TicTacToe_mcts.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "iA8rmV9drI2D"
   },
   "source": [
    "# Reinforcement Learning - Projet\n",
    "\n",
    "This is a notebook that implement Tic Tac Toe game with use of the Monte-Carlo tree search (MCTS) algorithm.\n",
    "This is part of the Reinforcement Learning course of E. LePennec, professor at École Polytechnique.\n",
    "\n",
    "Cédric Allain et Alexis Gerbeaux - février 2020"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sources:\n",
    "\n",
    "    - Tic Tac Toe game: https://gist.github.com/eaorak/3966315\n",
    "    \n",
    "    - MCTSPY package: https://pypi.org/project/mctspy/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "kF5jUqpK6_ul"
   },
   "source": [
    "## Packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set to True if use in Google Colab\n",
    "use_colab = False\n",
    "\n",
    "# Set to True if want to save results in Google Drive\n",
    "use_drive = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "raq4JCV-rI2K",
    "outputId": "8e62d5e7-636b-4921-a79d-09b5067d2413"
   },
   "outputs": [],
   "source": [
    "from __future__ import division\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random\n",
    "import itertools\n",
    "import sys\n",
    "import time\n",
    "from tqdm import tqdm\n",
    "from collections import Counter\n",
    "from copy import deepcopy\n",
    "from functools import reduce\n",
    "import operator\n",
    "\n",
    "# MCTSPY \n",
    "if use_colab:\n",
    "    !pip install mctspy\n",
    "from mctspy.tree.nodes import TwoPlayersGameMonteCarloTreeSearchNode\n",
    "from mctspy.games.examples.tictactoe import TicTacToeGameState\n",
    "\n",
    "# Graphes\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set(color_codes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "if use_drive:\n",
    "    from google.colab import drive\n",
    "    drive.mount('drive')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "y8FE-tdyrI2e"
   },
   "source": [
    "## Tic tac toe (jeu du morpion)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PDE3Q_vOQN8A"
   },
   "source": [
    "In this notebook, we implement a game of Tic Tac Toe.\n",
    "We create it so 2 automatic players can play together, with different level of difficulty: random, smart and mcts.\n",
    "We used the following implementation of Tic Tac Toe game, that we adapted to our needs: https://gist.github.com/eaorak/3966315\n",
    "\n",
    "The level 'mcts' corresponds to the level of difficulty where the player choose its actions based on a Monte-Carle tree search.\n",
    "Do to so, we use the package mctspy: https://pypi.org/project/mctspy/, with its own TicTacToe example (https://github.com/int8/monte-carlo-tree-search/blob/master/mctspy/games/examples/tictactoe.py)\n",
    "\n",
    "We also use some of the code in the package MCTS (https://github.com/pbsinclair42/MCTS) and in its Tic Tac Toe implementation (https://github.com/pbsinclair42/MCTS/blob/master/naughtsandcrosses.py)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "kcLC3SwFQN8D"
   },
   "outputs": [],
   "source": [
    "class TicTacToeState():\n",
    "    def __init__(self, board=None, first_player=1):\n",
    "        if board is not None:\n",
    "            if len(board.shape) != 2 or (board.shape[0] != board.shape[1]):\n",
    "                raise ValueError(\"Only 2D square boards allowed\")\n",
    "            else:\n",
    "                self.board = np.array(board)\n",
    "        else:\n",
    "            self.board = np.zeros((3,3))\n",
    "            \n",
    "        self.board_size = self.board.shape[0]\n",
    "        \n",
    "        if first_player not in [1,-1]:\n",
    "            raise ValueError(\"First player must be equal to 1 (player) or -1 (computer).\")\n",
    "        #self.currentPlayer = first_player\n",
    "        self.next_to_move = first_player\n",
    "        \n",
    "        self.corners = [[0,0],\n",
    "                        [0, self.board_size-1],\n",
    "                        [self.board_size-1, 0],\n",
    "                        [self.board_size-1, self.board_size-1]]\n",
    "        self.center = [int(self.board_size/2), int(self.board_size/2)]\n",
    "    \n",
    "    \n",
    "    def get_legal_actions(self):\n",
    "        \"\"\"Given a current state, return all the posssible actions\"\"\"\n",
    "        return [Action(player=self.next_to_move, x=idx[0], y=idx[1])\n",
    "                for idx in np.argwhere(self.board==0)]\n",
    "\n",
    "\n",
    "    def move(self, action):\n",
    "        newState = deepcopy(self)\n",
    "        newState.board[action.x][action.y] = action.player\n",
    "        newState.next_to_move = self.next_to_move * -1\n",
    "        return newState\n",
    "    \n",
    "    \n",
    "    def takeRandomAction(self):\n",
    "        return self.move(action=random.choice(self.get_legal_actions()))\n",
    "    \n",
    "    \n",
    "    def takeSmartAction(self):\n",
    "        \"\"\" Play a 'smart' action: \n",
    "            - win if it can\n",
    "            - block opponent if it can win next\n",
    "            - otherwise, play a action following the rule :\n",
    "                first the corners then the center, otherwise a random available position \"\"\"\n",
    "        possible_actions = self.get_legal_actions()\n",
    "        \n",
    "        # win if it can\n",
    "        for action in possible_actions:\n",
    "            newState = self.move(action)\n",
    "            if newState.gameResult() == [True, action.player]:\n",
    "                return self.move(action)\n",
    "            \n",
    "        # block opponent if is going to win\n",
    "        for action in possible_actions:\n",
    "            actionOpponent = deepcopy(action)\n",
    "            actionOpponent.player = action.player * -1\n",
    "            newState = self.move(actionOpponent)\n",
    "            if newState.gameResult() == [True, actionOpponent.player]:\n",
    "                return self.move(action)\n",
    "\n",
    "        # play randomly\n",
    "        return self.move(action=random.choice(possible_actions))\n",
    "\n",
    "    \n",
    "    def gameResult(self):\n",
    "        for row in self.board:\n",
    "            if abs(sum(row)) == self.board_size:\n",
    "                is_terminal = True\n",
    "                reward = sum(row) / self.board_size\n",
    "                return [is_terminal, reward]\n",
    "        \n",
    "        for column in self.board.T:\n",
    "            if abs(sum(column)) == self.board_size:\n",
    "                is_terminal = True\n",
    "                reward = sum(column) / self.board_size\n",
    "                return [is_terminal, reward]\n",
    "\n",
    "        first_diag = self.board.trace()\n",
    "        second_diag = self.board[::-1].trace()\n",
    "        \n",
    "        for diagonal in [first_diag, second_diag]:\n",
    "            if abs(diagonal) == self.board_size:\n",
    "                is_terminal = True\n",
    "                reward = diagonal / self.board_size\n",
    "                return [is_terminal, reward]\n",
    "        \n",
    "        if np.all(self.board != 0):\n",
    "            is_terminal = True\n",
    "            reward = 0\n",
    "            return [is_terminal, reward]\n",
    "        \n",
    "        return [0, None]\n",
    "\n",
    "    \n",
    "    def is_game_over(self):\n",
    "        return self.gameResult()[0]\n",
    "    \n",
    "\n",
    "    def game_result(self):\n",
    "        return self.gameResult()[1]\n",
    "    \n",
    "\n",
    "    def __repr__(self):\n",
    "        \"\"\"print game board\"\"\"\n",
    "        res = '---------\\n'\n",
    "        for row in self.board:\n",
    "            for idx, val in enumerate(row):\n",
    "                end = ' | '\n",
    "                if idx == len(row)-1: end = ' \\n'\n",
    "                res += self.getMarker(val) + end\n",
    "            res += '---------\\n'\n",
    "                    \n",
    "        return res\n",
    "    \n",
    "    \n",
    "    @staticmethod\n",
    "    def getMarker(val):\n",
    "        \"\"\"Transform an integer (1, -1 or 0) into a marker ('X', 'O' or ' ').\"\"\"\n",
    "        if val == 1:\n",
    "            return 'X'\n",
    "        elif val == -1:\n",
    "            return 'O'\n",
    "        elif val == 0:\n",
    "            return ' '\n",
    "        else:\n",
    "            print('Value must be equal to 1, -1 or O.')\n",
    "        return None\n",
    "            \n",
    "        \n",
    "\n",
    "class Action():\n",
    "    def __init__(self, player, x, y):\n",
    "        self.player = player\n",
    "        self.x = x\n",
    "        self.y = y\n",
    "\n",
    "    def __str__(self):\n",
    "        return str((self.x, self.y))\n",
    "\n",
    "    def __repr__(self):\n",
    "        return str(self)\n",
    "\n",
    "    def __eq__(self, other):\n",
    "        return self.__class__ == other.__class__ and self.x == other.x and self.y == other.y and self.player == other.player\n",
    "\n",
    "    def __hash__(self):\n",
    "        return hash((self.x, self.y, self.player))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ekejzf-WQN8M"
   },
   "source": [
    "We modify the MonteCarloTreeSearch class to add the parameter c_param in the function best_action(), that control for the exploration constant.\n",
    "\n",
    "source: https://github.com/int8/monte-carlo-tree-search/blob/master/mctspy/tree/search.py  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Au84nm-TQN8Q"
   },
   "outputs": [],
   "source": [
    "class MonteCarloTreeSearch(object):\n",
    "\n",
    "    def __init__(self, node):\n",
    "        \"\"\"\n",
    "        MonteCarloTreeSearchNode\n",
    "        Parameters\n",
    "        ----------\n",
    "        node : mctspy.tree.nodes.MonteCarloTreeSearchNode\n",
    "        \"\"\"\n",
    "        self.root = node\n",
    "\n",
    "        \n",
    "    def best_action(self, simulations_number=1000, c_param=1.4):\n",
    "        \"\"\"\n",
    "        Parameters\n",
    "        ----------\n",
    "        simulations_number: int\n",
    "            number of simulations performed to get the best action\n",
    "        c_param: double\n",
    "            exploration constant\n",
    "        Returns\n",
    "        -------\n",
    "        \"\"\"\n",
    "        for _ in range(0, simulations_number):            \n",
    "            v = self._tree_policy()\n",
    "            reward = v.rollout()\n",
    "            v.backpropagate(reward)\n",
    "        # to select best child go for exploitation only\n",
    "        return self.root.best_child(c_param=c_param)\n",
    "    \n",
    "    \n",
    "    def _tree_policy(self):\n",
    "        \"\"\"\n",
    "        selects node to run rollout/playout for\n",
    "        Returns\n",
    "        -------\n",
    "        \"\"\"\n",
    "        current_node = self.root\n",
    "        while not current_node.is_terminal_node():\n",
    "            if not current_node.is_fully_expanded():\n",
    "                return current_node.expand()\n",
    "            else:\n",
    "                current_node = current_node.best_child()\n",
    "        return current_node"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Copcj7S3QN8Y"
   },
   "source": [
    "Then we define a function that plays a game of Tic Tac Toe between a 'player' and a computer.\n",
    "Both have their own policy to play.\n",
    "\n",
    "There are 3 possible policies:\n",
    "\n",
    "    - random: play a random move among those possible\n",
    "    \n",
    "    - smart: win if can, block its opponent if it can win next, or moves randomly\n",
    "    \n",
    "    - mcts: select the best action to do using Monte Carlo tree search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zFNZGWKWQN8a"
   },
   "outputs": [],
   "source": [
    "def playTicTacToe(board=None,\n",
    "                  method_player='mcts',\n",
    "                  method_computer='random',\n",
    "                  first_player=1,\n",
    "                  print_board=False,\n",
    "                  return_board=False,\n",
    "                  simulations_number=1000,\n",
    "                  c_param=1.0):\n",
    "    \"\"\"\n",
    "    Play a game of Tic Tac Toe automatically between a 'player' and the computer.\n",
    "    \n",
    "    Parameters:\n",
    "        board: 2D array, the board to play Tic Tac Toc,\n",
    "            if None, initialize with np.zeros((3,3))\n",
    "        method_player: string, method to play against the computer\n",
    "            default 'mcts' (see Methods for other values)\n",
    "        method_computer: string, method used by the computer to play\n",
    "            default 'random' (see Methods for other values)\n",
    "        first_player: integer, default 1\n",
    "            (1: the \"user\" plays first, -1: the computer plays first)\n",
    "        simulations_number: integer, the number of iterations for a MCTS.\n",
    "        c_param: double, the exploration constant for the MCTS\n",
    "            \n",
    "    Methods:\n",
    "        random: play a random move among those possible\n",
    "        smart: win if can, block its opponent if he can win next, or moves randomly\n",
    "        mcts: select the best action to do using Monte Carlo tree search \n",
    "    \"\"\"\n",
    "    \n",
    "    game = TicTacToeState(board=board, first_player=first_player)\n",
    "    \n",
    "    list_board = []\n",
    "    \n",
    "    while not game.is_game_over():\n",
    "        if game.next_to_move == 1:\n",
    "            method = method_player\n",
    "        elif game.next_to_move == -1:\n",
    "            method = method_computer\n",
    "        else:\n",
    "            raise ValueError(\"Current player must be equal to 1 (player) or -1 (computer).\")\n",
    "        \n",
    "        if method == 'random':\n",
    "            game = game.takeRandomAction()\n",
    "        elif method == 'smart':\n",
    "            game = game.takeSmartAction()\n",
    "        elif method == 'mcts':\n",
    "            initial_board_state = TicTacToeGameState(state=game.board, next_to_move=game.next_to_move)\n",
    "            root = TwoPlayersGameMonteCarloTreeSearchNode(state=initial_board_state)\n",
    "            best_node = MonteCarloTreeSearch(root).best_action(simulations_number=simulations_number, c_param=c_param)\n",
    "            game.board = best_node.state.board\n",
    "            game.next_to_move *= -1\n",
    "        else:\n",
    "            raise ValueError(\"Method unkown. Please select a method in ['random', 'smart', 'mcts'].\")\n",
    "    \n",
    "        if print_board: print(game)\n",
    "        if return_board: list_board.append(game.board)\n",
    "            \n",
    "    if return_board:\n",
    "        return game.game_result(), list_board\n",
    "    else:\n",
    "        return game.game_result()   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5PmfeQyQQN8x"
   },
   "source": [
    "An example of automatic game of a mcts player against a 'smart' computer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "wKwkRqTGQN80",
    "outputId": "b7c58226-5fa7-419d-fe8e-30202d5652fa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   |   \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | O \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | O \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "X |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | O \n",
      "---------\n",
      "O | O |   \n",
      "---------\n",
      "X |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | O \n",
      "---------\n",
      "O | O | X \n",
      "---------\n",
      "X |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | O \n",
      "---------\n",
      "O | O | X \n",
      "---------\n",
      "X |   | O \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | O \n",
      "---------\n",
      "O | O | X \n",
      "---------\n",
      "X | X | O \n",
      "---------\n",
      "\n",
      "---------\n",
      "X | O | O \n",
      "---------\n",
      "O | O | X \n",
      "---------\n",
      "X | X | O \n",
      "---------\n",
      "\n",
      "CPU times: user 3.86 s, sys: 98.5 ms, total: 3.96 s\n",
      "Wall time: 4.29 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "playTicTacToe(board=None,\n",
    "              method_player='mcts',\n",
    "              method_computer='smart',\n",
    "              first_player=-1, # 1: player first, -1: computer first\n",
    "              print_board=True,\n",
    "              simulations_number=1000,\n",
    "              c_param=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "3UD9HtnPQN8-"
   },
   "source": [
    "We now explore the difference between the different game policies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Number of games to play to compute statistics\n",
    "n_games = 1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "3LZeSfjhQN8_"
   },
   "outputs": [],
   "source": [
    "def getResultProp(counter):\n",
    "  \"\"\"Compute proportion of win/loss/deuce for a given Counter object\"\"\"\n",
    "  n_games = sum(counter.values(), 0.0)\n",
    "  counter_prop = counter.copy()\n",
    "  for key in counter_prop:\n",
    "      counter_prop[key] /= n_games\n",
    "  return counter_prop\n",
    "\n",
    "\n",
    "def getDictResult(counter, method='random', player=1, first_player=1):\n",
    "  \"\"\"For a given player (most of the time player '1', not the computer),\n",
    "  return a dictionnary with the results of several games\"\"\"\n",
    "\n",
    "  if first_player==player:\n",
    "    play_order = 1\n",
    "  else:\n",
    "    play_order = 2\n",
    "\n",
    "  dict_res = {'method':method,\n",
    "              'play_order':play_order,\n",
    "              'victory':counter[player],\n",
    "              'draw':counter[0],\n",
    "              'defeat':counter[player*-1]}\n",
    "\n",
    "  return dict_res"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "P1DEa3y3IgG1"
   },
   "source": [
    "We will store the results in DataFrames. Two DataFrames will thus be built: one where we compete against a random computer and one where the computer is 'smart'. We also create two other DataFrames for when we will modify the value of the hyper parameters of the MCTS."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "sNEwe7FdJDze"
   },
   "outputs": [],
   "source": [
    "columns = ['method', 'play_order', 'victory', 'draw', 'defeat']\n",
    "df_random = pd.DataFrame(columns=columns)\n",
    "df_smart = pd.DataFrame(columns=columns)\n",
    "df_simulations_number = pd.DataFrame(columns=columns)\n",
    "df_exploration = pd.DataFrame(columns=columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "HNUyjkYCQN9D"
   },
   "source": [
    "## Two random players"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 102
    },
    "colab_type": "code",
    "id": "WLvKTaGRQN9Q",
    "outputId": "9045157b-4d6d-447d-a5fc-85a9ce40eaf4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "Two random players (computer plays second): \n",
      " Counter({1.0: 0.573, -1.0: 0.298, 0: 0.129})\n",
      "CPU times: user 1.4 s, sys: 32 ms, total: 1.43 s\n",
      "Wall time: 1.58 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "res_list = []\n",
    "for _ in range(n_games):\n",
    "  res = playTicTacToe(board=None,\n",
    "                      method_player='random',\n",
    "                      method_computer='random',\n",
    "                      first_player=1, # computer plays second\n",
    "                      print_board=False)\n",
    "  res_list.append(res)\n",
    "    \n",
    "print(\"Number of games: \", n_games)\n",
    "count_res = Counter(res_list)\n",
    "print('Two random players (computer plays second): \\n',\n",
    "      getResultProp(count_res))\n",
    "\n",
    "# Save results into csv\n",
    "dict_res = getDictResult(counter=count_res,\n",
    "                         method='random',\n",
    "                         player=1,\n",
    "                         first_player=1)\n",
    "df_random = df_random.append(dict_res , ignore_index=True)\n",
    "df_random.to_csv('df_random.csv', sep=';')\n",
    "if use_drive:\n",
    "    !cp df_random.csv \"drive/My Drive/RL/\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Z_U2SXNqQN9a"
   },
   "source": [
    "When two random players play together, there is not an equilibrium between the number of win/loss/deuce.\n",
    "Indeed, the player that plays first has an advantage and wins most of the time (almost 60% of the time).\n",
    "\n",
    "In the game of Tic Tac Toe, the board has 9 places, so the first player can have one more marker than the second player."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xrSFdAw-QN9e"
   },
   "source": [
    "## A 'smart' player against an random computer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 102
    },
    "colab_type": "code",
    "id": "QPAr7yJZQN9g",
    "outputId": "e2e747c8-ced3-48f3-eb64-7fead72c6807"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "\n",
      " first player:  -1\n",
      "A 'smart' player against an random computer: \n",
      " Counter({1.0: 0.684, 0: 0.258, -1.0: 0.058})\n",
      "\n",
      " first player:  1\n",
      "A 'smart' player against an random computer: \n",
      " Counter({1.0: 0.911, 0: 0.07, -1.0: 0.019})\n",
      "CPU times: user 11.8 s, sys: 213 ms, total: 12 s\n",
      "Wall time: 12.6 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print(\"Number of games: \", n_games)\n",
    "for first_player in [-1, 1]:\n",
    "    print(\"\\n first player: \", first_player)\n",
    "    res_list = []\n",
    "    for _ in range(n_games):\n",
    "      res = playTicTacToe(board=None,\n",
    "                          method_player='smart',\n",
    "                          method_computer='random',\n",
    "                          first_player=first_player,\n",
    "                          print_board=False)\n",
    "      res_list.append(res)\n",
    "\n",
    "    \n",
    "    \n",
    "    count_res = Counter(res_list)\n",
    "    print(\"A 'smart' player against an random computer: \\n\",\n",
    "          getResultProp(count_res))\n",
    "\n",
    "    # Save results into csv and on the Drive\n",
    "    dict_res = getDictResult(counter=count_res,\n",
    "                             method='smart',\n",
    "                             player=1,\n",
    "                             first_player=first_player)\n",
    "    df_random = df_random.append(dict_res , ignore_index=True)\n",
    "    df_random.to_csv('df_random.csv', sep=';')\n",
    "    if use_drive:\n",
    "        !cp df_random.csv \"drive/My Drive/RL/\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "3_q_74pLQN9s"
   },
   "source": [
    "We can observe that, even when it is the random player that plays first, the smart player succeed to win almost 70% of the time.\n",
    "But, there is still 25% of draws, and the smart player can lose sometimes.\n",
    "We will give an example where the random player, thanks to the fact that he plays first, can, unwillingly, trick the smart player.\n",
    "\n",
    "But, we can also observe that when it is the smart player that plays first, this one wins more than 90% of the time, the rest of the time the game finishes with a draw.\n",
    "It is extremely rare that the random player beats the smart one when it is the latter that plays first.\n",
    "Though, the probability is not null."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "8mRWLPQDQN9t"
   },
   "outputs": [],
   "source": [
    "def display_board_history(list_board):\n",
    "    \"\"\"Display the hisory of boards for a game\"\"\"\n",
    "    for board in list_board:\n",
    "        print(repr_board(board))\n",
    "    return None\n",
    "\n",
    "\n",
    "def repr_board(board):\n",
    "    \"\"\"print game board\"\"\"\n",
    "    res = '---------\\n'\n",
    "    for row in board:\n",
    "        for idx, val in enumerate(row):\n",
    "            end = ' | '\n",
    "            if idx == len(row)-1: end = ' \\n'\n",
    "            res += getMarker(val) + end\n",
    "        res += '---------\\n'\n",
    "\n",
    "    return res\n",
    "    \n",
    "    \n",
    "def getMarker(val):\n",
    "    \"\"\"Transform an integer (1, -1 or 0) into a marker ('X', 'O' or ' ').\"\"\"\n",
    "    if val == 1:\n",
    "        return 'X'\n",
    "    elif val == -1:\n",
    "        return 'O'\n",
    "    elif val == 0:\n",
    "        return ' '\n",
    "    else:\n",
    "        print('Value must be equal to 1, -1 or O.')\n",
    "    return None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "CyOz1SAwQN9w"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | O | X \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | O | X \n",
      "---------\n",
      "O |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   | X \n",
      "---------\n",
      "  | O | X \n",
      "---------\n",
      "O |   |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   | X \n",
      "---------\n",
      "  | O | X \n",
      "---------\n",
      "O |   | O \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | X \n",
      "---------\n",
      "  | O | X \n",
      "---------\n",
      "O |   | O \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | X \n",
      "---------\n",
      "  | O | X \n",
      "---------\n",
      "O | O | O \n",
      "---------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "res = 1\n",
    "while not res==-1:\n",
    "    res, list_board = playTicTacToe(board=None,\n",
    "                                    method_player='smart',\n",
    "                                    method_computer='random',\n",
    "                                    first_player=-1, # computer plays second\n",
    "                                    print_board=False,\n",
    "                                    return_board=True)\n",
    "    \n",
    "display_board_history(list_board)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Q4gu4m33QN9z"
   },
   "source": [
    "This is an example where a random player, that plays randomly, succeed to unwillingly trick a smart player.\n",
    "\n",
    "We can observe that the same scenario can happen even when it is the smart player that plays first, as show after."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "8tukikxHQN90"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | X |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "O | X |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "O | X | X \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "O |   |   \n",
      "---------\n",
      "O | X | X \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   |   \n",
      "---------\n",
      "O |   |   \n",
      "---------\n",
      "O | X | X \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   |   \n",
      "---------\n",
      "O | O |   \n",
      "---------\n",
      "O | X | X \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | X \n",
      "---------\n",
      "O | O |   \n",
      "---------\n",
      "O | X | X \n",
      "---------\n",
      "\n",
      "---------\n",
      "X |   | X \n",
      "---------\n",
      "O | O | O \n",
      "---------\n",
      "O | X | X \n",
      "---------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "res = 1\n",
    "while not res==-1:\n",
    "    res, list_board = playTicTacToe(board=None,\n",
    "                                    method_player='smart',\n",
    "                                    method_computer='random',\n",
    "                                    first_player=1, # computer plays second\n",
    "                                    print_board=False,\n",
    "                                    return_board=True)\n",
    "    \n",
    "display_board_history(list_board)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "1vmIOvDwQN9_"
   },
   "source": [
    "## Two 'smart' players"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 102
    },
    "colab_type": "code",
    "id": "Ih-u077oQN-S",
    "outputId": "cfd79a47-76bd-4552-cf70-64cfd7588717"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "Two 'smart' players (computer plays second): \n",
      " Counter({0: 0.485, 1.0: 0.346, -1.0: 0.169})\n",
      "CPU times: user 11.5 s, sys: 160 ms, total: 11.7 s\n",
      "Wall time: 12.4 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# computer plays second\n",
    "res_list = []\n",
    "for _ in range(n_games):\n",
    "  res = playTicTacToe(board=None,\n",
    "                      method_player='smart',\n",
    "                      method_computer='smart',\n",
    "                      first_player=1, # computer plays second\n",
    "                      print_board=False)\n",
    "  res_list.append(res)\n",
    "    \n",
    "print(\"Number of games: \", n_games)\n",
    "count_res = Counter(res_list)\n",
    "print(\"Two 'smart' players (computer plays second): \\n\",\n",
    "      getResultProp(count_res))\n",
    "\n",
    "# Save results into csv and on the Drive\n",
    "dict_res = getDictResult(counter=count_res,\n",
    "                         method='smart',\n",
    "                         player=1,\n",
    "                         first_player=1)\n",
    "df_smart = df_smart.append(dict_res , ignore_index=True)\n",
    "df_smart.to_csv('df_smart.csv', sep=';')\n",
    "if use_drive:\n",
    "    !cp df_smart.csv \"drive/My Drive/RL/\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "NQrl3fYPQN-V"
   },
   "source": [
    "When two smart players play together, the game finishes most of the time with a draw (around 50% of the time).\n",
    "But, we can notice that the first player has an advantage over the other player, as the player that goes first wins around 30% of the time and loses only 15% of the time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "BBmCM3iXQN-W",
    "outputId": "1203ac32-7481-4091-d44c-7ecf005d2ad3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | X |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  |   | O \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | X |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  | X | O \n",
      "---------\n",
      "  |   |   \n",
      "---------\n",
      "  | X |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  | X | O \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "  | X |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  | X | O \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "X | X |   \n",
      "---------\n",
      "\n",
      "---------\n",
      "  | X | O \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "X | X | O \n",
      "---------\n",
      "\n",
      "---------\n",
      "X | X | O \n",
      "---------\n",
      "  | O |   \n",
      "---------\n",
      "X | X | O \n",
      "---------\n",
      "\n",
      "---------\n",
      "X | X | O \n",
      "---------\n",
      "  | O | O \n",
      "---------\n",
      "X | X | O \n",
      "---------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "res = 1\n",
    "while not res==-1:\n",
    "    res, list_board = playTicTacToe(board=None,\n",
    "                                    method_player='smart',\n",
    "                                    method_computer='smart',\n",
    "                                    first_player=1, # computer plays second\n",
    "                                    print_board=False,\n",
    "                                    return_board=True)\n",
    "    \n",
    "display_board_history(list_board)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QQXP7s-BQN-c"
   },
   "source": [
    "The example above show how a smart player can lose against another smart player, even when he goes first.\n",
    "\n",
    "Notice however that the second player 'tricks' the first one only by chance, as a 'smart' player only looks one step ahead. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "2iBenb7YQN-l"
   },
   "source": [
    "## A smarter player against a 'smart' computer\n",
    "\n",
    "We have shown that a 'smart' player can lose against a random player or another 'smart' player, even is it is the first one to play (we also have shown that the first player has an advantage in the game of Tic Tac Toe).\n",
    "We now try to determine if it possible to have a player even smarter, that can almost never lose.\n",
    "\n",
    "Using an MCTS algorithm to learn what to play in what situation might be a solution to overpast this problem.\n",
    "\n",
    "\n",
    "Now, the 'smarter' player plays by looking in the future in order to select the most adequate action to do.\n",
    "In order to do so, each time the player has to play, it runs a Monte-Carlo tree search.\n",
    "\n",
    "From now, we will use the following parameters for MCTS:\n",
    "\n",
    "    - simulations_number, the number of simulations for a search of MCTS\n",
    "    \n",
    "    - c_param, the constant for the importance of exploration during the MCTS phase\n",
    "    \n",
    "FORMULE OU INTERVIENT LA CONSTANTE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Oax_ugyuQN-m"
   },
   "outputs": [],
   "source": [
    "simulations_number = 500\n",
    "c_param = 1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 119
    },
    "colab_type": "code",
    "id": "0D1fJhCKQN-p",
    "outputId": "dbccf338-c6bd-4410-c980-4904b032bf31"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [15:43<00:00,  1.20it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "A smarter player against a 'smart' computer (computer plays first): \n",
      " Counter({0: 0.823, 1.0: 0.144, -1.0: 0.033})\n",
      "CPU times: user 16min 7s, sys: 23.5 s, total: 16min 31s\n",
      "Wall time: 15min 43s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print(\"Number of games: \", n_games)\n",
    "for first_player in [-1, 1]:\n",
    "    print(\"first player: \", first_player)\n",
    "    res_list = []\n",
    "    for _ in tqdm(range(n_games)):\n",
    "      res = playTicTacToe(board=None,\n",
    "                          method_player='mcts',\n",
    "                          method_computer='smart',\n",
    "                          first_player=first_player, # 1: player first, -1: computer first\n",
    "                          print_board=False,\n",
    "                          simulations_number=simulations_number,\n",
    "                          c_param=c_param)\n",
    "      res_list.append(res)\n",
    "\n",
    "    count_res = Counter(res_list)\n",
    "    print(\"A smarter player against a 'smart' computer: \\n\",\n",
    "          getResultProp(count_res))\n",
    "\n",
    "    # Save results into csv and on the Drive\n",
    "    dict_res = getDictResult(counter=count_res,\n",
    "                             method='mcts',\n",
    "                             player=1,\n",
    "                             first_player=first_player)\n",
    "    df_smart = df_smart.append(dict_res , ignore_index=True)\n",
    "    df_smart.to_csv('df_smart.csv', sep=';')\n",
    "    if use_drive:\n",
    "        !cp df_smart.csv \"drive/My Drive/RL/\"\n",
    "        \n",
    "    # save in df_simulations_number\n",
    "    dict_res = getDictResult(counter=count_res,\n",
    "                             method='mcts_'+str(simulations_number),\n",
    "                             player=1,\n",
    "                             first_player=first_player)\n",
    "    df_simulations_number = df_simulations_number.append(dict_res , ignore_index=True)\n",
    "    df_simulations_number.to_csv('df_simulations_number.csv', sep=';')\n",
    "    if use_drive:\n",
    "        !cp df_simulations_number.csv \"drive/My Drive/RL/\"\n",
    "        \n",
    "    # save in df_exploration\n",
    "    dict_res = getDictResult(counter=count_res,\n",
    "                             method='mcts_'+str(c_param),\n",
    "                             player=1,\n",
    "                             first_player=first_player)\n",
    "    df_exploration = df_exploration.append(dict_res , ignore_index=True)\n",
    "    df_exploration.to_csv('df_exploration.csv', sep=';')\n",
    "    if use_drive:\n",
    "        !cp df_exploration.csv \"drive/My Drive/RL/\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "b_1ozae4RGEQ"
   },
   "source": [
    "When a player with a MCTS policy plays agains a smart player, the result of the game highly depends on who plays first.\n",
    "\n",
    "Indeed, if it is the smart player that goes first, the game has more than 82% of chance to finishes with a draw. The MCTS can also win sometimes (15% of the time) and can almost never be defeated (loses only 3% of the time).\n",
    "\n",
    "But, when it is the MCTS players that starts to play, he wins 60% of the time, the chance to finishes with a draw are less than 40%, and the probability that the MCTS loses the game is almost null."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "xnbt3ci_VAfB"
   },
   "outputs": [],
   "source": [
    "# we fill df_random for a player with a mcts policy\n",
    "print(\"Number of games: \", n_games)\n",
    "for first_player in [-1, 1]:\n",
    "    print(\"first player: \", first_player)\n",
    "    res_list = []\n",
    "    for _ in tqdm(range(n_games)):\n",
    "        res = playTicTacToe(board=None,\n",
    "                            method_player='mcts',\n",
    "                            method_computer='random',\n",
    "                            first_player=first_player,\n",
    "                            print_board=False,\n",
    "                            simulations_number=simulations_number,\n",
    "                            c_param=c_param)\n",
    "        res_list.append(res)\n",
    "    count_res = Counter(res_list)\n",
    "    print(\"A smarter player against a 'smart' computer: \\n\",\n",
    "          getResultProp(count_res))\n",
    "\n",
    "    # Save results into csv and on the Drive\n",
    "    dict_res = getDictResult(counter=count_res,\n",
    "                             method='mcts',\n",
    "                             player=1,\n",
    "                             first_player=first_player)\n",
    "    df_random = df_random.append(dict_res , ignore_index=True)\n",
    "    df_random.to_csv('df_random.csv', sep=';')\n",
    "    if use_drive:\n",
    "        !cp df_random.csv \"drive/My Drive/RL/\"\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "omszt7IBQN-z"
   },
   "source": [
    "## Two MCTS players"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 119
    },
    "colab_type": "code",
    "id": "m4Tn8bKRQN-9",
    "outputId": "9f85a114-37cd-4820-a888-0754b3c407e3"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [33:06<00:00,  1.66s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "Two MCTS players (computer plays second): \n",
      " Counter({0: 0.918, 1.0: 0.082})\n",
      "CPU times: user 33min 56s, sys: 51.4 s, total: 34min 48s\n",
      "Wall time: 33min 6s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# computer plays second\n",
    "res_list = []\n",
    "for _ in tqdm(range(n_games)):\n",
    "  res = playTicTacToe(board=None,\n",
    "                      method_player='mcts',\n",
    "                      method_computer='mcts',\n",
    "                      first_player=1, # computer plays second\n",
    "                      print_board=False,\n",
    "                      simulations_number=simulations_number,\n",
    "                      c_param=c_param)\n",
    "  res_list.append(res)\n",
    "\n",
    "print(\"Number of games: \", n_games)\n",
    "print(\"Two MCTS players (computer plays second): \\n\", getResultProp(Counter(res_list)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "W5JBVCsmH3Yd"
   },
   "source": [
    "When two players have a MCTS policy, we see that most of the games finish with a draw (more than 90% of the time), but when it does not, the game is won by the payer that went first, thus confirming a player's advantage to play first."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>method</th>\n",
       "      <th>play_order</th>\n",
       "      <th>victory</th>\n",
       "      <th>draw</th>\n",
       "      <th>defeat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>random</td>\n",
       "      <td>1</td>\n",
       "      <td>562</td>\n",
       "      <td>116</td>\n",
       "      <td>322</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>smart</td>\n",
       "      <td>2</td>\n",
       "      <td>679</td>\n",
       "      <td>252</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>smart</td>\n",
       "      <td>1</td>\n",
       "      <td>900</td>\n",
       "      <td>90</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>mcts</td>\n",
       "      <td>2</td>\n",
       "      <td>914</td>\n",
       "      <td>82</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>mcts</td>\n",
       "      <td>1</td>\n",
       "      <td>980</td>\n",
       "      <td>20</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   method  play_order  victory  draw  defeat\n",
       "0  random           1      562   116     322\n",
       "1   smart           2      679   252      69\n",
       "2   smart           1      900    90      10\n",
       "3    mcts           2      914    82       4\n",
       "4    mcts           1      980    20       0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# run to read the csvs with results obtained\n",
    "#df_random = pd.read_csv('df_random_full.csv', sep=';', index_col=0)\n",
    "#df_random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>method</th>\n",
       "      <th>play_order</th>\n",
       "      <th>victory</th>\n",
       "      <th>draw</th>\n",
       "      <th>defeat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>smart</td>\n",
       "      <td>1</td>\n",
       "      <td>283</td>\n",
       "      <td>524</td>\n",
       "      <td>193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>mcts</td>\n",
       "      <td>2</td>\n",
       "      <td>144</td>\n",
       "      <td>823</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>mcts</td>\n",
       "      <td>1</td>\n",
       "      <td>613</td>\n",
       "      <td>386</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  method  play_order  victory  draw  defeat\n",
       "0  smart           1      283   524     193\n",
       "1   mcts           2      144   823      33\n",
       "2   mcts           1      613   386       1"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# run to read the csvs with results obtained\n",
    "#df_smart = pd.read_csv('df_smart_full.csv', sep=',', index_col=0)\n",
    "#df_smart"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "zAO40rLPQN_S"
   },
   "source": [
    "## Importance of the simulatons during the MCTS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "bwKvnVo7QN_U"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [00:34<00:00, 28.63it/s]\n",
      "  0%|          | 2/1000 [00:00<02:21,  7.08it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "A smarter player against a 'smart' computer (computer plays first): \n",
      " Counter({-1.0: 0.693, 0: 0.22, 1.0: 0.087})\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [02:40<00:00,  6.22it/s]\n",
      "  0%|          | 0/1000 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "A smarter player against a 'smart' computer (computer plays first): \n",
      " Counter({0: 0.482, -1.0: 0.381, 1.0: 0.137})\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [05:14<00:00,  3.18it/s]\n",
      "  0%|          | 3/1000 [00:00<01:23, 11.99it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "A smarter player against a 'smart' computer (computer plays first): \n",
      " Counter({0: 0.584, -1.0: 0.241, 1.0: 0.175})\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [00:45<00:00, 22.10it/s]\n",
      "  0%|          | 0/1000 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "A smarter player against a 'smart' computer (computer plays first): \n",
      " Counter({-1.0: 0.389, 0: 0.341, 1.0: 0.27})\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [03:32<00:00,  4.70it/s]\n",
      "  0%|          | 0/1000 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "A smarter player against a 'smart' computer (computer plays first): \n",
      " Counter({1.0: 0.482, 0: 0.459, -1.0: 0.059})\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [07:05<00:00,  2.35it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of games:  1000\n",
      "A smarter player against a 'smart' computer (computer plays first): \n",
      " Counter({1.0: 0.534, 0: 0.439, -1.0: 0.027})\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "c_param = 1.0\n",
    "\n",
    "for first_player in [-1, 1]:\n",
    "    for simulations_number in [10, 50, 100, 1000]:\n",
    "        method = 'mcts_' + str(simulations_number)\n",
    "        res_list = []\n",
    "        for _ in tqdm(range(n_games)):\n",
    "          res = playTicTacToe(board=None,\n",
    "                              method_player='mcts',\n",
    "                              method_computer='smart',\n",
    "                              first_player=first_player, # computer plays second\n",
    "                              print_board=False,\n",
    "                              simulations_number=simulations_number,\n",
    "                              c_param=c_param)\n",
    "          res_list.append(res)\n",
    "\n",
    "        print(\"Number of games: \", n_games)\n",
    "        count_res = Counter(res_list)\n",
    "        print(\"A smarter player against a 'smart' computer (computer plays first): \\n\",\n",
    "              getResultProp(count_res))\n",
    "\n",
    "        # Save results into csv\n",
    "        dict_res = getDictResult(counter=count_res,\n",
    "                                 method=method,\n",
    "                                 player=1,\n",
    "                                 first_player=first_player)\n",
    "        df_simulations_number = df_simulations_number.append(dict_res , ignore_index=True)\n",
    "        df_simulations_number.to_csv('df_simulations_number.csv', sep=';')\n",
    "        if use_drive:\n",
    "            !cp df_simulations_number.csv \"drive/My Drive/RL/\"\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>method</th>\n",
       "      <th>play_order</th>\n",
       "      <th>victory</th>\n",
       "      <th>draw</th>\n",
       "      <th>defeat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>smart</td>\n",
       "      <td>1</td>\n",
       "      <td>283</td>\n",
       "      <td>524</td>\n",
       "      <td>193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>mcts_10</td>\n",
       "      <td>2</td>\n",
       "      <td>87</td>\n",
       "      <td>220</td>\n",
       "      <td>693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>mcts_50</td>\n",
       "      <td>2</td>\n",
       "      <td>137</td>\n",
       "      <td>482</td>\n",
       "      <td>381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>mcts_100</td>\n",
       "      <td>2</td>\n",
       "      <td>175</td>\n",
       "      <td>584</td>\n",
       "      <td>241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>mcts_500</td>\n",
       "      <td>2</td>\n",
       "      <td>144</td>\n",
       "      <td>823</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>mcts_1000</td>\n",
       "      <td>2</td>\n",
       "      <td>90</td>\n",
       "      <td>899</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>mcts_10</td>\n",
       "      <td>1</td>\n",
       "      <td>270</td>\n",
       "      <td>341</td>\n",
       "      <td>389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>mcts_50</td>\n",
       "      <td>1</td>\n",
       "      <td>482</td>\n",
       "      <td>459</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>mcts_100</td>\n",
       "      <td>1</td>\n",
       "      <td>534</td>\n",
       "      <td>439</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>mcts_500</td>\n",
       "      <td>1</td>\n",
       "      <td>613</td>\n",
       "      <td>386</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>mcts_1000</td>\n",
       "      <td>1</td>\n",
       "      <td>525</td>\n",
       "      <td>475</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       method  play_order  victory  draw  defeat\n",
       "0       smart           1      283   524     193\n",
       "1     mcts_10           2       87   220     693\n",
       "2     mcts_50           2      137   482     381\n",
       "3    mcts_100           2      175   584     241\n",
       "4    mcts_500           2      144   823      33\n",
       "5   mcts_1000           2       90   899      11\n",
       "6     mcts_10           1      270   341     389\n",
       "7     mcts_50           1      482   459      59\n",
       "8    mcts_100           1      534   439      27\n",
       "9    mcts_500           1      613   386       1\n",
       "10  mcts_1000           1      525   475       0"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# run to read the csv with results obtained\n",
    "#df_simulations_number = pd.read_csv('df_simulations_number_full.csv', sep=';', index_col=0)\n",
    "#df_simulations_number"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmgAAAG6CAYAAAC88brwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdd3gVVfrA8e/clkBCJxAgIP1AAoRO6EWkKiKgosLqImvX9Se6oIuKIruWtbsqKFZAVFix0SWIFJEiNXCoSoAEQichyW3z+2MuMYEUkPS8n+fxkdyZOXNmzr0z75xz5hzDNE2EEEIIIUTxYSvqDAghhBBCiKwkQBNCCCGEKGYkQBNCCCGEKGYkQBNCCCGEKGYkQBNCCCGEKGYkQBNCCCGEKGYcRZ0BUfSUUvWBvcDWTB8bwOta6w8uM626wALAC9yrtV6TX/ksjpRS7wADgFla63/mU5rJQAugOjBBaz3iwvMK/CW/95tLfgYDnbTWTxXkfrLZ70fANq31fwppf5dVlkqpe4DKWuvnCzhfDYD/aK2HK6VqA3O01l3yKe1+wHvAEaCn1jr1T6RxB/Ah8KzW+ulMnxtY15VzWusWgc8qA88CvQA/YAJvaa2nK6X6AufLOhywA4cCf/9ba/25UuofwK1Y1yc7sBB4QmvtziOPrYG5wClguNb6t8s9zkxpBQHfAVO11nNyWGcM8CjgBJYCD2mtPUqp8sD7QBusCpLxWut5gW06AW8BocBhYJTWOuHP5lOUfBKgifNStdatz/+hlKoDbFNKrddab7mMdHoDiVrrvvmew+LpbqCe1vpgfiestV4PjAj8meW8KqVWFdR+s9EBqFoI+ylql1WWWut3Czg/510FqMA+DwP5EpwFjATe01o/d4XpHABGAU9n+qw7UB44B6CUCgZ+BGYCbbXWXqXUVcAPSim01tOB1oF1JwHVtdYPnE9MKXUjcAPQWWudGkhvDjAJeCKP/A0BYrXWY6/kIJVSnYH/As2AqTms0wJ4BmgLHMc63v8DXgzkNVlr3VwpVQ9Yo5RaDxwNHMtIrfUqpdS9wHRg0JXkV5RsEqCJbGmtDymldgNNgS1KqTuB+7Ce+o4DD2itdwZqOaoCjYBkoBZQSSkVq7XurZS6C3gI8GE9pT+gtd51wXbfATWxLuQtA//+JrCf67CepsdqrZcppZpiXSArBPa1CbhZa52mlEoDngf6BZa9qLV+B0Ap9ThwO1YN1G7gDq316VyOqxvwCtZTuon1BD838zlSSv2E9SS/QCl1H3AC6wm4WmCbl7XWnyilegGvAylYT8cdtNbpmdLpDrwZ2GZdIC8EtnsLeBB47vx5xfrdZt7vvsB69bCe2Gdrrf8VqBn9CdgB1Ad6Ag2AF4CQQJk8o7X+LlALcgNWrUaTQFncDlQE7gHsSqnTF9Ys5XTOA+mN0FpfG1gv4+9A2edZ1oFddFNKjQjkYzHwaODG3jxwTqsFyugNrfUHl3Cuo3IooyxlqbX+KdM2zbBulsGBdd7XWr+dOYhQSv0GzAL6AFWwbsZdgXaABxiitT4cWG9EIPgm899KqSeA64FygfJ5NHBu3gfqKKUWYQWR27TWoUopJ9Z39OpAWa4F/k9rfTaQ7keBZfWAT7TWT15Qdo8BQ4FUpVQlrCAnt/TWAq2waqy+IqutQF2lVBet9erAZ7cDM7BqJQFuxgpOXjy/kdb6d6XUTYCLvNXCKutyWA+UaUqpB4AauW2klLoN6zduV0qV01rfppR6ErgF63qwC+t3n6iUWo71O24GvKO1fvOC5B4CJpB7QHg98I3WOimw/6nAG1jfiRuwagDRWh9QSi0BbsI6t2e01qsCaUwHXlNKVdNaH7/geO4I5CEVWAb8XWvtUErVxAoaa2L9jn4HbtJaH72M72cdsr+WOLCuUV0D6+8D/qq1Ts7lPIgrJH3QRLYCT4qNgbVKqZ5YF9vuWus2WD/uzBfo8lrrKK11J+Ap4KdAcNYH+AfQW2sdjXWBmBdo+si83fjA322xLiA9gHFYF/MuWDfcCYF1/gZ8rLWOCeSvATA4sCwIOBbYZgTwqlIqWCk1BLgD68m7BbAfeCCP43oGeEVr3Q4YE8hXFlrr7oF/9gbWYN1M39RatwIGAv8KnEewmixv0Vq3uiBgcAFfAuMCeYjFugFl3k9s5vOaeb+BQOJT4INAXjsCfQM3PYAIYLLWuimQhtUUNVpr3RbrRvJO4EkerADuwcA5WovVvLoWeBf4PIdmv2zPeTbrXehSyvp8/q/GqlmJBv4WuFnMCeSvXSDfjyqlYgLb5HSuHeRQRtmc08weA74N7GsQ0EMpld21MzjwvXwKmIbVRSAaiMf6/uUoUJPUF+gVyNs/sZoMfcBYYK/Wuv8Fm00EagfOSzTW9fylTMtDA8fVJXB+GmTeWGv9UuB8vKq1fuwS0tumtW6eTXB23ifA6MDxlMeqQVuYaXl7YNWFG2mtN2qtf84hzcw+xmqiTFRKrVFKvYxV4/lLbhtprWfyx3f4NqXUX7HKvkPgXG/DCmbPO6m1jswmOENrfYvWenEe+ayLVebnHcT6Hue2LMvn2mqyTQLqZE5YKRWJ9YDVN3C9OIMVtIJVG7pGa90ZaIj1EDQ60+aX8v3M6VrSGatZOjqwbB9WsC4KkARo4rxySqlNgf+2Af8GbtNax2MFQI2B1UqpTViBTBWl1Plmr5U5pDkA66KYBKC1/gjrglM/h+2+1Vp7tNaJWDUg5y/ue/mjiW08kBToi/IO1g0lNFMaXwf+vxEreAjBuvF9qbU+GcjHI1rrKXkc1xfAf5VSM7GeMvNqQmmKdQH8X2Afh7H6vJyvPYjXWv+ezXYtAY/W+ofAdp8BZ/PYVwalVAhWgDI5cAw/Yz39nm+u9mIFj2BdZGthBcmbgPlYtUjnL7Qb9B/Nexu59GbN7M55Xi6lrAE+1VqnBG5YM4BrsM51I+CDwHH8iBXUtglsk9O5zquMcvIV8A+l1P+AYVj9ifzZrHe+hnUvVnP05hyO6SKB/P4FuE0p9TxWrWVobttgBRnvBs6jH6uGY2Cm5V8H0j6E1YSWV3nmld6FgeuFZgLDAg8dN2AFf95My/1cwT1Ha31aa90Pq3brfayas++VUi9cZlIDgQ+11imBv18Hrg7kG/I+zrzYsH5X5xlYNZK5Lbvw8wu3O68/sDjT7zQjiNRav451LXsEeBvrQSXzdyjX72ce15KtgbysVUpNBuZmqikVBUSaOMV5WfqgXcCOdaMcDxCoPagNnAwsz6ma2w5c2HnXwKo6z2679Av+9mST5mdY39svgO+xLiBGpuWpAFprUyl1fn9eMl38lNVRuXJux6W1nqqU+har6W4AMEkppbTWabkc64UXWFsux5qZccHf3mzXynm/BtBFa32+r091rNqy6kC61tqbad0dgZpOAuvWxnpSv43AuQsws8lXTrI75xduf2ET1qWUNWS9QdkC69mB0zprn8mawGkghty/j7mVUba01QTcBCs4vBp4WinVLptVMx9TTseT7XlRSrXFCqhexWrK/RHrASQ3Fx7PhcdyueWZV3q5Nmdpq4lwI1YAdDvwCNZ38Lyfgfsv3C5Qw909UIuXo8BD2cpAYLAPmB7oirAQ68HtUmV3nOe7DUAex3kJDmBdR86rjVVTlnnZkUzLNl24TaD5uhp/vCRxnpes5ejLtM0LWLVeH2DVxDsvWDev72eO1xKtdbJSKhqribMP8LlS6iWt9dvZpCPyidSgiUuxCLhFKVUr8Pc9wA+XsN1CYKRSKgwg0LRwHNhzBXnpj9X083ng7078UcWfk6VYT/YVA39Pwrp55HhcSqnVQJtArd9dWAFdeC772Al4lFLDAtvXBoYDS/LI2xbAUEoNCmw3BKuPyCXRWp/BuvE9Eti+MlYz0vXZrP4z0EQp1SOwbmus/nh1slk3My95BDHZSAJaBJqYnfzxssPlGqmUCgo0m96O9Sarxuo3NQoy3hzehlXTmZs/VUZKqVlY/RxnY/VlOoNVg/dnJGE19Z3vY3j+u9cDWK+1fgUrOBvKH9/rnM7/QuBepZQz8HBxf17Hkof8SO8TrCbrSlrrbRcsm4vVj/IfSik7gFKqIVa/tx2XkHZ54PlMNfdg1UBvvMw8LgTGBGqMwOpXtiJzc/gV+gYYopSqEejOcRcwL7Ds68DfKKUisB7+vsPqUlBNKXX+BZAxWM2Vpy5IexFWs+P532zmlx76A69prT/FqjG9hryvjRlyu5Yopa7Fujau1lpPwirnDpeatvhzJEATeQr0uXgBWKKU2oLVyXWY1vrC2ogLt1uCVSOwTCm1HesGe20OzUOX6gngK6XUVqwOsT9iNVPmlo/5WH2vVgW2Cwf+mcdx/QN4Vin1K7AcqzP9b7nsw4N1U/17IK2lWIFkbB55O7/d+WaFYVgX18txKxATOLa1wGeBfjcX7isJKyB5SSm1Gau/yejcjitgGdBfKXVRn5xcnK8F2gmsANZfxraZ7cdqcvo1kM7HgebO64GxgXO9GHhS/9HBOlt/toyAyVhNj5uxzu9Xgbz8GeMD+9+E1T9oQ+Dzz4DqSqkdQBxWLU5VpVSFwN9pSqlfyFoj8hyQiFUDswMriPv7n8xXfqU3D6v/2qcXLgiUW18gCtgaKIO5wHP60obzmYxVZquVUjuUUruAblid7FFKtQ+c17xMD6TzS+B8t8WqQc4X2nrr/Vms381OrFqu882wTwOhgevhUuAxrfXewHdzGNaLAdsD+flrNmnvwnojdJGy3v5sTuAt2cA+/xM4r99gdSHJ9dqYjZyuJQuA7QTe7Mfq1/jMZaYtLpNhmrneY4UQQogSQSk1V2s9vKjzUVACL3r8BevFH3+gNnh85m4LovSQPmhCCCFKvECz32UNrF0CHcTqq7ZVKeXF6nc5pmizJAqK1KAJIYQQQhQz0gdNCCGEEKKYkQBNCCGEEKKYKQ190IKwXvdN4OJB/YQQQgghihM71hA767h4TMgMpSFA68CVj/wshBBCCFGYupPzTDylIkBLADh5MgW/v2BfeKhWLZTjx2Vu2LJGyr3skrIvu6Tsy66CLnubzaBKlRAIxC85KQ0Bmg/A7zcLPEA7vx9R9ki5l11S9mWXlH3ZVUhln2u3LHlJQAghhBCimJEATQghhBCimCkNTZw58vm8nDyZhNfrzpf0jh614fdfyTSSorhyOFxUqRKG3V6qfxJCCCFKiFJ9Nzp5Mong4PKEhIRjGEbeG+TB4bDh9UqAVtqYpklKyhlOnkyievVaRZ0dIYQQonQ3cXq9bkJCKuZLcCZKL8MwCAmpmG81rUIIIcSVKtUBGiDBmbgk8j0RQghRnJT6AE0IIYQQoqSRAK2QbNy4ngceuKvA0p8//1umTJkEwKOPPsSxY0kFtq/8smrVT8yePaOosyGEEEIUO6X6JYGy6j//eaOos3BJdu6MK+osCCGEEMWSBGiF6PTpUzzyyIMcO3aUyMgWPPLIeFwuF3Pnfs7ChfNJS0vF6XQyadIU6tWrz1tvvca6dWux2Qy6d+/FmDF3ce7cOV555QX27duL3+/nttv+wjXXDMiynxEjruPNN6fy668bWLt2NWfOnOHw4UN06BDDo49OAODTTz8iNnYJPp+fTp1iuPfeh7L0w0pJSWbSpH9y/PhxAMaM+RvduvXkgQfuQqlmbNmyCbfbzT33PMiXX87mt9/2cfPNt3LzzbeRlHSUf/97MsnJZzl2LIlBg65j7Nh7mD//WxYs+I7Tp08REVGPbdu2ABAeXovBg4cUUikIIYQQxV+ZCdBWbU1g5ZZcp73Kk2GAmc3sD91a1aJry7yHZ0hIOMy//vUfIiLq8vTTTzBv3lwGD76OFSt+5K23phIUFMz777/L3LlfcMsto/n559XMmPEFaWlp/Otfz5Cens7HH09HqeZMnPgMKSnJ3HPPGCIjW+S4z61btzBjxhfYbHZuvXU4e/eOICnpKFrv4L33PsEwDCZPforFixfQv/+gjO1WrFhOeHhtXnrpdXbv1ixevJBu3XoC1rAU7733CR98MI3XXnuJjz+ezalTJ7njDitAW7JkEddc05+BA68lOTmZYcMGM2LESACSko4yY8aXOBwOpk+fCiDBmRBCCHGBMhOgFQfR0W2pW7ceAP36DeD777/lpptuYdKk51i6dDHx8QdYu3Y1TZooqlcPIygoiHvvHUOXLt25994HCQoKYv36X0hPT+P7778BIC0tjf379+W4z5YtW1G+fAgAtWvX4cyZ06xf/wtxcdu4887RAKSnp1GzZniW7Vq0aMXUqf/l2LGjdO7cjTvuuDNjWUxMV8Cq+YqKaklwcDDh4bVITj4LwK23jmbjxvXMmvUp+/fvxev1kJaWCkDTps1wOORrJ4QQQuSmzNwpu7a8tFqu3FzpQLV2uz3j336/icPh4MiRRB588G6GD7+JmJguVK1ajd27NQ6Hg2nTPmLTpo2sWbOKe+75K2++OQ2/38eTT05GqWYAnDhxnIoVK7F48YJs9+lyBWX52zRN/H4fN910CyNHjgLg7NmzWfIGULduPWbNmsPPP69h1aoVzJ49gxkzvgychz++NhduB/Dmm69y+PAhrrlmAD169GL9+l8wA1WPQUFBF60vhBBCiKzkLc5CtGXLJhITE/H7/Sxc+D3t23dk5844IiLqcvPNt9G8eSQrVsTi9/vYtWsnDzxwF9HRbXjggYepX78hBw78Ttu2HZg3bw4Ax44d4/bbb+HIkcTLykfbth1YtGg+586dw+v18vjj41i+/Ics68yd+znTp0+lT5++jBs3gZMnT5KSknJJ6a9fv5Zbbx1Nnz59OXDgd5KSjmY7RZbdbsfn811W3oUQoiD5k0+QvmEeJ2Jn4k8+XtTZEWVYmalBKw4aNGjIv//9LMePH6Ndu/Zce+31pKen89VXcxg16kZM06R167bs27eXpk2b0aJFK/7yl5sJDg6mZctoYmK60KZNW15++QVGj74Jv9/Pffc9RJ06EWze/Osl56Nbtx7s2bOLu+66A7/fR6dOXRg48Nos6wwYMJhJk/7JX/5yM3a7nfvvf4gKFSpcUvqjRt3B5MlPERQURI0a4TRrFsnhw4cuWq9167ZMmTKJqlWrZvRRE0KIwmaafnwHt+OJW4b3wCYwwW0YsOZrHI064ooeiL1avaLOpihjDDO7Xu8lS31g//Hjyfj9WY8lMfF3wsOvyrcdyVycpVtO35ewsAokJZ0tghyJoiZlX7r5087i1T/h3rEc88xRjOAKOFV3nM17U616BRJ+/ArPzh/Bk4a9TpQVqNWJkplHSrmC/t3bbAbVqoUCNAB+y2k9qUETQghRZpimie/IHqu2bN868HuxhzfF2X4YjgbtMOxOAByVKhDc+RaC2g7BvSMWz9YlpM7/D7Zq9XC1GoCjUUcMm9xCRcGRb5cQQohSz3Sn4tmzBk/cMvwnDoIzGGeznjgje2OvGpHjdkZQCEGtr8XVsj/e3Wtwb1lIWuw0jF/m4GrZD2eznhiucoV4JKKskABNCCFEqeU7Ho8nbhmePWvAk4atWj2Cut+Bs3EMhjP4ktMx7E6czXrgUN3wxW/BvXkB6T/PJn3j17ia98bZ4hpsIVUK8EhEWVOgAZpS6jrgaSAEWKy1/rtSqi/wClAO+FxrPTGwbmvgfaAisAK4R2vtLcj8CSGEKH1Mrxvv/vW445bhP7IH7E6rs39kH2xhDa+oD5lh2HDUa42jXmt8R/fh3rLA+m/rIhyNu+BqNQB71Tr5eDSirCqwAE0p1RB4F+gEHAGWKaUGAlOBnkA88L1SaqDWegEwAxirtf5ZKTUd+BvwTkHlTwghROniP3MUd9wyvHolZnoyRqWaBMWMxNm0G0ZwaL7vz16jIeX63m/td8siPPonvLt+wl63lfVCQa1m8kKB+NMKsgbtBqwasoMASqmbgSbAbq31/sBnM4AblVJxQDmt9c+BbT8CnkECNCGEELkw/T68BzbhiYvFd3AbGDYc9dvibN4be53mGEbBD/dpq1iD4G6jcbUfajWnbltK6ncvYAtrgKvVQOvlA9vFg3oLkZuCDNAaA26l1DdAPeA7YDuQeULMBCACqJ3D50IIIcRF/Ckn8excgWfnj5gpJzBCquBqdwPOZj2KrC+YLbgCQW2vx9VqIJ5dq3BvXUjaD29jVAizXihQPTCcMpuKuDQFGaA5gB5ALyAZ+AZIBTIPVmYAfqwZDbL7/JIFxhTJ4uhRGw5H/j495Xd6U6Y8y7BhI2jePDLb5fPm/Y/y5cvTr9+AfN2vz+fjiSf+wTPPPEdwcPZvIL333rsA/O1v91zRfh555EESEhK4/vob+O233/jnP5+6pG1XrlzBgQMHuPXWUcTG/sCxY0nceGPBDWhrs9kIC8t+MN6cPheln5R98WGaflJ/28qZDYs4t2sdmH7KNYymYtuxlG/SPt9rqa6o7GsNwew+mHO71nPq569JXz0Tz69fU7HtACq2H4gjtHL+ZVTku+Lwuy/IAC0RWKq1TgJQSn0F3AhkntsnHDgMHARqZfP5JctuoFq/35+vA8sWxEC148dPBMgx3c2bN9GmTbt83+/cuV/QoUMMDkdQjmmfP59Xsu/ExCPs2bOHr79emPHZpaa3ffv2jPW7d+/NQw/dQ69efalSpeqfzk9u/H5/toMTymClZZeUffFgpiXj2bUS945YzNNHMIJCcbbsh6t5b2yVanIOOHf8XL7uM9/KvlokrsGR2BN3496ygFOr5nLq53k4m3TD1ao/tspXNke0yH+FOFBtrgoyQPsO+FgpVRk4CwwE5gATlFKNgf3ArcAHWuvflVJpSqmuWutVwGgg+9m//yTPrlV49IorSsMwDLKbecGpeuBs2jXXbZ944jH69RtAr15XAzBmzCgmTJjIG2+8wpgxd9GmTTveeedNVqxYjsNhZ8iQYTRo0JCVK1ewYcM6qlWrTpMmTXn++ckcOZKI3W7nrrvuJyamC9OnT2X79m0cPZrI0KHDmTXrU+bM+RabzcbGjeuZOfMTXn75jYy8mKbJnDmf8957nwAwZcokgoKC2LEjjpSUFO64404GDBicJf9z537OwoXzSUtLxel0MmnSFI4ePcr06e/yzjsfADB//rfExW3j0Ucfz9hu/PiHOX36FHfeOZr77/87H3wwjbfemsYDD9xFxYqV2L9/L089NZk5c2azb99eAG644UZatozm66//B0B4eC0GDx5Cz569+d//vuTOO+++3KITQpQwpmniT9pndfrf+wv4PNhqNiao7fU4GrTHcLiKOouXxR7ehHLhTfCfSrBeKNi9Es/OH3Fc1Rpn9CAc4U2KOouimCmw3pNa67XAi8BKIA74HavT/x3A3MBnO7GCNoDbgFeVUjuBUOANSpH+/QexdOkiAOLjD+B2u2natFnG8tjYH9i6dTOffDKbadM+Zv78b2nYsBHduvVg7Nh76NSpM6+++hJt27bn449nM3nyC/z7389y4oQ1ma/bnc6MGV8yYsRIateuw6+/bgBg4cLvGTQo6zybe/bsIjQ0lNDQPyL4Q4cOMnXqh7zxxjv897+vc/z4sYxlKSnJrFjxI2+9NZVPP/2CLl26M3fuF7Rr14Fjx45z6NDBjH0NHHhdln09//wrVK8exvTpn150Tho1asxnn/2PtLQ0zpw5w4cfzuKll15n8+ZfadCgIddfP4zrrx/G4MFDAIiObsvKlT/+6TIQQhR/picd947lnPvfJM7Nm4x3/wacTbtRfvizhFw/EWeTLiUuOMvMVrkWwT3uIOSWl3G1vQ5v4i5Sv5lCytfP4dm/AdMv0wkKS4GOg6a1/gD44IKPfwCis1l3M9CxoPLibNo1z1quvFxJE2eXLt149dUXOXcuhaVLF9G//8Asyzdt2kCfPtfgcrlwuVx89NGsi9LYuHFdRpNonToRREa2IC5uGwCRkS0y1hs8eAiLFs0nKqolGzasY9y4CVnSiY+PJyysZpbPBg26DofDQY0aNWnZMpotWzZlLAsJCWXSpOdYunQx8fEHWLt2NU2aKAzDYODAwSxaNJ9Bg4Zw4sQJoqJacKnO57lhw0YcOPA7jzzyADExXbn//r9nu354eC0OHoy/5PSFECWH78Qh6w3I3avBk4qtagRB3f6Cs3HnUjlSv618JYLaD8MVPRiP/gn31kWkLXkTo1JNXC37W0ODlOBAVFw5mUmgkDidTrp27c7KlStYtmwJL730epblDoeDzMPlJCQcpnLlrG8iXdjHDkx8PqtLX1DQH28G9e7dl2nT3iY2dimdO3fNsgzAMKz9ZWa3//G3afqz/H3kSCIPPng3w4ffRExMF6pWrcbu3RqwArtx4x7E5XIxYMCgSzsZAefzValSZT799AvWrVvLmjWrGDNmFJ9++sVF61vnqOBfmRdCFA7T58G7fwOeuGX4EneBzYGjYQeckX2w12xcJsYQM5xBuFr0xRnZG+/+Dbi3LCB95Se413+FM6ovzqg+2IKLvsO6KHxytytE/fsPYvbsGVSqVJnw8KwdQ6Oj27J8+TK8Xi9paWmMG/cgSUlHsdvtGUFYu3bt+e67eYDVJLl162aiolpdtJ/g4GBiYrowbdrbFzU5AkRE1CUhIes7GMuWLcE0TRITE4iL20Z0dOuMZTt3xhERUZebb76N5s0jWbEiFr/fylN4eC3Cwmowb97ci/qtXaqVK39k8uSn6NKlGw8//CjlypXj6NEjWY4d4PDhQ0REyOgrQpR0/jNJpP/yJSkzHyFt2bv4U04S1OkmQka9Srk+d+MIb1ImgrPMDJsdZ6OOlB/6FOWunYCtRkPcG74iZeY40lZ+iv/M0aLOoihkUoNWiFq1ak1ycjJDh464aFnPnr3ZuTOOMWNuw+83ufHGW6hX7yrat+/I1KlvExoaysMPP8aLL05h/vxvMQyD8eMnUr169Wz3dfXV/QIB3MVNjo0bN+X06VMkJydn9ENLT0/jzjtH4/G4eeyxf1Kp0h+vgHfoEMNXX81h1KgbMU2T1q3bZnToB+jbtx/Lly+jevWwP3VeYmK6snz5MkaPvgmXy0X//oNo1KgxZ8+eYcqUSVStWpURI0by66/r6dat55/ahxCiaJl+P774zbjjYvHFbytKxOsAACAASURBVAUDHPVaW7VlEVFSOx5gGAaO2s1w1G6G78Qh3FsW4tm5HM+OZTjqt8MVPQh7jYZFnU1RCIzs3kosYeoD+7MbZiMx8XfCw6/Ktx0VxDAbBcHn8zFt2ttUqVKFkSNHZbvOl1/OxmYzGD78ZqZMmUSbNu0YNOji2ra8eL1eJk9+ij59+tKzZ58rzXqu7r33Tv71r5cKbJiNnL4vMtRC2SVlf+X85079MaBs8nGM8pVxNuuBs1lPbKHVijp7OSpOZe9POYln+1LcccvAnYq9lsLVaiD2eq0ksC0AhTjMRgPgt5zWkxq0Umjs2NFUqlSZF154Jcd1brhhBBMn/oPBg6//0/sxTZOhQwfSoUMnunfv9afTuRSxsUvp3fvqAgvOhBD5xzRNfAk78cTF4t2/AUwf9jqROGNG4qjfBsMmt57LYQupQlDHG3G1vhbPzhW4ty4iddFr2CrXxtmqv/Vmq91Z1NkU+Uxq0C5DSalBE3+O1KCJC0nZXx4zPcUac3JHLP5TCRAUgrNpN1zNe5W4AVmLc9mbfi/evb/g3rIA//F4jHKVcLa4Bldkb4ygkKLOXoknNWhCCCFKBV/SfmuIjD1rwefGFtaQ4J534mjUSYaKKACGzYGzSRccjTvjOxSHe8sC3Ovm4P71W5zNeuJq2Q9bhez7J4uSQwI0IYQQl830puPdsxb3jlj8SfvB4cLZpDPOyN7Yq9cv6uyVCYZh4IiIwhERhe/4AdybF+DZ/gOe7UtxNOpo9VOrnn+tSKJwSYAmhBDikvlOHcYTF4tn1ypwn8NWpTZBXUbhbNoFw1W+qLNXZtmr1aNcn7vxdxyBe9sSPDuW493zM/Y6kVagFtGizA1dUtJJgCaEECJXps+L97eN1oCyCTvBZsfRoL01REZ4U7nxFyO20GoEx4wkqM11uHf8iGfbYlIXvIytal1crQbgaNxJXtIoIeT93CL2/POT2bkzLsfl33zzFUuWLMz3/fp8Pp544jHS0tJyXGf69KlMnz413/dd2M4fh9/v5/HHH+XcuXNFnSUhSgR/8nHS180lZdYjpP3wNv7kY7g6jiDktlcpd/W9OGopCc6KKSMohKDWgwi55T8E97wTTD9py98j5bN/4N6yANOdWtRZFHmQMLqITZjwZK7Lt27dTJs27fJ9v/PmzaFjxxiCg4PzPe3iymazMWTIUD766D3uuy/7+T6FKOtMvx/fwW2445bhi98MJtjrtcIV2Qd7REsMmzzXlySG3YFTdcfRtBu++K3WVFI/f076hm9wNu9lvVAQUiXvhEShkwCtkDzxxGP06zeAXr2uBmDMmFFMmDCRN954hTFj7qJNm3a8886brFixHIfDzpAhw2jQoCErV65gw4Z1VKtWnSZNmvL885M5ciQRu93OXXfdT0xMF6ZPn8r27ds4ejSRoUOHM2vWp8yZ8y02m42NG9czc+YnvPzyGxl5MU2TOXM+5733PgFgypRJBAUFsWNHHCkpKdxxx50XTds0d+7nLFw4n7S0VJxOJ5MmTeHo0aNMn/4u77zzAQDz539LXNw2Hn308Yzt9uzZzYsvTsHn8+FyuXjiiaepW7ceP/+8munT38Xr9VKrVh3Gj7dmL1i3bi1vvfUapuknPLwWTz/9HOXKleeNN15m/fp1GIY1ZdaoUXewceN6Pv30Q4KDg/ntt/00atSYp5+egtPpZNasT/jmm6+oVKkyFSpUoHnzKAA6duzMa6/9h9tvv5OQkNACLXMhShJ/6hk8egWeHcsxzx7DKFcRV/RgnM17yRuBpYBhGDjqtcJRrxW+pP3WCwVbF+LZthhH4xirn1pVmUqvOCkzAdrahA2sSVh3RWkYBmQ3bFznWh3oVCv3Wq7+/QexZMkCevW6mvj4A7jdbpo2bZaxPDb2B7Zu3cwnn8zG6/Vy331jefnlN+jWrQdt2rSjU6fOPPnkBNq2bc/IkaM4dOgg9903lg8/nAmA253OjBlfArB8+TJ+/XUD7dp1YOHC7xk06NosedmzZxehoaEZ0zyBNbfn1KkfcuLEce68czQdOnTKWJaSksyKFT/y1ltTCQoK5v3332Xu3C94+OHHeOGFKRw6dJA6dSJYuPB77r77gSz7+uKLWYwcOYo+ffqyYMF3bN++ldDQCrz77lu88ca7VKxYkXnz5vLOO2/yyCPjefbZJ3nllTdp0kTx7rtvsWDBd9hsdo4cOcLHH3+Gx+PhwQfvomHDxgQHB7Nt2xZmzpxD9eph3H33Haxdu4bq1avz/fff8MEHMzEMg3vu+WtGgGa322nUqAkbN64v8MF1hSjuTNPEl7grMKDsOvD7sNdSODveiKN+Owx7mblFlCn2sAaU63sf/jNJuLctxrPzR7y7VmGv29IK1Go3l6brYkB+fYWkS5duvPrqi5w7l8LSpYvo339gluWbNm2gT59rcLlcuFwuPvpo1kVpbNy4jvHjJwJQp04EkZEtiIvbBkBk5B9zbg4ePIRFi+YTFdWSDRvWMW7chCzpxMfHExZWM8tngwZdh8PhoEaNmrRsGc2WLZsyloWEhDJp0nMsXbqY+PgDrF27miZNrL4nAwcOZtGi+QwaNIQTJ05cNPdn585deeWVF1m7djVdu/aga9fu/Pzzao4cSeShh+4BwO/3UbFiJfbt20NYWBhNmigA7rnHCvYmTvwHgwZdi91ux263c801A9mw4Re6du1BgwaNqFHDOparrmrA2bNnOHDgd2JiulK+vPVGWe/efbNMuh4eHk58fHyu5SVEaWa6U/HsXoUnbjn+kwfBVQ5nZB+czXtjr1K7qLMnComtYhjBXW4jqO31uOOW4dm+lNTvX8RW/SpcrQbiaNgBw2Yv6myWWWUmQOtUq12etVx5uZKZBJxOJ127dmflyhUsW7aEl156/YK0HWR+YElIOEzlyln7BVw4UwKYGYFHUFBQxqe9e/dl2rS3iY1dSufOXbMsA6sm0OHIWvT2TE/KpunP8veRI4k8+ODdDB9+EzExXahatRq7d2vACuzGjXsQl8vFgAGDLjru3r370qJFK1at+okvvpjFmjUr6dKlG61aRfPCC68CkJ6eTmpqKklJR4E/TkJycjLnzqXketwu1x+DYBqGgWmagSe/P7ax2+1ZAjS73YHNJk+HouzxHfvdGiJjzxrwpmOrXp+gHn/F2SgGwxmUdwKiVDKCQwlqOwRXqwF4dq/Gs2UhacvexVg3B1fL/jhVdwxn2emvXFxIb89C1L//IGbPnkGlSpUJD8867Ul0dFuWL1+G1+slLS2NceMeJCnpaJbgol279nz33TzAapLcunUzUVGtLtpPcHAwMTFdmDbtbQYOvHgC9IiIuiQkHM7y2bJlSzBNk8TEBOLithEd3Tpj2c6dcURE1OXmm2+jefNIVqyIxe+38hQeXouwsBrMmzf3on5rAE899Tg7dsQxdOhwxo69B613EhnZgu3bt3LgwO8AfPTR+/z3v69Rr95VnDp1kv379wEwc+bHzJs3l3bt2rNgwff4fD7S0tJYvHghbdq0z/E8t2/fgVWrfiI5OZn09HRWrIjNsjwh4TAREXVz3F6I0sT0uvHsWknKvGc597+n8exejaNhR8oPfYqQYZNwNespwZkAwHC4cDXvRfmb/kW5fn/HFlKV9NUzSZ41jvRf5uA/d6qos1imlJkatOKgVavWJCcnM3ToiIuW9ezZm5074xgz5jb8fpMbb7yFevWuon37jkyd+jahoaE8/PBjvPjiFObP/xbDMBg/fiLVq2ffeffqq/sFArgWFy1r3Lgpp0+fIjk5OaMfWnp6GnfeORqPx81jj1kd9s/r0CGGr76aw6hRN2KaJq1bt2Xfvr0Zy/v27cfy5cuoXj3son2NHv1XXnjhOT766D0cDiePPjqBatWqM2HCUzz11OP4/T7Cwmry1FPPEhQUxJNPPstzzz2N1+uhdu0InnzyWVwuF/HxB7jjjlvwer306zeQnj17s3Hj+myPvUkTxY033sLYsX+hQoUK1Kz5RzDs8/nYtWsnEyc+k0MpCVE6+E8n4o6LxbNrJaSnYKsUTlDnW3E27SrzNYpcGYYNR/02OOq3wXdkD+7NC3Bv+h73loU4m3bB2WoA9srSFF7QZLL0y1BSJkv3+XxMm/Y2VapUYeTIUdmu8+WXs7HZDIYPv5kpUybRpk07Bg26uLYtL16vl8mTn6JPn7707NnnSrNe4H76aTlbtmzm/vsvHmZDJksXFyppZW/6vXh/34QnLhbfoe1g2HE0aGsNKFurmXT8vgwlrewLmv90Iu4ti6yA3+fBXq81ruiBpXKgYpksXRSYsWNHU6lSZV544ZUc17nhhhFMnPgPBg++/k/vxzRNhg4dSIcOnUrEG5F+v59vv/2ap5+eXNRZESJf+ZNP4Nn5I56dP2KeO4URUhVX+2E4m/XAVr5y3gkIkQdbpXCCu9+Oq/0Ngfk+fyD1239jq9HQeqGgfjsZIy+fSQ3aZSgpNWjiz5EaNHGh4lz2punHdygOT9wyvL9vAtPEXrcFruZ9sNeLlpvlFSrOZV8cmN50PHol7i0LMc8mYVSsiatVf6sJ3VGy+zRKDZoQQojLZqYl49E/4d6xHPPMEYzgCrhaDbAGlK1Yo6izJ8oIwxGEK+pqnM174/1tA+7NC0hf+Qnu9V9ZQ7ZEXY2tXMWizmaJJgGaEEIUc6Zp4j+6F3fcMrz7fgGfF3t4U5zth+Jo0B7D7izqLIoyyrDZcDbsgKNBe3yJu6wXCjZ+jXvzfJyqO66W/bFVqpl3QuIiEqAJIUQxZbpT8exZg2dHLP7j8eAMxql64Izsjb2qDBUjig/DMHDUUjhqKXwnD+PZshDPzhV44mJxNGiHq9UA7DUbF3U2SxQJ0IQQopjxnYi3BpTdvRo8adiq1SOo+x04G8fIgKGi2LNXqY295xhcHYbh2bbUqvndvx57eFNrKqmrojEM6SOZFwnQhBCiGDC9brz711tDZBzZDXYHjoadcEX2xlajUakbykCUfrbylQnqOAJXm2vx7FyBe+siUhe/jq1yLZytBuBs3BnD4co7oTJKQtgi9vzzk9m5My7H5d988xVLlizM9/36fD6eeOIx0tLS8j3t3Pj9fh5//FHOnTtXqPsVorjynzlK2s+fkzJrHGmx0/CnniEo5mZCb3uNcr3/hr1mYwnORIlmOINxtexHyMgXCe5zD9hdpK/4kJTPHiV94zeYaclFncViSWrQitiECU/munzr1s20aXNlc4hmZ968OXTsGENwcOE2l9hsNoYMGcpHH73HffddPFisEGWB6ffhPbAZT9wyfAe3gWHDcVUba0DZOs2l+UeUSobNjrNxDI5GnfAd3oF7ywLc6/+He9P3OJv1wNWyH7YKF89IU1aVmQDtzOpVnF654orSOD8Z94UqdetBxS5dc932iSceo1+/AfTqdTUAY8aMYsKEibzxxiuMGXMXbdq045133mTFiuU4HHaGDBlGgwYNWblyBRs2rKNateo0adKU55+fzJEjidjtdu66635iYrowffpUtm/fxtGjiQwdOpxZsz5lzpxvsdlsbNy4npkzP+Hll9/IyItpmsyZ8znvvfcJAFOmTCI4uBy7du0kOfksd911P4sWzWfPnl10796LBx/8P3w+H2+//Tq//roBn8/PoEHXcvPNt+H1enn55efZt28vJ06coHHjxkyaNIUTJ07wxBOP0rBhI3bt0lStWo3Jk5+nYsVKdOzYmdde+w+3334nISGhV1QmQpQk/nOnrAFld/yImXICI6QKrnZDcTbriS2kSlFnT4hCYRgGjjqROOpE4jsRj3vzQjzbl+HZ/gOOhh2sfmph9Ys6m0WuzARoRa1//0EsWbKAXr2uJj7+AG63m6ZNm2Usj439ga1bN/PJJ7Pxer3cd99YXn75Dbp160GbNu3o1KkzTz45gbZt2zNy5CgOHTrIffeN5cMPZwLgdqczY8aXACxfvoxff91Au3YdWLjwewYNujZLXvbs2UVoaGjGPJwAx44lMXXqhyxY8B3//vczfPbZ/wgKCmLo0EH89a9/Y+lSq5n1gw9m4na7eeSRB2jWLBLTNHE4nEyd+iF+v5+HHrqHNWtWoVRz9uzZzeOPP0XTps345z8fY/HiBYwYMRK73U6jRk3YuHF9iZiBQIgrYZomvsM7rAFlf/sVTB/2OlE4u9yG46rWGDZ7UWdRiCJjr1qXcr3/hr/DcNzbFuPZsRzv3rXYaze3ArW6LctsE3+ZCdAqdumaZy1XXq5kJoEuXbrx6qsvcu5cCkuXLqJ//4FZlm/atIE+fa7B5XLhcrn46KNZF6WxceM6xo+fCECdOhFERrYgLm4bAJGRf0yKPnjwEBYtmk9UVEs2bFjHuHETsqQTHx9PWFjWcWliYroAULNmOA0aNKJKlaoAVKxYkbNnz7B+/S/s3r2LDRusCcpTU8+xd+8ehg27kYoVKzF37hccOPAbBw/Gk5qaCkCVKlUzgtCGDRtz5syZjP2Fh4cTHx9/mWdRiJLDTE+xRlrfEYt5OhGCQnC2vAZX817YKoUXdfaEKFZsoVUJjhlJUNsheHb8iHvbYlIXvoKtSgSu6AE4GsVg2MtMyAKUoQCtqDmdTrp27c7KlStYtmwJL730epblDoeDzA8JCQmHqVw5a5PHhVNZgYnP5wMgKOiPqTV69+7LtGlvExu7lM6du2ZZBmAY1v4uzN95dvvFT/Q+n5/77nsoY0L0U6dOUa5cOVau/JH335/KjTeOZNCgIZw6dSqjGdjlyvp2TubmYbvdgc1WNp+KROllmib+pP3WsAJ714LPg61mY4La/A1Hww7yxpoQeTBc5XFFD8TZ4hq8e9fi3rKAtOXvY/wyB2eLfrgie2G4yhd1NguF9EQtRP37D2L27BlUqlSZ8PBaWZZFR7dl+fJleL1e0tLSGDfuQZKSjmK32zOCsHbt2vPdd/MAOHToIFu3biYqqtVF+wkODiYmpgvTpr3NwIHXXbQ8IqIuCQmHLyvv7dq155tv5uH1ejl37hz33Xcn27dvZf36X+jTpy+DBw8hNDSUX3/dgN/vyzO9hITDRETIQJuidDA96bh3/si5ryZxbt6zePetw9m0K+WHP0vI9RMD8xNKcCbEpTLsjsBvaDLlBj6CrUpt3L98QfLMR0j7eTb+5ONFncUCJzVohahVq9YkJyczdOiIi5b17NmbnTvjGDPmNvx+kxtvvIV69a6iffuOTJ36NqGhoTz88GO8+OIU5s//FsMwGD9+ItWrV892X1df3S8QwLW4aFnjxk05ffoUycnJWfqh5Wbo0BEcPBjPX/96Kz6fj0GDrqNt2/ZUqlSZZ575J0uXLsLhcNKyZSsOHz5Mu1xePPX5fOzatZOJE5+5pH0LUVz5Th7CE7cMz67V4EnFViWCoK6jcTbpguEqV9TZE6LEMwwDR91WOOq2wnfsN+uFgq2L8WxdgqNxJ6ufWrXS+bBvZPdWYglTH9h//HjyRU2AiYm/Ex5+Vb7t6Er6oBUmn8/HtGlvU6VKFUaOHJXtOl9+ORubzWD48JsLOXfw00/L2bJlM/ffX7yG2cjp+xIWVoGkpLNFkCNR1LIre9PntQaU3RGLL0GDzYGjYQdriAwZs6zUkN998eU/m4R762I8O1eANx17RAsrUKsTmS+/v4Iue5vNoFq1UIAGwG85rSc1aKXQ2LGjqVSpMi+88EqO69xwwwgmTvwHgwdfX6hjofn9fr799muefnpyoe1TiPzgP5uEZ8ePePQKzNQzGBXCcHW8Cafqhq1cxaLOnhBlhq1CGMFdbiOo3VDccbF4ti0hdf5L2KrVwxU90OrvaSv54Y3UoF2GklKDJv4cqUETF6perTyJv67GHReL78AWMMBRr7U1WXlECxlQthST333JYfo8eHavxrNlIf5TCRih1XC16IezWY8/1dVAatCEEKIYMr3p+BJ24T20nfjfN+A9nYRRrhKuttdZA8qGVivqLAohMjHsTlzNeuJU3fEd2Ix7y0LSf/6M9I1f44rsjbPFNdjKVy7qbF62Uh+gmaYpfUJEnkpBTbL4k0zTj//YAbyHtuE7uB1f4m7we8HmoFz9KBwdbsJRv02paDIRojQzAlOmOa5qg+/oXtybF+DePB/3lkU4m3TG2Wog9iq1izqbl6xUX3EcDhcpKWcICakoQZrIkWmapKScwSHDIJQZ/rPH8B7abgVkh+Iw063Jmm1V6+Js0RdHnSjstZpSo1Z1aeYSogSy12hEuWsewH/mKO4tC/HolXj0T9jrRVsvFNRSxT4uKNUBWpUqYZw8mURy8ql8Sc9ms+H3Sx+00sjhcFGlikzSW1qZ7nN4D+/Ad3A73kPbMU8fAcAoXxn7VdFWQFYnClv5SkWcUyFEfrJVrEFwt7/gan9DYL7PpaR+9zy2sAbWCwX122PYimdf0lIdoNntDqpXr5X3ipdIOo0KUTKYfi++o/syAjL/0X1g+sERhL12MxyRV2OPiMJWuXaxf4oWQlw5W3AFgtpdjyt6IJ5dK3FvWUTa0rett7Fb9cepumM4gvJOqBCV6gBNCFE2mKaJ/3RCRpOl9/AO8KSBYVhPyq0HY49ogb1GozI3n58Q4g+Gw4Ursg/OZr3w/r4R9+YFpK+agXv9PJxRfXBG9QUqFHU2AQnQhBAllD/1DL5DcfgObcd7cDtmygkAjIo1cDbujD0iCkft5hhBIUWcUyFEcWPYbDgbtMfZoD3exN14Ns/HvfFb3JsXkDzkQQi7eBrFwiYBmhCiRDC9bnyJuzMCMv/x360FQSE4ajfHHjEER50obBWlL6EQ4tI5wpvgCP87/lMJuHcsx16+eAw8LQGaEKJYMk0//uPxGQGZL3EX+Dxgs2Ov2RhX+2E4Ilpgq16/2HbyFUKUHLbKtQjufAvlwiqQXAz6m0uAJoQoNvzJJ/4IyA5tx0yzLpK2KnVwNu+NIyLKej3eWXjTkwkhRFGQAE0IUWRMdyq+hJ0ZAZn/VAIARrlK2CNa4Ihogb1OJLaQKkWcUyGEKFwSoAkhCo3p9+FP2p8RkPmO7AXTB3YX9tqKoGY9reEvqkTI8BdCiDJNAjQhRIExTRPzzJGMgMx7eAe4UwEDW1h9XNEDsUdEYa/ZGMPuLOrsCiFEsSEBmhAiX5lpyXgPxeE7tM0a/iL5OABGheo4G3YMDH8RiREcWsQ5FUKI4ksCNCHEFTF9HnxH9uA7aAVk/mO/Aya4yuGoHYm99WAcdaIwKtaQZkshhLhEEqAJIS6LaZr4Tx7MmEbJl6DB6wbDjr1mI1zth1rjkYU1wLDZizq7QghRIkmAJoTIkz/lpDWF0sFt+A7FYaaeBqxxg5yqR2D4i2YYrnJFnFMhhCgdJEATQlzE9KRnHf7i5CEAjOAK2OtEWQFZnShsoVWLOKdCCFE6SYAmhMD0+/Ef+y1QQ7Yd35E94PeB3Yk9vClBTbtaAVm1uhiGjNovhBAFTQI0Icoo/5mjWYe/SE8BwFbtKlwt+2OvE4U9vAmGw1XEORVCiLKnQAM0pVQsUAPwBD66G2gETAScwGta6/8G1u0LvAKUAz7XWk8syLwJUdaY6SmB4S+sqZTMs0kAGCFVcdZvawVkdSKxlSseEwULIURZVmABmlLKAJoCV2mtvYHP6gCzgXZAOrA6EMTtBz4AegLxwPdKqYFa6wUFlT8hSjvT57WGvwgEZP5j+8E0wRmMo3Zz7C3744iIwqgULsNfCCFEMVOQNWgq8P/FSqlqwHvAWWCZ1voEgFJqDjAC+BHYrbXeH/h8BnAjIAGaEJfIGv7icMYAsdbwF+lg2LDVaIirzRDsES2w12iAYZPeDUIIUZwV5FW6CvAD8CBWc+Zy4HMgIdM6CUBHoHY2n0dczs6qVSucUcnDwioUyn5E8VJcy92bfJLU/Vsy/vMlnwDAWbU2IdG9KdcgmnJXRWELDininJZcxbXsRcGTsi+7ikPZF1iAprVeA6w5/7dSajpWH7PnMq1mAH7ABpjZfH7Jjh9Pxu83817xCoSFVSAp6WyB7kMUP8Wp3E1vOr6EXdYAsQe34z8RD4ARFIq9TiRBbaOsQWIrVMcEzgHnzvrhbPHIf0lTnMpeFC4p+7KroMveZjMuqVKpIPugdQOCtNY/BD4ygN+AWplWCwcOAwdz+FyIMs00/fiPHcB7aBu+g9vxJe4GvxdsDuzhTXB1HIGjTgts1evJ8BdCCFGKFGQTZ2XgWaVUF6wmztuBUcAMpVQYkAIMB+4CtgBKKdUY64WBW7FeGhCizPGfPZZRQ+Y7FIeZngyArWpdnC364qgThb1WUwxHUBHnVAghREEpyCbO75RSnYBfATvwX631KqXUP4FYwAW8r7X+BUApdQcwFwgG5gNzCipvQhQnpvsc3sM7Mua2NE8fAcAoXxn7VdFWQFYnClv5SkWcUyGEEIXFMM2C7bdVCOoD+6UPmigo+V3upt+L7+i+jIDMf3QfmH5wBGGv3cwKyCKisFWuLcNfFDH5zZddUvZlVyH2QWuA1fUrW/KuvRAFzDRN/KcTrIDs4HZ8CTvBkwaGgS2sAa7WgwPDXzTCsMtPUgghhARoQhQIf+oZfIfiMqZSMlOs4S+MijVwNu6MPSIKR+3mGEEy/IUQQoiLSYAmRD4wvW58ibsyAjL/8QPWgqAQa9T+iCHW8BcVw4o2o0IIIUoECdCE+BNM04//eHxGQOZL3AU+D9js2Gs2xtV+GI6IFtiq18ewyfAXQgghLo8EaEJcIn/y8YyO/b5DcZhpVidSW5U6OJv3xhERhb2WwnAGF3FOhRBClHQSoAmRCzMtmWNL5pCyawP+U9ZsZMb/s3fncXKVZd7/P7V2V+/7lu6s3X0ngYRASEISdtmCIjqIiIiCiuO4DaPobM4+88zz+83gzDiOjoZNjCLIomxR2RTIAoRVSLizELL3nqT3rqqu8/xRlUp31krS1VXV9X2/XnnRfboqdXUO1blyG4YdkAAAIABJREFUn/t8r0AxnvrT8dafjmfSbNz5pSmuUkREJho1aCJH4YQG6V95O5HOHXgmzSJn5gXR+IvSesVfiIhIUqlBEzkCZzjMwFPfI9KxjeqPfYv+0pmpLklERLKIdi+LHMJxIgz+7g6Gd75N7vk3k9+8INUliYhIllGDJjKC4zgMrbmP8Ja1+Bdei8+cl+qSREQkC6lBExkh+OYThN5+Ct+cy/GfcWWqyxERkSylBk0kJvTu8wRffhBv42JyzrlONwKIiEjKqEETAcLvv87gC3fjqT+d3As+h8ult4aIiKSO/haSrBfeYxl45vu4K6YRuPQrGlguIiIppwZNstpw1w4GfvOfuAvKCSz7M00BEBGRtKAGTbJWpKedgSdvx+XLJfDBb+LOLUx1SSIiIoAaNMlSkYFu+p+8HWc4RGDZbbgLylNdkoiISJwaNMk6TnCAgV//B05vF3mX34qnbFKqSxIRERlFDZpklZEjnAKXfAlPTVOqSxIRETmMGjTJGvERTrveIff8m/FOmZfqkkRERI5IDZpkhdEjnD6uEU4iIpLW1KBJVgi+8fiIEU7LUl2OiIjIMalBkwkv+O7vCb7ykEY4iYhIxlCDJhNa6P1XGXrhHjwNc8m9UCOcREQkM+hvK5mwwnssg8/8AHflNAKXfBmXWyOcREQkM6hBkwlpuDM2wqmwkrwrvo7Ll5PqkkRERBKmBk0mnEh3OwMrb8flCxC48jZcuQWpLklEROSEqEGTCSUy0E3/yn+PjnC68hsa4SQiIhlJDZpMGE5wgIGV38Hp3UveFX+Gp1QjnEREJDOpQZMJwRkORUc4dW4ncOmX8FQ3prokERGRk6YGTTKe40QYfG55dITTBZ/DO1kjnEREJLOpQZOM5jgOQ6t/Svi9l8lZdB2+5qWpLklEROSUqUGTjBZ8/TFC7zyDb+4VGuEkIiIThho0yVjBDb8juO5hvE1LyFn08VSXIyIiMmbUoElGCm19laEXfxwd4XTBZzXCSUREJhT9rSYZJ7z7XQaf/QHuyuka4SQiIhOSGjTJKMOd2xn4zX/hLqwi74o/0wgnERGZkNSgScaIdLcx8OTtuPwBAld+QyOcRERkwlKDJhkhMtBN/5O340TCBK68TSOcRERkQlODJmkvOsLpdpy+AyOc6lJdkoiISFKpQZO0Fh3h9N9EOncQuPTLGuEkIiJZQQ2apC0ncmCE0/rYCKczUl2SiIjIuFCDJmlp1AinczTCSUREsosaNElLwdcfJbT+GXxzl+GfqxFOIiKSXdSgSdoJrn+O4LpH8DYv1QgnERHJSmrQJK2Etq5jaNW9eCafQe75N+NyuVJdkoiIyLhTgyZpI7x7A4PP/C/uqhkELvmSRjiJiEjWUoMmaWG4YxsDv/ku7uIq8i6/FZdXI5xERCR7HXeJwhjzGvB94GfW2v7klyTZJtLdxsDK2AinZbdphJOIiGS9RFbQvgqcB2wxxnzPGHN6kmuSLBLp30//k/+OExmOjXAqS3VJIiIiKXfcBs1au8pa+xlgFrAe+JUx5gVjzLVJr04mtOgIp+/g9O8jb9nXNcJJREQkJqE9aMaYEuBG4AvAfuAB4EZjzPIk1iYTmDMcYuC33yXStZPApV/BUzUj1SWJiIikjeM2aMaYFcBWYCHwJ9bas6y1/w38EXBNkuuTCciJRBh89ocM795A7oWfw9swN9UliYiIpJVEcgzWA39mrW0fedBaGzbGaP6OnJDoCKcVhLeuI+ecT+BrWpLqkkRERNJOIpc4P31oc3aAtXbDGNcjE1zwtUcJrX8W/xlX4p97RarLERERSUuJNGjbjDFLjDHKTJNTElz/LMFXH8HbfC7+hbrHRERE5GgSucQ5C3gRCBljhgAX4Fhri5JamUwoofdeYejFn2iEk4iISAISadDOO5UXMMb8O1Bhrb3JGDMPuAMoAp4HvhjbyzYZWAFUARa4wVrbeyqvK+kjvHsDg8/+EE91Y2yEkyfVJYmIiKS1RHLQtgGLgL8H/gW4IHbsuIwxHwA+M+LQCuAr1tpmoitxt8SOfx/4vrV2JrAO+JtEvwFJb9ERTv+Fu7iawBUa4SQiIpKIRGI2bgP+CngTeA34ujHm2wk8r4xoQ/d/Yp9PAQLW2rWxh9wDXGuM8QHnAw+OPH5C34WkpfgIp5x8AlfehisnP9UliYiIZIRELnF+GjjXWtsNYIy5E1gL/PNxnvdD4K+BhtjndcCeEV/fA9QDFUC3tTZ8yPETUl4+PvMbKysLx+V1Ml24dy+7H7gdFw51N/wt/ooTPqVpRec9e+ncZy+d++yVDuc+kQaNA81Z7OP9xpjQsR5vjPk8sMNa+4wx5qbYYTfgjHiYC4gc4Tix4yeks7OXSOTQ32ZsVVYW0t7ek9TXmAic4AD9j/0rkd695H3oz9nvFEMG/7npvGcvnfvspXOfvZJ97t1uV0KLSok0aO8bY/6U6D4xgC8D24/znOuAWmPMG0AZUEC0Casd8ZgaYDfQBhQbYzzW2uHYY3YnUJekISccjI1w2kXgils1wklEROQkJJJt9ifAR4H+2K9rgC8d6wnW2kuttadba+cBfws8aq29GRgcMX3gRmCltTYEvEC0qYPoJdWVJ/ydSMo5kQiDz/1oxAinOakuSUREJCMl0qDVWWsvBIqBUmvteUDzSb7eDcB/GGPeJbqq9t3Y8S8BXzDGrCca63HcmxAkvTiOw9Cqn0RHOC2+XiOcREREToHLcY68b8sYcybRfWI/BT4Z+xjAB6yw1jaNS4XHNxXYqj1oqTW07hGCr/0K/7wPkjPBpgTovGcvnfvspXOfvcZxD9o04P2jPe5Ye9D+BLiU6N2XD484Hj7kc8lywfXPEnztV/jMefgXfCzV5YiIiGS8ozZo1tovABhj/tlaq0uOckSh916OjXCaR855N2mEk4iIyBhIZA/aA8aYJwCMMXOMMW8YY0yS65IMEN61nsFnfxQb4fQnGuEkIiIyRhJp0H5AdH4m1to/EB359MMk1iQZYLhjGwO//S7u4hqNcBIRERljiTRo+dbaRw58Yq39JdFh55KlIvtbR4xw+oZGOImIiIyxRBo0xxgz98AnxphZwHDySpJ0FunfR/+T/w6OQ96Vt+HOL011SSIiIhNOIpME/gb4vTHmD7HPZxLNM5Ms4wT7GVh5O85AN3kf+nPcJbXHf5KIiIicsOM2aNbax2M3BSwlGrHxkrW2LemVSVpxwkEGfvNfRLp2E1j2Z3iqpqe6JBGRMRMJhQh3dRLq6CDU2YGrppzIVIPb7091aZKljtqgGWMuttY+a4z5oxGHfcC5xhistcpCyxJOJMLgsz9keI8l9+I/xlt/eqpLEhE5IZFQkHBnF6HODkIdHYQ7O+Ifhzo7GN63b9Tj2wB3IEDB2QsoWryUQFOzYoRkXB1rBe164Fngq0f4moPCarOC4zgMvXgv4fdfJWfxJ/E1Lk51SSIih4kEg6NWwKJNWGf84+H9oxswPB58pWV4KyrIP20OvooKfOUVeMvL8ZWXkxfqY8fKp+l5+SW6X3geX0UlhYuXULR4Kf6qqtR8k5JVjhVUe0vswy9aa+041SNpJvjqLwm9+zv88z6Ef85lqS5HRLJUJBgcveoVXwWLNmHD+/ePfoLHg6+sDG95Bflz5uArr8BXUYG3PNaIlZTg8hw9u7Gkcjqh2qlU3XAjva+9SveaVXQ9/ihdj/2K3MYmihYvpXDBAjx5uotdkiORmwSeMca8B/wIeNBaO5jkmiRNBN95JjbC6Xz8C65JdTkiMoFFhoYIdXZGm65Rq2CxS5Dd3aOf4PHgKyvHV1FB/pwz4itg0SasHG9JKS53IkEFx+bOyaFo8RKKFi8h1NVFz0tr6F69iraf3EP7fSvIn3cWRUuWkD/7dFzeRP5KFUlMIv83TQYuB24C/s0Y8xCw3Fr7ZjILk9QKbXmZoVUr8E45k5zzPqO9FyJySqIN2Iim65BLkcM9oxswl9eLt6wcX3kF+WfMizdfvvJKvBUVeIuLx6QBOxG+sjLKln2Q0iuuZGjbNrpXv0jPyy/Ru+5lPIVFFC46h6IlS8lpmKyfmXLKErmLMwKsBFYaY2YC9xAdpK65PhNUeNd6Bp/7IZ6aJnI/oBFOInJ8kcHB2OXGdsIdIzfgdxLu6GC4t2fU411eb2y/VwW5kydHLz2OWAXzFI1/A5Yol8tF7tSp5E6dSuXHP0Hf23+ge80q9v/uWfY9/Vv8k+opWrKUokWL8ZaUpLpcyVDHbdCMMV7gw0RX0BYB9wO3HOs5krmG29+PjnAqqSVw+Z/i8uoWcxGByOBAvOEKdXYc0oR1EOntHfV4l9eLN9Zw5U6ecnD/V+yYp6gobRuwE+HyeimYdyYF885kuLeXnldepnvNKjp+cT8dDz5A3mmnU7R4KQXzzsSdo5F4krhELnHuAd4G7gSutdYOJbckSZXI/paDI5yWaYSTSDYZHhg42HQdqQHr6xv1eJfPF7/rMXfq1OjHI1fACidGA3YiPAUFlFx0MSUXXUywpYXuNavoXrOaluX/izs3l4KzF1K0ZCmBxqas+7ORE5dIg7bEWrsp6ZVISsVHOAF5V35TI5xEJpjh/v6Ddz12HN6ERfoPacD8/lgDVkHutOkHYyhGroBpn9VR+WtqqPjoNZRf/VEGNlq616yOrq69+DzeigqKzoneeOCvrkl1qZKmEtmDpuZsgnOC/Qw8eTvOQE9shJN+YIhkmuH+vtHRE4c0YZH+/lGPd/n98aYrd8YMfGUVoy5DegoL1YCNAZfbTd7MWeTNnEXVJz9F7+uv0r1mNV1PPEbX44+SO6ORoiVLKTx7IZ58XbWQg3RPcJaLj3Dat5vAFRrhJJKOHMch0t9/8K7Hjg5CXaPviIwMDIx6jisnJ365MbexCV95+ahVME+BGrDx5s7Jia6cnbOE0N699KxdQ/eaF2n7yY9pv++n5M87k6LFS8k/TZEdcuxRTx+x1v7SGJOjfWcT08ERThs1wkkkhRzHIdLXd0j2V+eozw9vwHJjDVc5gabmeADrgSbMXVCgBiyN+UpLKVt2JaVXLGNo+za616yi56W19K57BU9hYSyy41xFdmSxY7Xo/wT8ElgDnDU+5ch4iY5w+nF0hNOSG/A1npPqkkQmrHgD1tFBqLP9sEuR4c4OIoOjM8DjDVhFBXnNZlQKvq+iAnd+vv7ingBcLhe5U6aSO2UqlR+7bkRkx3Pse/qpaGTH4iUUnbMYb4n2BmeTYzVo3caYjcAkY8xbh37RWjs3eWVJsgXXPUzo3d/jP/Mq/KdfmupyRDKa4zhEentH3fV4aBPmDI1uwNy5uXgrKvFVVpI3c9aIDfjRbDA1YNnnsMiOdS/TvWY1HQ8+QMdDvyBv9mkULVlKwbyzFNmRBY7VoF0BnEk0XuNIA9MlQwXfforg64/hm3k+/rP/KNXliKQ9x3EY7u05LHriYCxFJ87Q6J0g7kAgugJWWUnerFmHzYJ05+WpAZOj8hQUUHLhxZRceDHB1pGRHT+MRnbMXxCN7GhqVmTHBOVyHOeYDzDGNAO7gfmAD3jJWttzzCeNr6nA1s7OXiKRY38vp6qyspD29nT61k9caMtLDD7zv3innknuJV/WlIAETITzLokb7u9n8L3NDGzahNOyk77drYQ6O3CCwVGPc+flHZb9NXIWpIZoZ7Z0fN87kQgDmzbSvWYVveteITI4iLe8PHYJdCn+Gt2BPxaSfe7dbhfl5QUA04D3j/a4RG4TKQZ+B7QSHe9Ub4z5kLV29amXKeMpvPMdBp/7UXSE08VfVHMmAoQ6OxnYvJGBzZsY2LSJ4K6d4DjgdpM/ZTL+mlryTp8zqgmLNmB5qS5dsozL7SbPzCTPzCRy/afofeM1ulevouuJx+l6/LFoZMfiJdHIjoKCVJcrpyiRBu124AZr7XMAxpiLge8A2lWeQYbbtzLw1H/jLqnTCCfJWk4kQnDXTgY2bWRg82YGNm8k3NUFRDflBxobKZx/NoHGJnKnz6C6viLtVlFEIBbZsWgxRYsWE963l+61a+hevYq2FffS/vOfkX/GvGhkx+lzFNmRoRI5a4UHmjMAa+2zxhj90zGDREc4fQdXbgGBKzXCSbJHZGiIwa3vRVfHNm9icMvmeFyFp6SEvKZmcq9oJtDYRM6kelwerSpL5vGWlFJ2xZWUXr6MoR3b6V69ip6X1tD76jo8BbHIjsVLyZkyRfseM0giDZpjjJlird0GYIyZCgwntSoZM5G+vSNGON2GO68kxRWJJE+4uzvaiG3ayMCWTQxu2wbD0R9X/kn1FC48h0BTE4HGJrzlFfrLSiYUl8tF7uQp5E6eQuXHPk7fO29HIzt+/xz7nnkKf90kihYvpfCcxfhKFdmR7hJp0P4RWGuMeRpwgMuBLyW1KhkTzlAfAytvxxnsJe9Df4G7WBtIZeJwHIdQa2t0/9im6ApZqLUFiMYV5E6bTtnly8htbCIwo1FjdCSruLxeCs6YR8EZ8xju64tGdqxeRcdDD9Dx8C/ImzU7Gtlx5nxFdqSpRGZx/tIYswG4mOhNAv9qrd2Q9MrklBwc4bSHwBVfx1M5NdUliZwSJxxmcPu22P6xTQxu3sRwT3R/mLuggEBjE8XnnU+gqZmcyVNw+3wprlgkPXjy8ym54CJKLrgoGtmxdg3da1bRcsePcOXkUjj/7GhkR7NRZEcaSWjnoLXWAjbJtcgYcSLDDD7zA4ZbNpH7gS/irT8t1SWJnLCRcRcDmzYy+P7WeNSFr6qa/DlzCTQ2E2hqwldTq8uVIgnwV9dQcfVHKb/qagY2b6J79Sp6171M9+oX8ZbFIjsWL8FfU5vqUrOebu2YYOIjnLa9Ts6ST+GbsSjVJYkkJBp3sSl+yXJk3EXO5CkUn3/hwf1jxdpLKXIqXG43ec2GvGZD5Pob6H3jdbrXrKLrycfpeuIxcqdPp2jxuRQuUGRHqqhBm2CiI5yej41wuiTV5YgcUTzuIpY9dljcxYwZFM7/SDTuYtp03Lm5Ka5YZOKKRnacQ9Gicwjv20f3S7HIjp/eS/v9PyN/7hnRyI45cxXZMY6O+ydtjLnXWvvp8ShGTs3BEU4XaoSTpJXjxV0EGpsJXN4U3T+muAuRlPGWlFB2+TJKL7siGtmxZjU9a9fQ+9qr0ciOhYsoWrKUnClTta0gyRJphecZY1zW2uTOUZJTEtq8lqHVP8M7dT45535abxxJqXjcReyS5eFxF4vi+8cUdyGSfkZFdlxzLX3r36Z79Wr2P/879j37NP66umhkx6LF+MrKUl3uhJRIg7YbeMcYsxboPXDQWvu1pFUlJyS8820Gf7ccT20zuRf/se7CkXGVSNxF6WVXRPePTW/UfhaRDOPyeimYO4+CufMY7u+j55VX6F6zio6HfkHHww+SN3M2RUuWUHDW2YrsGEOJNGhrYr8kDQ23vcfAb2MjnC77mkY4SdIdM+4iP59AU3M07qKxiZwpUxV3ITKBePLyKbngQkouuJBgayvda1dHIzvuXI5rxb3RyI7FSwmYmVosOEWJ5KD9gzEmADQC7wC51tr+pFcmxxXZt4eBX/8HrkCRRjhJ0hwz7qKyKh53kdvYhL+mRj+URbKEv7p6dGTHmlX0rnuF7tWr8JaVUXTOEoqWLFVkx0lK5CaBRcAjQBhYArxpjLnKWrs62cXJ0Y0e4fQNjXCSMaO4CxE5EaMjOz5F7xuv0b16NV0rn6DrycfJnTadosVLKFx4jrY4nIBELnH+O3AJ8FNr7U5jzI3AfwELklqZHFV8hNNQn0Y4ySlR3IWIjCW330/RwnMoWnhIZMfPVtB2/30UzJ1H0ZIl5M85Q5Edx5HIn06etXa9MQYAa+2Txph/SW5ZcjQHRzi1EFimEU5yYhR3ISLj5UBkR9nly6KRHatX0f3SGnpffxV3QQFFCxdRtHgpOVOn6U7uI0ikQQsZY0qJDkrHHOjUZNyNGuF0yZ/gnTQ71SVJmjtm3EXdJMVdiMi4yGmYTOV1k6n42Mfpe+dtetasYv/zv2ffs8/gr62LXgI9ZzG+svJUl5o2EmnQ/hn4PVBrjLkPuAz4QlKrksM4jsPQC/dERzgtvRHf9IWpLknSzKi4i9glS8VdiEg6cXk8FMw9g4K5Z0QjO9a9Qs+a1XQ8/CAdjzxE3sxZFC1eSsFZ87N+S4XLcY6fP2uMaQQuBTzAM9baDcku7ARMBbZ2dvYSiSQ3S7eyspD29p6kvsbRDL38IME3Hsd/1ofJ0ZSAcZXK834sI+MuBjdvZmDzxsPiLgIzmgg0Ke7iZKXruZfk07kfX8G2NnpikR2h9nZcOTkUnnU2RUvGP7Ij2efe7XZRXl4AMA14/2iPS3SHno9ocxaK/ZJxFPzDbwm+8Xh0hNP8j6a6HEmRROIuchubCDQ2K+5CRDKKv6qK8g9/hLKrrmYwFtnR88rLdK9Zhbe0jKLFSyhavAR/bV2qSx03icRs3Az8K/Abok3a3xtjvmKtfSjZxQmENq9haI1GOGWjUFdn/M7Ko8ZdNMbiLkoUdyEimc/lckVX/5uaqfzEDfS98Trda1bR9esn6XrycXKmTqNoyVKKFizCU1iY6nKTKpEVtK8DZ1pr9wAYYyYDjwNq0JIsvPNtBp+7A0/tTI1wmuCicRe7RoxLUtyFiGQ3t99P4cJFFC5cRHj/PnpeeonuNS/S/rMVtN9/H/lzz4juV5s7MSM7EvmOggeaMwBr7XZjjC5zJll8hFNZHYHLNcJpookMDTH4/taD45KOFnfR2EROfYPiLkQkq3mLSyi97HJKL7s8GtmxZjXda1fT9/pruPPzKVx4DkWLl5I7beJEdhy1QTPGnBX78E1jzPeAHwLDwE3AquSXlr0i+/YwsPI70RFOy76By5+X6pLkFIV7uqNRF5uOE3fR2IS3QnEXIiJHk9MwmcqGyVRccy3969+he80qul98nv3PPYOvpoaixUspOmcJvvLMjuw41graoZcwPzjiYwf42tiXI/ERTm43eVfephFOGehg3MXB/WOKuxARGVsuj4f8OXPJnzOX4f5+el+NzgHtfOQhOn/5MAEzk6LFSymcf3ZGbgtJKGYjzU1lgsRsOEN99D/6r0R6O8i76i/wVExN2mtJ4o533o8bdxG7s1JxF5lHUQvZS+c+cwXb2+hZGx0xFWpvw+X3UzD/bIoWLyVv5qzj7ufOmJgNY0wN0cuaZSOPW2u/dUoVyihOeIiBX/8nkf2xEU5qztLWqLiLzZsY3Pre6LiL0+eS26S4CxGRVPBXVlF+1dWUfejDDG7eHIvseImeNavxlpZSuGgxRUuWklM3KdWlHlMiNwk8CuwEtiS5lqzlRIYZePr7DLduJveSL2mEU5oZ6uik+6XXonvHNm9iaOehcRcXHNw/prgLEZG0EI3siIZ1V17/SfreeIPuNavY+9tfs/fXT0YjOxYvoXDhIryFRaku9zCJNGh+a62i65PEcRwGn7+H4e1vknPup/FNX5DqkiRmeGCA9gfuY+MLzwMH4y7Kr7qaQFOz4i5ERDKE2+encMFCChcsJLx/Pz0vr6V79Sra7/sp7Q/8nPw5cylavJT8uWekutS4RBq0V40xp1tr3056NVko+MqDhDe+gP+sq/HPvjjV5UhM/7sbaLn7DsJdXdRdfRXeufMVdyEiMgF4i4spvfRySi+9nKEdO+heu4rutWvoe+N13Hn5eG/9CkyfleoyE2rQVgFvGGP2MGLMk7V2etKqyhLBP/yG4BtP4Jt1Ef75H0l1OUI0n6zj4QfZ98xT+Kqqafjzv2Ly4rO0WVhEZALKaWigsuETVPzRtfRveIeeda/g9vmIpLowEmvQvgl8kpPYg2aM+UfgY0RjOe601n7HGHMJ8B0gANxvrf127LHzgDuAIuB54IvW2vCJvmamCG1azdCa+/BOO5ucpTcq9yoNDGzZTMtdywm1tlJy8SVUXHMt7pycVJclIiJJ5vJ4yD99Lvmnz6U0Te7gTaRB22etfeBEf2NjzAXAxcBcosPW1xtjngHuAi4AdgBPGGOWWWtXAiuAz1tr1xpj7gRuAX5woq+bCcI73mLwd3dGRzhd9AXd5ZdikVCIzkd/yd5fP4m3rIz6b3yLvFm6UUNERFInkQbtWWPMvxMNrh06cNBa+9qxnmSt/b0x5iJrbdgYMyn2WiXAJmvtVgBjzArgWmPMeiBgrV0be/o9wD8wARu04bYtDDz1Pdxlkwhc/qca4ZRig9u30XLncoK7dlJ07vlUXnc9nkAg1WWJiEiWS6RB+2Tsv9eMOOYAx92DZq0NGWP+AbgN+AVQB+wZ8ZA9QP0xjk8ow/t2M7DyP3AFigks+zouvxqBVHHCYbpWPkHn44/iKSik7mu3UjB3XqrLEhERARJo0Ky1007lBay1f2eM+f+Ax4Bmos3dAS4gAriPcjxhsVTepKusLDyp54W7O9n16+/g9nqpu/Hv8ZXWjHFlkqj+7TvY+J//Td+WLVScfx7Tv/A5fIXHPq8ne94l8+ncZy+d++yVDuc+kUkCXz/ScWvtd47zvJlArrX2DWttvzHmYaI3DAyPeFgNsJtoEG7tEY4nLJ1HPTmDvfQ/9q9EBnrJu+ov2RfOhzTYgJhtnEiEvU/9hs5HHsKdG6D2i1+m8OwF7BsEBo9+PjTyJXvp3GcvnfvsNY6jno4pkUucc0Z87Ce6wf+ZBJ43HfgHY8y5RFfHrgZ+CPybMaYR2Er08uld1tptxphBY8xSa+0q4EZgZQKvkfac8BD9v/lPIvtbCVz5DTwVU1JdUlYKtrbScvcdDG7eRP6ZZ1H9qc/gLS5OdVkiIiJHlMglzptHfm6MqQPuTOB5TxpjFgKvE101e8ha+3NjTDvRGw5ygSeBB2NPuQFYbowpAl4Dvnsi30g6ciJhBp7+PpHWLdERTnWpD77LNk4kwv7fP0f7L+7H5fFQ87lbKDxniWJNREQkrSXVEWqxAAAgAElEQVSygjaKtXa3MWZqgo/9e+DvDzn2DHDYLAVr7ZvAwhOtJ11phFPqhTo7ab3nLvo3vEPeaadT/ZnP4isrS3VZIiIix3Wie9BcwNlAW9IqmiCCL/+C8MYX8c//iEY4jTPHcehe/SLtP/8ZTiRC1Y2fofj8C7VqJiIiGeNE96A5wHai0wXkKIJv/Zrgm0/im30x/rOuTnU5WSW8fx+t995D35tvEGg2VN/8OfyVVakuS0RE5ISc8B40ObbQptUMrf15dITTkk9p1WYc9bz8Eq0/vRcnGKTyuusp+cClmtIgIiIZ6agNmjHmbkZnk43kWGs/l5ySMld4e2yEU90sci/+YzUH42S4p4fWn/6E3nUvkzttOjWf/Tz+2rpUlyUiInLSjrWC9vYRjlUAtwLvJ6WaDDbctoWBp7+Hu7yewGVfw+XxpbqkrND7xuu03ns3w319lH/0GsquuBKXx5PqskRERE6Jy3ESC3c1xlwC/JhoPtnXrLX9ySzsBEwFtiY7qHbVH/bw0rtthILDh32tNNLFRwceZMiVw8OBaxlw5SWtDonyhoeYu+l5prRsYH9+BetmX0p3QWVSXsvn9xzxvMvEp3OfvXTus9eV505jzpTSpP3+I4Jqp3GMBa9E7uL0Av8K3AR80Vr70NiUODEURHr40OAvibjcPJb7ETVn46Cyaztnvfs0gaE+3p2ygHenLsRxa9VMREQmjmOuoBljmoD7gF7gU9baneNV2AmYyjisoMHh4x+iI5z+D5HeveRd9ReaEpBkkcFB2h96gP3PPYuvpoaaz95CYPqMpL+uRr5kL5377KVzn73GcdTTya2gGWNuBm4HbrfW/stYF5jp4iOcutsILNMIp2Qb2LSRlrvuINTRTsmll1Px0Wtw+/2pLktERCQpjnWJ804gAvyFMebPRxx3Eb2LsyiplaWx+Ainti3kXvJljXBKokgoSOcjD7P3qd/gK6+g/pt/QV6zSXVZIiIiSXWsBm3auFWRQaIjnO6OjnA67yZ8085OdUkT1uD7W2m5cznBPbspvuAiKq+9DndubqrLEhERSbqjNmjW2m3jWUimGHrpAcIbV+E/+6P4Z12Y6nImJCccpvPxR+l68nG8xcVMuvUb5J8+5/hPFBERmSBOeFh6Ntu39lFCb63EN/sD+M/8cKrLmZCGdu6g5c7lDO3YTtHipVRe/0k8efmpLktERGRcqUFLUGjjKnp+92O80xeQs+QGjXAaY87wMHt/s5KOXz2CJy+fui9/lYIz56e6LBERkZRQg5YAJxxk8Pm7CEydg+eiL2iE0xgLtrTQctdyBt/bQsH8s6n61KfxFmbtPSgiIiJq0BLh8voJXPlNqmfNoXN/KNXlTBhOJMK+Z5+m4+EHcXl91NzyRQoXLtLqpIiIZD01aAny1s3E7c8F1KCNhVBHOy1338mAfZf8uWdQ/emb8JYkb7SGiIhIJlGDJuPKcRy6X3ietvvvw+WC6ps+S9HS87RqJiIiMoIaNBk3ob17af3x3fS//RaBmbOouflz+MorUl2WiIhI2lGDJknnOA49L62h7WcrcMJhKj/5KUouvFg3W4iIiByFGjRJqnB3N20rfkzva6+SO6ORms9+Hn91TarLEhERSWtq0CRpel5dR9uKHxMZGKDimo9TevkVWjUTERFJgBo0GXPDfX203beCnrVryJk8hZrbbiFnUn2qyxIREckYatBkTPW9/RYt99zFcE8P5R/+CGVXfgiXV/+biYiInAj9zSljIjI4QPsD97P/+d/hr6tj0lduJXfq1FSXJSIikpHUoMkp6393Ay333Em4s5PSK66k/OqP4Pb5U12WiIhIxlKDJictMjRExyMPsu/pp/BVVdPwrb8i0NSU6rJEREQynho0OSkDWzbTctcdhFpbKLn4A1Rc83HcOTmpLktERGRCUIMmJyQSCtH12K/oWvkE3tIy6r/xLfJmzU51WSIiIhOKGjRJ2OD2bbTcuZzgrp0UnXs+ldddjycQSHVZIiIiE44aNDkuZ3iYrpVP0PnYr/AUFFD31VspOGNeqssSERGZsNSgyTEN7d5Ny13LGXp/K4ULF1H1yRvxFBSkuiwREZEJTQ2aHJETibD3qd/Q+chDuHJzqf3ilyg8e2GqyxIREckKatDkMMG2NlrvvoOBTRvJn3cm1TfehLe4ONVliYiIZA01aBLnOA77f/cc7b/4OS6Ph5rP3kLh4iW4XK5UlyYiIpJV1KAJAKGuTlrvuYv+9e+QN/s0qm/6LL6y8lSXJSIikpXUoGU5x3HoXr2K9p//FCcSoepTn6b4gou0aiYiIpJCatCyWHj/Plp/8mP63nidQFMz1Td/Hn9VVarLEhERyXpq0LJUz7qXaV1xL87gIJUf/wQll1yGy+1OdVkiIiKCGrSsM9zbS9tP76XnlZfJmTqNms/eQk5dXarLEhERkRHUoGWR3jdep/Un9zDc20v5R/6IsmUfxOXxpLosEREROYQatCww3N9P+/330b3qBfz1DUz606+TO3lKqssSERGRo1CDNsH1b1hPy913EN67l7IrP0TZVVfj9vlSXZaIiIgcgxq0CSoyNET7gw+w/7ln8FXX0PCX3yYwfUaqyxIREZEEqEGbgAY2baLlruWE2tsoueQyKj56De6cnFSXJSIiIglSgzaBREJBOn/5CHt/+2u85eXU3/bn5M2cleqyRERE5ASpQZsgBt/fSstdywnu3k3xBRdSee11uHMDqS5LREREToIatAznhMN0PvEYXU88hre4mEm3fp380+emuiwRERE5BWrQMtjQrp203Lmcoe3bKFy8hKpP3IAnPz/VZYmIiMgpUoOWgZxIhL2/WUnnrx7BHQhQ+6WvUnjW/FSXJSIiImNEDVqGCba00HL3HQxu2UzBWfOpuvEzeAuLUl2WiIiIjCE1aBnCiUTY99wzdDz0C1xeLzW3/DGFC8/B5XKlujQREREZY2rQMkCoo52We+5i4N0N5J0+l5qbbsZbUprqskRERCRJ1KClMcdx6H7hedruvw+A6s/cTNG552vVTEREZIJTg5amwvv20vrju+n7w1sEzExqbv4cvorKVJclIiIi40ANWppxHIeel9fS9tMVOOEQlZ+4gZKLP4DL7U51aSIiIjJO1KClkXBPN20r7qX31XXkTp9BzWdvwV9Tk+qyREREZJwltUEzxvwd8PHYp09Ya79ljLkE+A4QAO631n479th5wB1AEfA88EVrbTiZ9aWT3tdfpfXee4gMDFBxzbWUXr5Mq2YiIiJZKmkdQKwRuww4E5gHzDfGXA/cBVwNzAIWGGOWxZ6yAviKtbYZcAG3JKu2dDLc18eeO3/E7v/5b7ylZUz+9t9RtuyDas5ERESyWDJX0PYA37DWBgGMMRuAZmCTtXZr7NgK4FpjzHogYK1dG3vuPcA/AD9IYn0p1/f2H2j98V2E9++n7KqrKf/gVbi8uuosIiKS7ZLWDVhr3znwsTGmieilzv8m2rgdsAeoB+qOcnxCigwO0P7A/ex//nf4a+uY/OWvkTt1WqrLEhERkTSR9OUaY8xpwBPAN4Ew0VW0A1xAhOilVucIxxNWXl5waoUmqLKy8JSev//td9j0X99jqL2duo98mCk3XI/b7x+j6iRZTvW8S+bSuc9eOvfZKx3OfbJvElgKPATcaq39uTHmAqB2xENqgN3AzqMcT1hnZy+RiHP8B56CyspC2tt7Tuq5kWCQjocfZN8zT+GrqKThW39JoKmZzv1DwNDYFipj6lTOu2Q2nfvspXOfvZJ97t1uV0KLSklr0IwxDcAvgeustc/GDr8U/ZJpBLYCnwTustZuM8YMGmOWWmtXATcCK5NV23gbeG8LLXctJ9TSQvFFH6DyYx/HnZOT6rJEREQkTSVzBe02IBf4jjHmwLH/BW4iuqqWCzwJPBj72g3AcmNMEfAa8N0k1jYuIqEQXY/9iq6VT+AtLWXS179J/uzTUl2WiIiIpLlk3iTwp8CfHuXLZxzh8W8CC5NVz3gb2rGdPXcuJ7hzB0VLz6Pyuuvx5OWluiwRERHJAMp0GGPO8DBdK5+g87Ff4SkooO6rt1JwxrxUlyUiIiIZRA3aGAru2c2eO5cz9P5WChcspOqGT+MpGJ+7S0VERGTiUIM2BpxIhH1PP0XHIw/i8vup/cKfULhwUarLEhERkQylBu0UBdvbaL37TgY2WvLPmEf1p2/CW1yS6rJEREQkg6lBO0mO47D/98/R/ov7cbndVN/8eYqWLMXlcqW6NBEREclwatBOQqirk9Yf303/O2+TN+s0qm/+LL6y8lSXJSIiIhOEGrQT4DgO3atX0XbfCpzhYapu+DTFF16kVTMREREZU2rQEhTev593l/8PXS+9QqCpmeqbP4+/qirVZYmIiMgEpAYtAZFQkG3/9HdE+vqouPY6Si+9HJfbneqyREREZIJSg5YAl9dH2bIPUr9kAX0B3aEpIiIiyaUGLQEul4vSD1xKXmUhfUmccC8iqbd/qJsdPbvY0bOL0PYhCl3F1ORXUZNXRUlOsfacisi4UIMmIlnJcRz2De1ne6wZ29Gzkx09u9gfjP4jzIWLXF8OA6HB+HNyPTlU51VFG7ZY01aTX0VFoBy3S9seRGTsqEETkQnPcRy6BveOaMZ2sb1nJ72hPiDajNXkVzGzrJmGwkk0FE6ivqCW+poK3tu9h5a+tuiv/jZa+lp5t2sTL7W8Gv/9vS4PVXmVVOdXURtv4KqpClTg8/hS9W2LSAZTgyYiE4rjOLQPdMYbsQO/+sL9ALhdbmrzqzm9YhaTC+vjzZjf4z/s93K5XBT5CynyF9JcOmPU1wbCA7T0tdPS30ZrXxt7+lrZ0bOLN9r+gIMTfT4uygNl1ORVUZtfTfWIVbeANzf5fxgikrHUoIlIxoo4Edr6O0atiu3s3c1AOHpZ0uvyUFdQw7yq02korGdy4STq8mvGZFUr4A0wrXgy04onjzoeGg7RNtBBS1/riFW3Nt7t2kjYGY4/rthfNOJSaXX840Jfgfa5iYgaNBHJDMORYVr72w9rxoaGgwB43V4mFdRydvWZNBTWMbmwntr8arzu8f0x5/P4mFRQy6SC2sPq7xzsGtW0tfS1sXbPuvj3AJDnDYzY31Yd/7g0t0T73ESyiBo0EUk7w5Hh+CXDA/vGdvbuJhQJAeB3+6gvrOOc2gU0FE5icuEkavKq8Lg9Ka786Dzu6D61qrxK5nJa/PiBmxXiTVvskulbHetZveeV+OP8bh/VsX1uNXnV1MZW3CoDFWn9fYvIyVGDJiIpFYqE2dPbEl8V29Gzm119ewhHwkD0zsn6wjrOnbQovmesOq9ywqwmuVwuSnNLKM0tYVZZ86iv9Yb6aOmLNmwHGrgt+95nXesb8ce4XW4qAxWj7iqtya+iOq+KnCPsqxORzKAGTUTGTXA4xK7ePSM27+9kd18rw7G9WQFvgIbCSVxQvyTejFVmcYRFgS+fxpJpNJZMG3V8MDxEW3/7qFW3lr5W/tCxnogTiT+uLLf0YNM24pJpvi9vvL8VETlBatBEJCmGhoPs6t3N9u6De8Za+tviDUS+L4+Ggkl8YPL58cuU5bll2iCfgFxvDpOL6plcVD/qeDgSpn2gc0QsSCutfW1s2vde/PIwQKGvILrKFmvcamONW7G/SH/+ImlCDZqInLKB8CA7e3azo3dXtCHr3UVrX1s8bqLQV0BD0STmVp4Wb8ZKc0rUDIwxr9tLbX41tfnVo45HnAhdg/uid5b2H7xk+mrrmwyEB+KPy/XkUp1fGV91q82vpjqviopAWdauYoqkiho0ETkh/aGB6OXJ3l1s797Jjt5dtPV3xL9eklNMQ2EdZ1XNZXIs9FUrM6nldrmpCJRRESjjdGbFjzuOQ3ewl9b+QyNBDgnidXupOmyfWzVVeZX4xvkuWZFsoXeWiBxVb7BvVKzFjp5ddAx2xb9emlPC5KJ6FlbPZ3JRtBkr8hemsGI5ES6Xi+KcQopzCmkubRz1tWgQb9uoxm17zy5ePySItyJQFs9yq86vojZ2g4KCeEVOjRo0EQGgO9gTbcS6D66O7R3aF/96RW4ZDUX1LK1bREPRJBoKJlHgz09hxZJM0SDeKUwrnjLqeHA4dNgNCq19bazv3Bi/2QOiK6k1eVWjpifU5ldT4MvXaqpIAtSgiWQZx3HYH+yONWM74/vG9ge744+pyqtgevEUJhctpaFgEg2FdeTpzj8B/J5oBl19Yd2o48ORYTpiQbwjY0HW7nllVBBvvjdvVNN2YPWtNLdY+9xERlCDJjKBOY7D3qF98bDXA5cpe4K9QPQSVXV+Fc2ljdFLlAWTqC+s0+UpOWEetycapJtXCZVHCOIdEQfS0t/GWx3vsHrPy/HH+d2+0Y1bnoJ4JbupQROZIBzHoXOw62AzFlsd6wsdHBJek1fFaWUzo3dSFk1iUkGdwkwlqUYF8ZYfEsQb7BvVtLX0tbF531ZeaX09/hi3yz3qBoXq/IMN3JEG3ItMFGrQRDJQxInQPtA5Kn1/R8+ueGSCx+WhLr+aMypOpyF2J+Wkglr8YzAkXGSsFPjzafQfOYi39ZA9bnv6WnlrRBCvCxdluSWHrLopiFcmDjVoImku4kTiQ8IPXKLc2bObweEhIDYkPL+W+VVz4+n7tQU1ij+QjJXrzWFKUQNTihpGHQ9HwrT1d4zKcmvpa2PT3i2EYqPBAAr9BQcnJ4zY66a4F8kk+gkukkaGI8O09LfFL1Pu6NnJzp7dBGMp8D63j/qCOhbWzI9njNXmV2uPjmQFr9tLXUENdQU1o45Hg3j3jooEaelrY13rG4cF8R5pZqmCeCUdqUETSZFwJMyevtb4JcrtPTvZ3bsnvhKQ4/FTXzApGmsRa8aq8yrVjIkcIhrEW05FoPyIQbwj97i19LexocuytmVd/HEHgnhr86vjl0znBQwedLOMpI4aNJFxEBoOsbuvJbYyFr1Mubu3hXAsNyrXk0tDYR3nT1oSH4VUmVehf9WLnIKRQbymbHQQb39oYETTFp1Zuq17B6+1vRUN4n0HKgLlzC4zzC5vpqlkBrnenBR9J5KN1KCJjLHgcJCdvXtG7Rnb09ca39yc5w0wubCeixrOi6+M6RKLyPjK8wWYXjyF6UcI4m3tb6dteA+vbH+LtS3reH7XajwuDzNKpjG7rJnZ5Ya6/BrtZ5OkUoMmcgoGw0Ps7N09ahxSy4gh4QW+fCYX1nN6+az4nrGy3FL9YBdJU36PLzpLttIwv2Q+oUiY9/a9z4aujazvsvxyy5P8csuTFPsLmRVbXZtZ1qw7R2XMqUETSdBAeCC+V+xAQ9bW3xFvxor9hTQU1jOvck78MmVJTrGaMZEM5nN7MWWNmLJGPsKV7Bvaz4bOjWzo2shbHe+wtmUdLlxMKWqIr65NKWrQiricMjVoIkfgOA6t/e1s3LuZHZt2sKnjfdoHOuNfL80poaFwEguqz4xfpizOKUphxSIyHkpyillct4DFdQuIOBG2de9gfddGNnRaVr7/DE++/zR53gAzy5qYXWaYVd5MSU5xqsuWDKQGTSSma3Avtmszdu8WNu7dHJ9NWZ5XSkNBPefULohfpiz0F6S4WhFJNbfLHR8o/8Fpl9IX6ufdrk2s77Js6LS81vYWAHX5NcwuN8wuM0wvmaqMQkmI/i+RrNUT7GXj3s3YvdGmrCO2QlboK6C5dAamNHpZY2bDFDo6elNcrYiku3xfHvOrz2B+9Rk4jsPuvhbWd1rWd23kuR0v8vT23+N3+2gubWR2uWFWWTNVeRWpLlvSlBo0yRoD4QE27X2PjXu3YPduZndfCwABby5NJTO4sH4pprSR2vzqUfvGtIdMRE6Uy+ViUkEtkwpquXTKhQyGh9i0bwvrO6M3G7zduQFQlIccnRo0mbCCw0He278ttkK2me3dO3Fw8Ll9zCieyoLqMzFljdQX1Cn8VUSSKtebw5yK2cypmA1AW39H9M7QTqsoDzkiNWgyYQxHhnm/e0f8suXW/dsIO8O4XW6mFk3miqkfwJTOYGrxFO0BEZGUqsqroCqvggvql8SjPNZ3WTZ0bRwR5VHErPJmZpcpyiMb6W8pyVgRJ8Ku3j3xFbLN+7YSHA7iwkV9YR0XNCzFlDYxo3iqLhuISNoaGeXxUT4Yj/JY32V5q/0d1u5RlEc2UoMmGWNk9IXdu4VNe7fQF+4HoCavinNqzsaUzqCpdIb+pSkiGeuIUR6d0dU1RXlkDzVoktaOFn1RmlPCnMrZmNJGmktn6IeTiExIo6I8pl9Gb6gP27Upnr2mKI+JS2dQ0srB6IvonZZHi74ozy3T5lkRyToFvnzmV89jfvW8o0d5ePw0l8yIN2yVeeWpLltOgho0SamjRV/kenJpKp1+1OgLEZFspyiPiU0NmoyrRKIvmstm0FAwSdEXIiIn4JhRHnteUZRHhlGDJkml6AsRkdQ4WpTH+k6rKI8MoL8RZUwlFn3RyIziaVpmFxEZJ6OiPBqPHuUxtaiBWYrySAtq0OSUHCv6ojqvinNq5mNKG2ksnU6BLz/F1YqICBw9ymO9ojzShho0OWGKvhARmTiOGuXRuZENXYrySBX96cpxHS36osCXH429KG2kubSRioCiL0REMt1Rozw6raI8xpEaNDnMQHiAzfu2xlbJFH0hIpKtjh7lEW3YFOWRPGrQ5BjRF15mFE9T9IWIiACK8hhPatCy0PGjLy7GlDYq+kJERI7p0CiPLfu2xhs2RXmcGv3tmwWOGX1RUKvoCxEROWU+t5eZZU3MLGtSlMcYUIM2ASn6QkREUm1klMdwZJhtPTvZcIQoj1llzdFfivIYJekNmjGmCFgNfMha+74x5hLgO0AAuN9a++3Y4+YBdwBFwPPAF6214WTXN1F0De6N3mXZtVnRFyIiklY8bg/Ti6cw/QhRHuu7LK+2vQnApILa6OqaojyS26AZYxYBy4Hm2OcB4C7gAmAH8IQxZpm1diWwAvi8tXatMeZO4BbgB8msL5Mp+kJERDKVojyOL9mt6S3Al4GfxD5fCGyy1m4FMMasAK41xqwHAtbatbHH3QP8A2rQ4hKJvmgunaG7ZUREJKMoyuPIktqgWWs/D2CMOXCoDtgz4iF7gPpjHE9YeXnBSdd5IiorC8fldYLhIO92bOHtNsvbrZYte7fhOA4+j4+ZFTO4YPoi5lTPZFppg6IvxsF4nXdJPzr32UvnPlUKaait4GIWAdDS08YbLet5Y887vNSyLhrl4fYwq6KRebWzmVdzGg3FdWO6OJEO5368L+66AWfE5y4gcozjCevs7CUScY7/wFNQWVlIe3tPUn7v6AbKHfEVssOiL6ZcTHNpI9NGRl9EoKuzPyn1yEHJPO+S3nTus5fOffrwEGB+yXzml8wnZKJRHuu7LBs6N7LizUdY8eYjI6I8DDPLmk4pyiPZ597tdiW0qDTeDdpOoHbE5zXA7mMcn7COG31RvxRT1siM4qnkenNTXa6IiEjKjYzyoBH2De2Pzwx989Aoj3LD7LLmjI3yGO8G7SXAGGMaga3AJ4G7rLXbjDGDxpil1tpVwI3AynGuLakcx6Gtvz3WkB05+qK5tJEmRV+IiIgkpCSnmCV1C1hypCiPrU/z5NanMjbKY1wbNGvtoDHmJuAhIBd4Engw9uUbgOWxWI7XgO+OZ23JcMzoi4rZmDJFX4iIiIyFI0V5vNu1KR6Wm2lRHi7HSe6+rXEwFdiaDnvQjhd90Vw6A1PapOiLDKO9KNlL5z576dxPLI7jsKt3T3wM1Zb97zPsDB8xymMc96BNA94/2uPSs23MEMeOvpim6AsREZE04HK5qC+so76wLhblMcjGvVvY0LWRdw6J8vjCguuZ5J2c4orVoJ2QYDjIu12bsHs3s3HvFrZ178DBwef2MqN4GmdXz8OUNdJQMEnRFyIiImkq15vL3MrTmFt5Go7j0D7QwfqujWzau4Vh54RCJJJGDVoCQpEwd7/zM97pfJdwJByLvmjgiqmx6Iuiyfg8vlSXKSIiIifI5XJRlVdJVV4lF9YvTZvL22rQEuBxucn15HBF04U05DTQWDJN0RciIiKSNGrQEuB2ufn07OvSpqsWERGRiS3zkttEREREJjg1aCIiIiJpRg2aiIiISJpRgyYiIiKSZtSgiYiIiKQZNWgiIiIiaUYNmoiIiEiaUYMmIiIikmbUoImIiIikGTVoIiIiImlGDZqIiIhImlGDJiIiIpJm1KCJiIiIpBk1aCIiIiJpxpvqAsaAB8Dtdo3Li43X60h60XnPXjr32UvnPnsl89yP+L09x3qcy3GcpBUxTs4FXkh1ESIiIiIn4DzgxaN9cSI0aDnAAmAPMJziWkRERESOxQPUAq8AQ0d70ERo0EREREQmFN0kICIiIpJm1KCJiIiIpBk1aCIiIiJpRg2aiIiISJpRgyYiIiKSZtSgiYiIiKQZNWgiIiIiaUYNmoiIiEiamQizOMecMWYhcI219s9P4ff4J2DYWvv3sc9LgJ8C04F24OPW2pYxKFfGyKmcd2PMZ4D/C7TGDj1hrf1rY8xkYAVQBVjgBmtt71jVLCdvPN/nxhg/cCdwNjAAfNJa++4pfgtyksbzva6f/amXDu91Y4wL+DfgQ0AEuMVau+pYr6kVtCObDVSfzBONMcXGmDuBbxzypX8GXrDWzgKWA/91aiVKEpz0eSf6Zvy6tXZe7Ndfx45/H/i+tXYmsA74mzGoU8bGeL7Pvwb0xY7fCtxzUhXLWBnP97p+9qdeOrzXrwFmxWr5CHCPMeaYi2QTegXNGHMh8NdAEJgGPAr0Ev3DcQFXAh8Avg04ROdifRP4R6DAGPPXwGPAj4j+WQ0CN1trNx3jZa8GNgG3H3L8g8D5sY/vA/7HGOOz1oZO7buUQ6XovC8AmowxfwW8CXw19prnx14Xom/U3wMn/a84OVwmvM9jx/8WwFr7vDGm0hgz2XrCms8AAAjsSURBVFq7/eS/c8mQ97p+9o+RTH6vx47/3FobATYaY7YDS4Dnj/bC2bCCtgj4ItF/9XwFaLfWng28FTv+H8Bl1trTiA4wXUr0D/dRa+2/AH8G3B57znLgnGO9mLX2Xmvt/+Xwwe11RAe6Y60NA91A5Zh8h3Ik43reiZ7bfwLmAjuA7wEVQHfsfB94TP2YfYcyUrq/z+PHY/T/wthJ9/e6fvaPrUx9r5/wz4AJvYIW87a1dgeAMaYDeCZ2fBtwLbDKWrsTwFp7Y+xxN414/hNEO+MriHbej51kHa4jfB45yd9Ljm9cz7u19qMHPjbG/P/AFuBbRP8VN5LOeXKk+/vczej/F/T+Hzvp/l7Xz/6xlanv9RP+GZANK2jBQz4Pj/g4xIg/sNhS5Kh/2dj/1979B1lZ1XEcf6+Lu4AWZM0AWQZlfRltEiFHMJgaSp0YNeyHjQPJkkxNVFqNYE5CTo1ov3VsSkoNSxgrwrEMNYWcMkkljAKmDw3WZFHkNOKPRJLY/jjnsg+XvffuXi7wtPt5zeyw997nOec5z7PP7pdzznO+0kpgIvAIKfK+scnj+BswOtczBHgJ8K8my7LGDtt1z3MUPll4qy3X909gRES05/fHANv72Q7rm7Lf538lXf+K0fhnoVXKfq/7d39r/b/e6/3+HTAYArR6HgUmR8To/PprpPHmPeTexYj4PnCapKWkSZ8Tm6xrNXBR/v79pMmFnoNwZLT6uj8HLIyI0/PrjwF35Ov7S9L1hnT9725lQ6xPynCf73s/IqYCL3j+2WFRhnvdv/sPnzLf66uBWRHRHhEnAm/Ix1vTYBjirGc7cClwb/6fzzrgO8DrgKsi4lpgCXBTRCwmRe4fabKuRaSnNjYDO4FZB3vw1rSWXndJ/42IC4BvRsQwYCs9N+584NaIuBL4C3DhIWqT1VaG+/wGYGl+fzfwgWYbY/1Shnvdv/sPnzLf6ytJ8+d+l19fLGlXvQraururh83NzMzM7Ega7D1oTYmInwMv6+WjGyU1O55tJefrPrj4eg9evvaDS1mvt3vQzMzMzEpmsD8kYGZmZlY6DtDMzMzMSsZz0MwGkIgYC/wJmCfp5sL7lwFvlNTVonr+DLxX0vpWlNegrpeSliwYCSyStKoP+9xESqtyfwvq7yK19ZwG232bNGflNy2uv9/tr1FON2kxz3GSimtFXQV8lrT0wPr83hzSquzDgA7gQWChpJ0R8RAwPL8fwO9zUZslzYqIycA1wMtJnQBPAJdJ2tzMcZsNVg7QzAaevcBXIuJBSTrSB9MCE4BRkk7s6w6S5h3C46nlTGDpIai/3+2vow2YRs7/FxFtwAXAU5UNco7JdwIzJe3I+QWvI624Pk3SGXm7saRV3ScU9u0E7iKl2tmQ35sN3B0R4yRVp8sxsxocoJkNPLtIiX1XRMQUSfutvB0Ry0h/WL9c/Tr3jK0AppOeavoiKZfdJNIq3edJqqx+/dGIOAXoJOW2uyWXdy4pWXEH8Dyp92Rd7qmZQspJt1HS7KrjmknqyTkKeBb4FPA0cAtwfET8FphSXDsoIt6d69pLypW3ICcofoCUI3E9sBa4L7dhCCkv34eB8fnzC4ET8jk4Npc7tvi6UN/kfE46SauC3yfp4oi4OrdreURcBHwB+Lqklb21S9Ij+XyMzeW8hrQy+WxJfy/UF9XtB86uU17N85vdBsymJ0HzVGALqUeMiDgGuAKYKGkHgKQXI2IBcH5EdFT/PFUZTurpK5635aQ8he0cmM/QzGrwHDSzgelq0qrnS5rYd6ikyaRA5lvA9ZJOIQ1VdRW22yVpIqnn6JqIODkiXp/rnCHpVOBDwKr8hx9SIHJqL8HZeFLKlffkuhYDd5ISCs8Dtkma0MvCjl8C5ufEx4uAt/XSnnHAT/M264DrSUHZyaTepEbJsYsuBRZLOh04CTgvIiZJ+gxpkcxZkh5u1K48bEmu/32SxgP/Jg0r7pN7QPe1n3T+6pXX6/ktWEEKtDrz6znAssLn40nX9Y9Vx/G8pOUNgjMkPUXKS3lPRDweEd8D5gL3N9rXzPbnAM1sAJK0l9RTMjcizuzn7j/K/24D/iFpY+H1cYXtKsN524GfAW8nBWtjgDW5x2c5qXerMjz3a0nF3HkV04E1kh7PZa4l5Tec1OBYbwfuyHO+Kj1+1V6kJyHyNuAhSc9IeoEUVB3Xyz61zAFG5mHAb5DmaB1bZ/tG7XpA0jP5+8f6cCyNyqt1fit2kHIQnpNXwp8G3FP4vJLUuWmSvgqMAi4hBdiXA49FxIiDKddssHGAZjZASXqCNJR3K/CKwkfdpLlIFR1Vu+4ufF8vZ2BxuOqovG07KYCYUPki9VBtyts9V6OsdgpJjgtlHl2nfnLP1VTSUGUXPUN3Rf8pToqn9zY1OicVvwBmAH8APkcalmyrsS00blexR7D6GJopr9b5LfouKXg/H/hJVUC3BTg694TuExFDI2J1RLyyXsER8ZaIWCDpWUl3SVpI6qnsJgXvZtZHDtDMBjBJK0lPAH6i8PaTwJsB8h/ctzZZfFcu4wTgHcCa/HVWHtojImaQcs8Na1DWGuDsiHht3m868Grg4Vo7RMSQPGdueF7tez7wpsLwXX/sBDoi4qT8+oCcqRExEjgNuDw/SfkqUs9ge95kDwcGlP1uVwOtKO9O4AxSou9lxQ8k7SbNn7s5IkblOjpJSaePKcw/rOVJ4MqcJLpiDDCCnqc9zawPHKCZDXyXkJZXqLgBGBMRIiUSXttkuUMjYgOwGvi4pK2StpDmnd0eERuBz5MeLKjbs5P3m0+ar7YJuBY4V9LTdfbZQwo8V+Tj+CHwwRxk9EuuZyHpacNH2b9nq7LNTtLyERvyMX4a+BU9w7ergNsi4qyDaVeD4zzo8vLQ7o+BDkmbevl8CWmY+948TL2R1LP3rj6UvRWYCSzJc9C2AD8A5g6QJ4rNDhunejIzMzMrGfegmZmZmZWMAzQzMzOzknGAZmZmZlYyDtDMzMzMSsYBmpmZmVnJOEAzMzMzKxkHaGZmZmYl4wDNzMzMrGT+B69CUthUc/ooAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "baseline_smart = df_simulations_number[['victory', 'draw', 'defeat']][df_simulations_number['method']=='smart']\n",
    "\n",
    "idx_smart = df_simulations_number.index[df_simulations_number['method'] == 'smart'].tolist()\n",
    "df_simulations_number.drop(idx_smart, inplace=True)\n",
    "\n",
    "df_simulations_number_1 = df_simulations_number[df_simulations_number['play_order']==1]\n",
    "df_simulations_number_1 = df_simulations_number_1.set_index('method')\n",
    "df_simulations_number_2 = df_simulations_number[df_simulations_number['play_order']==2]\n",
    "df_simulations_number_2 = df_simulations_number_2.set_index('method')\n",
    "\n",
    "for field in ['victory', 'draw', 'defeat']:\n",
    "    plt.figure(figsize=(10,7))\n",
    "    plt.plot(df_simulations_number_1.index,\n",
    "             [baseline_smart[field] for _ in range(len(df_simulations_number_1.index))],\n",
    "             label='baseline smart')\n",
    "    plt.plot(df_simulations_number_1.index,\n",
    "             df_simulations_number_1[field],\n",
    "             label=field+' (play first)')\n",
    "    plt.plot(df_simulations_number_2.index,\n",
    "             df_simulations_number_2[field],\n",
    "             label=field+' (play second)')\n",
    "    plt.plot(df_simulations_number_2.index,\n",
    "             (df_simulations_number_1[field] + df_simulations_number_2[field])/2,\n",
    "             label=field+' (mean)')\n",
    "    plt.xlabel('Number of simulation for MCTS')\n",
    "    plt.ylabel('Number of '+field)\n",
    "    plt.title('Performances for different number of simuation for MCTS, for 1 000 games')\n",
    "    plt.legend(loc='best')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importance of the exploration constant"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|          | 5/1000 [00:09<30:23,  1.83s/it]\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-72-2b27e9c4c891>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     12\u001b[0m                               \u001b[0mprint_board\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     13\u001b[0m                               \u001b[0msimulations_number\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msimulations_number\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 14\u001b[0;31m                               c_param=c_param)\n\u001b[0m\u001b[1;32m     15\u001b[0m           \u001b[0mres_list\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-4-4d2c0c4d828a>\u001b[0m in \u001b[0;36mplayTicTacToe\u001b[0;34m(board, method_player, method_computer, first_player, print_board, return_board, simulations_number, c_param)\u001b[0m\n\u001b[1;32m     47\u001b[0m             \u001b[0minitial_board_state\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mTicTacToeGameState\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mboard\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnext_to_move\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnext_to_move\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     48\u001b[0m             \u001b[0mroot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mTwoPlayersGameMonteCarloTreeSearchNode\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minitial_board_state\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 49\u001b[0;31m             \u001b[0mbest_node\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mMonteCarloTreeSearch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mroot\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbest_action\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msimulations_number\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msimulations_number\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mc_param\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mc_param\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     50\u001b[0m             \u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mboard\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbest_node\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mboard\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     51\u001b[0m             \u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnext_to_move\u001b[0m \u001b[0;34m*=\u001b[0m \u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-3-911678b25359>\u001b[0m in \u001b[0;36mbest_action\u001b[0;34m(self, simulations_number, c_param)\u001b[0m\n\u001b[1;32m     23\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0m_\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msimulations_number\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     24\u001b[0m             \u001b[0mv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_tree_policy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 25\u001b[0;31m             \u001b[0mreward\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mv\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrollout\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     26\u001b[0m             \u001b[0mv\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackpropagate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreward\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     27\u001b[0m         \u001b[0;31m# to select best child go for exploitation only\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.7/site-packages/mctspy/tree/nodes.py\u001b[0m in \u001b[0;36mrollout\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    108\u001b[0m         \u001b[0mcurrent_rollout_state\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstate\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    109\u001b[0m         \u001b[0;32mwhile\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mcurrent_rollout_state\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mis_game_over\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 110\u001b[0;31m             \u001b[0mpossible_moves\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcurrent_rollout_state\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_legal_actions\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    111\u001b[0m             \u001b[0maction\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrollout_policy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpossible_moves\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    112\u001b[0m             \u001b[0mcurrent_rollout_state\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcurrent_rollout_state\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmove\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0maction\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.7/site-packages/mctspy/games/examples/tictactoe.py\u001b[0m in \u001b[0;36mget_legal_actions\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     98\u001b[0m         return [\n\u001b[1;32m     99\u001b[0m             \u001b[0mTicTacToeMove\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcoords\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcoords\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnext_to_move\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 100\u001b[0;31m             \u001b[0;32mfor\u001b[0m \u001b[0mcoords\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mindices\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindices\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    101\u001b[0m         ]\n",
      "\u001b[0;32m~/anaconda3/lib/python3.7/site-packages/mctspy/games/examples/tictactoe.py\u001b[0m in \u001b[0;36m<listcomp>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m     98\u001b[0m         return [\n\u001b[1;32m     99\u001b[0m             \u001b[0mTicTacToeMove\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcoords\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcoords\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnext_to_move\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 100\u001b[0;31m             \u001b[0;32mfor\u001b[0m \u001b[0mcoords\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mindices\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindices\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    101\u001b[0m         ]\n",
      "\u001b[0;32m~/anaconda3/lib/python3.7/site-packages/mctspy/games/examples/tictactoe.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, x_coordinate, y_coordinate, value)\u001b[0m\n\u001b[1;32m      7\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mx_coordinate\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx_coordinate\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      8\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0my_coordinate\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0my_coordinate\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 9\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalue\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mvalue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     10\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m__repr__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "simulations_number = 500\n",
    "\n",
    "for first_player in [-1, 1]:\n",
    "    for c_param in [0, 0.2, 0.4, 0.6, 0.8, 1]:\n",
    "        method = 'mcts_' + str(c_param)\n",
    "        res_list = []\n",
    "        for _ in tqdm(range(n_games)):\n",
    "          res = playTicTacToe(board=None,\n",
    "                              method_player='mcts',\n",
    "                              method_computer='smart',\n",
    "                              first_player=first_player,\n",
    "                              print_board=False,\n",
    "                              simulations_number=simulations_number,\n",
    "                              c_param=c_param)\n",
    "          res_list.append(res)\n",
    "\n",
    "        print(\"Number of games: \", n_games)\n",
    "        count_res = Counter(res_list)\n",
    "        print(\"A smarter player against a 'smart' computer (computer plays first): \\n\",\n",
    "              getResultProp(count_res))\n",
    "\n",
    "        # Save results into csv\n",
    "        dict_res = getDictResult(counter=count_res,\n",
    "                                 method=method,\n",
    "                                 player=1,\n",
    "                                 first_player=first_player)\n",
    "        df_exploration = df_exploration.append(dict_res , ignore_index=True)\n",
    "        df_exploration.to_csv('df_exploration.csv', sep=';')\n",
    "        if use_drive:\n",
    "            !cp df_exploration.csv \"drive/My Drive/RL/\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "# run to read the csv with results obtained\n",
    "#df_exploration = pd.read_csv(\"df_exploration_full.csv\", sep=',', index_col=0)\n",
    "#df_exploration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmYAAAG6CAYAAACiOIpDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeXxM1//H8dcsWSSxRhBL/Sy11hpbEjtFYou1inzVUm0t328VtcVSsQRVqmqtUqlqa29RW+1rkba0qpYWQSQRIrJnZs7vjyQjk1VIdMrn+Xh4mNzl3DP33Jl5z7ln7tUopRRCCCGEEOIfp/2nKyCEEEIIIZJJMBNCCCGEsBISzIQQQgghrIQEMyGEEEIIKyHBTAghhBDCSkgwE0IIIYSwEhLMrMDNmzepXr06Xbt2Nf/r0qULGzduzHVZISEhdOrUia5du/Lzzz/nQ22ty9SpU2ndujULFizIszLr1avHzZs3OX/+PP/973+BjPs1P7ablYMHD/Lxxx/nSVmbN2/mrbfeypOycpLVPgoODmbkyJFA8rFfr169Z1Kfx1W1alXu3buXZ+X5+fnx22+/ATBp0iSOHz+eZ2U/C4MGDXqq/XHu3DmmTJmS6byjR4/SqlUrevbsSXx8/BOVv3nzZqpWrcqiRYsspiulaNOmDZ06dTJPi4qKYsaMGXTu3JmuXbvi4+PDhg0bADh+/Lj5/dfT05MmTZqY/965cycAK1euNL8/d+rUiTlz5pCYmJhjHf/44w/atm1L9+7duXnz5hM9z1SJiYkMHDiQXbt2ZbnMxo0b8fb2pl27dkydOpWkpCQA4uLiGD16NF5eXrRv3559+/aZ1/n111/p0aMHXl5eDBgwgLCwsKeqp3gKSvzjgoODVd26dS2m3blzRzVo0ED98ccfuSpry5YtasCAAXlYO+tWtWpVFRISkqdl1q1bVwUHB1tMS79f82O7WVm0aJH64IMP8qSsTZs2qaFDh+ZJWTnJah+dPHlSdezYUSmV+bH/T6tSpYqKiIjIs/JatWqlzp07l2flPWtPuz+yO+bGjx+vPv300ycuO7X8li1bqjZt2lhM/+mnn5SHh4f5WIuPj1edO3dWK1asUElJSUoppW7evKnatm2rvv32W4t1M3vN7dy5U/Xu3VvFxcWZyxs6dKiaP39+jnX85JNP1MSJE5/4OaYKCgpSPj4+qnbt2uqHH37IdJk///xTNW/eXEVERCij0ahGjRqlVqxYoZRSas6cOcrPz08ppdStW7dU06ZNVUhIiEpISFDNmzdXZ86cUUoptW7dOjVkyJCnrq94Mvp/OhiKzJUsWZLy5ctz7do1qlWrxoYNG1i/fj0mk4kiRYowefJkKlWqxPjx44mMjCQ4OBgHBwfCw8N5+PAhvr6+BAYG8s033xAYGIhWq6V48eJMnjyZChUqWKzXsmVLIiIisLe359KlS0RERNC6dWuKFCnCgQMHCA8PZ8aMGbi7u/P3338zffp0YmJiCA8Pp1q1aixcuBA7Oztq1arF0KFDOXbsGGFhYQwZMoS+ffsCsHz5crZs2YJer6d8+fIEBARQsGDBLJ/XmTNnCAgIwGQyAfDWW2/Rvn17i33Ut29flFK8+eabTJ06lcKFCzN9+nQiIyPRaDQMGjQIHx8fTp06xcyZM3FwcCAmJoZNmzZha2trLufMmTP4+/uj0WioVauWeZunTp3C398fPz8/Fi5caN6vRqPRYrvlypVj+vTphISEkJSURMeOHXn77be5efMm/fr1o1KlSty6dYvAwEBu3rzJhx9+SFxcHFqtlhEjRtCqVSs2b97M3r170Wq1XL9+HXt7e+bMmUN0dDRff/01RqORggULMmrUKHO9v/nmGw4cOMCyZcsAuHr1Km+88QYHDx5ky5YtfPPNNyQlJfHgwQPefPNNc1uk8vX1pV+/fnTo0CHD31evXmXmzJlERkZiNBrx9fWlZ8+eGY7Ty5cvZ7rP07dNgwYNADAajfj5+REaGsrgwYP54IMPMBqNTJkyhfPnz/Pw4UPGjh1rbuulS5eyZ88eTCYTZcqUYerUqZQsWTJDPTI7jipUqMDAgQOpWbMm77//PsePH2f8+PFs3ryZDz/8EDs7Oy5evEhERASenp74+flhY2NjUe6nn37Kjh070Ol0VKhQgcmTJ+Pi4oKvry+FCxfmr7/+4vXXX6dWrVrMmzePxMREwsPD8fDwYNasWSxYsICwsDDGjBnD3Llz+fDDD837eN++fSxevBiTyYSjoyMTJkygdu3afPLJJ9y6dYvw8HBu3bpFyZIlmTdvHiVKlLCom8FgYN68eRw8eBCdTke9evWYOnUqGo2GgIAATpw4gU6no3bt2kyYMAEnJydat25Nt27dOHHiBCEhIXTt2pV3332XmJgYJkyYwPXr19FqtdSsWZPp06czadIkAAYMGMCKFSu4ePEiy5cvJzExkXv37uHj48O7777LqVOnWLBgAeXKlePy5csYDAY++OADSpcuzaJFi3j48CETJkxg9uzZ5vp/9tln/Pjjj9jZ2fHw4UPee++9bOtdu3Zt/vzzT9577z1effVVi31RpUoVQkJCCAoKon79+gBs2bKFLl26cOTIEQB27tyJg4MDb775pnm9MmXKsHDhQnOPUnbCw8MxGo3Ex8djb2+PnZ0dkydPzrE38bvvvmP9+vXmdefPn//Yx5Wvr69FWYGBgYwePZrly5dnub0ff/yR1q1bU6xYMQBee+01ZsyYwZtvvsm+ffv48MMPAShdujSenp788MMP1K5dGycnJ9zc3ADo2bMns2bN4v79+xQtWtSi/M2bN7NixQrs7e1p0qQJa9eu5cKFC9y9e5cpU6YQERFBeHi4ed86OzvTunVrOnXqxMmTJ3nw4AFDhgwhKCiI33//Hb1ez9KlSylZsiShoaGZvpcaDAb8/f0JCgrCxsaGsmXLMnv2bBwdHXNst3+lfzgYCpV5r0FQUJBq2LChun37tjp16pTq27evio2NVUopdeTIEdWhQwellFLjxo2z6MlJ++30+PHjqm3btuZvu5s2bVJeXl7KZDJlWG/cuHGqV69eKjExUYWFhakqVaqotWvXKqWUWrNmjRo4cKBSSqmAgAC1detWpZRSiYmJqlOnTmrXrl1KqeRv1oGBgUoppc6fP69eeeUVFR8fr/bt26fatWunIiMjlVJKzZo1Sy1ZsiTb5/Wf//xHbd++XSml1B9//KGmTZuW6b5L/TaflJSk2rRpo3bv3q2USu5xbNasmQoKClInT55U1apVUzdv3sywfkJCgvLw8FDHjx9XSin1/fffqypVqqjg4GCLnp303/rT9iL4+vqqH3/8USmV/C3a19dX7dixQwUHB6sqVaqo06dPK6WUioyMVO3atTP3xt25c0c1b95c3bp1S23atEm5ubmZe5imT5+u3n//faVU1j1mDx8+VA0aNFBhYWFKKaXmzp2rPvroIxUdHa169+6t7t27p5RS6ueffzYfX2mfR//+/S2+daf+nZSUpLy9vdVvv/2mlFIqKipKeXl5qZ9//tli+9nt8/T7KK30PWZVqlQxH0N79uwx93xs2bJFvfvuu+beja+//jrTb/HZHUehoaHKw8ND7d27VzVr1kz99NNPSqnk493Hx0dFR0erhIQE1a9fP/Oxm1rvjRs3qtdee03FxMSY22HQoEHmfTVhwgRzHUaNGqVOnjyplFIqOjpaNW7cWJ0/f14pZdljlrqPr1y5ojw8PNSNGzeUUsmvVU9PT/Xw4UO1aNEi1aZNG/Xw4UOllFJvvfWW+vjjjzM87y+++EL169dPxcXFKaPRqP73v/+pLVu2qI8//liNGDFCJSYmKqPRqMaPH68mT55srktAQIC5vWrVqqVu3LihtmzZYn5uBoNBTZo0SV27ds1if5hMJtW/f3/1999/m9evXr26ioiIUCdPnlTVq1dXFy5cUEoptWrVKtWvXz+lVPY9ZuPGjVOfffaZUkrlWO/FixdnWkZq+atWrVJTpkxRSikVGxur2rVrp44dO2Y+1qZPn67mzJmTaRnpZfaai4qKUgMHDlQ1a9ZUvXv3VrNnzzYfT7kpLzfHVVbSv3bTmjx5slq+fLn572vXrqmGDRsqpZR65ZVXzO8XSin10UcfqVmzZqnt27eb65CqWbNmGc7YXL58Wbm7u5vfpz755BNVpUoVpVTy50Tqdk0mkxoyZIhatWqVUiq5/WbNmqWUUmrHjh2qWrVq5rKHDRumli5dqpTK+r309OnTqkOHDspkMimlkt/rzp49m+N++reSHjMrER8fT9euXYHkXoWiRYsyb948XF1dCQwM5Pr16/Tp08e8fFRUFJGRkQDmbznpHTlyBG9vb/M3p+7duzNz5kzzGIf067Vq1QobGxtcXFxwcHCgWbNmALz00kvmbY0dO5Zjx46xcuVKrl27RlhYGLGxseYy2rRpA0DNmjVJTEwkNjaWEydO0KFDBwoXLgzAhAkTAJg7d26Wz8vLy4vp06ezf/9+PDw8eO+997Ldf9euXSMhIYF27doByT2O7dq148iRIzRu3BhXV1fKlCmTYb1Lly6h1+txd3cHoFOnTlmOh8lMbGwsp0+f5sGDB+ZxYLGxsVy8eJHatWuj1+upW7cuAL/88gvh4eEMHz7cvL5Go+HPP/8077NSpUoBUKNGDfbu3Zvttp2cnHj11Vf57rvveOONN/j+++9Zt24djo6OLFu2jEOHDnHt2jUuXrxo0UY5uXbtGjdu3GDixInmafHx8Vy4cMH8XFKXy2qf52bcmI2NjbmHrFq1akRERABw4MABzp8/T48ePQAwmUzExcVlWP/gwYNZHkclSpTA39+fYcOGMXLkSBo2bGheplu3buZv3F27duXHH3+kf//+5vmHDx+me/fuODg4APCf//yHZcuWmccUpfYCAgQEBHD48GGWLVvGX3/9RUJCQrb7/OTJkzRp0oRy5coB4O7uTrFixcxj0Ro1aoSTkxOQfCw8ePAgQxmpY6Ls7e0BWLhwIZDc2zFq1Chz75+vr6/FMZf6Gi1ZsiTOzs48ePAANzc3FixYgK+vLx4eHgwYMIDy5ctbbE+j0bBs2TIOHjzI9u3buXr1Kkopc5uULl2a6tWrm+u8ZcuWLJ9/Zg4fPpxtvdPu78ykjhubNGkSe/fupXXr1uh0Oov6q6e4A2HBggX5/PPPCQ4O5uTJk/z0008MHTqUvn37Mnbs2McuJzfH1ZNI/xyVUmi1WvNjjUZjMV+r1WIymTJMV0pZ7D9IHhPo6elpfp/q378/n3zyCZDcq3rmzBlWr17NtWvXuHz5MnXq1DGvm/o+Ua5cOYoXL061atWA5M+XBw8eZPte2rRpU3Q6Hb169aJp06a0b9+e2rVrP9V+smYSzKyEvb0927Zty3SeyWSia9eu5he/yWQiLCzMHHRSX+CZrZeeUgqDwZDpemlP7wHo9RkPj/feew+j0YiXlxctW7YkJCTE4o3Azs4OwPwiT31xp33RR0VFERUVle3z6tOnD61ateLYsWMcOXKExYsXs2vXLnP56RmNxkzfWLJ6rumXy+l5Z8VkMqGU4uuvv6ZAgQIA3Lt3Dzs7O+7fv4+tra25PKPRSKVKlcyDjQFCQ0MpVqwY33//vfkDFh7/Q6R3797m07+VKlWiXLly3Llzh9dee43evXvj5uZGhw4dOHDgQI7PPfV0Tupp07TH4927dylYsKDFujnt88eV9vRh2vJMJpPF6fDExMRMA0pOr48rV65QvHhxzp07Z7Fe2g+dtB9eactNX5+0zy3tMdW/f3+qVq1Ks2bN8PLy4tdff822/bL6IEwt/3GOhfTH6d27dzGZTJnWO+2purSvodSyy5Urx969ezl16hQnT55k4MCBTJ8+ndatW5uXjY2NpVu3brRt25YGDRrQo0cP9u3bZ67bkxy/2e2T9PXO7jUM4OLiQo0aNTh8+DBbt25l/Pjx3L9/3zy/bt26rFu3LsN6P/74I2fOnGHcuHHZlr9y5Urc3NyoX78+5cqVo1evXpw5c4Y333wzV8EsN8fVk3B1dbUYuB8WFmYOUqnzihcvbp5XrVq1DOskJSURGRmZYdiATqezaNe0r6F58+Zx7tw5evToQePGjTEYDBbLpv18ST9kALJ/L3V0dGTbtm0EBQVx8uRJ3n33XQYPHky/fv2eaB9ZO/lV5r9A06ZN2bFjh/mFs379egYMGJDjes2aNWPnzp3mMRCbNm2iSJEiGb4J58bRo0cZPnw43t7eQPIveYxGY7breHh4sHfvXqKjowH45JNPWLNmTbbPq0+fPvzxxx90794df39/oqKiCA8Pz3IbFStWRK/Xs2fPHiA58OzevRsPD49s61a1alWUUhw6dAhIfpPO7MM/K05OTtStW5fVq1cDyaHz9ddf58cff8ywbN26dbl+/TqnT58Gkn+p1b59e0JDQ7Pdhk6nyzLspPZgffrpp/Tq1QuA3377jWLFijFs2DCaNm1qDmXp2yltD82VK1fMPXcVKlSw+KKQ+ovU1GVTPek+1+l0jzWmp2nTpmzcuNF83Hz88ce8//77mS6X1XF07tw51q5dy6ZNm3j48CFffPGFeb0ffviBxMREEhIS2LJlC61atbIot1mzZmzatMnc8xUYGEjDhg0zfIGJiori/PnzjBkzhnbt2nHnzh1u3Lhh/mKUWfu5u7tz9OhRgoODAcxjvtL2MOTE3d2d7du3k5iYiMlkYtq0aezYsYNmzZqxfv16kpKSMJlMrFu3Dk9Pz2zL+uqrr5gwYQJNmzZl7NixNG3alAsXLljU//r160RHR/Puu+/SunVrTp06Zd52drI7ftN6knqn5+Pjw+rVq3n48CFVqlSxmNeuXTuio6NZuXKl+bUQHBxMQEAAlSpVyrHs1PFhqWcPILnHvUaNGrmq4+MeV0+qdevW7N+/n4iICJRSfPPNN7Rt2xZI7i395ptvALhz5w5HjhyhVatW1KlTh8jISIKCgoDkz4q6detSqFAhi7KbNm3KiRMnzO9Zab9kHj16lAEDBuDj44OzszPHjx/P8bMhrezeSw8cOMAbb7xBvXr1GDlyJD4+Phnej54n0mP2L9C0aVPefPNNBg0ahEajwcnJicWLF2f4xp2ep6cnb7zxBgMGDMBkMlGsWDGWL1+eoWcgN0aNGsXw4cNxcHDAycmJhg0bcuPGjWzXadGiBVeuXOH1118HoHLlyvj7++Pk5JTl8xozZgyzZs1i4cKFaDQaRowYQdmyZbPcho2NDUuWLGHGjBl88sknGI1Ghg8fTpMmTTh16lS263366adMmzaNjz76iOrVq+Ps7JyrffLhhx/i7+9P586dSUxMpFOnTnTp0iXDz+KLFSvGokWLmDt3LgkJCSilmDt3LmXLluWnn37KsvwmTZowZswY/P39mTx5cob5vXr1YsmSJeY3X09PTzZu3EiHDh3QaDQ0atSIYsWKcf36dYv13nnnHcaPH8+hQ4eoWLGi+RSKra0tS5YsYebMmXz22WcYDAb+97//ZTj1nd0+z07lypWxs7OjZ8+e2V5upFevXoSGhtK7d280Gg2urq4EBARkWC6r10dMTAzvvfcefn5+lCxZkoCAAHr16mU+nWlvb0/fvn2Jioqiffv25lOmqXr27ElISAi9evXCZDJRvnx588DptAoVKsTQoUPp1q0bDg4OlCxZkvr163P9+nXc3d159dVXGTt2LNOmTbPYB1OnTmXEiBEYjUbs7e1ZtmxZhl7J7PTp04dbt27RvXt3lFI0atQIX19fDAYDc+bMwcfHB4PBQO3atTM9btLy8fHhp59+wtvbmwIFCuDq6moeeN6hQwd8fX35+OOPadmyJV5eXtja2lKlShUqV67M9evXsw0VdevW5dNPP2XEiBEsXrw4y+XeeeedXNc7vbZt2zJ16lSLH8mksrW1ZfXq1cybN4/OnTuj0+nQ6XS88847dO/ePceyhw0bhkajoU+fPmg0GkwmE6+88or5FPL58+fx8/PL8sxHqsc9rp5UtWrVGD58OAMGDCApKYk6deqYf/AwcuRIpk2bRseOHTEajYwdO5aXXnoJgMWLFzN9+nTi4uIoUqQIc+bMyVB2hQoVmDBhAoMHD8bW1pbq1aube7eGDx/O3Llz+fjjj7GxsaF+/fo5fjakl9V7qdFo5PDhw3Tq1AkHBwcKFy6Mv7//U+4p66VRT3PSXQgh/oXGjx/Pyy+/zODBg//pqojnyMiRI81jrp5HwcHBbNu2jWHDhqHVatmzZw8rV6606DkTT096zIQQQoinFBoamqHX9XlTqlQpwsLCzD2OBQsWZNasWf90tZ470mMmhBBCCGElZPC/EEIIIYSVkGAmhBBCCGElJJgJIYQQQlgJCWZCCCGEEFbiuflV5v37MZhM+fs7BmdnJyIiovN1GyJ3pE2sk7SL9ZE2sT7SJtYpv9tFq9VQtGjWN2B/boKZyaTyPZilbkdYF2kT6yTtYn2kTayPtIl1+ifbRU5lCiGEEEJYCQlmQgghhBBW4rk5lSmEEELkN6PRwP374RgMiU9dVliYNsebwItnLy/bRa+3pWhRF3S6x49bEsyEEEKIx3T/fjj29g44OpZCo9E8VVl6vRaDQYKZtcmrdlFKERMTxf374RQv7vrY68mpTCGEEOIxGQyJODoWeupQJp5/Go0GR8dCue5dlWAmhBBC5IKEMvG4nuRYkWAmhBBCCGElJJgJIYQQz5GAAH8uXryQ5fzvvtvC3r278ny7RqORiRPHEh8fn+Uyq1YtZ9Wq5U+9nffeG8Hrr3fnq68CCQjwf+x1jx07wtdffwnAoUP72bTpm6eqS36Qwf9CCCHEc2T8+MnZzj9//lfq1XPL8+1u3bqRRo2aYG9vn+dlpxUeHs7Vq1fYti334TJtYG3RojX//e/btG79KkWLFsvLKj4VCWZCCCHEv9DEiWNp164DLVu2AWDQoP6MH+/HokUfMWjQUOrVc2Pp0k84fPgger2OLl26U6FCRY4ePczZs6dxdi7Oyy9XISDAn9DQO+h0OoYOHU6TJh6sWrWc33//jbCwO/j49OCrrwLZuPF7tFotQUFnWLduLfPnLzLXRSnFxo3fsHLlWgBmzpyGnZ0df/xxgZiYGN54YzAdOnS0qP+mTd+wa9dO4uPjsLGxYdq0mYSFhbFq1TKWLv0cgJ07v+fChd8YM2aCeb1x497lwYNIBg/2Zfjw//H55ytYvHgFI0YMpVChwvz991WmTPFn48av+euvqwB069aLWrXqsG3bZgBKlXKlY8cutGjRis2bNzB48Fv511C5JMFMCCGEeELHzodw9FzIE62r0YDK5s4/TWu74lkr68sstG/vzd69P9CyZRuCg2+QmJhIlSrVzPMPHPiR8+d/Ze3arzEYDAwbNoT58xfRtGlz6tVzo3FjdyZPHk/9+g3o06c/t27dZNiwIaxevQ6AxMQEvvxyAwAHD+7n55/P4ubWkF27duDt3cmiLleuXMLJyQknJyfztFu3brJ8+Wru3Ytg8GBfGjZsbJ4XExPN4cOHWLx4OXZ29nz22TI2bfqWd98dy5w5M7l16yZlypRl164dvPXWCIttBQR8xMiRb7FqVSBBQWcs5lWqVJlZs+bx889niYqKYvXqr7h7N5ylSz+hS5dudO3aHYCOHbsAUKdOfWbOnGpVwUzGmAkhhBD/Qh4eTfntt/PExsawb99u2rf3spj/yy9nad36VWxtbXFwcGDNmq9wdi5usUxQ0Gk6dfIBoEyZstSo8QoXLvwGQI0ar5iX69ixC7t37yQ+Pp6zZ0/TtGkLi3KCg4NxcSlpMc3buzN6vZ4SJUpSq1Ydzp37xTzP0dGJadNmsG/fHpYtW8yxY4eJi4tDo9Hg5dWR3bt3cufOHe7du0fNmq/wuFLrXLFiJW7cuM57741g//59DB/+v0yXL1XKlZs3gx+7/GdBesyEEEKIJ+RZK/terew87YVMbWxs8PRsxtGjh9m/fy/z5n2crnw9aa/WEBJymyJFilosk/Fm3Qqj0QiAnZ2deWqrVm1ZsWIJBw7sw93d02IeJPf+6fWWkSLt1e6VMln8HRp6h5Ej36JHj940aeJBsWLOXL78J5Ac6EaPHomtrS0dOng/3s5IkVqvwoWLEBj4LadPn+LEiWMMGtSfwMBvMyyfvI+sq4/KumojhBBCiMfWvr03X3/9JYULF6FUKcuAWKdOfQ4e3I/BYCA+Pp7Ro0cSHh6GTqczhy83twZs374VSD71eP78r9SsWTvDduzt7WnSxIMVK5bg5dU5w/yyZcsREnLbYtr+/XtRSnHnTggXLvxGnTp1zfMuXrxA2bLleO21flSvXoPDhw9gMiXXqVQpV1xcSrB166YM49Ie19Gjh/D3n4KHR1PefXcMBQoUICws1OK5K6UIvhlM6dJleBibyL2oeMLuxxEbb3iibeYV6TETQggh/qVq165LdHQ0Pj49M8xr0aIVFy9eYNCgfphMil69Xuell8rToEEjli9fgpOTE+++O5a5c2eyc+f3aDQaxo3zo3jx4plsCdq0aZcS3DKeWqxcuQoPHkQSHR1tHmeWkBDP4MG+JCUlMnbsJAoXLmJevmHDJmzZspH+/XuhlKJu3frmgfoAbdu24+DB/RQv7vJE+6VJE08OHtxP//69sbW1pXWbDri4vkTFl19h4fwZaG2daNuhGwcOH6d2fXciHsSDBvQ6LSq7gX/PgEb90zXIIxER0Zl0yeYtF5eChIc/zNdtiNyRNrFO0i7WR9okb9y5c51SpcrnSVn/pntlGo1GVqxYQtGiRenTp3+my2zY8DVarYYePV5j5sxp1Kvnhrd3xt61nBgMBvz9p9C6dVtatGid4/JKKQxGEwajIslgIslowpD6v9EEaaKBRqNBr9Og12ux0WkZP3YY0z4IwKV4cfQ6TfL8PG6X9MeMVqvB2dkpy+Wlx0wIIYQQ2RoyxJfChYswZ85HWS7TrVtP/Pzep2PHrk+8HaUUPj5eNGzYmGbNWlpMNxhNJBlSAliO4UuLrV6Lg50eG70WvU6LjV6LTqsx3ybpwIF9tG3TltKlSjxxffOD9JjlgnzjtD7SJtZJ2sX6SJvkjRe1x+xZMCmFwZActJKD16MAZoS/pY4AACAASURBVDBa7ieNRpMmcCUHMRudFn268PUkpMdMCCGEEC8Ek0mlCV6WAcyYPnxpNdjotNja6nDU2WCTcgpSr3v68GXNJJgJIYQQIs+Yw5dF75eJJKPKEL602uTeLntbHXqdTUovWHIg0z7H4Ss7EsyEEEIIkSsmU7pxXmnGexmNlsOKtNrk0472tjps9DZpTjtq0Gnlql3pSTATQgghRAZGU5pxXml7wAymDGO6ddrk04z2tsmD7VPHe6WedhSPT4KZEEII8QJSSmFSyvxLx9Rer9QAliF86ZJPMzrY6y0G26eedhR5Q/oQhRBCiOdIQIA/Fy9eAJLDl9FoIj7RQHRsIvcfJrDu62/4dvM2gsOiCQ6N5k5ELHcj44iMTiA+0YhWkxy+iha0w6VoAUoXd+SlkgUpV8IJV2dHihcuQBEnOxwL2GBnozOHMqPRyMSJY4mPj8+ybqtWLWfVquXPZD/kp9TnYTKZmDBhDLGxsXlWtvSYCSGEEP9iSimMaQbcvzV8LEkGxe27MSQZTah0PV+//36OWrXr42hvg16fPNBen9L7pX2KwfZbt26kUaMm2NvbP+1T+tfQarV06eLDmjUrGTYs8xul55YEMyGEEOIJJV06RtKfh59oXY1Gk+3tf2yqNsemiifwKHyl/aXjzOkT8Gz2Km6NmqGUwu/9txjyzhi+XLOEXq+/Qe3a9dn01UpOnjyCXqejc5duVK5UiZ/PnODi779Q8aXSvPxyFWYG+BMaegedTsfQocNp0sSDVauW8/vvvxEWdgcfnx589VUgGzd+j1arJSjoDOvWrWX+/EXmuiql2LjxG1auXAvAzJnTsLOz448/LhATE8MbbwzOcN/LTZu+YdeuncTHx2FjY8O0aTMJCwtj1aplLF36OQA7d37PhQu/MWbMBPN6V65cZu7cmRiNRmxtbZk4cSrlyr3EyZPHWbVqGQaDAVfXMowbl3wbqNOnT7F48UKUMlGqlCtTp86gQAEHFi2az5kzp9Foku852r//GwQFneHLL9dgZ2fHtWt/U6lSZaZOnYmNjQ1ffbWW777bQuHCRShYsCDVq9cEoFEjdxYu/JABAwbj6Jj19ckeV76eyty/fz/du3fHy8uLGTNmAHD8+HE6d+5Mu3btWLBggXnZP/74g+7du9O+fXsmTZqEwfDP3kRUCCGEeJYUYFJgNCmMRkV0XBKh92O5FR7NjdBoboZFE3ovlogH8UTFJOLZvB3Hj/6Ik4MNcQ/DUSYDno3qYm+rw6VwAf749QSXL/3Ol4Hf8Nlna9m9awcVK1amadPmDBnyNo0bu7NgwTzq12/AF198jb//HGbPns69exEAJCYm8OWXG+jZsw+lS5fh55/PArBr1w68vTtZ1P3KlUs4OTmZ75MJyTdFX758NYsWLeXTTz8mIuKueV5MTDSHDx9i8eLlBAZ+i4dHMzZt+hY3t4bcvRvBrVs3zdtKf9P0b7/9ij59+rNqVSBdunTj99/Pc//+fZYtW8z8+YtZvforGjVqwtKln5CYmMj06ZPx85vG2rXfULFiZX74YTtbt24iNDSUL75Yz8qVazl0aD/Hjx8F4Pz5Xxk16n3WrdtIaOgdTp06wcWLF9ix4zs+/3wdCxcuITw8zFwfnU5HpUovExR0Jk+Og3zrMQsODmbq1Kls2LABZ2dnBgwYwKFDh5g6dSqBgYG4urry1ltvcejQIVq0aMHYsWOZMWMGdevWZeLEiXz77bf07ds3v6onhBBCPDWbKp7mXq30UgfXK1PyVe2VSv1fYVKg0UBCojFNL5gCZXFvIWwMJvQ6LQXsHt1WSJ8yCL9Mh7asXbWIAnojJ4/tp0MHb2z0OvPqv/xyltatX8XW1hZbW1vWrPkqQx2Dgk4zbpwfAGXKlKVGjVe4cOE3AGrUeHSz8o4du7B7905q1qzF2bOnGT16vEU5wcHBuLiUtJjm7d0ZvV5PiRIlqVWrDufO/WKe5+joxLRpM9i3bw/BwTc4deo4L79cFY1Gg5dXR3bv3om3dxfu3buX4abp7u6efPTRXE6dOo6nZ3M8PZtx8uRxQkPv8N//vg2AyWSkUKHC/PXXFVxcXHj55aoAvP32CAD8/N7H27sTOp0OnU7Hq696cfbsT3h6NqdixUqUKJH8XMqXr8DDh1HcuHGdJk08cXBwAKBVq7YYjUZznUqVKkVwcHCmx0Fu5Vsw27t3L97e3pQqVQqABQsWcP36dcqXL0+5cuUA6Ny5M7t27aJy5crEx8dTt25dALp3786iRYskmAkhhMh3KuWXifFJRhISk/+lPo5PNJKQZDBPq1rSwL2o+EdBy5QxcKUGMh7jLoGpN9W20SeHL/NthnRadLrsL7BqY2ODp2czjh49zP79e5k372OL+Xq9nrSrh4TcpkiRohbLZLyVoTIHDjs7O/PUVq3asmLFEg4c2Ie7u6fFvOTnkby9tHS6R38rZbL4OzT0DiNHvkWPHr1p0sSDYsWcuXz5TyA50I0ePRJbW1s6dPDO8LxbtWrLK6/U5tixI3z77VecOHEUD4+m1K5dhzlzks/EJSQkEBcXl9Kz9WgnREdHExsbk+3ztrV99NxSTzcnt8OjdXQ6nUUw0+n0efbL1HwLZtevX8fGxoa3336bkJAQWrZsycsvv4yLi4t5mRIlShAaGkpYWJjFdBcXF0JDQ/OrakIIIf6llFIkJqWGKAPx5vCUNkgZiU80pPyfErbSPDYHsDTTTI952+j3upVBF5uEVpN8yyCtRoNGk3wR1eTHKf9rQatJMz/t/1oNWg3Y2OhRJtNTXd2+fXtvFi6cR+HCRShVytViXp069dmwYT0+Pj0xGAyMHj2SgICPLEKFm1sDtm/fSp8+/bl16ybnz//K6NETuHLlskVZ9vb2NGniwYoVS5gxY26GepQtW46QkNsW0/bv30urVm0IDb3DhQu/MX78ZHP4unjxAmXLluO11/qRkBDPZ58to2TJ5F6qUqVccXEpwdatm1i27PMM25oyZQJt27bHx6cH//d/FVi06COGDHmbOXNmcOPGdV56qTxr1nzG3bvhjBkzgcjI+/z9919UqFCRdeu+QKPR4ObWgB9+2IGHRzOSkpLYs2cXvr4Ds9zPDRo0ZPLk8QwcOBQbGxsOHz5AkyaPekpDQm5Tp07d7JrqseVbMDMajZw5c4bAwEAcHBx45513sLe3tzgAU1OoKd2B+SidPr7sbgial1xcCj6T7YjHJ21inaRdrM+zbhOjSZGQaCAuITlAxSUYiE/zOPXvuEQD8QlGi8fJ/6d7nBLEHjNDodVAATs99nZ67G31FLDTYW+np6CTXfJ0W12a+SmPbfUp03TY2+pxsE+eZm+no4CtnqtXLlG6dOG820k6Xc7LZKN+/frExMTQvXsv9PrkYeMajQadTkubNm24dOkPBg/uj8lk4rXX+lKxYgUaN27C0qWLKVy4EKNHjyMgwJ8fftgOwMSJUyhVqoS59ye1TIB27dpz/vyv1KlTO0M9qlWrRlTUA+LjY3ByKohGoyExMYEhQ3xJSkpi/Hg/nJ2Lmct1d/dg69ZN9O/fC6UU9eu7cfXqVfP22rXrwIEDP1KqVMkM2xo4cDCzZk1nzZrPsLHRM27cREqWLMGkSVOZOnUCJpMJF5cSfPDBDBwdCzBt2gxmzpxKUlISZcuWY+pUf2xtbbh5M5iBA/tiMBho396LNm3acPbsGYvnrdFo0Go1VK9enT59+vLmm/+hYMFCuLqWTr6dlF6L0Wjk8uU/mTZtusX+SqXVanP12tOo7H4S8hQWLlxIdHQ0fn7J567XrVvHrl270Ol0rFmzBoCtW7dy6tQpRowYwRtvvMHevXsBOHPmDIsWLWLt2rWPvb2IiOhMuibzlotLQcLDH+brNkTuSJtYJ2kX65NTmxiMpkx6nVJ7mAyZnuJ71FNlSHfqL/lxosGU5fbS02k12NvqsLPVYWeTHIrszY/TTk9+bG+TOk2fcVrKYxu9Ns/vtXjnznVKlSqfJ2Xp9VoMudhH/ySj0ciKFUsoWrQoffr0z3SZDRu+RqvV0KPHa8ycOY169dzw9u6c6bLZMRgM+PtPoXXrtrRo0fppq55ruW2XI0cOcu7crwwfnvnlMtIfM1qtJtvOpHzrMWvVqhXjxo0jKioKR0dHjhw5QocOHVixYgXXr1+nbNmybN++nR49elCmTBns7Ow4e/Ysbm5ubNu2jebNm+dX1YQQ4rkXHZdESEQMIRGx3L4bQ6JR8eBhfMZAlfLYYHz8L7Y2em2mIamQgy32tsmhKrOg9Oh/y/nJN7CW651bsyFDfClcuAhz5nyU5TLduvXEz+99Onbs+sTbUUrh4+NFw4aNadas5ROX86yYTCa+/34bU6f651mZ+dZjBrBx40bWrFlDUlISnp6e+Pn5cerUKWbPnk1CQgItWrRgwoQJaDQaLl68iJ+fH9HR0dSsWZPZs2dja2v72NuSHrMXk7SJdZJ2eTaUUjyISeT23ZQAFhFDyN0YbkfEEhWTaF7ORq/FubA9NjptJiHJMjDZpwSljD1VeuxsdNjZal/oG0+/qD1mL5K8bpfc9pjlazB7liSYvZikTayTtEveMilFxIP4TANYXMKjaz4WsNNT2tkBV2dHShd3xNXZAdfijhQvZE/JkoWkTfKABLPn3z8dzOTK/0IIYSUMRhOh9+MIuRtjcRryzr1Yi/FahRxtKe3sQJMaJR8FMGdHijjZ5vmYKiHEsyXBTAghnrGEJCN3Unu+ImK4fTeWkIgYwu7HYUzT8+9cyB7X4g5UK1/UIoA5FbD5B2svhMhPEsyEECKfxMQnEXI3YwCLeBBvvlSlVqOhRNECuDo7UL+KC6WdHXEt7kCpYg7Y28pbtBAvGnnVCyHEU1BKEZUyAP92RGxKAEs+Dfkg3QD8UsUcqFi6EE1ruyYHMGcHShZzkF8kijwVEOCPj08PqlWrken8777bQoECBXj11Q55ul2j0cjkyeOZMsUfe3v7PC07OyaTiUmT3mfy5OnmWyb9m0kwE0KIx2BSinsP4rmdpucreRB+LLEWA/B1uDo78krFYimnHx0p7exA8cIF8uyWLUJkZ/z4ydnOP3/+V+rVc8vz7W7dupFGjZo801AGyRdw7dLFhzVrVjJsWObXEvs3kWAmhBBpGIwmwu7HpQSv2JRfP6YMwE9KMwDfwQZXZ0ca1SiZ/EvI4o6UlgH44hmaOHEs7dp1oGXLNgAMGtSf8eP9WLToIwYNGkq9em4sXfoJhw8fRK/X0aVLdypUqMjRo4c5e/Y0zs7FefnlKgQE+BMaegedTsfQocNp0sSDVauW8/vvvxEWdgcfnx589VUgGzd+j1arJSjoDOvWrWX+/EXmuiil2LjxG1auTL4w/MyZ07C3L8ClSxeJjn7I0KHD2b17J1euXKJZs5aMHDkKo9HIkiUf8/PPZzEaTXh7d+K11/phMBiYPz+Av/66yr1796hcuTLTps3k3r17TJw4hooVK3Hp0p8UK+aMv38AhQoVplEjdxYu/JABAwbj6Phs7gSUXySYCSFeSKkD8NMHsIwD8O1wdXak2ktFzYPvSxeXAfgi2amQs5wIOf1E62o0ZHt7KXfXhjR2zbpnq317b/bu/YGWLdsQHHyDxMREqlSpZp5/4MCPnD//K2vXfo3BYGDYsCHMn7+Ipk2bU6+eG40buzN58njq129gvlfmsGFDWL16HQCJiQl8+eUGAA4e3M/PP5/Fza0hu3btwNu7k0Vdrly5hJOTE05Oj0LR3bvhLF++mh9+2M7s2R+wfv1m7Ozs8PHxZuDAN9m3bxcAn3++jsTERN57bwTVqtVAKYVeb8Py5asxmUz8979vc+LEMapWrc6VK5eZMGEKVapUY9KksezZ8wM9e/ZBp9NRqdLLBAWd+VdcmDY7EsyEEM+12Pgkc/BKvQbY7bsZB+C7FC1A6ZQB+KkBzNVZBuAL6+Xh0ZQFC+YSGxvDvn27ad/ey2L+L7+cpXXrV7G1tcXW1pY1a77KUEZQ0GnGjUu+dWKZMmWpUeMVLlz4DYAaNV4xL9exYxd2795JzZq1OHv2NKNHj7coJzg4GBcXy/taNmniAUDJkqWoUKESRYsWA6BQoUI8fBjFmTM/cfnyJfP9KePiYrl69Qrdu/eiUKHCbNr0LTduXOPmzWDi4uIAKFq0mDl8VqxYmaioKPP2SpUqRXBwcC73ovWRd5zH9Jxch1eI55JSiqjYpJRB9zHmX0LejojhQfSjAfh6XZoB+LVccU25BEXJog7YZHLzYSFy0tjVLdterew87YVMbWxs8PRsxtGjh9m/fy/z5n2crnw9ac+qh4TcpkiRohbLZLwwu8JoNAJgZ2dnntqqVVtWrFjCgQP7cHf3tJgHyb1/er1lpLCxedSrrMvkZu1Go4lhw/5rvh9mZGQkBQoU4OjRQ3z22XJ69eqDt3cXIiMjzZ/B6e8IlPazWafTPxfjOCWYPYYkg4mxS48Tn2jEwU6Ho70NDvZ6HO1tcLTX42D+X49jgfTTkv+XX10J8fQeDcCPTbkA66OB+DHxjwbg29umDMCvUCzl14/Jl6BwkQH44jnTvr03CxfOo3DhIpQq5Woxr06d+mzYsB4fn54YDAZGjx5JQMBH6HQ6c/hyc2vA9u1bzacyz5//ldGjJ3DlymWLsuzt7WnSxIMVK5YwY8bcDPUoW7YcISG3c1V3N7cGfPfdVjw9m5OYmMiwYYMZM2YCZ878ROvWbenYsQu3bt3k55/P0rBhoxzLCwm5TZ06dXNVB2skwewx6HQaejSvSGScgbv3kj8AYuOTiIiKJzgsiZj45BsBZ8fORpcS5h6FtUcBLznQOVhMexT2XuT70okXk8FoIjwyLsOvH0PuxVgMwC+YMgC/YfWSuDo7UDpl/JcMwBcvitq16xIdHY2PT88M81q0aMXFixcYNKgfJpOiV6/Xeeml8jRo0Ijly5fg5OTEu++OZe7cmezc+T0ajYZx4/woXrx4pttq06Yd58//Ss2ar2SYV7lyFR48iCQ6OtpinFl2fHx6cvNmMAMH9sVoNOLt3Zn69RtQuHARPvhgEvv27Uavt6FWrdrcvn0bt2w6Jo1GI5cuXcTP74PH2rY1k3tl5kJ29/8zGE3EJhiIjTcQE59ETFxyeEsNcTEp05PnW05L+0GTGXtbXRaBzgbHAul67NL01DnYPR/dutmRezJap8dtl8QkI3fuxZqD1+2U2xCF3ou1GIBfLGUAfurFV1OvAVbQwTab0kVa8lrJGy/qvTKNRiMrViyhaNGi9OnTP9NlNmz4Gq1WQ48erz3j2sGRIwc5d+5Xhg9/+stlyL0ynxN6nZZCDrYUeoIPCoPRZBnW4pLMAS823kB0yv+p0+7cizVPS8rh4Clgp8/QO+eQEujSBjwHez1OaZaxt9OjlR4HkUdi4w2Per5S7v8YEhHD3chHA/A1GihRpACliztSt3Lx5B6w4o6UKuZAATt5qxLinzRkiC+FCxdhzpyPslymW7ee+Pm9T8eOXZ/5BWa//34bU6f6P7Nt5ifpMcsFa/zGmWQwpvS8ZQx0Mel67NJPMxiz3l8aDTjYZRboLHvnHOz0GabZ2+qe2Wkka2yT541SCqNJYTQqjCYThjSPjSaFwagwGpMfG02KxCQjMUkmLl27Zw5gkRkG4BcwX3w1NYDJAPz8Ja+VvPGi9pi9SKTHTDwVG72OIk46ijjZ5bxwGkopEg2m5LAWl2RxmjVteEs77W5UgnmaMZsQrNVoMo6nK5Am4Nk9mpb+FK2tjfa5GhtkMikMaUJLaoAxpD42qpS/Hz02GjOGH3MoSl0m7fKp07NbPl0dsiw/7XRz+U/2hcfOVkdpZwdq/l8x868fSxd3lAH4QgiRDQlmLyiNRoOdjQ47Gx1FC+Y+1CUkGYmJM2QIdJn12MXEJxEWGWeell0frU6ryTzQ2WX+Q4n7cQYiImLShJDchR1DugBiDk7GtIHlUbm5Xf5ZdEdrNRp0Og06rQa9TotO++hvnVaLTqdBn/J/8jQNBWx05sd6nTbD8hZlaTXodFr0aR6bl0lbbsoyep2W6pVdMCUmPVchW4hUSik5tsVjeZKTkhLMRK5pNBrsbfXY2+pxLpy7cQQmpYhPMOb4w4jolP8fxCQSEhFjHmOXl0FHAymhQpsSRFJCR9oAkjo9ZRk7Gx06u0fhJ314sQg7mQUki+VTgk2asKNPF44yBKK06+k0aLUaqxwLWLxIAcLDDTkvKMS/jF5vS0xMFI6OhSSciWwppYiJiUKvz93Ycwlm4plKPc3pYK8n8x9kZ82kFHEJlr9qdXC0IyY63hycLMNO5sEmNWjJ6TQhRG4VLerC/fvhREdHPnVZWq0Wk0nGmFmbvGwXvd6WokVdcrdOnmxZiGdAq9GkXA7EBigAyIBmIcSzpdPpKV7cNecFH4O8f1mnf7pd5CdQQgghhBBWQoKZEEIIIYSVkGAmhBBCCGElJJgJIYQQQlgJCWZCCCGEEFZCgpkQQgghhJWQYCaEEEIIYSUkmAkhhBBCWAkJZkIIIYQQVkKCmRBCCCGElZBgJoQQQghhJSSYCSGEEEJYCQlmQgghhBBWQoKZEEIIIYSVkGAmhBBCCGElJJgJIYQQQlgJCWZCCCGEEFZCgpkQQgghhJWQYCaEEEIIYSUkmAkhhBBCWAkJZkIIIYQQVkKCmRBCCCGElZBgJoQQQghhJSSYCSGEEEJYCQlmQgghhBBWQoKZEEIIIYSVkGAmhBBCCGElJJgJIYQQQlgJCWZCCCGEEFZCgpkQQgghhJWQYCaEEEIIYSUkmAkhhBBCWAkJZkIIIYQQVkKCmRBCCCGElZBgJoQQQghhJSSYCSGEEEJYCQlmQgghhBBWQoKZEEIIIYSVkGAmhBBCCGElJJgJIYQQQlgJfX4W7uvry71799Drkzczffp0bty4wdKlSzEYDAwYMIB+/foBcPz4cWbPnk1CQgJeXl6MGjUqP6smhBBCCGF18i2YKaW4du0aBw4cMAez0NBQRo0axebNm7G1taVPnz40btyYsmXLMnHiRAIDA3F1deWtt97i0KFDtGjRIr+qJ4QQQghhdfItmP31118ADBo0iMjISHr37o2joyNNmjShSJEiALRv355du3bRqFEjypcvT7ly5QDo3Lkzu3btkmAmhBBCiBdKvo0xi4qKwt3dnU8//ZQ1a9bw9ddfc/v2bVxcXMzLlChRgtDQUMLCwjKdLoQQQgjxIsm3HrN69epRr1498989e/Zk9uzZvPPOO+ZpSik0Gg0mkwmNRpNhem44Ozs9faUfg4tLwWeyHfH4pE2sk7SL9ZE2sT7SJtbpn2yXfAtmZ86cISkpCXd3dyA5bJUpU4bw8HDzMuHh4ZQoUYJSpUplOj03IiKiMZlU3lQ+Cy4uBQkPf5iv2xC5I21inaRdrI+0ifWRNrFO+d0uWq0m286kfDuV+fDhQ+bOnUtCQgLR0dFs2bKFefPmceLECe7du0dcXBx79uyhefPm1KlTh7///pvr169jNBrZvn07zZs3z6+qCSGEEEJYpXzrMWvVqhW//vorPj4+mEwm+vbti5ubG6NGjeI///kPSUlJ9OzZk9q1awMQEBDAyJEjSUhIoEWLFnTo0CG/qiaEEEIIYZU0Sqn8Pf/3jMipzBeTtIl1knaxPtIm1kfaxDo9t6cyhRBCCCFE7kgwE0IIIYSwEhLMhBBCCCGshAQzIYQQQggrIcFMCCGEEMJKSDATQgghhLASEsyEEEIIIayEBDMhhBBCCCshwUwIIYQQwkpIMBNCCCGEsBISzIQQQgghrIQEMyGEEEIIKyHBTAghhBDCSkgwE0IIIYSwEhLMhBBCCCGshAQzIYQQQggrIcFMCCGEEMJKSDATQgghhLASEsyEEEIIIayEBDMhhBBCCCshwUwIIYQQwkpIMBNCCCGEsBISzIQQQgghrIQEMyGEEEIIKyHBTAghhBDCSkgwE0IIIYSwEhLMhBBCCCGshAQzIYQQQggrIcFMCCGEEMJKSDATQgghhLASEsyEEEIIIayEBDMhhBBCCCshwUwIIYQQwkpIMBNCCCGEsBISzIQQQgghrIQEMyGEEEIIKyHBTAghhBDCSkgwE0IIIYSwEhLMhBBCCCGsRI7BrFu3bmzYsIG4uLhnUR8hhBBCiBdWjsHMz8+PM2fO8OqrrzJ9+nQuXbr0LOolhBBCCPHC0ee0gJubG25ubkRFRfH9998zbNgwSpQoga+vL15eXs+ijkIIIYQQL4THGmMWFRXFtm3b+PbbbylYsCBeXl5s27YNPz+//K6fEEIIIcQLI8ceszFjxnDo0CFatmzJtGnTqFevHgCvv/46Hh4ezJgxI98rKYQQQgjxIsgxmFWuXJmJEydSrFgxyxX1etavX59vFRNCCCGEeNHkeCpz69atGUJZqkqVKuV5hYQQQgghXlQ5BrMyZcoQFBSEyWR6FvURQgghhHhh5Xgq8+rVq/Tt2xe9Xo+trS1KKTQaDUFBQc+ifkIIIYQQL4wcg9m6deueRT2EEEIIIV54OQazMmXKsHPnTo4cOUJSUhJNmzbFx8fnWdRNCCGEEOKFkuMYs1WrVrF8+XKqVq1KzZo1Wb16NUuWLHkWdRNCCCGEeKHk2GO2detW1q9fj5OTEwA9e/akd+/eDBs2LN8rJ4QQQgjxInmsK/+nhjKAggULotfnmOeEEEIIIUQuPdblMr744guSkpJISkpizZo1lC5d+lnUTQghhBDihZJjMPvggw/Yt28fdevWpW7duuzZs4epU6c+i7oJIYQQQrxQcjwnGRoaSmBgIHFxcZhMJhwdHTl+/Lj0mgkhhBBC5LEsg9mFCxdQSjFu3Djmz5+PUgoAg8HAtGnT2LNnzzOrpBBCCCHEiyDLYLZ+/XqOHTtGWFgYI0aMeLSCXs+rr776TConhBBCCPEiyTKY+fv7A7BgwQJGjRr1zCokhBBCCPGi/lX9ggAAIABJREFUynHwv5eXF0OHDgXgzz//pGvXrvz111/5XjEhhBBCiBdNjsFs2rRp9OrVC4CqVasycuTIXP0qc86cOYwfPx6AP/74g+7du9O+fXsmTZqEwWAA4Pbt2/Tr148OHTrwzjvvEBMT8yTPRQghhBDiXy3HYBYXF2cxpqxt27ZER0c/VuEnTpxgy5Yt5r/Hjh3LlClT2L17N0opvv32WyD5khx9+/Zl165dvPLKK3LLJyGEEEK8kHIMZhqNhosXL5r/vnr1KlptzjcMiIyMZMGCBbz99tsA3Lp1i/j4eOrWrQtA9+7d2bVrF0lJSZw+fZr27dtbTBdCCCGEeNHkeB2z//3vf/j6+lKlShUA/vrrLz788MMcC54yZQqjRo0iJCQEgLCwMFxcXMzzXVxcCA0N5f79+zg5OZlv85Q6PbecnZ1yXigPuLgUfCbbEY9P2sQ6SbtYH2kT6yNtYp3+yXbJMZi1atWKXbt2ERQUhE6no06dOjg7O2e7zoYNG3B1dcXd3Z3NmzcDYDKZ0Gg05mWUUmg0GvP/aaX/+3FERERjMqlcr5cbLi4FCQ9/mK/bELkjbWKdpF2sj7SJ9ZE2sU753S5arSbbzqQsg9mJEydwd3e3uJCswWDg7NmzALRr1y7LQnfu3El4eDhdu3blwYMHxMbGotFoCA8PNy9z9+5dSpQoQbFixXj48CFGoxGdTkd4eDglSpTI1ZMUQgghhHgeZBnMduzYgbu7O4GBgRnmaTSabIPZ6tWrzY83b97MTz/9xOzZs+nUqRNnz57Fzc2Nbdu20bx5c2xsbGjQoAH/396dxzdVJXoA/90sTboBXSkiIjvaYXVEOuO0Ag7wmdoBOk/Ap7g74FPEx2NvAVlExCr7qiKOD0UUCsrDMgroPARkcRRx8MkMi1CxlO5r0uSe90fSm6RJmtI26S35fT+ffprc9dyctvn1nJNz9+7di7S0NOzatQvJyclNvCwiIiKi1sdrMFu8eDEA2ycmu3bt2iwny8rKQmZmJsrLy5GYmIiHH34YADB//nzMmjUL69evR4cOHfDaa681y/mIiIiIWhNJ1N4E04vk5GR06tQJY8eOxciRI2EwGAJVtuvCMWbBiXWiTqwX9WGdqA/rRJ1aeoyZz3kvPv/8c/z5z3/GwYMHMWzYMCxcuNBl+gwiIiIiah4+g5lGo0FKSgpWrFiBt99+G6dPn8aYMWMCUTYiIiKioOIzmFksFvz1r3/FpEmTMGHCBPTt2xe7du0KRNmIiIiIgorPeczuvvtu9OjRA//2b/+GVatWISQkJBDlIiIiIgo6PoPZtm3bcOuttwagKERERETBzWdXJkMZERERUWD4vhs5EREREQWE12D22WefAQDMZnPACkNEREQUaEIIlJnLcaksFzXWmhYti9cxZitXrsS9996LcePGITs7O5BlIiIiImo2spBRYipFQXURCquLUFhdbP9ue1xUXQSzbAtkf6r4A4Ym3NNiZfUazMLDwzFixAjk5eUhLS3Nbf3HH3/s14IRERERNUSNbEGRh8BV+7jIVAJZyC77ROjDEW1shw7h8UiM6YVoYxSije3wu553oLTI1EJXUk8we+ONN3DmzBlkZGRg7ty5gSwTERERkaLKUu0xcNU+LjW73kJJgoS2hjaINkahS9vOuMMYZQ9eUYgxtkOUMQoGrefpvwy6EAAqDGYRERG48847sXHjRsTHx+P777+HxWJB3759ERHh/R5PRERERA0lhEB5TQUKq4u8djVWWapc9tFJWkQZ2yHaGIXEmN6Itj+u/YoytIVWo22hK2oan/OYlZWVYcKECYiNjYXVakVeXh42bNiAgQMHBqJ8RERE1IpZZStKzKV1wpZz+CpGjew64N6oNShdi93a3qo8rg1ekSER0Eg35sQSPoPZyy+/jKysLAwePBgAcOTIESxduhTbt2/3e+GIiIhI3WqsNSg0eR7bVVhdjGKv47ui0CE8wd7i5drVGKoLhSRJLXRFLctnMKuoqFBCGQAkJSVhyZIlfi0UERERqUOVpUoJW566GsvM5S7bS5DQztDWa2tXtLEdQryM76IGBDNJkpCbm4uOHTsCAC5fvgyttnX22xIREZGDEAJlNeUeWrocz6ss1S776DQ6RBtsQatPzG0ugSvaGIV2rXh8lxr4DGbPPPMMxo0bh6SkJEiShEOHDmH+/PmBKBsRERE1gVW2othU6rmb0VSEoupi1MgWl32MWqMSsrq17eI2sD4yJPyGHd+lBj6D2b333ouuXbvi6NGjkGUZEydORLdu3QJRNiIiIqqH2VrjNn9XgdPjEnOp2/iuSH0Eoo1R6BjeAX1ibnfragzTh7bQ1RDQgGAGAF27dkXXrl39XRYiImpGFtmCwupiFFQX4rxJi7LSamgkjccvraSBJEnQwPs6raR120bZjy0oflFZU+U0rsu91au8psJle42kQdsQ2/xd3dt1RUydsV1RxiiEaPUtdDXUEA0KZkREpD6144OuVRWioKrQ9r26ENeqClBQbeumEhABKYsESQloGkkDDZwDnQaSEuI00EgSNJIWGrivU7aBBI3G+TheAqKm7rm0Ludscjlq94MGWo39u9M2GklybF9nG43T6+HYRlI+bVh7myDvE6cWo9rqOr5Lr9HZ5u8yROHm2JvcWrvaGdpwfFcrx2BGRKRiJqvZHrpsYcv2vVAJY+Y68z+1CYlEbGg0urW9FbEJ0YgxRiM2NBo3xcWgoLAcMmTIQoYsBGRhhSwErEKGELXLZcgQkGUrZDjWObax7wcBWZYho+465+PIdfavs95lG6dzyRbUCBmyxXEur+WoPaaHcwUqlF4vjT14Cvt1OQvVGW1TRoRGoUdUV7eB9ZH6iKCdRiJY+AxmM2bMwLJlywJRFiKioGMbnF3iErauVTtawMpqXKciCNGGINYYjbjQWPSO7oFYYwxiQqMQGxqDGGOU12kI4qIiEW4p87juRiU7hUGrkCHgOSC6BDoIWIXVdT/hHB6tSqCS3dY5njsCsPu62u/h4QborUbEOIWvUB3HdwU7n8HszJkzEEIwoRMRNYIQAhWWSkdXo3Pwqi5EYXWRy+BsjaRBlKEtYkJj0Cf2dsSE2lq8alu+IvTh/HvcQBpJA0iAFlqocVRVXFwk8vODKyyTbz6DWXx8PFJTU9GvXz+Eh4cryzMzM/1aMCKi1qLGWoOC6iKl1cul27GqyG2cUIQ+HDGh0egceTMGxvdFrDFaCWBRhnYcI0QUxHwGswEDBmDAgAGBKAsRkSrJQkapucxpkH2BS3djibnUZXu9Rqe0cHVv18UpeNm6G406YwtdCRGpnc9g9uyzz6K6uhoXL15Ejx49YDKZEBrKPnAiurFUWapwraoIBXVCV0F1IQqqi2BxmoRTgoS2hjaIDY3GbdE9ncZ42cLYjXyDZSLyL5/B7Ntvv8UzzzwDnU6Hbdu2YdSoUVi/fj0GDhwYiPIRETUL5zm9XMd62bobKyyVLtuH6oyINUajQ3iCbayXPXTFhEYj2hgFvYYfaiei5ufzL8vLL7+MLVu2YNq0aUhISMCyZcvw4osvYseOHYEoHxFRg3ia0+tadYHyuNhU4jJ9glbSIsYYhZjQaNzSppPLOK9YYzTC9GEteDVEFKx8BrPq6mp0795deZ6SkoLly5f7tVBqI4TAN/mnIRebUVlh9jAztuQ02aBjckWN8wSGzhMvKtvUs85p4kPHBIX8JBYFN09zeindjfXM6dW9XRd7a1cMYo22bse2hjbsbiQi1fEZzHQ6HUpKSpRQcO7cOb8XSm0ssgVbf/gQVZaqFi2HBKlOcKsbAusJg/Cwn8eg6GVd7b4NOZdyPu/Bs3YG7PpuDaOVtJCcZuT2dGuYcLMOZqsZWknL8HoD8DanV+3junN6GbQhiKkzp1dtd2N9c3oREamVz2D29NNP46GHHkJ+fj6mTp2KL7/8EgsXLgxE2VRDr9VjyW8zEdlOj6vXytxmm3ZMVFhnJmqnma1dZ7v2PPu113XOM227rauzn8dz1SkTZNTINd7LWzvzt9dzOdarjVbSQitpoNVo7Y+19scaL49rt9HU2b4hx/Hw3Hkbl+VOyyQNtBpdneM7jnMjt+LUO6dXVQEKTcUe5vRqh5jQaPSJvV0JXbXzenFOLyK60UhCCJ/3rLh48SK+/PJLyLKMpKQkdOvWLRBluy4FBeWQZf/efoOTAbqqveWJp5CozLIt2767rXOeiVvZxvPs2I6ZtB1BsnZZWLgeJWWVsMpWWIUVViHXeWyxfRdW+3Lnx/YvWXZ67Ok4jn0DcYsXCVIDA15jA6UGWklXZ19fx3EKk/UGXNv8W1JYDX7M/cnpFkK1c3oVotpqcrne2jm9nMd42QbaxyDK0JZzejUT/v1SH9aJOvm7XjQaCTExEV7XN+hjRRaLBbIsQ6fTQafjJ5HIRpIkpTuypQT6D5tcJ/hZZCvkOuHNKqyw2LeR6wQ/i9O+snMIdN7HOUA67yPLjnMpy23lscgmWC2OMtWWR/Z0/AC3dHqe08ve5cg5vYiIXPhMWTt27MBrr72Gu+++G1arFWvWrMHcuXMxYsSIQJSPSFU0kgYarQZ6Vd7gpWEc9wB0D35KmBQyLLWtjc5h0mV71wBZux0E0CmuPQyWMMQYo9EmJJLdjUREDeQzmG3ZsgXZ2dmIj48HAPz888+YOHEigxlRK1X74Qkt/NdFyC4aIqLG8dkHpdfrlVAGADfddBP0+tbbWkBERESkVl5bzL7//nsAQK9evbBw4UKMGzcOWq0WO3fu5Kz/RERERH7gNZhNnjzZ5fnnn3+uPJYkCZmZmX4rFBEREVEw8hrMDhw4EMhyEBEREQU9n4P/8/PzkZ2djeLiYpflM2bM8FuhiIiIiIKRz8H/Tz/9NE6dOmWbTNTpi4iIiIial88Ws5qaGqxZsyYQZSEiIiIKaj5bzBITE/Hjjz8GoixEREREQc1ni9nAgQMxevRoxMXFudyOaf/+/X4tGBEREVGw8RnM3nzzTWRlZeGWW24JRHmIiIiIgpbPYNamTRv84Q9/CERZiIiIiIKaz2A2ePBgvPzyyxg+fDhCQkKU5YmJiX4tGBEREVGw8RnMPv74YwDAvn37lGWSJHGMGREREVEz8xnMeAcAIiIiosDwGczeeustj8sfe+yxZi8MERERUTDzGcyc5zAzm804fvw4kpKS/FooIiIiomDkM5i99NJLLs/z8vKQkZHhtwIRERERBSufM//X1b59e+Tm5vqjLERERERB7brGmAkhcPr0acTExPi1UERERETB6LrGmAFAhw4dMGPGDL8ViIiIiChYXfcYMyIiIiLyD6/BbPbs2V53kiQJS5Ys8UuBiIiIiIKV12DWo0cPt2VFRUV4++230bFjR78WioiIiCgYeQ1mjz/+uMvzw4cPY+bMmUhLS0NmZqbfC0ZEREQUbHxOl2GxWPDyyy9j6tSpyMzMxIsvvojQ0NAGHXzlypX4wx/+gNTUVOXTnYcPH0ZaWhqGDx+O5cuXK9ueOXMG6enpGDFiBDIyMmCxWBp5SUREREStU73B7MKFCxg7dixOnz6NXbt2YcSIEQ0+8LFjx3D06FF89NFH2LFjB9555x388MMPmDNnDtatW4e9e/fi9OnT+OKLLwAA06dPx7x587Bv3z4IIbB9+/amXRkRERFRK+M1mO3YsQNjx47F73//e7zzzjtISEi4rgMPGjQIf/nLX6DT6VBQUACr1YrS0lJ07twZnTp1gk6nQ1paGnJycpCbm4vq6mr0798fAJCeno6cnJymXRkRERFRK+N1jFlGRgY0Gg02bdqE119/XVkuhIAkSfj66699Hlyv12PVqlXYvHkzRo4ciatXryIuLk5ZHx8fj7y8PLflcXFxyMvLu64LiYmJuK7tGysuLjIg56GGY52oE+tFfVgn6sM6UaeWrBevwWz//v3NcoLnnnsOTz31FCZNmoQLFy5AkiRlXW3Ik2XZ4/LrUVBQDlkWzVJmb+LiIpGfX+bXc9D1YZ2oE+tFfVgn6sM6USd/14tGI9XbmOQ1mDV1Sox//etfMJvNuO222xAaGorhw4cjJycHWq1W2SY/Px/x8fFISEhAfn6+svzatWuIj49v0vmJiIiIWpvrvol5Q12+fBmZmZkwm80wm83Yv38/xo8fj/Pnz+PixYuwWq3Ys2cPkpOT0bFjRxgMBpw8eRIAsHv3biQnJ/uraERERESq5POWTI2VkpKCU6dOYfTo0dBqtRg+fDhSU1MRHR2NyZMnw2QyISUlBSNHjgQAZGVlITMzE+Xl5UhMTMTDDz/sr6IRERERqZIkhPDvwKwA4Riz4MQ6USfWi/qwTtSHdaJOLT3GzG9dmURERER0fRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJRjMiIiIiFSCwYyIiIhIJfwazNasWYPU1FSkpqZi2bJlAIDDhw8jLS0Nw4cPx/Lly5Vtz5w5g/T0dIwYMQIZGRmwWCz+LBoRERGR6vgtmB0+fBiHDh1CdnY2du3ahe+//x579uzBnDlzsG7dOuzduxenT5/GF198AQCYPn065s2bh3379kEIge3bt/uraERERESq5LdgFhcXh1mzZiEkJAR6vR7dunXDhQsX0LlzZ3Tq1Ak6nQ5paWnIyclBbm4uqqur0b9/fwBAeno6cnJy/FU0IiIiIlXyWzDr0aOHErQuXLiATz75BJIkIS4uTtkmPj4eeXl5uHr1qsvyuLg45OXl+atoRERERKqk8/cJzp49i4kTJ2LGjBnQarW4cOGCsk4IAUmSIMsyJElyW349YmIimqvI9YqLiwzIeajhWCfqxHpRH9aJ+rBO1Kkl68WvwezkyZN47rnnMGfOHKSmpuLYsWPIz89X1ufn5yM+Ph4JCQkuy69du4b4+PjrOldBQTlkWTRb2T2Ji4tEfn6ZX89B14d1ok6sF/VhnagP60Sd/F0vGo1Ub2OS37oyr1y5gmeeeQZZWVlITU0FAPTr1w/nz5/HxYsXYbVasWfPHiQnJ6Njx44wGAw4efIkAGD37t1ITk72V9GIiIiIVMlvLWZvvvkmTCYTli5dqiwbP348li5dismTJ8NkMiElJQUjR44EAGRlZSEzMxPl5eVITEzEww8/7K+iEREREamSJITwb/9fgLArMzixTtSJ9aI+rBP1YZ2o0w3blUlERERE14fBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEgldC1dACIiIqJAEVYLRGUxRGUx5MpiiIrax0UQVaWo/M0fgbY9Wqx8DGZERETU6gnZAlFZ6ghZ9sBVN4CJ6jL3nSUtpLC2kMLaAZIU+MI7YTAjIrrBCSFDCLmli0HUKEK2QlTZApeosIeuyhKIyiLITuFLVJUBEK47SxolcGkiYyG17w4pvJ3teViUbV14FCRjBCTJNrorLC4SFfkewluAMJgREbVSQggIU7m9JcD2ZiVXFDlaCZQ3rRKUQ0AKCQOMEbY3IUO4/XuE0/fwOs8jIOlCWvoy6QYlZBmiutT951UJX7afXVFVAoi6gUuCZGxjC1nhUdDGdXUKXO2Ux5KxDSRN6xpOz2BGRKQyQgjAVGHrfnF507J1z8hVJRAVtiAG2eJ+AEO47c0prB00UR2gCWuHsHAjKgoLbUGuuhyiohBywSUIUzlgMXsvjDbEFtA8hbba73XWwRCmtD5Q8BFCtv2M1Q1ZFU7divbQBQ8tuVJoG1uoCmsHbWxn5bEmLMoRuELbQNJoW+Dq/I/BjIgoQIQQgLkScmWJ/Y2qyCV82d7EbAEMVg+BKyRUeXPSdOhlbxmwd8eERdnDWFuPrVzRcZGweumeERYzhKnC9mZaG9ycHyvfKyAXXAKqyyHMFe6tGArJFs48hbjaljpP69g6p2qOfxiKXEOW8+D5itrAZXXbXzJG2kJVeDtoom6GpjZkhTt1K4a1haQJ7mgS3FdPRNRMhLnK6Y2qqE5XouMNDFYPrVN6oxKytO27u3XHaGrDl87gl7JLuhBbKAqPavA+QsiAuaqeMOcU9CqKmtA65/TYQ5hDSFir66pSG+UfBpef1yL3Fq6KYt8ttB072ENWbeCyh6+wtpC0+sBfXCvEYEZ+JYSw/SJbzBA1Jtt3iwnCYgbcvtdu42F53e3tx6rQSBD6UEghYfYv22OEhEIyhCnLERLqWG9fDn0o/6CTT6LG5DbI2NFKUOQIXBaT+866EEjhtpYsbVwXSJ3b2VsJouyBy/6mpTcG/sKaSJI0gCEckiEcQPsG76e0zrmEOS+tdQWXlG2br3XOvsxPIVdNhBBATZVrq5YyHrHYZUwirDXuB3BuoU3o6QhZ9p/h+lpoqfEYzIKcqA1NFjNQ4yUw2Zc7ljkClvu+tcsdzz2NIaifBOgNtl92XYjtD6j9P3rJGAlJFwPoDJD0BhgNOlSVlkCYK20tFhVF9seV9f9nXktvdIQ6Q50AFxJm+2/c4HguhYTa3gRqA59WD6mFP1pNjSMsJqc3Jg+f8KooglxZAtRUue+s1TsCV0xnSJ36OXXL2Fq3NGFRtp8v/ny4aN7WOQ8Br6IYcuFliOpyz2G5llbvObC1ktY5UVPtNn7L05hEj38H3Vpo27qM3/J3Cy3Vj8FMxYQs12k1MnkMPo6WqDrBqTYsWb1vA9l9HIBP2hBIekdYgs4WoiRDGKTwKEeY8hKubMsNyvK621xP2ImLi0S+t3EzsgXCXGX7g26qdAS2Os9t29QGu2LIRT8r2/kMlRqt9wDnFPgcz0NdgyBb7ZqdsJjtQcu5C6Z2LJcjgMFc6b6zVucYaBx9M7Sd+ji6FZ26ZhASxsAVQI1unbPWeA5zyrIKwL5OLspV1nv/va/TOldvmHNqrbuOgCMsJteQ5SV8oabafWeXFtpbIYX1d22hrW3hCgltcHko8BjMGkkIGbDWeAg8rl1t7l1yJreWKNfuPcc2Hgf/+qLV2YOSUyuTzgBJb4QU2sYpLLluA+eApDc4Pa+zXKtvNZ+2kjQ6SMZIwBjZqP2FELa6sgc5OAU5JeSZnJ/bv1cW24NfVf3/sddybrVTWuRC6++GdXoObUhQhARhtUDYP43oOuDY9Q0Mpgr3nTVaR+Bq2wHam26zd8XY5zCyv2nBEB4Ur2WwkOwtm41qnasNcJ4+BFEb8KpKbIHOVOE5KNXy0jpX0CYSVQXX7D/X9p9fs5cWWntLljbmFkid+jpaaJ0Gz7OF9sbAYNYAwmpB1Sev4qfyfFjN9tDlaQCvLxqtU8uQa+CRDOHQ6Jxboty3cQ1RBveWKJ2BrS/NSJIk2x86vfG6/rA7E7IFMFc7WudMzi10judKS5492MlFPyvPfbbaSdpGdcO6jMdrwY+dN222bo1T4IqHtkPPOgPn7d0zhvBW8w8FtSyX1rk28Q3eT1hrfIQ559a5nyFM5Si1mAD7JxU1UTdBe3Oio1uRLbRBi8GsISQNNDG3wBDbHiaLxi04ObrnQtxbmpy784L8I8DBSNLolAk9G8PWamf20kJXt2vWaXnJFUdLX4Nb7ULrtND56IZ1btXTubfauc/WXRu4GjJbtwQp1N6aFREDbftuSshyfMqrHaTQSAYuUgXJ3qqFsHYN3qe+oRgUvJgUGkDSaGBMeoC/RBRwtlY72wcdGt9qZ3W0xtV2udbtmq3bkldVArnkFyUIepqTyLWgWkeLnC4EleYKWCuKPXySTnJMHhkeZfukov2xyye+WuFs3UREzYHBjOgGJ2m0zdtqV9+HJ+wtdKHtesGkDXeasdv+acUbeLZuIqLmwGBGRPVqTKsdW5eJiBqHfQVEREREKsFgRkRERKQSDGZEREREKsFgRkRERKQSDGZEREREKuH3YFZeXo777rsPly9fBgAcPnwYaWlpGD58OJYvX65sd+bMGaSnp2PEiBHIyMiAxdKI2xERERERtWJ+DWbffvstHnjgAVy4cAEAUF1djTlz5mDdunXYu3cvTp8+jS+++AIAMH36dMybNw/79u2DEALbt2/3Z9GIiIiIVMevwWz79u2YP38+4uNt9xs7deoUOnfujE6dOkGn0yEtLQ05OTnIzc1FdXU1+vfvDwBIT09HTk6OP4tGREREpDp/hh0fAAAR+klEQVR+nWD2xRdfdHl+9epVxMXFKc/j4+ORl5fntjwuLg55eXn+LBoRERGR6gR05n9Zll1udCyEgCRJXpdfj5iYxt1u5nrFxUUG5DzUcKwTdWK9qA/rRH1YJ+rUkvUS0GCWkJCA/Px85Xl+fj7i4+Pdll+7dk3p/myogoJyyHLdGyY3L95mRn1YJ+rEelEf1on6sE7Uyd/1otFI9TYmBXS6jH79+uH8+fO4ePEirFYr9uzZg+TkZHTs2BEGgwEnT54EAOzevRvJycmBLBoRERFRiwtoi5nBYMDSpUsxefJkmEwmpKSkYOTIkQCArKwsZGZmory8HImJiXj44YcDWTQiIiKiFicJIfzb/xcg7MoMTqwTdWK9qA/rRH1YJ+rU0l2ZAW0x8yeN5vo+LKD281DDsU7UifWiPqwT9WGdqJM/68XXsW+YFjMiIiKi1o73yiQiIiJSCQYzIiIiIpVgMCMiIiJSCQYzIiIiIpVgMCMiIiJSCQYzIiIiIpVgMCMiIiJSCQYzIiIiIpVgMCMiIiJSCQYzIiIiIpVgMPPh1KlTeOWVVxq1788//4wHH3wQI0eOxNNPP42KiopmLt2Nz9+v/9WrV/HEE09g1KhRGDNmDI4cOdLUIgeFQP1elJeX495778VXX33V2KIGDX/XidlsxuLFizF69Gikpqbi0KFDTS1yUAhEvfzXf/0X0tLSMGrUKBw+fLipRQ5KTamnWitWrMDq1aubXBYGMx/++c9/oqCgoFH7LliwAP/+7/+OnJwc/OpXv8K6deuauXQ3Pn+//suWLcPQoUOxe/duvPrqq5g2bRqsVmtTi33DC9TvxaJFi1BaWtrYYgYVf9fJG2+8gaKiImRnZ2PFihWYPXs2eKtl3/xdL7t374Ysy/j444+xbNkyzJo1q6lFDkpNqaeysjLMmTMHb731VrOUJWhuYv7VV19hw4YN0Ov1uHz5MoYOHYqwsDB89tlnAIBNmzbhyJEjWL9+PSRJQp8+fTBjxgykp6ejsrISjz32GIYMGYJ58+bBYrHAYDDgpZdewq233urxfDU1Nbjrrrtw7Ngx6HQ6XLlyBQ899BD2798fwKtWD7W+/vv27cPgwYPRtm1bWK1W3HXXXTh48CAiIyP9/ZKoglrrBQD27t2LEydO4OzZs3j22Wdx1113+fOlUA211klaWhpeeeUV9O7dG4Dtjaxr167QaILj/3u11sv777+PgwcPYu3atfjuu+8wY8YM/PWvf/X3y6Faga4nANi1axeuXr2KyspKaLVaTJ48uWkXIYLE0aNHxYABA8TPP/8sKisrRf/+/cV7770nhBBi1qxZYvXq1SIpKUlcuXJFCCHEtGnTxKeffip27NghZs6cqWy3d+9eIYQQO3fuFNnZ2V7Pl5eXJ373u98pz2tqakRiYqK/Lk/1WsPrv3HjRvHQQw816TpbG7XWS25urhg7dqyoqqoSDz30kDh69GizXbPaqbVO+vTpI/7yl7+IBx54QIwdO1YcOXKk2a65NVBrvZhMJjFu3DiRlJQkEhMTxb59+5rtmlujQNeTs1WrVolVq1Y1+Rp0TYt1rUvPnj3RoUMHAEBUVBSSkpIAADfddBNycnIwcOBAJCQkAIDS17xz505l/5SUFCxcuBD/+7//i6FDh2LIkCFezyWEgCRJLsvqPg82an79t2zZgvfffx///d//3biLa8XUVi+yLCMjIwNz586F0Whs+gW2QmqrEwCwWq345ZdfsHXrVvzf//0fnnzySXzyySdB07oMqLNe1qxZg/79++O9997DhQsX8OijjyIxMREdO3Zs2sW2YoGsJ38IjjZoO71e7/Jcq9Uqj3U6ncsPfWFhIQoLC122HzlyJLKzs9G3b19s2bIF8+fP93qu6OholJWVKeOV8vPzER8f3xyX0Wqp9fVftmwZPvjgA2zdulX5ZQ4maquXc+fO4dy5c8jIyMCoUaNw+vRpZGZm4ujRo42+xtZGbXUCALGxsUhNTYUkSejduzcSEhJw/vz5Rl1fa6XGetm/fz/S09MhSRK6dOmCfv364dSpU426vhtFIOvJH4IqmNWnT58++Oabb5Cfnw8AWLJkCfbv3w+tVguLxQIAeP755/Hdd99h/PjxmDJlCv7xj394PZ5er8evf/1r7N27F4CtDzo5Odn/F9JKtdTrv2XLFnz11Vd47733lP+gyKEl6qV79+744osvsHv3buzevRu/+tWvsHjxYgwePNhPV9m6tNTvypAhQ5RtLl26hCtXrqBLly7NfXmtVkvVS+/evZXxU4WFhTh9+jRuu+225r68G0Zz15M/BFVXZn3i4+ORkZGBJ554ArIso3///khPT8dPP/2ENWvWICsrC5MmTUJGRgbWrl0LvV6PF154od5jzp8/H7NmzcL69evRoUMHvPbaa4G5mFYokK//e++9h6tXr+K5557D2rVrERERgQkTJij7bdq0Ce3bt/fn5bYaLVEvU6ZMCcCVtV4tVSfTpk3DwoULkZqaCgBYvHhxUHVj+tJS9TJ79mzMnTsXqamp0Gg0mDp1ar0D1YOdP+qpuQXNpzKJiIiI1I4tZk00YcIEj/MsjR8/Hg888EALlCi48PVXJ9aL+rBO1In10joEsp7YYkZERESkEhz8T0RERKQSDGZEREREKsExZkQ3sF69eqFnz55ut81Zu3Ytbr755us+3qxZs9CjRw888cQTzVK+zz//HN9++y2mTJmC/fv348iRI8jMzGyWYwfCBx98ALPZjAcffLDRx3j88ceRlZWF6Ohol+Xl5eV48sknUVZWhilTpmD48OGNOn6vXr3QsWNH7N+/32X+ptWrV2PNmjX48MMP0adPHwBAdnY2tm3bhurqatTU1OCOO+7A9OnT0aZNG4wfPx5VVVWoqanB+fPn0bNnTwC26U1effVVfPPNN3j11VdRXFwMIQQSEhIwc+ZM9OjRo5GvDFFwYjAjusG9/fbbbm/6avHdd9+hpKQEADBs2DAMGzashUt0fU6ePNnk4PHll196XH7mzBkUFBTg008/bdLxAdss8idOnMCdd96pPP/kk0/Qtm1bZZsNGzbgb3/7G9auXYvY2FjU1NRgyZIlmDRpEt59911s27YNAHD58mWkpaVh9+7dyr5msxkTJ07E5s2bkZiYCMB2c+2nnnpKmSOKiBqGwYwoSGVnZ2Pt2rXYvXs3JEnCn/70J0ycOBEdOnRAVlYWbrrpJpw7dw5GoxFLly5Ft27dXPY/ceIEli1bhqqqKuj1ejz//PNITk7Gzp078eGHH6KqqgoRERHYuHEjXnjhBVy8eBHFxcUIDw9HVlYWysrKsG3bNlitVkRGRqJz587Yt28fNm7ciF9++QUvvPACcnNzIYTA6NGj8eSTT+Ly5ct49NFHkZKSgm+//RalpaWYPn06fv/737td38GDB7FixQrIsoywsDAsWLBAmYxzzZo1kGUZ4eHhmD17Nvr27YvVq1cjNzcX+fn5yM3NRfv27fHKK68gPj5eCSZ6vR4GgwELFy7E+fPnceDAAXz55ZcwGo0YMWIE5s2bh4KCAuTn56Njx45YsWIFYmJiMHToUIwZMwZHjhzBlStXMGrUKDz//POYPXs2AOCRRx7Bpk2blDtPnDt3DnPmzEFeXh5GjRqF999/H4cOHfJa7m+++QZXr15Fr169kJWV5fZa/PGPf8RHH32kBLOTJ0+ie/fuqK6uBgBUVlZi48aNyM7ORmxsLADbBKczZszAp59+CrPZjJCQEK8/S1VVVSgrK0NlZaXLOSMiImC1WhnMiK5Hk++2SUSq1bNnT3HfffeJP/7xj8rXf/zHfyjrp06dKubPny9mz54tMjMzhRC2mwD37t1bHD9+XAghxLvvvivGjBkjhBBi5syZ4o033hCFhYUiKSlJfPPNN0IIIX788UcxaNAg8dNPP4kdO3aIO++8U5SVlQkhhPjkk0/EokWLlHPOnTtXLFy4UAhhu+nvggULhBBC7NixQ/z5z38WQgjx4IMPis2bNwshhCgtLRVpaWliz5494tKlS6Jnz57iwIEDQgghcnJyxD333ON23fn5+eKOO+4Q33//vRBCiH379oknnnhC/POf/xS/+c1vxE8//SSEEOLw4cPit7/9rSgrKxOrVq0Sw4YNU8o9ceJEsXLlSmGxWERiYqLIy8sTQgiRnZ0ttm3b5vJ6CCHEli1bxMaNG4UQQsiyLJ588knx5ptvCiGEGDJkiFi6dKkQQohffvlF9OnTRylDz549RUFBgds1HD16VKSmpgohhM9yjxgxQtTU1Hj6ERA9e/YUP/74o7jrrruEyWQSQggxZ84cceDAATFkyBBx6tQp8d1334nBgwd73L+uS5cuif79+7st37x5s+jbt68YOnSomDZtmvjggw9EZWVlg45JRA5sMSO6wdXXlblgwQKMGjUKRqPR5Sa+vXv3xq9//WsAwJ/+9CcsXLgQRUVFyvpTp07hlltuQb9+/QAAPXr0wMCBA3Hs2DFIkoRevXohIiICgO2+c506dcI777yDixcv4tixYxgwYIDX8lZWVuLrr7/G5s2bAQCRkZFIT0/H3/72N/Tr1w96vR4pKSkAgNtvvx3FxcVux/j666/Ro0cP3H777QCA4cOHY/jw4di6dSsGDx6MTp06AQCSkpIQHR2N06dPAwAGDRqklPv2229HSUkJtFotRo4cifHjx+Oee+7B3XffrZzf2SOPPIITJ07grbfewoULF3D27Fnl9QGgdNO2b98eMTExKCkpUcrhy9GjR+std//+/aHTef9zHhMTg759++LgwYNISUnBiRMnsGDBAmW9RqOBLMsNKos3jz32GO6//34cP34cx48fx+uvv47XX38dH374Ie8QQHQd+KlMoiBWUFAAk8mE0tJSXL16VVnuqevJeZnVanUZSA7Yxi3V3msuLCxMWf7uu+8iIyMDRqMRaWlpuO+++yDqmT5RlmW39bIsK8fW6/XKhxnqlsG5rM7rhBD44YcfIMtyveU2Go3KckmSlHJkZWVhw4YNuOWWW7Bp0yZMnTrV7ZyvvPIKVq5ciaioKIwbNw6//e1vXa7DYDB4PHZD+Cq38+vtzejRo/HRRx/h008/xdChQ12CXPfu3WGxWHDhwgWXfUwmE5566ink5eXVe+yTJ0/ijTfeQEREBIYMGYIZM2bgf/7nfyBJktcxdETkGYMZUZCqqanB1KlTMWXKFDz77LP4z//8T9TU1AAAfvjhB/zwww8AgPfffx8DBgxAmzZtlH379++Pc+fO4dSpUwCAs2fP4vjx4xg0aJDbeQ4dOoQxY8bg/vvvR5cuXXDgwAFYrVYAcLlxcK2IiAj069cPW7duBQCUlZVh165d+M1vftPga+vXrx/+9a9/4ezZswCA/fv3Y/r06UhKSsKhQ4dw6dIlAFDGfDm3bNVVWFiIlJQUtGvXDo8++qhyg+O65T906BAeeeQRjB49GjExMTh8+LBynfXx9BrU1Zhy1zVs2DD8/e9/x9atWzFmzBiXdSEhIXjqqaeQkZGBa9euAbAN6F+yZAmqqqp83js2Ojoa69evx4kTJ5Rl+fn5KC8vVz69SUQNw65MohvcI4884jZdxtSpU3H06FHExsbi/vvvBwB89tlnWL58OVJSUhAbG4sVK1YgNzcX0dHRWLZsmcv+0dHRWLlyJRYtWoTq6mpIkoSXXnoJXbp0wd///neXbR9//HHMmzcPH374IQBbqPvxxx8BAIMHD8a0adOwaNEi5dN8gK2FauHChdi5cyfMZjPS0tKQnp6O3NzcBl1zbGwssrKyMHPmTFitVkRERGD58uXo3r075s+fj2effRZWqxVGoxEbNmyot6stOjoaTz/9NB599FEYjUZotVosXrwYAJCcnIylS5cCAJ555hksW7YMK1euhF6vx8CBA/HTTz/5LOvIkSMxYcIErF692muIaUy56zIYDBg6dCj+8Y9/eDzPpEmTEBoaqkyFYjKZMGjQIKxbt87nsbt06YK1a9di+fLl+OWXX2AwGBAZGYklS5aga9euDS4jEfGWTERUx1dffYVFixZhz549LV0UIqKgw65MIiIiIpVgixkRERGRSrDFjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVILBjIiIiEglGMyIiIiIVOL/AXoxLMpX4M/mAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_exploration_1 = df_exploration[df_exploration['play_order']==1]\n",
    "df_exploration_1 = df_exploration_1.set_index('method')\n",
    "df_exploration_2 = df_exploration[df_exploration['play_order']==2]\n",
    "df_exploration_2 = df_exploration_2.set_index('method')\n",
    "\n",
    "for field in ['victory', 'draw', 'defeat']:\n",
    "    plt.figure(figsize=(10,7))\n",
    "    #plt.plot(df_exploration_1.index,\n",
    "    #         [baseline_smart[field] for _ in range(len(df_exploration_1.index))],\n",
    "    #         label='baseline smart')\n",
    "    plt.plot(df_exploration_1.index,\n",
    "             df_exploration_1[field],\n",
    "             label=field+' (play first)')\n",
    "    plt.plot(df_exploration_2.index,\n",
    "             df_exploration_2[field],\n",
    "             label=field+' (play second)')\n",
    "    plt.plot(df_exploration_2.index,\n",
    "             (df_exploration_1[field] + df_exploration_2[field])/2,\n",
    "             label=field+' (mean)')\n",
    "    plt.xlabel('Exploration constant for MCTS')\n",
    "    plt.ylabel('Number of '+field)\n",
    "    plt.title('Performances for different value of the exploration constant for MCTS, for 1 000 games')\n",
    "    plt.legend(loc='best')\n",
    "    plt.show()"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "RL_project_v2.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}