqVp37ayuQ1B2gQGEzElC70/uUnfklpxpuN36bg5ldBU6uL3FSealpuNBPH5Vk+3v2mGZZhf+Lp425b6OnoWvsVKiIO7//nydUzBr00eyQn2n1BDeteLozvQFqcGhcRuV7Zwj4U8vFg7poz9J+wnnmbTvBEu0r89EwtIva5MX9SMKmdJzE1rS2uq8Zi/v6kFskWuYk807TMmGESUiH9hjNxXJ7lI7BSRRjyJ/gUhqk9bGu6iIj8PzfLk6tnDCp114Osb/olJVKPEP31PVrUVkSuM3MmnNhakI2nIll7OJpepWrxeLuKGIZxJU96NStBZMt3GZfWBWPjZPjtUTUuIjeQJ5qWqCgYNtJKQM+V+PYNv/VMHAVKw6AF4FccpvaEw8tzvV4RcVwZyZNmHR5gVoUP8b94jAvf3APxkblfsIg4pMtZYgmKASAtzpNxL5S4YZb8r10l1pV7lC/Se8DmqfDLI2BNz92CRRxcnmhaIiLAq2AS7kXiAHAvEnfrmTj8i9sal4BSML0PHFmdW6WKiIPLaJ6EhQ3kvcA3cL1wnORJ90LC2VyrVUQc1+UscQtIAsCtQOJNs8TFxWBs33rM8R/ANy73w9bp8OtjuuIicpU80bSEhEDieU9SIv0ASIn0IzHG89YzcfgVhUG/g39JmNYbjm/IjVJFxMFlNE/cXV14ZPBgnnR9Hs5HkP59N7h4LtfqFRHHlNEsCfB245sH6zM2tQezfPvDlqnwx1O6OV/kkjzRtPw3s09z4meEEju3ORPHu9x+1WvfIjDwV9s9Lj/2gFOaVUwkv8tMnhTx92TYgwMZmfok1si9mFN7QHJc7hUtIg4nM1lSpZg/7/eqxbNnO7K86IOwYTL89YIaFxHySNMC187sE3HAQliYcWcv9C8BA38DDz/4sTucPZCzhYqIw8tMnjQICST03jBGpjyGeXIbzAiD1KRcqFZEHFVmsqRr7RIMa1GOB4904EC5/rDma1j2YS5UK+LY8kzTAtfO7JMhBUrBgPmACT/eB7HHs784EXEqmcmTAU3L4F+7K0+mjoCI5TBnMKSn5VyRIuLwMpMlz3WsQpNyhei8rxPnK/aCJW/D2m9yrkgRJ5CnmpYsKVwR+s+DpFjbrGIaky4iGWQYBu90r8neIp14myGw9w/4/XEN7RCRDHG1uPDlA/Uo6OPJfcf6klKhEywcAzvm2rs0EbtR03K1EnWg7zQ4dwhmPgCpifauSEScjJe7hW/612e20ZFpnmG26UuXvGPvskTEyRT29WB8//qciktjZNIozNJN4eeRcHiZvUsTsQs1Lf9f2ZbQ/Rs4ugbmDtM86SKSYaULeTO2bx1eiu3M6oB7YdkHsH6SvcsSESdTu1QB3ryvOosPXOCzIm9AYHmY2Q9O77B3aSK5Tk3LjdToAR3ehT2/w98v27saEXFCrSsX4cl2lel/pi/HCt8FfzwN+/62d1ki4mTub1iaBxqXZuyKKJY0+BrcfW1rzF04ae/SRHKVmpabafIwNH4Y1nylm99EJFMeaV2B1lWLc+/JISQEVoOfBsGprfYuS0SczKtdqlG3dAFGL4jkSIcptvtvp/fR1OqSr6hpuZV73obK98Kfz8G+v+xdjYg4GRcXg0/ur0OhwEB6xv6PdM8CMP1+nSEVkQzxcLUwrl99vNwtDF6YSMJ9k+HMLpgzRMPYJd9Q03IrLhbo+S0Uq2kLhjM77V2RiDgZf083xvevz5FkP55xexEzOQ5m9IWUBHuXJiJOpFiAJ189UI+j5y7yvw2FsHb6CPb/DX+/ZO/SRHJFrjctu05e4JmftnIwKj63d5057j4QNtO2+OT0vhAfae+KRMTJVC7mxwe9ajHvZAGmBb8Kp7fDzyPAarV3aSLiRBqXK8SL91bln11n+OrCXZeGsX8NGybbuzSRHJf7TcupC/y69STtPlnKI9M2seNEbG6XkHH+JSBsBiREwawHIS3Z3hWJiJPpUrsEw1qU5aVdJdlW7WnY/RuEv2vvskTEyQxqFsJ9dUrwyb/7WBLyOFS4G/54BiJW2Ls0kRyV601Lr/rBrHi2DQ+3Ks+yfVF0/mIFg75bx/oIB1/MsURduO8rOLYGFjylxeJEJMNsq1wH0ntrXc5X6mObCnnnz/YuS0SciGEYvNujFlWK+fP4rG0cbfMFBJaD2QPg/BF7lyeSY+xyT0uQnwdjOlRhxXNteOaeymw7Hkvv8avpM3414XsjMR21IajRE+56Gjb/COsm2LsaEXEyV1a59vag57FepJVoCPNH2YaLiYjcocuL2BqGwfDZ+0ns+SOkp9kWxk52kuH3Ihlk1xvxA7zceKR1BVY+24ZXu1Tj+PmLDPpuPZ2/WMGCbadItzpg89L6RajcCf58XpdiRSTDCvt6MK5/PY5fsPKE8TSmV0HbgcZFB7/aLCIOpXQhbz4Pq8veM3E8G56I2WsyRO6CX0drNIjkSQ4xe5iXu4XBzcsS/kxrPuhVi8SUdB6Zvom7P1nK7A3HSElzoJtVXVyg+zeXLsUOhNjj9q5IRJxM3dIFea1rdX47mM60Mm9D3BmYM9h2plRE5A61qhTE0+0r8+vWk0w+Ux7avmobcrryM3uXJpLtHKJpuczd1YU+DUrxz5Ot+OqBeni6WRgzZxuhHy7hu5WHSUxxkLnIPf2h73TbDfmz+kNqkr0rEhEnE9aoFH0aBPPSend21nsNDoXDotfsW5SIOJ1RoeW5p3pR3vljN6uL9Ydq98Gi1+HAInuXJpKtHKppucziYnBvreIseKwF3w1uSHBBb17/bRct3l/MV0sOEJuYau8SIagS9PgGTm6Ghc/YuxoRcTKGYfBGtxrUCg6g77ryxNYYCKu+gJ3z7V2aiDgRwzD4qHdtQgp5M3rGZk63/hiCqsDcoboxX/IUh2xaLjMMg9aVizB7ZFNmj2hKzeAAPvxrLy3eW8wHf+7hbLydpx6ucq/txvxNP8DG7+1bi4g4HU83C1/3q4erxaDvka6kl2gAvzwCUXvtXZqIOBE/Tze+ebAByWlWRszeS0qv78GabptRTKNBJI9w6Kblao3KBjJlcCN+f7QFLSsHMW7pQZq/t5hXf9nB8fMX7VdY6xegXGvbHOknNtmvDhFxSsEFvfkirB57zybziscYTDcv27DT5Dh7lyYiTqRCEV8+6l2brcdieHVFEnQfD6e2aDSI5BlO07RcVqNkAF89UI9/n2xFtzolmLb2KKEfhvPU7K0ciLTDNH8uFug5CXyCbDfmawYgEcmgFhULM6ZDFabtTmNBpbch+gD89rhmABKRDOlQoxiPtC7PjHXHmHGhJrR40jYaZPM0e5cmkmVO17RcVj7Ilw961WbpmNb0b1KGBdtPcvenS3l46ka2H4/N3WJ8CkGf7yHuFMx/GKwONNuZiDiFES3L0bFGMR5f68eROk/BjrmwfqK9yxIRJ/Pk3ZVpWSmIV3/ZyeYKoyDkLtui2Kd32Ls0kSxx2qblspIFvHita3VWPtuGUaHlWXHgLF2+XMGAyetYeyg69xaqDG4A97wD+/6ElZ/mzj5FJM8wDIMPL91M23NrQ5LK3m1bD+r4RnuXJiJOxOJi8HnfOhQN8ODh6ds42+Fr8Ayw3d+SdMHe5YlkmtM3LZcV8vXgmXuqsPK5NozpUJmdJ2K5f8Iaeo9fzZI9kbnTvDR6CKr3gMVvQcTKnN+fiOQpvh6ufPNgA5LSYciFYZh+xeGnQRp2KiIZUsDbnfH96xOTmMIjv5wgreckOB8Bvz2mYafitPJM03KZv6cbo0IrsOLZNrzetTqnYpMYPGU9nT5fwe/bTpJuzcEPq2FAl8+gYFmYMwTio3JuXyKSJ12+mXbViXS+DnrZNuz0l0d0oCEiGVK9RADv9qjJ2sPneGdHILR5ybbwpIadipPKc03LZV7uFgY2CyH8mVA+7FWL5LR0Rk/fTLtPljJr/VFS0nLovhNPf9v9LYnnYd5Dur9FRDKsQ41iPBxang93+LCpylOw9w9Y/aW9yxIRJ9O9bjCDmoUweeVhfvHtDRXbw18v2NaYE3EyebZpuczN4kLvBqX454lWjOtXDx8PC8/O3U6rD5cwecVhLqakZf9Oi9WEju/DoSWw4pPs376I5HlPt69MiwqF6butNjEhHeDf1+D4BnuXJSJO5sV7q9IoJJBn5+1gb9OPbLOd/jRY97eI08nzTctlFheDjjWL89voFvwwpBGlA7154/ddtHh/CV8u3k9sYmr27rD+IKjRE5a8DUdWZe+2RSTPs7gYfB5WlyBfT/qc7Ee6bwnbgUbieXuXJiJOxM3iwpf96hLg5cawOQeJ6zweYo5qWnVxOvmmabnMMAxaVgpi1oimzBnZlDqlCvDR3/to/t5i3v9zD1Fxydm1I+g8FgqGwJyhkBCdPdsVkXwj0Md2M23ERTde93gKM+4k/DJaBxoikiFF/DwZ178+p2OTGL3CE2vrF2HnPNj4nb1LE7lj+a5puVqDkEAmD2rIgsdaEFo5iG+WHqTF+4t55ZcdHD9/Mes78PSH3lPg4lnb+i060BCRDKoZHMDb99Xgh2NBLAoeBXt+1420IpJh9UoX5PWuNVi6L4pPEztB+Ta2adXP7LR3aSJ3JF83LZdVLxHAlw/UY9FToXSvW5IZ644S+mE4T83eyoHIuKxtvHhtaP8W7P8L1ozLnoJFJF/p3aAU/RqXZti+xkQWa2W7kfbUNnuXJSJOJqxRKe5vUIovlhxiSbW3bOu3/DQYUhLsXZrIbalpuUrZwj6817MWy8a0ZkDTEBZsP8ndny7j4akb2X48NvMbbjQcKneCf17RjB0ikimvdKlG3dIF6XGyP2keBW3TqifH27ssEXEihmHwerfq1A4O4NFfj3OyzVg4uw/+fM7epYnclpqWGyge4MUrXaqx8tk2jG5dgRUHztLlyxU8OGktaw5FZ3yhSsOAbl+Bb5FLBxpZvHojIvmOh6uFcf3qk+RekDHmaMzoA7BwjL3LEhEn4+lmYVz/+ni4ujAg3IeUpo/Dph9gx1x7lyZyS2pabqGQrwdPta/Mqufa8GyHKuw+FUffCWvoNX41i/ecyVjz4h0IPb61rUj7xzM5VrOI5F3FAjz58oF6/BJbgQUF+sGWabDtJ3uXJSJOpkQBL754oC6HzybwvzOdMIMbwm//sx2jiDgoNS13wM/TjYdDy7Pi2da82a06p2OTGDJlAx0/W86vW0+Sbr3D5iWkObQcA1tnwNZZOVu0iORJTcoV4vmOVXj89D2c8q8Nvz8B5w7ZuywRcTLNyhfm+Y5V+GPXWaaWfNn2w7nDID2bl4AQySZqWjLA083Cg01DCH8mlI971yY13cpjMzbT9uNwZq47SnJa+u030vIZKN0UFjwJ0QdzvmgRyXOGtihLp9ql6B01hDQT27TqOtAQkQwa2qIsXWqX4NVl8exu8CYcXw/h79q7LJEbUtOSCW4WF3rWD+afJ1oxvn99/DzdeG7edlp9EM6kFYe5mJJ28xdbXG3DxFwstjMaaSm5V7iI5AmGYfB+z5r4FCnHc2kPwclNtoVsRUQy4HKWVCrqR9jqEsRXC4Pln8DhZfYuTeQ6alqywMXFoEONYvw6ujk/Dm1ESGFv3vx9F83fW8wXi/YTe/EmZz4LlIKuX9gONMLfyd2iRSRP8HZ35ZsH6/OX2YS/PNpjrhgLh5bauywRcTKXs8RqNXnwZE+sgeVh3gi4eM7epYlcQ01LNjAMg7sqBjFzeFPmPtyUeqUL8vE/+2j+/mLeXbibyLik619UrRvUGwgrxsKh8NwuWUTygJDCPoy9vw7/i+1LlHspzJ9HQEK0vcsSESdTppAPn/Wty5YzKXzq/yxmQhT8MlqLYotDUdOSzeqXCWTSoIYsfPwu2lQpwrfLDtHi/SW8PH8Hx85dvPbJHd6DwhXh55E60BCRTGlbtSgPta3J4LiRWOPPwq+P6kBDRDKsdZUiPNGuEl/s8WFDhcdg7wLYMNneZYlcoaYlh1Qt7s/nYXVZ/FQoPeuVZOb6o4R+FM6Ts7ew/8yldVrcvaHnJLgYDb/qjIaIZM7/2lakSKWGvJfWVwcaIpJpo1tXoF3Vojywox4xJVrCXy9A5B57lyUCqGnJcSGFfXi3Ry2Wj2nDoGYhLNx+mrs/XcaIHzew9VgMFK8F7V6HvX/A+on2LldEnJCLi8HY++vyt293Vht1MP96ASJ327ssEXEyLi4Gn9xfm+BAX+4/MwCrm49tUezUGwxzF8llalpySbEAT17uXI2Vz7XhsbYVWX0wmm5freTBSWtZHdQbs0I7+PslHWiISKYEeLsx7sGGPJM2kgtWD8w5Q3WgISIZ5u/pxjcP1udYqh/vuD8GkTvh39fsXZaImpbcFujjzpN3V2Llc214vmMVdp+KI2ziOobEDCbF4oOpMxoikknVSvjzdI+WPJ40HEMHGiKSSZWK+vFR79pMPFORVYV7wdpxsO9ve5cl+dxtmxbDMB41DKNgbhSTn/h5ujGiVXlWPNuaN++rwf4Eb4bHDcGI3MWhGU+Slm61d4mSV9lxbSDlSc67r25JQpp057u0e3SgITnv7AG77Vp5krM61SzOiFblGHy8MzF+FWH+wxAfae+yJC/b88ctH76TKy1FgfWGYcw2DKODYRhGthQmAHi6WXiwSRmWPB1Kt96DmOfehXKHpvHCBx8zY91RktPS7V2i5DX2XYRQeZILXry3Kv+WfIQ9ZmnSftaBhuSQtBSYO9SeFShPctgz7SvTsEIJ+p0fjjU5zta4WHVSVXLA2QO3zZPbNi2mab4EVAQmAYOA/YZhvGMYRvnsqFFs3CwudK8bzH1Pf0tcQGVeSPmSj+etoOUHS5i4/BAJyWn2LlHygkPhsPIzu+1eeZI73CwufNq/Ma+6PUl64gVS547UgYZkvyVvw6ktdtu98iTnuVpc+DysLjG+FfjEGAAH/oV139i7LMlrLp8AcfW45dPu6J4W0zRN4PSlrzSgIDDHMIwPslqnXMvF3Qu/ft8TYEni77IzqFDYm7cW7Kb5+4v57N/9xFy039AecU5RUbB+PZw9Gm1bE6hwRbvWozzJHUX8PBnzYDfeSe+P2+FFWNeMs3dJ4uQuZ0lUFP+dAKk/yK41KU9yXqCPO988WJ9vk9qw0bMJ5j+vwOnt9i5LnNw1ebLkLdsJkK5f3PI1d3JPy+OGYWwEPgBWAjVN03wYqA/0zHrZcp0iVTHueZvAU8uYVnMz80Y1o0GZQD79dx/N31vMu3/sJvKCbtaX25sxwySkQjode8ez+oXRpMdHQ0/7Ta2tPMld9csEUr7jY/yTXh/rP6/CqW32Lkmc1H9ZkkD9GpFcnH7pBMg979itJuVJ7qlRMoC3u9diWMwg4l38Yc5QSLl4+xeK3MDVeTKo9WLMlZ9D/cFQtcstX3cnV1oCgR6mad5jmuZPpmmmApimaQU6Z710uaEGQ6FyJ/j3Neq5HWPiwAb8+b+7aFetKN8uP0SLD5bw0vztHDun0JAbi4qCYSOtBPRcyYh+L9Olwh+8GP4yUa617VmW8iSXPdg0hGVVXyHa6kPCjIE60JAMuzpLfPsuYcLA/liSojnXZhK4+9izNOVJLupVP5guTWsyMuEhOLsX/n7R3iWJE7qcJwX7L6FU31+ZdN9Q9kZXIKr+7U+A3Mk9La+apnnkJo9pUZGcYhjQ9UvwCrSN80u5SJVi/nzWty5Lng6lV/1gZq8/TuhH4Twxawv7zsTZu2JxMBER4FUwiWpFd/Oy648sTa/FpEODiYiwX03Kk9xnGAYv9r6LsX5P43XhMHG/jrF3SeJkLmeJe5E4+lkW0cF7LW9sfJGDCbXsWpfyJPe9dG81kku1ZJK1C2yYDLt/t3dJ4mQiIsArMBFXvyQ+cPuWgpYLjFzxFREnvG/7Wq3T4sh8CkGPb+Dsfvjr+Ss/LlPIh3e612T5s60Z3CyEv3aepv2nyxj+wwa2HouxX73iUEJCwHoBPjO+Jg4vHjv1NBdjvAkJsXdlkts83SyMGjKU7+mK344fSd4+394liRMJCYHE856ExJ3lZdcfWZxYj89WPaIsyYfcXV34ul89Jrr3Y69Leay/jIbYE/YuS5xISAikeV6kv+Vf7rZs5O2YQaw/1OCO8kRNi6MrFwrNH4ONU2DXr9c8VNTfk5c6V2Pls214vG1F1h4+R7evVtJ/4lpWHTyL7f5Eya+CgmD1K69R3T2CR//9igM/dWbieBeCguxdmdhDqUBvyt//LtusZUn7+VHM2OP2LkmcRFAQTPwqmXEFPibO9GbY1B/5dryrsiSfKuLvyRf9GzMq6RFSkhMx5w0Hq5ZnkDsTFATDBq3mJdep/HsilE+mvHfHxyZqWpxB65egRF349VG4wYFGQR93nri7Eiufa8MLnaqw90wcD3y7lu5fr+KfXWewWtW85Et7/6RyzAQu1nqYp8Z2JeKAhbAwLWOQn7WsWpKtjT+G9GROTxmgAw25Yw1Sn6Sq2xFWlf+QtVuClSX5XIOQQAZ1acvLKQMwjqyAFZ/auyRxEmlJCYxMfIlkiy+FBn7L4QNud5wnalqcgas79JwE1jSY+9BNDzR8PVwZ3rI8y8e05u3uNYhOSOahHzbQ8bPlzN98grR0rdOQb1w4Bb+MgqI18e76Og0borOiAkC/jm2YXeQxip/fyPFf37J3OeIEknb8TsUjM/jN+z66DuijLBEA+jcpg7VWP35Lb4J1yTtwbL29SxInEDlvDBU4yv5mH1K3RZEM5YmaFmdRqDx0+giOroJlH93yqZ5uFvo1LsOSp0IZe38dTEz+N2sLbT5eyrS1R0hK1dnVPM2aDj8Pt80S1WvSbRdrkvzFxcWg55AxLHa9i2JbxnJuz3J7lySO7MJJzPmj2GktQ3Dv99Gi83KZYRi83aMmUwv/j5NmIKmzB0NSrL3LEke2ZwEl9k1litmZGq16ZPjlalqcSe2+ULMPLH0Pjqy+7dNdLS7cV7ckfz7ekgkP1qegjzsv/ryDlh8s4dtlh0hITsuFoiXXrfwMDi+DTh9AUGV7VyMOyN/LndIDvuGUWZi02UNIiT9v75LEEVnTSf5pGGZqEnPKvUHdssXsXZE4GE83Cx8PaMWLxuO4xJ0g7ZfHQffTyo3EnsD85RF2UZadVR7H082S4U2oaXEmhgH3fgwFysDcYZB46wONy6uNRkcbtK9ejPmjmjFtWGMqFvXl7T920+y9xXz6zz7OJ6Tk0huQHHdsPSx+C6p3h7oP2rsacWAVSpfkaOsvKJgezb6JQ257oHHN6sWSP6z4BI9jK3kjfTBDura3dzXioIILevPQA2F8mtYL190/Y27+8ZbPV5bkQ9Z0mDec9NRkRiWPpmPdMpnajJoWZ+PpbxvyE3/admP+TQ40rl5tNKRCOjNmmBiGQfMKhZk2rAk/j2pGo7KBfLZoP83fX8zbC3Zx5kJSLr8ZyVaJMTB3CPiXhM5jbU2uyC00D+3AilIjqBGzmI0/j73p826UJ5LHHV2DueRdfk1vin/TgZQKvP0aCpJ/tahYGL92z7AyvTppvz8DUXtv+DxlST617CM4soLZQY9z3rM0LSpk7sY4NS3OqGR9aPsq7P4NNky67uFrVy8OJ6DnSoaNtF5zVqNu6YJ8O6ABf/2vJe2rFWXSisPc9f4SXvh5O0ejtWK20zFN+O1x23z5vSaDVwF7VyRO4q5Bb7LdvS7Vt77N/u3rrnv8TvJE8piL5zDnDuW0EcQnHg/zaNuK9q5InMDwVhX5vcKrXEh3I2HaAEhNvOZxZUk+FbESlr5Heo0+vHOyDvdUL4q7a+baDzUtzqrpaKjQDvPPF9ixaPs1H/qrVy8GsPgm4+aVwubN12+mcjE/xvatS/jTrenVIJg5G44T+tESHp+5mT2nL+TOe5Gs2zgFds2Hti8T5dlQl97ljrm6ulJi6A9cNLyxzBtC+NKYTOeJ5AGmCb8+innhNCMSH2HwXfXZs91NeSK3ZRgGL/ZtyyfeT+ATs4cj3z2rLMnvLp6z3c5QMISFpZ8hPjmdyp4lM50nalqclYsL88xxnI4tiMfvA6leNfbKZdbLqxenRPqRsLs4Jya0IiHFSvdeN78UW7qQN+90r8nyZ1sz7K5y/LPrDB3GLmfY9xvYfFQ36Tqyc3t2Yv3jOVJKtWbG0cd06V0yrFDR0vwb9AXlzGOc/HUEIRXTspQn4pyiouDInImw53c+4wFivRowqmsx5YncMV8PV4oWHc6E5M6UOfk9L7zyPm3eXM2Dk9YydtM63BrvwprkSuo5b05+11xZkodFRZrETBmJmXAWen3H+39GkXbeh2dHe2Y6T4zcWDW9QYMG5oYNG3J8P/lJVBS2g9MHxjG7yEvMSWjN4G/nEHHAQlCQbdzo0OFWkpJNivVfhXuROFIi/Yid2/zKc27lfEIK36+O4LuVEcQmptKsfCEeaV2BZuULacpLBzJ7Wjy11oRS0DuGxlOWczq+CIH3r7zt79swjI2maTawT9WZpyzJGZfz5L3hw3jUZx6Pxj3O5ImvZlueiOObMcPki5e3sCSsPUvjGjDCcxTnfmqOT6sdyhO5Y5ezJKjvv8wv+hxljdO0P/YVRWoXBReTqGgrx09asRRMwJrkisUrTVmSB82YYbJ1whe81+plnlryHr49+/L9yVWkxXng6pec6TzRlRYndfky64aAsnye3p0+vosZ1OAHIiJsj4eFGcyfayGgaPKVS7HuReLwKpB05Tm3UtDHnf+1q8TK59rwYqeqHIiMp9/Etdz39Sr+2nkaq1VnRewtKgpS5j9NpcADPGF5iPhGp0j3SMrU71vyt8t5MtbSnfXWSrzr+w01G67MtjwRxxYVBU88FsPU+8M45+LLGI8BJGwvg2uah37fkiGXs4SCaYxOfQwDGOf1Ga82achvj7ZgzWstmdwrlOSdpbF42ZZd0N+tvCUqCsa9up43W77Gn+kNmVG+GuMXR2CmWnD1SwYy/ztX0+Kkrh6y8XlaD1Ym1eSDJi9Q3n/PlefUrQspF2zPAUiJ9CMxxpOQkDvfj6+HKw+1LMeyMa15p3tNziekMOLHjdwzdhk/bz5OWro1e9+Y3LG4FdPpX30GX6TfxyprDbzKRpMe75Gl37fkT5fzJDGyAI+lPEoqrkxs+iguPmevPCc78kQcU8Rhk687Pk4Z19M8ljKaGMOf9N2luRit37dkzNXHJsfNIjxx9nHqBW2h2snXrjynbl2IW1kFa6rtEDTlrI/+buUhx/af58cuQzhDQcakPoR7kXg8K54icVdwlvNETYuTCgqCieNdiJ3bnAsz2jDgh1lYPH0I/HcgpCRc95z4GaHEzm3OxPEumbr86ulm4YHGpVn8VCs+61sHF8PgiVlbaf1xOFPXHCEpNT2b36HcUuRuyu58kqXHWvDhyYEApMd74OZqEDsn679vyV+uzor9M3ry0F9fUcMSwcn5j1xZhDY780QcS5WEyfQo9wsfxoex3qxCWpwHSdF+jP1Ev2/JmP+fEzOnPsO+Ag/hvfUr2P37f8/52pWYn5pjWg3STxXS3628wjSpvv8RinufZsSZ57iAL2nx7hgWkyc6h2Q5T3RPi5OLirJdjg0JgaC4cPjhPqgdBt3H3fg52RQKVqvJoj2RfLXkAFuOxRDk58GwFmXp16QMvh6u2bMTubGUBJjQGhLP83OxZfQfXQSvAkkkxngycbwL7doZt/19awy63MjVWZG85BmCd01gcrGXGTziqSv3suVEnogdndoKE+/mmHtzWpwciTXBi3NzmjFxnIWwMOOOft/KE/n/rvl7UzAZJt8D5w7BiGVQMOTKc56bu41lx06wfExrigV42rNkyQ6rv4a/nmdT0Dvc9cJIvAIv4tV5LdWK+fPXi42ynCdqWvKaJe/C0veg6xdQb0CO7840TVYfjObr8IOsOHCWAC83BjYLYXCzEAr6uOf4/vMd04SfR8K2WTBgPpQLzdRBpA4y5LbSUzn1WVv8YveysOkMendoY++KJLslxcI3rSAtmddKjOfH7Qm836YFrWr7Z6ghVZ7IbZ07bPu7VqgcDPkLXD0AOHbuIq0/CueBxqV5o1sNOxcpWXJsPXzXASp1gPunEnXW4MdlJ/h8/RamDG5IaOUid7QZ3Yifn7QaA+VC4Y9n4NS2HN+dYRg0q1CYqcMaM/+R5jQuG8jni/bT/P3FvPX7Lk7HJuV4DfnKximwbSaEPmf7PWNrVBo21FlvyWYWN4oNnY5p8aD2qkdZu+eYvSuS7GSaMH8UxB5je9NPmbI1nuEty9GrXcYaFpE7ElgW7vsaTm6Gv1648uNSgd70qh/MzHXHOBWbeIsNiENLOAs/DQT/ktDtKzAMgoJg+ZkIyhX2oWXF7AkVNS15jYsFekwEr4Iwe4DtTFouqVOqABMGNODvJ1pyT/VifLcqgpYfLOH5eduIOJuQa3XkWSc3w8IxUL4ttBxj72okHzACgrH0nkQFlxNEzxrFqZiL9i5Jssvqr2DP76S2eY1HV3pQppA3j2vle8lJVTtDs0dh/UTY9tOVHz/SugJW02Rc+EE7FieZZk2HeQ/ZGpc+P4BXAQC2Hoth89EYBjQtg4tL9iyVoaYlL/INgt5TIOao7UxaLgwBvFqlon58en8dwp8OpU/DYOZuOkGbj8N5bMZmdp+6kKu1OKOoKK5f0f7iOZg9EHyCoMe34KKPruQO76p3c77hk3Qyl/Hzt29p0g0ncsMsATiyCv55Bap2YWx8OyKiL/L2fTXxdLPYpU7JR9q+CqWbwm+PQeRuwHa1pXcDXW1xdDfNk6Xvw8HF0OkDKFEH0zSZu/E4Q79fj5+nKz3rB2dbDTryyatKN4H2b8Ge32HlWLuUUCrQm7fuq8mKMa156K5yLNp9ho6fLWfolPVsPHLeLjU5uhkzzOtXtLdaYd5wuHDSdhbDp5C9y5R8plDHl4gq1pKh8d8wbtpscuNeSMmaG2YJQNxp+GkQBJZlX5P3+GbZYXrWC6ZFxcJ2rVfyCYsb9PoO3H1hVn9Isp3IHBVagXTTZMrKCPvWJzd00zzZ95etaan9ANQbyN7Tcdw/YQ1P/bSVUoHezBreFD9Pt2yrQzfi52WmCXOGwK758ODPV+6BsJfYi6l8vzqCySsPE3MxlSblAnmkdQVaVCh8ZWai/OzySsIBPa9d0T5y+of4rH8X7v0EGg7Nln3pxlnJsIvniP2sOfFJyfzbYjYD73a6vz75xs2yJGKvlaA/usCpraQO+Zcec85zIiaRf59sRWAWJk5RnkiGRayE77tA5Y5w/1QwDEb8uIH1EedZ/XwbPFx11c9R3CxPjm48SqE5raBAadIG/cXH4ceYsOwQfp6uPNehCn0alMrUsDDdiJ9fGYZtFrFCFW3NS8yxm1/eywUB3m481rYiK59tw0v3VuXw2QQenLSObl+t5M8dp7Fa8/fZ28srCV+9AnXXGr/jvf4921mMBkPsW6Dkb96B+A+cThGXC1RY/hiLdpywa57Izd0oS7wKJJH+58twdDV0/YJPt1rYfiKWd7rXzFLDIpIpIc2h/Zu20SArPgXg3iplOJeQwpzVZ+xcnFztRnkSWOgcXr8+CEDcfd8xZPpOxoUfpEfdkix+KpS+jUpn230sV1PTktd5+ELfaZCeyrlx/ahSOf76y3u5zMfDlWF3lWPZmNa826MmsYmpjJy6kfZjlzFv03FS0612qcverl5JGKDkuTjG3TWatMI14d6PbU2oiB0ZJepi3vspzV12EjFrDOUbxdg9T+R6/z9LUiL9uK/4LxSLGAdNRrHWpzXjlh7k/gal6FCjmH2LlfyrySio0RMWvcGSif8S1q4g1nhPnplwRHniQK7PE18+avgcXhd2EHn3V3SffpJVB87yfs+afNi7do6eBNHwsHwidt1CAv7oy+z4NoxxHUpKpL9tuMABi92nt0xLt7Jg+ynGhR9kz+k4ggt6MaJVeXrXD853N4bOmGEybKSVoMJn+aNTF8oXicLj0aVQoHS27kfDOSQromc8QaG9k3k0aTS/0ey/4UcOkCdiczlLvAokUdl9D0v7d8K1TENie/9Epy9W42YxWPDYXfhkw2LAyhPJtJQE0r5pT9yJY3Q69yGnAn0BODejBQc3BChPHMTVeTK07GTeD32JY3WfpuvWxpjAuH71aVo+e+631fAwYZ/Rkfc3P0Uf38UMsfx5ZbhARIS9KwNXiwvd6pRk4eN3MWlgA4L8PHh5/g5avL+E8UsPEpeUau8Sc01YmEHEfoPNTz9J1cL78eg3JdsbFpGsOlzuA1afbcAHHhOobhx2qDwRm7Awg4gDFv6ZlsCyEQNw9SsMvafwym97OH0hiU/vr5MtDYtIlrj7sKuW7Z6W74q9iQ+22cN860YoTxzI5TxZOX4d77V+lePF2tFmXT0Kervz86jm2daw3I6alnwiJATeDH+ePy425UXXqTS5cIDEGE9CQuxd2X8Mw6Bt1aLMe7gZMx5qQtXifry3cA/N31vMJ3/v5VxCir1LzBVBO9+j4JnfMdq/DeVaXfOY7iEQR1CmnBthc6ZxzurPBPdPCEyOJzHWw6HyRCCoYAp19z6IJSka+k7jl/3J/LLlJI+3rUjd0gWVJ+IQilcry4O/T6aicYJP3MZhJrngGnKG4sGaXt2RBLkcpNLWQUR5hnBPRBiNyhbi51HNKVvYB8id4xM1LflEUBB8O96VQZNnsTemMuP8P2bm54evXHp1pH+8DMOgaflC/Di0Mb+Obk6z8oX5fPEBmr+3mDd/38Xp2CR7l5hzdv1imz6wTj9o8vA1D910ykGRXBYUBO9+WpTeM+YQaI1ngt9H9HpxO0FBjpUl+Zppwh9PwbE1cN9XHPeqxEs/76B+mYKMCi2vPBGHERQE/V5ow/NL3+QeywZGn1yGi2cqG06fVp44iqQLWKf3JSElnZ4xo7mvcWWmDG5EgLdtOuPcyhPd05LPREXBqV2HqbGqNS6+hWHYP8yYX8A2VrFgEonnPZk43oWwMMe66Xv/mTjGLT3IL1tO4mJAr/rBjGhZnpBLHX6ecHIzTO4IxWrAoAXg6nHloZtOYZrJewg0Bl2yQ1QUXFg1l/Kbh/BTWkvCg8Yy5ZUQh86SfGP11/DX89DiSdLbvELYhDXsOnWBhY/fhWe6t/IE5YmjiYo04dfRBB2fyuseTxGedjfrPmqkPLE3azrmjL5Y9//LgJTnaXdvbwY1C7myVEVuHp/oSks+ExQEtVqVxSVsKpyPIGXaIEaOSiag50p8+4YT0HMlw0ZaHe6sRsWifnzSpw7hT4cS1qg0czedoM3H4Tw6YzO7T12wd3lZd+EkzAgDn8LQd/o1DQvcfApTjfkVewoKgvLdemJtOYberssoGfkhBR4Id+gsyRf2/Q1/vwhVOkOblxm/9CDrIs7xetfqlAr0Vp6IQwoqYhA06BMo3YwXU74g4OI6AvosU57Y298vYez/m1dSB9GhSx8GNy97zdp6uZknalryq5AW0PlT3I+HM/aeZ3AvYjvwd/R/vEoFevNGtxqseLY1w1uWZ8meSDp+tpyhU9az8ch5e5eXOSkJtoYlOQ7CZoJvkeuecqMpTB3tniTJv1xCnyeq2H085zqTewsuBxw/S/KsM7ts63IVrQE9JrDt5AU+/WcfnWsVp0e9koDyRByYqwfcP5UUj2JMcPuYkEJHAeWJ3Wz4DtZ8zXQ6sje4N/0al7nuKbmZJ2pa8rN6D3Kx7mMMrvIDgxIXA87zj1cRP0+e61iFlc+24am7K7Hp6Hl6jltF3wmrWbYvitwY9pgtrOkw9yE4vQ16TrQNDbuBoCCYON6F2LnNiZ8RSuzc5kwc76LpIMUxuLhAt/FsjKzDZ25fUcs4SEq0t1NkSZ5y4RRM6w3uPhA2k4t48L+ZWyji58Hb99W8cnZUeSIOzacQSd1/wtNMY7L7h/hx0WmOTfKUA//CgqfY5dOYN1L78W6PmjdcMDI380RNSz7n3eV1jvp25ZUCkwldleB0/3gFeLvxaNuKrHyuDa90rkbE2YsMmLyOrl+u5M8dp7BaHbx5+etF2LsAOrwHlTve8qmXpxxcOMeHiAMWje0VhxJU3IvjzWZxJr4oE90+JsTnJG+PTXCaLHF6yfEw435IPA/9ZmP6l+DFn3dwODqBj/vUuXLD7GXKE3FkgZWr8GuRcZQ3TvLlxW9J+LmRUx2bOL3T22H2IOIDKtE7ejjDQytTsajfTZ+eW3mipiW/c3Gh9OMTSC/WgMntHuX48g1O+Y+Xt7srQ1qUZemYUN7vWZO4pFRGTt1E+7HLmLvxOKnpVnuXeL3VX8Na2wrVNB5xRy8JCoKGDVFwi0Pq/mBR/IfPpYC7ydQC7/DLsX/ZezrO3mXlfelptiFhp7dD7++geG0mrTjMz5tP8NTdlW66hoLyRBxZv9Fd+ch9FK0C13J63BOE9bV3RflE7HGY1gerhy8PJj1J0aDCjAotf9uX5UaeqGkRcPPCbcBMXAqUoODC+yFqn70ryjQPVwv3NyzNoqdC+SKsLm4WF576aSuhH4bz4+oIklIdZN73HXNtM/tU6Qzt37J3NSLZJrByZdz7TaOsSyRjzfcZNGEpe07ngckyHJVpwu+Pw/6/oNNHUOkelu+P4p0/dtOxRjEeaV3B3hWKZIphGITcPZyxaT3w2jMdFuvfyhyXeB6m9oSUeCaX+YDNMT68070mnm4We1cGqGmRy3wKQ/954OIKU3vYZrNyYhYXgy61S/DHYy34blBDigV48vIvO2nx/hLGhR8kLik1S9vP0tzxh5bCzyOhdDPoOQlcHCMMRLJN2bswek6glrmXd83P6D9hVd6Y5S8HZHkdiiXvwOap0HIMNBzKkegERk/fTKWifnzUu/Y1s/yIOJv7G5RiXenhzDHbwPKPYN239i7JoWUpT1ITYcYDcO4Qh9tN4J2NFvo2LEWTcrmz2v2dUNMi/wksC/3m2DrtH3vAxXP2rijLDMOgdZUizBnZlFnDm1C1uB/v/7mH5u8t5uO/9xIdn5zhbWZpEaUTm2BmPwgsD2HTwc0zw/sXcQrVu2N0fJ9Qcx2vMIEHJqxm10k1LlfL8oJsa8bDsg+g7oPQ+gUSktMY/sNGACY82AAfD9ccqFok97i4GLzfqzZvmMPY6NkE849nbCMV5DpZypP0VPhpMBxdzcXOXzN8uTeBPh4837FqzhWcCWpa5Fol6tjWCTl30DYLTXK8vSvKFoZh0LhcIX4c2pjfRregeYXCfLnkAM3fX8zrv+3kZEziHW0nKgqGjbRmbl2bqH0wrRd4FYQH59n+K5KXNR4BLcfQ1bqIp12m0e/b1ew8GWvvqhxClrIEYOtM+PNZ2xDTzmMxgad/2sr+yDi+eqAepQt552T5IrmmVKA3z91bk34xI4gsWBfmDYf9/9q7LIeSpTyxWuGX0bBvIekdP2TEptIcPpvA532vn8DD3tS0yPXKtYJe39lWaJ/5AKQm2buibFUzOIBx/evzzxOt6FyrBD+uPkKrD5fw7JxtHD6bcMvXZnoRpfNH4Mf7wLDAgPngXyIb3omIE2j9AjR8iH7pvzDSZT79Jq5lxwk1LllakG337zB/FJRtaRtianHlg7/2snDHaV7oVJUWFQvnZOkiuS6sUSkaVixJl+hHSSlUBWb1h4iV9i7LYWQ6T0zTdvJj20xo8xKvn27K8v1neeu+GjSr4Hg5oqZFbqxqZ7jvazi8DGY/CGkp9q4o21Uo4stHvWsT/kwoDzQqzfwtJ2j7cTiPTN9007PBmVpEKfY4fN/Ztojkgz9Dof9m4cjyeHYRR2cY0PEDqHU/I9KnM9RlgRoXsrAg276/4adBUKKu7aq4mydfLTnAuPCD3F22NF0rl83hykVyn2EYvN+zFomGD6NdXsIsUAqm94Fj6+1dmkPIVJ6YJvz9EqybAE1HM8WlJz+sPsLwluVoW7a0Qx6bqGmRm6vdFzp/Cvv/hjmDbWMe86Dggt683q0GK55tw4hW5Vm6N4p7P1/B4O/WsSHi2vt6MryI0oVT8H0XSIyxNSxXLR6Z5fHsIs7CxQW6fQ3V7uPRtCkMtPxNv4lr2X48/zYumVqQ7eAS2xnmotWg/1zw8GPyisN8+NdekvaW4Od3y1K2olVZInlSiQJevNy5Gn8fsfJTta/AJ8g209WJTfYuze4ynCemaZuNbfWX0Gg4S8o8xhsLdnN3taKUianssMcmRm6sHN6gQQNzw4YNOb4fySFrv4GFY2xjp3t9B67u9q4oR8UmpvLj6ggmr4zgXEIKjcoG8kjrCrSsWPjKTDxRUbbLsSEht5iT/MJJmNIZ4s/YGpZSja48FBUFIRXSCei5EvcicaRE+hE7tzkRByy5smaCYRgbTdNskPN7yl7KEieXngqzB8LeBXzqNozvUtszdVhjagUXsHdldnNHWQJwYJFtuG5geRj0O3gHMmv9UZ6du53kQ0UxfBNwLxKf61kCyhPJPaZpMnjKetYeOsffQ8pSan4vSI61/Rtbsr69y7O7O8qTyw3L8o+g3kD2NnyTnuPXUDrQm3G9mlKtmmG3YxO4dZ7oSovcXuMRtuEde36H2QMgLeMzbjmTAC83RrepyIpnW/Nql2ocO3eRgZPX0eXLFfyx/RTpVvP2iyjFHofvOkF8pO2M6FUNC2RxPLuIs7K4Qe8pUKUzT6ROZJjbQvpPXMvWYzH2rsxu7mhBtv3/wIwwKFQRBv4G3oH8suUEz83bTp2iQaRuqIx7EdukKcoSycsMw+DdHjVxtRg8+Vc01oG/g2cB+OE+DRXjDvLENOHf1640LFGh7zPk+414u1uYNKgBkSddHfrYRE2L3JnGI2wLl+1baPvHM+XWN6znBd7urgxuXpalz7Tmg561SEhOZ9S0Tdz96VJ+2nCM1HTrjV8YfRC+64g1IZpdDX4myqvJdU/J9Hh2EWfn6m5rXKp25bHU7xjtNp/+k9awJR83Lre061eY+QCpgVXYVOtXoi4W4u+dp3ly9lYahQTyxf31SYz2VpZIvlE8wItXu1RnfcR5vttlhcF/gHch22Q3h5fbuzzHZbXCwmdh5VgiSw/lWP2PGT51E9EJyUwa2JDiAV4Of2yipkXuXKOHoOuXcGiJbR2XxJg7fqkz33Du7upCn4al+PfJVnz5QF08XC08M2cboR+GM2XlYRJT0v978pmd8F1HkuPiaTnlZ1oOr3bDMaGZGs8ukldY3GxDTWvdz/DU6TzvOoMHJ65h89Hzd/RyZ86TDNkyHX4aSJRbHUJe/5n2/TyoGHqGh6dupmbJACYNakipEhZlieQ7PeuVpG2VInzw5x4OpRSAwQshINh2j8vePzO0rXyRJ+lp8MsjsO4bvtj4MNXeeJ2Gj21j89EYxt5fh5rBAYDjH5vonhbJuJ0/w9yHIKgK9PsJ/Ivf8ukzZpgMG2nFq2ASiec9mTjehbAw512l2TRNwvdG8dWSA2w4cp5CPu4MaVGWQcGn8Jnbn3SLF42+nsup1tG3HRN6x+PZs5nGoItDsFph4TOwfiILXNvyYupQPu/XiJaVbv5hyGt5ckOmabtB9u+XSAkOJfi5H3DtuhX3InGYJqRH+7H01aaUL/XfGgr2yhJQnoh9nLmQRPtPl1E+yIefRjbDkngOpvaAMztsJ1jrhN12G/kiT1IuwtxhsHcBb6x6nvHBTa8MJ41fU4ndP1V0mGMT0D0tkt2qd4cHZsK5QzDpbojcc9OnZnkBNQdkGAatqxRhzsPNmD2iKTVKBrD9nx9xndqdc/ixuvavHDFD7mhM6B2NZxfJq1xcbMNOW47h3rRFTHD7mEemLGfqmiM3fHpezJPrWNPhz+dtU5FW68a2GrOw+rpeyRPDgORlNTl3+tpF35Qlkt8U9ffk9a7V2XQ0hkkrDoFPIds9X2Waw/yRsOwj2wmAm8gXeZIQDT90hb1/cKT6h3x+6OErDQsAh4s71bGJmhbJnArtbONI05Jhcns4tPSGT8vrN5w3CinI91XWMc79M054VaRd7IsMXXYQl2qHSTnrAzjemFARh2IY0OZF6DyWhumb+c33Hb6Yv4wXft5OStq1943l9TwhJQF+Gghrx0GTUdBrCmXKe5LmnYB56X9FylkfLp7xV56IAN3qlKB9taJ89Pc+DkTGgac/9JsDNfvA4jfht8dvulxDns+TswdsJ5ZPb4c+P7CuZFe8Oqy/0selRPqSGOPlVFmipkUyr0QdGPYP+BW3XZLd8N2Vhy6PEfX1vbObupxyTGlaCvz2GPz9IkbVzpR7chGzn+hCl9ol8K11FLeCF0k9VIz4RbUdakyoiENqMBgjbCZlOMW/fq+xfV04D3y7hsi4pPyRJ7EnYHIH2LMA7nkXOryLaRj8sT+CwK4bSI/xJeHXxsT+dJfyROQSwzB4u3tNfNwtPPXTNtLSrbbJPrp/A3c9BZu+hx+7w0XbmmtXZ8Od3nTulHlyKBwmtoGkGM72mM1DG0rwzK8bKVnYnZh5TS7dr9LC6bJE97RI1iVdgDlD4MA/0HAYs2LfYcjDrlfGiA4ZYDD5BxOvAkkkxlw/ZtQpx5TGnbGtSn10FbR8BkJfsA11ueRETCKf/XmQn7cdI81qpVPN4jwcWp4aJQPsV/NVNAZdHNaZnTCjL+lxkTyXOpS/zLYcmVkHt1TPvJsnR1bbrrCkXIRek6FSe1LSrLz6605mrDtKu6pFeLFtHc6ecrPLGPPbUZ6Ivf229SSPztjMmA6VGRVa4b8Hts6CX0eDfwkW+k+l1/+qXZMNYBsilmfyxDRh7Xj460XMwpWYXekjXl8ej9U0eaJdJYa0KEvMORe73a9yJ26VJ2paJEuu3KxVOp2gra/Bqs9Ze7IBo8xHiC7kceUm9A1rLMTHX/8hsfcii5lyZDX8NAgz6QKHan6Bf7NeN601Ki6Z71Ye5sfVR4hLTiO0chCPtK5Aw5DA3K35/9FBhjiaa2789IqCOYMhYjlTLnbkLSOMNMOVlEhfYue2yDt5cvkA4++XSPMrw+7a0yhWsyouXsk8PHUT6yLO8Ujr8jx1d2VcXBz3QEl5IvZmmiaPTN/Ev7si+f2xFhR08fsvT5LWkT5zAMkxMTwbO5LffBteyYZl65OYv+0oZ89bKVbYQuECrvi4W/D2cCU9yZVRDxt4N92NW+BFx8+T5Hjb6I8dczlb4h4GRA9lV6ztuOPNbjUoFeht7wrvyK3yxDW3i5G84/ozEG/QsF59qqeO4h/XR3kmdQSLitTHq0AS8fE+NGx4/TZuPqbUx/FCwZoOKz6FJe8Q51qGtpPncMhajsTz6Tc9+xLk58GYDlUYGVqeH1cfYfKKw/Qev5pGIYGMal2eVpWCMAzHPRgRyQ3XZ0lhwh6cz6lprzHo0BfUte7l8dRHiChSHO8SMcTHF3L+PLl4Dn59FPb8znHfTjR852tSfdxJcTlPuX6bSTST+axvHbrVKWnvSkUcnmEYvNmtBmsPLWPw+C1sGdsEr4CUS3nSkCpNlpIwdSBfFP+UFmmhvFZkAEanjfSaHIWrxcDTzcLFw+mkW689kR9w739/dug8ObUV5gzFPHeQn70e5YkDTTGTXYhfXpv2tUpQKjBv3A2ipkUy5epZN9yLxOEe6cewkc15/eVuTJpUmZ+GPMAk94+ZHNeZMfEtb3qj1+Uxpe6RflfOjDrkTeuxx+HnkRCxnKQKPaj0xCcYXbbjWyT8yntv1+7mZ1/8Pd14pHUFhjQvy6z1R5mw7BCDvltP9RL+jAqtQIcaxbA48JlUkZxysyxp186VXxPf5K85DZnUfRS/u7/IqykD+amDhX1JtWhIieu25TR5cniZLU/iI4lv9hZVeo/Ev+dqfC9NaRx93oOpI5rSqmYBe1cq4jQK+Xowpk0Nnv1tEwX7hePqm3LVsUkRnv9uPh8/+gijfefSwGUvj5V8jPoN2/Bo+7IE+XlgmibJaVYupqRzMSWN46fTuadrCv4dNmLxSSU9wd3x8sRqtU3c8e9rmF6FeL/QB4w/XgIzxRUXrxQ86xzioYeLc/fdjjkULKPyRuslue5GZzQ9/JJ4+TUrp6uYNPx8GZ9vHcIQv9858L9WBCWuueF2HH0hI0wTNn4PXzWBExuh21dsrzSZVB/3TM044uVuYVDzsoQ/05oPe9UiMTWdR6Zv4u5PljJ7w7HrZksSyetudnVk82Z4coyVJYUrUuebFWw6WZuPPb5huv/HfLZgMaOmbeT4+YvXbMvh8yQ5HhY8Dd93AVdPGPYPuws+imdg0jVTGif9Uw/vpAL2rVXECZVzL07qkSK4+qYAtjzxrnSCDxbtoNiopXzi1oOwhJfxSknjN9+XeKPgHIK8bCcMDcN2xSXQx53ggt40qerH+DcCOfdDG1IPFMfik8LQt445Tp5EH4Qp98JfL5BatjXDfMbyzfESJG8qj4tnGpD3ZkRT0yKZcqNZNy7GeOJdKImAJocpPGQN7/r3pNuC2Xi5p9pmxVnwNCTGXLetsDCDiAMWFs7xIeKAxXFucjt7AH68zzZGtHhteHgV1O1PSFnjjmYcuRV3Vxd6NyjFP0+04ut+9fBytzBmzjZCP1zCdysPk5iSniNvScTR3GwGH7A1MwFNDmPte4CwlBd5cc1rNLNsY5nvcwTtnUa7j5fw0V97SUhOu7I9h82TfX/BuGawfiI0fhhGroASdUnzuYBHmy1XnpYS6cvF0wGOdTZXxEmEhEDsohqY6f997n3abser+nEsnmlYUywsO9uI5rOXcq5kH1j2IXzTEo6suuH2bHniytxn6nBXuSL8dnInf+44lUvv5ibSUmxD1ce3gDM7ievwOV3PjmbZcStv3VuXC2srZun4xJHpRnzJtCvj0C/NuvHJBy48OcZ63U2wR3ZdpPDWN2H9t+BdGNq9BrXDrplty6Ekx8GKsbDqc9vZ0LavQIOh19T7/997VmcUMU2Tpfui+HrJQdZFnKOQjztDWpSlf5MyBHi53X4DGaQbZ8WR3Ojz1K6dccOb6o9tOEzgyv9BxHKOeFbhf7FhnPCtwZgOVehRt6Tj3bB+7hD8/TLs+R0KV4LOYyGkOYkp6Xy2aD8Tlx/Cw3Dj5IJqWM4XyJY8yW3KE3EkM2aYjHjuAj6VT5Ga4sLQbgX48s0CBNy35voJOs79CX88DbHHbMclbV8F/+I33G5iSjr9Jq5hx8kL/DikEY3LFcrldwYcXAILn4Wze6HyvVy8+z16Tz/Cwah4JjzYgJaVgrL9+CS3afYwyTHXzPgTdJuD+ZNbYMFTcGIDFK0Jd78O5dvYxkM4gvRU25zu4e9BQpRtcar2b4Ff0Rs+/f+/9+yy7vA5vg4/QPjeKPw8XHmwaRmGtChLYV+PbNuHDjLE0dzo83TTPDFN2D4H/n4R4s+w0qMFL17ojl+JKrzSpZrdZ+cDbCtRL/vQdmXF4gatxkCTR4hLM1iw7RRfhR/g2LlEetcP5oVOVUm76O7Q05DeivJEHE2Gjk1SEmyf1VVfgosrNBsNzR6zLVT5/5xPSKHX+FVExiUzZ2QzKhfzy503dHo7/PMqHFwEBcpAxw+wVryHEVM3smj3GSYNakjrykVu+v6diZoWyVW3/LBYrbBzHix6HWKOQnAj2zonFe+2X/OSmgRbptqursQeg9LNoP2bEGzff4N3nIhl3NKD/LH9FO4WF8IaleahluUoWcAry9vWQYY4i1vmSXIcrPoSc9XnkJrEny538VFiZ6rUbMBTd1eiXJBv7hccd8Z2lXbDZEhLgroPYm31PKuj3Jiz8TgLd5wiKdVKlWJ+vNKlGs3KF879GrOZ8kScwW0P5M8dgkVv2o5RPANswzgbjwDva0+CHD9/kR5fr8LFMJg3qhklsuHf5Js6uRmWfWS7UutZwHa81OghcPXg3YW7+WbpIV7vWp2BzUJyroZcpqZFHE9aMmz+8b9GoXBl2wexdl/wyKUzF7HHYcN3sHEKXDwLwQ2h5Rj7NlA3cDAqnm+WHmTephMA3Fe3JCNbladCkcwfkOkgQ/KUS42CuWEyRupFlpm1+S61PQnBrejZsAz31iqBr0cOTpZpmraJOtZNgJ0/gzUNavbmRM1RzDzsxbxNJzgRk4ifpytda5egV/1g6pQqkGemO1eeSJ5ydaPg5g21+kCj4VC0+pWn7D51gT7jV1M0wJM5I5tSwNs9+/afnmrb99oJtgWsLzdQTUaCV0EAZq8/xpi52xjQtAxvdKuRfft2AGpaJEdl6TJkWgrsmGNbYO3UVltAVLnXNjSrbEtw88zeYi+eg70LYdss27SjAJU7QuORtv058EHEiZhEvl12iJnrj5KcZqVjjWKMCq1AjZIBGd6WDjLEUWUpTxLOwvpJpK+fhCXhDFFGIeamNuUfozkhNZrQq0EZGpcNzL77XqIPwq75tlW3z+4Fdz9SavblX7/uTNnjwrqIc7gYcFfFIHrVD+buakXxdLNkz74diPJEHFWW8uT0Dtt0wtvn2K6aFq8NtfpC1S5QoBSrDp5l0OT11AoOYOqwxln7bFvT4fh62DHX9nUxGgqUhoYPQf1B1wxVW30wmgcnraVp+UJ8N6ghrhYHvT84k9S0SI65flG4TN7wZZpwfANsmWY7U5kUY2tgyraCsndBqSZQrAa4ZvC+jqQLtntojq6FQ+FwfB2YVigYArXuhzoP2P58G440PvRsfDLfrTzMD6uPEJeURstKQTwSWj5DNwXqIEMcUbblSVoK7F2AuXUm7P8Xw0wj0ixIeHotDnrXpETN1rRt3oTgwAxcrTRNiDsFx9bC0TVw4F+IPmB7qFRTDhXvxMSY+szfHUdiajrlgnzoVT+YHnWDKRZw7ckXR8qT7KA8EUeUbXly8RxsnWk72Xlqi+1nRapDhTast1Zi1FJX6lerzLj+9e786qk1Hc7us2XJkZVwYBEknrNN/lO5o21SgArtwOXaRujw2QS6f72Swr4ezH24GQFebvkqT9S0SKZFRXHD2X0iDtx8kcU7kpZsuwqy7y/Y/zfEHAHANCxc9CxHgmdF/EoUwyuoKLj72D7kptV2JiTpAsSfhtgTcHY/XDh+aaMGFK8FFe+BSh2gZL07vqpyOfg8/JO4eN6Tzz52YcQI+1+RuZCUytQ1R5i84jBn41NoUKYgj7SuQGjloNsGpw4yxNHkWJ4kRMP+v0nbsxDrwSW4p14A4KLpwVGXUkS7luSiZyFSPAMxXb0wXN3xcLHiRSq+1lh806LxTzpFQPzBK69Nt3iSUKwhF0q348+UOny303pl+FeXS8O/6t5k+Jej5klWKE/E0eRYnpzdD/v+tB2fHFsL6bb1YM6a/iT5VcQzoCLeRYvhHVgAXL1sTUdaMqQm2ib4iTsF5w7bGpb0ZNs2fYKgfFuo1N7WqHjeePRE7MVUun+9kvMXU/jlkRaULuTNjBkmQ4dbcfdNISXenUkTnGumsBu5VZ7k4CBfyetutihcRIRP1kLB1cN2X0nFu23fXzjFiplrWDZnO1UK7aFSoYO4nFiDl/e5619ruIBvUfArBiHNbVOMlqhru1/lBjOB3M7l1brd6+4ldm0FLP7xPDzaFwwXRgy3bzD4e7oxKrQCQ5qXZfaGY3yz9BCDp6ynanF/RoWWp1PN4lgcbfpXkZvIsTzxKQR1wnCtEwZWKwt+2MPCH5dTo9wmqgTso5THAYolr8UnLum6lyabbpwxC3CIQuy3NmS/GcwWa3l2miGkHXSFg2AYCbSoUJgxHSpzT/Vitxwi4sh5IpKX5FieFK5o+2r2KKQm8c8Pm/n3xw3UqreCyumHKBq5FP+jZ8GSev1rPfxtxyYFykD5UChSDUo1hsBytz2JmppuZdT0jRw/n8i0hxpTupA3UVEwaKiVVGs66WYKqakWBg2Fdu2y2Jg5MDUtkmmXF4Vzj/S7ciYjJxYxikouTvunupKUei/FHliDu69tXwnzGnFwZwqFA5LAsNiaHTev6y6nZkVEBHj4JxG7tgJFw/6b4/1/TzanR3fHCAZPNwsDmoYQ1qg0v2w5ydfhB3h0xmY++WcfI1uVo3vdYNxd89aYV8l7ciNPoqJd6PN4ZXzvjWXc1MH/faZP+ZE0vz77tqVSOCAZLG6kWzxIM93xSbMSnGYlKDWdeqlWuqWlk5SaTnKqleQ0K7WCA+549iBnyBORvCBX8iTGk/uebITvvXB+6gCKD1yJYTGxXnAhcU4jdq8xKFwg9b9jk4wOb7/ENE1e+WUnKw9E83Hv2lemdN+8GVLTTIr1/y9LTk9txubN0L599r1PR6KmRTItKAgmjndh2Mjm18x9nt3/8EZEgLtvCulmyjVnTS76pnP4pD+FS2b8CsqdCgmBi+c9sfjHX3vGJjAbzthkMzeLC73qB9O9bkn+3nmar8IP8Ozc7Yz9dz8f965NswrOP7Wq5F25kSeXz766uKXjGpB4zWc6xQ8OnypA4WDbcy2Az6Wv7OJMeSLizHI7TwwXMCy22y1cvKx499nAxNVVGdOrVJYn/pi8MoIZ644yKrQ8PesHX/OYxTf5miyx+CSTlw/tdfpVsiQszCDigIWFc3yIOGDJkbGUISGQEu9OaowXKZG26ZBTIv1IvZDxsyZRUbB+ve2/dyIoCD772IW0c77X7Ds5NvuvKGUXi4tBx5rF+W10C34Y0oiQQj4EF/S2d1kit5XTeXL57Ks11UJa7LV5kpmzsPkhT0Sclb3yJDXam9RoP8Zv2s79E1az70zcbbd1syxZtPsMby3YRYfqxXi6feVrHqtbFyzJntdkiSXFk7p1s+XtOSbTNHP8q379+qZIVkyfbjXdvdJMwy3VdC0Qb3p4p5nTp1szvA1v/zSzUJl409s/Y68f/43V9PRJMwuWyvhrHRGwwcyFz352fylLJKsu54BvoUTTcE0z/Ypn7jOtPPmP8kTyq5vlybRp6eas9UfNOq//ZZZ/foH53sLd5sXktFtu4/9nya6TsWa1lxeanT9fbiYkp978tX5pZkDJeNPbz/mzxDRvnSeaPUycRlSUbQwn2M4wZOQyb3bMJJKXphXUbD+Sn13+LPv6Qnx8xj/TypNrKU8kP7tVnpxLSOHdP3bz08bjBBf04o1u1WlTpeg1r71RlmzYmsqwmatIt5r8Mro5Rf1vvmZdXsoS0OxhkkcEBWX+5rLsmEkkKChvBIJIfpfVz7LyREQuu9VnOdDHnQ9716ZX/WBenL+DIVM20LFGMV7pUo3iAV43zpJCCTw2ezvnElL4aWTTWzYst9t/XqN7WiRfuDz2NKtj2EVElCcikhGNyxXij8fu4pl7KrN4TyTtPl7K5BWHCS5t/X9Z4otr3f3sPxfDp/fXoUbJG6/Zkl/pSovkC7k105mI5H3KExHJKHdXFx5pXYEutUrw8i87eOP3XcwtcZxXP6nB60/assSocAyfhgd5tkMVOtQoZu+SHY6aFsk3wsIM2rWzEBHhk2fGfoqIfShPRCQzShfyZsrghvyx/TSv/7aTcadWMeqbMhQ0/Jmw+SC96gczslU5e5fpkNS0SL6Sn8Z+ikjOUp6ISGYYhsG9tYrTslJhPv57Hz+sjsBqQqOygbzTvSaGkf3LR+QFalpERERERHKZn6cbr3WtTo96JZm/+SSj21TA3VW3m9+MmhYRERERETupFVyAWsEF7F2Gw1M7JyIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yIiIiIiIg5NTYuIiIiIiDg0NS0iIiIiIuLQ1LSIiIiIiIhDU9MiIiIiIiIOTU2LiIiIiIg4NDUtIiIiIiLi0NS0iIiIiIiIQ1PTIiIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yIiIiIiIg5NTYuIiIiIiDg0NS0iIiIiIuLQ1LSIiIiIiIhDU9MiIiIiIiIOTU2LiIiIiIg4NDUtIiIiIiLi0NS0iIiIiIiIQ1PTIiIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yIiIiIiIg5NTYuIiIiIiDg0NS0iIiIiIuLQ1LSIiIiIiIhDU9MiIiIiIiIOTU2LiIiIiIg4NDUtIiIiIiLi0NS0iIiIiIiIQ1PTIiIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yIiIiIiIg5NTYuIiIiIiDg0NS0iIiIiIuLQ1LSIiIiIiIhDU9MiIiIiIiIOTU2LiIiIiIg4NDUtIiIiIiLi0NS0iIiIiIiIQ1PTIiIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yIiIiIiIg5NTYuIiIiIiDg0NS0iIiIiIuLQDNM0c34nhhEFHMnxHYnInSpjmmaQvYvIKGWJiENSnohIdrlpnuRK0yIiIiIiIpJZGh4mIiIiIiIOTU2LiIiIiIg4NDUtIiIiIiLi0NS0iIiIiIiIQ1PTIiIiIiIiDk1Ni4iIiIiIODQ1LSIiIiIi4tDUtIiIiIiIiENT0yK3ZRhGQ8MwthmG4WkYho9hGDsNw6hh77pExPkoT0QkOyhL8h/DNE171yBOwDCMtwBPwAs4bprmu3YuSUSclPJERLKDsiR/UdMid8QwDHdgPZAENDNNM93OJYmIk1KeiEh2UJbkLxoeJneqEOAL+GE7qyEiklnKExHJDsqSfERXWuSOGIbxKzATKAsUN01ztJ1LEhEnpTwRkeygLMlfXO1dgDg+wzAGAKmmaU43DMMCrDIMo41pmovtXZuIOBfliYhkB2VJ/qMrLSIiIiIi4tB0T4uIiIiIiDg0NS0iIiIiIuLQ1LSIiIiIiIhDU9MiIiIiIiIOTU2LiIiIiIg4NDUtIiIiIiLi0NS0iIiIiIiIQ1PTIiIiIiIiDu3/AMfe9awCteW2AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 1008x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "def true_fun(X):\n",
    "    return np.cos(1.5 * np.pi * X)\n",
    "\n",
    "np.random.seed(0)\n",
    "n_samples = 30\n",
    "degrees = [1, 4, 15]\n",
    "\n",
    "X = np.sort(np.random.rand(n_samples))\n",
    "y = true_fun(X) + np.random.randn(n_samples) * 0.1\n",
    "\n",
    "plt.figure(figsize=(14, 5))\n",
    "\n",
    "for i in range(len(degrees)):\n",
    "    ax = plt.subplot(1, len(degrees), i + 1)\n",
    "    plt.setp(ax, xticks=(), yticks=())\n",
    "\n",
    "    \n",
    "    linear_regression = LinearRegression()\n",
    "    polynomial_features = PolynomialFeatures(degree=degrees[i])\n",
    "    polynomial = polynomial_features.fit_transform(X[:, np.newaxis])\n",
    "    \n",
    "    linear_regression.fit(polynomial, y)\n",
    "\n",
    "    X_test = np.linspace(0, 1, 100)\n",
    "    plt.plot(X, linear_regression.predict(polynomial), label=\"Model\")\n",
    "    plt.plot(X_test, true_fun(X_test), label=\"True function\")\n",
    "    plt.scatter(X, y, edgecolor=\"b\", s=20, label=\"Samples\")\n",
    "    plt.xlabel(\"x\")\n",
    "    plt.ylabel(\"y\")\n",
    "    plt.xlim((0, 1))\n",
    "    plt.ylim((-2, 2))\n",
    "    plt.legend(loc=\"best\")\n",
    "\n",
    "    plt.title( \"Degree: \" + str(degrees[i]))\n",
    "    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regresión no lineal: un ejemplo\n",
    "\n",
    "Cuando queremos resolver un problema usando un modelo más complejo podemos elegir alguno de los siguientes:\n",
    "\n",
    "- Árboles de regresión\n",
    "- Bosques de regresión\n",
    "- SVM para la regresión\n",
    "- Y todos los modelos que existen...\n",
    "\n",
    "Realizaremos un ejercicio con los árboles de decisión debido a que son fáciles de entender y que en sesiones futuras\n",
    "trabajaremos con los análogos de los Bosques de regresión y las SVM para problemas de clasificación.\n",
    "\n",
    "Los **árboles de decisión (DT)** son un método de aprendizaje supervisado no paramétrico que se utiliza para\n",
    "clasificación y regresión. El objetivo es crear un modelo que prediga el valor de una variable objetivo mediante\n",
    "el aprendizaje de reglas de decisión simples inferidas de las características de los datos. Un árbol puede verse como\n",
    "una aproximación de una función a trozos.\n",
    "\n",
    "En sci-kit la clase que modela este tipo de árboles se llama `DecisionTreeRegressor` [enlace](https://scikit-learn.org/stable/auto_examples/tree/plot_tree_regression.html)\n",
    "\n",
    "Veamos un ejemplo sencillo:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Importamos las librerias necesarias\n",
    "import numpy as np\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Generamos un conjunto de datos (función seno)\n",
    "rng = np.random.RandomState(1)\n",
    "X = np.sort(5 * rng.rand(80, 1), axis=0)\n",
    "y = np.sin(X).ravel()\n",
    "y[::5] += 3 * (0.5 - rng.rand(16))\n",
    "\n",
    "# Entrenamos el modelo\n",
    "\n",
    "regr_1 = DecisionTreeRegressor(max_depth=2)\n",
    "regr_1.fit(X, y)\n",
    "\n",
    "# Realizamos una predicción\n",
    "X_test = np.expand_dims(np.arange(0.0, 5.0, 0.01), 1)\n",
    "y_1 = regr_1.predict(X_test)\n",
    "\n",
    "# Mostramos los resultados\n",
    "plt.figure()\n",
    "plt.scatter(X, y, s=20, edgecolor=\"black\", c=\"darkorange\", label=\"data\")\n",
    "plt.plot(X_test, y_1, color=\"cornflowerblue\", label=\"max_depth=2\", linewidth=2)\n",
    "plt.xlabel(\"data\")\n",
    "plt.ylabel(\"target\")\n",
    "plt.title(\"Decision Tree Regression\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Vamos a ver la estructura del primer árbol que hemos entrenado para entender mejor el proceso que ha seguido:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0.5, 0.8333333333333334, 'X[0] <= 3.133\\nsquared_error = 0.547\\nsamples = 80\\nvalue = 0.122'),\n",
       " Text(0.25, 0.5, 'X[0] <= 0.514\\nsquared_error = 0.231\\nsamples = 51\\nvalue = 0.571'),\n",
       " Text(0.125, 0.16666666666666666, 'squared_error = 0.192\\nsamples = 11\\nvalue = 0.052'),\n",
       " Text(0.375, 0.16666666666666666, 'squared_error = 0.148\\nsamples = 40\\nvalue = 0.714'),\n",
       " Text(0.75, 0.5, 'X[0] <= 3.85\\nsquared_error = 0.124\\nsamples = 29\\nvalue = -0.667'),\n",
       " Text(0.625, 0.16666666666666666, 'squared_error = 0.124\\nsamples = 14\\nvalue = -0.452'),\n",
       " Text(0.875, 0.16666666666666666, 'squared_error = 0.041\\nsamples = 15\\nvalue = -0.869')]"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn import tree\n",
    "plt.figure(figsize=(9,9))\n",
    "tree.plot_tree(regr_1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Ejercicio**\n",
    "\n",
    "- Cambia el parámetro _max_depth_ del modelo. Que pasa si es 1? Que pasa si es 5 o mayor?\n",
    "- Visualiza el árbol resultante. \n",
    "- Cuál es la medida de Score de este método (mirar documentación)? Como cambia al permitir árboles más profundos?\n",
    "- Busca en la documentación de Sci-kit de los Bosques de Regresión ( _RandomForestRegressor_ ) compara el score obtenido cuando incrementamos el número de árboles, manteniendo la misma profundidad."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Métricas para problemas de regression\n",
    "\n",
    "Las métricas más comunes en los problemas de regresión son el **error cuadrático** y el **error absoluto**, y sus distintas modificaciones.\n",
    "\n",
    "### 3.1 Errores cuadráticos\n",
    "\n",
    "El **error cuadrático (Squared Error)** de un valor predicho con respecto al valor real, se calcula cómo:\n",
    "\n",
    "$$ SE = \\sum_j\\left[\\hat{y}_j - y_j\\right]^2$$\n",
    "\n",
    "\n",
    "**Error cuadrático medio (Mean Squared Error)** Da una idea del error de nuestras predicciones dando más peso a los errores grandes.\n",
    "\n",
    "$$ MSE = \\frac{1}{m}\\sum_j^m \\left[\\hat{y}_j - y_j\\right]^2 $$\n",
    "\n",
    "**Raiz del error cuadrático medio (Root Mean Square Error)** La raíz cuadrada del MSE produce el error de la raíz cuadrada de la media o la desviación de la raíz cuadrada media (RMSE o RMSD). Tiene las mismas unidades que la cantidad que estima. Para un estimador sin sesgo (bies), el RMSE es la raíz cuadrada de la varianza, es decir la desviación estandar.\n",
    "\n",
    "$$ RMSE = \\sqrt{\\frac{1}{m}\\sum_j^m \\left[\\hat{y}_j - y_j\\right]^2} $$\n",
    "\n",
    "\n",
    "\n",
    "A pesar de ser una de las métricas más utilizadas, tiene el inconveniente de ser sensible a los valores extremos (outliers). Cuando este comportamiento pueda suponer un problema, los **errores absolutos** pueden darnos una mejor medida de rendimiento."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "sns.set_style('darkgrid')\n",
    "sns.set_context('notebook')\n",
    "sns.set(rc={'figure.figsize':(12, 8)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "ename": "SyntaxError",
     "evalue": "invalid syntax (820556737.py, line 7)",
     "output_type": "error",
     "traceback": [
      "\u001b[0;36m  Input \u001b[0;32mIn [34]\u001b[0;36m\u001b[0m\n\u001b[0;31m    sq_error = ##\u001b[0m\n\u001b[0m               ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m invalid syntax\n"
     ]
    }
   ],
   "source": [
    "## Escribe una función que devuelva el MSE y el RMSE dados dos arrays de numpy\n",
    "\n",
    "def MSE(x1, x2):\n",
    "    \"\"\"\n",
    "    Returns the Root Mean Squared Error of the two input vectors\n",
    "    \"\"\"\n",
    "    sq_error = ##\n",
    "    mean_sq_error = np.##\n",
    "    return mean_sq_error\n",
    "\n",
    "def RMSE(x1, x2):\n",
    "    \"\"\"\n",
    "    Returns the Root Mean Squared Error of two vectors. Depends on the MSE function\n",
    "    \"\"\"\n",
    "    mse = MSE(x1, x2)\n",
    "    root_mse = np.##\n",
    "    return root_mse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2 Errores absolutos\n",
    "\n",
    "El **error absoluto** (Absolute Error) se define cómo:\n",
    "\n",
    "$$ AE = \\sum_j \\left|\\hat{y}_j - y_j\\right| $$\n",
    "\n",
    "#### Error absoluto medio (Mean Absolute Error)\n",
    "\n",
    "Es más robusto a los valores extremos y su interpretabilidad es más alta que la del RMSE ya que también está en las unidades de la variable a predecir con la ventaja de que el dato no ha sufrido ninguna transformación.\n",
    "\n",
    "$$ MAE = \\frac{1}{m} \\sum_j^m \\left|\\hat{y}_j - y_j\\right| $$\n",
    "\n",
    "#### Error absoluto medio porcentual (Mean Average Percentage Error)\n",
    "\n",
    "A pesar de su simpleza, presenta varios inconvenientes a la hora de usarlo de forma práctica. Por ejemplo, **no puede usarse cuando el valor de referencia es 0**. Además, **si se usa para elegir métodos predictivos seleccionará de forma sistemática un metodo que prediga valores bajos.**\n",
    "[wiki](https://en.wikipedia.org/wiki/Mean_absolute_percentage_error)\n",
    "\n",
    "$$ MAPE = \\frac{1}{m} \\sum_j^m \\left|\\frac{\\hat{y}_j - y_j}{y_j}\\right| $$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3 Generalización\n",
    "\n",
    "Las dos **métricas** expuestas anteriormente pueden considerarse como **distancias entre el vector de valores reales y el predicho**. De hecho, el RMSE corresponde a la **distancia euclidiana**, también conocida como norma $l_2$ o $\\lVert{v}\\rVert_2$.\n",
    "\n",
    "Por otro lado, el MAE corresponde a la norma $l_1$ o $\\lVert{v}\\rVert_1$. A esta distancia se la conoce como **distancia de manhattan**, porque sólo se puede viajar de un bloque a otro de la ciudad a traves de calles ortogonales.\n",
    "\n",
    "De forma general, una norma $l_k$ o $\\lVert{v}\\rVert_k$ se calcula:\n",
    "\n",
    "$$\\lVert{v}\\rVert_k = \\left(|v_0|^k + ...+ |v_m|^k \\right)^\\frac{1}{k}$$\n",
    "\n",
    "El concepto de distancia será particularmente útil en los problemas de segmentación (aprendizaje no supervisado)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.4 Coeficiente de determinación ($R^2$)\n",
    "\n",
    "El coeficiente determina la calidad del modelo para replicar los resultados, y la proporción de variación de los resultados que puede explicarse por el modelo. El valor más alto obtenible será 1, aunque hay casos en los que puede presentar valores negativos. De forma intuitiva, $R^2$ compara el \"fit\" de nuestro modelo al de una linea recta horizontal. Dada una regresión lineal simple, un $R^2$ negativo sólo es posible cuando la ordenada en el origen o la pendiente están restringidas de forma que el mejor modelo es peor que una linea horizontal.\n",
    "\n",
    "Si representamos la **varianza de la variable dependiente** por $\\sigma^{2}$ y la **varianza residual** por $\\sigma _{r}^{2}$, el coeficiente de determinación viene dado por la siguiente ecuación:\n",
    "\n",
    "$$ R^{2}=1- \\frac{\\sigma_{r}^{2}}{\\sigma ^{2}}$$\n",
    "\n",
    "\n",
    "Siendo $\\hat{y}_i$ el valor predicho de la muestra i y $y_i$ el valor real, el $R^2$ estimado sobre $n_{\\text{muestras}}$ se define como:\n",
    "\n",
    "$$ R^2(y, \\hat{y}) = 1 - \\frac{\\sum_{i=0}^{n_{\\text{samples}} - 1} (y_i - \\hat{y}_i)^2}{\\sum_{i=0}^{n_\\text{samples} - 1} (y_i - \\bar{y})^2}$$\n",
    "donde $$ \\bar{y} =  \\frac{1}{n_{\\text{samples}}} \\sum_{i=0}^{n_{\\text{samples}} - 1} y_i$$\n",
    "\n",
    "#### 3.4.1 Consideraciones R²\n",
    "\n",
    "1. R² no puede determinar si los coeficientes y las predicciones tienen bies: Hay que checkear los residos --> Si observamos patrones en los plots de residuos es indicativo de un mal ajuste a pesar de un R2 elevado\n",
    "\n",
    "1. Cada vez que añadimos un predictor a un modelo, el R² aumenta aunque sea por suerte, pero nunca decrece. Por consiguiente, un modelo con muchos terminos puede parecer mejor simplemente por el hecho de tener más terminos. Para prevenir este efecto, podemos usar el **adjusted R²**, una versión modificada que se ajusta al número de predictores en el modelo. De ésta forma, el R² solo aumenta si el nuevo término mejora el modelo más que por mera suerte. Siempre es más bajo que el R²\n",
    "\n",
    "$$\\bar{R}^2 = 1 - \\frac{N-1}{N-k-1}(1-R^2)$$\n",
    "\n",
    "\n",
    "n – numero de observaciones\n",
    "\n",
    "k – numero de parametros"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[![License: CC BY 4.0](https://img.shields.io/badge/License-CC_BY_4.0-lightgrey.svg)](https://creativecommons.org/licenses/by/4.0/) <br/>\n",
    "Isaac Lera and Gabriel Moya <br/>\n",
    "Universitat de les Illes Balears <br/>\n",
    "isaac.lera@uib.edu, gabriel.moya@uib.edu"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.12.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}