Blackjack- Q Learning (Doubling Down).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### RANDOM ACTIONS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "HELLO\n",
      "{0: 4574, 1: 4691, 2: 3286}\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEKCAYAAADTgGjXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x109fed9e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average payout after 1000 rounds is -58.55\n"
     ]
    }
   ],
   "source": [
    "from blackjackenv import BlackjackEnv\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "KEY = {0: 'STAND', 1:'HIT', 2:'DOUBLE DOWN'}\n",
    "env = BlackjackEnv()\n",
    "env.reset()\n",
    "\n",
    "num_episodes = 10 # num of episodes the simulation will run\n",
    "num_rounds = 1000 # num of rounds/hands played per episode\n",
    "\n",
    "# PERCENTAGE OF ACTIONS TAKEN\n",
    "ACTIONSD = {0: 0, 1:0, 2:0}\n",
    "\n",
    "average_payouts = []\n",
    "\n",
    "for sample in range(num_episodes):\n",
    "    round = 1\n",
    "    total_payout = 0 # to store total payout over 'num_rounds'\n",
    "    \n",
    "    while round <= num_rounds:\n",
    "        obs1, payout1, is_done1, ddown1 = env._get_obs()\n",
    "\n",
    "        action = env.action_space.sample()  # take random action \n",
    "        while action == 2 and not ddown1:\n",
    "            action = env.action_space.sample()\n",
    "\n",
    "        ACTIONSD[action] += 1\n",
    "\n",
    "#         print(action)\n",
    "        \n",
    "        obs, payout, is_done, info, ddown = env.step(action)\n",
    "#         if action == 2:\n",
    "#             continue\n",
    "#             print('ACTION: ' + str(action))\n",
    "#             print('OBS1: ' + str((obs1, payout1, is_done1, ddown1)))\n",
    "\n",
    "#             print('OBS2: ' + str(obs))\n",
    "#             print('PAYOUT: ' + str(payout))\n",
    "#         print('ID_DONE: ' + str(is_done))\n",
    "#         print('INFO: ' + str(info))\n",
    "\n",
    "        total_payout += payout\n",
    "        if is_done:\n",
    "            env.reset() # Environment deals new cards to player and dealer\n",
    "            round += 1\n",
    "    average_payouts.append(total_payout)\n",
    "\n",
    "print(ACTIONSD)\n",
    "\n",
    "plt.plot(average_payouts)                \n",
    "plt.xlabel('# of Samples Ran')\n",
    "plt.ylabel('Payout after ' + str(num_rounds) + ' rounds')\n",
    "plt.show()     \n",
    "print (\"Average payout after {} rounds is {}\".format(num_rounds, total_payout/num_episodes))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### BASIC STRATEGY- Based on Thorpe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{0: 7193, 1: 6407, 2: 984}\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEKCAYAAAAMzhLIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x109fdf320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average payout after 1000 rounds is -42.3\n"
     ]
    }
   ],
   "source": [
    "#### BASIC STRATEGY #####\n",
    "def basic_strategy(player_hand, dealer_up_card):\n",
    "    \"\"\"\n",
    "    2D Arrays Storing Actions- Rows represent player hand, and column represents dealer's up card\n",
    "    Action Hard- actions appropriate when player has a hard hand (no aces)\n",
    "    Action Soft- actions appropriate when player has a soft hand (at least one ace)\n",
    "    \"\"\"\n",
    "    actionsHard = [[1]*10]*7 # 2 to 8\n",
    "    actionsHard.append([1]*2 + [2]*4 + [1]*4) #9\n",
    "    actionsHard.append([1]*1 + [2]*8 + [1]*1) #10\n",
    "    actionsHard.append([1]*1 + [2]*9) #11\n",
    "    actionsHard.append([1]*3 + [0]*3 + [1]*4) #12\n",
    "    actionsHard.extend([[1]*1 + [0]*5 + [1]*4]*4) #13-16\n",
    "    actionsHard.extend([[0]*10]*5) #17-21\n",
    "    \n",
    "    actionsSoft = [[2]*10] # A,A\n",
    "    actionsSoft.append([1]*4 + [2]*2 + [1]*4) # A,2\n",
    "    actionsSoft.append([1]*4 + [2]*2 + [1]*4) #A,3\n",
    "    actionsSoft.extend([[1]*3 + [2]*3 + [1]*4]*2) #A4, A5\n",
    "    actionsSoft.append([1]*2 + [0]*4 + [1]*4) #A,6\n",
    "    actionsSoft.append([1]*1 + [0]*1 + [2]*4 + [0]*2 + [1]*2) #A,7\n",
    "    actionsSoft.extend([[0]*10]) #A,8\n",
    "    actionsSoft.extend([[0]*10]) #A,9\n",
    "    actionsSoft.extend([[0]*10]) #A,10\n",
    "    \n",
    "    player_sum = sum(player_hand)\n",
    "\n",
    "    if 1 in player_hand and len(player_hand) == 2:\n",
    "        return actionsSoft[player_sum-2][dealer_up_card-1]\n",
    "    else:\n",
    "        return actionsHard[player_sum-2][dealer_up_card-1]\n",
    "\n",
    "# Make sure actions have been stored correctly mainly when dealer's upcard is A\n",
    "assert (basic_strategy([5,10], 2)) == 0\n",
    "assert (basic_strategy([1,4], 1)) == 1\n",
    "\n",
    "################################################ SIMULATION ##########################\n",
    "\n",
    "num_episodes = 10 # num of episodes the simulation will run\n",
    "num_rounds = 1000 # num of rounds/hands played per sample\n",
    "total_payout = 0 # to store total payout\n",
    "\n",
    "average_payouts = [] # store payout per episode\n",
    "\n",
    "# PERCENTAGE OF ACTIONS TAKEN\n",
    "ACTIONSD = {0: 0, 1:0, 2:0}\n",
    "\n",
    "for x in range(num_episodes):\n",
    "    round = 1\n",
    "    round_payout = 0\n",
    "    while round <= num_rounds:\n",
    "        player_hand, dealer_up_card, is_done, ddown = (env._get_obs())\n",
    "        \n",
    "        # Take action based on normal strategy stored above\n",
    "        action = basic_strategy(player_hand, dealer_up_card) \n",
    "        if ddown and action == 2:\n",
    "            pass        \n",
    "        elif not ddown and action == 2:\n",
    "            action = 1\n",
    "                \n",
    "        ACTIONSD[action] += 1\n",
    "\n",
    "        obs, payout, is_done, _, ddown= env.step(action)\n",
    "        \n",
    "        round_payout += payout\n",
    "        total_payout += payout\n",
    "        if is_done:\n",
    "            env.reset() # Environment deals new cards to player and dealer\n",
    "            round += 1\n",
    "    average_payouts.append(round_payout)\n",
    "\n",
    "print(ACTIONSD)\n",
    "plt.plot(average_payouts)                \n",
    "plt.xlabel('# of Samples Ran')\n",
    "plt.ylabel('Payout after ' + str(num_rounds) + ' rounds')\n",
    "plt.show()     \n",
    "print (\"Average payout after {} rounds is {}\".format(num_rounds, total_payout/num_episodes))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "import json\n",
    "\n",
    "class Agent():\n",
    "    def __init__(self, env, epsilon=1.0, alpha=0.5, gamma=0.9, num_episodes_to_train=50000):\n",
    "        self.env = env\n",
    "\n",
    "        # Looks like n is number of valid actions from the souce code\n",
    "        self.valid_actions = list(range(self.env.action_space.n))\n",
    "\n",
    "        # Set parameters of the learning agent\n",
    "        self.Q = dict()         # Q-table which will be a dictionary of tuples\n",
    "        self.epsilon = epsilon   # Random exploration factor\n",
    "        self.alpha = alpha       # Learning factor\n",
    "        self.gamma = gamma       # Discount factor- closer to 1 learns well into distant future\n",
    "\n",
    "        # epsilon will reduce linearly until it reaches 0 based on num_episodes_to_train\n",
    "        # epsilon drops to 90% of its inital value in the first 30% of num_episodes_to_train\n",
    "        # epsilon then drops to 10% of its initial value in the next 40% of num_episodes_to_train\n",
    "        # epsilon finally becomes 0 in the final 30% of num_episodes_to_train\n",
    "        self.num_episodes_to_train = num_episodes_to_train # Change epsilon each episode based on this\n",
    "        self.small_decrement = (0.1 * epsilon) / (0.3 * num_episodes_to_train) # reduces epsilon slowly\n",
    "        self.big_decrement = (0.8 * epsilon) / (0.4 * num_episodes_to_train) # reduces epilon faster\n",
    "\n",
    "        self.num_episodes_to_train_left = num_episodes_to_train\n",
    "        \n",
    "\n",
    "    def update_parameters(self):\n",
    "        \"\"\"\n",
    "        Update epsilon and alpha after each action\n",
    "        Set them to 0 if not learning\n",
    "        \"\"\"\n",
    "        if self.num_episodes_to_train_left > 0.8 * self.num_episodes_to_train:\n",
    "            self.epsilon -= self.small_decrement\n",
    "        elif self.num_episodes_to_train_left > 0.6 * self.num_episodes_to_train:\n",
    "            self.epsilon -= self.small_decrement\n",
    "        elif self.num_episodes_to_train_left > 0.4 * self.num_episodes_to_train:\n",
    "            self.epsilon -= self.small_decrement\n",
    "        elif self.num_episodes_to_train_left > 0.2 * self.num_episodes_to_train:\n",
    "            self.epsilon -= self.small_decrement\n",
    "        elif self.num_episodes_to_train_left > 0:\n",
    "            self.epsilon -= self.small_decrement\n",
    "        else:\n",
    "            self.epsilon = 0.0\n",
    "            self.alpha = 0.0\n",
    "\n",
    "        self.num_episodes_to_train_left -= 1\n",
    "\n",
    "    def create_Q_if_new_observation(self, observation):\n",
    "        \"\"\"\n",
    "        Set intial Q values to 0.0 if observation not already in Q table\n",
    "        \"\"\"\n",
    "        if observation not in self.Q:\n",
    "            self.Q[observation] = dict((action, 0.0) for action in self.valid_actions)\n",
    "\n",
    "    def get_maxQ(self, observation):\n",
    "        \"\"\"\n",
    "        Called when the agent is asked to find the maximum Q-value of\n",
    "        all actions based on the 'observation' the environment is in.\n",
    "        \"\"\"\n",
    "        self.create_Q_if_new_observation(observation)\n",
    "        return max(self.Q[observation].values())\n",
    "\n",
    "    def choose_action(self, observation):\n",
    "        \"\"\"\n",
    "        Choose which action to take, based on the observation.\n",
    "        If observation is seen for the first time, initialize its Q values to 0.0\n",
    "        \"\"\"\n",
    "        self.create_Q_if_new_observation(observation)\n",
    "\n",
    "        # uniformly distributed random number > epsilon happens with probability 1-epsilon\n",
    "        # EXPLOIT WITH 1 - P(epsilon)\n",
    "        if random.random() > self.epsilon:\n",
    "            maxQ = self.get_maxQ(observation)\n",
    "\n",
    "            # multiple actions could have maxQ- pick one at random in that case\n",
    "            # this is also the case when the Q value for this observation were just set to 0.0\n",
    "            action = random.choice([k for k in self.Q[observation].keys()\n",
    "                                    if self.Q[observation][k] == maxQ])\n",
    "            if action == 2 and observation[3] == False:\n",
    "                action = 1\n",
    "        # EXPLORE WITH P(epsilon)\n",
    "        else:\n",
    "            action = random.choice(self.valid_actions)\n",
    "#             print(action)\n",
    "#             print(observation)\n",
    "            if action == 2 and observation[3] == False:\n",
    "                action = 1\n",
    "\n",
    "        self.update_parameters()\n",
    "\n",
    "        return action\n",
    "\n",
    "\n",
    "    def learn(self, observation, action, reward, next_observation):\n",
    "        \"\"\"\n",
    "        Called after the agent completes an action and receives an award.\n",
    "        This function does not consider future rewards\n",
    "        when conducting learning.\n",
    "        \"\"\"\n",
    "\n",
    "        # Q = Q*(1-alpha) + alpha(reward + discount * utility of next observation)\n",
    "        # Q = Q - Q * alpha + alpha(reward + discount * self.get_maxQ(next_observation))\n",
    "        # Q = Q - alpha (-Q + reward + discount * self.get_maxQ(next_observation))\n",
    "#         print(self.alpha)\n",
    "        self.Q[observation][action] += self.alpha * (reward\n",
    "                                                     + (self.gamma * self.get_maxQ(next_observation))\n",
    "                                                     - self.Q[observation][action])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "HELLO\n",
      "{0: 61157, 1: 59297, 2: 16863}\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAEKCAYAAAA8QgPpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzs3Xl83HWd+PHXe5JMJvfRXE3a9C6l\nLbSUUg5BLpFDFDwXFxUVl3VVZNc9FNld711dFVdd9bfsgqLrguiCHIoICCJ3S0vvll5pm6TN0RyT\nzCRzfn5/fL/fySSZTCaTzKRJ38/HI49mvjOZ+Uym+b6/n/f7c4gxBqWUUmqiXNPdAKWUUjOTBhCl\nlFJp0QCilFIqLRpAlFJKpUUDiFJKqbRoAFFKKZUWDSBKKaXSogFEKaVUWjSAKKWUSkvudDcgk6qq\nqszChQunuxlKKTWjvPbaa53GmOrxHjerA8jChQvZtGnTdDdDKaVmFBE5nMrjNIWllFIqLRpAlFJK\npUUDiFJKqbRoAFFKKZUWDSBKKaXSogFEKaVUWjSAKKWUSosGEKWUyqLBUIQHNh1lNmwnrgFEKaWy\n6A972vmHX21jW3PvdDdl0jSAKKVUFnX5ggAc9w5Oc0smTwOIUkplUe9ACICOvsA0t2TyNIAopVQW\nOQGkXQOIUkqpiejxWymsjj5NYSmllJqAHr/dA/FqD0QppdQExGog/RpAlFJKTUCsBqI9EKWUUhPh\npLA6+wNEozN7MqEGEKWUyqLegRD5uS7CUUOXXVCfqTSAKKVUlgyGIgyEIiypLgZm/lwQDSBKKZUl\nXrv+sbzWCiAzfS6IBhCllMoSp4C+rLYEgPYZvpyJBhCllMqSHjuALK3RHohSSqkJcEZgzS3zUJKf\nm7EaSK8/RCAcychzx5uWACIi7xWRnSISFZH1I+67XUT2i8heEbky7vhV9rH9IvK57LdaKaUmx0lh\nlRe4qS7Nz1gA+etfbOFdP3wxI88db7p6IDuAdwHPxR8UkZXADcAq4CrghyKSIyI5wA+Aq4GVwPvt\nxyql1IzhrINVVpBHTUk+7RlaD6u9L0BNSX5GnjvetAQQY8xuY8zeBHddB9xvjAkYYw4B+4EN9td+\nY8xBY0wQuN9+rJpC//XcQd73ny9NdzOUmrV6B0KIQIknl+oST8ZqIB19AWpKPBl57ngnWw2kATga\nd7vZPjbWcTWFdh3zsqmpi1AkOt1NOSU8uauNK7/zHL12XlzNfj3+EGUFebhcQk1JZlJYkaihsz9A\nTekM7oGIyFMisiPBV7KegyQ4ZpIcT/S6t4jIJhHZ1NHRkU7TT1n9gTBRA20zfGhhNnX5gnSmsShe\nU6ePz/zidfa29fFqU1cGWqZORr0DVgABqCnJxx+M0B8IT+lrnPAFiBqyksLKzdQTG2PeksaPNQPz\n427PA1rt78c6PvJ17wLuAli/fv3MXmgmy/xB6z9ya88g8yoKp7k1M8Pf/XIr/mCY+285P+WfGQxF\n+MTPN+NyCbkuYfORbq5YWZvBVs5s+9r6eGzbMerKPDSUF7CiroSa0synZzKhZyBEuR1Aqu0TfLt3\nkGJ7ZvpUcBZprM5CCitjASRNjwD/KyJ3AvXAMuBVrB7IMhFZBLRgFdr/fNpaOUv5Atawv9aegWlu\nycxxqNNHZ38AYwwiiTrKo33p0V3sOublng+v59+f2seWI90ZbuXUa+r0sbCqKCuv9eMXm/jfV47E\nbpd4ctn6z2/F5Rr6fUejhqPdfhbMyU6b0tXrD1JW6AaI1Sja+wIsTjOAPLi5mcbKQtYvrIwdc9Ji\n1bO1iC4i7xSRZuB84Dci8gSAMWYn8ACwC/gd8EljTMQYEwY+BTwB7AYesB+rppDP7kq3aABJiTGG\n472D9A2GU97b4endbdz36hH+6pIlXLailrPml7OtuZdwFupOgXCEwdDk5wa8frSHS771LK8f7ZmC\nVo2vxx9kcXURz3/2Uj524SL6BsN4B4fXjZ7a3cal33qW/e19U/ratz+4jR+/cGjKnq83rgfi1CjS\nrYOEI1HueGgH//ncwWHHnZFds3kU1kPGmHnGmHxjTK0x5sq4+75mjFlijDnNGPN43PHfGmOW2/d9\nbTraPdv5g9bJRQNIaryDYQbsE/L+9v6UfuYPe9opyc/lb69YDsC6BRX4gxHeaEvt5yfjU/+7hb/4\n6aZJP88bbdZJ+lBn5tsMVuG5stDNvIpCVjeUAXDCN3wV2yNdfqIGnt07dXXPna293PfqUZ7c1TZl\nz9kzogYC6c9G33O8j4FQhJbu4X+vQymsWRpA1MnJKeadSimscCTKVx/bxZET/gn/bPxggwMdvpR+\nZmtzD2fOLyM3x/rTO2t+BQCbM5zG2t/ez5O72mg6kVo7k3FOWG1Z2hApvvBcUWSlf7pHBJAu+/af\n9nVO2eve/bzV8zjeOzWDSqJRY/VACq33UlaQhzvHlfZcEOf/zMgLvva+AGUFeXjycibX4BRoAFEx\nQ0X0UyeAvH60h/9+/hD3pJGmiD+xHEihBzIYirDnWB9nziuPHZtfWcCcIjdbjmQ2HfTTl5oApmTI\nsPP/I1s76vX4Q5TZJ91Ku37QNSKAdNsT9F45dGJKlvBo7xvk0a2t5LqEY72DGDP58Th9gTDGEAuG\nIkJ1ST4dcb/Hke8rmc2HrQDSOxAaNpKrvW8wK+kr0ACibIFwhFDE+iNp6R6Ykj+YqbCztZdXDp7I\n2PNvsv8In9h5fMzd4Ywx/MOvtvLg5uZhx4/bPZCKwjwOdAwPIN7BEEe7hvdqdh/zEo4a1sQFEBHh\nrMYKthydWA/EGMPf/OJ1nntj/JSNdzDEr15rxiXWSWyyu+A5V7xtGZpFPZJ3WA/E+rd7xEZMJ/qt\n24OhKK8dnnxv7n9eOkw4arhhw3wGQhG8A5MfausEb+e9gJVmcupnLx7oZP1Xnxw2YCCZzUd6yM+1\nTuHxaaz2vuzMAQENIMrmt0dgNZQX4AtG8A5O7dj0dHT5gnzw7ld5/3+9zGPbEo7anrRN9hyMY72D\nbGvpTfiYR7a28sCmZh7bdmzYcWcp7vMWzxnVA/nX3+7h7f/xPMHwUHF8q110XjO/bNhjz2os52CH\nL7bMRSp8wQgPbWnh0a3j/15+uakZfzDCO9bUYwz0TfKzbYn1QJIHkJ+/cnhUEJ2ocCRKXyBMeYHV\n85hTZJ0Yu3zDe1JdviBnNJSR4xKen2QaazAU4X9eOcLlK2o4f3EVAK29k++V9wxYn2+53YsCqw7S\n7rW2tv3ab3YTNXDnk3tjA1rG0tkf4EiXn8tPrwGgpWfo99zuzc4sdNAAomxOF3iZvdHNyZDG+spj\nu/AOhFhVX8Zt97/O49uPjf9DExCNGl473M2Vq2rJdQm/23F81GMGghG+/vgewCrUxjvuHaS8MI9V\n9aW09g4O+6N/fn8HPf4Qrx4amiS4rbmXmpJ86kbMYTir0eqRbJnAqKYT9lXrvnFSZ9Go4acvNXH2\nggouWGqdDEeOYJqIaNRwrMcKHMlqIAc7+rnjoR3c8esdab8WELuQKSuwZhwUuHPw5LlG9UC6/EEa\nKws5a345z++fXAD59ZYWunxBPnrhIuaWW59VqnUQYwyRMXp4sYUUC4f3QNr7BnloSws7W7189E2L\n6OwP8t9/Sp5SddJXbz+zHoBmuwdijLGXMdEeiMoiZwTWspqTI4A8u7edh7a08IlLlnDfLeexdn45\nt963hd/vHH2ST9fBzn66/SEuX1HLBUur+N2OY6NSd//vjwc41jvI+gUV1kifuJPD8d4AdaWe2Pak\nB+1CemvPAEe7rN/fU7uHRvC83tzDmfPKR80XWTOvHJcwoTqIM/t9f3t/0nTjs2+0c/iEnw9fsDCW\nOnFOZOno6A8QjEQpcufQ5h27NvD07nYAnnujI9bLA+sE950n3+Dp3amNbIotPhh30q0sdMdSVo4u\nX5DKIjcXLqtie0vvqCJ7KowxPLO3ne89vY/T55Zy/uI5zC2zAsixFAPID589wGXffjbh78VZyr08\nLoVVU+Kh2x/i357Yw5p5Zfzj207n6tV13PXcgaQrHGw+0kNejnDJaTW4c12xFFbvQIhgJJqVEVig\nAUTZhnog1k5p0zmU1xcIc8dDO1hSXcQnL1tKcX4uP/nIOaxuKONT/7tl0ikKx6Ym6yru7IUVXLWq\njqYTfva2Dc0jaOkZ4D+fO8DbzpzL9Wc1EAxHhw25bPMOUlvqiW0O5NRBXjlk1WwaKwt5ancbxhi8\ngyEOdvhYOyJ9BVCUn8tpdaUTmlDYaZ9A+wPhhCc3YwzP7GnnS4/uorY0n6tW101JAHH+X6xtLCcQ\njo5ZG3h6TxuLq4uoKnbznafeiB2/f+NRvvv0Pn7yYlNKrxe//Lmjosg9rAcSjkStob5Fbi5aVoUx\n8OKBidXNtjf38v7/epmP/HgjuTkuvnr9KqvIXZyPS+B4CimscCTKvS82cfiEP2EK2NlMKr4G4tQq\n2rwB7njbSlwu4e+uPI3BcJTvP71vzNfafKSblfVlFLhzaCgvoNn+XJw5Jdmaqa8BRAFDI7AWVBbi\nznFNawD5zpNv0NIzwDfefSb5udZQxBJPHvd+ZAOLq4u45WebpmT29sambiqL3CyuKuKKlbWIwOPb\nrR5ONGr46mO7MAZuv3oFjZXW0i6H44bBHvcOUlfqoXFOITkuGQogB7so8eTylxcvprl7gL1tfWxv\ntuor8SOw4p3VWM7rR3pSLnDHX52OTGPtavVyw10v85GfbATgzvetJS/HlVIAOd47yIObm8fsWThX\nus7w40SF9F5/iI1N3Vy9uo6PX7yEF/af4JWDJ9jX1seXHt2JCOxo6U1poIbT1tK4k25lkXvYaKVu\n+8q+ssjNmnnllOTn8vz+1OeD+INhbrjrJfa19fOld6ziqc9czNkLrJnduTkuako8KfVA/rSvM3aB\ncSxBwOm1g17psB6IFUCuXFXLhkXWay6pLubPzpnPz185Muz/myMUibKtuYd1duqzobwg9rk4r19d\nrD0QlUVO/r7Ek8fccg+tPdOzoGIkavjV5mauPXPusOUZwEpj/PTmDVSX5PPhH29k7/HJzTp+7XAX\nZy+oiA2nPGdBJU/sPE6vP8THfrqJx3cc59bLljKvopAFc6wA4tRBQpEonf0Bass85Ofm0FhZGJtM\n+MqhLjYsrOSK0631rZ7a1cbWZis9dea80T0QgHWNFfQFwuzvSG1yXnwKZ1/b8N/DX/7PJva39/OV\n61bx5N9czJvs2kcqAeTHLx7iMw9sTVgPgqEeiFO3SbTw5h/3dRCJGi5bUcsHzltAdUk+3/r9Xm69\nbwtF7lxuvXQp3f4QrSmclBPVDSpH9ECc7yuL3OTmuDhvyRz+tK8z5ZGEL+4/gS8Y4bs3nMVNFyzE\nnTv8tFhX5omNuEvml68dxclOJkoB9w6EKMjLGTY/Y+38ci5fUcMd1wzf3ujWy5YSjhoeT/A57D7m\nZTAUZV2jFcQbyguGBjY4s9B1FJbKJmcdrKL8HOrLCqatBrK9pZcef2jMxQVrSjz8z83nkp/r4qM/\n2Zh0CZD+QJgHNh5NWNTs6AvQdMLP+gUVsWNXra5jz/E+rvruc/xpXwdfvm4Vn7x0KQD15QXkuCQW\nQDr6AhhDrCC+pLqYAx39tHsHOdTp49zFldSUelgzv5wnd7ez9WgPC+cUDhuBEy9WSB/Rs2rvG2T9\nV58cVowHqwdS6smlqtjNvrhZ7C12/eXTly/jg+cPPxmmEkB2tngB+OKjOxOuEtvSPUCpJzdW90lU\nSH96dxtzitysnV+OJy+HT1yyhI1N3ew53se33reGS1fU2K81fNTbjpbeUcuj9CZI+1QUDu+BOMF0\njj3J8KJlVTR3D9CU4uTQZ/a2U+TO4ZxFFQnvn1vmGffvocsX5MldbVxzxlyAhBdgzlLu8eYU53P3\nh8+hcc7wxUvnlhVQVezmYIILCqeAvs7+vzuvooCOvgCDoUhsbo4W0VVW+ewUVlF+LvXl0xdA/ri3\nAxG40L5qTmR+ZSFfuX41LT0D/DHJPIjv/2Ef//B/2xJeTb922Dohx/dyrlxdhwgYA7/4y/P50PkL\nYwXvvBwX9eWeWABxrkjryqw/1CU1RTR1+mO593MXzQHgitNr2Hq0h5cPdrFmfuL0FcCiOUWUFeSN\nKqS/dOAEnf3B2BBgx4n+IFUl+SytKWZf3PpPTsF6/cLRJ8NCdw45LsE7RgAxxrCjtZc188po7wvw\n7d+P3vOtpWeAhorCuNz98BNlOBLl2b0dXHJaDTn2Yofv39DI6oZSPn35Mi49rYYVdaW4BHa0eof9\n7N/9citfeHj4qK2eBHMnKovc9A2GY/vWOMHEmaV+6WlWgHoihQEXxhie3dvBm5ZWxdKlI80tKxg1\nmXDzkW4e3doaO/bw6y2EIoZPXLLEnnw4+u+nJ24WeioWVxXHBmbE23ykh7pSD/V2gb+hogCwej3t\nfQEK8nIozs/OOrkaQBQQ1wNx59JQ7qHNO5jxjaWee6Nj1BIiz+3r4IyGMuaMk8O9bEUNVcVuHth0\nNOH9Xb4gP3vpMDA0CzvepqZu3LkuVjeUxo41lBfw4F9dwG9vuyiWHojXWFnIYbu9bXb6pdbugSyt\nLiYYifKr15opzs9lVb31vJfbaazegdCY9Q8Al0s4q7F8VADZGDdPJV5Hf4CqonyW1ZSwr21oJNar\nh7ooyc9lRV0pI4kIZQV5Y/ZAWnoG6PGHeM/6+dx4biP3vtjEjhG9hNaeARrKCyh051LiyR01F2Tz\nkR56B0K8xZ6fAODJy+HRT13IZ+z1vwrcOSytKR7WA2nvG2TP8b5RqaLegRBF7hzycoZOVSOXM+ny\nD++BzK8sZM28Mn6zbfxh3/va+2npGYj1ihKZW+bBH4zQF9cj+5ff7ObW+7bw+Yd2EIpE+eWmZs5o\nKGNVfRm1pYlTwPFLsqRiSU0RBztHB5AtR7tZt2BoNF9DuRVAWuwAUlOan/LK0JOlAUQBVg3EJeDJ\nc1FfXpDxjaV6/SFuvncjn/2/bcOObTnSzZuXVY/783k5Lq5f28DTu9tjcyLi3f38QQZCEd5z9jxe\nOdQ1ql6y8XA3a+eVj7rqPKuxgsqixGmmxsqi2MS4WA/ESWHZI7Ge39/J+oUVsbWuVtSVxP7AE43A\nGvba8yt4o71v2DwNZ6TYce/wK9oT/QGqStwsry2mLxCOpZI2NXWzbkFF7Op/pGQBZIedvlpdX8rf\nX7mCyqJ8Pv/Q9mGF/ZbuAebZV7y1pZ5RKaynd7eRlyNcuGx4D3LkCW11fRk7WocCyEt2z62zPzgs\n5djjD41K+8WWM7EDR1f/8B4IwNvOnMv2lt5x1zh7Zo813PiS08b+P1dXNnwuSCRq2HXMS0N5Afe9\neoR3/vAFdh3z8t718wCoL0+c8ur1T7wH0uULDhuS3OMPcrRrgDMahi5GnB5IS/cA7d7sLWMCKQQQ\nEXmTiBTZ339ARO4UkQWZb5rKJl8wTJE7FxGhvtzpEmcugDyx8zihiOGlgydiI5ReONBJ1MDFSf6Y\n4713/XzCUcOvXx8+G7vHH+TeFw9zzRlz+fw1p+POdfGzl5ti9w8EI+xs6eXsBGmeZBorCznhC9If\nCHPcO0hejlBhn8yWxO3n4KSvwDpxXrmqjvxcFyvnjhNAGssxBrYdtX4fvf5QbFjxyB5IZ3+QOUX5\nLK2xhl3va++jxx9kb1sf5yR5X6VJAsjO1l5yXMLpc0spK8jj769czrbm3tgAgN6BEH2BMPX25Lra\n0vxRo7Ce3tPOuYvmUOJJfqJc1VBGmzcQK/q+YE/+i0TNsAJ570Bo2KglGFrOxElddfkClHhyh/VS\nnFrEb8aZfPrs3g5W1JUwt6xgzMeMnAtyqLMffzDC31yxnG+/dw1vHO/HnePiHWusSX315QUJR231\nDAQn1ANZXG3tbXIwbtXjXcesIO/0cMG6iMlxCS09A1nbC92RSg/kR4BfRNYA/wAcBn6a0VaprPMF\nwhTZedP6WJd4cstQgHVC+OIjO9lzfHi++9FtrTSUF1Ccn8t//cnaz+C5Nzooyc9lbZJaQbzT6kpY\nM6+MX246Oiw/fc8LTfQHwtx62VIqi9y8/cx6HtrcQt9giGjUcMdD2wlHTUo9nXixkVgn/LT1DlJT\n4oltalRWkBebvHXu4uGjx/72rct59NYLKXAnXx11bWM5IkOF9M1HujHGKpLGz4QOhqP0DoSoKs5n\nub1ywBtt/bE1oEaOXotXVpA3Zg1kZ6uXJdVFsVFCb11Zh0uGlkh3hoo2lFu/h9oSz7AFFVt6Btjf\n3p80HeRYbZ8Ad7Z6Mcbw/L5OCuzXjX/O3oHgsIl3MLScSbe9nEmXPxRLXznmVRSyZn45v9k+9lIv\nfYMhNjZ1cclpyds71AOx3n+sp9ZQyrvPnseDn7iA/7ppfaynZNVMBkYNybZW4k3cu03EuSiJX+l5\nl103On3uUADJzXFRV+qhudtKYWVrEiGkFkDCxvrrvA74rjHmu0BJZpulss0XjFCYb/0BO1eYU9ED\n2dnay09ebOIrj+2KHTvRH+DFAye4/qx63r9hPr/Zfozmbj/PvWEVM+OvJMfznvXz2XO8L/ZH3esP\n8eMXDnHlqtpYHeBD5y/AF4zw4OYWvvzYLh7c0sLfvXU55y+Zk+ypR3Hmghzp8llzQMqGX+ktqS6i\n0J3DGQ3DexpF+bksrx3/T6bUk8fS6uLYMt2vNnWR6xKuWlVHe18gNuLMufKeU+xmTnE+lUVu9rf3\nsbGpm7wcSRqAk6ewelldP9T2iiI3a+aXxwYqOENFnZRJTamH9r7B2InSCXzJekCOlXYA2dXqpemE\nn9beQa5aXQcwbHnzRHWDWA/EP9QDSZR2vPaMuexo8SacSwFWryccNVw6To+3ttSDyFAPZEdLL/m5\nLpbaJ/jVDWVcvHzoOerLPYQihk7fUCAcDEUYDEUn1AOZV1FAXo4MK6TvOualpiR/VJBoqChgX3sf\n/YFw1obwQmoBpE9Ebgc+gLV7YA6Q+m9BzQi+QDg2cqPQnUtlkXtKJhM6w09f2H8iNkLo8R3HiUQN\n155Zz4fftAiAf/z1Dlp7B3nz8on1Ct6xpp78XBd3/ekgX398D5d86xm797Es9pg188tZM6+Mrz++\nh5+82MTNFy6KDc+diMa4uSBt3sCoNa1uvnAxn71qxYQC4EhnNZaz5WgPxhg2NXWxuqGMRdVFRKIm\nNvvcmURYZQ80WFpTzBtt/Wxs6uKMhrKk+0CUFeQmnCXd7h2kvS/AqhHB7+Ll1Wxt7qHLF4zl9Z2a\nTm1pPqHIUMpp69Ee3LmuhAX8kUo8eSycU8iOlt7Y2lXvPKsBGL5DX0+CuoGTNnRqAyf6gwkDyNVn\nWAFprDTWM3s6KPHkxobDjiUvx0V1cX5sDbDtLb2cPrc0VucayUmHHesZHgiBCQWQ3BwXC+YUDRvK\nu6vVOyx95ZhXXsCeY1a682RLYf0ZEABuNsYcBxqAb2a0VSrr/IEIhXEplrEKgRO1qamb+jIPc4rc\nfO8P+wF4dGsrS2uKYwXma8+cG0uTvHn52MN3EykryOPKVXU8urWVu547wDkLK7n/L86L7Vzn+ND5\nC2NF9TuuOT2tUSqlnjzKC/M4fMLP8d7B2AgsxxUra7npgoUTft546xor6LFrH1uP9nLOwoq4HLz1\neTgBpLrEOmkury1m73Frtvs5SdJXMNQDGTnJbmfrUAE93iWn1WAM/GlfBy09A7hzXVQVW6/rvH+n\nkP760R5W15eOmog3llUNViH9hX2dNJQXxNoev1xMoh5IXo6LEk9uXA0kcQCZV1HI2vnlCUdjdfuC\nPLO3nTcvq04p4M8t83DMa/W2drV6h43eG2moBz/095NoQmQqFlcVxVY4CIQj7G/vj/Xe4s2rKCBs\n9wSzmcIad7CwHTTujLt9BK2BzDq+YDh2ogKoLyuY9O51xhg2He7izcurWV5bwtcf38MTO4/zalMX\nt12+LHYS/4uLFvPw660sqS5iXkXhOM862t9feRqrG0p525n1savjkd61roHGOdZqra4xRiilYkFl\nITtbvQyEIrE5IFPpLHv48L0vHiYYiXLOwkrqSq335NRBhibOWa+/rKYkNukvWf0DrCAYiRp8wciw\nuQI77RFRI09OZzSUUVGYxx/3dhAIR2koL4h9brWlzpasgyyLFLO9pZc/35D6+JrV9dZQ267+IG87\ncy4F7hxK8nNjPZDBUIRAODqqiA5Dy5kYY/WAKosSfxbXnjmXr/5mN3c/f4irV9dRW+rhFxuP8m9P\n7KFvMMx77JFT46kr83Co08fhLj99gfCoNGW8ersHEj/TfmghxdRrIGCN7ntmbzvhSJR9bf2Eoybh\nYAwnrQjZm0QISQKIiPQBY64FYIwZv5+qZgxfIEyhe+i/Q315AU/tbuOSbz5DiSePRVVFfOHtK8ed\nnxGv6YSfzv4g5yys5B1r6vnPPx7gb37xOsbAtfYy1GDlkD98wUJOq0uvtDa/spBb3rwk6WNEZNyr\n81Rfy5mYOLIHMhWW1hRTnJ/L/9mbVzlLrcBQDj6WwipxAsjQCLD146Rj4mejxweQHS1eFlUVjRo9\nleMS3ry8muf2ddBQXjAsQDupknZvgL3H+xgMRVnbmNoACCB2Fe8LRmLLrVSX5McCSLKrdmc5k75A\nmFDEUFmU+Mr++rMa+NVrzXzlsV185bFdVBW76ewPsmFRJV++blVK6Taw0lIvHjgRmxezqn7sAFJe\nmIcnz8WxuB5IbFXhCaSwwOqBhCKGo90DsQJ6oh6IM7ABTpIAYowpARCRLwPHgZ8BAtyIFtFnHV8w\nQlH+UArrA+c1ErH3cPYOhnhi53G2t/Ty049uYH5lar2EjXb945yFFRTl5/KxixbzzSf2cvrc0tgK\nto4vvmPV1L2ZDFowpzCWKhhZA5kKOS6rCP78/k6WVBcxpzgfYwzuXFds7skJX5D8XBdFdsrRWUF5\nWU3xsLkQicQCiD80LBjsaO0dc6b8JadV8/DrrXT5grz37Pmx4/Gz0YN2gX9tksmSI8WfhC9YMhRA\nnCJ6olnojspCN8e9g7E5IGP1QKqK83n8tos40OHj6d1tvHa4m2vOmMt1a+snlMasK/PQNxjmlUMn\ncOe4kg6KcIbCx29C5aTbJpzCBtwlAAAgAElEQVTCim0V0M+uY14K3TksSPD35/RAcl1DQ8uzIZX5\n7lcaY86Nu/0jEXkF+LcMtUlNA1/AmgfiWFpTwleuXx27vampi5vv3cS7fvQiP/7wOaNqDIlsbOqi\nojAvNhzxQ+cv4OcvH+bPN8wf5ydPXo1xf7wjR2FNlbMarQDirM4qIlYO3umB9AWoKh6abVxV7Ka+\nzDNq8l4iidbD6vEHae4e4MZzE6efLrKHO0fN8FRJfm4OFYV5tPUNcrjLT2WRm/mVY8+nGKmyyGp3\nadwQ6JpSD9vj5p1A4rRPRZGb3ce8o2ahJyIiLK0pHnXRMhFOevfp3e2cVlcybp3HWk9uKIX12uFu\nSj25sSHyqVpizwU50NHPrlYvp88tTZiCdeou1SX5k0rRTlQq1a6IiNwoIjki4hKRG4HJ71qvThrR\nqMEfjFCYZP2c9Qsr+b+/Op88l3DDXS+ntCfHpsPdrF9YGTvRlXjyeOFzl/HB8xdOVdOzrrGyKPZ9\nJlJYMLSw4voFQym3ulJPbB5Cpy8YK2SDdYJ85NYL+exVK8Z97tIEAcRJjYxVGK4qzo+tIjyyxuTM\nRn/9aA9r54/eLGs8X7puNf/89qGVaKuL82NF9GRpn8oiN13+YMJZ6JkQG1nVO5i0gD70+KFBKMYY\nXjxwgvOXzBlzhYCxlBe6mVPk5kC7j13HvKycm/i183NzqCnJz2r6ClILIH8OvA9os7/eax9Ts4Q/\nZF0PFOcnn+i2tKaEBz/xJhrKC/jIT17l4ddbxnxsR1+AQ52+UXMCsrVGT6Y4kwmtPHfy31e63rys\nmq+9czVvO3Nu7FiiHki8quL8lNrjnIzjl0txlhRJltd35jmMvIKuKfVwoL2fAx39KU8AjXfFytpY\n+sp6vnz8wQi+QDhpDaSi0M1gKBobap6sBzIV4geYJPs9OerLC6zdG8NRjnT5aekZiNV5JmpxdRHP\n7eugPxBOWP9wrF9YkXS9tUxIZRRWE9YkQjVL+e0RPPFF9LHUlXl44OPn8xc/3cRt979OR1+Amy9c\nNCowDK0KO/nC9cmkttSDO8dFbQbH2ufmuEalk+rKCmjzHiMaNZzwBVK6Ck7E2Ro2fjb6vrZ+qkvy\nx1wDDOC9Z89n7/E+zhixn0ltST7P2RMN0wkgIzkbIbX3BRJuJuVwAoazB0uytk+F+Ml5yUZgOerL\nPRh7PbkX9lvrfMUHyolYXFXMRntNtLF6IAA/vPHstJ5/MsY9Y4hINfAXwML4xxtjPpq5ZqlscoaA\nproEdFlBHj/96AY+88DrfPU3u9nW3MtX37ma0rgRPBubuvHkuYbNbJ4NclxC45zCWM45W+aW2bOb\n+wPWUu5p7jhX7M5FZHgK63CXn4Vzkg+MaJxTyF0fWj/qeHwab80UXP06J+p27yC9AyFcAiUJ/l9W\nxAUQd65r2BymTMjPzaGq2E23P5TSaEEn5dXaM8ALBzqpLc2P1TMmylkTyyWkPVIxU1I5YzwM/Al4\nCq19zEr+oPWxTuSP0JOXw/ffv47T6/bz70/v47XD3Xz3hrWxHsemw12snV+e8qSymeQ771s77rpW\nU80p2O9t6yMcNRMaTh3P5RJKPcOXMzl8wseFSye2AoDDmQuyuKoo1ruZDGdocEd/ILaQYqKisDNs\nd39HP3OK3FlJjVqbPKWWKoxNJuwd4OUDJ7h4eXXabXQGoSypLs5Y2jRdqQSQQmPMZzPeEjVtJtoD\nceS4hFsvX8abllVx2/1beM//e4k5RW4aKgrY2erlE5ckn5sxU41M42SDk4Pfbs9DiC+iT1T8elgD\nwQht3sC4PZCx1Ng9kKlIX8HQLOp2b8BaxmSMeRPOUNWOvkDCpT0y4fZrVuBKMQg4PZBn9nRwwhfk\ngjTrHzDUA0lW/5guqVwePiYi10zli4rIN0Vkj4hsE5GHRKQ87r7bRWS/iOwVkSvjjl9lH9svIp+b\nyvac6vz2boTJRmEls66xgt9++iJuv3oFb11VS1lBHqfPLYktqa0mz+mBOFvOppvCguEBxNlhceSW\nqqlyUlgTmUCYTEVhHnk5EquBjDXxLr7mken6h+OCJVWctzi1BTiL8nMpK8iL7Yr4pqUTW7gzXmNl\nIYuri5LuWTJdUjlj3AZ8XkQCQAhrMqGZ5Ez0J4HbjTFhEfkGcDvwWRFZCdwArALqgadEZLn9Mz8A\nrgCagY0i8ogxZleC51YT1B9IbRRWMiWePP7y4tnZ4zgZVBXlk+uSuB7I5AKIU0R3VqpdMCe9/PwZ\nDWV87uoVXG8vhDhZIkJ1sTUbvWcgRNkYk+JKPXnkuIRI1GQtgEzU3DIPe473sbiqKOl+I+PJzXHx\nh7+9ZOoaNoXG7YEYY0qMMS5jTIExptS+Pam+lDHm98YYZ0nQlwFnQZrrgPuNMQFjzCFgP7DB/tpv\njDlojAkC96Mjw6bMREZhqenhcgm1pUN7ss+ZohSW83zpprByXMLHL14ybADFZDmz0b1JeiAul1Bh\n11xO1gDizJm5YBK9j5NdKqOw3pzouDHmuSlqw0eBX9jfN2AFFEezfQzg6Ijj8bPjY0TkFuAWgMbG\nxilq4uzm1ECK0kxhqeyYW+ahpWcAlzCp5SqsXQmtz7zphI9ST+6ENjrKtOoSD83dfnr8QcoKxv4/\nWVForWtVeRK1Pd5cu5D+pjSH784EqZwx/j7uew9Wb+A14LJkPyQiTwF1Ce66wxjzsP2YO4Aw8HPn\nxxI83pC4p5RwoUdjzF3AXQDr168fczFINcQZhVWU5ZFFamKcOkhlkXvCM5rjlRbk4rWXdD98wp92\n+ipTqkvy2Xyk29rBL8nqtc5Q3spJ9MYyaUl1Mfm5rpTrJjNRKhMJ3x5/W0Tmk8I6WMaYtyS7X0Ru\nAq4FLjdDmxM0A/ELJc0DnD0pxzquJskXCJOf6xpzgxx1cnBGYk2m/gFWCisYiTIYsmZJpzIxLptq\nSvJjiw8mW73W6XlkehZ6um48dwFvOb0248usTKd0zhjNwOpxH5WEiFwFfBZ4hzEmfuPtR4AbRCRf\nRBYBy4BXgY3AMhFZJCJurEL7I5NpgxriC4Y1fTUD1NmF2MnUP2DopHzCF6C5eyC2PMvJIn7Wd7K5\nJU7PI5urz06EO9eV8srVM1UqNZDvM5QucgFrga2TfN3/APKBJ+3JNS8bYz5ujNkpIg8Au7BSW580\nxkTsdnwKeALIAe4xxuycZBuUzTdiN0J1cprKHghYiyhGoubkS2HFvb+UeiAnaQrrVJDKZeemuO/D\nwH3GmBcm86LGmDE3pDbGfA34WoLjvwV+O5nXVYnF74euTl5ODWTOGHtfpMo5KTtDghPtLzGdauKW\nRxlrIiFYS8vn5QjVWdwDXA2XSg3kXjtt5MzH2JvZJqls8wXD2gOZAWI9kJKpSWHFAsjJ1gMpSS2F\n9a51DaxfUDHhXf7U1EklhXUJcC/QhDVKar6I3DSFw3jVNPMFIpR4tAdysqsr9fCld6ziylWJBjem\nzjnhbmvuxZPnyvoeEuOJT2ElG4WVn5sT241RTY9UzhrfBt5qjNkLYM8Mvw/I/trBKiN8gXBGtmdV\nU0tEuOmChZN+HieAdPmCLK8tzuoOdqlw57qoKMyj2z/2REJ1ckhlFFaeEzwAjDFvAPqpziL+YERH\nYZ1CSuJmjcfvsHgyqS7Jx53rwpOnQ8tPZikV0UXkbuBn9u0bsSYSqlnCGsarNZBTRY5LKMnPpS8Q\nTnsJk0yrKfHQ7Q/N+B0sZ7tUAshfAZ8EPo1VA3kO+GEmG6WyyxfQeSCnmtKCPPoC4ZNuDojj3EWV\nmr6aAZKeNUQkB7jbGPMB4M7sNEllUzAcJRQxuozJKaasII+WngEaT7IRWI5bL1823U1QKUiaYLQn\n8VXbw3jVLOTThRRPSc7V/cmawlIzQypnjSbgBRF5BPA5B40x2iOZBXz2ZlJFupT7KaWswNpPo748\n/X0qlErlrNFqf7kAHXQ9y/jszaS0B3JqWVVfSs9AkDxdQFNNQioz0b+UjYao6eGLbWerNZBTya2X\nL9M6g5o0vfw4xTk1EF0LSyk1URpATnFOCkvXwlJKTZQGkFNcbBSWFtGVUhM05llDRHKBm4F3AvVY\ne4K0Ag9jzQ0JZaWFKqP8QR3Gq5RKT7Kzxs+AHuCLWLsQgrWV7E3A/wB/ltGWqazoj43C0hSWUmpi\nkgWQdcaY00YcawZeFpE3MtgmlUX+YBgRKMjTAKKUmphkNZBuEXmviMQeIyIuEfkzoDvzTVPZ0B8I\nU+TO1UXrlFITliyA3AC8B2gTkTdEZB/QBrzLvk/NAv5ARNNXSqm0jJnCMsY0Ydc5RGQOIMaYziy1\nS2WJLxjWEVhKqbSMtxrvCuA6oAEwItIKPGyM2ZONxqnM8wcjOgtdKZWWMVNYIvJZ4H6sPUBeBTba\n398vIp/LTvNUpvkCYQrztAeilJq4ZGeOm4FVI+d7iMidwE7g65lsmMqOgVCEikJdrV8pNXHJiuhR\nrAmEI82171OzgD8Y0WVMlFJpSdYD+WvgaXv01VH7WCOwFPhUphumsmMgGKFQi+hKqTQkG4X1OxFZ\nDmzAKqIL1kTCjfZOhWoW8AfD2gNRSqVlvEtPE/cVjftXzRI+TWEppdKUbDHFtwI/BPYBLfbhecBS\nEfmEMeb3WWifyqBI1BAMRynQAKKUSkOyHsh3gbfYEwpjRGQR8Fvg9HRfVES+gjW/JAq0Ax82xrSK\ntZ7Gd4FrAL99fLP9MzcB/2g/xVeNMfem+/rK4tf90JVSk5BsFFYuQ6vwxmsB8ib5ut80xpxpjFkL\nPAb8s338amCZ/XUL8CMAEakEvgCci1WT+YKIVEyyDae8gaBVytIeiFIqHckuPe8BNorI/QyNwpqP\ntQ7W3ZN5UWOMN+5mEVZtBaxeyU+NMQZr1d9yEZkLXAI8aYzpAhCRJ4GrgPsm045TnS+ouxEqpdKX\nbBTWv4rIr7FO6uczNArrRmPMrsm+sIh8DfgQ0Atcah9uYChYYb9eQ5LjiZ73FqzeC42NjZNt5qzm\npLA0gCil0pE0+W2M2Q3sTueJReQpoC7BXXcYYx42xtwB3CEit2PNK/kCVpAa1YwkxxO1+S7gLoD1\n69cnfIyyDMR6IFoDUUpNXFpnDhF53BhzdbLHGGPekuLT/S/wG6wA0oyVJnPMw9pGtxkrjRV//NkU\nn1+Nwa8pLKXUJCQbxrturLuAtZN5URFZZozZZ998B+Cs7vsI8Cm77nIu0GuMOSYiTwD/Elc4fytw\n+2TaoIZSWFpEV0qlI1kPZCPwRxKnj8on+bpfF5HTsIbxHgY+bh//LdYQ3v1Yw3g/AmCM6bKH/m60\nH/dlp6Cu0ufXFJZSahKSnTl2A38Z11OIEZGjCR6fMmPMu8c4boBPjnHfPVgjw9QU0RSWUmoyks0D\n+WKS+2+d+qaobBvQAKKUmoRkw3h/leS+X2emOSqbNIWllJqMZD0QNcv5g2HcuS5yXInKXEoplZwG\nkFOYbiallJqMpAFERFwickG2GqOyyx+M6EKKSqm0JQ0gxpgo8O0stUVl2UAorHNAlFJpSyWF9XsR\nebe91LqaRXwBTWEppdKXSv7iM1gr5kZEZABrYqExxpRmtGUq4waCEQryNIAopdIzbgAxxpRkoyEq\n+/yhMDUlnuluhlJqhho3hSWWD4jIP9m354vIhsw3TU3W7mNe2vsGx7zfH4xoDUQplbZUaiA/xNoP\n5M/t2/3ADzLWIjVlPnTPq9z5+zfGvH8gGKFQU1hKqTSlUgM51xizTkS2ABhjukXEneF2qUnq9Yfo\n6AvQ0jMw5mN8gbAW0ZVSaUulBxISkRzsDZxEpBprFV11Ems64QPgeO/YKayBUITCfJ0HopRKTyoB\n5HvAQ0CNvQ3t88C/ZrRVatKcANLmTRxAQpEooYjRFJZSKm3jBhBjzM+Bf8AKGseA640xD2S6YWps\nwXCUT9+3hf3tfWM+pqnTD4B3MBxbdTees5CiFtGVUulKZRTWz4wxe4wxPzDG/IcxZreI/CwbjVOJ\nHe3288jWVh7demzMxzg9ECDhSCxnN0JdiVcpla5UUlir4m/Y9ZCzM9MclQrvQAiAna29Yz6m6YQP\nd4718Saqg+hmUkqpyRozgIjI7SLSB5wpIl77qw9oBx7OWgvVKN5Bq/ews9U75mOaOn2smV8GQFtf\nYNT9upmUUmqykvVAnrNnoX/LGFNqf5UYY+YYY27PVgPVaE4P5FjvIF2+4Kj7e/0huv0hzl00B4D2\nBIV03UxKKTVZyQLI9+x/35qNhqjUeQdDse8TpbGc+seZ88rw5LkSjsTy2TUQLaIrpdKV7PIzJCI/\nBhpE5Hsj7zTGfDpzzVLJeAfCse93tnq5aFn1sPudALKoqojaUg/HvZrCUkpNvWQB5FrgLcBlwGvZ\naY5KhXcwRF6OUFPiYUdLgh5Ipx8RmF9ZSG2JJ2EPxElh6YZSSql0jXn2MMZ0AveLyG5jzNYstkmN\nwzsQotSTx6r6UnYlKKQ3nfBRX1aAJy+HmtL8hEFmQFNYSqlJSuXyc6+IfBJrOG9s7W9jzEcz1iqV\nlHcwTGlBHqvqy3hydxu+QJiiuCVJmk74WDCnEIC6Ug9P727HGEP8nmA6jFcpNVmpzAP5GVAHXAn8\nEZgHjD0FWmWc1QPJZXVDKcZYy7bHa+r0sbCqCIDaUg8DoQh9gfCwx/icmei6lIlSKk2pBJClxph/\nAnzGmHuBtwFnZLZZKhnvYCjWA4Hh80GcIbwL7R5ITWk+AG0jJhMOBMMU5OXgculOxUqp9KS0Gq/9\nb4+IrAbKgIUZa5Eal1MDqS3NZ06Re9hQXmcE1sI5Vg+krtTKOraNGInlD+p+6EqpyUklgNwlIhXA\nPwKPALuAb0zFi4vI34mIEZEq+7aIyPdEZL+IbBORdXGPvUlE9tlfN03F689UVg0kFxFhZX0pO1qG\neiCxABKXwoLRq/IO6G6ESqlJSmVP9P+2v30OWDxVLywi84ErgCNxh68Gltlf5wI/As4VkUrgC8B6\nrH1JXhORR4wx3VPVnpnE6YEArKov4+7nDxIMR3HnumJDeBsrR6SwRiyo6AvqZlJKqclJpQeSKd/B\nWibexB27DvipsbwMlIvIXKwC/pPGmC47aDwJXJX1Fp8EBkMRAuEopQVOACklFDG80WaNa2g64WNu\nqQePXRwvdOdS4skdVQOxUlg6B0Qplb5pCSAi8g6gJcH8kgbgaNztZvvYWMdPOX32QoqlHuvkv7rB\nKqR/9+l9PLOnnf3t/bH0laOu1DOqBjKgNRCl1CSNewkqIvnGmMB4xxL83FNYw39HugP4PInX2Eo0\nJMgkOZ7odW8BbgFobGxM1sQZyVkHy+mBLKgs5N3r5vH4jmM8uasNgPdvGP6+a0s9o1JY/mCE8sK8\nLLRYKTVbpZLDeAlYl8KxYYwxb0l0XETOABYBW+2JbfOAzSKyAatnMT/u4fOAVvv4JSOOPzvG694F\n3AWwfv36hEFmJnNW4nVqIC6X8O33reFr71zNK4e6eHF/J9etHd45qynN55WDvmHHBkIRCjSFpZSa\nhDHPICJSh5UmKhCRsxjqBZQChem+oDFmO1AT9zpNwHpjTKeIPAJ8SkTuxyqi9xpjjonIE8C/2KPB\nwOq9nJJLyjt7gZQWDP/oPHk5XLy8mouXV4/6GSuFNUg0amLzPnyBsO6HrpSalGSXoFcCH8a62r8z\n7ngfVgoqE34LXAPsB/zARwCMMV0i8hVgo/24LxtjujLUhpOa0wMpK0g9/VRb6iEcNXT5g1QVW6Oy\nBoIRCvM1gCil0pdsMcV7gXtF5N3GmP/LVAOMMQvjvjfAJ8d43D3APZlqx0wRq4F4JhJA7KG83kGq\nivMxxuAPaRFdKTU5qSTBV4vIqpEHjTFfzkB71Dh6B4YX0VNRY08mbPcGWFUPwUiUSNToMF6l1KSk\ncgbpj/veg7VPyO7MNEeNxzsQxp3jIj839RHYdSNmo/sDupCiUmryUpmJ/u342yLyLawlTdQ0sBZS\nzB22NPt4qkusFNZxJ4CE7M2ktAailJqEdCYSFjKFS5qoiYlfxiRVeTkuqordscmEQ5tJaQpLKZW+\nVCYSbmdo0l4OUA1o/WOaeAfDlEyg/uFoqCjkUKeVjYxtJqUpLKXUJKRyCXpt3PdhoM0YEx7rwSqz\nnM2kJursxgp+/sphAuEIvoDuRqiUmrxxU1jGmMNAOfB24J3Aykw3So3N2Uxqos5dXEkgHGVbcy8D\nISv+F+ZrCksplb5xA4iI3Ab8HGv2eA3wcxG5NdMNU4l5B8ITroEAbFhYCcArB0/ofuhKqSmRyiXo\nzcC5xhgfgIh8A2strO9nsmEqMWcU1kRVFLlZUVfCK4e6YvNCdBivUmoyUhmFJUAk7naExKvjqgwb\nDEUIhqNp9UAAzl1Uyaam7thyKNoDUUpNRioB5MfAKyLyRRH5IvAycHdGW6USGrmU+0RtWDSHgVCE\nVw5Zy4jpTHSl1GSkMpHwThF5FrgQq+fxEWPMlkw3TI3mHRi+mdREbVhk1UFe2N+JCHjypnNDSqXU\nTJfKPJBvAT82xnwvC+1RSUy2B1Jdks+S6iIOdPgocudMaDa7UkqNlMol6B7gLhF5RUQ+LiJlmW6U\nSmzkZlLpOHfxHEBnoSulJi+VeSD/bYx5E/AhYCGwTUT+V0QuzXTj1HDOZlJlaYzCcpxrp7G0gK6U\nmqyUkuAikgOssL86ga3AZ+ydA1WWTEUP5Dy7B6IBRCk1WalMJLwT2Iu1U+C/GGPONsZ8wxjzduCs\nTDdwtvIOhnjtcPeEfwbSr4GAtTvhwjmFFOksdKXUJKVyFtkB/KMxxp/gvg1T3J5Txs9eOsw3n9jL\nl96xipsuWJjSz6SzF0giX7l+9aR+XimlILVhvPeISIWIrMbaUMo5/pwxpjejrZvFOvqspdW/8MhO\nXC7hg+ctGPdn0tkLJJGLllVP6ueVUgpSG8b7MeA2YB7wOnAe1lIml2W2abNbtz9IQ3kBp88t4Z9+\nvQOM4cZzF+ByjR0c0tkLRCmlMiWVXMhtwDnAYWPMpVh1j46MtuoU0OMPUVXs5gc3ruPS06r5p4d3\nsuFfnuLvf7mVP+1L/OtNdy8QpZTKhFQCyKAxZhBARPKNMXuA0zLbrNmvxx+kvNBNfm4O//nB9Xz3\nhrVcsKSKJ3Ye54N3v8qe495RP+MdCFGmAUQpdZJIJYA0i0g58GvgSRF5GGjNbLNmv56BEOWFVjBw\n57q4bm0D33v/Wfz2tosAeNVeryqedzC9zaSUUioTUplI+E5jTI8x5ovAP2EtpHh9phs223X7glQU\nukcdbygvoKYkn80Jhvh6B8KTGsKrlFJTaczLWRHxAB8HlgLbgbuNMX/MVsNms0jU4B0MJ0xHiQjr\nGivYfKRn1H1WD0QDiFLq5JCsB3IvsB4reFwNfDsrLToF9NozyisKEweDdQvKOdLlp7M/EDsW2wtk\nEsuYKKXUVEp2NlppjDkDQETuBl7NTpNmv25/EIDyBCksgHWNFQBsPtzNW1fVAXGz0LUHopQ6SSTr\ngYScb4wx4Sy05ZTR47d+teVj9EBWN5SRlyPD0lixdbC0BqKUOkkkCyBrRMRrf/UBZzrfi8joMaYT\nYO9u2CIir9tf18Tdd7uI7BeRvSJyZdzxq+xj+0Xkc5N5/enWM04PxJOXw8r6MjYfGSqkP7vXmhuy\nvLY48w1USqkUjJnCMsZkernW7xhjvhV/QERWAjcAq4B64CkRWW7f/QPgCqAZ2CgijxhjdmW4jRnh\n9EDGqoEArGss575XjxCKRHGJcO9LTWxYWMmKutIstVIppZI72fY0vQ643xgTMMYcAvZjLdi4Adhv\njDlojAkC99uPnZFiNZCCxD0QsOogg6Eoe4718Yc97RztGuDDb1qYpRYqpdT4pjOAfEpEtonIPSJS\nYR9rAI7GPabZPjbW8RmpdyCES6AkyaTAdQvsQvqRbn7y4iHmlnl468rabDVRKaXGlbEAIiJPiciO\nBF/XAT8ClgBrgWMMDRFOtJKgSXI80eveIiKbRGRTR8fJuWRXtz9IWUFe0oUT68s81Jbm86vXmnlh\n/wk+cN4CcnNOtg6jUupUlrFJBcaYt6TyOBH5L+Ax+2YzMD/u7nkMLZsy1vGRr3sXcBfA+vXrEwaZ\n6dbjDyWchR7PmVD4+I7juHNdvH9DY5Zap5RSqZmWS1oRmRt3851Ym1YBPALcICL5IrIIWIY1/2Qj\nsExEFomIG6vQ/kg22zyVevwhypIU0B3OfJDr19ZTWZQ84CilVLZN17TmfxORtVhpqCbgLwGMMTtF\n5AFgFxAGPmmMiQCIyKeAJ4Ac4B5jzM7paPhU6BkIUlPiGfdxl66o4ScvNvGxixZnoVVKKTUx0xJA\njDEfTHLf14CvJTj+W+C3mWxXtnT7QiyvKRn3cUtrinnhc7pvl1Lq5KRV2WnQOxAacxKhUkrNFBpA\nsiwYjtIfCI+5jIlSSs0UGkCyrGfAmkSYbBa6UkrNBBpAsqw3tpCiprCUUjObBpAs6x5nJV6llJop\nNIBkmbMS73gTCZVS6mSnASTLnJV4E21nq5RSM4kGkCyLFdF1ZrlSaobTAJJl3f4QuS6hyJ3p7VaU\nUiqzNIBMkb3H+2jq9I37uB6/NYlQZOyVeJVSaibQADIGXyD1beB3H/Pyrh++wO0Pbh/3sT3+oI7A\nUkrNChpAEujoC3DxN5/l27/fSyAcGfexH7t3E75ghB2tvRiTfAV5ayl3DSBKqZlPA0gC7hwXFy+v\n5vt/2M+133ueLUe6Ez5uMBTh4//zGid8Ad6/oZG+wTDN3QNJn9vaTEoL6EqpmW+6lnM/qZUV5vHt\n963h2jVz+fyD23n3j17kktNquPz0Gi5bUYMvEOHVQ108urWV1w5388Mb11FX5uG+V4+w+5iX+ZWF\nYz5370CIMxq0B6KUmvk0gCRx6Wk1PPE3b+YHf9jPb7Yf4w972ofdX1Xs5gtvX8k1Z8zFHwwjAruO\neXnrqroxn7NbayBKqeLy6+4AAAskSURBVFlCA8g4Sj153H7N6Xzu6hXsa+/nj3s7KC3I5ZyFlSyq\nKoqNpip057JoThG7j3nHfK7BUITBUFTXwVJKzQoaQFIkIiyvLWF57dgbQZ1eX8q25p4x7+/RdbCU\nUrOIFtGn0Mq5pRztGsA7GEp4f7eug6WUmkU0gEyhlXNLAdhzrC/h/bEeiK6DpZSaBTSATKHT7QCy\nq7U34f3OSrxaA1FKzQYaQKZQbWk+lUVudsf1QNq9g7x+tAdjDD0DWgNRSs0eWkSfQiLC6XNL2GWP\nxIpGDTffu4ntLb2cvaCC6uJ8QGsgSqnZQXsgU2zl3FL2tvURjkR5eGsL21t6ee/Z8zjeO8jvdh7H\nnevCk6e/dqXUzKc9kCl2+txSguEou455+ebv9nLmvDK+8e4ziRjDw6+3EolGdSVepdSsoAFkiq2s\ntwrpn39oO629g9z5Z2txuQQXwnvOnjfNrVNKqamjuZQptqS6GHeOix0tXq5YWct5i+dMd5OUUioj\nNIBMsbwcF8tqi8l1CbdfvWK6m6OUUhmjKawM+PTly+j1h1hcXTzdTVFKqYyZth6IiNwqIntFZKeI\n/Fvc8dtFZL9935Vxx6+yj+0Xkc9NT6tTc+WqOt53zvzpboZSSmXUtPRARORS4DrgTGNMQERq7OMr\ngRuAVUA98JSILLd/7AfAFUAzsFFEHjHG7Mp+65VSSsH0pbD+Cvi6MSYAYIxxNtq4DrjfPn5IRPYD\nG+z79htjDgKIyP32YzWAKKXUNJmuFNZy4CIReUVE/igi59jHG4CjcY9rto+NdVwppdQ0yVgPRESe\nAhJtzXeH/boVwHnAOcADIrIYSDTDzpA40JkxXvcW4BaAxsbGiTdcKaVUSjIWQIwxbxnrPhH5K+BB\nY4wBXhWRKFCF1bOIrz7PA1rt78c6PvJ17wLuAli/fn3CIKOUUmrypiuF9WvgMgC7SO4GOoFHgBtE\nJF9EFgHLgFeBjcAyEVkkIm6sQvsj09JypZRSwPQV0e8B7hGRHUAQuMnujewUkQewiuNh4JPGmAiA\niHwKeALIAe4xxuycnqYrpZQCEOu8PTutX7/ebNq0abqboZRSM4qIvGaMWT/u42ZzABGRDuDwJJ6i\nCiu1dio5Fd8znJrv+1R8z3Bqvu+JvucFxpjq8R40qwPIZInIplSi8GxyKr5nODXf96n4nuHUfN+Z\nes+6mKJSSqm0aABRSimVFg0gyd013Q2YBqfie4ZT832fiu8ZTs33nZH3rDUQpZRSadEeiFJKqbRo\nAElgJu09MhkiMl9EnhGR3fa+LLfZxytF5EkR2Wf/WzHdbZ1qIpIjIltE5DH79iJ7cc99IvILe8WD\nWUVEykXkVyKyx/7Mz5/tn7WI/I39f3uHiNwnIp7Z+FmLyD0i0m5PznaOJfxsxfI9+/y2TUTWpfu6\nGkBGEJEcrL1HrgZWAu+39ymZjcLA3xpjTsda2PKT9nv9HPC0MWYZ8LR9e7a5Ddgdd/sbwHfs99wN\n3Dwtrcqs7wK/M8asANZgvf9Z+1mLSAPwaWC9MWY11ioWNzA7P+ufAFeNODbWZ3s11jJRy7AWnv1R\nui+qAWS0Ddh7jxhjgoCz98isY4w5ZozZbH/fh3VCacB6v/faD7sXuH56WpgZIjIPeBvw3/ZtwVqb\n7Vf2Q2bjey4F3gzcDWCMCRpjepjlnzXWck0FIpILFALHmIWftTHmOaBrxOGxPtvrgJ8ay8tAuYjM\nTed1NYCMdkruPSIiC4GzgFeAWmPMMbCCDFAzfS3LiH8H/gGI2rfnAD3GmLB9ezZ+5ouBDuDHduru\nv0WkiFn8WRtjWoBvAUewAkcv8Bqz/7N2jPXZTtk5TgPIaGPtSTJriUgx8H/AXxtjvNPdnkwSkWuB\ndmPMa/GHEzx0tn3mucA64EfGmLMAH7MoXZWInfO/DliEtUV2EVb6ZqTZ9lmPZ8r+v2sAGS3ZniSz\njojkYQWPnxtjHrQPtzldWvvf9rF+fgZ6E/AOEWnCSk9ehtUjKbfTHDA7P/NmoNkY84p9+1dYAWU2\nf9ZvAQ4ZYzqMMSHgQeACZv9n7Rjrs52yc5wGkNFOmb1H7Nz/3cBuY8ydcXc9Atxkf38T8HC225Yp\nxpjbjTHzjDELsT7bPxhjbgSeAd5jP2xWvWcAY8xx4KiInGYfuhxr24RZ+1ljpa7OE5FC+/+6855n\n9WcdZ6zP9hHgQ/ZorPOAXifVNVE6kTABEbkG66rU2Xvka9PcpIwQkQuBPwHbGaoHfB6rDvIA0Ij1\nR/heY8zIAt2MJyKXAH9njLnW3lL5fqAS2AJ8wBgTmM72TTURWYs1cMANHAQ+gnUROWs/axH5EvBn\nWCMOtwAfw8r3z6rPWkTuAy7BWnW3DfgC1sZ9oz5bO5j+B9aoLT/wEWNMWvteaABRSimVFk1hKaWU\nSosGEKWUUmnRAKKUUiotGkCUUkqlRQOIUkqptGgAUbOOiPz/9s4vxKoqisPfLzBNDOtBekhKmJR5\nsbmCMihGAxFiZIhgPYiNlVBPgjAEEcj0FCraQ1IQPVgiUgaJBUk0MTja5NCfO16LDESFwpekXsJE\nZlYPa93meLv3qkdj4ro+2Jx19zl778UdZq+zzrn7t1+X1Cdp7Y2qKUuaF0qt30t6pOHck1E/LulH\nSS/eWs//5cugpIGb7GOBpEuSquHz+7F4NElumgwgSSfSi69leRRf53IjPAb8ZGZLzOyftjHpvgOs\nMbMeXDds+Na4+59zxswqwGJ81fHT0+xP0iFkAEk6Bkk7JZ0ElgGj+KKxtyVta3Ltg5KGYj+EIUkP\nxEK7HcATccd+V6HJ3bie1EUAM7tsZqejrzWFrOULSfdF/aCk9yR9LumcpHWSdkiqSTpSzwTi3HZJ\nY1EeauJvV7T5VtKIpO6oXy/f62Jc0tF234+ZTQBjhHBeZCcjkr6LsiLq+yQNa2rvkP2x+CxJrsbM\nsmTpmILL8b8JzACOt7nuE6A/7OeBQ2FvAva0aPMurid0ANgA3BH19zK1KHczsCvsQeBY+NKDr/pd\nHec+BtaGfQ54NexngU8L7QfCHgIWht2LS7CAqwjcH/Y9TXxeAJwKexYu4/FwfJ4NzAp7IfBN2H24\ncu18/CZzFFg53X/bLP+/UhcUS5JOYQlQBbpx3aNWLAfWhb0PzzzaYmabJS3GRfoGgMfxgDMf+CAE\n6+4EzhaafWZmVyTVcGmcI1Ffwyf3OgcKxzeK44Za8grgYCERmBnH48BeSR/iYoHN6JJUxYPER2Z2\nMupnAHsi85oAFhXajJnZLzF+NXw91qL/5DYlA0jSEcQkuBefzH/D764Vk99yM7t0jS6uS9PHzGpA\nTdI+PFBswjOe3WZ2OPS1BgtNLke7SUlXzKw+ziRX//9ZCxs8C/jD/D1Goz8vSerFN8iqSqqY2cWG\ny86YWSUC3LCkp8zsMLAV103qiTH+avQ7mCDniqQJ+Q4k6QjMrBoT7M/4VsRfAqvMrNIieHyFq/GC\nP45qe3ctaU4EhzoV4HzYc4Ffw+6nHM8UjqPFE+Z7tJyVtD58kaSesLvM7ISZbcMDZ1Gmm4Z+LuB7\ngLxS8PuCmU0CG/EMKUmumwwgSccgaR7we0yI3WbW7hHWFuC5eOm+Ed8jvW33wMuSTkdW8xqefYBn\nHAcljeCTeBlmSjoRfmxtcn4D8IKkceAHprZZ3hkv5U8BR4Hxa4xzCJgdP1F+C+iX9DX++OrPkr4n\ntympxpsk04x8c6ulZlY2+CTJtJAZSJIkSVKKzECSJEmSUmQGkiRJkpQiA0iSJElSigwgSZIkSSky\ngCRJkiSlyACSJEmSlCIDSJIkSVKKvwFsClmExO7TgQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x109fd2908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average payout after 1000 rounds is -214.85\n"
     ]
    }
   ],
   "source": [
    "from blackjackenv import BlackjackEnv\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "KEY = {0: 'STAND', 1:'HIT', 2:'DOUBLE DOWN'}\n",
    "\n",
    "env = BlackjackEnv()\n",
    "env.reset()\n",
    "\n",
    "# env = wrappers.Monitor(env, './logs/blackjack-Q', False, True)\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "num_rounds = 1000 # Payout calculated over every episode\n",
    "num_episodes = 100 # num of episodes the simulation will run\n",
    "\n",
    "agent = Agent(env=env, epsilon=1.0, alpha=0.01, gamma=0.1, num_episodes_to_train = 25000)\n",
    "\n",
    "total_payout = 0\n",
    "average_payouts = []\n",
    "\n",
    "# PERCENTAGE OF ACTIONS TAKEN\n",
    "ACTIONSD = {0: 0, 1:0, 2:0}\n",
    "\n",
    "observation = env.reset()\n",
    "for sample in range(num_episodes):\n",
    "    round = 1\n",
    "    round_payout = 0 # to store total payout over 'num_rounds'\n",
    "    # Take action based on Q-table of the agent and learn based on that until 'num_episodes_to_train' = 0\n",
    "    while round <= num_rounds:\n",
    "#         print(env.decks)\n",
    "\n",
    "        action = agent.choose_action(observation)\n",
    "\n",
    "#         print(action)\n",
    "\n",
    "        next_observation, payout, is_done, _, ddown = env.step(action)\n",
    "#         print('OBS: ' + str(observation))\n",
    "#         print('ACTION: ' + str(KEY[action]))\n",
    "# #         print('DEALER: ' + str(dealer_up_card))\n",
    "#         print('OBS2: ' + str(next_observation))\n",
    "#         print('Payout: ' + str(payout))\n",
    "\n",
    "#         print('------')\n",
    "#         if num_episodes > 20:\n",
    "        ACTIONSD[action] += 1\n",
    "        agent.learn(observation, action, payout, next_observation)\n",
    "        round_payout += payout\n",
    "        observation = next_observation\n",
    "        if is_done:\n",
    "            observation = env.reset() # Environment deals new cards to player and dealer\n",
    "            round += 1\n",
    "    total_payout += round_payout\n",
    "    average_payouts.append(round_payout)\n",
    "# print(agent.Q)\n",
    "# for x in agent.Q:\n",
    "# #     if agent.Q[x][2] > 0:\n",
    "# #         print(x, agent.Q[x])\n",
    "#     if not x[2]:\n",
    "#         print(x, agent.Q[x])\n",
    "print(ACTIONSD)\n",
    "\n",
    "\n",
    "# Plot payout per 1000 episodes for each value of 'sample'\n",
    "plt.plot(average_payouts)                \n",
    "plt.xlabel('# of Samples Ran')\n",
    "plt.ylabel('Payout after ' + str(num_rounds) + ' rounds')\n",
    "plt.show()     \n",
    "print (\"Average payout after {} rounds is {}\".format(num_rounds, total_payout/num_episodes))\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}