{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Data Literacy - Project\n", "## Gender Share in Movies\n", "#### Tobias Stumpp, Sophia Herrmann" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### README & TODO\n", "\n", "This file analyzes the time-frame 2000-2020, if due to the introducton of the Bechdel test,\n", "if it is possible to find a relationship between the share of actresses on the pricipal cast and the average movie rating.\n", "Additionally, if it is possible that the average movie rating can be predicted by a linear regression model and the predictors:\n", "- share of actress on the principal cast \n", "- share of actress on the principal cast, genre and movie duration" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import pandas as pd\n", "import os\n", "import matplotlib.pyplot as plt\n", "import statsmodels.api as sm " ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "path = '../dat/'\n", "os.chdir(path)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Read data, keep only the years 2000 - 2020 and include the share of actresses on the principal cast" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "data_movie = pd.read_csv('data_movie.csv')\n", "\n", "# Keep only the years of 2000 to 2020 & sort the data frame according years\n", "dat_2000 = data_movie[(data_movie.startYear >= 2000) & (data_movie.startYear <= 2020)].sort_values(\"startYear\")" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/tobi/anaconda3/lib/python3.8/site-packages/pandas/core/indexing.py:1637: SettingWithCopyWarning: \n", "A value is trying to be set on a copy of a slice from a DataFrame\n", "\n", "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n", " self._setitem_single_block(indexer, value, name)\n" ] } ], "source": [ "# Compute share of actresses on the principal cast and include it into the data set\n", "\n", "# 1)\n", "# Number Actress\n", "number_actress = dat_2000[dat_2000.category == \"actress\"].groupby([\"tconst\"]).category.count().reset_index() \n", "number_actress = number_actress.rename(columns = {\"category\" : \"nactress\"})\n", "# Number Actors\n", "number_actor = dat_2000[dat_2000.category == \"actor\"].groupby([\"tconst\"]).category.count().reset_index()\n", "number_actor = number_actor.rename(columns = {\"category\" : \"nactor\"})\n", "\n", "# Merge number of actress & actors to data_movie, and delete original category-column & delete row-duplicates\n", "dat_2000 = pd.merge(dat_2000, number_actor, on=\"tconst\", how='left')\n", "dat_2000 = pd.merge(dat_2000, number_actress, on=\"tconst\", how='left')\n", "dat_2000.drop([\"category\"], axis = 1, inplace = True)\n", "dat_2000 = dat_2000.drop_duplicates()\n", "\n", "# 2)\n", "dat_2000[\"proportion\"] = dat_2000[\"nactress\"] / (dat_2000[\"nactress\"] + dat_2000[\"nactor\"])\n", "\n", "# Having NaN's, for films w/o actress or w/o actors\n", "# Replace 0 or 1. 1 if no actor is given. 0 if no actress is given\n", "propor_1 = dat_2000.index[(dat_2000.proportion.isnull()) & (dat_2000.nactor.isnull())]\n", "propor_0 = dat_2000.index[(dat_2000.proportion.isnull()) & (dat_2000.nactress.isnull())]\n", "\n", "dat_2000.proportion.loc[propor_1] = 1\n", "dat_2000.proportion.loc[propor_0] = 0" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "### Descriptive Analysis: Share actress on pricipal cast & average rating" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ " lower quantile: 0.25\n", " upper quantile: 0.5\n", " mean: 0.33\n" ] } ], "source": [ "# It is not important for further analysis,\n", "# only to have better understanding of the underlying data set\n", "\n", "# Boxplot:\n", "plt.boxplot(dat_2000.proportion)\n", "plt.title(\"Boxplot: Actress share 2000 - 2020\")\n", "plt.xticks([])\n", "plt.show()\n", "\n", "# See:\n", "# - Proportions are right skewed, having the mean value by 0.33.\n", "# - Upper quantile 0.75: 0.5; lower quantile 0.25: 0.25\n", "print(f\" lower quantile: {np.percentile(dat_2000.proportion, 25)}\")\n", "print(f\" upper quantile: {np.percentile(dat_2000.proportion, 75)}\")\n", "print(f\" mean: {np.percentile(dat_2000.proportion, 50).round(2)}\")\n", "# Some Outliers between 0.9 - 1.0\n", "\n", "# - Porportions -> 50% of the actress shares lies between 0.25 & 0.5.\n", "# Conclude: Actresses are less precence than actors" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ " lower quantile: 5.1\n", " upper quantile: 6.9\n", " mean: 6.1\n" ] } ], "source": [ "# It is not important for further Analysis,\n", "# only to have better understanding of the underlying data set\n", "\n", "# Boxplot:\n", "plt.boxplot(dat_2000.averageRating)\n", "plt.title(\"Boxplot: Average rating\")\n", "plt.xticks([])\n", "plt.show()\n", "\n", "# See:\n", "# - Average ratings are normal distributed\n", "print(f\" lower quantile: {np.percentile(dat_2000.averageRating, 25)}\")\n", "print(f\" upper quantile: {np.percentile(dat_2000.averageRating, 75)}\")\n", "print(f\" mean: {np.percentile(dat_2000.averageRating, 50).round(2)}\")\n", "# Some Outliers between rating of ~9.2 - 10 and ~0 - 2.2 \n", "\n", "# - Average ratings -> 50% of the average ratings lie between 5.1 & 6.9." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Analyze possible realtionship between average rating and the share of actresses on pricipal cast. Additionally analyze if average ratings can be predicted by the use of linear regression models" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6wAAAHgCAYAAAChPuB6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAADkgklEQVR4nOydeXhU1f3G3zuEEEISQiDsS2TfGRFRXEAF3Frbamm1ivxsWCJKa4sWUbClFXGhtrXV2lqNuIIbFiGIERWoxQ1x2HdEQJYkhJCNLJO5vz9gJmM6czL33HPOPXfy/TwPT/ra5N4z957lnrnnfY9hmiYIgiAIgiAIgiAIQjc8TheAIAiCIAiCIAiCICJBE1aCIAiCIAiCIAhCS2jCShAEQRAEQRAEQWgJTVgJgiAIgiAIgiAILaEJK0EQBEEQBEEQBKElNGElCIIgCIIgCIIgtCTB6QLEQrt27cysrCyni0EQBEEQBEEQBEFI4MsvvywyTTOz4X93xYQ1KysLGzZscLoYBEEQBEEQBEEQhAQMw/gm0n+nJcEEQRAEQRAEQRCEltCElSAIgiAIgiAIgtASmrASBEEQBEEQBEEQWkITVoIgCIIgCIIgCEJLaMJKEARBEARBEARBaAlNWAmCIAiCIAiCIAgtoQkrQRAEQRAEQRAEoSU0YSUIgiAIgiAIgiC0hCasBEEQBEEQBEEQhJbQhJUgCIIgCIIgCILQEpqwEgRBEARBEARBEFpCE1aCIAiCIAiCIAhCS2jCShAEQRAEQRAEQWgJTVgJgiAIgiAIgiAILZE2YTUMI9cwjALDMLaG/bcMwzDeNwxjz9mfbWSdnyAIgiAIgiAIgnA3Mt+wLgJwdYP/NhvAB6Zp9gHwwVlNEARBEARBEARBEP+DtAmraZrrABQ3+M8/BPDC2f/9AoAfyTq/Sqqq/Hh381H4/X6ni0IQBEEQRBwTCJgoLKuGaZpOF4UgCEIJCYrP18E0zaMAYJrmUcMw2is+v3CqqvzoP++9kN457yokJam+rARBEARBxDuBgImf/etTfPnNSZzXow0WT70QHo/hdLEIgiCkom3okmEY0wzD2GAYxobCwkKnixOVj3YXMjVBEARBEIQITlTU4MtvTsIfMPHlNydxoqLG6SIRBEFIR/WE9bhhGJ0A4OzPgmi/aJrmM6ZpjjBNc0RmZqayAlpl/MBMpiYIgiAIghBBu5REnNejDRI8Bs7r0QbtUhKdLhJBEIR0VK9dfQfA/wF45OzPZYrPL5yEhATsnHcVPtpdiPEDM5GQQMuBCWcIBEycqKhBu5REGIZ+S8R0Lx9BEITuGIaBxVMvpL6UIIgmhbTZlWEYiwFcBqCdYRiHAfwOZyaqrxuGMRnAQQA/kXV+lSQlJeCaoZ2cLgbRhNHd16R7+QiCINyCx2MgM7WF08UgCIJQhrQJq2maP4vyf42VdU6CaKpE8jXp9ECje/kIgiAIgiAIPdE2dIkgiNjR3deke/kIgiAIgiAIPTHcsI/XiBEjzA0bNjhdjKj4/QHsLSxH3w4p8HjoOwDCGXT3iFI7+S663y/COahuECyoftRD14IgYsMtbcUwjC9N0xzR8L9TQpBN/P4Azp3/Psqq/EhNSsBXc8cjIYEexgn16OxrCgRM3PLcZ+RhPQt5eoloUN0gWFD9qIeuBUHERjy0FZpZ2WRvYTnKqvwAgLIqP/YWljtcIoLQD9o78LvQ9SCiQXWDYEH1ox66FgQRG/HQVmjCapO+HVKQmnTmRXVqUgL6dkhxuEQEoR/kYf0udD2IaFDdIFhQ/aiHrgVBxEY8tBXysAqAvHkE0Thu8U+ogq4HEQ2qGwQLqh/10LUgiNhwS1uJ5mGl2ZUA/P4Avi6sQCAQcLooBKEtgYCJE+XVcMOXZARBEA0JBEwUllEfRhCE+3D7MxiFLtmkqsqP/vPeC+md865CUhJdVoIIh8LJvks8BCAQcqC6oSe63BddyqEDdC0IIjbi4RnMXaXVkI92FzI1QRAUTtaQeAhAIORAdUNPdLkvupRDB+haEERsxMMzGE1YbTJ+YCZTEwRB4WQNiYcABEIOVDf0RJf7oks5dICuBUHERjw8g1HokgAOnyjFzDe24NmbhyAtLc3p4hBNFN0N9eUVNViy4RAmjeqGxER6sKCwNiIalZW1WL7lKG4Y3gnNmzd3ujjEWXTpY3Uphw7QtSCI2Kiq8uOj3YUYPzATCQn6WhejhS7RhNUmRacqMeLhj0J6w32Xo13rZAdLRDRFdPfykNf7u+h+vwjnoLZCEARBiMRNHlZKCZbEglW7mJogVKC7l4e83t9F9/tFOAe1FYIgCEIk5GElMP+6fkxNECrQ3ctDXu/vovv9IpyD2gpBEAQhEvKwKkLnJcHAmWXBC1btwvzr+iE5mZYDE86gu5fHLf4JVeh+vwjnoLZCEARBiMQtuRm0JFgi5TU12H28FDU1tKzP7bh5Y3iPx0BmagttJz81/jrsOlaKuro6p4uiBbrfLx1xc/u0QkKCB+dkttL6oYJwjqbSDgiCEIffH8DXhRUIBAJOF4ULesNqkwOFJbjs8f+G9Jq7L0ZWZrpzBSK4oSAceZSWV2Po/NUhvXnuOKSltHCwRITbaCrt003hGIR6mko7IAhCHG4K86M3rJKYsXgTUxPugYJw5JG7/gBTE0RjNJX2GQ/hGIQ8mko7IAhCHPEQ5kcTVpu8PGkYUxPugYJw5DF9TBZTE0RjNJX2GQ/hGIQ8mko7IAhCHPEQ5kdLggXwr0824KFlx/HmhA4YMeJ/3mITLkJ0EA4F69Sz93gxbn/Jh9ezvcjIyFByTp2vv85l05Wmcs0OFp3CXa9txqKJQ9G6dWuni0NoRlNpB7FAAWUEERvlFTVYsuEQJo3qhsREfb/oirYkmCasNlm1Yy9uf6F+79V//F8/XD2gt4MlInSBvEb1FJRUYOQja0L689mXoX16K6nn1Pn661w2wlmOFJfhosfWhfT6WaPROSPVwRIRhJ64yZdHEE7ipmwE8rBK4o6wyWokTTRdyGtUz7wVO5haBjpff53LRjjLrKVbmZogiDPEgy+PIFQQD9kINGG1Sf6UfkxNNF3Ia1TPwusHMLUMdL7+OpeNcJanbxzM1ARBnCEefHkEoYJ4yEagJcECWPjuajy1thrPjmuBcePGOV0cQiPIa1TP1m8LMPl5H5ZP86J9+/ZKzqnz9XeLn4RQz8aDR5Gd68O7OV506tTJ6eLEhI5tLV4zCXQphw6UlFXhqbX7cPe4XkhKSnK6OAShLZWVtVi+5ShuGN4JzZs3d7o4USEPqyTW7DmA257bFtKLJg/CZX2ynCsQQWiIEx5WnamsrMXAP+SH9PbfXonkZH0HEEIdx06W48JH14b0p/eOQcc2en8brqMnW3SZdPmMupRDB8jDShCxUVNThwG/W4U6E2hmADt+fzUSE5s5XayIkIdVEtNytzE1QRDOeFh1ZvmWo0xNNF3mLNvG1DqioydbdJl0+Yy6lEMHyMNKELHx5cGTqDv7frLOPKPdBk1YbZKfM4ipCYJwxsOqMzcM78TURNPliQmDmFpHdPRkiy6TLp9Rl3LoAHlYCSI2Rp7TBs3OLsRoZpzRboOWBAtgzZ4DmJa7Dfk5g5CVleV0cQhCSwpKKjBvxQ4svH4AWrVqusuBg7jFT0Ko59jJcsxZtg1PTBiElBS9lwMH0dFXSR7W+If2YSWI2KipqcOXB0+embw203M5MEBLgqVy8NQp1JjAqVOnnC4KoRmBgInCsmqo+GLI7w9g59FSBAIB6efiofj0aWw5XILTp08rO6fO16Smrg57C8tQV1fndFGEIrPO63w/RbKzqAj/2VWEoqIip4sSMx6PgczUFnE9gWoKn5EgiPikprYO27495dpnDnrDapMXN2zCb988HNJ/mNAVk0YMc7BEhC6oDMfQfVPonUeLcPUTn4X0qrsuQP9O7aSeU+drUlJWBe9DH4S0b85YpKe6P+FSZp3X+X6KhIL8xBCv4UTx+rl4oNAlgoiN8ooaDH7w/ZDe+sB4pLTS005Ab1glET5ZjaSJpovKcAzdN4We+oKPqWWg8zV5au0+pnYrMuu8zvdTJBTkJ4Z4DSeK18/FA4UuEURsLNlwiKndAE1YbbLsZ12Zmmi6qAzH0H1T6LeneJlaBjpfk7vH9WJqtyKzzut8P0VCQX5iiNdwonj9XDxQ6BJBxMakUd2Y2g3QkmAB/Pq5PLy9B/jDQGDSpO85XRwC/KEUosMs/P4A9haWo2+HFHg8cr8fEmWolxXosXrXftz+/A7kTxuAnj17CjsuC5XX3yq6hFC5KZiGp47rXAei8a9PNuChZcfx5oQOGDHif1ZGETHiprptBQoaqqe0vBq56w9g+pgstGjRwuniEIS2uCXML9qSYJqw2oQ8rPrB6/Fx80bzos4lq8yHT5TikoX/CemPf3MpurZNs31ct6LLJt5u8sPxlNWNvlcn/N5E4+jSVnTpO3SArgVBxAZ5WAnysGoIr8fHzRvNizqXrDLPfGMLUzc1dNnE201+OJ6yutH36oTfm2gcXdqKLn2HDtC1IIjYIA8rQR5WDeH1+Lh5o3lR55JV5mdvHsLUTQ1dNvF2kx+Op6xu9L064fcmGkeXtqJL36EDdC0IIjbIw6oInZcEA/XLgpf9rCuGDaPlwDqgi4dVpedJ1LlklfnwiVLMfGMLnr15CNLSmu5y4CC6bOKtiy8vFnjK6kYP686jRZj6gg9vT/GiXTtaDqwLurQVXfoOHaBrQRCxUV5RgyUbDmHSqG5ITNT3y2laEiyR/QfOLAM+cICWA8skEDBRWFaNWL5kCQRMnCiP7XfDEb0xvMqN5nk/syp2FhVh44ESFBUVKTunlTqjmsrqWny48zhqa2sdLUdVlR8fbj8Ov98v5Hg1NXX4ZG+RlM3Jeeq4x2OgbYqaNiiKnUVFOFRSrbSt2EXHtlZV5ce7m48Kq9sq+3MWCQke9O6Q6povYGRSeroab3x5CNXV1cKPLapO69g2iKZHeU0NPvm6CDU1+lp/WNAbVpvMezMPi8KKdtsIYN4ESgoWjZXACzcGrdhF1GeWFSyyetd+THl+R0g/+/MBGNdPblKwLiEpkSgpq4L3oQ9C2jdnLNJTk5SXo7KyFgP/kB/S2397JZKTm3MfT2YICk8d17kOROPfW3biV6/U78v7l1t64UdD+jtYosbR8TpXVfnRf957Ib1z3lVISnJ/oq6O19opik5VYsTDH4X0hvsuR7vWyUKOrXuQIUFY4djJclz46NqQ/vTeMejYRk+LDL1hlcSiDWxNiMFK4IUbg1bsIuozywoWuT1sshpJy0CXkJRIPLV2H1OrYvmWo0xtFZkhKDx1XOc6EI3wyWokrSM6XuePdhcytVvR8Vo7xYJVu5jaDroHGRKEFeYs28bUboAmrDZ5YhRbE2KwEnjhxqAVu4j6zLKCRfKnDWBqGegSkhKJu8f1YmpV3DC8E1NbRWYICk8d17kORGPVbb2YWkd0vM7jB2YytVvR8Vo7xfzr+jG1HXQPMiQIKzwxYRBTuwFaEiyA4bPzUAxgLIDnHqHlwLKwEnjBG8SgS6gGD6I2k5d1DZ74YB3+/H4ZXrw2FaNHjxZ2XBY6B+4cKCzBjMWb8PKkYUhPT3esHJWVtVi+5ShuGN4JzZvzLwcOIjPYged+urFNL3jnXTyzPoCnx3hwzTXXOF2cmNDx3ojqE4PoUpd0DU9x4vpsP1KIKYt8WDbVi8xMsV9KiBo/dKk3RNPGd+gYsnN9WJnjRceOHZ0uTlSiLQmmCatNfv9WHp7/ol7//Hzgdz+mSauT8HpG3Ow10d23++nXh3HTPzeF9JKcYbjwHLlbQOl8P3XxsIpGpoe1qfDh7q+Rnbs9pHOzB+KKvuc4WKLG4WlrOrfPSOhSXtG+c1E4cX0KSiow8pE1If357MvQPr2VkGPrcr8JQgTkYSW+M1mNpAn18HpG3Ow10d23m527ialloPP91MXDKhqZHtamQk7YZDWS1hGetqZz+4yELuUV7TsXhRPXZ96KHUxtB13uN0GIgDysBP56EVsT6uH1jLjZa6K7bzc/ZxhTy0Dn+6mLh1U0Mj2sTYX8nIFMrSM8bU3n9hkJXcor2ncuCieuz8LrBzC1HXS53wQhAvKwKkLnJcEAkDU7DwDQF0A+eVi1gNd7ItrPp9JDKWoDdVl+m/tfzcOrm4GF5wE/+YmadmL1fqr0Gh0sOoW7XtuMRROHonXr1lLPxUL0Z9bVX+cm7lmUhzd3Ag8NAW65xR1jCk890tljHgldvIgFJRWYt2IHFl4/AK1aiVkCKwIn7ufmw8eRnetDXo4XHTp0EHpsXe43QYgguGXaqtt6oX9/fbdKIw+rJCb/OQ8fHK/XYzsAz/3aHQ8Y8Qqv90T0vn0qfaW67xn393Wf4LGVxSE969oM3DFabqS21fup0rNUWl6NofNXh/TmueOQltJCyrlYiP7M5GG1z1Nr12Phu/VLqX9zTRvcOSb+lu6QR5CP8ooaDH7w/ZDe+sB4pLRy/oshJ+6nm3x5BOEkTuSI8EIeVkmET1YjaUI9vN4T0fv2qfSV6r5nXPhkNZKWgdX7qdKzlLv+AFOrQvRnJg+rfcInq5F0vEAeQT6WbDjE1E7hxP2MB18eQajAiRwR0dCE1Sb3dGVrQj283hPR+/ap9JXqvmfc4h9mMLUMrN5PlZ6l6WOymFoVoj8zeVjt8+oP2jB1vEAeQT4mjerG1E7hxP2MB18eQajAiRwR0dCSYAEElwXf0xWYMYOWA+sAr/dE9L59Kj09ovw2snw7wWXBi3+YgVGj5C4HDmL1fqr0LJWWVyN3/QFMH5OFFi3ULwcOIvozi/JSN2WCy4Jf/UEbXHRR/C0HDkIeQT509Yk7cT+PnSzHnGXb8MSEQUhJoeXABBGNT78+jOzcTcjPGYauXfV9u0ZLgiUSHC80GjeaPB6PgczUFpYHzaSkBFwztJOQyWqwHG1TrJcjHjl16tR3fqrA4zGQntw85uvPW294CJgmKmr8cPpLw6oqPz7cfhx+v1/I8TweA21ayXlgDQRMFJZVO37NZBMMCBMR/KYKvz+AnUdLEQgEnC5KCB3LJIKkFgm4pK+YL1VForL/DFJZW4tvT1aitrZW2Tl1xo19pBvL7EY6pqSgZ2aKa7/YoTesNpn+VB7eDbOQXNMNePpOestKqA2h0D106eHlq/DP/9aFdM7FzXDfdVfbPi4LnQOASsqq4H3og5D2zRmL9NQk5eWorKzFwD/kh/T2316J5GT+SZLMoLGmEtLzzPovsOCdgpC+/wftMe2i8x0sUePw3HfZ91Nl6J1K4vVz8bC/4CSu+NP6kP5w5kXo2T4+l9DHghv7SDeW2Y0cKCzBZY//N6TX3H0xsjLTnSsQA3rDKol3D7E10XRRGUKhe+hS+GQ1kpaBzgFAT63dx9SqWL7lKFNbRWbQWFMJ6QmfrEbSOsJz32XfT5WhdyqJ18/Fwx2v+Ji6qeHGPtKNZXYjMxZvYmo3QBNWm9zfk62JpovKEArdQ5f+cVkzppaBzgFAd4/rxdSquGF4J6a2isygsaYS0vPGj9sztY7w3HfZ91Nl6J1K4vVz8bD4Ni9TNzXc2Ee6scxu5OVJw5jaDdCSYAFkzc4L/e8Dj9ByYFlYCTDSJfhFZTiGqM8s69r1np0HP4DzAbyhqJ1UVtZi+ZajuGF4J+38gFu/LcDk531YPs2L9u2dm5QUl57G46v3YO41fdCyZUvbx5N5zXmO7cZgn7Gz87APwAQAf3TJmMLTb8gOpRMdbKZLXdIlsE0Hlny1FbNf+wYrJvbA4MGDnS5ORNwYvqgSN5bZjazasRd3vLAL+VP6oXfv3k4XJyrRlgTThNUmF8zOQ/jWqx0AfOaSBww3YcW3o4t3UbQ/kIUoX5Msf9SMf+RhxYF6/f0s4Mnb5bYTnb1eh0+U4pKF/wnpj39zKbq2TVNejqoqP/rPey+kd867CklJ/EEuMtseT1nd6I+a/XIelmyt1zcNBh6ZqPeYwnOdZd8b0f2vLnVJ5biiO58f+BY//YcvpF+/3YuRWV2cK1AEdB6HiKbDzqNFuPqJz0J61V0XoH+ndg6WKDrkYZXE8UY0IQYrvh1dvIui/YEsRPmaZPmjwierkbQMdPZ6zXxjC1Or4qPdhUxtFZltj6esbvRHhU9WI2kd4bnOsu+N6P5Xl7qkclzRnexcH1PrgM7jENF0mPqCj6ndAE1YbdIw51Ru7mnTxYpvRxfvomh/IAtRviZZ/qgH+rC1DHT2ej178xCmVsX4gZlMbRWZbY+nrG70Rz16LlvrCM91ln1vRPe/utQlleOK7ryX42VqHdB5HCKaDm9P8TK1G6AlwQIILgu+GsA/aDmwNNzoYVXpoRTlk5HltwkuC36gDzB5spp2otI7ZJXDJ0ox840tePbmIUhLU78cOEhVlR8f7S7E+IFi9nWU2fZ4yupGf1RwWfCj5wI33uiOMYXnOsu+N6L7X13qks7efNV8fuBbZOf68F6OF1266LUcOIjO4xDRdNh5tAhTX/Dh7SletGun53JggJYES6X92XGrvf5f3rsaj8dA25TYNiVPSPCgd4dUy4NDVZUf724+Cr/fz1vM7+DxGEhr2dw1D8oyCe5VrXLP6kDAxMmKGi03JD9WXo6dR0+hvNzZJWIej4H0ZHF1lLftxQJPWQMBEyfK3bUpfXHxd3+6AZ7rLPveiO5/PR4DmamxjUEy8QcCOHrqNAKBgKPl0IG0Fi3QJrmF1uFTKsehQMBEYZnz/Z0u5SDq8QcCqPEHXNtv0BtWm1x3Xx62hF3CIQaw/GF3fCPuJqwEXvCGY4gOoBF9PBa6hy45ESSjS/hWJDZ8cwQTnv4qpN+cfi5G9OisvByir5HMYBqesrox8GTaX/OQf6ReX9kZeOaXeo8pPNdZ9r1R2f+qpLS8GkPnrw7pzXPHIS1F38maTNwQJKNyHNIlGEyXchD1bP22AN//2xchveIX52NwFz23TKM3rJLYYrI1IQYrgRe84RiiA2hEH4+F7qFLTgTJ6BK+FQldwkJEXyOZwTQ8ZXVj4En4ZDWS1hGe6yz73qjsf1WSu/4AUzcl3BAko3Ic0iUYTJdyEPVMft7H1G6AJqw2mZTI1oQYrARe8IZjiA6gEX08FrqHLjkRJKNL+FYkVjUIB2moVSH6GskMpuEpqxsDT+7twdY6wnOdZd8blf2vSqaPyWLqpoQbgmRUjkO6BIPpUg6inuXTvEztBmhJsACyZueF/vcBCl2ShpXgAt5wjPKKGizZcAiTRnVDYqL9TrakrApPrd2Hu8f1QlJSku3jsRC1mbys0JxgO2kNYJOidqLy+ltFlw3vRdf54tLTeHz1Hsy9pg9atmwpoIT18IQuuTHwxI1jCk+/ITs8SHSgmC51qaCkAvNW7MDC6wegVatWjpWjIU5cn2fWf4EF7xTgjR+3x/nnn6/knFZRGQKpSx3VJaCMqOfp/3yKR/NOYMmP2uLCCy90ujhRibYkmCasNuk/Ow9VYToJwE6XPGC4CRWeCNGeKpUeKlGbycu6zj/5fR6+OF2vz28JvPE7ue1E1DWRwd7jxRj3509CevWvR6F3hwzl5RDtryopq4L3oQ9C2jdnLNJTxXxR0FR8UT+amwdfWOabNwH493y9xxSee6OzxzwSuvihdfXmOnF9Ptz9NbJzt4d0bvZAXNH3HKnn1Jmm0kcS1lm9az+mPL8jpJ/9+QCM69fTwRJFhzyskqhqRBNiUOGJEO2pUumhErWZvKzrHD5ZjaRlIOqayOD2l3xMrQrR/qqn1u5jajs0FV+Uz8/WOsJzb3T2mEdCFz+0rt5cJ65PTthkNZJuajSVPpKwzu1hk9VI2g3QhNUmoxrRhBhUeCJEe6pUeqhEbSYv6zpPS2NrGYi6JjJ4PdvL1KoQ7a+6e1wvprZDU/FF/bwlW+sIz73R2WMeCV380Lp6c524Pvk5A5m6qdFU+kjCOvnTBjC1G6AlwQJwo9/IjVjxRPB6l9zk52uIKD+YLA+ME+1E5fW3yrq932Bq7lbk5wxGjx7OJeuI9hEeO1mOOcu24YkJg5AieNNdnnbtRi+VG8cUnn5DlO9eFSq9iCyOFJdh1tKtePrGwUhNTXWsHA1x4vo8tvJ9/H1dDf41NhHjx49Xck6rqOyDRJ7LzrHc2O/GO098sA5/fr8ML16bitGjRztdnKiQh1USfWbnoTZMNwewxyUPGPEKr8dHtDdIZw9lNGR5YLL/lIcPC+r1Fe2B3Jly24muXi/gzAPnRY+tC+n1s0ajc4b6B083+bZ5PI9u9HS5cW9vnuusc/uMhC51qbj0NIYv+DCkN95/BTLSnP8yzonrs2rHXtz+wq6Q/sf/9cPVA3pLPadVdKk3VrFTbrd+5njm068P46Z/bgrpJTnDcOE5XR0sUXTIwyqJ2kY0oR5ej49ob5DOHspoyPLAhE9WI2kZ6Or1AoBZS7cytSrc5Nvm8Ty60dPlxr29ea6zzu0zErrUpcdX72Fqp3Di+twRNlmNpHVAl3pjFTvldutnjmeyczcxtRugCatNRjaiCfXwenxEe4N09lBGQ5YH5u4ubC0DXb1eAPD0jYOZWhVu8m3zeB7d6Oly497ePNdZ5/YZCV3q0txr+jC1UzhxffKn9GNqHdCl3ljFTrnd+pnjmfycYUztBmhJsACCy4JHAnidlgNrAa+HVfS+fbL3GZSBLO9JcFnw3V2AX/xCTTsRfT9FoosPTbRnWeY15/HIudFLFVwWPCkR+MMf3DGm8FxnndtnJHSpS7p68524Pqt27MUdL+xC/pR+6N1br+XAQXSpN1YhD2t88enXh5Gduwn5OcPQtauey4EBWhIsleAyYD0W5riTQMBEYVk1WF+gVFX58e7mo/D75e3zkJDgwTmZrYSFDSUkeNC9bbKSDbyLS09jztLNOH3a3p4xgYCJE+Xse8FDcBnwO98KPSyTGn8ddh0rRV1dnbqTxsjBU6ew8ZtinDp1ytFy+P0BfF1YgUAg4Gg5YqGyuhYf7jyO2trYzRd+fwB7j5e54vMFOXW26Z1y0Uo6j8dAZmoLSw+novvbhtTU1OGTvUXC2j/PZ5SBPxDAycoaJXU6lrE5/HdljB0siiorEQBQWVmp7JxW0aXeqKQpfmbdqaytRW0tLI2fOkFvWG0SnuYYxC2pjroQi0HfSkAHb5iH6AAanpAYXkQFccja/N2JdlJaXo2h81eH9Oa545CWokcSqS4BCKKDb2QG6ZSUVcH70Ach7ZszFumpScy/UdkGRTF6dh4OhunuANbF4Zgiq68J4sZ7HwsFJRUY+ciakP589mVon95KyrmshOfIvp+RePnLzZj7xqGQnv+Tbph43lCp52wqUHBSfPHh7q+RHbZPcW72QFzR9xwHSxQdesNKaEssBn0rAR28YR6iA2h4QmJ4ERXE4cTm77LIXX+AqZ1ElwAE0cE3MoN0nlq7j6kjobINiuJgIzpekN3XuPHex8K8FTuYWiRWwnOcGDvCJ6uRNMEPBSfFFzlhk9VI2g3QhNUmDWM/9N76XE9iMehbCejgDfMQHUDDExLDi6ggDlmbvzcsjYqYkOljspjaSXQJQBAdfCMzSOfucb2YOhIq26AoftiIjhdk9TVB3HjvY2Hh9QOYWiRWwnNk389IvHNzN6Ym+KHgpPgiP2cgU7sBWhIsADdu8q4bsRj0yytqsGTDIUwa1Q2JiezO08rvhiM6BKSgpALzVuzAwusHoFUrOcu2guwvOIk7XvFh8W1etGnD/3AmOoQniBPtROX1t8qzn36J+f8+hrd+0hHnnXeeY+UoLa9G7voDmD4mCy1a2F8yLTNMau/xYtz+kg+vZ3uRkZER09+4LdgHaDpjiui61xDR954n9EsGPO2AFyvhOU5cn9++nocXNwJ/vgC4/vr4bSuxIjLsiEKX4os/rvoAT66pwnPjkzB27FinixOVaEuCacJqE/KwqsGKP4bXSyPag8PjuXPDuXjoNTsP4bEnzQDsk9xOZPop7bL58HH84Mn6Pu2dGSMwtGsH5eUQ7fMT5aWORNGpSox4+KOQ3nDf5WjXOpn5N270YfWbnYfqMN0CwK44HFMqK2sx8A/5Ib39t1ciOVlckrroe6+LJ1ZXb74TbS1/5z5MW7QzpJ+5rT+u7N/4yot4RZf+TpdyEPWs3rUfU56vtw88+/MBGNevp4Mlig55WAlXY8Ufw+ulEe3B4fHcueFcPDTM6FSR2SvTT2mX7FwfU6tCtM9PlJc6EgtW7WLqSLjRh1XdiI4Xlm85ytR2EX3vdfHE6urNd6KtTQ+brEbSTQ1d+jtdykHUc/vzO5jaDTgyYTUM49eGYWwzDGOrYRiLDcPQ51WQRTo1ogkxWPHH8HppRHtweDx3bjgXDw2/Kvufr84kINNPaZe8HC9Tq0K0z0+UlzoS86/rx9SRcKMP68JGdLxww/BOTG0X0fdeF0+srt58J9pa/tT+TN3U0KW/06UcRD350wYwtRtQviTYMIwuAD4GMNA0zdOGYbwOYKVpmoui/Y3OS4KB+mXBnQB8EodLt3TBireS14cp2r9ZUlaFp9buw93jeiEpSe73MirPxUNwWfAIAG8qaic6+xc3Hz6O7Fwf8nK86NBB/XLgIKJ9Z8Wlp/H46j2Ye00ftGwpZjlwkKJTlViwahfmX9cPycns5cBB3OilCi4LvhDAkjgeUyora7F8y1HcMLwTmjcXtxw4iOh7r4uHVbb3lxcn2lr+zn2Yvmgn8qf2R69een1R6wS69He6lIOoZ/Wu/bj9+R3InzYAPXvquRwY0G9JcAKAloZhJABIBnDEoXIIReyCJqIhHo+BtimxbURdU1OHTQdLLG8Yz/t30UhqnoDh3TOUTJaSWzTHFf072H7w8/sD2Hm0VPim9MEr+pXQo7KpqvVj48Fi+P3+mH4/EDBRWKZm4/sjZWUorvSjrKxM+rlYVNX48dn+EzFfo8ZI8HiQmdJCaGBX+LHbtkqUcmydCC4DdmazIz6qqvx4d/NRS/UoKSkBVwzsIK1/rDxdi2UbD6O2tlbI8QIBEyWVtUr6B2Y5TBMVNX7Hy9GQ8soavPLJAdTUqFsCWlBRgToAFRUVys5pFVljqmzsjIcej4HM1Nie1wg17CkogB9AQUGB00XhwpHQJcMw7gLwEIDTAPJN07yF9fs6v2Gl0CU1WDHx84Z5iA4BURn6IyoQRNbm7060E6tBVCqDIlbt2IvbX6j3YP7j//rh6gG9pZyLhegAF5mBMDzBYm4M/3DjmMLT18m+N+UVNRj84PshvfWB8Uhpxb80UZcQN10D9pwIg3r5y83f2Xt1/k+6YeJ5Q6We0yqyxtRIiGxTbuw7ieg8/Z9P8WjeiZC+93ttMf1SPQ0n2rxhNQyjDc5sLXcOgM4AWhmGMTHC700zDGODYRgbCgv1CUshnMGKiZ83zEN0CIjK0B9RgSBObP4uC6tBVCqDIu54YRdTq0J0gIvMQBieYDEK/1ADT18n+94s2XCIqa2iS4ibrgF7ToRBhU9WI2kdUDmmimxT1HfGF+GT1UjaDTixrmocgK9N0yw0TbMWwFIAFzX8JdM0nzFNc4RpmiMyM/UJSyGcwYqJnzfMQ3QIiMrQH1GBILI2f2/4rleF+8tqEJXKoIj8Kf2YWhWiA1xkBsLwBIu5MfyjoetXrAtYDjx9nex7M2lUN6a2ii4hbroG7DkRBvXOzd2YWgdkjamRENmm3Nh3EtFZ8qO2TO0GnAhdugBALoDzcWZJ8CIAG0zT/Fu0v9F5STDQdDZ5dxorgUi8oRQHi07hrtc2Y9HEoWjdurXdInOFxPAiKhBEVhCKE+3kQGEJZizehJcnDUN6enqjv68yKEKXTbwLSiowb8UOLLx+AFq1amX7eMdOlmPOsm14YsIgpKSIfTjbe7wYt7/kw+vZXmRkZMT0N+UVNViy4RAmjeqGxER3PHS5cUzh6XNlh6KJDqKTHRIVK1b7NVU4Efx3+5N5WHUYuO8cICdHz7YiOsyRhcgxjIKT4otfPpOHd/YDv+sH/PznerYVIPqSYKc8rL8HcCMAP85ksEwxTTPqdnM6T1jd6DdyI1b8FLyekaJTlRjx8EchveG+y9GuNf8kU7SHSgWivLANcYOHVSXr9n6DSc9uDekXpwzG6N49lJdDtC9PtA88HJ72KbM8snDjmMLTb8jqa4KI9g7q4unTtV9z4vr8fd0neGxlcUjPujYDd4weJfWcBOFGlm7egZmv7g/pP93cEzcM1XNrG208rABgmubvTNPsb5rmYNM0b2VNVgkCsOan4PWMLFi1i6mtItpDpQJRXlgd0NXrBQBTc7cytSpE+/JE+8DD4WmfMstD1MPTb8jua0R7B3Xx9OnarzlxfcInq5E0QRBnCJ+sRtJuIL73BlBAwwtIF1QOVvwUvJ6R+df1Y2qriPZQqUCUF1YHdPV6AUB+zmCmVoVoX55oH3g4PO1TZnmIenj6Ddl9jWjvoC6ePl37NSeuz+IfZjA1QRBnePf/ejK1G3BkSbBVdF4SDLjTb+RGrPgpiktP4/HVezD3mj5o2TL22JLdx05g2os+vDXZi7Zt7ZvSjxSXYdbSrXj6xsFITU21fTwWvJ+5IbI8f060E6v3U6XX6Lev5+HFjcCfLwCuv965fkO051Smh5XHu6eL79AKbhxTeDz0sj3+oo8vKifALlu/LcDk531YPs2L9u3bO1aOhjjheRw1Ow9HAVwL4O+C24qoz+NWL6hby01E5o6/52HlQWBOL2DqVH3HFa08rFbRecLqRr9RvMO7H1xx6WkMX/BhSG+8/wpkpPFP/FTuSyeq7PHkYS0oqcDIR9aE9OezL0P79OihQir3y3vus4148O365akPXN8Jky8YLuVcLET74WTWeTd6wnloKmOK7P5RdN1W2T+wOHyiFJcs/E9If/ybS9G1bZrycujAzNw8LN1dr2/oC/wpW0xbEeXJ1cX7bBW3lpuIzN8++hiPv3cqpO++qjV+cfklDpYoOlp5WAlCJrz7wT2+eg9TqyoHD6LKHk8e1nkrdjB1Q1Tulxc+WY2kVSHaDyezzrvRE05ER3b/KLpu67JH9cw3tjB1UyJ8shpJ20GUJ1cX77NV3FpuIjLhk9VI2g3QhJWIO3j3g5t7TR+mVlUOHkSVPZ48rAuvH8DUDVG5X97Sn3ZialWI9sPJrPNu9IQT0ZHdP4qu2yr7BxbP3jyEqZsSDw5iazuI8uTq4n22ilvLTUTm5e+3Zmo3QEuCBeBGv1G8w7sPqygfqN1y8CCq7LJ8Wk60E6t7jKr0sAaXBS/9aScMH65+OXAQ0XsnyqzzbtxTlYemMqbI7h9F122V/QOLwydKMfONLXj25iFIS2uay4GDBJcFPzgIuPVW8rCKxK3lJiITXBb88vdb45JL9FwODNCSYEJzAgEThWXVYH2BUlNTh0/2FqGurq7R4yUnNcfVQ6yHrCQ2a4Zz2rYSNlELmCYqavzMzyWKytpa7CkoQ21tra3jBAImSiprlZRZNimJiRjTp33Mk5uEBA/6d0pT8jDa4WwgkehgIqskNmuG3pmprqjzxadPI2/rEZw+fTrmv7HSbxD8xNKHN8QfCKCwvBqBQEBKmRI8HnRITRLWnmtq6rDpYInjdan49GnsLyi31A5UwFMH7NKlS8vv/BSJx2MgM7WF7cmaqOO4CSfqAhHf0BtWmzSVgAyZxGLutxIGxBuOITrUReUG76KCOKqq/Og/772Q3jnvKiSdXQZnByfaSWVlLQb+IT+kt//2SiQn65ESu2Lbbsx4qd5n/OStffD9QX2Vl0P0NZJZ5w8WncLoP34c0uvuuQTd27GXNckKEZOJG8cUnoAW0SF3DRHdn+vSn2w+fBw/eLL+eeidGSMwtGsH5eVoiBMhPY+/9yH+9lH9pP0Xl7fE3VddIfWcTQU795MCm/SDQpcIQgCxmPuthAHxhmOIDnVRucG7qCCOj3YXMrWbWL7lKFM7SfhkNZJWhehrJLPO3/XaZqaORDyFiOkMT0CL6JC7hojuz3XpT7JzfUztFE6E9IRPViNpgh8795MCm/SDQpcIQgCxmPuthAHxhmOIDnVRucG7qCCO8QMzmdpN3DC8E1M7SX52H6ZWhehrJLPOL5o4lKkjEU8hYjrDE9AiOuSuIaL7c136k7wcL1M7hRMhPblXtmRqgh8795MCm/SDQpcUofOSYKDpBGTIJBZzf2VlLZZvOYobhjfuTeUN8zhYdAp3vbYZiyYORevW9hv0gcISzFi8CS9PGob09HTbx2PhO3QM2bk+rMzxomPHjtzHsXKdreBEO9l5tAhTX/Dh7SletGvXrtHfVxkycf+reXh1M7DwPOAnP3Gu3xAd4CKzzm88eBTZuT68m+NFp06xTRjcGNTkxjGFJ6zt2MlyzFm2DU9MGCTFy110qhILVu3C/Ov6ITk52fbxVIbosVi1Yy/ueGEX8qf0Q+/evaWey0qfKGvsYDF8dh6KAYwF8JymbcWt4UV2Qsbc+pnjmZ/8Pg9fnAampQH3369nWwGiLwmmCatN3Og3ciNWfKm8XiPRniqVHtaiU5UY8fBHIb3hvsvRrrX1BzRZ3hMn2olVz6NK382LGzbht28eDuk/TOiKSSOGSTkXCzfVeZ46Th5WNfDkBsi+N6L9+LzZCKI5UlyGix5bF9LrZ41G54xUKeey0ifKyj9gcc+iPLy5s15P6A/88Ta92opb/ZxuLTcRmSc+WIc/v18W0r8en4q7xo52sETRIQ8r4Wqs+FJ5vUaiPVUqPawLVu1i6liJJ++JVc+jys8ePlmNpFXhpjrPU8fJw6oGntwA2fdGtB+fNxtBNLOWbmVqkVjpE53IPwifrEbSOuDWMdWt5SYiEz5ZjaTdAE1YCVdgxZfK6zUS7alS6WGdf10/po6VePKeWPU8qvzsy37WlalV4aY6z1PHycOqBp7cANn3RrQfnzcbQTRP3ziYqUVipU90Iv/goSFsrQNuHVPdWm4iMi9em8rUboCWBAvAjX4jN2LFT8HrpSkuPY3HV+/B3Gv6oGVL+wEOojeuZyHKryXLe+JEO7HqSVbpuwkuC172s64YNkz9cuAgbqrzPHWcx1vpNG4cU3j8brLvTVWVHx/tLsT4gZlISLC/PNWOp08kR4rLMGvpVjx942Ckpsp98LTSJ4q+3rEQXBb80BDgllv0bCtu9XO6tdxEZILLgl+8NhWjR+u5HBigJcFEHODxGGibEtvm2zV1ddhbWGZ5g3ePYSA5sZmwzrn49Gn8Z0+Bkg3eA6aJqto62xt119TUYcPXxZavnY6kJCZiYKfWMYftqNzgvbq2FgBQe/anUyQlJGBol3RhD5j+QAAnKmoQCASEHC+cBI8HbVslOjpZIMQRCJgoqay13WdFw18XwLcnK6XURSepqatDSWWNkj7aSp/o8RhIT26udHJTU/PdnzoSCJg4UV4trZ43PFdhmZpzEe7i8OGy7/x0G/SG1SZuDMhwI1YCAHiDX0QHxuwvOIkr/rQ+pD+ceRF6tpezJLGgpAIjH1kT0p/Pvgzt01tZPo6s4Awn2onoQCGR/OuTDXho2fGQnvPDDpg66n++UJQOb0BZNESFf0WCp31S6JIaeAKJZIf0lFfUYPCD74f01gfGI6UV/7JGXUKXDhSW4LLH/xvSa+6+GFmZ6crL0RAn2tovn8nDO/vr9Q96An+dpldbUVlvRAYlUehSfPGbF/Lwxo56/ZMBwML/06utBKE3rISrsRIAwBv8Ijow5o5XfEwtknkrdjB1rDgRnCEL0YFCIgmfrEbSquANKIuGqPCvSPC0TwpdUgNPIJHsvmbJhkNMbRVdQpdmLN7E1E7hRFsLn6xG0jqgst6IDEqi0KX44o0dbO0GaMJKuAIrAQC8wS+iA2MW3+ZlapEsvH4AU8eKE8EZshAdKCSSNyd0YGpV8AaURUNU+FckeNonhS6pgSeQSHZfM2lUN6a2ii6hSy9PGsbUTuFEW/tdP7bWAZX1RmRQEoUuxRcLhrK1G6AlwQJwY0CGGymvqMGSDYcwaVS3Rj2JvKEUu4+dwLQXfXhrshdt27a1W2RsPHgU2bk+vJvjRadO9iYDjfHp14eRnbsJ+TnD0LUrf+qsletsBSfayc6jRZj6gg9vT/GiXbt2Ss4ZK3c+nYe8b4C5vYEpU5zrN0TX+XV7v8HU3K3IzxmMHj16CChhPVu/LcDk531YPs2L9u3bx/Q3suqzTNw4pvAE3e0vOIk7XvFh8W1etGkjfoIjKoguCG+Yn2g+3P01cnK3Iz9nIM455xzHytEQJ66PG9qKyuA3kcFgdo5FgU360X92HqoAjAKwWNO2AkRfEkwTVpu40W/kRqx47Xh9eaJ8oEFEe2JZiPLLyvLbONFORN9PkSz6wod5b30b0vN+3AW3ne9VXo7DJ0pxycL/hPTHv7kUXdumcR9v+5FCXPvXz0N65S9HYmBnMW/OeMpKHlY18FznI8VluOixdSG9ftZodM4Ql3gr2iOrS11SmY1gBSc8vt+bnYdtYXoQgDzN2opKL6guHlbyv+rHH5bmIbd+aEb2SOC3N+jVVoKQh5VwNVa8dry+PFE+0CCiPbEsRPlldfFpiUD0/RRJ+GQ1klbFzDe2MLVVpizyMbUdeMpKHlY18FznWUu3MrVdRHtkdalLKrMRrODE2LGtEa0DKr2gunhYyf+qH+GT1UjaDdCElXAFVrx2vL48UT7QIKI9sSxE+WV18WmJQPT9FMm/b+rC1Kp49uYhTG2VZVO9TG0HnrKSh1UNPNf56RsHM7VdRHtkdalLKrMRrODE2DExga11QKUXVBcPK/lf9eNvF7O1G6AlwQJwg4ciHigtr0bu+gOYPiYLLVq0YP5ucelpPL56D+Ze0wctW8a+lcn2I4WYssiHZVO9yMy0v5Tx8IlSzHxjC569eQjS0viXWsbC+v2HMDl3M/JzhqJbN/6AEd5r1xhOtBOrHjmVvpt49bCu3rUftz+/A/nTBqBnz54CSljP5sPHkZ3rQ16OFx06xBZUVVJWhafW7sPd43ohKUnOknzRuHFM4ek3RPe3Ddl7vBi3v+TD69leZGRk2D5eVZUfH+0uxPiBmcL2LeZhzZ4DmJa7Dfk5g5CVleVYORqi0qsZRGZbEeUHFekrbQyRY5id+0keVv0YMjsPZQBGA3hR43GFPKyScKPfyI1Y8cfwepdE7yEpeh9AFsdOluPCR9eG9Kf3jkHHNta/4S4tr8bQ+atDevPccUhLYX85EAtOtBOr11+l7yZe92GV6Uvk2VdX9OdTgRvHFJ6+TrbHX7Q/Wxdf3sGiUxj9x49Det09l6B7u9bKy9EQJ67PlbPzsDtM9wWQL6itiPLk6lJvrKLLvsOEGO5/NQ+vbq7XNw8FFtys57hCHlbC1Vjxx/B6l0TvISl6H0AWc5ZtY+pYyV1/gKndhNXrr9J3E6/7sMr0JfLsqyv68xGR4enrZHv8RfuzdfHl3fXaZqZ2Cieuz+5GtB1EeXJ1qTdWiac8CwLfmaxG0m6AJqyEK7Dij+H1LoneQ1L0PoAsnpgwiKljZfqYLKZ2E1avv0rfTbzuwyrTl8izr67oz0dEhqevk+3xF+3P1sWXt2jiUKZ2Cieuz42NaDuI8uTqUm+sEk95FgSw8Dy2dgO0JFgAbvQbuRErPhBer5HofftU7gF57GQ55izbhicmDEJKCv/gYsUrbAUn2onV66/SdxNcFvzmhA4YMUL9cuAgovdO5N0DORZ4fJK67J1pBTeOKTx9nWx/segMAV18eQeLTuGu1zZj0cShaN3a+eXAQZy4PsFlwTcCeFRTD6su9cYqKr23hHyCy4IXngf85Cf6jiu0JJhwPYGAiZMVNYjlS5ZAwETp6dqYfvc7f2eaqKqts/x30SivqcEnXxehpkb+MqDS6mrsKyhHdXW1reN4DAMpLRJcNbBGo6SqCvk7jqGqqiqm3w8ETJworxZ2/1nU1tZ+56dTlNfUYO2eAmF1NGCaqPEHpFzDgGnitMX2mZjYDMO6pysLgWmqJLdsjh8O72rpS4EEjwcdUpOkPQwnJSSga5tkYQFJfn8Ae4+XIRAICDkeLwdPncKOI6U4deqUo+VoiMr+M0hwGbCMFY4JCR7075Rmu36qrDd+fwA7j5YKOZedzy+yHIQYgt2FZt1GzNAbVpu4MSDDjVjZtJ03aKWgpAIjH1kT0p/Pvgzt01txl1lUEFIs7D52Alf+5dOQzv/Vhejb0Xriq6ygKCfaidXQFZUhE39f9wkeW1kc0rOuzcAdo0dJORcL0XVedNBNODyhaG4MDnHjmMITLCM7lE50iJ6VMUgmH+87iIn/qvfjvjx1CC7p1V15ORriRFtzQ1tRWW906e90KQdRT3BXgiDf6wE8NV2vthKE3rASrsbKpu28QSvzVuxgaquICkKKhWkv+pg6VlQGRcnGauiKypCJ8MlqJK0K0XVedNBNODyhaBQcogaeYBnZfY3oED0rY5BMpuRuYWqnoLYWGZX1Rpd7oEs5iHrCJ6uRtBugCSvhCqxs2s4btLLw+gFMbRVRQUix8NZkL1PHisqgKNlYDV1RGTKx+IcZTK0K0XVedNBNODyhaBQcogaeYBnZfY3oED0rY5BM8nOGMLVTONHWGvZW9novOaisN7r0d7qUg6hnbm+2dgO0JFgAbgzIcCNWgpR4wzZ8h44hO9eHlTledOzY0W6RsfnwcWTn+pCX40WHDnKTYFdu34M7X9yN/Ml90adP4wmq0SgoqcC8FTuw8PoBaNWKf3loQ5xoJ+v2foOpuVuRnzMYPXr0aPT3VYZM3PJwHv57CpjRDrjnHuf6jZ1HizD1BR/enuJFu3btbB9PdBsKZ8M3R5Cd68OqHC86d+4c099Q6JIaeK6z6FCkhhwoLMGMxZvw8qRhSE9Pt308XerSk2v+iz+uKsEr16Xj4osvdqwcDZEV2MfCDW1FdrhYODU1dfjy4MkzE2Wbvn07YVFWxlK3hlK5jWBb6QTgE03bChB9STBNWG3iBg9FPGDFI1VcehrDF3wY0hvvvwIZaY0nih4pLsNFj60L6fWzRqNzBn/KqWh/IIu9x4sx7s+fhPTqX49C7w7W39rx+n8bw4l2sr/gJK740/qQ/nDmRejZ3pm3Ig0JPnAGuefqdMy4TP2DJ29biYZo32A4BwpLcNnj/w3pNXdfjKzMdObf6OI7tIIbx5SqKj/6z3svpHfOuwpJSewvFUvKquB96IOQ9s0Zi/RUcQ/zoj2yuvjy1u8/hJufqY8YenXaUFzU0/mVMLLGDhZuaCsqrwuPl1zFsXQ4T1Nnxj/ysOJAvf5+FvDk7Xq1lSDkYSVcjRWP1OOr9zB1NGYt3crUVhHtD2Rx+0s+po4VXv+vjtzxio+pnSR8shpJq4K3rURDtG8wnBmLNzF1JHTxHcY7H+0uZOpIPLV2H1PbRbRHVhdf3uTczUztFPE0dohE5XXh8ZKrOJYO52nqhE9WI2k3QBNWwhVY8UjNvaYPU0fj6RsHM7VVRPsDWbye7WXqWOH1/+rI4tu8TO0kr1yXztSq4G0r0RDtGwzn5UnDmDoSuvgO453xAzOZOhJ3j+vF1HYR7ZHVxZeXnzOUqZ3CibGj4ft4uYtt+VB5XXi85CqOpcN5mjoP9GFrN0BLggXgBg+F7sTiYXCjh3XjwaPIzvXh3RwvOnWSO4Bb9WtGo+hUJRas2oX51/VDcrKY5ZyAM+3Eqq9XpZfmxgfz8FkFcHs6MHu2c/3GwaJTuOu1zVg0cShat25t+3ii6mEk/r1lJ371yj6suq0X+vfvH9PfqPSPicKNYwqPf3Hv8WLc/pIPr2d7kZEhPnhs97ETmPaiD29N9qJtW+vbfDWkKXpYrfSJ5GGNjKwxNRIicxjs+GGt1BvysKoh2FZSAGzVtK0A5GGVhhs8FLoTi4fBjfuwytyTsiGi/Lfx5GHdebQIVz/xWUivuusC9O8UPVRIpZfmrx/+B3/KLw3pmVem4ZdXXCrlXCxE+/y2HynEtX/9PKRX/nIkBnZu/G1bLPgOHcOPnvoypP9953nwdmN/qeSEr84ubhxTeLzCMv3OgPhMAl380Cr3YbXSJ9I+rJFxq4fVzv0kX6p+THosD+vCds8bnQG8OEuvthKEPKyEtsTiYXDjPqwy96RsiCj/bTz5kKa+4GPqhqj00oRPViNpVYj2+U1Z5GNqO2Tn+pg6EvFUn3WGxyss0+8MiM8k0MUPrXIfVit9oi4eX91wq4fVzv0kX6p+rCtmazdAE1bCcWLxMLhxH1aZe1I2RJT/Np48rG9P8TJ1Q1R6aV76XhpTq0K0z2/ZVC9T22FljpepIxFP9VlneLzCMv3OgPhMAl380Cr3YbXSJ+ri8dUNt3pY7dxP8qXqx10d2NoN0JJgAbjBQ6E7sXgYrPgpeL1Govcglb3PYDhHisswa+lWPH3jYKSm8i99k+XTcqKdWN1jVKWXJrgs+KXvpeHSS9UvBw5SXlGDJRsOYdKobkhMtP9gsf1IIaYs8mHZVC8yM8UsBw7C4zHXxXdoBTeOKTx+N9nePlF9YhCRe1za4eN9BzEldwvyc4age3c5y4GDWOkTVe5jHcQNbUVlHyRyDLNzP8mXqh/BZcF3dQB+/Ws92wpAS4IJzfF4DGSmtmB2bB6PgTatYuv8EhI86N422XIn6zEMJDVvJqyD9RgGEhM8SjrsBI8HaUnNbT8oeDwG0lo2j4tBJiUxEZ3SWwqZiImmpqbmOz+dorK2FhsPnURtba2Q46UkJqJD6yQp17x9q1YY2KW1pS+TePsCwhpW+ucgic2aoXubZGmTv5TERIzoniGsLvJ8Rhm0b9UKHVsnCflSVSR+fwBfF1YgEAg4XRQhBAImCsuqYffFjsoxNZZnqVhJSPCgf6c0rr5TZDkIMQRz5wTkzzkCvWG1iRtM//GAlQAA3nAM0SEgokM/WIgKjKqq8qP/vPdCeue8q5CUxE5kjgUn2onV0CuVQRF/XPUBnlxTFdIzLkvCPVePlXIuFqKDxg4WncLoP34c0uvuuQTd29lPHgb42pMuQTlWcOOYwhPQUlpejaHzV4f05rnjkJYiLllW9PGdCBWKxO5jJ3DlXz4N6fxfXYi+HeU8gVrpE2WNHSxkthVR44ET14UgGvLr5/Lwdtg269f3Af48Wc9xhd6wEq7GSgAAbziG6BAQ0aEfLEQFRn20u5Cp3YTV0CuVQRHhk9VIWhWig8buem0zU9uBpz3pEpQT7/AEtOSuP8DUdhF9fF1Chaa96GNqkVjpE+Np7ADEjQfxdl0IdxI+WY2k3QBNWAlXYCUAgDccQ3QIiOjQDxaiAqPGD8xkajdhNfRKZVDEc+OTmFoVooPGFk0cytR24GlPugTlxDs8AS3Tx2QxtV1EH1+XUKG3JnuZWiRW+sR4GjsAceNBvF0Xwp38YSBbuwFaEiwAN5j+4wEr4TC8YR6iN5pXGY6xZs8BTMvdhvycQcjKyuI+TnHpaTy+eg/mXtMHLVu2FFY+J9rJO1t34Zcv78V7P++Nfv0a/wJCZajKsNl5OAXgMgCLHOw39hecxB2v+LD4Ni/atLE/obN6za2w8eBRZOf68G6OF506xZa2KTp4RwVuHFN4gmW2fluAyc/7sHyaF+3btxdeJtH9ueiAMl6eWf8FFrxTgDd+3B7nn3++1HNZCc+RNXawkNlWRAUH6VJvrGLn8+sSUEbUE2wrqQC2aDyuRFsSTBNWm7jRb+RGrPhAeL1LxaWnMXzBhyG98f4rkJHGP+juLziJK/60PqQ/nHkReraX84Zn+5FCXPvXz0N65S9HYmBn69/kytrk3Il2svHgUdzw940hvfSO4RjePfokR6VHbWZuHpburtc39AX+lK2+3yivqMHgB98P6a0PjEdKK/4HKt+hY/jRU1+G9L/vPA/ebrGl+TbGsZPluPDRtSH96b1j0LEN+y2XaF+6Ctw4pvB4hWX6nQG++sJCFy/i6l37MeX5+qX7z/58AMb166m8HA2RNXawcENb0cX7bBU7Hl43ZgfEOzcvyMP6sO3eL0oDXr1fr7YShDyshKux4gPh9S49vnoPU1vljld8TC2SKYt8TB0rKjc5l012ro+pG6LSoxY+WY2kVbFkwyGmtorVa26FOcu2MXUkRPvSicjweIVl+p0BvvrCQhcv4u3P72Bqp4insUMkunifrWLHw0vZAfoRPlmNpN0ATVgJV2DFB8LrXZp7TR+mtsri27xMLZJlU71MHSsqNzmXzbs5XqZuiEqP2oOD2FoVk0Z1Y2qrrGxwjRtqOzwxYRBTR0K0L52IDI9XWKbfGeCrLyx08SLmTxvA1E4RT2OHSHTxPlvFjoeXsgP04862bO0GaEmwANzoN3IjVVV+fLS7EOMHZiIhgb0Uq7S8GrnrD2D6mCy0aBH7VgaiPTii/YEsth8pxJRFPiyb6kVmJv/DlKxNzp1oJ1Y9jyo3vg8uC35wEHDrrc71G6L9Vb5Dx5Cd68PKHC86dhSzHDjIsZPlmLNsG56YMAgpKbE9+PH62Z3EjWMKj2ftYNEp3PXaZiyaOBStW4tbDhyEp76wsDIGyWT1rv24/fkdyJ82AD17Or8cOIissYOFG9qKynFFJORhjS+Cy4LvbAv85jd6thWAlgQTcUBCggfnZLaKqcMvr6nBhoPFqKmxFkWf4PEgM6WFsEElweNBy8QEJYNUYrNmSE5sRoNDGN1bt8YVAzrG/DBsZ6N0qwS/R7HwfYoUKmtrsfHQSdTW1go5XlqLFmif1sLSF0WxUnz6NHYcKcXp06dj/pvk5s1xQVZbZQ/QTRWPx0CbVtYebNNatMCIHhlS6gpwZhz4pqjC8jgQDX9dAN+erEQgEBByPF4CpgkTgG4vHBISPOjeNtlVk7J4IxAwUVhWLaRuBAImTpTzHSsxsRlG9W4X0/OIyDIT0QnmzgnIn3MEesNqEzeY/uMBK8EFR4rLcNFj60J6/azR6JzReDqo6I3mZYeKhLP3eDHG/fmTkF7961Ho3SHD8nHiKXRJ58Cd+17Jw+KwbWF/NgR4+Bb1/UZBSQVGPrImpD+ffRnap7fiPp7MoDGeYDEngmDs4sYxhSdYpqSsCt6HPghp35yxSE8Vt72TqD4xiOiAMl7yd+7DtEU7Q/qZ2/rjyv69lJejIU4E7bihragMXbITlNQQVeUWWWYiOr/4Zx6Wf12vrzsH+FuOXm0lCL1hJVyNleCCWUu3MnU0RG80LztUJJzbX/IxdazEU3CGzoE74ZPVSFoV81bsYGqryAwa4wkWi6f6rDM8wTJPrd3H1HYR1ScGER1Qxsv0sMlqJO0UFLQTGZWhS3aCkhqiqtwiy0xEJ3yyGkm7AZqwEq7ASnDB0zcOZupoiN5oXnaoSDivZ3uZOlbiKThD58Cdx4aztSoWXj+Aqa0iM2iMJ1gsnuqzzvAEy9w9rhdT20VUnxhEdEAZL/lT+zO1U1DQTmRUhi7ZCUpqiKpyiywzEZ3f9mVrN0BLggXgBtO/7sRi7rcSDnOgsAQzFm/Cy5OGIT09PeZy7D1ejNtf8uH1bC8yMviXjwX59OvDyM7dhPycYejatavt47FYsW03Zry0B/nZfdC3L39vVFBSgXkrdmDh9QPQqhX/8tCGONFOrAZRidooPhYGzs5DJYCLAbziYL8hKqwryJKvtmL2a99gxcQeGDw4ti+LYmXVjr2444VdyJ/SD717947pb0SHSqnAjWMKT2Dd1m8LMPl5H5ZP86J9+/bCyySqTwyiS4DX3CV5eNkHPH4+8OMf61M/Ssqq8NTafbh7XC8kJYlb3s3CDW1FdJgjC5FhR6qCk1SOu02ZYFvpCOBTTdsKEH1JME1YbeIGD4XuxOJhsLJpO6+XRrSfj9dLy8POo0W4+onPQnrVXRegf6d2lo8jy6flRDuxej9VemnmvZmHRWFd2m0jgHkT1Pcbon2+nx/4Fj/9hy+kX7/di5FZXewUMQSPJ9yN/ig3jik8/v/DJ0pxycL/hPTHv7kUXdumCSuTqD4xiJUxSCaLvvBh3lvfhvS8H3fBbed7lZejIU74xd3QVkRnY7AQ6Tt1Y99JRGf6U3l4N8zFcE034Ok79WorQcjDSmhLLB4GK5u283ppRPv5eL20PEx9wcfUsaKLT0sEVu+nSi/Nog1srQrRPt/sXB9T24HHE07+KDXw+P9nvrGFqe0iqk8MYmUMkkn4ZDWSdgryi0dGdDYGC5G+U+o744t3D7G1G6AJK+E4sXgYrGzazuulEe3n4/XS8vD2FC9Tx4ouPi0RWL2fKr00T4xia1WI9vm+l+NlajvweMLJH6UGHv//szcPYWq7iOoTg1gZg2Ty75u6MLVTkF88MqKzMViI9J1S3xlf3N+Trd0ALQkWgBs8FLoTi4fBin+Id8P4nUeLMPUFH96e4kW7dvzLx4J8vO8gpuRuQX7OEHTv3t328Vi87tuGWUsOIO/WLAwaNIj7OLJ8Wk60kzV7DmBa7jbk5wxCVlZWo7+vcoP34bPzUAxgLIDnHOw3NnxzBNm5PqzK8aJz5862j/fM+i+w4J0CvPHj9jj//PMFlLAeHk84eVjVwNNv7D52AtNe9OGtyV60lbA54Pr9hzA5dzPyc4aiWzf7X7454dGMxP2v5uHVzcDC84Cf/ESf+kEe1sio9LCK7O9UjoeEfNzQVgDysErDDR6KeMCKt7K49DSGL/gwpDfefwUy0hofJHj/Lhoy96RsyObDx/GDJ+vbyDszRmBo1w6WjyPLp+VEO7G6b6dKz87v38rD81/U65+fD/zOgfCUA4UluOzx/4b0mrsvRlZmOvfxNnxzBBOe/iqk35x+Lkb0sD8JBvg84bQPqxp4cgNk3xvR/bkuHtaV2/fgjhd3h/TfJ/XFtQP7KC9HQ2gf1sio3IdVZB0lD2t88ZPf5+GL0/X6/JbAG7/Tq60EIQ8r4WqseCsfX72HqUX/XTRk7knZEFHeQV18WiKwum+nSs9O+GQ1klbFjMWbmNoqMj2sPJ5w8tWpgSc3QPa9Ed2f69I33hk2WY2knYL2YY2Myn1YRdZR8rDGF+GT1UjaDdCElXAFVryVc6/pw9Si/y4aMvekbEheA69gQx0ruvi0RGB1306Vnp2/XsTWqnh50jCmtsqqBvWuobYDjyecfHVq4MkNkH1vRPfnuvSN+ZP7MrVT0D6skVG5D6vIOkoe1vhiWhpbuwFaEiwAt6wLdztWvBm8nhHRXpP9BSdxxys+LL7NizZt5A7gmw8fR3auD3k5XnToYH05cJCqKj8+2l2I8QMzkZAgbskb7cP6XYLLgv96EfCDHzjXb/DuWRwN0Z7YcI4Ul2HW0q14+sbBSE2NbYuoyspaLN9yFDcM74TmzfVeDhzEjWMKz56Nsu+N6P5cVt9olZXb9+DOF3cjf3Jf9Onj/HLgIKr27QzHDW1FpRdUZB2l/VHji+Cy4GlpwP3369lWAFoSTMQBCc086NImOaYOPyUpEd8f2tly6EBlbS32FJShtraWt5jfIbl5c3Rpk6zkQTnB40FigofCEcKorK1FeXWtsPspkmCRNCyaLar8ftT4Tfj9fuHHLqiowPZvT6GiokL4sQn1lFRVYfmWb1FVVSXl+KXV1dh0+CSqq6uFHC8hwYNzMls53scmeDzwAI6XoyF+fwAHT1QiEAg4XRSt8PsD+LqwQsl18XgMpCc3FzLBDARMnCivBs+LrUDARGEZ398ScsjI+O5Pt0FvWG3iBtN/PGAlTIA3/EH0RvbHTpbjwkfXhvSn945BxzZylgNZDRiKRjyFLlkNAFIZMjF3SR5e9tXriV5g/k3uD136eN9BTPxX/X6aL08dgkt6iUnH9h06hh899WVI//vO8+Dt1pH5NxS6pAaePpcnRMsKouu2yvAcFvk792Haop0h/cxt/XFl/17Ky9EQJ9qaG9qKyrAukcFXduo7BTbpR87f8vBe2JbNV3UB/vkLvdpKEHrDSrgaK2ECvOEPojeyn7NsG1OLxGrAUDR0CRYRgdUAIJUhE+GT1UhaFaJDl6bkbmFqO/AEOlHokhp4+lyeEC0riK7bKsNzWEwPm6xG0k5BbS0yKsdUkcFXduo7BTbpR/hkNZJ2AzRhJVyBlTAB3vAH0RvZPzFhEFOLxGrAUDR0CRYRgdUAIJUhE4+fz9aqEB26lJ8zhKntsLLB/WuoI0GhS2rg6XN5QrSsILpuqwzPYZE/tT9TOwW1tcioHFNFBl/Zqe8U2KQfs7PY2g3QkmABuMH0Hw9YCV3i3cBc9Ebza/YcwLTcbcjPGYSsrCzbx2Ox5KutmP3aN1gxsQcGD+Z/+DtYdAp3vbYZiyYORevWrYWVz4l28txnG/Hg20ex9KedMHz48EZ/X2XIRO/ZefADOB/AGw72GxsPHkV2rg/v5njRqZP9h8x//vdzPLy8EK/fkImRI0cKKGE972zdhV++vBfv/bw3+vXrF9Pf8PYFTuLGMYUnQEl2eJDoADBd6tIfluYh93PgbxcD112nT/1w4vrIbCuiwpKsPLvYRWSQmZ0QLQps0o9gW8kE8IXG40q0JcGNTlgNw5gZ4T+fAvClaZo+McVjo/OE1Q0einjAip+C10tzsOgURv/x45Bed88l6N6Of8K282gRrn7is5BeddcF6N+pHffxWGw+fBw/eLK+jbwzYwSGdrWeFFxQUoGRj6wJ6c9nX4b26a1sl8+JdsLjeVTFvS/l4bWwFeI3DgIevVV9v1FcehrDF3wY0hvvvwIZafxpqjKvOU/7FOnpUoUbxxQev5vs9ll0qhIjHv4opDfcdznatU7mPl55RQ0GP/h+SG99YDxSWql/c/Tmpu24Z/HXIf3Hn52DCcMGKi9HQ5xoazLbiijPskrvs8hzkQ81vpjxjzysOFCvv58FPHm7nuOKHQ/rCAC3A+hy9t80AJcB+JdhGLNEFpIgomHFT8Hrpbnrtc1MbZWpL/iYWiQ8/r5IzFuxg6ndhKhrIoPXtrG1Kh5fvYeprSLzmvO0T5GeLiI6PH432e1zwapdTG2VJRsOMbUqwierkbRTxFtbE+VZVul9Fnku8qHGF+GT1UjaDcQyYW0LYLhpmnebpnk3zkxgMwGMBnCbxLIRRAgrfgpeL82iiUOZ2ipvT/EytUjyGvj5GupYWXj9AKZ2EzyeR1U84mVrVcy9pg9TW0XmNedpnyI9XUR0ePxustvn/Ov6MbVVJo3qxtSqWDnpHKZ2inhra6I8yyq9zyLPRT7U+OKBPmztBmJZErwDwDDTNGvO6hYAfKZpDjAM4yvTNM+VXUidlwQD7vQbuRErfhJeH4do/+bOo0WY+oIPb0/xol07OcuBg2w+fBzZuT7k5XjRoYP15cBBCkoqMG/FDiy8fgBatbK/HDiIE+3Ed+gYsnN9WJnjRceOeiwHDhJcFvyIF7jJgS1tghSXnsbjq/dg7jV90LIl/3LgIDKvOU/7tOPDcgo3jik8fj/Z7bPoVCUWrNqF+df1Q3Iy/3LgICq9iCyCy4JXTjoHAwc6vxw4iBNtzQ0eVlHHUX0u8qHGF8FlwQ/0ASZP1ndcsbMk+FUAnxqG8TvDMH4H4L8AFhuG0QrAdsHlJIioBAImTlbUxLQRdWl1NfJ3HBW2YTwvBRUVKCitRkVFhfRzJXg8SEzw2B6kkhIS0Ld9KhIS5OwVpxJ/IAB/IKDlRvZduqR956dT+AMBnKys0fIaNaSkqgqHTlSiqqoq5r+pqa3Dtm9Poa6uTmLJCB46pqTg3B5tkJIi561TSVUVNh86aam+sAiYJsqr/TGNQTJJOts369ZHV1bX4sOdx1FbW+t0UYSQkOBB/05ptsfUqmo/Pt5dCL/fL6hkas5VXlmDVz45gJoaucuB/f4Adh4tdcUY5GZOnfruT7cRU0qwYRgjAFwMwADwsWmaSl936vyG1Y0BGW7ESqDDsZPluPDRtSH96b1j0LFN4w9EokOX1u39BpOerd9b8MUpgzG6dw/u47HYfqQQ1/7185Be+cuRGNjZenx+aXk1hs5fHdKb545DWkoL2+Vzop1s+OYIJjz9VUi/Of1cjOhhPylUBH/98D/4U35pSM+8Mg2/vOJS5eUQHbIlM0iHJ1hMl6AcK7hxTOEJe+Htp2Nl7/FijPvzJyG9+tej0LtDBvfxZPWNVlmxbTdmvFTvNX/y1j74/qC+ysvRkJKyKngf+iCkfXPGIj1VblKwG9qKyj5I5LlU1XeVoVRNmVsfzcN/wmzll7YBXrpXr7YSxM4bVgD4CsAbAJYCKDAMo7vIwhFEY1gJdJizbBtTR0N46FLuVqYWyZRFPqaOldz1B5jaTegcuhQ+WY2kVSE6ZEvmNec5ti5BOfEOT9gLbz8dK7e/5GNqq+jSN4ZPViNpp3hq7T6mbqqo7INEnktVfVcZStWU+c9JtnYDjU5YDcP4BYDjAN4HsAJA3tmfBKEMK4EOT0wYxNTREB26lJ8zmKlFsmyql6ljZfqYLKZ2E6sahLg01E7y0vfSmFoVokO2ZAbp8ASL6RKUE+/whL3w9tOx8nq2l6mtokvfmJ/dh6md4u5xvZi6qaKyDxJ5LlX1XWUoVVPml+3Z2g3E8ob1LgD9TNMcZJrmUNM0h5imae9JPo5ouPxEt+UobiEQMFFYVh3VG9SsWTN8cd8VuGVkN2z97RXMQIeUlBR8OPMi9O+QjK/uvShmf1Tr1q2x9I7hSE/y4JO7htsOXerRowd+e8OZJahv39gZPXrIWQ4MAJmZmfj7pL4wALw/uS8yM60vBwaAFi1aYMN9l+OGcztj+28vR4sWYpYAOdFOOnfujPk/OTNgv3NzN3TurMdyYAC49NJLMeTsFzCTEs9oJ2jVqhVW/OJ8dEhpjs9nnm87ZKtjx4547KYsAEDerVlCg3Q6dOiAJ28983Cen90npmCxxMTEM0tB27XExlmjHA3KiRU3jikejwdrZ47GLSO74fN7R8fk+UtJScGSnGFIbg58PGOYcB9rRkYGXp02FC0TgP/cORQZGfzLgYEzfeP6WaNxSe8MbJkzWljfaJW+ffti2kVnru/TYzzo29f55cAAkJSUhE/vHYOx/dth69wxSEqSuxwYkN9WRHgrExMTv3NdZPZBiYmJ8M0Zi6mXZGHnvLG2ztWiRQtsnjsOv7qiN3b9fpy0+u7xePDF7LF4+ubh+Gru2Kh9R2PPiASbmTPr20bPBtotxJIS/BGA8aZpyneLR4E8rPFNLBtUW/FmVFbWYuAf8kN6+2+vRHJy40nBojeaF+2hYnGgsASXPf7fkF5z98XIyky3fBxZm7870U52Hi3C1U98FtKr7roA/TvJTWqOlcdWvo+/r6sPsrhjdCJmXTteeTlEe1h3HzuBK//yaUjn/+pC9O3Y1k4RQ/B4zGX7JGXgxjGFx+8mqs+Khui6rYsfOn/nPkxbtDOkn7mtP67s7/zbzKoqP/rPey+kd867CklJckOhZLYVUd5KlfXGjX7QWMocyzMiwebBt/PwXP3jECZfADxwvZ7jih0P634AawzDuM8wjJnBf+KLSDRVYtmg2oo3Y/mWo0wdDdEbzYv2ULGYsXgTU8dKPG3+PvUFH1M7SfhkNZJWhWgP67QXfUxtBx6PuWyfJHEGHr+bqD4rGqLrti5+6Olhk9VI2ik+2l3I1G5DlLdSZb1xox80ljLH8oxIsAmfrEbSbiCWCetBnPGvJgJIDfvHjWEY6YZhvGkYxk7DMHYYhjHKzvEIdxPLBtVWvBk3DO/E1NEQvdG8aA8Vi5cnDWPqWImnzd/fnuJlaif519hEplaFaA/rW5O9TG0HHo+5bJ8kcQYev5uoPisaouu2Ln7o/Kn9mdopxg/MZGq3IcpbqbLeuNEPGkuZY3lGJNg8eQlbu4GYtrURflLDeAHAf0zTfNYwjEQAyaZplkT7fZ2XBAPu3ORdN2LZ7LqkrApPrd2Hu8f1atQfc6S4DLOWbsXTNw5Gamrs36/sLziJO17xYfFtXrRpY3+y9vmBb5Gd68N7OV506dLF9vFYrNv7DabmbkV+zmBbftnS8mrkrj+A6WOyhPpWnGgnH+87iCm5W5CfMwTdu+sVbj5idh6KAIwH8C8H+w3RdXTJV1sx+7VvsGJiDwweLDZobOX2Pbjzxd3In9wXffrEFjaz/UghpizyYdlUL7e3WzVuHFOs9M9BZPePovrEIEWnKrFg1S7Mv64fkpP57SJ2+VP+R/jrh5V4/qpkXH755Y6VoyGyxg4WMttKLM8lsVBQUoF5K3Zg4fUDbOcENEZNTR2+PHjyzJfPjKyPWAgETJyoqEG7lEQYhrUluFb+NpYy2ykLcYZzZ+fhJIArAORqPK5EWxIcdcJqGMZfTNP8lWEYywH8zy+ZpvkDzoKkAdgEoKcZ42xZ5wmrG/1GuhGLP8GKN4N3PzjR+44Vl57G8AUfhvTG+69ARlpL7uOxELWHrCwfkhPtRLQnWST3vZKHxVvq9c+GAA/for7fEO0j3HjwKG74+8aQXnrHcAzvHtsKh8bY+m0Bvv+3L0J6xS/Ox+Au7KhDlW1QFG4cU3g8ZqI9pg0R7WF3wqMZiY/3HcTEf9V3Hi9PHYJLejn/ZZys/AMWbmgrvJkaPIj0eto5lpW/JX+qGh54LQ8v1W9Lj1vPBR68Ua+2EoTHw/rS2Z9/BPB4hH+89ARQCOB5wzC+MgzjWcMw/meUMgxjmmEYGwzD2FBY6G4vBMEmFn+CFW8G735wovcde3z1HqYWiag9ZOPJhyTakyyS8MlqJK0K0T5CmfuwTn7ex9SRUNkGmzI8HjPRHtOGiPaw69I3TsndwtROEU/5ByLhzdTgQaTX086xrPwt+VPVED5ZjaTdQNQJq2maX579n17TNNeG/wPgtXHOBADDATxtmua5ACoAzI5w/mdM0xxhmuYItyzjIviIxZ9gxZvBux+c6H3H5l7Th6lFImoP2XjyIYn2JIvkseFsrQrRPsJ3G+yN2lDbYfk0L1NHQmUbbMrweMxEe0wbItrDrkvfmJ8zhKmdIp7yD0TCm6nBg0ivp51jWflb8qeq4U8j2doNxLKtzUbTNIc3+G9fnZ1sWj+hYXQE8Klpmlln9aUAZpumGfXdtM5LggF3+o10IxZ/ghU/CY+fChDvwSkuPY3HV+/B3Gv6oGVLuUsRDxadwl2vbcaiiUNt7SFbVeXHR7sLMX5gJhISxC15c6Kd6OI5i0RwWfBjw4Gf/tS5fuNAYQlmLN6ElycNQ3p6uu3jbTx4FNm5Pryb40WnTmIfzrZ+W4DJz/uwfJoX7dvHtvO5yjYoCjeOKTweM9nevp1HizD1BR/enuJFu3b2t7SS1TdaRVdvvkj/ZKy4oa1UVtZi+ZajuGF4JzRvLmc5cBCRXk9VHlbyp6ohuCz4TyOBG27Qs60AHEuCDcP42Vn/6jmGYbwT9u8jACd4C2Ka5jEAhwzDCL7uGAtgO+/xiPjA4zGQmdqC2VkFAiZOVtTEtHF0UvMEDO+eYfmhorK2FluOlKC2ttbS30WjpKoKX35zAlVVVUKOxyKxWTNktEq0/aDg8RhIT24eFwNHUWUlvvj6BCorK2P6fREbxcdKRkaL7/x0igSPB8mJCbZCRcJpl5yMczJTpHxBUFNXh2p/Herq6mL+m+TE5rikd6b0B8WmTlWVHx9uPw6/P/Yt2xObNUOX9JbSJjeJzZqhVYtmwo5fWVOLj/cWChsfeElu3hxJzT3a1emEBA96d0gV1pfECwkJHnRvm6zkuvj9Aew9XiZkDLNzrFie6Xh+l+CnVauE7/x0G6zQpR4AzgHwML67ZLcMwGbTNGMflf732F4Az+LMVjn7AfzcNM2ohged37C6wfQfD1gJdOANxzh2shwXPro2pD+9dww6tuGPhd9fcBJX/Gl9SH848yL0bC9nmZSosssKznCinVgNXVG56frCd1fjqbXVIX3nmBb4zTXjpJyLxeETpbhk4X9C+uPfXIqubdO4jycq/CsSPIFOugTlWMGNYwpPsAxvOF6siO5/dQnwkhlsZgcnwnPc0FZUhlGJPJcTIVqEPB5Z8R7+8XH9tO32SxIw+/tXOVii6Fh+w2qa5jemaa4xTXNUAw/rRjuT1bPH9p31pw41TfNHrMkqQQDWAh14wzHmLNvG1Fa54xUfU4tEVNnjKTjDauiKyk3XwyerkbQqZr6xhamtIir8KxI8gU66BOXEOzzBMrzheLEiuv/VJcBLZrCZHSg8JzIqx1SR54qnZwEC35msRtJuoNHXB4ZhXGgYxheGYZQbhlFjGEadYRilKgpHEEGsBDrwhmM8MWEQU1tl8W1ephaJqLLHU3CG1dAVlZuuPzuuBVOr4tmbhzC1VUSFf0WCJ9BJl6CceIcnWIY3HC9WRPe/ugR4yQw2swOF50RG5Zgq8lzx9CxAAP+8PIGp3UAs692eBPAzAHsAtAQwBcDfZBbKTTRcfqLbcpR4oVmzZtg0dzzmXtMfO34/nulJSkhIwOezL8O1gztg2wOXxexjTUlJwYpfnI8OKc3x+czzkZJib8LSpk0bPHZTFgAg79YstGkjr8NPSUnBm9PPRVoLA+t/eS532Zs1a4b/zhqDsf3bYdOcMcK8X060k3bt2mHej7sAAP59U5dGQ1c8Hg++mjseq+66FJt+O16q32jcuHFIP/u/Lz+rnSAtLQ1/n9QXBoD3J/dFWhr/cmAAaN26Nf5yy5mJx6rbetkK/2pIp06d8IcJXQEAy37WNaZAp4SEBKyfNRqX9M7AljmjHQ3KiRU3jinNmzf/Tp8bi7cyKSkJ6+65BOd2S8Om+y6xFI4XC23atMHrt3uRkgj89xde2/1vy5Ytv1OXnArw6tSpE2ZdmwEAWPzDDOHBZg0JBEwUllU3mh1hGAaem3geZl/VDy/9/DwlfkTZbSXWz86iWbNm+PL+sZh6SRa2/W6s1DCqZs2a4as54/CrK3pj+7xxts4l8liE81x11VUIju5jzmq3EUtK8AbTNEcYhrHZNM2hZ//betM0L1JSQpCHlbDmpygtr8bQ+atDevPccUhLafwNVtGpSox4+KOQ3nDf5WjXmj84ZsM3RzDh6frNrt6cfi5G9OjMfTwWoryIsnxaTrSTT78+jJv+Wb+v6JKcYbjwnK5Szxkr9yzKw5s76/WE/sAfb1Pfb2w+fBw/eLK+b31nxggM7dqB+3gy/XVbvy3A9//2RUiv+MX5GNyFnRSsi+/QCm4cU8orajD4wfdDeusD45HSiv2GjbefjhXR/bkufug1ew7gtufqLR+LJg/CZX2ypJzLii+Vx8dsF5ltRZQnV+V1EVlHdanvhBh+80Ie3gjb6vonA4CF/6fnuGLZwxpGpWEYiQB8hmE8ZhjGrwGIz58nCAZW/BS56w8wdTQWrNrF1FZR6TUS5UXUxaclguzcTUztJOGT1UhaFaLrqMw6P/l5H1NHIp7qs84s2XCIqSPB20/Hiuj+XBc/9LTcbUwtEiu+VB4fs86I8uSqvC4i66gu9Z0QQ/hkNZJ2A7FMWG89+3szAFQA6AbgBpmFIoiGWPFTTB+TxdTRmH9dP6a2yqoG3qKGWiSivIi6+LREkJ8zjKmd5KEhbK2KvAZ1sqG2ikx/3fJpXqaORDzVZ52ZNKobU0eCt5+OFdH9uS5+6PycQUwtEiu+VB4fs86I8uSqvC4i66gu9Z0Qw4KhbO0GGl0S/D9/YBhtANxhmuZDcor0v+i8JBhwx8bV8YCVTclLy6uRu/4Apo/JQosWsS8zKzpViQWrdmH+df2E7CO54ZsjyM71YVWOF507y1kOHOTwiVLMfGMLnr15iC0vYnHpaTy+eg/mXtNHqE/LiXby6deHkZ27Cfk5w9C1qx7LgYMElwU/NAS45Rbn+o3Nh48jO9eHvBwvOnTgXw4cZOPBo8jO9eHdHK9wf93Wbwsw+Xkflk/zon179nLgILLqs0zcOKaUV9RgyYZDmDSqGxITY3u45+2nY0V0f15V5cdHuwsxfmCmo37oNXsOYFruNuTnDEJWVpbUcwUCJk5U1KBdSmKjvtTKylos33IUNwzvpGyPWJltxcpnZ6Hyuoiso7rUd0IMwWXBC4YCN9+s77hieUmwYRjdDMN4xjCMFYZhTDEMI9kwjMcB7AIQ25MCQQgkEDBRUlkbUwBCclJzXD3E+uBQU1eHgrIq1NXV8RbzOyR4PEjweJRsGB4wTdTWBWwFRABAcmJzXNI7U7tN6XlIbt4cyc0TtPwsHTq0+M5Pp2iXnIz+ndOEPNADgMcw4DEMKaErxadPo6SyFqdPn475bxKbNcM5bVtRaIhkEpp50KVNsqW+rsrvx+6CMvj9crZYSPB40LZVorD+1+MxkJ7cXEmgEIszYwqUjCsej4HM1BYxfeaEBA+6t7VWB3TGymdnkZjYDMO6pyvpg0TW0YQED87JbMV1P60EVokItyIap0eP9O/8dBtR37AahvERgLUAPgFwNYCxALYB+LVpmseUlRB6v2F1Y0CGG7ESAOD3B3Du/PdRVuVHalICvpo7HgkJjXe4R4rLcNFj60J6/azR6JyRyl1m36Fj+NFTX4b0v+88D95uHbmPx+Jg0SmM/uPHIb3unkvQvZ31hFZZQQtOtBPRgUIiWfju6u/svXrnmBb4zTXqk4LdVOfX7f0Gk57dGtIvThmM0b17MP+GJwzIadw4pvD0GwUlFRj5yJqQ/nz2ZWifLi4eo6SsCt6HPghp35yxSE/lTyK2Evwnk4/3HcTEf9VnFLw8dQgu6dVdeTka4sT1cUNb4X0e4UHkPbBTbiuBVaLCrQg2T675L/64qiSk77k6HTMuu9i5AjHgCV3KME1znmma75mm+WsAHQDcpnqyShCAtQCAvYXlKKs68419WZUfewvLYzrHrKVbmdoqKkOX7nptM1PHSjwFLai8/lYJn6xG0qpwU52fmruVqSPBEwZEWIen35i3YgdT2+WptfuY2ipWgv9kMiV3C1M7hS7XRzd4n0d4EHkP7JTbSmCVqHArgk34ZDWSdgPMr0sMw2hjGEaGYRgZAI4BSA7TBKEMKwEAfTukIPXst/upSQno2yG2PUmfvnEwU1tlZYPAmYZaJIsmDmXqWImnoAXRgUIieXZcC6ZWhZvqfH7OYKaOBE8YEGEdnn5j4fUDmNoud4/rxdRWsRL8J5P8nCFM7RS6XB/d4H0e4UHkPbBTbiuBVaLCrQg2r1yXztRugLUk+ACAAIBI7+ZN0zR7SizXd9B5STDgzoAMN2Il1IM3zEN0YMyKbbsx46U9yM/ug759+9o+HgtRZS8oqcC8FTuw8PoBaNVK3BI9J9rJ675tmLXkAPJuzcKgQfLSNHm4/oE8fFULZCcDv/2tc/3G5we+RXauD+/leNGlSxfbx3t14xbc//pBLL+lO4YMEfswHVzW9Mp16bj44tiWMx0pLsOspVvx9I2DkZrKv9xZJW4cU3iCZWQGdAHA3uPFuP0lH17P9iIjw/737CVlVXhq7T7cPa4XkpL4lxfb5Zn1X2DBOwV448ftcf755ztWjobwBG/ZxQ1tRWV4kch7YCXosiFWAqtEhVsRbMbNzsNeAD8BsFDTtgJEXxJsOSXYCXSesLrBQxEPWPFT8G7UfexkOS58dG1If3rvGHRsw/9t6OETpbhk4X9C+uPfXIqubfnTe1kUnarEiIc/CukN912Odq2th+iUlldj6PzVIb157jikpdh/++dEO9l5tAhXP/FZSK+66wL079RO6jlj5am167Hw3frlWr+5pg3uHHOR8nKIrqPbjxTi2r9+HtIrfzkSAzuLeUv/6deHcdM/6/fSXZIzDBeew05+llWfZeLGMYXHhya7fxTVJwbhHVdEo6s3nzyskXGrh5W8pfHF797Iwwv18RL4v/OA3/9Er7YShMfDShDaYMVPwbtR95xl25jaKjPf2MLUIlmwahdTx0ru+gNM7SamvuBjaicJn6xG0qoQXUenLPIxtR2yczcxdSTiqT7rDI8PTXb/KKpPDMI7rohGV28+eVgj41YPK3lL44vwyWok7QZowkq4Ait+Ct6Nup+YMIiprfLszUOYWiTzr+vH1LEyfUwWU7uJt6d4mdpJXv1BG6ZWheg6umyql6ntkJ8zjKkjEU/1WWd4fGiy+0dRfWIQ3nFFNLp688nDGhm3eljJWxpf/OVCtnYDtCRYAG7wUMQDVvwUvD7MnUeLMPUFH96e4kW7dvaXj364+2vk5G5Hfs5AnHPOObaPx8J36Biyc31YmeNFx478W4kcPlGKmW9swbM3D0Famrglek60kzV7DmBa7jbk5wxCVlaWknPGypS/5GH1MeA33YA773Su39h+pBBTFvmwbKoXmZn2l+/mfv4V/rD0CN6+sTPOPfdcASWs581N23HP4q+xctI5GDhwYEx/I8uTLRM3jik8HlbR/umGiOoTgxSXnsbjq/dg7jV90LJlSwEl5ONfn2zAQ8uO480JHTBixP+snHMM8rBGRqX3WaRf1o63lHyp+nHV7DzsAnATgEc0bSuATQ+rYRiXAOhjmubzhmFkAkgxTfNrCeWMiM4TVjd4KOIBK34K3r0XRe/bt/vYCVz5l09DOv9XF6Jvx7bcx2Mhyg8WTx5W0Z5kkQQndUF+e0NnZI8UO7mLBdH3m8dnGis8dVzWvsIyceOYwnOdRe8B3BDRHlld6tL6/Ydw8zP125a9Om0oLurpfPo1eVgjo3IvaF18p7qUg6jn6f98ikfzToT0vd9ri+mX6vmaldvDahjG7wDcC+C+s/+pOYCXxRaPINhY8VPw7r0oet++aS/6mFokovxg8eT5E+1JFkn4ZDWSVoXo+83jM40VnjoeT/sK6wzPdRa9B3BDRHtkdalLk3M3M7VTkIc1Mir3gtbFd6pLOYh6wierkbQbiMXDej2AHwCoAADTNI8AcMfeAETcYMVPwbv3ouh9+96a7GVqkYjyg8WT50+0J1kkb9/YmalVIfp+8/hMY4WnjsfTvsI6w3OdRe8B3BDRHlld6lJ+zlCmdgrysEZG5V7QuvhOdSkHUc+SH7VlajfQ6JJgwzA+N01zpGEYG03THG4YRisAn5imqayX1HlJMOAOD0U8YMUTweulEe012X3sBKa96MNbk71o21ZuByHKe8q7h21jONFOjp0sx5xl2/DEhEFISdFjOXAQmV5PK4i+359+fRjZuZuQnzMMXbuKWQ4chKeOq9wDURRuHFN4rrPsPXJF+/F1qUvr9x/C5NzNyM8Zim7dnF8OHMTOvp28uKGtqPT26uId1aUcRD3BZcFLftQWF16o53JgwN62Nq8bhvFPAOmGYUwFsBrAv0QXkGjaBAImCsuqwfoCxe8PYO/xMgQCgUaPV1NXh69PVKCurs5aOUwTlTV1zHJYIT0pCQM7t3Z0k3mrJCUm4IKebV3zcM+isrYW356sRG1tbUy/H0s9FEVVVdV3fjqFPxBAYXl1TO0qFtKTktC2VQspdT4pIQFd2yRbqpv+ugC+PVkp7PMRkamq9WPjwWL4/f7Y/8bvR1FZlaW/sUJKYiKGdU1XFgCkiuTmzdGyebOYw61UUVNbh23fnrI87oogXfkZY8djGEhNaq5k4mblOakxVI6HhHy+/fbEd366jVhDl8YDuBKAAeA90zTfb+RPhKLzG1Y3mP51JxaDvpVAh+LS0xi+4MOQ3nj/FchIazzRkffvolFQUoGRj6wJ6c9nX4b26XJSSkUFjMgKznCinewvOIkr/rQ+pD+ceRF6to++TE1lUMTf132Cx1YWh/SsazNwx+hRUs7FQnSd33m0CFc/8VlIr7rrAvTvZD9tGwCKTlVixMMfhfSG+y5Hu9bJzL9RGXgiCjeOKTyBdVbbp4oysdAldMl36Bh+9FT9Jor/vvM8eLvZT0C2ixNtzQ1tpbKyFgP/kB/S2397JZKT5XzRIHL8tjMeUuiSfsxdkoeXffV6oheYf5NebSWInTesME3zfdM0f2Oa5j2qJ6tE/BOLQd9KoMPjq/cwtei/i8a8FTuYWiSiAkbiKTjjjld8TN0QlUER4ZPVSFoVouv81Bd8TG2HBat2MXUkVAaeNGV4Auustk+riA7R0yV0KTvXx9ROQW0tMsu3HGVqkYgcv+2MhxS6pB/hk9VI2g3EkhJcZhhGaYN/hwzDeNswjJ4qCknEN7EY9K0EOsy9pg9Ti/67aCy8fgBTi0RUwEg8BWcsvs3L1A1RGRSx+IcZTK0K0XX+7SleprbD/Ov6MXUkVAaeNGV4Auustk+riA7R0yV0aWWOl6mdwom2lt6I1oEbhndiapGIHL/tjIcUuqQfj5/P1m4gltCl3wM4AuBVnFkSfBOAjgB2AZhumuZlksuo9ZJgwB2mf92JxaBvZWN63jCPnUeLMPUFH96e4kW7dvaXMn64+2vk5G5Hfs5AnHPOObaPx2LjwaPIzvXh3RwvOnXiHxRlBUQ40U4+3ncQU3K3ID9nCLp3797o7/v9AewtLEffDinweGJagMLNtL/mIf8IcG8PYPp05/oN0XV+5fY9uPPF3cif3Bd9+tibADdk3d5vMDV3K/JzBqNHjx4x/Y3OwVvRcOOYUnSqEgtW7cL86/ohOZm9VDsIz/20wv6Ck7jjFR8W3+ZFmzb2v3xTGZ7D4pn1X2DBOwV448ftcf75+jx58tQBu8hsK6KCg2QFGUZCZPCVnc9PoUv6cfXsPOwE8DMDePhhfceVaEuCY5mwfmaa5gUN/tunpmleaBjGJtM0xe1bEAWdJ6xu8FDEA35/AOfOfx9lVX6kJiXgq7njkZAQeULB6zUS7Tnd+m0Bvv+3L0J6xS/Ox+Au7bmPx0JU2ePJw3qkuAwXPbYupNfPGo3OGdG/vFDpu1n0hQ/z3vo2pOf9uAtuO98r5VwsRNd5mb7E7UcKce1fPw/plb8ciYGd2W+5yMOqBiv9cxBRvvtoiPZn83xGGazffwg3P1O/9+qr04biop7OrxxwwuMrs62IGg90qTdE0+Zfn2zAQ8uOh/ScH3bA1FH/MyfUAjse1oBhGD81DMNz9t9Pw/4/ig4jlLC3sBxlVWeSJMuq/NhbWB71d3m9RqI9p5Of9zG1SESVPZ48rLOWbmXqhqj03YRPViNpVYiu8zJ9iVMW+Zg6EuSrU4OV/jmIKN99NET7s3k+owwm525maqfQxeMrClHjgS71hmjahE9WI2k3EMuE9RYAtwIoAHD87P+eaBhGSwAzJJaNIEL07ZCC1LPf1qYmJaBvh+hL+3i9RqI9p8uneZlaJKLKHk8e1qdvHMzUDVHpu/n3TV2YWhWi67xMX+KyqV6mjgR5WNVgpX8OIsp3Hw3R/myezyiD/JyhTO0Uunh8RSFqPNCl3hBNmzcndGBqNxDTtjZOo/OSYMCdfiM3YsVfyLvBe0FJBeat2IGF1w9Aq1b2t6DZ+m0BJj/vw/JpXrRvL2c5cBBRZZe1+bsT7cSql1ml7ya4LPjfN3WB1+uVei4Wouu8aN9gONuPFGLKIh+WTfUiMzO2B2JdfIdWcOOYwuP/PnyiFDPf2IJnbx6CtDRxy4GDFJeexuOr92DuNX3QsiX/cuAgKj3uLNbvP4TJuZuRnzMU3brp8yUM77hrBzd4WHWpN0TTJrgs+M0JHTBihJ7LgQEbS4INw0gyDONOwzD+bhhGbvCfnGISRHQ8HgNtU1rENHD46wL49mSl5c2zA6aJan+dKzfKrvL7cfTUafj9flvH8fsDOHjC+rXTkQSPB2lJzbV8SOialoYEQMqDuhXKa2qwv7AMNTVilkAXnz6NIyWncfr0aSHHs0tCMw+6tEnWsg7EE4GAiZMVNZb6zoBporYuIK2/LaqsxH/3FqKyslLI8aqq/fh4d6HtPtYuyc2bo2XzZo2GD6rG4zGQntzckZAdNRFP+lNeUYNn1+4T0p8HAiYKy6pd+TxE/C/Bfsvp/ouXWEKX3gCwE8DNAP6AM0uEd5imeZf84p1B5zesbgzIcCNWAhB4g1aOnSzHhY+uDelP7x2Djm34l++oDF06WHQKo//4cUivu+cSdG/X2vJxZG1y7kQ7sRoopDJ0afWu/ZjyfL1f9NmfD8C4fup3CRMdkrThmyOY8PRXIf3m9HMxokdnW2UMwhO65EQQjF3cOKbwhLWJ6rOisfvYCVz5l09DOv9XF6Jvx7bcx9MlwMt36Bh+9NSXIf3vO8+Dt1tH5eVoiKzAPhYUuvRdRNZRleMhIZ+n//MpHs07EdL3fq8tpl96oYMlio6d0KXepmk+AKDCNM0XAHwPgFizCUE0gpUABN6glTnLtjG1VVSGLt312mamjhWVm5zLxmqgkMrQpduf38HUqhAdkpSd62NqO/CELsVbEIyu8IS1ieqzojHtRR9TW0WXAC+ZbcwO8RTYB7gzdElkHVU5HhLyCZ+sRtJuIJYJa+3ZnyWGYQwG0BpAlrQSEUQErAQg8AatPDFhEFNbRWXo0qKJQ5k6VlRuci4bq4FCKkOX8qcNYGpViA5JWpXjZWo78IQuxVsQjK7whLWJ6rOi8dZkL1NbRZcAr5UN2lRD7RROBPY1XAYsclmwG0OXRNZRleMhIZ8lP2rL1K7ANE3mPwBTALQBMBrAfpxJC85p7O9E/jvvvPNMnelx74rQP4KPurqAWVBaZQYCgai/U1tbZ+44csqsq6tr9HiFJRXmr5dsNCsqKiyVY8vh4+bIB98zjx8/bunvopG3bbeZde8Kc/fu3UKOx2LToWPmeb9fZR47dszWcU6cqjTvf2uTWVlZKahkZ3CinXz29WFz0AMrzMOHD8f0+9XVfnP9nkLT7/dLLplpzsw9cy1eesnZfuO/+w6a/eesMA8ePCjkeMu37jJ73LvC3LVrl5Dj2T02b1/gJG4cU8rKq81/rdlrVldXx/w3Xx08ap47713z6NGjUspktf03hi516YUvfGaPe1eYPp/P0XI05FRZlfnn93aaVVVVys4ps62cPl1rrtx0xKytrbV1HJXjytHiMjP7+U/NsrIy28eK5bmMcA9T/nKmnTz1lN7jCoANZoS5INPDahiGB8AE0zRflz91jg55WOObWLwSVvwUvL61olOVGPHwRyG94b7L0a41/3e2sj1a4ZSUVcH70Ach7ZszFumpSZaPI8vz50Q7OVBYgsse/29Ir7n7YmRlpkf9fZVeo6fWrsfCd+uXzP3mmja4c8xFUs7FQrRvW/TxwuHx2+riO7SCG8cUHv9icelpDF/wYUhvvP8KZKTZT/INIqpPDKKLH3rz4eP4wZP1z0PvzBiBoV2d36LCiesjs62I8uSq9IK6sb8j1PDcZxvx4Nv1Fq8Hru+EyRcMd7BE0eHysJqmGQDttUpIJhavhBU/Ba9vbcGqXUxtFdkerXCeWruPqWMlnjx/MxZvYuqGqPQahU9WI2lViPZtiz5eODx+W118h/EOj3/x8dV7mNouovrEILr0jbp6WHW5PqIQ5clV6QWl/o6IRvhkNZJ2A7G8PnjfMIx7DMPoZhhGRvCf9JK5hIb5l2LyMJsWsXglrPgpeH1r86/rx9RWke3RCufucb2YOlbiyfP38qRhTN0QlV6jV3/QhqlVIdq3Lfp44fD4bXXxHcY7PP7Fudf0YWq7iOoTg+jSN+Y18Kw21E7hxPVpmKsuMmddlCdXpReU+jsiGkt/2omp3UAs29p8HeE/m6ZpKtuDQeclwYA7N3nXjVg26K6srMXyLUdxw/BOje4/x7th/IHCEsxYvAkvTxqG9PR0Kx8hIhu+OYLsXB9W5XjRubPcrzN8h44hO9eHlTledOzIv80B77VrDCfayfr9hzA5dzPyc4aiW7fGB++amjp8efDkmYeVZnK3ZJj+VB7ePQTc3xOYNs25fmPrtwWY/LwPy6d50b69/W2XrF5zK6zZcwDTcrchP2cQsrKyYvqbI8VlmLV0K56+cTBSU1OFlkcWbhxTSsqq8NTafbh7XC8kJcW29PbYyXLMWbYNT0wYhJQU8V8QFZRUYN6KHVh4/QC0ahV9S6tYKa+owZINhzBpVDckJjq31PLfW3biV6/sw6rbeqF///6OlaMhTlwfmW1F1Hig8rroUkcJ/ZiZm4elu4EHBwG33qrvuBJtSXCjE1Yd0HnC6ka/kRux4o/h3UtUtP/D6j6gdrDq14xGPO3DeqS4DBc9ti6k188ajc4Z0ScsKr1GS77aitmvfRPSj9zYAzedO1jKuViI9m2L3tc1HB5PeGl5NYbOXx3Sm+eOQ1pKCyHlkYUbxxSefkP2vp2ij6/S485CZhuzgxPXZ/DsPIQbN1IAbNWsrcgaUwnCCq/7tmHWkgMh/dhNWfipV9wKKJFw78NqGEayYRhzDcN45qzuYxjG92UUkiCiYcUfw7uXqGj/h9V9QO1g1a8ZjXjah3XW0q1M3RCVXqPwyWokrQrRvm3R+7qGw+MJz11/gKkJMfD0G7L37RR9fJUedxYy25gdnLg+Dc/gzB1hE09jKuFewierkbQbiOXrr+cB1AAIRlgeBjBfWolcRkPXjVgXDhHEij+Gdy9R0f4Pq/uA2sGqXzMa8bQP69M3Dmbqhqj0Gq2Y2IOpVSHaty16X9dweDzh08dkMTUhBp5+Q/a+naKPr9LjzkJmG7ODE9fn0ka0DsTTmEq4l7xbs5jaDcTiYd1gmuYIwzC+Mk3z3LP/bZNpmnxPxBzovCQYqF/C1QfA+5otR4knqqr8+Gh3IcYPzERCAjsu34rfNRzR/g/RHioWovy3vNeuMZzw5Vn1L8bipRZFcFnwiok9MHiw+uXAQYpOVWLBql2Yf10/JCfzLwcOsr/gJO54xYfFt3nRpo3YScjBolO467XNWDRxKFq3jm2LqNLyauSuP4DpY7LQooXey4GDuNHDytNvyPaMiz6+3x/A3sJy9O2QAo9H/XLgIDLbmB2cuD7BZcGXAnhJ07Yia0wlCCsElwXn3ZqFQYP0XA4M2FgSDKDGMIyWAMyzB+oFoFpw+VxNMF6inaOliH8SE5thxDkZMT14eDwG0lo2lz7paAx/IIDSqloEAgHp50pr0QLDurax/VCekOBB97bJjj6QicIfCKCi2h/z9fd4DGSmtnC83qgkuXlzXJDVVtiDVFJCAtqlJjX6pRIP5TU1OH6qCjU1sS/XTklOxC2jsiiARDLJyc1x4wXdLdWjqho/Ptt/An6/X0qZamrrsO3bU6irqxNzvJo6bDpYIux4hH2MBj9FEgiYKCyrht2sF38ggKOnTit5DhBV5ngpB1GP5+xzjVufb2J5w3olgDkABgLIB3AxgNtM01wjvXRn0fkNa//ZeagK00kAdmr6LZ+bsRKIw7uBuejQJauhP3YoLj2N4Qs+DOmN91+BjDTrCb+yglCcCJLhCelRhS6hS6IDQWTW+e1HCnHtXz8P6ZW/HImBndlbZ6gM0hKFG0OXeJAdiCW6P9clPIdCl+oZMjsPZWE6FcAWQW1FVN+hMvhNl/5Ol3IQ9by5aTvuWVy/6csff3YOJgwb6GCJosP9htU0zXwANwC4DcBiACNUTlZ1p6oRTYjBSiAO7wbmokOXrIb+2OHx1XuYOlZkB6GohCekRxW6hC6JDgSRWeenLPIxdSRUBmkR1pAdiCW6P9clPIdCl+opa0TbQVTfoTL4TZf+TpdyEPWET1YjaTcQS0rwOwCuBLDGNM0VpmkWyS+WexjViCbEYCUQh3cDc9GhS1ZDf+ww95o+TB0rsoNQVMIT0qMKXUKXRAeCyKzzy6Z6mToSKoO0CGvIDsQS3Z/rEp5DoUv1jG5E20FU36Ey+E2X/k6XchD1rJx0DlO7gViWBI8BcCOA7wH4HMBrAFaYpqnsZaLOS4IBdwZkuBErgUi8G9KLDknafqQQUxb5sGyqF5mZsU2cedl97ASmvejDW5O9aNu2LfdxZAVFOdFO1uw5gGm525CfMwhZWVmN/r7K0JBfP5eHt/cAfxgITJrkXL8hOsDl8wPfIjvXh/dyvOjSpYuAEtZj9X4CaoO0ROHGMYUnsO7wiVLMfGMLnr15CNLS0oSXiSekiwXvuCKa/J37MH3RTuRP7Y9evXpJPZeV9uNEuJDMtmIl6JFFcelpPL56D+Ze0wctW1q36lhBZH9n51i6BJQR9dyzKA9v7gQeGgLccou+40q0JcGNTljDDtAMwBUApgK42jRN8aNLFHSesDYVv5HTWPEPlZRVwfvQByHtmzMW6alJEX+X9xyxwFsOHopOVWLEwx+F9Ib7Lke71tYTX2WV2Yl2YtXrpdKD9dxnG/Hg2/VLCh+4vhMmXzBcyrlYHDtZjgsfXRvSn947Bh3b8D+IHz5RiksW/iekP/7NpejaVsxQUVBSgZGPrAnpz2dfhvbpcpO3ncCNYwqPX1S2t0+Urz+ISi8ii73HizHuz5+E9Opfj0LvDhlSzmXFiygr/4CFzLYi6vM44e0VgR0fKnlY9SP386/wh6VHQvq3N3RG9shzHSxRdOykBONsSvCPAdwO4HwAL4gtHkGwseIfemrtPqYWcY5Y4C0HDwtW7WLqWFFZZtlY9Xqp9GCFT1YjaVXMWbaNqa0y840tTG2HeSt2MDXhHDx+UdnePlG+/iAqvYgsbn/Jx9QiseJFjKf8A0Dc53HC2ysCOz5U8rDqR/hkNZJ2A7F4WF8DsANn3q4+BaCXaZq/kF0wt9DQFWPPJUNEw4p/6O5xvZhaxDligbccPMy/rh9Tx4rKMsvGqtdLpQdr6U87MbUqnpgwiKmt8uzNQ5jaDguvH8DUhHPw+EVle/tE+fqDqPQisng928vUIrHiRXQi/6B3I9oOoj6PE95eEdjxoZKHVT/evrEzU7uBWDysVwN43zTNurP6YgA3m6Z5p4LyAdB7STBQvyylG4D/aL50y81Y8ceUlFXhqbX7cPe4XkhKin1Jq2gPDm85eCg6VYkFq3Zh/nX9kJxsfTlwEFlldsKXZ9WfqdJ3E1wWvPSnnTB8uPrlwEFE+/Jk+hJl+at1o6l4WEvLq5G7/gCmj8myvX90JER7B2WXN1b2Hi/G7S/58Hq2FxkZcpYDB7HiY6ypqcOXB0+emezFsF+6CIJtpTeA1YLbiqjP41Y/px0PqxuzA+Kd4LLgt2/sjHPP1XM5MGBvW5tVAIYYhvGoYRgHAMwHsFN8Ed1LcJdB92VuuYvExGYY1j09poEjuUVzXNG/g+VJp8oNvkWTlJCAgR3TbIVDAEBS8wQM755h+zg6kNisGdJaJsb8sJGQ4EH/TmlKHiranv1Swc6XCyJISkhAz3Yprrjf/kAApVW1ltpnVZUf724+Cr/fL7FkBA9JiQm4oGdbaXUvYJo4XVuHWLM6GiMlORG3jMqKeUIui5q6OlTW1KGurk76uTweA5mpLWKadPj9ARw8Ual0/Ax+bdBOwrETE5thVO92tiffHo+BtimxXUO7BAImCsuqhdV5XqzUG13KHO9knP3STnbwlyyivmE1DKMvgJsA/AzACZxJB77HNE3l+y/o/IZV5sbVRD1Wggt4wxJEh2qoDF3iCT2JRFWVH/3nvRfSO+ddhaQk+w+TTgTJyAwAssu/t+zEr16p9wf/5ZZe+NGQ/srLIbqOyrzmR4rLcNFj60J6/azR6JyRyvgLefVZJk0ldEl2SI+oILogugTJbD9SiGv/+nlIr/zlSAzsLDeBPhZEhxbGQr/ZeagO0y0A7NKsraisNyLPparcurSreEeXZ45Y4HnDuhPAWADXmaZ5iWmafwMg/+s8lyFz42qiHivBBbxhCaJDNVQGGPGEnkTio92FTO0mZAYA2SV84IikVSG6jsq85rOWbmXqSMRTfdYZnv5HdkiPqCC6ILoEyUxZ5GNqpxAdWhgL1Y1oHVBZb0SeS1W5dWlX8Y4uzxx2YE1YfwzgGICPDMP4l2EYYwHQ1x4NkLlxNVGPleAC3rAE0aEaKgOMeEJPIjF+YCZTuwmZAUB2WXVbL6ZWheg6KvOaP33jYKaORDzVZ53h6X9kh/SICqILokuQzLKpXqZ2CtGhhbFwYSNaB1TWG5HnUlVuXdpVvKPLM4cdYgldagXgRzizNPgKnNnS5m3TNPNZfycSnZcEA+4MyHAjVgIQeIODDhSWYMbiTXh50jCkp6fbLDGw+9gJTHvRh7cme9G2bVvbx2NhNWAoGrKCbZxoJ+v3H8Lk3M3IzxmKbt0af4hWGRqy4J138cz6AJ4e48E111wj9Vwsdh4twtQXfHh7ihft2tl3gn1+4Ftk5/rwXo4XXbp0EVDCenyHjiE714eVOV507Ngxpr/RJSjHCm4cU3gCjkSHIjVEdP+rS3jOm5u2457FX2PlpHMwcOBAx8rRECfamhvaisp6I3IMUxWcRAFNavj9W3l4/gvgrxcBP/iBnm0FiL4kuNEJa4ODZAD4CYAbTdO8QmD5mOg8YR06Ow+lYToNwGZNO003Y8XnwOulEb3RfEFJBUY+siakP599Gdqny0k2FXUu0T7eIE748g4WncLoP34c0uvuuQTd27WO+vsqN77/cPfXyM7dHtK52QNxRV/1sW3HTpbjwkfXhvSn945Bxzb8ScE8PtNY4anjKu+pKNzoYeW5zqJ899EQ3Z/r4rXT1cNqJWdCFNfMzkP4bswDALyrWVtRWW+cuAeEO1ixbTdmvFS/F/WTt/bB9wf1dbBE0eFOCQ7HNM1i0zT/qXKyqjuljWhCDFZ8DrxeGtEbzc9bsYOpRSLqXKJ9vE5y12ubmbohKje+zwmbrEbSqpizbBtTW4XHZxorPHVc5T1tyvBcZ1G++2iI7s918drp6mG1kjMhioY9gLwRlh+V9caJe0C4g/DJaiTtBuirF5uMaUQTYrDic+D10ojeaH7h9QOYWiSiziXax+skiyYOZeqGqNz4Pj9nIFOr4okJg5jaKjw+01jhqeMq72lThuc6i/LdR0N0f66L105XD6uVnAlR3Oxhax1QWW+cuAeEO8jP7sPUbsDSkmCn0HlJMOAOD0U8YMWbUVlZi+VbjuKG4Z0s7cUq2r8pyleq8lyyfGVOtBOrfsqqKj8+2l2I8QMzpe9L+qf8j/DXDyvx/FXJuPzyy6Wei8XBolO467XNWDRxKFq3jr5kOlZk+ra3HynElEU+LJvqRWZmbMsgyytqsGTDIUwa1c3x/TNjxY1jCk/bEV33GiK6Lurih1639xtMzd2K/JzB6NFD+U6DUVHZfwaR2VZEeStVelhF3gPe5yirkIdVDbrkZjSGEA+rU+g8YR09Ow8Hw3R3AOtc8oDhJlR4M0SfQ7ZHS8a5ZF1nJ3x5Vv2ZKr1Ga/YcwG3P1S+/XTR5EC7rkyXlXCxEezxl1nmefTXJw6oGnrYj2mPaENH7sKrsz1mI/lyicMI/+dM/5OHzyno9Mhl4/bdi2oqo8cCtHlZV++rq4g2Pd/J37sO0RTtD+pnb+uPK/nomBQvxsBL/y8FGNCEGFd4M0eeQ7dGSca548sBY9Weq9BpNy93G1KoQ7fGUWed59tUkD6saeNqOaI9pQ0Tvw6qyP2ch+nOJwomxI3yyGknbQdR44FYPq6p9dXXxhsc708Mmq5G0G6AJq01+2IgmxKDCmyH6HLI9WjLOFU8eGKv+TJVeo/ycQUytCtEeT5l1nmdfTfKwqoGn7Yj2mDZE9D6sKvtzFqI/lyicGDtyWrO1HUSNB271sKraV1cXb3i8kz+1P1O7AVoSLIDgsuAfAnhC86VbbkaFD0T0OVT650SdS9Z1dsKXd+xkOeYs24YnJgxCSkrjg7dKL82aPQcwLXcb8nMGISsrS+q5WIjee1ZmnS86VYkFq3Zh/nX9kJwc2zJIlXvrisKNHlaetiN7H1ae+sJCFz+06M8lCif2qQ0uC85pDdx3n54eVpXjish7QB7W+CJ/5z5MX7QT+VP7o1cvPZcDA7QkmIgDPB4DbVNaxNShBQImCsuqYfULmdLKarz0yQFUV1fzFtMxqvx+bD9WCr/fb+s4gYCJkxU1lq9dPBAImDhRbr3e8FBaXY0aE47XtaoaPz7bf8J2vQlSWVuLjYdOora2Vsjx7FJZXYsPdx7XpjzxisdjIDM1tv45SFJCAoZ2SZcW0BMwTVTV1glrzwnNPOjSJlnZZCwa5TU12H28FDU1ei2frKyqxaotR5W2tYqzy4ArKsQfm6dOR8LvD2Dv8TIEAgFBJYuOleekxkhObo4bL+jONVm18gwWy3XmfaYj6klu3hzNm0Hqlw8yoTesNqHQJTVYMebzmvhFh4CoDOkQFcQhK6TGDaFLKkND3tm6C798eW9I/3Vib/xgsPqlfaXl1Rg6f3VIb547Dmkp/OmnBSUVGPnImpD+fPZlaJ9uP20b4KvjJWVV8D70QUj75oxFemqSkPLIwo2hSzzIDnURXRerqvzoP++9kN457yokJalJwg3nQGEJLnv8vyG95u6LkZWZrrwcDRHdl8TC92bnIdz9PwhAnmZtRWXwmy4BRqLLocvncjMf7zuIif/aEtIvTx2CS3p1d7BE0aE3rJKg0CU1WDHm85r4RYeAqAzpEBXEEU8hNVZDl1SGhoRPViNpVeSuP8DUVpm3YgdT24Gnjj+1dh9TE84hO9RFdF38aHchU6tixuJNTO0UovuSWGjYozsTXcdG5ZiqS4CR6HLo8rnczJTcLUztBmjCahMKXVKDFWM+r4lfdAiIypAOUUEc8RRSYzV0SWVoyHs/783Uqpg+JouprbLw+gFMbQeeOn73uF5MTTiH7FAX0XVx/MBMplbFy5OGMbVTiO5LYmFiAlvrgMoxVZcAI9Hl0OVzuZn8nCFM7QZoSbAA3BiQoRuxmO6thAnwBq1YDelpDJXhGAcKSzBj8Sa8PGkY0tPTuY8jK2jBiXay8eBRZOf68G6OF506Nf5ArDI05JEV7+EfH/vxz8sTcNVVV0k9F4sjxWWYtXQrnr5xMFJTU20fb3/BSdzxig+Lb/OiTRuxD2fbjxRiyiIflk31IjMztgmD7GAfGbhxTOFpO6LrXkN2Hi3C1Bd8eHuKF+3atbN9vNLyauSuP4DpY7LQooXc5a4sdh87gWkv+vDWZC/atm3rWDka4kRbk9lWRI0HqsKLAGeCryIhOkiJgpns89Ta9Vj47km8+oM2uOiii5wuTlSiLQmmCatNvj87D1vD9GAAK1zygKELsfgTrHgYeL2Ior0mKj1PovyysnycTvjyjhSX4aLH1oX0+lmj0TlD/EMxD2v2HMBtz9UvYFs0eRAu65OlvByifdYyfds8nkQ3ep/c6GHl6Tdk+4tFZxKo9CK6oRwNUZkBEGTGP/Kw4kC9/n4W8OTtYtqKqM+j8rq4sb8j1PDp14dx0z/r7QNLcobhwnO6Olii6JCHVRJbG9FE48TiT7DiYeD1Ior2mqj0PInyyzqx+bssZi3dytROMi13G1OrQrTPWqZvm8eTSN4nNfD0G7L9xaIzCXTx9+tSjoY4MXaET1YjaTuI+jwqrwv1d0Q0snM3MbUboAmrTW5tztZE48TiT7DiYeD1Ior2mqj0PInyyzqx+bssnr5xMFM7SX7OIKZWhWiftUzfNo8nkbxPauDpN2T7i0VnEuji79elHA1xYux4oA9b20HU51F5Xai/I6KRnzOMqd0ALQkWgBv9Rrqhi4e1oKQC81bswMLrB6BVK/vbcYg+HovDJ0ox840tePbmIUhLS+M+TnlFDZZsOIRJo7ohMVHcgOdEO9n6bQEmP+/D8mletG/fXsk5Y+UfH3+GR1YU4bXr2+GCCy5wrByifYQHi07hrtc2Y9HEoWjdurWAEtaz+fBxZOf6kJfjRYcOHWL6G5X+MVG4cUypqvLjo92FGD8wM+Z9VfceL8btL/nwerYXGRkZwsskum7L6hutwuPlVoETHt9gW+kE4BNNPawqfaW8zz5E/PPHVR/gyTVVeG58EsaOHet0caJCHlZJjJmdh2/CdA8Aa13ygOEmVOzDKtpTpXJfOlFll+W7dcKXJ9rDJpLNh4/jB0/W92nvzBiBoV1jm4CJRHQdlXnNre6rC+izd6YVmoqHVdTe0dEQfe918Y7uPV6McX/+JKRX/3oUencQP9m3iux9dSNx59N5yAt7APteD+Cp6Xq3FZk44SMm3IEue7/HAnlYJfFNI5oQg4p9WEV7qlTuSyeq7LrsNSgC0R42kWTn+phaFaLrqMxrbnVfXSC+6rPO8Pj0RO0dHQ3R914X7+jtL/mY2ilk76sbibxv2LqpEU8ZFIRYdNn73Q40YbXJjxrRhBhU7MMq2lOlcl86UWXXZa9BEYj2sIkkL8fL1KoQXUdlXnOr++oC8VWfdYbHpydq7+hoiL73unhHX8/2MrVTyN5XNxJze7N1UyOeMigIseiy97sdaEmwAILLgn8E4C+aL91yM1b24eLds6ukrApPrd2Hu8f1QlKS/S0WVHp6RJWdx4sWC0748nTeg5PHjykD0XVU5jXn2SdZVn2WiRs9rDw+Pdn7VIu+97r4A2V7f3lxwi8eXBY8tzcwZYo72opMdNmHldCP4LLg937eG/366bkcGKAlwVIZcHYrowHnOFsONxMImCgsqwbrCxSPx0BmaouYJqCVp2uxbONh1NbWWiqHxzDQKjHBlRtTJ3g86JCaZHuQSkpKwDVDO7nm4Z5FYrNmOKdtKy3DJ/yBAPyBAAKBgNNFEUpBRQXW7S5ARUWF8GMneDxo3TLRUh33eAykJzd3ZZt2E4GAiZMVNcw+vCEpLRIxfkBHaQFG/roAvj1ZGXdtLCUxET3atXI0+CkSHo+BtJZq21owy1BypqFrSEjwoH+nNCGT1Vieywj3MLRDBwzunOLoF+R2oDesNrn9yTysOlyvr+4K/GMGfctnBdGbXZdX1GDwg++H9NYHxiOlVeMDu+gAGpWhS7yfWRVOBMnofE02HjyKG/6+MaSX3jEcw7vLX0LXENF1dOfRIlz9xGchvequC9C/UztbZQzCE9KjS1COFdwYusRznWUHYolu/7rUJZ7wMRU4EXA268U8vL69Xv90IPDYJL3bilsQ/VxGOMuBwhJc9vh/Q3rN3RcjKzPduQIxoDeskgifrEbSROOI3ux6yYZDTB0N0QE0KkOXeD9zPKPzNYnX0KWpL/iY2g48IT26BOXEOzzXWXYgluj2r0td4gkfU4ETAWfhk9VImuBH9HMZ4SwzFm9iajfg2ITVMIxmhmF8ZRjGCqfKIIL7zmFronFEb3Y9aVQ3po6G6AAalaFLvJ85ntH5mrzbIGSpoVaF6Dr69hQvU9uBJ6RHl6CceIfnOssOxBLd/nWpSzzhYypwIuDs4WFsTfAj+rmMcJaXJw1jajfg2JJgwzBmAhgBIM00ze+zflfnJcFA/RKudgA2aL50S1diCUmyEiZQUFKBeSt2YOH1A9DKgrlF9Ebzh0+UYuYbW/DszUOQlpZm+3gsRAWYxFPokuj7KZJ//vdzPLy8EK/fkImRI0c6Vg7RIUkygx12HzuBaS/68NZkL9q2bRvT3+gSlGMFN4Yu8QTuyO4fRbd/J0KFIrHx4FFk5/rwbo4XnTqptxJEQ2XIYJDes/PgB3A+gDc0bSu8IZA8iOzv7AQ4qfzMRGy8unEL7n/9IJbf0h1DhgxxujhRibYk2JEJq2EYXQG8AOAhADPdPGG94+95WHmwXl/bHfj7HXp2mm7GyobYvBuYi/Y8qfSwivIPyfIhOeHLU3n9rbJ+/yHc/MzmkH512lBc1FP9G2DRG83vPnYCV/7l05DO/9WF6NsxtollYxSXnsbwBR+G9Mb7r0BGml7JzyJwo4eVpx7xeJKtILr9i24rvBwpLsNFj60L6fWzRqNzhvNfxjnh8X3gtTy89FW9vvVc4MEb9WorKr2gIu+BnXKT/1U/dHnmiAXdPKx/ATALQNToPsMwphmGscEwjA2Fhfpu9h4+WY2kCTFY2RCbdwNz0Z4nlR5WUf4hJ3xIslB5/a0yOXczU6tC9Ebz0170MbUdHl+9h6kJ5+CpRzyeZCuIbv+i2wovs5ZuZWqncMLjGz5ZjaR1QKUXVOQ9sFNu8r/qhy7PHHZQPmE1DOP7AApM0/yS9XumaT5jmuYI0zRHZGbqu9n7nF5sTYjByobYvBuYi/Y8qfSwivIPOeFDkoXK62+V/JyhTK0K0RvNvzXZy9R2mHtNH6YmnIOnHvF4kq0guv2Lbiu8PH3jYKZ2Cic8vn8aydY6oNILKvIe2Ck3+V/1Q5dnDjsoXxJsGMbDAG4F4AeQBCANwFLTNCdG+xudlwQD9cuC5/QCpk7VazlKPGHFT8HrNSqvqMGSDYcwaVQ3IXvcqfT0iPKexpOH1QlPVays338Ik3M3Iz9nKLp1c25pjuiN5nl8prEi2m+rK270sPLUI1G++2iIbv+i2wovunrznfCLB5cF/2kkcMMNerYVt3pY7ZSbPKz6ocszR2NosyTYNM37TNPsappmFoCbAHzImqy6gTZtvvuTkIPHY6BtSouYOj/eDcw9hoHUJLUbn4ui5HQV3tx4EFVVVbaOk5DgwTmZrRx9IBNFYkIz9OuYpmXYTkbLlmiX0sLxiVdlVS1WbTmK2tpaIcdr36oVxg/saCnsLFb8gQBOVtYgEIjqJvkfamrq8MneItTV1Qkvj2zc1AslJHjQv1OapX4jpUUixg/oKOTLwYhl8njQqXVLYX2ZlTFIJh7DQIsEj5JyBAImCsuqEcvLjYQED3p3SFU6dvj93/0pEiufnYXHYyAz1fl6YxU75XbrZ45nuqaloX+nNOkBoLJwLCUYAAzDuAzAPW4OXZqzOA+vhG1ndMsw4KGf6fktn5uxYuLnDQ7iDWuKhsrQH1GbycsKFnEiSMaJjexjZefRIlz9xGchvequC9C/Uzvl5RBdR0vKquB96IOQ9s0Zi/TUJFtlDFJQUoGRj6wJ6c9nX4b26exJsRNBMHZxY+gSD7Lbp+j+XJcgGVF9fSxY+cxOXJ/7XsnD4i31+mdDgIdvEdNWdLnfVnBjf0eo4WDRKYz+48chve6eS9C9XWsHSxQdbd6whmOa5prGJqu6Ez5ZjaQJMVgx8fMGB/GGNUVDZeiPqM3kdQkWEYHOAVJTX/AxtSpE19Gn1u5jajvMW7GDqSPhRBAMERuy26fo/lyXIBlRfX0sWPnMTlyf8MlqJG0HXe63Fai/I6Jx12ubmdoNuH/Nn8P8cQRbE2KwYuLnDQ7iDWuKhsrQH1GbyesSLCICnQOk3p7iZWpViK6jd4/rxdR2WHj9AKaOhBNBMHZp+A5H73c6/Mhun6L7c12CZET19bFg5TM7cX0eG87WdtDlflvBjf0doYZFE4cytRtwdElwrOi8JBioX8I1CEBeHC7dUkEsYRZWwoAKSiowb8UOLLx+gCUvXUlZFZ5auw93j+uFpCT7Sxl5y8GDqLAb0cFTQZwIkrF6P1UGRTz32UY8+PZRLP1pJwwfLvBJyyKiA1wOFp3CXa9txqKJQ9G6tdglRzx1XOfgrWi4MXSJB9H9bUNE979OhApFYvuRQkxZ5MOyqV7I3kXBSp/oxPXpOTsPAQDnAXhLcFsRNR6oDOsSGZpIwUnxxTtbd+GXL+/Fez/vjX79xKayiyTakmCasNrkl8/k4Z399foHPYG/TovfBwwZxOKbtOLN4PXlifZ/qPRQivIiyiqzGzysKj1LH+87iIn/ql+/9vLUIbikV3cp52JRXlGDwQ++H9JbHxiPlFb8X1LI9G3z+GN19jFHgzysYhBdt2X5+61SXHoawxd8GNIb778CGWnOJ2Y7cX1mv5yHJWHb0N40GHhkol5tReV1ETmGudHDS0THd+gYfvRU/W6i/77zPHi7dXSwRNHR0sMaD4RPViNponFi8U1a8Wbw+vJE+z9UeihFeRF19n1axepnUelZmpK7halVsWTDIaa2ikzfNo8/Np7qc7wh+96Irtu6+PsfX72HqZ3CiesTPlmNpHVA5XUROYa50cNLRCc718fUboAmrDb5XT+2JhonFt+kFW8Gry9PtP9DpYdSlBdRZ9+nVax+FpWepfycIUytikmjujG1VWT6tnn8sfFUn+MN2fdGdN3Wxd8/95o+TO0UTlyfR89lax1QeV1EjmFu9PAS0VmZ42VqN0BLggUQXMLVH8CqOFy6pQLRHlbeDelF+914y8GDKC9iPHlYrV4TlV6jv6/7BI+tLMbiH2Zg1KhRUs/F4vCJUsx8YwuevXmIkP3Z9hecxB2v+LD4Ni/aCN6cmqeOy6rPMnGjh5Wn7Rw7WY45y7bhiQmDkJIi/kFe9PErK2uxfMtR3DC8E5o3598ixy6i8gpEo7L/DDJ0dh5KAYwB8IKmbUXldRF5LjvHIv+rfrzu24ZZSw4g79YsDBokL6zNLuRhlYQbPBTxgBU/Ba8XVbSnSrSHSsW54mkfVqteL5Wenc8PfIuf/sMX0q/f7sXIrC5SzsVC9L6pRacqMeLhj0J6w32Xo11rMV/U8OyrqYvv0Apu9LDyXGeZe/YC4vtfXfa4lNnG3MZDy1biX5/UP8NOHWVgzg+vdbBEzqKLh5X8r/qx+9gJXPmXT0M6/1cXom9Hfb7sCoc8rJJwg4ciHrDip+D1oor2VIn2UKk4ly4+LRFY9Xqp9Ozo4icRvW/qglW7mNoOPPtqxlN91hme6yxzz15AfP+ryx6XMtuY2wifrEbSTQ1dPKzkf9WPaS/6mNoN0ITVJm7wUMQDVvwUvF5U0Z4q0R4qFeeS5bdpeAdU7A5n1eul0rPzXgP/SEOtCtH7ps6/rh9T24FnX01dfIfxDs91lrlnLyC+/9Vlj0uZbcxt/H20wdRNDV08rOR/1Y+3JnuZ2g3QkmABBJcFP3oucOONei/dcjMq9oMTuYcZoNY/J+pcsvw2waWObQB8pWiJY3HpaTy+eg/mXtMHLVs2vvWDSt/N5we+RXauD+/leNGli/rlwEFE74Up07fN4yF0wldnl6biYZW9D6vo/leXfVhVZiPoTnBZ8N9HG7j22qa7HDiIyDHMzrHIw6ofunrfG0JLgiUSHGcljLdEGB6PgczUFjF1fh6PgTatnO8oPYaB1KTmSsoh6lwej4G2KbFdZx5ULqRL8HiQmdIi5gdoK3XMLu1btUJWuxS0atVK+rlYeAwDrRIThH3mxGbN0L1NspQH+sTEZhjWPd3Ssf3+AL4urEAgEBBeHqIenn4juUVzXNG/g7QAI9H9r07jSsvmzRwvhw4E646TIViNEQiYKCyrhooXRCrHMMJdpCQmomtGsmvCBxtCb1htMu/NPCwKK9ptI4B5E9zxjXi8whu0Ijp0iSckxulzyQpLcCJIprS8GkPnrw7pzXPHIS3FfvKzCA4UluCyx/8b0mvuvhhZmenKyyH6Gsm85jztWnSbVoEbQ5d4+g3ZIUai+19dAryshsnFM4+tfB9/X1fvj7xjdCJmXTvewRL9L24NIKLQpfjiSHEZLnpsXUivnzUanTP4d5OQCb1hlcSiDWxNqIc3aEV06BJPSIzT54qnsITc9QeY2klmLN7E1KoQfY1kXnOedi26TROR4ek3ZIcYie5/dQnwshomF8+ET1YjaR1w65hKoUvxxaylW5naDdCE1SZPjGJrQj28QSuiQ5d4QmKcPlc8hSVMH5PF1E7y8qRhTK0K0ddI5jXnadei2zQRGZ5+Q3aIkej+V5cAL6thcvHMv8YmMrUOuHVMpdCl+OLpGwcztRugJcECGDg7D5UALgbwiuZLt3QlFoO+itAlqyE9jaEyHENUgIno4KkgTgTJ6BxOokvo0uETpZj5xhY8e/MQpKWl2T7ekeIyzFq6FU/fOBipqWKXHJWWVyN3/QFMH5OFFi1iW2osO9hHBk0ldOnYyXLMWbYNT0wYhJQU8RNA0f25rL7RKgcKSzBj8Sa8PGkY0tPTHSuHDlw1Ow+7ANwE4BFN24pbA4hUhS659fq4jXe27sIvX96L937eG/366ZsuHm1JME1YbXLfK3lYvKVe/2wI8PAtenaauhKL38GKJ4LXPyHa86TSwyrKDybLV+aEL0+2R84OBSUVGPnImpD+fPZlaJ+uPnxJtOe0vKIGgx98P6S3PjAeKa3EfLvO057Iw6oGnj5XthdT9L3XxcOqszdfNX9c9QGeXFMV0jMuS8I9V491sESEVcjvqoYN3xzBhKe/Cuk3p5+LET06O1ii6JCHVRLhk9VImmicWPwOVjwRvP4J0Z4nlR5WUX4w2b4ylej8Weat2MHUqhDtOV2y4RBT24GnPZGHVQ08fa5sL6boe6+Lh1Vnb75qwierkTShP+R3VUN2ro+p3QBNWG3y2HC2JhonFr+DFU8Er39CtOdJpYdVlB9Mlq+s4XsNFe+4ZHvk7LDw+gFMrQrRntNJo7oxtR142pMbPaxZjWgd4elzZXsxRd97XTysOnvzVfPc+CSmJvSH/K5qWJXjZWo3QEuCBRBcFvzYcOCnP9V76ZauiPaw8noiKitrsXzLUdwwvJOQfd1EH4+FqE3tRR2nIcGljgkA9ipa4ijrs4igoKQC81bswMLrBzi6FyuPL5RFeUUNlmw4hEmjugnf742nPeniO7RCsK1kAVij+XLgIDx9rmiPaUNE33sen64MRLdZNxNcFvzc+CSMHUvLgd0IeVjVsOGbI8jO9WFVjhedO+u5HBigJcFSSU9v/p2fhHV02ew6IcGD7m2ThT2M+AMBHD11GoFAQMjx4gGV7xJLT1fjjS8Pobq6OqbfV7nBe2l1NXYfK425bLJI8HjQqXVLYXU+qUUCLukrZ3LY1NrTCacLYAG/P4C9x8ss3ZvkxOa4pHem9C/zROH3B/B1YUWTqX9uIPhln5Nf+jWG3x/AzqOlrqs3dsZDlWMpERspiYlo3TJR+BfJqqA3rDZ5NC8fT/+nNqSnX9oc937vSgdLFJ+oCF0SHdKjMhyDQpf+l6JTlRjx8EchveG+y9GudfSkYJXhD3uPF2Pcnz8J6dW/HoXeHTKknIuF6GAwmcE0PO2JQpfUwNNvyL43oo+vS12i0KV6nv7Pp3g0r/5rnXu/1xbTL73QwRL9L7qEdVnFznio4nmNsMbOo0W4+onPQnrVXRegf6d2DpYoOvSGVRLhk9VImhCDitAl0SE9KsMxKHTpf1mwahdTN0Rl+MPtL/mYWhWig8FkBtPwtCcKXVIDT78h+96IPr4udYlCl+oJn6xG0jqgS1iXVeyMhyqe1whrTH3Bx9RugCasNnnmiuZMTYhBReiS6JAeleEYuocuDWlEy2D+df2YuiEqwx9ez/YytSpEB4PJDKbhaU9uDF1quHOt2J1s5cDTb8i+N6KPr0tdotClepb8qC1T64AuYV1WsTMeqnheI6zx9hQvU7sBWhIsgLGz87APwAQAf9R86ZabsRKgwxuOITqk49jJcsxZtg1PTBiElBS5A1VJWRWeWrsPd4/rhaQk/rREWUFR4UsdVS1x3F9wEne84sPi27xo06bxh2iV4Q+6BCAUnarEglW7MP+6fkhOjr5kOlZkBsIcPlGKmW9swbM3D0FaWlpMfyOqXajEibZiF56AM9F1ryGi66IuYUcqxxUrOBFK9fPH8/BRITCzM/DLX+rZVnQO/2NhZzxUEZJJWGPNngOYlrsN+TmDkJWV5XRxohJtSTBNWG1CHlY1WPGB8HoiRHspRPsDVZxLlt/m8tl5+DpMnwPgI8kP4jp7vUrKquB96IOQ9s0Zi/RU9ZMp0XVUZp1vKh7W3rPz4A/TKlO1VVJeUYPBD74f0lsfGI+UVuLeroj248vy91tF5bhiBSe8mv/87+d4eHn90uz7rstEzsUjpZ7TKuTRJHTg2MlyXPjo2pD+9N4x6NhGny+7wiEPqyTIw6oGKz4QXk+EaC+FaH+ginPJ8tt83YiWgc5er6fW7mNqVYiuozLrfFPxsPob0fHCkg2HmNouov34uvj7VY4rVnDCqxk+WY2kdYA8moQOzFm2jandAE1YbUIeVjVY8YHweiJEeylE+wNVnEuW3+aGRrQMdPZ63T2uF1OrQnQdlVnnm4qH9fxGdLwwaVQ3praLaD++LH+/VVSOK1Zwwqv5+g2ZTK0D5NEkdOCJCYOY2g3QkmABfG92HrYBmJgAzJ8ff0u3dEGFh1W0f5M8rPU44cs7UlyGWUu34ukbByM1Va/4ms2HjyM714e8HC86dOjgWDlE+whlekaLS0/j8dV7MPeaPmjZsmVMfyPal64CN3pYefpc2R5W0XVRFy8iTzvgxYq/0Am/+OQ/5+GD48A9XYEZM/RsK271aLq13ERkVu/aj9uf34H8aQPQs2dPp4sTFfKwSuKJD9bhz++XhfSvx6firrGjHSxRfKJiXy/Rfjfah7UeJ/aWlO2Rs8PBolMY/cePQ3rdPZege7vWyssh2g8n0+fn1v0MreJGDyvPvZHtL9bV62kXlb5sK2OpE/2tGzysboW8t/HF1m8L8P2/fRHSK35xPgZ3ae9giaJDHlZJhE9WI2lCDCr29RLtd6N9WJ1FtkfODne9tpmpVSHaDyez/rh1P0OruNHDynNvZPuLdfV62kWlL9vKWOpEf+sGD6tbIe9tfDH5eR9TuwGasNrkxWtTmZoQg4p9vUT73Wgf1noauiVUuCdke+TssGjiUKZWhWg/nEyfn1v3M7SKGz2sPPdGtr9YV6+nXVT6sq2MpU70t27wsLoV8t7GF8uneZnaDdCSYAEElwW/eG0qRo+m5cCyULGvl2i/m8p9+0T5q2T5tILLggcByFO0xLG8ogZLNhzCpFHdkJio14B7sOgU7nptMxZNHIrWrdUvBw4i2rMs0+fnxD6PThBcFnw+gDc0Xw4chOfeyPYXy/LjO41KX7aVsdSJ/ja4LPj1GzIxciQtBxYJeVjji63fFmDy8z4sn+ZF+/Z6LgcGaEmwVFq1avWdn4QcPB4DmaktpHacHo+B9OTmws6RlJiAC3q2VRL24vcHcPBEJQKBgK3jJCR40LtDqvAJQf+zP4cqHPeSWiTgkr56hu14DAPNm3kcfxBITGyGYd3TXbWhvRUCAROFZdVww5ezQTo0+OkGEhI86N8pTasvEkTXbV3qUo2/DruOlaKurk76uayMu4nNm2FQl9ZK+5IeGRnwAMjIyFB2Tp0RWUdVPHMR6ujeujV+NrKHo1+Q24HesNrk2U+/xPx/HwvpuT/qiCkXnudgiQjesAA3bzQvKmBEVtDC1bPzsDNM9wewSvKbI51Deg6fKMUlC/8T0h//5lJ0bZumvByir5FuoUtuDA65eHYevg3TXQD81yVvWa0gOzxIdN3WpS6pDPOzgsrxLsiqHXtx+wu7Qvof/9cPVw/oLfWcOqNLHSX0Q9d+IxL0hlUS4ZPVSJpQD29YgJs3mhcVMCIraGFnI1oGOof0zHxjC1OrQvQ10i10yY3BId82ouMF2eFBouu2LnVJZZifFZwI7LsjbLIaSTc1dKmjhH7o2m9YgSasNnnrJx2ZmlAPb1iAmzeaFxUwIito4WcGW8tA55CeZ28ewtSqEH2NdAtdcmNwyPcb0fGC7PAg0XVbl7qkMszPCirHuyD5U/oxdVNDlzpK6Ieu/YYVaEmwAKb9NQ/5R4B7ewDTp8ff0i0VxGLutxIAwBvOIjo0QuVG86I2bZcVtBC+F6vsPViD6By64jt0DNm5PqzM8aJjR+e+6BIdDCYzeIWnrCrboCicaCt24ek3ZIfSib73utSlYyfLMWfZNjwxYRBSUvT5Is6J6/PAa3l46SvgTyOBG24Q21bcGDokssx2juXGaxfv7D1ejNtf8uH1bK/Wnu9oS4JpwmoT2rjaPrH4Lqx4M3TxsKpE97J/f3YetobpwQBWSH4Q1/maFJ2qxIiHPwrpDfddjnatk5WXQ5T3OYjMa85TVp19zNH43uw8bAvTKlO1eeHpc2W3T9F+Pl3qkug262Ze/nIz5r5Rv9/r/J90w8TzxGwR1tT9oHY+f1O/djpSUFKBkY+sCenPZ1+G9ul6BsWSh1UStHG1fWLxXVjxZujiYVWJ7mXf2oiWgc7XZMGqXUytClHe5yAyrzlPWXX2MUdjWyNaR3j6XNntU7SfT5e6JLrNupnwyWokbYem7ge18/mb+rXTkXkrdjC1G6AJq01o42r7xOK7sOLN0MXDqhLdy35rc7aWgc7XZP51/ZhaFaK8z0FkXnOesursY47GxAS21hGePld2+xTt59OlLolus27mnZu7MbUdmrof1M7nb+rXTkcWXj+Aqd0ALQkWAG1cbR/RHlZe/4QuHiUedC97cFnwrc2BBx9Us8RR52tSdKoSC1btwvzr+iE5Wf1y4CCifb4yrzlPWXn97E4SXBY8MQGYP1/v5cBBePpc2e1TtI9Ol7qkszdfNcFlwe/c3A1Dh4pZDhykqfswycMaXxSUVGDeih1YeP0AtGql53JggJYESyUzNRUAkHr2J2GdWDaotrKJNe+G14GAiZLKWmEbw9fU1OGTvUVKNnhPSPCgd4dU2w9Sssrcvft3f6pA1DWJZzweA2ktmwt7qKiq8eOz/Sfg9/uFHC+cxMRmGNY93dLkxuMx0DbFel/gJF27nP3Z1dlyWCEQMHGivNpS3+nxGGjTSt4Drd8fwN7jZQgEAlKO7xSi2yyLQMBEYVls99XK74qiZ5s2aJkAtGkjfgUN73NEvGDn8zf1a6cjCR4P2iQnuvZ5iN6w2mTp5h2Y+er+kP7TzT1xw1D3vWonxG9krzL0R1TIgawy3/H3PKw8WK+v7Q78/Q65b450Dn7QJXRJdJ2XuTk5T+iNznUgGjl/y8N7YZuvXtUF+Ocv9H7LynNvZIcYie7LdAldEt1mWagIO7TD+v2HcPMzm0P61WlDcVFPccuCCSJeKC49jeELPgzpjfdfgYy0lg6WKDr0hlUS4ZPVSJpwD6I3slcZ+iMq5EBWmcMnq5G0DHQOftAldEl0nZe5OTlP6I3OdSAa4ZPVSFpHeO6N7BAj0X2ZLqFLotssCxVhh3aYnLuZqQmCOMPjq/cwtRugCatN3v2/nkxNuAfRG9mrDP0RFXIgq8xzerG1DHQOftAldEl0nZe5OTlP6I3OdSAas7PYWkd47o3sECPRfZkuoUui2ywLFWGHdsjPGcrUBEGcYe41fZjaDdCSYAHMfjkPS7YCj54L3Hij3ku3CDaiwyyKS0/j8dV7MPeaPmjZUu7yC1EhB7KCULJm5wEAOgD4TNG+kjqHk+w8WoSpL/jw9hQv2rVr51g5RF8jmWFSpeXVyF1/ANPHZKFFi9iWGutcB6IRbCsZADZqvgdrkKoqPz7aXYjxAzORkBDbElXZoUvlFTVYsuEQJo3qhsRE+xMoXULc/r+9+4+zqir3B/55hgGGYYZfgjgkiqKChDghmnRJSMHEMs1beVMu14s/SK0wLTPFpDSy1LIfaLf8zgsrEzMtUwQBRdTQCGkcUFHRFAjUGX4NMzDAzFnfP86ew2mas+f82GvvZ+3zeb9e84Jn5szZ6+y99jpnzV7Ps8Ms2BZGscNC/Gz587jzyV347af7YsKECaFss1iwcFK8bHh/O770m1r8fkY1BgwYEHVzMsq0JJgT1gLZvHE1hSvoHJym5v0YfcvSVLzupimo6K3/Co8NV92zEAvfPRh/6khg3hV2P4iHmeuVKy05rEH3eZt523v2HMCo7y5Jxa9++0yUl/tPQDX3gUyumLcQi9JuJzl1KHDPVbonrRrzi4Pui1ryofm+ctCzG97F9HsP3tX715eOxmnHHBlhi+JDS3+nYDCHlazeuJrCFXQOzoLVm3zjYpI+We0stiHMXK9caclhDbrP28zbfmztVt+4M5r7QCaLNvnHGmnMLw66L2rJh+b7ykGX1azzjSl/Wvo7BYM5rGT1xtUUrqBzcKaPH+obF5PZx/jHNoSZ65UrLTmsQfd5m3nb54+t8o07o7kPZHLD0f6xRhrzi4Pui1ryofm+ctCSmaN9Y8qflv5OwWAOa0g0LwkGgB8vfQY/eaoZ88/qjUmTJkXdnNgKI58i6G24mMNqaz8fff1CJACcBODhkPLydu5uwbwVb+HaycNRVlYWyjaztfrdLZhRU4vFM6sxZMiQyNqRT+6hH5v7PJ/zKeg8xjC057COAvCEIzmsra0JbKhvwnGDK7K+z59rOaxa8qG3bN+N6x5Zh3suGK3q/u/59IFC3fPci/jBwm1YcN4hOPXUU0PZZq5czQV1td3UuY0NuzDrwTrMnzYGffv2jbo5GTGH1ZIN72/H5B+/kIqXfW08jhmsN5nZVS7mU7h4H1Zb+3nunxfhlysTqfjyj5Xghs9MLfh5/WjOX3ynficm3fmXVPzMtf+BYYP6hd6OoPtoPnmm2crneGq5d2Yubn5oIe576WD8PycB3/m8G5PWXLh2H1Yt44nNex0XIopzbc3GrTj/7jWp+JErx2LsEV2vvAiTi59dKH5cyn1nDqslX/pNrW9MwXAxn8LF+7Da2s/pk9XOYhs05y9++YGXfeOwBN1H88kzzVY+x1PLvTNzkT5Z7SyOC9fuw6plPLF5r+NCRHGuzaip9Y01cPGzC8VPHHLfOWEt0O9nVPvGFAwX8ylcvA+rrf18z8QS39gGzfmLv51+om8clqD7aD55ptnK53hquXdmLu461T+OC9fuw6plPLF5r+NCRHGuLZpZ7Rtr4OJnF4qfOOS+c0lwAFy5t5HrXMynCPO+fdpzWNuXBd8zsQRTp9pdDtwu6PzMIL1TvxNffuBl/Hb6iejXr19k7Qi6j9rM88vneEaRV1eo9mXBd50KnHde/JYDt7N9bILu21rGk3zuRxyGKM61NRu3YkZNLRbNrEZVla7lwO1c/OxC8eNKPQcuCbbojW3b8FbDXmzbti3qpsRaImGwrWkfsvkjy/79bXhhQwPa2tpy2kZrawLrtzYikbC/ZDVoLS2tePrV99Ha2hp1UzrVftzC/CNZImHQuPdAqNvMVnn37jhiQO9Ii7cAQEmJoH9vNz5I9ejRDeOOGpDT5KOkRHBIRU8nXl+7/fv/9V8X5DN2JhIGO5r3Wzs/EwmDnXuCO/9LSgT9yrtH3pdKS0pQ1beXuj/AlJaWYGRVH3XtylciYVC/O7vPHH5KSgSDKsMZg4JqM8VPWc9STDhO3x/vs8UrrAV64tU3ceWv30jFd08/DmePcq9ctHa5FHTIt9hG0EUjwiy6FFSxG1sFIr736BP41QsHx5rLxgtuPPfsgp/Xj80CQIX6YGczTrntmVS86vpJOLRf79DbEXSft7nP8+mbLhY8+db9C/HA2oPxF08Avn+R7qus+fQj2+Nj0EWSwhzP/Wge18Jms+iSi2OHi22mcLhUgJBXWC25Km2y2llMwciloEO+xTaCLhoRZtGloIrd2CoQkT5Z7Sy2wWYBoELNefw13zgsQfd5m/s8n77pYsGT9MlqZ7FG+fQj2+Nj0EWSwhzP/Wge18Jms+iSi2OHi22mcLhYgLAjTlgLtOSS43xjCkYuBR3yLbYRdNGIMIsuBVXsxlaBiLtPE9/YBpsFgAp1+2eP943DEnSft7nP8+mbLhY8+eFY/1ijfPqR7fEx6CJJYY7nfjSPa2GzWXTJxbHDxTZTOFwsQNgRlwQH4M4nn8bPlu9FzZm9cPrpp0fdnNjKpaBDvsU2gi4aEeaN5rc37sWdy97E7KnHolevXnk/j60CEWdcvxBvAfgcgDtuC2eJ487dLZi34i1cO3k4ysrKQtlmtrTcxDvoPmpzn+fz3C4WXRp3/UI0AJgC4FchnSuFymfMtV3EKOjiRFqKLmkd16IoLrTg7+tw/YPv4vFpR2L06NGBPreLxZJcHO8oHFqLtXWUaUkwJ6wFWvn2Jlz4y7pU/LvLx+BjR7tXLpqCF2bOgJab2mfS/keddl/5RC9c+0m7f9zRknPWGS3HK+h9ZHOfNzbtw5hbl6XiutmT0afC/03XxZyuHz6xFHc/e3Ap35Wn9cB1Z0+JsEVd05hfHHRf1NKXtI5rUeyfN97bhjPvejEVL7n6VBx32CFWt6mZlj5K+riU+84cVksuqanzjal4hZkzoOWm9pmkT1Y7i23QknPWGS3HK+h9ZHOf16x8xzfujIs5XemT1c5ijTTmFwfdF7X0Ja3jWhT75/Jf1/rGxUZLHyV94pD7zglrgZbMHOMbU/EKM2dAy03tM6k5s5dvbIOWnLPOaDleQe8jm/v8ionDfOPOuJjT9aszevjGGmnMLw66L2rpS1rHtSj2z8OXVPvGxUZLHyV94pD7ziXBAVj59iZcUlOHJTPHYOhQLgemg8LMJ9GSX5VJFLne+eYyh0HL8Qp6H9nc5/nk4LiYh9a+LPhXZ/TAlCm6lwO3y2c/2z42QfdFLX1J67gWxf55471tuPzXtXj4kmocckjxLgdup6WPkj5h1lQpBJcEWzSgVy8MrOhZUKEbCla+N89ubU1g/dZGJBIJSy2zp7S0BEcN6l3wxNjWjcfbi4OEWSQkkTDYueeAypuotxxoxZqN29Ha2hppO1pbE9i4bU9gfb6kRNC/t50PSxXlPXDR+GHo0SP7KweJhMG2puD7s03tHyY0f6joqKREMKiyZ07H3faxKS0twTGDK2NXfCboczYo+fSBQpWVlqJ/7x4q/0gbhSCPga3PAhSNltZW1P1zZ+SfOfLFK6wFWr+1AWf95K+pePGsj2Jk1cAIW0T5Fh4IukhSmEWXgtqWraIN81asxO2LDuZafWNqf1w18WMFP68fLYWNOrNzdwuqv/dUKq698Qz0qwy/2mfQhRhs9vl8+qZLN0tv9+Olz+AnTzWn4lln9MbXpkyKrkGW2D42QY9lWgrauFQ8xbaNDbtw2h3Pp+Jnvz4BRwyMruJ6nGjp7xSM7Y17MXbu06l4zQ2nY0AfnRfZeIXVksvuq/WNKXz5Fh4IukhSmEWXgtqWraIN6ZPVzmIbtBQ26sy8FW/5xmEJuhCDzT6fT9908Wbp6ZPVzuK4sH1sgh7LtBS0iUPxlKDMerDON6b8aenvFIw7l73pG7uAE9YC/fHSat+Ywpdv4YGgiySFWXQpqG3ZKtrwu8/0941t0FLYqDPXTh7uG4cl6EIMNvt8Pn3TxZulzz+rt28cF7aPTdBjmZaCNnEonhKU+dPG+MaUPy39nYIxe+qxvrELuCQ4ADZvXE35ybfwQNCFcMIsjhFUgSdbbf7Szxdi8WbgW0cBM2d+KrDn9dPUvB8LVm/C9PFDc8p7DEPDrj2Yu/h13HrOCJSXl0fWjqBvJm6zmFQ+fTPMwmdB+e8fLMRzO4CvHgpcc00450oUbI+PQReg0dKXdu5uwbwVb+HaycNDrQmgkQtFL8MshBTktljAKV7e/mAHrry/Fg9cXI3+/XVUF+9MpiXBnLAWqHbTezhv3kup+E9XnYTqoYdF2CLKV9A5VS7mgNjKK/vdmrW44fcbU/HcLxyBC8eeUPDz+tGcv6ilbS7lbWvZZ7b9v7+uwS1/PLjM86bPVuGSj46NsEV2uDY+amnv/v1tOP7mxWgzydvavPads9Cjh55KwWHavK0RE25/LhU//42P4/BD+kTYon8XZr/R0kdJHy11M7LBHFZLZtTU+sbkjqBzqlzMAbGVV5Y+We0stkFz/qKWtrmUt61ln9mWPlntLI4L18ZHLe19aeMOtHnXGdpMMi5W1zy01jfWIMx+o6WPkj5a6mYUghPWAj0xs9o3JncEnVPlYg6Irbyyxy46wje2QXP+opa2uZS3rWWf2fbIF6p847hwbXzU0t5TjuqPbt5Fs26SjIvVvRee4BtrEGa/0dJHSR8tdTMKwSXBAfjzutfx1d9uwJP/ewxGjBgRdXOoAEHn34WZ8xRUvomtvLKbH1qI+14C7joVOO+8cPLyNN8oO+jc0XwFvY9s9vl89pmWvMNcXHrXQix7D/jGUOCqq9zIYc1n/LF9bILOwdOS06d5XAvb+q0NuOy+Wvzx0moMHKjzloKu5rBSvDCHNQSaJ6y8D1h8uJzDGtS2bLV5yfq3cPn89an4lxePxJkj7f6FT3Oul5Z7KWreRx3ls89czHu9+9kX8MMntqfi684egCtPGx9hi7qWz7hhe3yMaz6fi33alqbm/Rh9y9JUvO6mKajozauKRB3xPqzE+4DFiMs5rEFty1abr0ibrHYW26A510vLvRQ176OO8tlnLua9pk9WO4s1ymfcsD0+xjWfz8U+bcuC1Zt8YyJK4n1YifcBixGXc1iD2patNi+5bKRvbIPmXC8t91LUvI86ymefuZj3+sC5A3xjjfIZN2yPj3HN53OxT9syffxQ35iIkngf1pBoXhIMJJcFz3qwDvOnjUHfvlwO7LKgc6pczF2x1eYl69/CFfPXY8llIzF8eDgJ/2HeBzdXWvLQNO+jjvLZZy7msLYvC37g3AEYP173cuB2+YwbtsfHuObzudinbdF8r20iTbY37sWdy97E7KnHolcvncuBAS4JtqpEBN27lcTqDVGjRMKgfvc+2PwjS2lpCUZW9XHyQ0AiYbCtqfD9U1IiGFTZM/D+vOfAAbQBOHDgQKDP66e0tATHDK7M+niG0cfatSYS2LprLxKJhPVt+cl1H3WltTWB9VsbrbyusrJSnD5qcE4F0Vw8p08dNgz9ykowbNiwqJviNFtjWdRKSgSHVOh7XTbP/Uw+aG7GH9ZsRHNzc2jbzFWY7ytEmVSU9cCnxwxx9g87vMJaIBduXB0HLhbQCLPN2gtx/Gntelx9/8H7ft110XCcd4LdZcG57v8wj1dj0z6MuXVZKq6bPRl9KsKvFBz0a7bZD10cA/KxZuNWnH/3mlT8yJVjMfYI3be20Vh0Ka607rco3oPe/mAHTv/RylT89DUfw9GH6kpr0Hq8qLi4VGBRzRVWERkqIstF5DUReUVEZoXdhiC5cOPqOHCxgEaYbdZeiCN9stpZbEOu+z/M41Wz8h3fOCxBv2ab/dDFMSAfM2pqfWONNBZdiiut+y2K96Ar76/1jTXQeryouLhUYDGTKC7BtAK41hhzPIBTAVwlIqMiaEcgXLhxdRy4WEAjzDZrL8Sx+OLhvrENue7/MI/XFROH+cZhCfo12+yHLo4B+Vg0s9o31khj0aW40rrfongPeuDiat9YA63Hi4qLSwUWM4l8SbCIPArg58aYpZkeo3lJMACsfHsTLqmpw5KZYzB0KKvU2RJGAY2gtxFmQRvthThuX7QM81bsw72Te2Ly5MmhbDPX4xlmkZaGXXswd/HruPWcESgvL7e6LT9B95uWllYsf6MeU0YNyinXNBv5tNXFwju/emE1vvfo+/jD5wZj3Lh/WxmlUj5jne0xy8Vjnw2tryuK96C6ze9jRk0tFs6sxuDBg0PZZq60vzfboLWPFjMthR67kmlJcKQTVhEZBuBZAKONMY2ZHqd5wrqxYRdOu+P5VPzs1yfgiIGsFOwil/L5XPPiPzbjv/7v5VS8YOaJOPWowyNsUbS05JO41OeLJU9y1Tv/xBd+UZuKf/+lapwy7EPRNSgL+Rx328fGxWNPuWlq3o/Rtxy81rHupimo6K3rCmYx9sNifM3aufR5VE0OazsRqQDwMICrO5usisjlIrJaRFbX19eH38AszXqwzjcmd7iUz+eaGTUv+8bFRks+iUt9vljyJF3MYc3nuNs+Ni4ee8rNgtWbfGMNirEfFuNr1i4On0cjmbCKSHckJ6v3G2Me6ewxxphfGmPGGWPGDRo0KNwG5mD+tDG+MbnDpXw+1yyZeaJvXGy05JO41OeLJU/yyQ45qx1jjfI57raPjYvHnnIzffxQ31iDYuyHxfiatYvD59HQlwRLcjH7fQC2G2OuzuZ3NC8JBoDaTe9hRk0tnphZjcMOOyzq5lABgs67CDN3Jcx82XzUrPo7vvvIFvzxgiH4yEc+Eso2NecOabnhfdD7yGY/zOe5XcylYg5rMFw89tnQOq5Fsb83NuzCrAfrMH/aGPTtG2w6VlCvJ6790E8xvmbtbNaXCJKaHFYRmQDgOQBrAbTfXfoGY8wTmX5H84R1e+NejJ37dCpec8PpGNCnV4QtomKkJScyk1e31OPsn65KxU989RSMGmJ35YTmnA0txyvoXCObuUuaj2eQeB9W8qP1PIjieNr8/MX+SXGiddzojJocVmPM88YYMcaMMcZUe18ZJ6va3bnsTd+YKAxaciIzuXR+rW9sg+acDS3HK+hcI5u5S5qPZ5BczGEtlvxiDbSeB1EcT5ufv9g/KU60jhu50Dm9dsjsqcf6xkRh0JITmcmjl1X7xjZoztnQcryCzjWymbuk+XgGifdhJT9az4MojqfNz1/snxQnWseNXER+H9ZsaF4SDCSXpdy57E3MnnosevXicmCKhvYc1le31OPS+bV49LJqhFVITWuuF6DneAWda2Qzd0nz8QzSmo1bMaOmFotmVqOqSvdy4Hb5HHfmueVH63kQxfG0+fmL/ZPiROu40ZGaJcFxVFHWA58eMyTSwinFIJEwqN+9D9n8kaW1NYH1WxuRSCS6fGxclJaW4JjBlWoHon5lZTj60AqUlZWFts3S0hKMrOqjcp/sP9CGV/65C21tbVE3hZQpKy1FeY9S1YUxOiopEQyq7JnTB/t8foeS74U7mvdn9V4YdzY/fwXVP3P57EJkS0mJ4JAKd8dbXmEtkJbiKXGXSwEEl5LLg6K9QMSW7bvxsR8+m4pXXncahgyojLBF0dJyw3sWXdInigJl5A6tnzmieA/Sui/SaX9vpuLgUj/kFVZLtBRPibtcCiDEIbk8V9oLRFz3yDrfuNhoueE9iy7pE0WBMnKH1s8cUbwHad0X6bS/N1NxiEM/5IS1QFqKp8RdLgUQ4pBcnivtBSLuuWC0b1xstNzwnkWX9ImiQBm5Q+tnjijeg7Tui3Ta35upOMShH3JJcAA2b2vENQ+txb0XnoA+ffpE3ZzYyqUAgivJ5UHSUsQnk9XvbsGMmlosnlmNIUOGhLJNzf1g5+4WzFvxFq6dPDzUvN6Ogi4sYnOfa+/jQVn82gZced/rWHLpCBxzzDFRN4eU2bPnAB5buxXnj61C9+7do25OShRFipqa92PB6k2YPn6o2joiLN5EGrjSDzMtCeaEtUAf7GzGKbc9k4pXXT8Jh/brHV2DqChpz++LIodV8z7R3LZC2MyTcSkHpxAbG3bhtDueT8XPfn0CjhjYN8IWkSZxHTvy4UIOKxHlhjmslsx5/DXfmCgM2vP7oshh1bxPNLetEDbzZOKQg5ONWQ/W+cZU3OI6duTDhRxWIgoGJ6wFuv2zx/vGRGHQnt8XRQ6r5n2iuW2FsJknE4ccnGzMnzbGN6biFtexIx8u5LASUTC4JDgAH+xsxpzHX8Ptnz0evXtzOTBFQ3O+JpBcFnzdI+twzwWjUVkZzi1tNO8TzW0rhM08GVdycAq1sWEXZj1Yh/nTxqBvXy4Hpn8V17EjH8WS105ULLgk2KKy0lIcd2ilUzd5JyoGmm+UraVtra0JrN/aiEQiEWk76KASEXTvVhJ533BdImFQv3sfXPjDfC5aWxP4R30zz1kkx9H+veP9B6y4i+t5SsHiFdYCNTbtw5hbl6XiutmT0aeiZ4QtomKkvRBHFEWXNBfp0dK2oPsNiy4VbvO2Rky4/blU/Pw3Po7DD2H1+VzFtb+0tLRi5JwnU/H6OZ9EWVlx/rFc+/sedS2u5ynlj1dYLalZ+Y5vTBQG7YU4oii6pLlIj5a2Bd1vWHSpcNc8tNY3puzEtb8sf6PeNy4m2t/3qGtxPU8peJywFuiKicN8Y6IwaC/EEUXRJc1FerS0Leh+w6JLhbv3whN8Y8pOXPvLlFGDfONiov19j7oW1/OUgsclwQGIophM3BRLMRWbtBefqN30HmbU1OKJmdU47LDDQtmm5n2ipW1BF3Cx+bqKZZxYs3ErZtTUYtHMalRVVUXdHGfFtThRS0srlr9RjymjBqmqnRHF+dnUvB8LVm/C9PFD0aMHJzsuKpZxPWqu7OdMS4I5YS1QU/N+jL5laSped9MUVPTmoJkL5jAUTvs+/GBnM0657ZlUvOr6STi0n92K2przmzS3rRBxfV1hiuJciSPtY2LcRLG/mc9LlB2XxkPmsFqyYPUm35i6xhyGwmnfh3Mef803tkFzfpPmthUirq8rTFGcK3GkfUyMmyj2N/N5ibITh/GQE9YCTR8/1DemrjGHoXDa9+Htnz3eN7ZBc36T5rYVIq6vK0xRnCtxpH1MjJso9jfzeYmyE4fxkEuCA9Cwaw/mLn4dt54zAuXl5VE3x0murK3PVZivS3u+1up3t2BGTS0Wz6zGkCFDQtmm1lwvQE/bgu43Nvuh9j4elLrN72NGTS0WzqzG4MGDo26Os/i+Ei7msBLp5cr7J3NYLWEOBWUSZs6A9vyEjQ27cNodz6fiZ78+AUcM7Gt1m5rzKbW0TUs7suFSWwvx3o4mnPqDFan4xW9OxGH9eaWakrSP9WEqljGBqFAujRvMYbWEORSUSZg5A9rzE2Y9WOcb26A5n1JL27S0IxsutbUQNz76im9MxU37WB+mYhkTiAoVh3GDE9YCMYeCMgkzZ0B7fsL8aWN8Yxs051NqaZuWdmTDpbYW4ief+7BvTMVN+1gfpmIZE4gKFYdxg0uCA6AlF430CTOnR2teU7uNDbsw68E6zJ82Bn372l0O3E5zzoaWtmlpRzZcamsh3tvRhBsffQU/+dyHUVHBD+H0r7SP9WEqljGBqFCujBtcEmzR/tY2vP5eI9ra2qJuCilTUiIYVNkzlMEhzG3lo7x7dxw1sALdu3cPbZulpSUYWdVH5QcZLW3T0o5suNTWQvTp2RNnHl+Fnj17Rt0UqxIJg/rd+2DrD+e2nz8q2sd6ioe4nj/kJl5hLVBj0z6MuXVZKq6bPRl9KuL9IYMoVw279mDc95en4tXf+gQG9mVFbaKO9uw5gFHfXZKKX/32mSgvD++PPGGxXQTEpSIjlB8WXbKH50+8uHQ8eYXVkpqV7/jGRATMXfy6b0xESY+t3eobx4XtIiBxKDJC/lh0yR6eP/ESh+PJCWuBrpg4zDcmIuDWc0b4xkSUdP7YKt84LmwXAYlDkRHyx6JL9vD8iZc4HE8uCQ7A2x/swJX31+KBi6vRv3//qJsTW64kjFPnojhPNPeZuBYL2b+/DS9t3IFTjuqPbt26Rd0cJ23e1ohrHlqLey88AX369Im6OdbYPj+Dfv64nrMu27PnAB5buxXnj60KtT5CMdD8/km5c2X8yrQkmBPWAjE3Lxwurb+nfxdFrrfmPhPX3Kv9+9tw/M2L0WaAbgK89p2z0KMHJ6252N64F2PnPp2K19xwOgb06RVhiwiI7znrMo43RNnR/HmoI+awWsLcvHDEYf19MYsi11tzn4lr7tVLG3egzfsbaJtJxpSbO5e96RtTNOJ6zrqM4w1RdjR/HsoWJ6wFYm5eOOKw/r6YRZHrrbnPxDX36pSj+qOb90fbbpKMKTezpx7rG1M04nrOuozjDVF2NH8eyhaXBAegYdcezF38Om49ZwTKy7kc2BbmU7itsWkfala+gysmDgvt/pKa+4wr+SS5Yg5r4bY37sWdy97E7KnHolcvLgfWIq7nrMs43hBlR/PnoXRcEmxRn1498fmThsb+Ju9R483S3VbWoxQfPfoQlJaWhrZNzX2mtLQEI6v6xO6Db2lpCY4ZXGnldfFG9nppPDZBt0nLOatxX0fF5nhDFCeaPw9lI7xPjjHFpH+irvE8KQ42Czu4VDSiEOlFl+5ftcmJoksaj43GNgUhrq8rH9wXRMWDf5IqEJP+ibrG86Q42CzsEIeiEdlwseiSxmOjsU1BiOvrygf3BVHx4IS1QEz6J+oaz5PiYLOwQxyKRmTDxaJLGo+NxjYFIa6vKx/cF0TFg0WXAhBFMRlyQ5hJ7toT6rds343rHlmHey4YjcrKyqib06kw96GWG94H/Zpt7sNiKbDy9gc7cOX9tXjg4mr07+/GH3c0jj8u9W0X26FBS0srlr9RjymjBoVaH4GI7MhUdIkT1gK1tLRi5JwnU/H6OZ9EWRkHTQo3v0Z7Lk/Drj0Y9/3lqXj1tz6BgX11VdQOcx/u2XMAo767JBW/+u0zUV4e/qRVe79J19qawEduXYrdLa2oLCvF32dPQWlp/BYJ7dzdgurvPZWKa288A/0qyyJsEQFunSvFgrURiOKHVYItWf5GvW9MxSvM/BrtuTxzF7/uG2sQ5j58bO1W3zgs2vtNug31Tdjd0goA2N3Sig31TRG3yI55K97yjSkaLp0rxYK1EYiKByesBZoyapBvTMUrzPwa7bk8t54zwjfWIMx9eP7YKt84LNr7TbrjBleg0lu9UllWiuMGV0TcIjuunTzcN6ZouHSuFAvWRiAqHlwSHADmUFAmzGE9qGHXHsxd/DpuPWcEyst1LQduxxxWff0mXWtrAhvqm3Dc4IpY33dx5+4WzFvxFq6dPBxlZVwOrIVL50qxKJa8dqJiwRxWIiIiIiIiUok5rEREREREROQUTliJiIiIiIhIJU5YiYiIiIiISCVOWImIiIiIiEglTliJiIiIiIhIJU5YiYiIiIiISCVOWImIiIiIiEglTliJiIiIiIhIJU5YiYiIiIiISCVOWImIiIiIiEglTliJiIiIiIhIJU5YiYiIiIiISCVOWImIiIiIiEglTliJiIiIiIhIJU5YiYiIiIiISCVOWImIiIiIiEglMcZE3YYuiUg9gHejbkcXBgJoiLoRVPTYD0kD9kPSgn2RNGA/JA1c6IdHGmMGdfymExNWF4jIamPMuKjbQcWN/ZA0YD8kLdgXSQP2Q9LA5X7IJcFERERERESkEiesREREREREpBInrMH5ZdQNIAL7IenAfkhasC+SBuyHpIGz/ZA5rERERERERKQSr7ASERERERGRSpyw5kBEzhKR10Vkg4hc38nPRUR+6v28TkTGRtFOircs+uFFXv+rE5GVInJiFO2k+OuqL6Y97mQRaRORz4XZPioO2fRDEZkkIrUi8oqIrAi7jRR/Wbw39xWRx0TkZa8f/m8U7aR4E5EaEflARNZl+LmTcxVOWLMkIt0AzAMwFcAoAF8UkVEdHjYVwLHe1+UA7gm1kRR7WfbDfwCYaIwZA+AWOJyzQHpl2RfbH/cDAE+G20IqBtn0QxHpB+BuAJ8xxnwYwOfDbifFW5bj4VUAXjXGnAhgEoA7RaRHqA2lYjAfwFk+P3dyrsIJa/ZOAbDBGPO2MWY/gAUAzu3wmHMB/NokvQign4hUhd1QirUu+6ExZqUxZocXvgjg8JDbSMUhmzERAL4C4GEAH4TZOCoa2fTDCwE8YozZCADGGPZFClo2/dAAqBQRAVABYDuA1nCbSXFnjHkWyb6ViZNzFU5Ys/chAJvS4s3e93J9DFEhcu1jlwBYZLVFVKy67Isi8iEAnwXwixDbRcUlmzHxOAD9ReQZEXlJRKaH1joqFtn0w58DOB7AFgBrAcwyxiTCaR5RipNzldKoG+AQ6eR7HUssZ/MYokJk3cdE5BNITlgnWG0RFats+uJdAL5pjGlLXlQgClw2/bAUwEkAzgDQC8ALIvKiMeYN242jopFNP/wkgFoApwMYDmCpiDxnjGm03DaidE7OVThhzd5mAEPT4sOR/CtZro8hKkRWfUxExgC4F8BUY8y2kNpGxSWbvjgOwAJvsjoQwNki0mqM+VMoLaRikO17c4MxphlAs4g8C+BEAJywUlCy6Yf/C+A2k7yf5AYR+QeAkQBWhdNEIgCOzlW4JDh7fwNwrIgc5SXJ/xeAP3d4zJ8BTPcqcJ0KYJcxZmvYDaVY67IfisgRAB4B8N+8gkAWddkXjTFHGWOGGWOGAfgDgCs5WaWAZfPe/CiAj4tIqYiUA/gogNdCbifFWzb9cCOSV/khIoMBjADwdqitJHJ0rsIrrFkyxrSKyJeRrHTZDUCNMeYVEfmS9/NfAHgCwNkANgDYg+Rf04gCk2U//DaAQwDc7V3ZajXGjIuqzRRPWfZFIquy6YfGmNdEZDGAOgAJAPcaYzq95QNRPrIcD28BMF9E1iK5LPObxpiGyBpNsSQiDyBZhXqgiGwGcDOA7oDbcxVJrkwgIiIiIiIi0oVLgomIiIiIiEglTliJiIiIiIhIJU5YiYiIiIiISCVOWImIiIiIiEglTliJiIiIiIhIJU5YiYgoZyJyo4i8IiJ1IlIrIh/1vv+OiAyMsF23e+26vcDnudq7Z2fREpGVef7eOBH5aQHbbcr3d7t43mEicqGN5yYiInt4WxsiIsqJiIwH8CMAk4wx+7wJag9jzBYReQfAuHzuLygipcaY1gLb1ghgkDFmX4HP8w4yvA4R6WaMaSvk+TWL+vWJSJMxpsLC804C8HVjzKeDfm4iIrKHV1iJiChXVQAa2ieFxpgGY8yWtJ9/RUTWiMhaERkJACJyioisFJG/e/+O8L5/sYg8JCKPAVgiIr1FpEZE/uY99tyOG5ek20VknbeNC7zv/xlAbwB/bf9e2u9k2n43EbnDe546EfmKiHwVwBAAy0Vkufe4JhH5roj8FcB4EZkmIqu8q8v/5z1PNxGZn9aur3m/+1URedV7/gXe9zp9nSLy4bTnrRORYzt5/V/0nn+diPwg7ftNIvI9EXlZRF4UkcGd/O4cEfmNiDwtIm+KyGXe9yeJyHIR+R2Ate3Pl/azZ0TkDyKyXkTuFxHxfnaytz9f9tpd6T3+8S62VyEiT6X1k387zp20fbq3T14Wkd943ztHRP7q7cNl7a9ZRCZ6+7DW+1klgNsAfNz73te62h4RESlhjOEXv/jFL37xK+svABUAagG8AeBuABPTfvYOgK94/78SwL3e//sAKPX+PxnAw97/LwawGcAAL54LYJr3/37eNnp32P5/AlgKoBuAwQA2AqjyftaUoc2Ztn8FgIfTfjYg7XUMTPt9A+AL3v+PB/AYgO5efDeA6QBOArA07Xf6ef9uAdCzw/c6fZ0AfgbgIu/7PQD06vA6hnivdxCAUgBPAzgvrY3neP//IYDZneyHOQBeBtALwEAAm7znnASgGcBRaY9t8v6dBGAXgMOR/EP3CwAmeO17G8DJ6fvYe/zjXWyvFEAf7zEDAWzAwVVf/3YMAXwYwOvtxyTtOPVP+71LAdzp/f8xAP+R1l//pV384he/+MUvd754hZWIiHJijGlCcnJ2OYB6AA+KyMVpD3nE+/clAMO8//cF8JCIrAPwYyQnIO2WGmO2e/8/E8D1IlIL4BkAZQCO6NCECQAeMMa0GWPeB7ACwMldNDvT9icD+IXxliKntaOjNiQntgBwBpKv/29eO88AcDSSk7ejReRnInIWgEbv8XUA7heRaQDalzxnep0vALhBRL4J4EhjzN4O7TgZwDPGmHqvzfcDOM372X4Aj3v/T9/3HT1qjNlrksudlwM4xfv+KmPMPzL8zipjzGZjTALJP1YMAzACwFZjzN8AwBjTaDpf0t3Z9gTAXBGpA7AMwIeQ/ONDJqcD+IP3HOnH6XAAT4rIWgDfwMHj+hcAP/KulvfL0C4iInIAJ6xERJQzb7L4jDHmZgBfRvKqZ7v2/NE2JK9sAcAtAJYbY0YDOAfJCVq75rT/C4D/NMZUe19HGGNe67B5yaPJmbYvSF6Z7EqLOZjXKQDuS2vjCGPMHGPMDgAnIjkBvQrAvd7jPwVgHpKT3JdEpDTT6zTG/A7AZwDsRXIidnoOr/2AMab9taTv+446vt72uLnjA9Ok5wS3P3e2+66z7V2E5FXik4wx1QDex7/2iY4ybetnAH5ujDkBwMz25zDG3IbkFddeAF4Ub2k6ERG5hxNWIiLKiYiM6JBbWQ3g3S5+rS+Af3r/v9jncU8imQPbniP5kU4e8yyAC7yc0UFIXmFclef2lwD4kjeJhIgM8L6/G0Blhud6CsDnROTQ9t8RkSMlWXyqxBjzMICbAIwVkRIAQ40xywFch+Ty34pMr1NEjgbwtjHmpwD+DGBMh23/FcBEERkoIt0AfBHJK8y5OFdEykTkECSXyf4tx99vtx7AEBE52Wt7Zft+zGJ7fQF8YIw5ICKfAHBkF9t6CsAXvOdIP07px/V/2h8sIsONMWuNMT8AsBrASPgfUyIiUooTViIiylUFgPvaCwkBGIVkrqKfHwL4voj8Bcnc00xuAdAdQJ23fPeWTh7zRySX2b6MZA7ndcaY9/Lc/r1I5oTWicjLANpve/JLAIvEK7qUzhjzKoDZSBaJqkMyn7YKyWWtz3jLfOcD+Ja3rd96S1b/DuDHxpidPq/zAgDrvOcYCeDXHba91Xve5d7rX2OMebSL197RKgALAbwI4BbzrwWzsmaM2e+192fevluKzq+Sdra9+wGME5HVSF5tXd/Ftl4B8D0AK7xt/cj70Rwkl3o/ByC9ovPVXlGql5G8Wr0IyT7T6hVtYtElIiJH8LY2RERERUJE5iBZ1OiOOG6PiIjih1dYiYiIiIiISCVeYSUiIiIiIiKVeIWViIiIiIiIVOKElYiIiIiIiFTihJWIiIiIiIhU4oSViIiIiIiIVOKElYiIiIiIiFTihJWIiIiIiIhU+v8PqVx/e3yMXAAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# 1. Anaylze possible pattern/relationship between\n", "# actress share and average rating:\n", "plt.figure(figsize=(16,8))\n", "plt.scatter(dat_2000.proportion, dat_2000.averageRating, s = 5)\n", "plt.xlabel(\"Share of actresses on principal cast\")\n", "plt.ylabel(\"Average Rating\")\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Result:\n", "- No clear pattern identifiable\n", "- Many values/outcomes of the actress share covers a wide range of average ratings\n", "----" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
proportionaverageRating
proportion1.000000-0.072625
averageRating-0.0726251.000000
\n", "
" ], "text/plain": [ " proportion averageRating\n", "proportion 1.000000 -0.072625\n", "averageRating -0.072625 1.000000" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# We already see, that idea of linearly predicting the average rating on actress share cannot be done by a linear regression model.\n", "# (There is no linear relationship)\n", "\n", "# However, we will affirm this conclusion by some statistics:\n", "\n", "# 2. Compute pearson correlation coefficient for share of actresses on the principal cast and average movierating:\n", "dat_2000[[\"proportion\", \"averageRating\"]].corr()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Result:\n", "- the correlation coefficient has a negative sign, but is too low for presenting a meaningfullness correlation between average rating and actress share.\n", "---------" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
OLS Regression Results
Dep. Variable: y R-squared: 0.005
Model: OLS Adj. R-squared: 0.005
Method: Least Squares F-statistic: 518.6
Date: Mon, 07 Feb 2022 Prob (F-statistic): 1.72e-114
Time: 10:11:14 Log-Likelihood: -1.7020e+05
No. Observations: 97801 AIC: 3.404e+05
Df Residuals: 97799 BIC: 3.404e+05
Df Model: 1
Covariance Type: nonrobust
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
coef std err t P>|t| [0.025 0.975]
const 6.0990 0.008 760.018 0.000 6.083 6.115
x1 -0.4049 0.018 -22.772 0.000 -0.440 -0.370
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
Omnibus: 3067.568 Durbin-Watson: 1.978
Prob(Omnibus): 0.000 Jarque-Bera (JB): 3365.381
Skew: -0.445 Prob(JB): 0.00
Kurtosis: 3.181 Cond. No. 4.64


Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified." ], "text/plain": [ "\n", "\"\"\"\n", " OLS Regression Results \n", "==============================================================================\n", "Dep. Variable: y R-squared: 0.005\n", "Model: OLS Adj. R-squared: 0.005\n", "Method: Least Squares F-statistic: 518.6\n", "Date: Mon, 07 Feb 2022 Prob (F-statistic): 1.72e-114\n", "Time: 10:11:14 Log-Likelihood: -1.7020e+05\n", "No. Observations: 97801 AIC: 3.404e+05\n", "Df Residuals: 97799 BIC: 3.404e+05\n", "Df Model: 1 \n", "Covariance Type: nonrobust \n", "==============================================================================\n", " coef std err t P>|t| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "const 6.0990 0.008 760.018 0.000 6.083 6.115\n", "x1 -0.4049 0.018 -22.772 0.000 -0.440 -0.370\n", "==============================================================================\n", "Omnibus: 3067.568 Durbin-Watson: 1.978\n", "Prob(Omnibus): 0.000 Jarque-Bera (JB): 3365.381\n", "Skew: -0.445 Prob(JB): 0.00\n", "Kurtosis: 3.181 Cond. No. 4.64\n", "==============================================================================\n", "\n", "Notes:\n", "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n", "\"\"\"" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Even though there is no correlation, hence no linear relationship,\n", "# a linear regression would not be a suitable model.\n", "# However, just as safty check and regarding our research question we will implement it.\n", "\n", "# 2. Linear regression of average rating on the share of actresses on the principal cast\n", "y = dat_2000[[\"averageRating\"]].values\n", "x = dat_2000[[\"proportion\"]].values\n", "x_ = sm.add_constant(x) #adding a constant\n", "\n", "reg = sm.OLS(y, x_).fit()\n", "y_pred = reg.predict(x_)\n", "reg.summary()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ " Result:\n", " - the estimated coefficient for share of actresses on the principal cast is: -0.4049.\n", " - if the actress share increases by 10%-point (0.10), then on average the average rating decreases by 0.04. \n", " - The estimate is singificant and in line with our pearson-correlation coefficient, that also implies a small neg. relationship. \n", " \n", "- However, the model has no predictive power. The R-squared value is super low (0.005), hence our model does not explain the variation in the average rating. \n", " Therefore, the aim of well predicting the avarage rating on the share of actresses on the principal cast cannot be fullfilled.\n", " \n", " ---------------------" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Predicting average rating on share of actresses on the principal cast, movie duration and genre" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ " Idea: Increase predictive accuracy by including additional explanatory variables as RuntimeMinutes and Genre.\n", " (f.e. Persons who are watching certain genres differ regarding their preference/awareness for actress share. Hence, genre could have an impact on their rating" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### 1) Checking how many types of genre does the data set includes" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Number of genres in data set: 951\n", "\n", "Top 10 genres regarding their number movies:\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
genrecounts
0Drama19602
1Comedy8339
2Documentary4657
3Comedy,Drama4642
4Horror3678
5Drama,Romance2960
6Thriller2274
7Comedy,Romance2271
8Comedy,Drama,Romance2195
9Drama,Thriller1615
\n", "
" ], "text/plain": [ " genre counts\n", "0 Drama 19602\n", "1 Comedy 8339\n", "2 Documentary 4657\n", "3 Comedy,Drama 4642\n", "4 Horror 3678\n", "5 Drama,Romance 2960\n", "6 Thriller 2274\n", "7 Comedy,Romance 2271\n", "8 Comedy,Drama,Romance 2195\n", "9 Drama,Thriller 1615" ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "genre_counts = dat_2000.genres.value_counts()\n", "genre_counts = genre_counts.reset_index().rename(columns = {'index':'genre', 'genres' : \"counts\"})\n", "print(f\" Number of genres in data set: {len(genre_counts)}\")\n", "print(\"\")\n", "print(\"Top 10 genres regarding their number movies:\")\n", "genre_counts.head(10)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ " We can see that the data set consists on 951 genres. Where the majority of those are genre-overlaps such as Comedy-Drama.\n", " \n", " Splitting those combinations and allowing the movies to capture several genres\n", " leads to a dependencies. Further, includung all genres as dummy variables, leads to 950 dummy variables,\n", " that is messy.\n", "\n", " Therfore, we stick to only those movies who strictly capture only one gerne.\n", " This reduces the number of dummy variables and could lead to more discrimant power, because the movies and their type of viewers differ more from each other and hence, differ in their\n", " preference for actress share." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### 2) Using only movies that contain a single genreand prepare the data set for a linear regression" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [], "source": [ "# Using the csv file \"data_movie_genre\", that contains all single genres regarding their movie.id\n", "data_movie_genre = pd.read_csv('data_movie_genre.csv')\n", "\n", "# Drop movie.id and keep unique single genres in a list\n", "single_genres = list(data_movie_genre.genre.unique().tolist())\n", "\n", "# Keeping only those single genres in the list, that are covered between 2000 and 2020\n", "single_genres = dat_2000[dat_2000[\"genres\"].isin(single_genres)].genres.unique().tolist()\n", "\n", "# Keeping only single genres in the data set\n", "dat_2000_gen = dat_2000[dat_2000[\"genres\"].isin(single_genres)]" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "# Create Dummy variables for those single genres.\n", "# Dropping the last gerne \"western\", due to multicollineartiy/dummy trap\n", "dat_2000_gen = dat_2000_gen.join(pd.get_dummies(dat_2000_gen[\"genres\"] ))\n", "dat_2000_gen.drop(columns = \"Drama\", inplace = True)" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
OLS Regression Results
Dep. Variable: y R-squared: 0.218
Model: OLS Adj. R-squared: 0.218
Method: Least Squares F-statistic: 487.0
Date: Mon, 07 Feb 2022 Prob (F-statistic): 0.00
Time: 10:11:15 Log-Likelihood: -72342.
No. Observations: 43680 AIC: 1.447e+05
Df Residuals: 43654 BIC: 1.450e+05
Df Model: 25
Covariance Type: nonrobust
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
coef std err t P>|t| [0.025 0.975]
const 6.1531 0.040 152.550 0.000 6.074 6.232
Comedy 0.0025 0.000 6.342 0.000 0.002 0.003
Horror -0.1816 0.024 -7.551 0.000 -0.229 -0.134
Action -0.8941 0.040 -22.516 0.000 -0.972 -0.816
Thriller -0.6837 0.180 -3.805 0.000 -1.036 -0.332
Documentary -0.5107 0.087 -5.882 0.000 -0.681 -0.341
Romance -0.6602 0.066 -10.001 0.000 -0.790 -0.531
Adult 0.2165 0.109 1.990 0.047 0.003 0.430
Family -0.6678 0.017 -40.124 0.000 -0.700 -0.635
Sci-Fi -0.4855 0.067 -7.259 0.000 -0.617 -0.354
Fantasy 0.8538 0.022 39.271 0.000 0.811 0.896
Animation -0.4056 0.052 -7.735 0.000 -0.508 -0.303
Crime -0.4284 0.087 -4.913 0.000 -0.599 -0.258
Adventure 0.1616 0.127 1.277 0.202 -0.086 0.410
Mystery -1.7754 0.023 -76.888 0.000 -1.821 -1.730
Musical 0.9499 0.124 7.653 0.000 0.707 1.193
Biography -0.0975 0.105 -0.929 0.353 -0.303 0.108
War -0.2995 0.085 -3.522 0.000 -0.466 -0.133
History -0.3798 1.268 -0.299 0.765 -2.865 2.106
Western -2.5504 0.732 -3.483 0.000 -3.986 -1.115
Music -0.4086 0.044 -9.317 0.000 -0.495 -0.323
Sport -1.2113 0.066 -18.424 0.000 -1.340 -1.082
News -0.0012 0.178 -0.007 0.995 -0.350 0.347
Reality-TV -0.9539 0.028 -33.925 0.000 -1.009 -0.899
runtimeMinutes -0.3767 0.153 -2.462 0.014 -0.677 -0.077
proportion -1.5547 0.116 -13.431 0.000 -1.782 -1.328
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
Omnibus: 1068.241 Durbin-Watson: 1.969
Prob(Omnibus): 0.000 Jarque-Bera (JB): 1382.625
Skew: -0.305 Prob(JB): 5.84e-301
Kurtosis: 3.622 Cond. No. 1.96e+04


Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The condition number is large, 1.96e+04. This might indicate that there are
strong multicollinearity or other numerical problems." ], "text/plain": [ "\n", "\"\"\"\n", " OLS Regression Results \n", "==============================================================================\n", "Dep. Variable: y R-squared: 0.218\n", "Model: OLS Adj. R-squared: 0.218\n", "Method: Least Squares F-statistic: 487.0\n", "Date: Mon, 07 Feb 2022 Prob (F-statistic): 0.00\n", "Time: 10:11:15 Log-Likelihood: -72342.\n", "No. Observations: 43680 AIC: 1.447e+05\n", "Df Residuals: 43654 BIC: 1.450e+05\n", "Df Model: 25 \n", "Covariance Type: nonrobust \n", "==================================================================================\n", " coef std err t P>|t| [0.025 0.975]\n", "----------------------------------------------------------------------------------\n", "const 6.1531 0.040 152.550 0.000 6.074 6.232\n", "Comedy 0.0025 0.000 6.342 0.000 0.002 0.003\n", "Horror -0.1816 0.024 -7.551 0.000 -0.229 -0.134\n", "Action -0.8941 0.040 -22.516 0.000 -0.972 -0.816\n", "Thriller -0.6837 0.180 -3.805 0.000 -1.036 -0.332\n", "Documentary -0.5107 0.087 -5.882 0.000 -0.681 -0.341\n", "Romance -0.6602 0.066 -10.001 0.000 -0.790 -0.531\n", "Adult 0.2165 0.109 1.990 0.047 0.003 0.430\n", "Family -0.6678 0.017 -40.124 0.000 -0.700 -0.635\n", "Sci-Fi -0.4855 0.067 -7.259 0.000 -0.617 -0.354\n", "Fantasy 0.8538 0.022 39.271 0.000 0.811 0.896\n", "Animation -0.4056 0.052 -7.735 0.000 -0.508 -0.303\n", "Crime -0.4284 0.087 -4.913 0.000 -0.599 -0.258\n", "Adventure 0.1616 0.127 1.277 0.202 -0.086 0.410\n", "Mystery -1.7754 0.023 -76.888 0.000 -1.821 -1.730\n", "Musical 0.9499 0.124 7.653 0.000 0.707 1.193\n", "Biography -0.0975 0.105 -0.929 0.353 -0.303 0.108\n", "War -0.2995 0.085 -3.522 0.000 -0.466 -0.133\n", "History -0.3798 1.268 -0.299 0.765 -2.865 2.106\n", "Western -2.5504 0.732 -3.483 0.000 -3.986 -1.115\n", "Music -0.4086 0.044 -9.317 0.000 -0.495 -0.323\n", "Sport -1.2113 0.066 -18.424 0.000 -1.340 -1.082\n", "News -0.0012 0.178 -0.007 0.995 -0.350 0.347\n", "Reality-TV -0.9539 0.028 -33.925 0.000 -1.009 -0.899\n", "runtimeMinutes -0.3767 0.153 -2.462 0.014 -0.677 -0.077\n", "proportion -1.5547 0.116 -13.431 0.000 -1.782 -1.328\n", "==============================================================================\n", "Omnibus: 1068.241 Durbin-Watson: 1.969\n", "Prob(Omnibus): 0.000 Jarque-Bera (JB): 1382.625\n", "Skew: -0.305 Prob(JB): 5.84e-301\n", "Kurtosis: 3.622 Cond. No. 1.96e+04\n", "==============================================================================\n", "\n", "Notes:\n", "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n", "[2] The condition number is large, 1.96e+04. This might indicate that there are\n", "strong multicollinearity or other numerical problems.\n", "\"\"\"" ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Linear Regression Average rating on all genre dummies, runtimeMinutes and actress share\n", "\n", "# include runtimeMinutes and proportion into the list of single genres,\n", "# as reference for the predictor variables\n", "single_genres.remove(\"Drama\")\n", "single_genres.append(\"runtimeMinutes\")\n", "single_genres.append(\"proportion\")\n", "\n", "y = dat_2000_gen[[\"averageRating\"]].values\n", "x = dat_2000_gen[dat_2000_gen.columns.intersection(single_genres)].values\n", "x = sm.add_constant(x)\n", "\n", "single_genres = [\"const\"] + single_genres\n", "\n", "reg = sm.OLS(y, x).fit()\n", "reg.summary(xname = single_genres)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Result:\n", "- the estimated coefficient of the share on actresses on principal cast is: -1.55 and significant.\n", "Hence, if the share in actress by movies of genre drama increases by 10%-points,\n", "the average rating would decrease by 0.155.\n", "\n", "- The highest significant effect of actress share on average rating provides the genre Western. An increase of actress share by 10%-points, would lead to an decrease in average rating by (-1.55 + -2.55)/10 = 0.41.\n", "\n", "In comparison to the first linear model, now with more explainatory variables\n", "we can see a better model fit. However, the R-squarded is very low: 0.22.\n", "Hence, our model cannot predict well the variation in avarage rating.\n", "\n", "Moreover, we can see that including genres into the model reveals a higher model fit.\n", "This and the significance of some dummy variables incentivices to for further analysis, using only data within single genres." ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.8" } }, "nbformat": 4, "nbformat_minor": 4 }