examples/quickstart_ja/rtb/rtb_synthetic_discrete_basic_ja.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Quickstart: オフライン強化学習とオフ方策評価 (離散行動空間)\n",
    "このノートブックでは 入札オークション (Real-Time Bidding; RTB) における人工データセットでのオフ方策学習 (OPE) とオフライン強化学習 (Offline RL) の実行例を紹介します。\n",
    "\n",
    "特に、以下の5つの順番で解説していきます: \n",
    "\n",
    "0. 入札オークションのシミュレーション環境の設定とオンライン強化学習\n",
    "1. 人工データの生成\n",
    "2. オフライン強化学習\n",
    "3. オフ方策評価とオフライン強化学習手法の評価\n",
    "4. オフ方策評価手法の評価\n",
    "\n",
    "\\* このライブラリは，オンラインとオフラインの方策・モデルベース評価で [d3rlpy](https://github.com/takuseno/d3rlpy)のアルゴリズムを利用しています.   \n",
    "\\*また[Open Bandit Pipeline](https://github.com/st-tech/zr-obp)を参考に実装しています."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SCOPE-RL モジュールのインポート\n",
    "import scope_rl\n",
    "from rtbgym import RTBEnv, CustomizedRTBEnv\n",
    "from scope_rl.dataset import SyntheticDataset\n",
    "from scope_rl.policy import OnlineHead\n",
    "from scope_rl.ope.online import (\n",
    "    calc_on_policy_policy_value,\n",
    "    visualize_on_policy_policy_value,\n",
    ")\n",
    "\n",
    "# d3rlpy アルゴリズムのインポート\n",
    "from d3rlpy.algos import DiscreteRandomPolicyConfig\n",
    "from d3rlpy.algos import RandomPolicyConfig as ContinuousRandomPolicyConfig\n",
    "from d3rlpy.preprocessing import MinMaxObservationScaler, MinMaxActionScaler\n",
    "\n",
    "# その他のライブラリのインポート\n",
    "import gym\n",
    "import torch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "from glob import glob\n",
    "from tqdm import tqdm\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.2.1\n"
     ]
    }
   ],
   "source": [
    "# バージョン\n",
    "print(scope_rl.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# random stateの設定\n",
    "random_state = 12345"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = \"cuda:0\" if torch.cuda.is_available() else \"cpu\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# logディレクトリ\n",
    "from pathlib import Path\n",
    "Path(\"logs/\").mkdir(exist_ok=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# warnings\n",
    "import warnings\n",
    "warnings.simplefilter('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 0. 入札オークションのシミュレーション環境の設定とオンライン強化学習\n",
    "まず初めにRTB環境を紹介します．\n",
    "\n",
    "#### 強化学習でのRTB環境のセットアップ\n",
    "RTBにおいて強化学習エージェントの目的は，予算制約が与えられもとで，一つのエピソード内でのKPI(クリックやコンバージョン)を最大化することです. KPIを最大にするために，エージェントは入札価格関数パラメータ $\\alpha$ を選択します．\n",
    "入札価格関数パラメータ $\\alpha$ を利用することで以下の入札価格を決定します．\\\n",
    "$bid_{t，i} = \\alpha \\cdot r^{\\ast}$，\n",
    "ここで$r^{\\ast}$はKPIの予測値または期待値を表します.\n",
    "\n",
    "この問題を制約付きマルコフ決定過程(CMDP)として定式化します.\n",
    "\n",
    "- `タイムステップ`: 一つのエピソード(一日，一週間)に含まれるタイムステップ(24時間，数日など).\n",
    "- `状態`: それぞれのタイムステップで，環境からフィードバックが返ってきます.\n",
    "  - タイムステップ\n",
    "  - 残予算\n",
    "  - 一つ前のタイムステップでのインプレッション単位の特徴(予算消費率，一つのインプレッションあたりのコスト，オークション落札率，報酬)\n",
    "  - 一つ前のタイムステップでの入札関数パラメータ(強化学習エージェントが選ぶ行動) \n",
    "- `行動`: エージェントはKPIを最大化するために入札価格パラメータ$\\alpha$を選択します.\n",
    "- `報酬`: 一つのタイムステップあたりに得られたクリックまたはコンバージョンの合計数.\n",
    "- `制約`: あらかじめ決められた、一つのエピソードあたりの予算．\n",
    "\n",
    "より詳細に環境の引数を確認したい場合は，次のノートブックを参照してください．[examples/quickstart_ja/rtb/rtb_synthetic_customize_env_ja.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart_ja/rtb/rtb_synthetic_customize_env_ja.ipynb).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 環境のセットアップ\n",
    "env = RTBEnv(random_state=random_state)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ランダムに行動を選択するエージェントを定義\n",
    "agent = OnlineHead(\n",
    "    ContinuousRandomPolicyConfig(\n",
    "        action_scaler=MinMaxActionScaler(\n",
    "            minimum=0.1,  # 方策が取りうる最小の値\n",
    "            maximum=10,  # 方策が取りうる最大の値\n",
    "        )\n",
    "    ).create(device=device),\n",
    "    name=\"random\",\n",
    ")\n",
    "agent.build_with_env(env)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 環境とエージェントの相互作用\n",
    "# 6行で強化学習の相互作用を記述できる\n",
    "for episode in range(10):\n",
    "    obs, info = env.reset()\n",
    "    done = False\n",
    "\n",
    "    while not done:\n",
    "        action = agent.predict_online(obs)\n",
    "        obs, reward, done, truncated, info = env.step(action)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['timestep', 'remaining_budget', 'budget_consumption_rate', 'cost_per_mille_of_impression', 'winning_rate', 'reward', 'adjust_rate']\n",
      "(7,)\n"
     ]
    }
   ],
   "source": [
    "# 環境からのフィードバック\n",
    "print(env.obs_keys)\n",
    "print(obs.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 累積報酬と残予算の変遷を可視化しています\n",
    "# 目的は予算制約の中で強化学習を利用して, 累積報酬を最大化する方策を得ることです\n",
    "obs, info = env.reset()\n",
    "done = False\n",
    "remaining_budget = [obs[1]]\n",
    "cumulative_reward = [0]\n",
    "\n",
    "while not done:\n",
    "    action = agent.predict_online(obs)\n",
    "    obs, reward, done, truncated, info = env.step(action)\n",
    "    remaining_budget.append(obs[1])\n",
    "    cumulative_reward.append(cumulative_reward[-1] + reward)\n",
    "\n",
    "# プロット\n",
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "ax1.plot(remaining_budget[:-1], label='remaining budget')\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(cumulative_reward[:-1], label='cumulative reward', color='tab:orange')\n",
    "ax1.set_xlabel('timestep')\n",
    "ax1.set_ylabel('remainig budget')\n",
    "ax1.set_ylim(0, env.initial_budget + 100)\n",
    "ax2.set_ylabel('reward (coversion)')\n",
    "ax1.legend(loc='upper left')\n",
    "ax2.legend(loc='upper right')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "上記の例では，ランダム方策が早く予算を消費してしまっており，より多くの報酬が得られるかもしれない後半の時刻に予算が残っていません．そのため，予算制約を満たしつつも報酬を最大化するよう 入札価格パラメータ(行動) を選択する方策を得ることがここでの目標です．"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### オンライン強化学習と方策の比較\n",
    "まずは方策価値のオンラインでの推定に基づき，オンライン強化学習で学習した方策とランダム方策を比較します．"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.1         0.16681005  0.27825594  0.46415888  0.77426368  1.29154967\n",
      "  2.15443469  3.59381366  5.9948425  10.        ]\n"
     ]
    }
   ],
   "source": [
    "# SCOPE-RLで実装されている離散行動空間に対する標準的な環境\n",
    "env = gym.make(\"RTBEnv-discrete-v0\")\n",
    "print(env.action_meaning)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from d3rlpy.algos import DoubleDQNConfig\n",
    "from d3rlpy.models.encoders import VectorEncoderFactory\n",
    "from d3rlpy.models.q_functions import MeanQFunctionFactory\n",
    "from d3rlpy.dataset import create_fifo_replay_buffer\n",
    "from d3rlpy.algos import LinearDecayEpsilonGreedy\n",
    "\n",
    "# モデル\n",
    "ddqn = DoubleDQNConfig(\n",
    "    encoder_factory=VectorEncoderFactory(hidden_units=[30, 30]),\n",
    "    q_func_factory=MeanQFunctionFactory(),\n",
    "    target_update_interval=100,\n",
    ").create(device=device)\n",
    "# 再生バッファの設定\n",
    "buffer = create_fifo_replay_buffer(\n",
    "    limit=10000,\n",
    "    env=env,\n",
    ")\n",
    "# 探索\n",
    "explorer = LinearDecayEpsilonGreedy(\n",
    "    start_epsilon=1.0,\n",
    "    end_epsilon=0.3,\n",
    "    duration=1000,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 学習の開始\n",
    "# 事前学習したモデルを利用する場合はスキップ\n",
    "ddqn.fit_online(\n",
    "    env,\n",
    "    buffer,\n",
    "    explorer=explorer,\n",
    "    eval_env=env,\n",
    "    n_steps=100000,\n",
    "    n_steps_per_epoch=1000,\n",
    "    update_start_step=1000,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# モデルの保存\n",
    "ddqn.save_model(\"d3rlpy_logs/ddqn.pt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-07-30 17:48:26 [warning  ] Parameters will be reinitialized.\n"
     ]
    }
   ],
   "source": [
    "# モデルのリロード\n",
    "ddqn.build_with_env(env)\n",
    "ddqn.load_model(\"d3rlpy_logs/ddqn.pt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'episode reward')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# モデルの学習曲線\n",
    "log_path = sorted(glob(\"d3rlpy_logs/DoubleDQN_online_*/evaluation.csv\"))[-1]\n",
    "df = pd.read_csv(\n",
    "    log_path, \n",
    "    usecols=[1,2],\n",
    "    names=[\"train_episode\", \"episodic_rewards\"]\n",
    ")\n",
    "plt.plot(df[\"train_episode\"], df[\"episodic_rewards\"])\n",
    "plt.title(\"Learning curve of DDQN\")\n",
    "plt.xlabel(\"episode\")\n",
    "plt.ylabel(\"episode reward\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "random = DiscreteRandomPolicyConfig().create()\n",
    "random.build_with_env(env)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fdcc05ee3fa6411d9412b250f420654b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1b8c7418b60249a99c995c9abb7e2178",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# オンラインでの性能比較\n",
    "visualize_on_policy_policy_value(\n",
    "    env=env,\n",
    "    policies=[ddqn, random],\n",
    "    policy_names=[\"ddqn\", \"random\"],\n",
    "    n_trajectories=100,\n",
    "    random_state=random_state,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ff0061fa289047d5aee6fc0f25b99670",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d759903b14bc4c7a846cac845799ef94",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "relative improvement: 0.28032786885245903\n"
     ]
    }
   ],
   "source": [
    "# ddqn方策のオンライン評価による方策価値\n",
    "# ランダム方策より大きく改善（約20%）\n",
    "ddqn_performance = calc_on_policy_policy_value(env, ddqn, n_trajectories=100, random_state=random_state)\n",
    "random_performance = calc_on_policy_policy_value(env, random, n_trajectories=100, random_state=random_state)\n",
    "print(\"relative improvement:\", (ddqn_performance - random_performance) / random_performance)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`scope_rl.ope.online`では，以下の関数を用いて方策の性能をオンラインで評価できます.\n",
    "\n",
    "(統計量)\n",
    "\n",
    "* `calc_on_policy_policy_value`\n",
    "* `calc_on_policy_policy_value_interval`\n",
    "* `calc_on_policy_variance`\n",
    "* `calc_on_policy_conditional_value_at_risk`\n",
    "* `calc_on_policy_policy_interquartile_range`\n",
    "* `calc_on_policy_cumulative_distribution_function`\n",
    "\n",
    "(可視化)\n",
    "* `visualize_on_policy_policy_value`\n",
    "* `visualize_on_policy_policy_value_with_variance`\n",
    "* `visualize_on_policy_cumulative_distribution_function`\n",
    "* `visualize_on_policy_conditional_value_at_risk`\n",
    "* `visualize_on_policy_interquartile_range`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. 人工データの生成\n",
    "前章ではシミュレーション環境をオンライン環境とみなし，どのようにどのようにオンライン強化学習やオンラインでの性能評価を行うか確認しました．今回のようにシミュレーション環境上では簡単にエージェントを学習することができますが，実環境でオンラインのデータ収集や逐次的意思決定を行うことにはリスクを伴います．そこでオフライン強化学習ではデータ収集方策$\\pi_0$から集められた以下のオフラインのログデータのみを利用することで，新たな方策を学習することを考えます．\n",
    "\n",
    "$$\\mathcal{D}_0 := \\{ \\{ s_t, a_t, s_{t+1}, r_t \\}_{t=1}^T \\}_{i=1}^n \\sim \\prod_{i=1}^n d_{s_0}(s_0) \\prod_{t=1}^T \\pi_0(a_t | s_t) p(s_{t+1} | s_t, a_t) p(r_t | s_t, a_t),$$\n",
    "\n",
    "ここで $s_t \\in \\mathcal{S}$ は時刻$t$で観測された状態，$a_t$ はデータ収集方策 $\\pi_0$によって選択された行動，$r_t$は状態と行動に応じて観測された報酬を表します．\n",
    "\n",
    "`SyntheticDataset` は以下の引数を持っています:\n",
    "- `env`: 強化学習環境（RTBの環境）.\n",
    "- `state_keys`: 状態変数の名前\n",
    "- `max_episode_steps`: 一つのエピソードにおける連続意思決定の数（上限）．\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SCOPE-RLから必要なモジュールをインポート\n",
    "from scope_rl.policy import EpsilonGreedyHead"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "behavior_policy = EpsilonGreedyHead(\n",
    "    ddqn, \n",
    "    n_actions=env.action_space.n,\n",
    "    epsilon=0.3,\n",
    "    name=\"ddqn_epsilon_0.3\",\n",
    "    random_state=random_state,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# データセットクラスの初期化\n",
    "dataset = SyntheticDataset(\n",
    "    env=env,\n",
    "    action_meaning=env.action_meaning,\n",
    "    state_keys=env.obs_keys,\n",
    "    max_episode_steps=env.step_per_episode,\n",
    "    info_keys={\n",
    "        \"search_volume\": int,\n",
    "        \"impression\": int,\n",
    "        \"click\": int,\n",
    "        \"conversion\": int,\n",
    "        \"average_bid_price\": float,\n",
    "    },\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# データ収集方策によってログデータを集める\n",
    "# ログデータがある存在する場合はスキップ\n",
    "train_logged_dataset = dataset.obtain_episodes(\n",
    "    behavior_policies=behavior_policy,\n",
    "    n_trajectories=10000, \n",
    "    obtain_info=True,\n",
    "    random_state=random_state,\n",
    ")\n",
    "test_logged_dataset = dataset.obtain_episodes(\n",
    "    behavior_policies=behavior_policy,\n",
    "    n_trajectories=10000, \n",
    "    obtain_info=True,\n",
    "    random_state=random_state + 1,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'size': 70000,\n",
       " 'n_trajectories': 10000,\n",
       " 'step_per_trajectory': 7,\n",
       " 'action_type': 'discrete',\n",
       " 'n_actions': 10,\n",
       " 'action_dim': None,\n",
       " 'action_meaning': array([ 0.1       ,  0.16681005,  0.27825594,  0.46415888,  0.77426368,\n",
       "         1.29154967,  2.15443469,  3.59381366,  5.9948425 , 10.        ]),\n",
       " 'action_keys': None,\n",
       " 'state_dim': 7,\n",
       " 'state_keys': ['timestep',\n",
       "  'remaining_budget',\n",
       "  'budget_consumption_rate',\n",
       "  'cost_per_mille_of_impression',\n",
       "  'winning_rate',\n",
       "  'reward',\n",
       "  'adjust_rate'],\n",
       " 'state': array([[0.00000000e+00, 3.00000000e+03, 8.26436129e-01, ...,\n",
       "         9.56309981e-01, 1.00000000e+00, 1.24381775e+00],\n",
       "        [1.00000000e+00, 1.91000000e+03, 3.63333333e-01, ...,\n",
       "         9.90049751e-01, 6.00000000e+00, 2.15443469e+00],\n",
       "        [2.00000000e+00, 1.91000000e+03, 0.00000000e+00, ...,\n",
       "         0.00000000e+00, 0.00000000e+00, 1.66810054e-01],\n",
       "        ...,\n",
       "        [4.00000000e+00, 5.82000000e+02, 6.52951699e-01, ...,\n",
       "         1.00000000e+00, 4.00000000e+00, 5.99484250e+00],\n",
       "        [5.00000000e+00, 3.00000000e+01, 9.48453608e-01, ...,\n",
       "         4.38423645e-01, 5.00000000e+00, 1.29154967e+00],\n",
       "        [6.00000000e+00, 3.00000000e+01, 0.00000000e+00, ...,\n",
       "         4.28015564e-02, 0.00000000e+00, 1.29154967e+00]]),\n",
       " 'action': array([6, 1, 6, ..., 5, 5, 9]),\n",
       " 'reward': array([6., 0., 1., ..., 5., 0., 0.]),\n",
       " 'done': array([0., 0., 0., ..., 0., 0., 1.]),\n",
       " 'terminal': array([0., 0., 0., ..., 0., 0., 1.]),\n",
       " 'info': {'search_volume': array([201, 205, 217, ..., 203, 257, 179]),\n",
       "  'impression': array([199,   0, 216, ...,  89,  11,   8]),\n",
       "  'click': array([21,  0, 24, ..., 12,  0,  1]),\n",
       "  'conversion': array([6, 0, 1, ..., 5, 0, 0]),\n",
       "  'average_bid_price': array([116.91044776,   8.24390244, 112.33179724, ...,  62.2364532 ,\n",
       "          60.53307393, 463.00558659])},\n",
       " 'pscore': array([0.73, 0.03, 0.73, ..., 0.73, 0.73, 0.73]),\n",
       " 'behavior_policy': 'ddqn_epsilon_0.3',\n",
       " 'dataset_id': 0}"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_logged_dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"logs/train_dataset_discrete_ddqn.pkl\", \"wb\") as f:\n",
    "    pickle.dump(train_logged_dataset, f)\n",
    "with open(\"logs/test_dataset_discrete_ddqn.pkl\", \"wb\") as f:\n",
    "    pickle.dump(test_logged_dataset, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"logs/train_dataset_discrete_ddqn.pkl\", \"rb\") as f:\n",
    "    train_logged_dataset = pickle.load(f)\n",
    "with open(\"logs/test_dataset_discrete_ddqn.pkl\", \"rb\") as f:\n",
    "    test_logged_dataset = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b715f82baad04b41b6ea4096a2fe9d2d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "920cbc4ec6734758b51ae2f3c4a6338e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "83ad531943af419b8532429867435656",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# オンラインでの性能比較\n",
    "visualize_on_policy_policy_value(\n",
    "    env=env,\n",
    "    policies=[ddqn, behavior_policy, random],\n",
    "    policy_names=[\"ddqn\", \"ddqn (e=0.3)\", \"random\"],\n",
    "    n_trajectories=100,\n",
    "    random_state=random_state,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>adjust rate</th>\n",
       "      <th>average bid price</th>\n",
       "      <th>impression</th>\n",
       "      <th>click</th>\n",
       "      <th>conversion</th>\n",
       "      <th>winning rate</th>\n",
       "      <th>click / impression</th>\n",
       "      <th>conversion / click</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>70000.000000</td>\n",
       "      <td>70000.000000</td>\n",
       "      <td>70000.000000</td>\n",
       "      <td>70000.000000</td>\n",
       "      <td>70000.000000</td>\n",
       "      <td>70000.000000</td>\n",
       "      <td>70000.000000</td>\n",
       "      <td>70000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.338365</td>\n",
       "      <td>75.576464</td>\n",
       "      <td>68.830157</td>\n",
       "      <td>7.317400</td>\n",
       "      <td>1.823457</td>\n",
       "      <td>0.366040</td>\n",
       "      <td>0.051487</td>\n",
       "      <td>0.110896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.963969</td>\n",
       "      <td>105.805471</td>\n",
       "      <td>84.556172</td>\n",
       "      <td>9.566505</td>\n",
       "      <td>2.826144</td>\n",
       "      <td>0.411562</td>\n",
       "      <td>0.071177</td>\n",
       "      <td>0.167850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000106</td>\n",
       "      <td>4.045918</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.166810</td>\n",
       "      <td>8.036649</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.464159</td>\n",
       "      <td>36.129331</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.079218</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.291550</td>\n",
       "      <td>69.141414</td>\n",
       "      <td>161.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.823834</td>\n",
       "      <td>0.102410</td>\n",
       "      <td>0.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>10.000000</td>\n",
       "      <td>546.900000</td>\n",
       "      <td>277.000000</td>\n",
       "      <td>46.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.999997</td>\n",
       "      <td>0.999998</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        adjust rate  average bid price    impression         click  \\\n",
       "count  70000.000000       70000.000000  70000.000000  70000.000000   \n",
       "mean       1.338365          75.576464     68.830157      7.317400   \n",
       "std        1.963969         105.805471     84.556172      9.566505   \n",
       "min        0.000106           4.045918      0.000000      0.000000   \n",
       "25%        0.166810           8.036649      0.000000      0.000000   \n",
       "50%        0.464159          36.129331      3.000000      0.000000   \n",
       "75%        1.291550          69.141414    161.000000     16.000000   \n",
       "max       10.000000         546.900000    277.000000     46.000000   \n",
       "\n",
       "         conversion  winning rate  click / impression  conversion / click  \n",
       "count  70000.000000  70000.000000        70000.000000        70000.000000  \n",
       "mean       1.823457      0.366040            0.051487            0.110896  \n",
       "std        2.826144      0.411562            0.071177            0.167850  \n",
       "min        0.000000      0.000000            0.000000            0.000000  \n",
       "25%        0.000000      0.000000            0.000000            0.000000  \n",
       "50%        0.000000      0.079218            0.000000            0.000000  \n",
       "75%        3.000000      0.823834            0.102410            0.200000  \n",
       "max       19.000000      1.000000            0.999997            0.999998  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ログデータの統計量\n",
    "df = pd.DataFrame()\n",
    "df['adjust rate'] = train_logged_dataset['state'][:, 6]\n",
    "df['average bid price'] = train_logged_dataset['info']['average_bid_price']\n",
    "df['impression'] = train_logged_dataset['info']['impression']\n",
    "df['click'] = train_logged_dataset['info']['click']\n",
    "df['conversion'] = train_logged_dataset['info']['conversion']\n",
    "df['winning rate'] = train_logged_dataset['state'][:, 4]\n",
    "df['click / impression'] = df['click'] / (df['impression'] + 1e-5)\n",
    "df['conversion / click'] = df['conversion'] / (df['click'] + 1e-5)\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "より詳細にデータ収集と可視化について知りたい場合は，以下を参照してください．[examples/quickstart_ja/rtb/rtb_synthetic_data_collection_ja.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart_ja/rtb/rtb_synthetic_data_collection_ja.ipynb)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. オフライン強化学習\n",
    "この章では実環境と相互作用せずに，ログデータのみから新しい方策を学習する方法を紹介します．\n",
    "オフライン強化学習のアルゴリズムは [d3rlpy](https://github.com/takuseno/d3rlpy)に実装されているものを利用します．"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# d3rlpyのモジュールのインポート\n",
    "from d3rlpy.dataset import MDPDataset\n",
    "from d3rlpy.algos import DiscreteCQLConfig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "offlinerl_dataset = MDPDataset(\n",
    "    observations=train_logged_dataset[\"state\"],\n",
    "    actions=train_logged_dataset[\"action\"],\n",
    "    rewards=train_logged_dataset[\"reward\"],\n",
    "    terminals=train_logged_dataset[\"done\"],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Conservative Q-Learning 方策を利用\n",
    "cql = DiscreteCQLConfig(\n",
    "    encoder_factory=VectorEncoderFactory(hidden_units=[30, 30]),\n",
    "    q_func_factory=MeanQFunctionFactory(),\n",
    ").create(device=device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cql.fit(\n",
    "    offlinerl_dataset,\n",
    "    n_steps=10000,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# モデルの保存\n",
    "cql.save_model(\"d3rlpy_logs/cql_discrete.pt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-07-30 17:55:29 [warning  ] Parameters will be reinitialized.\n"
     ]
    }
   ],
   "source": [
    "# モデルのリロード\n",
    "cql.build_with_env(env)\n",
    "cql.load_model(\"d3rlpy_logs/cql_discrete.pt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c2403e3d858d4ce0a402244de8f48d69",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "48c9b338d75b4059b4a5b4b167c636b3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8886b23fef5c4a8c98c7320c649ff1e9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "50655b64964f4a40b9b4e87b92fe1374",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "[calculate on-policy policy value]:   0%|          | 0/100 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# オンラインでの性能比較\n",
    "# オフラインで学習した方策もオンラインで学習した方策と同等程度の性能を達成\n",
    "visualize_on_policy_policy_value(\n",
    "    env=env,\n",
    "    policies=[ddqn, behavior_policy, cql, random],\n",
    "    policy_names=[\"ddqn\", \"ddqn (e=0.3)\", \"cql (offline)\", \"random\"],\n",
    "    n_trajectories=100,\n",
    "    random_state=random_state,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. オフ方策評価とオフライン強化学習手法の評価\n",
    "前章ではオンラインの性能評価 (A/Bテスト) により学習した方策の性能を確認しました．しかし，実務においては方策の性能が悪い場合に顧客の満足度を毀損してしまうなど，A/Bテストにも大きなリスクが伴うことが知られています．そこで，学習した方策が実環境でどのように動作するかを事前に見積もるために，オフラインでの方策性能の評価を行います．\n",
    "\n",
    "#### 推定目標\n",
    "ここでは，方策の期待性能 (*方策価値*) を推定します．\n",
    "\n",
    "$$ V(\\pi) := \\mathbb{E}\\left[\\sum_{t=1}^T \\gamma^{t-1} r_t \\mid \\pi \\right] $$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SCOPE-RLからのopeモジュールのインポート\n",
    "from scope_rl.ope import CreateOPEInput\n",
    "from scope_rl.ope import OffPolicyEvaluation as OPE\n",
    "from scope_rl.ope.discrete import DirectMethod as DM\n",
    "from scope_rl.ope.discrete import TrajectoryWiseImportanceSampling as TIS\n",
    "from scope_rl.ope.discrete import PerDecisionImportanceSampling as PDIS\n",
    "from scope_rl.ope.discrete import DoublyRobust as DR\n",
    "from scope_rl.ope.discrete import SelfNormalizedTIS as SNTIS\n",
    "from scope_rl.ope.discrete import SelfNormalizedPDIS as SNPDIS\n",
    "from scope_rl.ope.discrete import SelfNormalizedDR as SNDR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 評価方策を定義\n",
    "cql_ = EpsilonGreedyHead(base_policy=cql, n_actions=env.action_space.n, name=\"cql\", epsilon=0.0, random_state=random_state)\n",
    "ddqn_ = EpsilonGreedyHead(base_policy=ddqn, n_actions=env.action_space.n, name=\"ddqn\", epsilon=0.0, random_state=random_state)\n",
    "random_ = EpsilonGreedyHead(base_policy=random, n_actions=env.action_space.n, name=\"random\", epsilon=1.0, random_state=random_state)\n",
    "evaluation_policies = [cql_, ddqn_, random_]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# まずOPEのインプットを準備\n",
    "prep = CreateOPEInput(\n",
    "    env=env,\n",
    "    model_args={\n",
    "        \"fqe\": {\n",
    "            \"encoder_factory\": VectorEncoderFactory(hidden_units=[30, 30]),\n",
    "            \"q_func_factory\": MeanQFunctionFactory(),\n",
    "            \"learning_rate\": 1e-4,\n",
    "        }\n",
    "    },\n",
    "    state_scaler=MinMaxObservationScaler(\n",
    "        minimum=test_logged_dataset[\"state\"].min(axis=0),\n",
    "        maximum=test_logged_dataset[\"state\"].max(axis=0),\n",
    "    ),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 計算に時間がかかります\n",
    "input_dict = prep.obtain_whole_inputs(\n",
    "    logged_dataset=test_logged_dataset,\n",
    "    evaluation_policies=evaluation_policies,\n",
    "    require_value_prediction=True,\n",
    "    n_trajectories_on_policy_evaluation=100,\n",
    "    random_state=random_state,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"logs/ope_input_dict_discrete.pkl\", \"wb\") as f:\n",
    "    pickle.dump(input_dict, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"logs/ope_input_dict_discrete.pkl\", \"rb\") as f:\n",
    "    input_dict = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "ope = OPE(\n",
    "    logged_dataset=test_logged_dataset,\n",
    "    ope_estimators=[DM(), TIS(), PDIS(), DR(), SNTIS(), SNPDIS(), SNDR()],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "# opeによる性能評価\n",
    "policy_value_df_dict, policy_value_interval_df_dict = ope.summarize_off_policy_estimates(input_dict, random_state=random_state)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cql':            policy_value  relative_policy_value\n",
       " on_policy     15.690000               1.000000\n",
       " dm            17.830122               1.136400\n",
       " tis            0.013578               0.000865\n",
       " pdis          10.765804               0.686157\n",
       " dr            14.412143               0.918556\n",
       " sntis         14.999998               0.956023\n",
       " snpdis        12.922537               0.823616\n",
       " sndr          13.945249               0.888799,\n",
       " 'ddqn':            policy_value  relative_policy_value\n",
       " on_policy     15.620000               1.000000\n",
       " dm            12.282429               0.786327\n",
       " tis            0.000000               0.000000\n",
       " pdis          12.686455               0.812193\n",
       " dr            16.830240               1.077480\n",
       " sntis          0.000000               0.000000\n",
       " snpdis        15.726926               1.006845\n",
       " sndr          17.240683               1.103757,\n",
       " 'random':            policy_value  relative_policy_value\n",
       " on_policy     12.570000               1.000000\n",
       " dm            12.010621               0.955499\n",
       " tis            4.857552               0.386440\n",
       " pdis          10.718063               0.852670\n",
       " dr            12.698753               1.010243\n",
       " sntis         11.730960               0.933251\n",
       " snpdis        11.914398               0.947844\n",
       " sndr          12.211570               0.971485}"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 推定した方策性能\n",
    "policy_value_df_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cql':                 mean  95.0% CI (lower)  95.0% CI (upper)\n",
       " on_policy  15.667900         14.840000         16.395750\n",
       " dm         17.830228         17.827503         17.833102\n",
       " tis         0.012627          0.000000          0.040734\n",
       " pdis       10.759151          9.630496         12.053074\n",
       " dr         14.214093         11.320457         16.318983\n",
       " sntis      13.949998          0.000000         44.999995\n",
       " snpdis     12.938436         10.765956         15.131035\n",
       " sndr       13.666319          9.728964         16.688969,\n",
       " 'ddqn':                 mean  95.0% CI (lower)  95.0% CI (upper)\n",
       " on_policy  15.569300         14.924750         16.220500\n",
       " dm         12.282437         12.281959         12.282890\n",
       " tis         0.000000          0.000000          0.000000\n",
       " pdis       12.877326          8.401726         18.559698\n",
       " dr         16.870486         15.428369         19.329362\n",
       " sntis       0.000000          0.000000          0.000000\n",
       " snpdis     16.036555          8.978773         24.831085\n",
       " sndr       17.307459         15.120284         21.308425,\n",
       " 'random':                 mean  95.0% CI (lower)  95.0% CI (upper)\n",
       " on_policy  12.497200         11.637000         13.475750\n",
       " dm         12.010628         12.010037         12.011211\n",
       " tis         5.012292          3.394177          6.685585\n",
       " pdis       10.877359          9.737495         12.146517\n",
       " dr         12.709577         11.850718         13.405091\n",
       " sntis      12.104656          8.196917         16.145648\n",
       " snpdis     12.124885         10.699847         13.801222\n",
       " sndr       12.243357         11.287623         13.216624}"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 推定した方策性能の信頼区間\n",
    "policy_value_interval_df_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. オフ方策評価手法の評価\n",
    "オフ方策推定量がどの程度正確に方策価値を推定できているかを評価します．"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJUAAAR0CAYAAADILCf5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAADpnElEQVR4nOzdeUBVdeL//9e97IvILiIiLqmRmWi5glqae5YtlK1W9unjUlNNTd/GKbVsyr7TWFM6Zc04LtVgjmuZX7Vy31dcwlwyFRdAQIQrAnJ/f/jzTgjq3e8Fn49/4t7zvve8rry5l16c8z4Gs9lsFgAAAAAAAGADo6cDAAAAAAAAoPahVAIAAAAAAIDNKJUAAAAAAABgM0olAAAAAAAA2IxSCQAAAAAAADajVAIAAAAAAIDNKJUAAAAAAABgM0olAAAAAAAA2IxSCQAAAAAAADajVAIAAKhjxo0bJ4PBoHHjxnk6CgAAqMMolQAAAAAAAGAzSiUAAAAAAADYjFIJAAAAAAAANqNUAgAAcDOTyaQPPvhAqampioiIUEBAgJo0aaK77rpLX375ZZWx+fn5euGFF9SkSRMFBAQoMTFRo0ePVn5+voYNGyaDwaB//etfnnkhAADguubr6QAAAADXk6NHj6pfv37au3evgoOD1a1bN0VFRSk7O1urV6/Wrl279PDDD0uSTp06pbS0NO3fv18REREaNGiQKisr9cUXX2jJkiW66aabPPxqAADA9YxSCQAAwE0qKyt17733au/everTp49mzZqlmJgYy/bS0lL98MMPltujR4/W/v37lZaWpkWLFql+/fqSLh69NGDAAC1cuNDtrwEAAOASTn8DAABwk0WLFmnLli1q2LCh/vOf/1QplCQpMDBQAwYMkHTxiKa5c+fKYDDok08+sRRKkhQZGalPPvnErdkBAAAuR6kEAADgJkuWLJEkPfzwwwoNDb3q2FWrVqmyslLt27dXcnJyte3t2rVT27ZtXZITAADAGpRKAAAAbvLrr79Kklq3bn3NsceOHZMkNW3a9IpjrrYNAADA1SiVAAAAAAAAYDNKJQAAADdJTEyUJGVlZV1zbKNGjSRJhw8fvuKYq20DAABwNUolAAAAN+nXr58k6auvvlJJSclVx3bv3l0Gg0Hbtm2rsYTauXOnMjMzXZITAADAGpRKAAAAbjJ48GClpKTo+PHjeuCBB3T69Okq20tLS/Xdd99JunhU05AhQ1RZWakRI0aoqKjIMq6goEAjR46U2Wx2a34AAIDf8vV0AAAAgOuF0WjUvHnz1LdvX3333XdKTExUamqqoqKilJ2drZ07dyo8PNxyWtvkyZO1c+dOrVixQk2bNlXPnj1lNpv1448/KioqSoMHD9bChQs9+6IAAMB1iyOVAAAA3KhJkybasmWLJk6cqJtuuknr16/X3Llz9euvv6pHjx6aOHGiZWxcXJw2btyo5557TsHBwfrmm2+0efNmPfTQQ9qwYYMiIiI8+EoAAMD1zmDmuGkAAIBaadiwYZo+fbqmTZumYcOGeToOAAC4znCkEgAAAAAAAGxGqQQAAAAAAACb1YmFuo8ePaqvv/5ahw4dUmFhoQICApSQkKC77rpLt956a5Wxx44d0/Tp05WVlSVfX1+1b99eTzzxhMLCwjyUHgAAAAAAoPapE2sqbdu2Td99951atmypiIgIlZWVaePGjfrpp5/0P//zP+rdu7ck6fTp0/rDH/6g4OBg9e/fX6WlpVq0aJGio6P1zjvvyNe3TnRsAAAAAAAALlcnSqWaVFZW6tVXX1V5ebk++OADSdLnn3+uFStW6IMPPlB0dLQkKTMzUxMmTKhSPgEAAAAAAODq6uyaSkajUVFRUSopKbHct3HjRrVv395SKElS27Zt1bBhQ61fv94TMQEAAAAAAGqlOnW+V2lpqcrKymQymbRlyxbt2LFDXbt2lSTl5+frzJkzat68ebXHtWjRQtu3b3d3XAAAAAAAgFqrTpVKM2bM0PLlyyVJBoNBnTp10lNPPSVJKigokCRFRERUe1xERISKi4tVXl4uPz+/atvLy8tVXl5uuW00GhUYGOiKlwAAAAAAAFAr1KlSaeDAgercubMKCgq0fv16VVZWqqKiQpJUVlYmSTUuxn2pSCorK6uxVJo3b57mzJljud2yZUtNmDDBFS8BAAAAAACgVqhTpVKjRo3UqFEjSVKPHj00YcIETZw4UX/+85/l7+8vSZaS6bcuHYV0aczlhgwZokGDBlluGwwGSVJubm6NzwfvYjAYFBcXp5MnT6qOrksPF2MOwVHMITiKOQRHMYfgKOYQHMUcqj18fX0VExNj3VgXZ/Gozp07a+rUqTpx4oTltLdLp8H9VkFBgUJDQ2s8Skm6eCTTlbbxw1B7mM1mvl9wCHMIjmIOwVHMITiKOQRHMYfgKOZQ3VJnr/4m/feUN5PJpMjISIWFhengwYPVxh04cEBJSUluTgcAAAAAAFB71YlS6cyZM9Xuq6io0MqVK+Xv76+EhARJUqdOnbRt2zbl5eVZxu3atUsnTpxQ586d3ZYXAAAAAACgtqsTp79NnTpV586d04033qjIyEgVFhZqzZo1ys7O1uOPP265UtuQIUO0YcMGjR8/XgMGDFBpaakWLlyoxMRE3X777R5+FQAAAAAAALVHnSiVunbtqh9++EFLly5VcXGxAgMD1axZMz3yyCO69dZbLeOio6M1btw4zZgxQ19++aV8fX2VkpKixx9//IprJgEAAAAAAKC6OlEqdevWTd26dbNqbOPGjTVmzBgXJwIAAAAAAKjb6sSaSgAAAAAAAHAvSiUAAAAAAADYjFIJAAAAAAAANqNUAgAAAAAAgM0olQAAAAAAAGAzSiUAAAAAAADYjFIJAAAAqMNMJpPi4+NlMBhkMpk8HQcAUIdQKgEAAAAAAMBmlEoAAAAAAACwma+tD8jJydGWLVuUlZWl7OxsFRUVyWAwqF69emrUqJFat26tW2+9VbGxsa7ICwAAAAAAAC9gdam0detWLVq0SFlZWTKbzYqLi1NsbKwaN24sSSopKdGvv/6qjRs3avr06WrdurUGDx6sDh06uCw8AAAAAAAAPMOqUmnMmDE6fPiwbrvtNr344ou6+eabFRwcXONYk8mkzMxMbdiwQZMmTVKTJk309ttvOzU0AAAAAAAAPMuqUummm27SK6+8ovDw8GuODQ4OVufOndW5c2cVFhZq8eLFjmYEAAAAAACAl7GqVHr44YftevLw8HC7HwsAAAAAAADvxdXfAAAAAAAAYDOrF+o+dOiQzU/erFkzmx8DAAAAAAAA72d1qfTaa6/Z/OQZGRk2PwYAAAAAAADez+pSacSIEa7MAQAAAAAAgFrE6lKpZ8+eLowBAAAAAACA2sSmhbrz8/OVn59/zTEFBQUOhQIAAAAAAIB3s7pUOnTokEaOHKl169Zdddy6des0cuRIHTlyxOFwAAAAAAAA8E5Wl0pLlixRw4YNNXDgwKuOGzhwoOLj47V48WKHwwEAAAAAAMA7WV0q7dmzR126dJHBYLjqOIPBoM6dO2v37t0OhwMAAAAAAIB3srpUKiwsVGxsrFVjo6OjWVcJAAAAAACgDrO6VAoICFBxcbFVY0tKSuTv7293KAAAAAAAAHg3q0ulJk2aaOvWrVaN3bp1q5o0aWJ3KAAAAAAAAHg3q0ul7t27a+/evfruu++uOm7JkiXau3evevTo4XA4AAAAAAAAeCdfawf26NFD69ev17/+9S9t375daWlpSkxMVFBQkM6dO6cjR45o9erV2rlzp9q2bauePXu6MDYAAAAAAAA8yepSyWg06uWXX9bMmTO1fPly7dy5s8Yxd955px5//PFrXiUOAAAAAAAAtZfVpZIk+fv76+mnn9aQIUO0fft2ZWdny2QyKTg4WPHx8UpJSVFUVJSrsgIAAAAAAMBL2FQqXRIZGalevXo5OwsAAAAAAABqCasX6gYAAAAAAAAuoVQCAAAAAACAzSiVAAAAAAAAYDNKJQAAAAAAANiMUgkAAAAAAAA2s+vqb5errKzU/v37lZ+fr/DwcLVs2VI+Pj7OeGoAAAAAAAB4IYdLpezsbE2cOFGnT59WaGioioqKFBkZqVdeeUVJSUlOiAgAAAAAAABv4/Dpb59//rnatWunadOm6dNPP9Wnn36qBg0a6NNPP3VGPgAAAAAAAHghq0ulqVOnqri4uNr9J06cUM+ePeXv7y9JCgsLU8eOHXXixAnnpQQAAAAAAIBXsbpUKigo0HPPPafFixersrLScn9ycrJmzpypn376SSdPntTWrVv17bffKjk52SWBAQAAAAAA4HlWr6n06quvaseOHZoxY4aWLVumJ554Qu3atdPw4cP1+eef66233tKFCxdkNBp12223afjw4a7MDQAAAAAAAA+yaaHudu3aqW3btvruu+/04YcfqmXLlho2bJief/55jR49WmfPnlW9evVkNDq8VBMAAAAAAAC8mM3tj9Fo1MCBA/Xhhx9arvI2Y8YMnT9/XvXr16dQAgAAAAAAuA7Y3ABVVFTIZDIpLCxMzz77rCZMmKCDBw/q+eef1/Lly2U2m12REwAAAAAAAF7E6tPfCgoK9Pe//12ZmZkym82Ki4vTs88+q+TkZI0fP17r1q3TrFmztHTpUg0bNoyFugEAAAAAAOowq49Umjp1qnJzc/XGG2/ovffeU1JSkt5//32dP39ektS1a1d98MEHuvXWW/XOO+/or3/9q8tCAwAAAAAAwLOsLpV++uknDRgwQMnJyWrSpIkeeeQRFRcX69ixY5Yx/v7+Sk9P16RJk2QwGFwSGAAAAAAAAJ5ndakUERGh/fv3W25f+jo8PLza2OjoaL344ouOpwMAAAAAAIBXsnpNpaFDh+rDDz9UVlaWQkJCdOjQIfXv319RUVGuzAcAAAAAAAAvZHWp1LFjR02aNEk7d+5UWVmZnnjiCbVu3dqV2ax24MABrVy5Unv27FFubq5CQ0N1ww036KGHHlJ8fHyVsceOHdP06dOVlZUlX19ftW/fXk888YTCwsI8lB4AAAAAAKD2sbpUkqTY2FjdeeedrspitwULFmjfvn3q3LmzmjRposLCQi1ZskSvvvqq3n77bSUmJkqSTp8+rbFjxyo4OFhDhw5VaWmpFi1apCNHjuidd96Rr69N/xwAAAAAAADXLae0KPn5+SouLlZYWFiNayy52qBBg/S73/2uSinUtWtXvfzyy5o/f76ef/55SdK8efN0/vx5TZw4UdHR0ZKkFi1aaMKECVqxYoV69+7t9uwAAAAAAAC1kdWl0rp169SyZUtLGSNJ27dv14wZM3T8+HHLfUlJSRo2bJhuvPFG5ya9ilatWlW7r2HDhkpISFB2drblvo0bN6p9+/ZVXkPbtm3VsGFDrV+/nlIJAAAAAADASlZf/e3SIt2XZGZmauLEiSopKdHdd9+tp59+WgMHDlRubq4mTJigw4cPuyKv1cxms86cOWNZKyk/P19nzpxR8+bNq41t0aKFfvnlF3dHBAAAAAAAqLXsPv3tq6++UlRUlCZOnKjQ0FDL/YMGDdIrr7yiuXPn6qWXXnJKSHusXr1a+fn5Sk9PlyQVFBRIkiIiIqqNjYiIUHFxscrLy+Xn51dte3l5ucrLyy23DQaDgoKCLF/Du136HvG9gr2YQ3AUcwiOYg7BEb+dNwaDgXkEu/A+BEcxh+omu0qliooKHTp0SMOGDatSKElSZGSkevfurR9++MEpAe2RnZ2tf/zjH2rZsqV69uwpSSorK5OkGhfjvlQklZWV1VgqzZs3T3PmzLHcbtq0qSZOnKiYmBgXpIerxMXFeToCajnmEBzFHIKjmEOwR0lJieXrBg0aKCQkxINpUNvxPgRHMYfqFrtKpQsXLki6+KFUk9jY2CofXu5UWFiod999V8HBwXrppZdkNF48w8/f31/SxULscpeOQro05nJDhgzRoEGDLLcvNau5ubk1Ph+8i8FgUFxcnE6ePCmz2ezpOKiFmENwFHMIjmIOwREmk8ny9alTpyxH3AO24H0IjmIO1R6+vr5WH0RjU6m0bds2FRYWSpICAwOVn59f47iCggKP/AXEZDLpz3/+s0pKSvTmm28qMjLSsu3SaW+XToP7rYKCAoWGhtZ4lJJ08UimK23jh6H2MJvNfL/gEOYQHMUcgqOYQ7DHb+cMcwiOYg7BUcyhusWmUmnt2rVau3at5famTZtqvGLarl27lJCQ4Hg6G5SVlWnixIk6ceKEXn/99Wr7j4yMVFhYmA4ePFjtsQcOHFBSUpKbkgIAAAAAANR+VpdKH3/8cbX7alpgq6ioSA0aNFC7du0cCmaLyspKffDBB/r555/1yiuvqGXLljWO69Spk1auXKm8vDxFR0dLuliAnThxQgMHDnRbXgAAAAAAgNrO6lLJ2vPpwsLCNHLkSLsD2WPGjBnasmWLOnTooOLiYq1atarK9u7du0u6uDbShg0bNH78eA0YMEClpaVauHChEhMTdfvtt7s1MwAAAAAAQG1m10Ld3ubw4cOSpK1bt2rr1q3Vtl8qlaKjozVu3DjNmDFDX375pXx9fZWSkqLHH3/8imsmAQAAAAAAoDqbS6Vt27Zp7dq1OnTokPLz81VWVqbAwEDFx8fr5ptv1p133qmoqChXZL2icePGWT22cePGGjNmjOvCAAAAAAAAXAesLpXOnz+v999/Xzt37qxyv4+Pj+Lj45WXl6d58+Zp8eLF+p//+R+lpqY6PSwAAAAAAAC8g9WlUkZGhnbt2qUnn3xS3bp1U0BAgPbu3avPPvtMycnJeuSRR3Ts2DHNmjVLH3/8sSIjI5WcnOzK7AAAAAAAAPAQo7UD161bp759+6pfv36qV6+e/P391a5dOw0fPlzffPONzpw5o4SEBL366qtq3ry55s6d68rcAAAAAAAA8CCrS6WzZ88qISGh2v0JCQmqrKzUiRMnJEkGg0Gpqak6cOCA81ICAAAAAADAq1hdKsXGxmr37t3V7t+zZ48kKTw83HKfj4+PzGaz4+kAAAAAAADglaxeU6lv376aNm2a/P391bVrV/n7+ysrK0sLFixQ69atFRcXZxl7+PDhKrfhvS48M9jTEdziqKcDuInPZws9HQEAAAAAcJ2wulTq16+f8vLy9O2332rlypWW+1u0aKEXXnihylh/f3/179/faSEBAAAAAADgXawulSTp0Ucf1aBBg/Tzzz+rvLxc8fHxatq0abVxw4YNc1Y+AAAAAABQi5lMJt1www2SpAMHDigoKMjDieAsNpVK0sW1kzp27OiKLAAAAAAAAKglrF6oGwAAAAAAALiEUgkAAMCLmUwmxcfHy2AwyGQyeToOAACABaUSAAAAAAAAbEapBAAAAAAAAJtRKgEAAAAAAMBmlEoAAAAAAACwma+9DywtLdU///lP3X333WrUqFGVbT///LOWL1+ukSNHOhzQm+0+VaIzpvOejuGQyqgbPR0BTmQ8dtbTEeokg8GgiBI/FRScldls9nQc1ELMITii9Nx/F+fekn1WAYEVHkyD2og5BGfgswyO4H2odqkfHKDbY2KsGmv3kUrl5eVauXKlCgoKqm07deqUVq5cae9TAwAAAAAAwMvZfaQSJGN5mYxlpZ6O4RhzpacTwIlq/Xz0UgaDQYbzpTKWlfKXOdiFOQRHGMvOK8jHaPnaaDR4OBFqG+YQnIHPMjiC96Haxehn/feHUskBraaNV/nBfZ6OAfzXO097OgEAwAX29U25+MWkUZ4NglqLOQTA03gfqj38mreSUr6waqzNpVJeXp4kqbi4WJJ05swZy32SFB0dbetTAgAAAAAAoJaxuVQaNapqq/i3v/2tyu2MjAzHEtUi+54cqzMl5zwdwyGVn77n6QhwIuOzf/B0hDrJYDAoPDxShYX5HO4NuzCH4IjSc+c0rH83SdL0JesUEBjo4USobZhDcAY+y+AI3odql/ohQYqzcqzNpdLYsWMlSSUlJfrLX/6ixx57TM2aNbP1aeqESj9/VfrX7jfUSoPda7XDG/nz5uwKBoNB5oBAVfoH8ksU7MIcgiMqL1Tq3IWLayBW+geokvd62Ig5BGfgswyO4H2odqn087d6rM2lUnJysiTp7NmLly5PSkqy3AcAAAAAAIDrA4epAAAAAACuqPScSfd0bKUeN8So9JzJ03EAeBGu/uaANg1CVF5u/WFh3ujC6Z88HQFO5JNQz9MR6iSDwaCGDaN14kQ5h3vDLswhOMJk8rF8fWujegoKCvJgGtRGzCE4ijkERzGHahc/Pz+rx9p9pJKPj4+Sk5MVEhJSbVv9+vU5JQ4AAAAAAKAOs/tIpeDgYMui3Zdr27at2rZta3coAAAAAAAAeDfWVAIAAAAAAIDNrDpS6fz58woICLBrB448FgAA4Fr+9re/eTqCS5WVlVm+njx5svz9a/d6jtfy/PPPezoCAACwklVHKo0YMUJz5sxRQUGB1U+cn5+vjIwMjRw50u5wAAAAAAAA8E5WHak0fPhwff3115ozZ45atWqlm2++Wc2aNVNsbKxCQkJkNptVUlKinJwcHTx4ULt27dL+/fvVsGFDPf30065+DQAAAIDdYg+85ukILlVyrtzydfSB1xUSZP1VfWqjnBbveDoCAFw3rCqVunbtqs6dO2vLli1asWKF5s2bp4qKipqf0NdXbdu21UsvvaRbb71VRiPLNgEAAAAAANQ1Vl/9zWg0qmPHjurYsaPKy8t16NAhZWdnq7i4WJIUGhqqRo0aqVmzZvLzq9t//QAAAAAAwFnid+zydASXKikttXwdt3O3QgIDPZjG9Y63u9nTEdzG6lLpt/z8/NSqVSu1atXK2XkAAAAAAABQC3BuGgAAAAAAAGxGqQQAAAAAAACb2XX6GwAAAADgoow9j3k6gkudP3fB8vXXe55WQJCPB9O43oM3zfR0BKDW4EglAAAAAAAA2IxSCQAAAAAAADajVAIAAAAAAIDN7F5TKS8vT3PnztWePXtUVFSkV155RcnJySoqKtKcOXN0++23q2nTps7MCgAAAAAAAC9h15FKx44d0x/+8AetX79esbGxMplMqqyslCSFhYVp3759WrJkiVODAgAAAAAAwHvYdaTSrFmzFBISorfffluS9Mwzz1TZnpKSovXr1zueDgAA4Drn7++vV1991dMxAAAAqrHrSKWffvpJd955p8LCwmQwGKptj46OVn5+vsPhAACo7Uwmk+Lj42UwGGQymTwdBwAAAHAau0qlyspKBQQEXHF7UVGRfH3tXq4JAAAAAAAAXs6uUqlZs2batm1bjdsuXLigdevWqWXLlg4FAwAAAAAAgPeyq1S65557tGPHDn322Wc6evSoJKmwsFCZmZmaMGGCsrOzdffddzs1KAAAAAAAALyHXeeopaSkaNSoUZo2bZqWL18uSfroo48kSUFBQRo1apSSk5OdlxIAAAAAAABexe6Fj7p3766OHTsqMzNTJ0+eVGVlpeLi4nTLLbcoKCjImRkBAAAAAADgZRxaTTswMFAdO3Z0Vha7lZaWauHChdq/f78OHDigkpISjRw5Uj179qw29tixY5o+fbqysrLk6+ur9u3b64knnlBYWJj7gwMAAAAAANRSdpVKeXl5Vo2Ljo625+ltVlRUpDlz5ig6OlpJSUnas2dPjeNOnz6tsWPHKjg4WEOHDlVpaakWLVqkI0eO6J133uGKdQDgAYsyCj0dwaXOnzdZvv726wIFBJz3YBrXu+vBcE9HAAA4WUCQj/669A5PxwDghexqUUaNGmXVuIyMDHue3mYRERGaOnWqwsPDdfDgQb322ms1jps3b57Onz+viRMnWgqvFi1aaMKECVqxYoV69+7tlrwAAAAAAFwvQgIDZf52iadjwAXsKpVGjBhR7b7Kykrl5uZq1apVCgsLU9++fR0OZy0/Pz+Fh4dfc9zGjRvVvn37KkdQtW3bVg0bNtT69esplQAAAAAAAKxkV6lU01pFl9x9990aM2aMTCbTFcd4Qn5+vs6cOaPmzZtX29aiRQtt377dA6kAAAAAAABqJ6cvIhQYGKiePXvq22+/1YABA5z99HYrKCiQdPFUuctFRESouLhY5eXl8vPzq7a9vLxc5eXlltsGg8FyhTuDweCixIDtmI+ucenflX9f4Nr4OYGjmENwFHMIjmIOwVHX0xxyycrUZrNZhYWFrnhqu5WVlUlSjYtxXyqSysrKaiyV5s2bpzlz5lhuN23aVBMnTlRMTIyL0rrPUU8HgFM1bNjQ0xHqtLi4OE9HqKMKPB0ATsT7EBzliTlUsd/tu4QLeeR9aLf7dwnX8cgc2p7p/n3CZa6n34ecWiqZTCb99NNPWrhwoZo2berMp3aYv7+/JKmioqLatktHIV0ac7khQ4Zo0KBBltuXWsfc3Nwanw/wlBMnTng6Qp1kMBgUFxenkydPymw2ezoO4NV4H4KjPDGHav+fCfFbvA/BUZ6YQ9dPBXF9qO3vQ76+vlYfRGNXqfTggw9edXt0dLSGDx9uz1O7zKXT3i6dBvdbBQUFCg0NrfEoJenikUxX2sb/YMKbMB9dy2w2828MXAM/I3AUcwiOYg7BUcwhOOp6mkN2lUr33XdftXMEDQaDQkJC1KBBA91yyy3y8fFxSkBniYyMVFhYmA4ePFht24EDB5SUlOT+UAAAAAAAALWUXaVSenq6s3O4RadOnbRy5Url5eUpOjpakrRr1y6dOHFCAwcO9HA6AEBdFBAQrE/+usfTMQAAAACnc8lC3Z6wZMkSlZSUWE5v27Jli06fPi1J6t+/v4KDgzVkyBBt2LBB48eP14ABA1RaWqqFCxcqMTFRt99+uyfjAwAAAAAA1CpWlUpTpkyx+YkNBoNGjBhh8+PstWjRIuXm5lpub9q0SZs2bZIkpaWlKTg4WNHR0Ro3bpxmzJihL7/8Ur6+vkpJSdHjjz9+xTWTAAAAAAAAUJ1VpdKePbYftn/5mkuuNnnyZKvGNW7cWGPGjHFxGgAAAAAAgLrNqlLJ2sIGAAAAAAAA1wejpwMAAAAAAACg9qFUAgAAAAAAgM3svvrb9u3b9c033+iXX36RyWSS2WyuNiYjI8OhcAAAAAAAAPBOdh2ptGHDBr377rs6c+aMunbtKrPZrG7duqlbt27y9/dXkyZNdP/99zs7KwAAAAAAALyEXUcqzZ8/Xy1atNBbb72l4uJiLVu2THfccYfatGmjnJwcjRkzRrGxsc7OCgAAAAAAAC9h15FKx44dU7du3WQ0GuXj4yNJqqiokCTFxsaqb9++WrBggfNSAgAAAAAAwKvYVSoFBATI1/fiQU4hISHy9fVVYWGhZXv9+vWVk5PjlIAAAAAAAADwPnaVSvHx8Tp27JjldlJSklatWqULFy6orKxMa9asUXR0tNNCAgAAAAAAwLvYVSrddttt2rx5s8rLyyVJ9957r/bs2aNhw4Zp+PDhysrK0j333OPMnAAAAAAAAPAiVi/UvXXrVqWkpMhoNGrw4MEaPHiwZVuHDh00btw4bdy4UUajUe3bt1ebNm1cEhgAAAAAAACeZ3Wp9N5776levXrq0qWL0tLS1LJlyyrbb7zxRt14441ODwgAAAAAAADvY3Wp9Mwzz2jt2rVatmyZli5dqpiYGKWmpio1NVUJCQmuzAgAAAAAAAAvY3Wp1Lt3b/Xu3Vv5+flau3at1q5dq3nz5mnevHlKSkpSWlqaunbtqsjISFfmBQAAAAAAgBewulS6JDIyUnfddZfuuusuHT9+XGvWrNHatWs1c+ZMzZo1S8nJyUpLS1OnTp0UHBzsiswAAAAAAADwMJtLpd+Kj49Xenq60tPTdeDAAa1du1br1q3TJ598on/+85+aOXOms3ICAAAAAADAixid9USxsbFq0KCBoqKiJEllZWXOemoAAAAAAAB4GYeOVCotLdXGjRu1Zs0a7d69W5WVlYqJidGQIUOUmprqrIwAAAAAAADwMjaXShUVFdq2bZvWrFmj7du3q6ysTKGhoerVq5dSU1PVunVrV+QEAAAAAACAF7G6VNq9e7fWrFmjjRs3ymQyyc/PTx06dFBaWppSUlLk4+PjypwAAAAAAADwIlaXSm+99ZYMBoNuuukmy9XdgoKCXJkNAAAAAAAAXsrqUumxxx5Tt27dFBER4co8AAAAAAAAqAWsLpUGDRrkyhwAAAAAAACoRYyeDgAAAAAAAIDah1IJAAAAAAAANqNUAgAAAAAAgM0olQAAAAAAAGAzu0qldevWqayszNlZAAAAAAAAUEtYffW33/rwww8VFBSkTp06KS0tTW3atHF2LgAAAAAAAHgxu0qlN998U6tXr9aGDRu0YsUKRUZGKjU1VWlpaUpMTHR2RgAAAAAAAHgZu0qlVq1aqVWrVnryySe1Y8cOrV69WkuWLNHChQuVmJio7t27KzU1VREREc7OCwAAAAAAAC9gV6l0iY+Pjzp06KAOHTqotLRUmzZt0ooVKzRr1ix9+eWXSk5OVo8ePdS1a1f5+jq0KwAAAAAAAHgRp1397ciRIzpw4ICOHDkiSYqPj1dxcbEmT56s5557TllZWc7aFQAAAAAAADzMocOHjh8/rjVr1mjNmjU6deqUwsLClJqaqu7du6tZs2aSpIMHD+qTTz7RZ599pvfff98poQEAAAAAAOBZdpVKixcv1urVq3Xo0CH5+fmpQ4cOGjZsmNq1ayejserBT82bN9fAgQP1ySefOCUwAAAAAAAAPM+uUmn69Olq1aqVnnnmGXXt2lXBwcFXHd+8eXPdd999dgUEAAAAAACA97GrVProo48UGxtr9fjGjRurcePG9uwKAAAAAAAAXsiuhbqjoqJkMpmuuN1kMunChQt2hwIAAAAAAIB3s6tUmjZtml5//fUrbn/99dc1Y8YMu0MBAAAAAADAu9lVKu3YsUOdOnW64vbOnTtr+/btdocCAAAAAACAd7OrVCooKFBkZOQVt0dERCg/P9/uUAAAAAAAAPBudpVKoaGhOn78+BW3Z2dnKygoyO5QAAAAAAAA8G52lUrt2rXT8uXL9csvv1TbdujQIS1fvlwpKSkOhwMAAAAAAIB38rXnQQ8++KB27NihP/7xj+rQoYMaN24sSTp69Ki2bt2qsLAwPfjgg04NCgAAAAAAAO9hV6kUGRmpd999V1988YW2bNmizZs3S5KCgoKUmpqqoUOHXnXNJQAAAAAAANRudpVK0sXFuEePHi2z2ayioiJJUlhYmAwGg9PCAQAAAAAAwDvZXSpdYjAYVL9+fWdkAQAAAAAAQC1hVak0Z84cSdK9994ro9FouX0t999/v/3JAAAAAAAA4LWsKpW+/vprSdI999wjo9FouX0tlEoAAAAAAAB1k1WlUkZGxlVvAwAAAAAA4Pri8JpKtU15ebkyMjK0evVqFRcXq0mTJnrooYfUtm1bT0cDAAAAAACoNYyeDuBukydP1rfffqvU1FQ9+eSTMhqNeuedd5SVleXpaAAAAAAAALWGVUcqjRo1SgaDwaYnNhgM+uijj+wK5SoHDhzQunXr9Oijj2rw4MGSpO7du+v3v/+9Zs2apQkTJng4IQAAAAAAQO1gVamUnJxsc6nkjTZs2CCj0ajevXtb7vP399cdd9yhr776Snl5eYqOjvZgQgAAAAAAgNrB6iOV6oJffvlFDRs2VHBwcJX7W7RoIUk6fPgwpRIAAAAAAIAVrquFugsLCxUREVHt/kv3FRQU1Pi48vJylZeXW24bDAYFBQVZvga8BfPR+Uwmk6V4PnjwoOVnH0DNeB+Co5hDcBRzCI5iDsFR19McsrtUqqys1KpVq7Rt2zbl5eVJkqKjo9WhQwelpaXJaPS+NcDLysrk5+dX7f5L95WVldX4uHnz5mnOnDmW202bNtXEiRMVExPjmqDu9O0WTycAvJ7ZbPZ0hDrt2RcbejoCarl3333X0xFQ2zWc7ukELuVbUiLpq4tfp34m35AQzwZyMU98qrzY8HsP7BV1SkN+H6pLrqfvpl2lkslk0ttvv60DBw4oKChIDRo0kCTt2rVLGzdu1NKlSzVmzJhqp5l5mr+/f5Ujji65dJ+/v3+NjxsyZIgGDRpkuX2pdczNzVVFRYULksKZDAaD4uLidPLkScoB2IU5BEcxh+Ao5hAcYTKZLF+fOnWKo25hF96H4CjmUO3h6+tr9UE0dpVKX331lQ4dOqSnnnpKvXr1kq/vxaepqKjQDz/8oGnTpunf//63nnrqKXue3mXCw8OVn59f7f5Lp73VdGqcdPFIppqOcJI4gqE2MZvNfL/gEOYQHMUcgqOYQ7DHb+cMcwiOYg7BUcyhusWuc9Q2bdqkPn36qG/fvpZCSbrYZvXp00d33nmnNm7c6LSQzpKUlKQTJ05U+WuNJO3fv9+yHQAAAAAAANdmV6lUXFys+Pj4K25v1KiRiouL7Q7lKp07d1ZlZaWWL19uua+8vFwrVqzQDTfcwJXfAAAAAAAArGTX6W9xcXHasmWL+vbtW+P2LVu2WNZZ8iY33HCDOnfurK+++kpFRUWKi4vTypUrlZubq//93//1dDwAAAAAAIBaw65SqU+fPvrnP/+pd955RwMGDFDD/3+l+uPHj+u7775TZmamnn76aacGdZbRo0crIyNDq1atUklJiRITE/Xqq68qOTnZ09EAAAAAAABqDbtKpb59++rMmTNasGCBduzYUfUJfX11//33q0+fPs7I53T+/v567LHH9Nhjj3k6CgAAAAAAQK1lV6kkSenp6erXr58yMzOVl5cnSYqJidHNN9+ssLAwpwUEAAAAAACA97GpVCosLNSKFSuUk5OjsLAwderUSampqa7KBgAAAAAAAC9ldamUk5Oj1157rcpV3ebPn6/Ro0dTLAEAAAAAAFxnrC6VZs+erdLSUj355JNq06aNTp48qWnTpmn69Onq2rWrjEajK3MCAAAAAADAi1hdKu3bt0+9e/dWv379JEkJCQkyGo2aOHGisrOz1bhxY5eFBAAAAAAAgHex+vCivLw8NW3atMp9zZo1kySdPXvWuakAAAAAAADg1awulSorK+XrW/XAJh8fH8s2AAAAAAAAXD9suvrbwYMH5efnZ7l97tw5SVJWVpZKSkqqje/UqZOD8QAAAAAAAOCNbCqVFi9erMWLF1e7/+uvv65xfEZGhn2pAAAAAAAA4NWsLpXGjh3ryhwAAAAAAACoRawulZKTk12ZAwAAAAAAALWI1Qt1AwAAAAAAAJdQKgEAAAAAAMBmlEoAAAAAAACwGaUSAAAAAAAAbEapBAAAAAAAAJtRKgEAAAAAAMBmvtYMWrlypV1P3qNHD7seBwAAAAAAAO9mVak0ZcoUu56cUgkAAAAAAKBusqpU+vjjj6vcLikp0eTJkxUcHKz+/fsrPj5ekpSdna0lS5bo3LlzGjVqlPPTAgAAAAAAwCtYVSrFxMRUuf31118rLCxMf/rTn2QwGCz3JyYmqlOnTnr77bf17bffauTIkc5NCwAAAAAAAK9g10LdmzdvVseOHasUSpYnNBrVsWNHbd682eFwAAAAAAAA8E52lUpms1nZ2dlX3H7s2DG7AwEAAAAAAMD72VUq3XbbbVq2bJm++eYbnT9/3nL/+fPntWjRIi1fvly33nqr00ICAAAAAADAu1i1ptLlnnzySeXk5GjmzJn68ssvFRERIUkqKCjQhQsX1KpVKw0bNsyZOQEAAAAAAOBF7CqVgoODNX78eG3evFnbt29XXl6eJOmWW25R+/bt1aFDhxrXWwIAAAAAAEDdYFepdMltt92m2267zVlZAAAAAAAAUEs4VCrl5+dr7969KioqUqdOnRQVFaXKykqZTCYFBwfLaLRrySYAAAAAAAB4ObtKJbPZrBkzZmjJkiWqrKyUJCUmJioqKkqlpaUaNWqU0tPTNXDgQKeGBQAAAAAAgHew61CihQsXavHixbrrrrv0pz/9qcq24OBgdezYURs3bnRKQAAAAAAAAHgfu0ql77//Xj169NDDDz+spKSkatubNGmiEydOOJoNAAAAAAAAXsquUun06dNq2bLlFbcHBATIZDLZHQoAAAAAAADeza5SKSwsTKdPn77i9kOHDik6OtruUAAAAAAAAPBudpVKnTp10rJly3Tq1Klq23bu3KkVK1aoS5cuDocDAAAAAACAd7Lr6m/p6enas2eP/vCHP6h169aSpAULFigjI0M///yzmjZtqiFDhjg1KAAAAAAAALyHXUcqBQcH6+2339bgwYOVn58vf39/7d27VyaTSQ888IDefPNNBQQEODsrAAAAAAAAvIRdRypJkr+/v+677z7dd999zswDAAAAAACAWsCuI5XGjx+vXbt2XXH77t27NX78eLtDAQAAAAAAwLvZVSrt3btXZ86cueL2oqIi7d271+5QAAAAAAAA8G52lUrXcvLkSQUFBbniqQEAAAAAAOAFrF5TacWKFVq5cqXl9ty5c/X9999XG2cymfTrr78qJSXFOQkBAAAAAADgdawulcrKylRUVGS5fe7cORkMhipjDAaDAgICdOedd+r+++93XkoAAAAAAAB4FatLpT59+qhPnz6SpFGjRunJJ5/Urbfe6rJgAAAAAAAA8F5Wl0q/NXnyZGfnAAAAAAAAQC1iV6n0W+fOnZPJZJLZbK62LTo62tGnBwAAAAAAgBeyu1RaunSpvvnmG506deqKYzIyMux9egAAAAAAAHgxoz0PWrp0qf7xj38oLi5ODz30kCRp4MCBuueeexQeHq6kpCSNGDHCqUEBAAAAAADgPewqlZYsWaJbbrlFf/zjH9W7d29JUvv27TV06FBNmjRJ586d09mzZ50aFAAAAAAAAN7DrlLp1KlT6tChgyTJx8dHklRRUSFJCg4O1h133KGlS5c6KSIAAAAAAAC8jV1rKgUHB+vChQuWr/39/ZWXl2fZHhQUpMLCQqcEvJaCggItXrxYBw4c0MGDB1VaWqqxY8fqpptuqnH8vn37NGvWLP3yyy8KCgpSly5d9PDDDyswMNAteQEAAAAAAOoCu45Uaty4sX799VfL7ZYtW2rZsmXKz89XXl6eli9froYNGzot5NUcP35cCxYsUH5+vhITE6869vDhw3rzzTdVVlamxx9/XL169dL333+vv/71r27JCgAAAAAAUFfYdaRSWlqali1bpvLycvn5+emBBx7QW2+9ZVmc29fXV7///e+dGvRKmjVrpn/+858KDQ3Vhg0brloQffXVVwoNDdXYsWMVHBwsSYqJidGnn36qnTt36pZbbnFLZgAAAAAAgNrOrlLp9ttv1+2332653bp1a/31r3/V1q1bZTQa1bZtW8XHxzst5NUEBQVZNc5kMikzM1MDBw60FEqS1KNHD02fPl3r16+nVAIAAAAAALCSXaVSTRo0aKABAwY46+mc7siRI7pw4YKaNWtW5X5fX18lJSXpl19+8VAyAAAAAACA2sfhUqmyslImk6nGbaGhoY4+vdNcWjg8IiKi2rbw8HBlZWVd8bHl5eUqLy+33DYYDJYjpAwGg3ODwukufY/4XsFezCE4ijkERzGH4IjfzhuDwcA8gl14H4KjmEN1k12lUkVFhRYsWKAff/xRp0+fVmVlZY3jMjIybHreyspKVVRUWDXWz8/PpslYVlZmedzl/P39LdtrMm/ePM2ZM8dyu2nTppo4caJiYmKs3j88Ly4uztMRUMsxh+Ao5hAcxRyCPUpKSixfN2jQQCEhIR5Mg9qO9yE4ijlUt9hVKk2dOlUrV65Uy5Ytddttt1VZo8gRP/30k8aPH2/V2EmTJqlRo0ZWP7e/v78kVTni6JKysjLL9poMGTJEgwYNsty+VGbl5uZaXYLBcwwGg+Li4nTy5EmZzWZPx0EtxByCo5hDcBRzCI747VkFp06dsnpNUuC3eB+Co5hDtYevr6/VB9HYVSpt2LBB3bt316hRo+x5+BU1atRII0eOtGpsTaexXU14eLgkqaCgoNq2wsLCqz6fn59fjUc4SeKHoRYxm818v+AQ5hAcxRyCo5hDsMdv5wxzCI5iDsFRzKG6xa5SKSAgQDfccIOzsyg8PFw9e/Z0+vNKUmJionx8fHTo0CF17drVcn9FRYUOHz6sLl26uGS/AAAAAAAAdZHRngd169ZN27Ztc3YWlwoODtbNN9+s1atX69y5c5b7V61apdLSUkolAAAAAAAAG9h1pNKjjz6qKVOm6N1339Xtt9+uqKgoGY3V+6lmzZo5HNAa//nPfyRJR48elXSxKLp0Nbf77rvPMu6hhx7S66+/rnHjxqlXr17Kz8/XokWLdMstt6hdu3ZuyQoAAAAAAFAX2FUqlZeXy2w2a/v27dq+ffsVx9l69Td7Xb6fH3/80fL1b0ulZs2a6fXXX9cXX3yh6dOnKygoSHfccYcefvhht+QEAAAAAACoK+wqlf7+979r06ZN6tatm1q0aOG0q7/Za/bs2VaPbd26td566y0XpgEAAAAAAKj77CqVdu7cqX79+mnYsGFOjgMAAAAAAIDawK6FuoOCghQXF+fsLAAAAAAAAKgl7CqVevXqpbVr16qystLZeQAAAAAAAFAL2HX6W0JCgrZs2aJXX31VPXr0uOLV3zp16uRwQAAAAAAAAHgfu0qlDz74wPL1zJkzrzjOXVd/AwAAAAAAgHvZVSqNHTvW2TkAAAAAAABQi9hVKiUnJzs7BwAAAAAAAGoRuxbqBgAAAAAAwPXNqiOVxo8fL4PBoDFjxsjHx0fjx4+/5mMMBoPeeOMNhwMCAAAAAADA+1h1pJLZbJbZbK5y25rHAAAAAAAAoG6y6kilcePGXfU2AAAAAAAAri92ram0d+9eFRUVXXF7UVGR9u7da3coAAAAAAAAeDe7SqXx48crMzPzitt3795t1bpLAAAAAAAAqJ1ccvW38vJyGY1cWA4AAAAAAKCusmpNJUnKy8tTTk6O5XZ2dnaNp7iZTCYtX75cMTExzkkIAAAAAAAAr2N1qfTjjz9qzpw5lttz587V3LlzaxxrNBr1zDPPOJ4OAAAAAAAAXsnqUqlLly5q3LixJGnSpEnq37+/WrduXWWMwWBQQECAkpKSFB4e7tSgAAAAAAAA8B5Wl0oJCQlKSEiQJI0YMULJycmKjY11WTAAAAAAAAB4L6tLpd/q2bNnjfefOnVK5eXllvIJAAAAAAAAdZNdpdJ3332nffv26YUXXrDcN2XKFK1cuVKS1LRpU7322muqX7++U0ICAAAAAADAuxjtedD3339fpTDasWOHVq5cqd69e+upp57SqVOn9PXXXzstJAAAAAAAALyLXUcq5ebmqlGjRpbb69evV2xsrOWKb4WFhVq1apVzEgIAAAAAAMDr2HWk0uUyMzPVrl07y+2YmBgVFhY646kBAAAAAADghewqlRo2bKjNmzdLunjqW35+vlJSUizb8/PzFRIS4pyEAAAAAAAA8Dp2nf5211136W9/+5uefPJJlZaWKiEhQbfccotl++7du5WUlOSsjAAAAAAAAPAydpVK3bp1U7169bRt2zaFhISob9++8vHxkSQVFxcrNDRU3bt3d2pQAAAAAAAAeA+7SiVJatu2rdq2bVvt/tDQUL388ssOhQIAAAAAAIB3s3pNpQMHDqi4uNiqsTk5OVq5cqXdoQAAAAAAAODdrC6VxowZox07dlhuFxcX69FHH9XevXurjd23b5+mTJnilIAAAAAAAADwPnZd/U2SzGazysvLVVlZ6cw8AAAAAAAAqAXsLpUAAAAAAABw/aJUAgAAAAAAgM0olQAAAAAAAGAzX1sG5+Tk6NChQ5Ikk8kkSTpx4oSCg4OrjQMAAAAAAEDdZVOplJGRoYyMjCr3ff75504NBAAAAAAAAO9ndak0YsQIV+YAAAAAAABALWJ1qdSzZ08XxgAAAAAAAEBtwkLdAAAAQB0WHBys48ePy2w2V1sLFQAAR1AqAQAAAAAAwGaUSgAAAAAAALAZpRIAAAAAAABsRqkEAAAAAAAAm1EqAQAAAAAAwGaUSgAAAAAAALAZpRIAAAAAAABsRqkEAAAAAAAAm1EqAQAAAAAAwGaUSgAAAAAAALAZpRIAAAAAAABsRqkEAAAAAAAAm/l6OkBt5uvLP19twvcLjmIOwVHMITiKOQRHMYfgKOYQHMUc8n62fI8MZrPZ7MIsAAAAAAAAqIM4/Q113rlz5/Tqq6/q3Llzno6CWoo5BEcxh+Ao5hAcxRyCo5hDcBRzqG6iVEKdZzab9csvv4iD8mAv5hAcxRyCo5hDcBRzCI5iDsFRzKG6iVIJAAAAAAAANqNUAgAAAAAAgM0olVDn+fn56f7775efn5+no6CWYg7BUcwhOIo5BEcxh+Ao5hAcxRyqm7j6GwAAAAAAAGzGkUoAAAAAAACwGaUSAAAAAAAAbEapBAAAAAAAAJtRKqFOmT17ttLT0z0dA3XEqFGjNHnyZE/HQB2Rnp6u2bNnW26vWLFC6enpysnJ8WAq1BZ8vsHVJk+erFGjRnk6BuqIcePGady4cZbbOTk5Sk9P14oVKzyWCbXDnj17lJ6erj179ng6CqxEqQTgurdv3z7Nnj1bJSUlno4CAIDL5Ofna/bs2Tp8+LCnowAA6ghfTwcAAE/bt2+f5syZo549eyokJMRy/wcffCCDweDBZKjLunfvrq5du3JZXQBuU1BQoDlz5ig2NlZJSUlVtj377LPiotBwlZiYGM2aNUu+vvzvJ1DX8FMNAFfA/+zDlYxGo/z9/T0dA3VEZWWlKioqmFOwG/+zD1cyGAy8P8Gpzp8/r4CAAE/HgCiVUItlZWVp+vTpOnLkiCIjIzV48OBqY9LT09W3b18lJydr9uzZysnJUVJSkp599lklJiZq2bJlWrhwofLz83XDDTdo5MiRio2N9cCrgafMnj1bc+bMkSSNHj3acv/HH3+s8ePHKzk52bLGREVFhebNm6fVq1fr9OnTCggIUKNGjfTAAw+obdu2HskP97k0VyZNmqSMjAzt3LlTPj4+SktL0yOPPGL5Zbm8vFxffPGFVq9erfLyct10000aPnx4tedbsWKFpkyZoo8//tjyvnPw4EH9+9//1qFDh1RaWqrw8HDddNNNGjlypFtfKzzLls+3li1bat68eTpx4oRefPFFdezY0QOJ4Wrnzp1TRkaGNm/erIKCAgUHB6tJkyZ65JFH1KxZM40bN05nz57Viy++qH/84x/av3+/QkJCNGDAAN19992SLq5TMn78eEnSlClTNGXKFEnSyJEj1bNnT02ePFl79+6tspbg2rVrtXDhQp04cUIGg0HR0dHq1auXBgwY4P5/BFyTM+aJ9N+58sILL+jw4cP68ccfVVpaqjZt2ujpp59WdHR0lf0uX75cCxYsUH5+vhITE/X4449Xy5aTk6PRo0db5pskFRYW6ssvv1RmZqaKiooUGhqqFi1aaNiwYfw+7sWcNc8k6fTp0/rHP/6hXbt2KSAgQKmpqWrXrl21fV56zlGjRmn69Ok6ePCgevfurWHDhrnnReOqKJVQKx05ckQTJkxQWFiYHnjgAV24cEGzZ89WeHh4tbFZWVnasmWL+vbtK0maP3++3n33XQ0ePFhLly5V3759VVxcrIULF+rvf/+7xo4d6+ZXA0/q1KmTTpw4obVr1+qJJ55QvXr1JElhYWHVxn799deaP3++7rjjDrVo0ULnzp3TwYMHdejQIUql68ikSZMUExOjoUOHav/+/fruu+9UUlJiKSU/+eQTrV69WqmpqWrZsqV2796td99995rPe+bMGcv72t13362QkBDl5uZq48aNrn5J8CK2fL7t3r1b69evV79+/VSvXj3+J6wO++yzz7Rhwwb169dPCQkJOnv2rLKyspSdna1mzZpJkoqLi/X222+rU6dO6tKlizZs2KAvvvhCiYmJSklJUaNGjSwXDOjdu7dat24tSWrVqlWN+8zMzNSHH36om2++WY888ogkKTs7W1lZWZRKXsoZ8+S35s6dK4PBoLvvvltFRUX69ttv9dZbb+n//t//a/lDyg8//KCpU6eqVatWGjBggE6dOqX33ntPoaGhioqKumre999/X0ePHlW/fv0UGxurM2fOKDMzU3l5ebyfeTFnzbOysjK9+eabysvLU//+/RUZGalVq1ZdcYHus2fP6s9//rO6du2qtLQ01a9f322vGVdHqYRaKSMjQ2azWW+++ablryWdOnXSyy+/XG3s8ePHNWnSJMuHU2hoqKZOnaq5c+fqww8/VFBQkKSLpw7Mnz9fOTk5fJBdR5o0aaKmTZtq7dq1uu222676vd+2bZtSUlL07LPPujEhvE1sbKz+8Ic/SJL69eunoKAgLV26VHfddZfMZrNWr16tPn36WI5O6tevn/72t7/p119/verz7tu3TyUlJfrTn/6k5s2bW+5/6KGHXPdi4HVs/Xx7//33lZCQ4O6YcLNt27apV69eVY4Aufwv/gUFBRo9erS6d+8uSbrjjjs0cuRI/fDDD0pJSVF4eLhSUlI0e/ZstWzZ0jLuavsMCgrSmDFjZDRybZ/awBnz5LeKi4s1adIky+/KTZs21aRJk7R8+XINGDBAFRUV+uqrr5SUlKSxY8daTqFMSEjQ1KlTr1oqlZSUaN++fXr00UerHI05ZMgQx/4R4HLOmmfLly+3HGXbpUsXSVKvXr30yiuv1LjfwsJCPfPMM7rzzjtd8bLgAD4hUOtUVlZq586duu2226ocfpuQkKBbbrml2vg2bdpUKQpatGgh6eIv6Zc+JCXphhtukCQu740rCgkJ0dGjR3XixAlPR4EHXTrq8ZL+/ftLkrZv367t27dLUrW/4lvzV/1Li8Rv3bpVFRUVzoiKWsbWz7fk5GQKpetESEiIDhw4oPz8/CuOCQwMVFpamuW2r6+vWrRoYffvNcHBwTp//rwyMzPtejzcz9nzpHv37lV+V+7cubMiIiIsn3WHDh3SmTNndOedd1ZZk6tnz54KDg6+alZ/f3/5+vpq7969Ki4utvo1wvOcNc+2b9+uiIgIde7c2XJfQECAevfuXeNz+vn56fbbb3fCK4CzUSqh1ikqKlJZWZkaNmxYbVt8fHy1+y4/7/vSh9zlfz25dD8fbLiS9PR0mUwm/e53v9Pvf/97zZw585pHn6Duufy9p0GDBjIYDMrJyVFubq4MBoMaNGhQZUxN702XS05OVqdOnTRnzhw9/fTTeu+99/Tjjz+qvLzcqfnhvWz9fOOo2uvHI488oiNHjmjEiBF67bXXNHv2bJ06darKmKioqGpXLA0JCbH795q+ffuqYcOG+vOf/6z//d//1ZQpU7Rjxw57XwLcwNnz5PL3IoPBoLi4OOXm5kqS5b+Xj/P19a32OXg5Pz8/PfLII9q+fbueeeYZjR07VgsWLFBhYaFVrxWe46x5lpubq7i4uGrjrvQ7U2RkJBcU8FJ8V1DnXemQbQ7lhq2Sk5P10UcfafPmzdq5c6d++OEHffvtt3rmmWfUq1cvT8eDh1z+y5Ajz/P73/9eP//8s7Zu3aqdO3fq73//u7755hu9/fbbCgwMdMp+UHdwJaXrR9euXXXjjTdq06ZN2rlzpxYtWqQFCxbo5ZdftpxK4uzfa+rXr6//+3//r3bs2KEdO3Zo+/btWrFihbp3717lwhbwHp6YJ44YOHCgOnToYPm9KiMjQ/Pnz9cbb7yhpk2bejoersBT84zPPO/lPe8qgJXCwsLk7+9f4ylIx48f90Ai1Ha2lAKhoaG6/fbb9cILL+jvf/+7mjRpoq+//tqF6eBtLn/vOXnypMxms2JjYxUTEyOz2VztL3a2vDe1bNlSQ4cO1bvvvqvnn39eR48e1dq1a52SHd6NzzdcTUREhPr27as//OEP+vjjj1WvXj3NnTvXpuewtQT39fXVrbfequHDh+ujjz5S7969tWrVKp08edKm54H7OGOeXHL5e5HZbNbJkycVExMjSZb/Xj6uoqLC6tMu4+LidNddd+lPf/qT3n//fVVUVGjRokV25YX7OGOexcTEWH6H+i0+72ofSiXUOkajUbfccos2b96svLw8y/3Hjh3Tzp07PZgMtVVAQIAkyWQyXXXc2bNnq9wODAxUgwYNOD3pOvP//t//q3L7u+++kyS1a9fO8he6xYsXVxlz+e2aFBcXV/vFKikpSZKYY9cJPt9Qk8rKymqfT/Xr11dERITN669d+rwrKSm55tjLP/OMRqOaNGkiifckb+TMeXLJqlWrdO7cOcvtDRs2qKCgwPJZ16xZM4WFhWnZsmVV9rFixYprzrHz58+rrKysyn0NGjRQYGAg6wp6MWfOs5SUFBUUFGjDhg2W+86fP6/ly5c7JSvch9PfUCulp6drx44deuONN9SnTx9VVlbqu+++U+PGjVnjBja7dPnTr776St26dZOPj486dOhQbdxLL72k5ORkNWvWTKGhoTp48KA2btxYbeFm1G05OTmaOHGi2rVrp59//lmrV69WamqqpQDq1q2bli5dKpPJpFatWmnXrl3VjlyqycqVK7V06VLddtttiouL07lz5/T9998rKChI7du3d/Grgrfg8w2XO3funP73f/9XnTt3VpMmTRQYGKhdu3bp4MGDVa6+ZI0GDRooJCREy5YtU1BQkAICAnTDDTfUuD7XJ598ouLiYrVp00ZRUVHKzc3VkiVLlJSUpEaNGjnr5cFJnDlPLgkNDdUbb7yhnj176syZM/r2228VFxdnOeXf19dXDz30kKZOnarx48era9euysnJ0YoVK665ptKJEyf05ptvqkuXLkpISJCPj482bdqkM2fOqGvXrnblhes5c5716tVLS5Ys0ccff6xDhw4pIiJCq1atspTfqD0olVArNWnSRGPGjNGMGTM0e/ZsRUVFKT09XQUFBfzSDZu1aNFCDz74oJYtW6YdO3bIbDbr448/rjauf//+2rJlizIzM1VeXq6YmBg9+OCDVS6Fi7rvhRde0OzZs/Xll1/KaDSqX79+evTRRy3bR4wYobCwMK1Zs0abN29WmzZt9H/+z//RiBEjrvq8ycnJOnDggNatW6czZ84oODhYzZs31/PPP8+CzNcRPt9wuYCAAPXt21c7d+7Upk2bVFlZqbi4OA0fPlx9+vSx6bl8fX01atQoffnll/rss8904cIFjRw5ssb3mLS0NH3//fdaunSpSkpKFB4eri5duig9Pd2r1uXBRc6cJ5cMGTJEv/76q+bPn69z587p5ptv1vDhw6v8T3/v3r1VWVmphQsXatasWUpMTNQf/vAHZWRkXPW5o6Ki1K1bN+3evVurVq2Sj4+PGjVqpBdffLHK1cDgXZw5zwICAvTGG2/on//8p5YsWSJ/f3+lpaWpXbt2+vOf/+yiVwBXMJgvP9YeAABUM3v2bM2ZM0eff/65wsLCPB0HAACX2LNnj8aPH6+XXnqJggfANfFnBgAAAAAAANiMUgkAAAAAAAA2o1QCAAAAAACAzVhTCQAAAAAAADbjSCUAAAAAAADYjFIJAAAAAAAANqNUAgAAAAAAgM0olQAAAAAAAGAzSiUAAAAAAADYjFIJAAAAAAAANqNUAgAAAAAAgM0olQAAAAAAAGAzSiUAAAAAAADYjFIJAAAAAAAANqNUAgAAcKGkpCQZDAYdPnzY6sf07NlTBoNBK1ascFkuAAAAR1EqAQAAAAAAwGaUSgAAAAAAALAZpRIAAAAAAABsRqkEAADgoL179+qBBx5QdHS0goKC1KZNG/3lL3/RhQsXrviYo0eP6qmnnlLDhg0VGBioG264QWPGjNG5c+euuq/8/Hy98MILatKkiQICApSYmKjRo0crPz9fw4YNk8Fg0L/+9a8qj/nt/b/88osee+wxxcXFKSAgQM2bN9ef/vQnnT9/3hn/FAAA4Dri6+kAAAAAtdmaNWvUr18/lZSUqFmzZrrzzjuVl5enP/7xj9qwYUONj8nKylKPHj2Uk5Ojhg0bavDgwSopKdGkSZP0448/XnFfp06dUlpamvbv36+IiAgNGjRIlZWV+uKLL7RkyRLddNNNV826Y8cO/e53v1NERIR69Oih/Px8rV27Vm+//bb27NmjefPmOfRvAQAAri+USgAAAHYqLS3Vww8/rJKSEr3wwgv6y1/+Ih8fH0lSZmamevXqpby8vGqPe/zxx5WTk6P09HRNnz5dgYGBkqQjR47ojjvu0MGDB2vc3+jRo7V//36lpaVp0aJFql+/vqSLRy8NGDBACxcuvGreDz/8UGPGjNH48eMtOXfv3q3OnTtr/vz5Wr9+vbp06WL3vwcAALi+cPobAACAnf7zn//o6NGjaty4sd577z1LUSNJbdu21ZgxY6o9Zu3atdq8ebNCQkI0ZcoUS6EkSYmJifrLX/5S476OHj2quXPnymAw6JNPPrEUSpIUGRmpTz755Jp5O3TooLfeeqtKzjZt2uixxx6TJC1fvvzaLxoAAOD/R6kEAABgpxUrVkiS0tPT5efnV237E088ccXH9OvXT1FRUdW233333VUKo0tWrVqlyspKtW/fXsnJydW2t2vXTm3btr1q3kGDBslgMFS7/8Ybb5QkZWdnX/XxAAAAv0WpBAAAYKdjx45Jkpo2bVrj9oiIiGoF0bUeYzAYlJSUZPO+rrVNungkVE3CwsIkXTydDwAAwFqUSgAAANcJo5Ff/QAAgPPwmwUAAICdGjVqJEk6fPhwjdsLCwt15swZmx4jSb/++qvN+7rWNgAAAGejVAIAALBTjx49JEmzZ89WeXl5te0zZsy44mOWLFmi/Pz8atsXLlyowsLCavd3795dBoNB27ZtU1ZWVrXtO3fuVGZmpq0vAQAAwG6USgAAAHa6//771ahRIx05ckSvvfaaKisrLdt2796tCRMmVHtMWlqa2rdvr+LiYo0aNUrnz5+3bDt69KhefvnlGveVmJioIUOGqLKyUiNGjFBRUZFlW0FBgUaOHCmz2ezEVwcAAHB1lEoAAAB2CgoK0hdffKHg4GC9//77atmypYYOHao+ffqoffv2SktLU5MmTao9bubMmYqJidG///1vNWvWTA8++KDuuusutW7dWlFRUerSpUuN+5s8ebKaN2+uFStWqGnTprrvvvt07733qlmzZjp16pQGDx7s6pcMAABgQakEAADggB49emjjxo269957VVBQoHnz5unYsWN68803lZGRUeNjkpOTtWXLFg0bNkwXLlzQ/PnztXfvXj333HP6/vvv5e/vX+Pj4uLitHHjRj333HMKDg7WN998o82bN+uhhx7Shg0bFBER4cqXCgAAUIXBzHHSAAAAdcKwYcM0ffp0TZs2TcOGDfN0HAAAUMdxpBIAAAAAAABsRqkEAAAAAAAAm1EqAQAAAAAAwGasqQQAAAAAAACbcaQSAAAAAAAAbEapBAAAAAAAAJtRKgEAAAAAAMBmlEoAAAAAAACwma+nA9RmBQUFqqio8HQMWCEmJka5ubmejoFajDkERzGH4CjmEBzFHIKjmENwFHOodvD19VVERIR1Y12cpU6rqKhQeXm5p2PgGgwGg6SL3y8udgh7MIfgKOYQHMUcgqOYQ3AUcwiOYg7VTZz+BgAAAAAAAJtRKgEAAAAAAMBmlEoAAAAAAACwGaUSAAAAAAAAbEapBAAAAAAAAJtRKgEAAAAAAMBmlEoAAAAAAACwGaUSAAAAAAAAbEapBAAAAAAAAJtRKgEAAAAAAMBmlEoAAAAAAACwGaUSAAAAAAAAbEapBAAAAAC4IpPJpPj4eBkMBplMJk/HAeBFKJUAAAAAAABgM0olAAAAAAAA2MzXXTvKycnRli1blJWVpezsbBUVFclgMKhevXpq1KiRWrdurVtvvVWxsbHuigQAAAAAAAA7ubxU2rp1qxYtWqSsrCyZzWbFxcUpNjZWjRs3liSVlJTo119/1caNGzV9+nS1bt1agwcPVocOHVwdDQAAAAAAAHZyaak0ZswYHT58WLfddptefPFF3XzzzQoODq5xrMlkUmZmpjZs2KBJkyapSZMmevvtt10ZDwAAAAAAAHZyaal000036ZVXXlF4ePg1xwYHB6tz587q3LmzCgsLtXjxYldGAwAAAAAAgANcWio9/PDDdj0uPDzc7scCAAAAAADA9bj6GwAAAAAAAGzm8oW6Dx06ZPNjmjVr5oIkAAAAAAAAcBaXl0qvvfaazY/JyMhwQRIAAAAAAAA4i8tLpREjRrh6FwAAAAAAAHAzl5dKPXv2dPUuAAAAAAAA4GZuWag7Pz9f+fn51xxTUFDgjjgAAAAAAABwkMtLpUOHDmnkyJFat27dVcetW7dOI0eO1JEjR1wdCQAAAAAAAA5yeam0ZMkSNWzYUAMHDrzquIEDByo+Pl6LFy92dSQAAAAAAAA4yOWl0p49e9SlSxcZDIarjjMYDOrcubN2797t6kgAAAAAAABwkMtLpcLCQsXGxlo1Njo6mnWVAAAAAAAAagGXl0oBAQEqLi62amxJSYn8/f1dnAgAAAAAAACOcnmp1KRJE23dutWqsVu3blWTJk1cnAgAAAAAAACOcnmp1L17d+3du1fffffdVcctWbJEe/fuVY8ePVwdCQAAAAAAAA7ydfUOevToofXr1+tf//qXtm/frrS0NCUmJiooKEjnzp3TkSNHtHr1au3cuVNt27ZVz549XR0JAAAAAAAADnJ5qWQ0GvXyyy9r5syZWr58uXbu3FnjmDvvvFOPP/74Na8SBwAAAAAAAM9zeakkSf7+/nr66ac1ZMgQbd++XdnZ2TKZTAoODlZ8fLxSUlIUFRXljigAAAAAAABwAreUSpdERkaqV69e7twlAAAAAAAAXMDlC3UDAAAAAACg7qFUAgAAAAAAgM0olQAAAAAAAGAzSiUAAAAAAADYjFIJAAAAAAAANnPr1d8uV1lZqf379ys/P1/h4eFq2bKlfHx8PBkJAAAAAAAAVvBYqZSdna2JEyfq9OnTCg0NVVFRkSIjI/XKK68oKSnJU7EAAAAAAABgBY+d/vb555+rXbt2mjZtmj799FN9+umnatCggT799FNPRQIAAAAAAICVXF4qTZ06VcXFxdXuP3HihHr27Cl/f39JUlhYmDp27KgTJ064OhIAAAAAAAAc5PLT3woKCvTcc8/pgQceUL9+/WQ0XuyxkpOTNXPmTKWnpysiIkLZ2dn69ttvlZycbPM+Dhw4oJUrV2rPnj3Kzc1VaGiobrjhBj300EOKj4+vMvbYsWOaPn26srKy5Ovrq/bt2+uJJ55QWFiYU14vAAAAAADA9cDlpdKrr76qHTt2aMaMGVq2bJmeeOIJtWvXTsOHD9fnn3+ut956SxcuXJDRaNRtt92m4cOH27yPBQsWaN++fercubOaNGmiwsJCLVmyRK+++qrefvttJSYmSpJOnz6tsWPHKjg4WEOHDlVpaakWLVqkI0eO6J133pGvr0fXLQcAAAAAAKg13NKitGvXTm3bttV3332nDz/8UC1bttSwYcP0/PPPa/To0Tp79qzq1atnOYrJVoMGDdLvfve7KqVQ165d9fLLL2v+/Pl6/vnnJUnz5s3T+fPnNXHiREVHR0uSWrRooQkTJmjFihXq3bu34y8WAAAAAADgOuC2hbqNRqMGDhyoDz/80HKVtxkzZuj8+fOqX7++3YWSJLVq1araUUYNGzZUQkKCsrOzLfdt3LhR7du3txRKktS2bVs1bNhQ69evt3v/AAAAAAAA1xu3lUoVFRUymUwKCwvTs88+qwkTJujgwYN6/vnntXz5cpnNZqfuz2w268yZM5a1kvLz83XmzBk1b9682tgWLVrol19+cer+AQAAAAAA6jK3LNT997//XZmZmTKbzYqLi9Ozzz6r5ORkjR8/XuvWrdOsWbO0dOlSDRs2zK6FumuyevVq5efnKz093ZJDkiIiIqqNjYiIUHFxscrLy+Xn51dte3l5ucrLyy23DQaDgoKCLF/Du136HvG9gr2YQ3AUcwiOYg7BUcwhOOK388ZgMDCPYBfeh+oml5dKU6dOVW5urt544w2FhIRo7ty5ev/99zVlyhQFBASoa9euuvXWWzV//ny98847SklJ0UsvveTQPrOzs/WPf/xDLVu2VM+ePSVJZWVlklTjYtyXiqSysrIaS6V58+Zpzpw5lttNmzbVxIkTFRMT41BOuFdcXJynI6CWYw7BUcwhOIo5BEcxh2CPkpISy9cNGjRQSEiIB9OgtuN9qG5xean0008/6ZFHHrEcgfTII4/oueee07Fjxyynovn7+ys9PV133HGHZs6c6dD+CgsL9e677yo4OFgvvfSSZa0mf39/SRdPw7vcpaOQLo253JAhQzRo0CDL7UvNam5ubo3PB+9iMBgUFxenkydPOv00S1wfmENwFHMIjmIOwVHMITjCZDJZvj516pTlrA3AFrwP1R6+vr5WH0Tj8lIpIiJC+/fv15133ilJ2r9/vyQpPDy82tjo6Gi9+OKLdu/LZDLpz3/+s0pKSvTmm28qMjKySg7pv6fB/VZBQYFCQ0NrPEpJungk05W28cNQe5jNZr5fcAhzCI5iDsFRzCE4ijkEe/x2zjCH4CjmUN3i8lJp6NCh+vDDD5WVlaWQkBAdOnRI/fv3V1RUlFP3U1ZWpokTJ+rEiRN6/fXXlZCQUGV7ZGSkwsLCdPDgwWqPPXDggJKSkpyaBwAAAAAAoC5zeanUsWNHTZo0STt37lRZWZmeeOIJtW7d2qn7qKys1AcffKCff/5Zr7zyilq2bFnjuE6dOmnlypXKy8tTdHS0JGnXrl06ceKEBg4c6NRMAAAAAAAAdZnLSyVJio2NtZz+5gozZszQli1b1KFDBxUXF2vVqlVVtnfv3l3SxbWRNmzYoPHjx2vAgAEqLS3VwoULlZiYqNtvv91l+QAAAAAAAOoat5RKV5Kfn6/i4mKFhYXVuMaStQ4fPixJ2rp1q7Zu3Vpt+6VSKTo6WuPGjdOMGTP05ZdfytfXVykpKXr88cevuGYSAAAAAAAAqnN5qbRu3Tq1bNnScrqZJG3fvl0zZszQ8ePHLfclJSVp2LBhuvHGG23ex7hx46we27hxY40ZM8bmfQAAAAAAANuZTCbdcMMNki6uacwVBOsOo6t3cGmR7ksyMzM1ceJElZSU6O6779bTTz+tgQMHKjc3VxMmTLAcdQQAAAAAAADv5fbT37766itFRUVp4sSJCg0Ntdw/aNAgvfLKK5o7d65eeukld8cCAAAAAACADVx+pNJvVVRU6NChQxo0aFCVQkmSIiMj1bt3b/3000/ujAQAAAAAAAA7uLVUunDhgiSpQYMGNW6PjY1VSUmJOyMBAAAAAADADm45/W3btm0qLCyUJAUGBio/P7/GcQUFBQoJCXFHJAAAAAAAADjALaXS2rVrtXbtWsvtTZs2qXfv3tXG7dq1SwkJCe6IBAAAAAAAAAe4vFT6+OOPq91nMBiq3VdUVKQGDRqoXbt2ro4EAAAAAAAAB7m8VIqJibFqXFhYmEaOHOniNAAAAAAAAHAGty7UDQAAAAAAgLrBLWsqSRcX6167dq0OHTqk/Px8lZWVKTAwUPHx8br55pt15513Kioqyl1xAAAAAAAA4ACXl0rnz5/X+++/r507d1a538fHR/Hx8crLy9O8efO0ePFi/c///I9SU1NdHQkAAAAAAAAOcnmplJGRoV27dunJJ59Ut27dFBAQoL179+qzzz5TcnKyHnnkER07dkyzZs3Sxx9/rMjISCUnJ7s6FgAAAAAAABzg8jWV1q1bp759+6pfv36qV6+e/P391a5dOw0fPlzffPONzpw5o4SEBL366qtq3ry55s6d6+pIAAAAAAAAcJDLS6WzZ88qISGh2v0JCQmqrKzUiRMnJEkGg0Gpqak6cOCAqyMBAAAAAADAQS4vlWJjY7V79+5q9+/Zs0eSFB4ebrnPx8dHZrPZ1ZEAAABqDZPJpPj4eBkMBplMJk/HAQAAsHD5mkp9+/bVtGnT5O/vr65du8rf319ZWVlasGCBWrdurbi4OMvYw4cPV7kNAAAAAAAA7+TyUqlfv37Ky8vTt99+q5UrV1rub9GihV544YUqY/39/dW/f39XRwIAAAAAAICDXF4qSdKjjz6qQYMG6eeff1Z5ebni4+PVtGnTauOGDRvmjjgAAAAAAABwkFtKJeni2kkdO3Z01+4AAAAAAADgQi5fqBsAAAAAAAB1D6USAAAAAAAAbEapBAAAAAAAAJtRKgEAAAAAAMBmlEoAAAAAAACwGaUSAAAAAAAAbOb2Uqm0tFRTpkxRdnZ2tW0///yzpkyZ4u5IAAAAAAAAsJHbS6Xy8nKtXLlSBQUF1badOnVKK1eudHckAAAAAAAA2MjX0wEAAKjLTCaTbrjhBknSgQMHFBQU5OFEAADA28Tv2OXpCC5VUlpq+Tpu526FBAZ6MI3rHW93s6cjuA1rKgEAAAAAAMBmbjtSKS8vT5JUXFwsSTpz5ozlPkmKjo52VxQAAAAAAAA4yG2l0qhRo6rc/tvf/lbldkZGhruiAAAAAAAAwEFuK5XGjh0rSSopKdFf/vIXPfbYY2rWrJm7dg8AAAAAAAAncluplJycLEk6e/asJCkpKclyHwAAAAAAAGoXrv4GAABqtctPqa9rysrKLF9PnjxZ/v7+Hkzjes8//7ynIwAAACtx9TcAAAAAAADYzO2lko+Pj5KTkxUSElJtW/369TklDgAAAAAAoBZw++lvwcHBlkW7L9e2bVu1bdvWzYkAAAAAAABgK05/AwAAAAAAgM1cWiqdP3/eI48FAAAAAACAa7m0VBoxYoTmzJmjgoICqx+Tn5+vjIwMjRw50oXJAAAAAAAA4AiXrqk0fPhwff3115ozZ45atWqlm2++Wc2aNVNsbKxCQkJkNptVUlKinJwcHTx4ULt27dL+/fvVsGFDPf30066MBgAAAAAAAAe4tFTq2rWrOnfurC1btmjFihWaN2+eKioqag7i66u2bdvqpZde0q233iqj0fuXe9p9qkRnTJym5+0MBoMiSvxUUHBWZrPZ03FQCzGH4IjScybL11uyzyogsObPQdgv1y/G0xFcqtz839818vyi5ecX4ME0rrfp2FlPR6iT+CyDI/gsc73IswZPR3Cpc6X/fX1rzhoUVF63X29+Lf8sqx8coNtjrPv9yuVXfzMajerYsaM6duyo8vJyHTp0SNnZ2SouLpYkhYaGqlGjRmrWrJn8/PxcHQcAAAAAAABO4PJS6bf8/PzUqlUrtWrVyp27BQAAAAAAgJO5tVQCAAAAgLpm9ZFJno7gUuWl5Zav1x75SH6BdfsMk7TEFz0dAag1vH/hIgAAAAAAAHgdSiUAAAAAAADYjFIJAAAAAAAANmNNJQcYy8tkLCv1dAxcg8FgkOF8qYxlpVxCF3ZhDsERxrLzCvIxWr42Guv2JXQ9we9CHb+0dWWFZQ75VVbI74KPhwO5Fr9buQafZa7lW37B0xFcq+LCf9+HKi7It7xuH5vgifchQ3mZ2/fpToaKMsscMlSUycAc8mpGP+t/XzWY+VSx28nnH1H5wX2ejgEAAAAAAOAUfs1bKe5vX1g11u1HKuXl5Wnu3Lnas2ePioqK9Morryg5OVlFRUWaM2eObr/9djVt2tTdsQAAAAAAAGADt5ZKx44d0xtvvCGz2awWLVro5MmTqqyslCSFhYVp3759On/+vEaMGGHT85aWlmrhwoXav3+/Dhw4oJKSEo0cOVI9e/asMm7y5MlauXJltcfHx8frgw8+sPn17HtyrM6UnLP5cXAvg8Gg8PBIFRbmc7g37MIcgiNKz53TsP7dJEnTl6xTQGCghxPVPYsXL/F0BJcqLz+vryaNlyQNfXGs/PwCPJzItQYM6OfpCHUSn2Wutfbox56O4FIV58v19+EzJUkjPn9MvgF+Hk7kWt0aj3b7PiMOH3X7Pt3p3PlS3fPsg5Kk+Z9mKCigbv8+VJDU2NMRHFI/JEhxVo51a6k0a9YshYSE6O2335YkPfPMM1W2p6SkaP369TY/76WjnKKjo5WUlKQ9e/Zccayfn5+effbZKvcFBwfbvE9JqvTzV6U/H8rezmAwyBwQqEr/QH6Jgl2YQ3BE5YVKnbtw8Q8olf4BqvSv279EeUK5T91eIrL8wgXLHCo3+kp1/PXyM+IafJa5VoVf3V7rrPw3n2Xlvj5SHX+9nngfMvv5u32f7mT+zRwy+/rX+ddb2z/LKm34/rj1t5KffvpJ9913n8LCwnT27Nlq26Ojo5Wfn2/z80ZERGjq1KkKDw/XwYMH9dprr11xrNFoVPfu3W3eBwAAAAAAAP7LrUuuV1ZWKiDgyodsFxUVydfX9p7Lz89P4eHhNuUwmUw27wcAAAAAAAAXufVIpWbNmmnbtm3q27dvtW0XLlzQunXr1LJlS5dmKCsr0xNPPKHz588rJCRE3bp106OPPqpA1rgAAAAAAACwmltLpXvuuUfvvvuuPvvsM3XrdnHR0sLCQmVmZmrevHnKzs7WU0895bL9R0REaPDgwWratKnMZrN27NihpUuX6tdff9W4cePk41PzucHl5eUqLy+33DYYDAoKCpLBYJDBYHBZXjiHQYb//pdvF+zAHIIjfvs5YRCfG8C18DPiGnyWAdbjfQiOqu1zyJb8bi2VUlJSNGrUKE2bNk3Lly+XJH300UeSpKCgII0aNUrJycku2//DDz9c5Xa3bt3UsGFD/fvf/9aGDRssRdfl5s2bpzlz5lhuN23aVBMnTlRYWJiMARdclhfOFR4R4ekIqOWYQ7DHud+c9l0/PFxBwSEeTAN4v4iISE9HqNP4LHORXz0dAM7kmfehwx7YJ1yltn+WhQZYvxi/2y8f0r17d3Xs2FGZmZk6efKkKisrFRcXp1tuuUVBQUHujqNBgwYpIyNDu3btumKpNGTIEA0aNMhy+1Jr1ziwQhW+FW7JCfsZDAbFxcXp5MmTXO0EdmEOwREmw38/J24IvaCgoPKrjIY9YspzPR3BpcrKyyxfR5fnyd9Qt6+Yc0MIPyOuwGeZayXU3+/pCC513v+/f0iPDzuogKC6ffU3T7wPNaxXt38uS/z++/pS65kVEli3X++JWv5Z5utr/ffHI9ekDQwMVMeOHT2x62r8/f1Vr149FRcXX3GMn5+f/Pz8atzGh3LtYTab+X7BIcwh2OO3c4Y5BFwbPyOuxfsQcG38jMBR19MccmuplJeXZ9W46OhoFyf5r3Pnzuns2bMKCwtz2z4BAAAAAABqO7eWSqNGjbJqXEZGhtP3XVZWpgsXLlQ7xe4///mPzGaz2rVr5/R9AgAAAAAA1FVuLZVGjBhR7b7Kykrl5uZq1apVCgsLU9++fe167iVLlqikpEQFBQWSpC1btuj06dOSpP79+6u4uFivvvqqunXrpvj4eEnSzp07tX37drVr10633nqrna8KAAAAAADg+uPWUqlnz55X3Hb33XdrzJgxMplMdj33okWLlJv734U6N23apE2bNkmS0tLSFBISovbt2yszM1MrV660LBA+dOhQ3XXXXTIajXbtFwAAAAAA4HrkkYW6axIYGKiePXvq22+/1YABA2x+/OTJk6855rnnnrMnGgAAAAAAAC7jNaWSdHGF9MLCQk/HAAAA8Br+/v569dVXPR0DAACgGq8olUwmk3766SctXLhQTZs29XQcAAAAAAAAXINbS6UHH3zwqtujo6M1fPhwN6UBAAAAAACAvdxaKt13330yGAxV7jMYDAoJCVGDBg10yy23yMfHx52RAAAAAAAAYAe3lkrp6enu3B0AAAAAAABcxOjpAAAAAAAAAKh9XHqk0pQpU2x+jMFg0IgRI1yQBgAAAAAAAM7i0lJpz549Nj/m8jWXAAAAAAAA4H1cWipNnjzZlU8PAAAAAAAAD2FNJQAAAAAAANiMUgkAAAAAAAA2c+npbzXZvn27vvnmG/3yyy8ymUwym83VxmRkZLg7FgAAAAAAAGzg1iOVNmzYoHfffVdnzpxR165dZTab1a1bN3Xr1k3+/v5q0qSJ7r//fndGAgAAAAAAgB3ceqTS/Pnz1aJFC7311lsqLi7WsmXLdMcdd6hNmzbKycnRmDFjFBsb685IAAAAAAAAsINbj1Q6duyYunXrJqPRKB8fH0lSRUWFJCk2NlZ9+/bVggUL3BkJAAAAAAC4UEhgoMzfLpH52yUKCQz0dBw4kVtLpYCAAPn6Xjw4KiQkRL6+viosLLRsr1+/vnJyctwZCQAAAAAAAHZwa6kUHx+vY8eOWW4nJSVp1apVunDhgsrKyrRmzRpFR0e7MxIAAAAAAADs4NZS6bbbbtPmzZtVXl4uSbr33nu1Z88eDRs2TMOHD1dWVpbuueced0YCAAAAAACAHVy+UPfWrVuVkpIio9GowYMHa/DgwZZtHTp00Lhx47Rx40YZjUa1b99ebdq0cXUkAAAAAAAAOMjlpdJ7772nevXqqUuXLkpLS1PLli2rbL/xxht14403ujoGAAAAAAAAnMjlpdIzzzyjtWvXatmyZVq6dKliYmKUmpqq1NRUJSQkuHr3AAAvtyij0NMRXOr8eZPl62+/LlBAwHkPpnG9ux4M93QEAAAAuInLS6XevXurd+/eys/P19q1a7V27VrNmzdP8+bNU1JSktLS0tS1a1dFRka6OgoAAAAAAACcxOWl0iWRkZG66667dNddd+n48eNas2aN1q5dq5kzZ2rWrFlKTk5WWlqaOnXqpODgYHfFAgAAAAAAgB3cVir9Vnx8vNLT05Wenq4DBw5o7dq1WrdunT755BP985//1MyZMz0RCwAAAAAAAFYyejpAbGysGjRooKioKElSWVmZhxMBAAAAAADgWjxypFJpaak2btyoNWvWaPfu3aqsrFRMTIyGDBmi1NRUT0QCAAAAAACADdxWKlVUVGjbtm1as2aNtm/frrKyMoWGhqpXr15KTU1V69at3RUFAAAAAAAADnJ5qbR7926tWbNGGzdulMlkkp+fnzp06KC0tDSlpKTIx8fH1RFwFReeGezpCG5x1NMB3MTns4WejgAAAAAAuE64vFR66623ZDAYdNNNN1mu7hYUFOTq3QIAAAAAAMCFXF4qPfbYY+rWrZsiIiJcvSsAAAAAAAC4ictLpUGDBrl6FwAAAAAAAHAzo6cDAAAAAAAAoPahVAIAAAAAAIDNKJUAAAAAAABgM0olAAAAAAAA2MytpdK6detUVlbmzl0CAAAAAADABVx+9bff+vDDDxUUFKROnTopLS1Nbdq0cefuAQAAAAAA4CRuLZXefPNNrV69Whs2bNCKFSsUGRmp1NRUpaWlKTEx0Z1RAAAAAAAA4AC3lkqtWrVSq1at9OSTT2rHjh1avXq1lixZooULFyoxMVHdu3dXamqqIiIi3BkLAAAAAAAANnJrqXSJj4+POnTooA4dOqi0tFSbNm3SihUrNGvWLH355ZdKTk5Wjx491LVrV/n6eiQiAAAAAAAArsLjV387cuSIDhw4oCNHjkiS4uPjVVxcrMmTJ+u5555TVlaWhxMCAAAAAADgch45DOj48eNas2aN1qxZo1OnTiksLEypqanq3r27mjVrJkk6ePCgPvnkE3322Wd6//33PRETAAAAAAAAV+DWUmnx4sVavXq1Dh06JD8/P3Xo0EHDhg1Tu3btZDRWPWiqefPmGjhwoD755BN3RgQAAAAAAIAV3FoqTZ8+Xa1atdIzzzyjrl27Kjg4+Krjmzdvrvvuu89N6QAAAAAAAGAtt5ZKH330kWJjY60e37hxYzVu3NiFiQAAAAAAAGAPty7UHRUVJZPJdMXtJpNJFy5ccGMiAAAAAAAA2MOtpdK0adP0+uuvX3H766+/rhkzZrgxEQAAAAAAAOzh1lJpx44d6tSp0xW3d+7cWdu3b3djIgAAAAAAANjDraVSQUGBIiMjr7g9IiJC+fn5bkwEAAAAAAAAe7i1VAoNDdXx48evuD07O1tBQUFuTAQAAAAAAAB7uLVUateunZYvX65ffvml2rZDhw5p+fLlSklJcWckAAAAAAAA2MHXnTt78MEHtWPHDv3xj39Uhw4d1LhxY0nS0aNHtXXrVoWFhenBBx+0+XlLS0u1cOFC7d+/XwcOHFBJSYlGjhypnj17Vht77NgxTZ8+XVlZWfL19VX79u31xBNPKCwszNGXBwAAAAAAcN1wa6kUGRmpd999V1988YW2bNmizZs3S5KCgoKUmpqqoUOHXnXNpSspKirSnDlzFB0draSkJO3Zs6fGcadPn9bYsWMVHBysoUOHqrS0VIsWLdKRI0f0zjvvyNfXrf8cAAAAAAAAtZbbW5SIiAiNHj1aZrNZRUVFkqSwsDAZDAaHnnPq1KkKDw/XwYMH9dprr9U4bt68eTp//rwmTpyo6OhoSVKLFi00YcIErVixQr1797Y7AwAAAAAAwPXErWsq/ZbBYFD9+vVVv359hwolSfLz81N4ePg1x23cuFHt27e3FEqS1LZtWzVs2FDr1693KAMAAAAAAMD1xKVHKs2ZM0eSdO+998poNFpuX8v999/v9Cz5+fk6c+aMmjdvXm1bixYttH37dqfvEwAAAAAAoK5yaan09ddfS5LuueceGY1Gy+1rcUWpVFBQIOniqXKXi4iIUHFxscrLy+Xn51dte3l5ucrLyy23DQaDgoKCLF8D3oL56BqX/l359wWujZ8TOIo55Bp8lgHW4+cEjrqe5pBLS6WMjIyr3nansrIySapxMe5LRVJZWVmNpdK8efOqHGXVtGlTTZw4UTExMS5K6z5HPR0ATtWwYUNPR6jT4uLiPB2hjirwdAA4Ee9DcBRzyLX4LHOR3Z4OAGfyyPvQ9kz37xMucz19ll03lzvz9/eXJFVUVFTbdukopEtjLjdkyBANGjTIcvtS65ibm1vj8wGecuLECU9HqJMMBoPi4uJ08uRJmc1mT8cBvBrvQ3AUc8g1+CwDrOeJ96Hrp4K4PtT2zzJfX1+rD6K5bkqlS6e9XToN7rcKCgoUGhpa41FK0sUjma60jQ9leBPmo2uZzWb+jYFr4GcEjmIOuRafZcC18TMCR11Pc8ilpdKoUaNsPpfQYDDoo48+cnqWyMhIhYWF6eDBg9W2HThwQElJSU7fJwAAAAAAQF3l0lIpOTnZqxao6tSpk1auXKm8vDxFR0dLknbt2qUTJ05o4MCBHk4HAAAAAABQe7j8SCV3WbJkiUpKSiynt23ZskWnT5+WJPXv31/BwcEaMmSINmzYoPHjx2vAgAEqLS3VwoULlZiYqNtvv91tWQEAAAAAAGq7OrOm0qJFi5Sbm2u5vWnTJm3atEmSlJaWpuDgYEVHR2vcuHGaMWOGvvzyS/n6+iolJUWPP/74FddMAgAAAAAAQHVuL5UqKyu1atUqbdu2TXl5eZKk6OhodejQQWlpaTIajXY97+TJk60a17hxY40ZM8aufQAAAAAAAOAit5ZKJpNJb7/9tg4cOKCgoCA1aNBA0sV1jTZu3KilS5dqzJgxCg4OdmcsAAAAAAAA2MitpdJXX32lQ4cO6amnnlKvXr3k63tx9xUVFfrhhx80bdo0/fvf/9ZTTz3lzlgAAAAAAACwkX3nmtlp06ZN6tOnj/r27WsplCTJ19dXffr00Z133qmNGze6MxIAAAAAAADs4NZSqbi4WPHx8Vfc3qhRIxUXF7sxEQAAAAAAAOzh1lIpLi5OW7ZsueL2LVu2WNZZAgAAAAAAgPdya6nUp08fZWZm6p133tHOnTuVk5OjnJwc7dixQ++8844yMzPVr18/d0YCAAAAAACAHdy6UHffvn115swZLViwQDt27KgaxNdX999/v/r06ePOSAAAAAAAALCDW0slSUpPT1e/fv2UmZmpvLw8SVJMTIxuvvlmhYWFuTsOAAAAAAAA7OCWUqmwsFArVqxQTk6OwsLC1KlTJ6Wmprpj1wAAAAAAAHABl5dKOTk5eu2116pc1W3+/PkaPXo0xRIAAAAAAEAt5fJSafbs2SotLdWTTz6pNm3a6OTJk5o2bZqmT5+url27ymh061rhAAAAAAAAcAKXl0r79u1T7969LVd1S0hIkNFo1MSJE5Wdna3GjRu7OgIAAAAAAACczOWHCeXl5alp06ZV7mvWrJkk6ezZs67ePQAAAAAAAFzA5aVSZWWlfH2rHhDl4+Nj2QYAAAAAAIDaxy1Xfzt48KD8/Pwst8+dOydJysrKUklJSbXxnTp1ckcsAAAAAAAA2MktpdLixYu1ePHiavd//fXXNY7PyMhwdSQAAAAAAAA4wOWl0tixY129CwAAAAAAALiZy0ul5ORkV+8CAAAAAAAAbubyhboBAAAAAABQ91AqAQAAAAAAwGaUSgAAAAAAAP9fe3ceHkWZ7n3811nJQkggCWFPWIJGBMIWdtlkE1BGiSjOiCMcBRxHHdTj4KCoKOh4GEdAD27HDU3IYRVkADWssod92EUEAklIQshGOnS/f/jSx5CwdDpVneX7uS4uuqqe6ufuzp2qzt1PPQWnUVQCAAAAAACA0ygqAQAAAAAAwGkUlQAAAAAAAOA0LyOffO3ateXa74477qjgSAAAAAAAAFCRDC0qzZ07t1z7UVQCAAAAAACo3AwtKs2ePbvEcl5enubMmSN/f38NGTJEDRs2lCSdPn1aK1euVEFBgSZNmmRkSAAAAAAAAKgAhhaVwsLCSiwvWLBAQUFBevHFF2WxWBzrmzZtqri4OE2fPl3Lly/XxIkTjQwLAAAAAAAALjK0qHS1bdu2afTo0SUKSld4eHioS5cu+vrrr80MCQAAQ/n6+uv9/9rv7jAAACg3Xz9P/deqfu4OA0AlZOrd3+x2u06fPn3N7adOnTIxGgAAAAAAAJSXqUWlzp07a/Xq1frmm2906dIlx/pLly5p2bJlWrNmjTp16mRmSAAAAAAAACgHUy9/e+SRR5SWlqbPP/9c8+fPV0hIiCQpKytLly9fVuvWrTV27FgzQwIAAAAAAEA5mFpU8vf317Rp07Rt2zalpKQoIyNDktSuXTt16NBBHTt2LHO+JQAAAAAAAFQuphaVrujcubM6d+7sjq4BAAAAAABQAdxSVMrMzNSBAweUk5OjuLg41atXTzabTfn5+fL395eHh6lTPQEAAAAAAMBJphaV7Ha7PvvsM61cuVI2m02S1LRpU9WrV0+FhYWaNGmS4uPjddddd5kZFgAAAAAAAJxk6pCgpUuXasWKFRo+fLhefPHFEtv8/f3VpUsXbdmyxcyQAAAAAAAAUA6mFpW+++473XHHHXrwwQcVGRlZanuzZs2UmppqZkgAAAAAAAAoB1OLSufPn1d0dPQ1t/v6+io/P9/EiAAAAAAAAFAephaVgoKCdP78+WtuP378uEJDQ02MCAAAAAAAAOVhalEpLi5Oq1ev1rlz50pt2717t5KTk9WtWzczQwIAAAAAAEA5mHr3t/j4eO3fv1/PPfecbrnlFknSkiVLlJCQoMOHDysqKkojR440MyQAAAAAAACUg6kjlfz9/TV9+nSNGDFCmZmZ8vHx0YEDB5Sfn69Ro0bplVdeka+vr5khAQAAAAAAoBxMHakkST4+Prr33nt17733mt01AAAAAAAAKoipI5WmTZumvXv3XnP7vn37NG3aNBMjAgAAAAAAQHmYWlQ6cOCALly4cM3tOTk5OnDggIkRAQAAAAAAoDxMLSrdyNmzZ+Xn5+fuMAAAAAAAAHADhs+plJycrLVr1zqWFy5cqO+++65Uu/z8fP3888+KjY01OiQAAAAAAAC4yPCiUlFRkXJychzLBQUFslgsJdpYLBb5+vrqzjvv1H333Wd0SAAAAAAAAHCR4UWlgQMHauDAgZKkSZMm6ZFHHlGnTp2M7hYAAAAAAAAGMryo9Ftz5swxszsAAAAAAAAYxNSi0m8VFBQoPz9fdru91LbQ0FBD+ty/f7+mTZtW5rbXXntN0dHRhvQLAAAAAABQ3ZheVFq1apW++eYbnTt37pptEhISDI1hyJAhatGiRYl1ERERhvYJAAAAAABQnZhaVFq1apU++ugjtWvXTn379tXXX3+tu+66S97e3kpOTlZwcLCGDBlieBy33nqrunbtang/AAAAAAAA1ZWHmZ2tXLlS7dq101//+lcNGDBAktShQwc98MADmjVrlgoKCnTx4kVTYikoKNDly5dN6QsAAAAAAKC6MXWk0rlz5zRo0CBJkqenpySpuLhYkuTv769+/fpp1apVGj58uKFxzJ07V4WFhfLw8NCtt96qhx56qNTlcAAAAAAAALg2U4tK/v7+jtFB/v7+8vHxUUZGhmO7n5+fsrOzDevfy8tLcXFxio2NVVBQkE6dOqVly5Zp6tSpeu211xQVFVXmflarVVar1bFssVjk5+fneAxUFuSjMa68r7y/wI3xewJXkUPG4FwG3Dx+T+CqmpRDphaVmjRpop9//tmxHB0drdWrV6tDhw6y2Wxas2aNGjRoYFj/rVu3VuvWrR3LnTp1UteuXTV58mTNnz9fU6ZMKXO/RYsWKSkpybEcFRWlmTNnKiwszLBYzfKLuwNAhTLy9wdM6G+cLHcHgArEcQiuIoeMxbnMIPvcHQAqkluOQyl7zO8ThqlJ5zJTi0q9evXS6tWrZbVa5e3trVGjRunVV1/VhAkTfg3Gy0t/+ctfzAxJERER6tSpk7Zu3SqbzSYPj9LTTI0cOVLDhg1zLF+pOqanpzsu3wMqg9TUVHeHUC1ZLBZFRETo7Nmzstvt7g4HqNQ4DsFV5JAxOJcBN88dx6GaU4KoGar6uczLy+umB9GYWlTq27ev+vbt61i+5ZZb9F//9V/asWOHPDw81LZtWzVs2NDMkCRJoaGhKi4uVmFhofz9/Utt9/b2lre3d5n7clJGZUI+Gstut/MeAzfA7whcRQ4Zi3MZcGP8jsBVNSmHTC0qlaV+/foaOnSoW2M4d+6cvL29VatWLbfGAQAAAAAAUFW4rahks9mUn59f5rbAwEBD+szJyVFQUFCJdSdOnND27dsVGxtb5qVvAAAAAAAAKM3UolJxcbGWLFmiH374QefPn5fNZiuzXUJCgiH9z5o1Sz4+PmrdurXj7m/fffedfH199eCDDxrSJwAAAAAAQHVkalFp3rx5Wrt2raKjo9W5c+cy5y8yUufOnbVhwwZ98803KigoUFBQkLp06aJRo0ZxJwwAAAAAAAAnmFpU2rx5s3r37q1JkyaZ2a3D0KFD3T5/EwAAAAAAQHVg6iRCvr6+atWqlZldAgAAAAAAwACmFpV69OihnTt3mtklAAAAAAAADGDq5W8PPfSQ5s6dqxkzZqhv376qV69emXdca968uZlhAQAAAAAAwEmmFpWsVqvsdrtSUlKUkpJyzXZG3f0NAAAAAAAAFcPUotJ7772nrVu3qkePHmrZsqXpd38DAAAAAABAxTC1qLR7924NHjxYY8eONbNbAAAAAAAAVDBTJ+r28/NTRESEmV0CAAAAAADAAKYWlfr376+NGzfKZrOZ2S0AAAAAAAAqmKmXvzVu3Fjbt2/X888/rzvuuOOad3+Li4szMywAAAAAAAA4ydSi0j/+8Q/H488///ya7bj7GwAAAAAAQOVmalHppZdeMrM7AAAAAAAAGMTUolJMTIyZ3QEAAAAAAMAgpk7UDQAAAAAAgOrB0JFK06ZNk8Vi0ZQpU+Tp6alp06bdcB+LxaKpU6caGRYAAAAAAABcZOhIJbvdLrvdXmL5ZvYBAAAAAABA5WboSKWXX375ussAAAAAAAComkydU+nAgQPKycm55vacnBwdOHDAxIgAAAAAAABQHqYWlaZNm6Y9e/Zcc/u+fftuat4lAAAAAAAAuFeluvub1WqVh0elCgkAAAAAAABlMHROJUnKyMhQWlqaY/n06dNlXuKWn5+vNWvWKCwszOiQAAAAAAAA4CLDi0o//PCDkpKSHMsLFy7UwoULy2zr4eGh8ePHGx0SAAAAAAAAXGR4Ualbt25q0qSJJGnWrFkaMmSIbrnllhJtLBaLfH19FRkZqeDgYKNDAgAAAAAAgIsMLyo1btxYjRs3liRNmDBBMTExCg8PN7pbAAAAAAAAGMjwotJv9enTp8z1586dk9VqdRSfAAAAAAAAULmZWlT69ttvdejQIT311FOOdXPnztXatWslSVFRUXrhhRdUp04dM8MCAAAAAACAkzzM7Oy7774rUTDatWuX1q5dqwEDBuiPf/yjzp07pwULFpgZEgAAAAAAAMrB1JFK6enpatSokWP5xx9/VHh4uOOOb9nZ2Vq3bp2ZIQEAAAAAAKAcTB2pdLU9e/aoffv2juWwsDBlZ2e7LR4AAAAAAADcHFOLSg0aNNC2bdsk/XrpW2ZmpmJjYx3bMzMzFRAQYGZIAAAAAAAAKAdTL38bPny4/vnPf+qRRx5RYWGhGjdurHbt2jm279u3T5GRkWaGBAAAAAAAgHIwtajUo0cP1a5dWzt37lRAQIAGDRokT09PSVJubq4CAwPVu3dvM0MCAAAAAABAOZhaVJKktm3bqm3btqXWBwYGavLkyWaHAwAAAAAAgHIwfE6lo0ePKjc396bapqWlae3atQZHBAAAAAAAAFcZXlSaMmWKdu3a5VjOzc3VQw89pAMHDpRqe+jQIc2dO9fokAAAAAAAAOAiU+/+Jkl2u11Wq1U2m83srgEAAAAAAFBBTC8qAQAAAAAAoOqjqAQAAAAAAACnUVQCAAAAAACA07zM6CQtLU3Hjx+XJOXn50uSUlNT5e/vX6odAAAAAAAAKj9TikoJCQlKSEgose7DDz80o2sAAAAAAAAYwPCi0oQJE4zuAgAAAAAAACYzvKjUp08fo7sAAAAAAACAyZioGwAAAAAAAE6jqAQAAAAAAACnUVQCAAAAAACA0ygqAQAAAAAAwGkUlQAAAAAAAOA0ikoAAAAAAABwGkUlAAAAAAAAOI2iEgAAAAAAAJzm5e4AzGa1WpWQkKD169crNzdXzZo10+jRo9W2bVt3hwYAAAAAAFBl1LiRSnPmzNHy5cvVs2dPPfLII/Lw8NAbb7yhgwcPujs0AAAAAACAKqNGFZWOHj2qTZs26YEHHtDvf/97DRgwQFOnTlVoaKi++OILd4cHAAAAAABQZdSootLmzZvl4eGhAQMGONb5+PioX79+Onz4sDIyMtwYHQAAAAAAQNVRo4pKP/30kxo0aCB/f/8S61u2bClJOnHihBuiAgAAAAAAqHpq1ETd2dnZCgkJKbX+yrqsrKwy97NarbJarY5li8UiPz8/eXlV/bfPo0Vrd4eACuTp7e3uEKoli8UiSfL29pbdbndzNNVPaHgtd4eACuTthuNQw4YNTe8TxnFHDtUEnMuMFV67lbtDQAVyy3EoqLb5fcIwVf1c5kyto+pXRZxQVFRU5g/3yrqioqIy91u0aJGSkpIcyz169NCf//znMgtUVc4/v3R3BECVERoa6u4QqqV7x4S5OwRUcU8++aS7QwCqDM5lxhgT9r67Q0BVF8bnoeqkJv00a9Tlbz4+PiVGHF1xZZ2Pj0+Z+40cOVL/8z//4/g3fvz4Mp8HlVNBQYGef/55FRQUuDsUVFHkEFxFDsFV5BBcRQ7BVeQQXEUOVU81aqRScHCwMjMzS62/ctnbtUYeeXt7V/nhazWZ3W7XTz/9xFBvlBs5BFeRQ3AVOQRXkUNwFTkEV5FD1VONGqkUGRmp1NRU5efnl1h/5MgRx3YAAAAAAADcWI0qKnXt2lU2m01r1qxxrLNarUpOTlarVq24xhwAAAAAAOAm1ajL31q1aqWuXbvqq6++Uk5OjiIiIrR27Vqlp6fr8ccfd3d4MIi3t7fuu+8+LmFEuZFDcBU5BFeRQ3AVOQRXkUNwFTlUPVnsNeyCxqKiIiUkJGj9+vXKy8tT06ZNdf/996t9+/buDg0AAAAAAKDKqHFFJQAAAAAAALiuRs2pBAAAAAAAgIpBUQkAAAAAAABOo6iEaiUxMVHx8fHuDgPVxKRJkzRnzhx3h4FqIj4+XomJiY7l5ORkxcfHKy0tzY1Roarg/AajzZkzR5MmTXJ3GKgmXn75Zb388suO5bS0NMXHxys5OdltMaFq2L9/v+Lj47V//353h4KbRFEJQI136NAhJSYmKi8vz92hAABgmMzMTCUmJurEiRPuDgUAUE14uTsAAHC3Q4cOKSkpSX369FFAQIBj/T/+8Q9ZLBY3RobqrHfv3urevTu31QVgmqysLCUlJSk8PFyRkZEltj322GPi/j0wSlhYmL744gt5efHnJ1Dd8FsNANfAH/swkoeHh3x8fNwdBqoJm82m4uJicgrlxh/7MJLFYuH4hAp16dIl+fr6ujsMiKISqrCDBw/q008/1cmTJ1W3bl2NGDGiVJv4+HgNGjRIMTExSkxMVFpamiIjI/XYY4+padOmWr16tZYuXarMzEy1atVKEydOVHh4uBteDdwlMTFRSUlJkqQnnnjCsX727NmaNm2aYmJiHHNMFBcXa9GiRVq/fr3Onz8vX19fNWrUSKNGjVLbtm3dEj/McyVXZs2apYSEBO3evVuenp7q1auXxowZ4/iwbLVa9eWXX2r9+vWyWq267bbbNG7cuFLPl5ycrLlz52r27NmO486xY8f09ddf6/jx4yosLFRwcLBuu+02TZw40dTXCvdy5vwWHR2tRYsWKTU1VU8//bS6dOnihohhtIKCAiUkJGjbtm3KysqSv7+/mjVrpjFjxqh58+Z6+eWXdfHiRT399NP66KOPdOTIEQUEBGjo0KG6++67Jf06T8m0adMkSXPnztXcuXMlSRMnTlSfPn00Z84cHThwoMRcghs3btTSpUuVmpoqi8Wi0NBQ9e/fX0OHDjX/TcANVUSeSP+XK0899ZROnDihH374QYWFhWrTpo0effRRhYaGluh3zZo1WrJkiTIzM9W0aVP94Q9/KBVbWlqannjiCUe+SVJ2drbmz5+vPXv2KCcnR4GBgWrZsqXGjh3L5/FKrKLyTJLOnz+vjz76SHv37pWvr6969uyp9u3bl+rzynNOmjRJn376qY4dO6YBAwZo7Nix5rxoXBdFJVRJJ0+e1GuvvaagoCCNGjVKly9fVmJiooKDg0u1PXjwoLZv365BgwZJkhYvXqwZM2ZoxIgRWrVqlQYNGqTc3FwtXbpU7733nl566SWTXw3cKS4uTqmpqdq4caMefvhh1a5dW5IUFBRUqu2CBQu0ePFi9evXTy1btlRBQYGOHTum48ePU1SqQWbNmqWwsDA98MADOnLkiL799lvl5eU5ipLvv/++1q9fr549eyo6Olr79u3TjBkzbvi8Fy5ccBzX7r77bgUEBCg9PV1btmwx+iWhEnHm/LZv3z79+OOPGjx4sGrXrs0fYdXYBx98oM2bN2vw4MFq3LixLl68qIMHD+r06dNq3ry5JCk3N1fTp09XXFycunXrps2bN+vLL79U06ZNFRsbq0aNGjluGDBgwADdcsstkqTWrVuX2eeePXv0zjvv6Pbbb9eYMWMkSadPn9bBgwcpKlVSFZEnv7Vw4UJZLBbdfffdysnJ0fLly/Xqq6/qrbfecnyR8v3332vevHlq3bq1hg4dqnPnzunNN99UYGCg6tWrd9143377bf3yyy8aPHiwwsPDdeHCBe3Zs0cZGRkczyqxisqzoqIivfLKK8rIyNCQIUNUt25drVu37poTdF+8eFGvv/66unfvrl69eqlOnTqmvWZcH0UlVEkJCQmy2+165ZVXHN+WxMXFafLkyaXanjlzRrNmzXKcnAIDAzVv3jwtXLhQ77zzjvz8/CT9eunA4sWLlZaWxomsBmnWrJmioqK0ceNGde7c+bo/+507dyo2NlaPPfaYiRGisgkPD9dzzz0nSRo8eLD8/Py0atUqDR8+XHa7XevXr9fAgQMdo5MGDx6sf/7zn/r555+v+7yHDh1SXl6eXnzxRbVo0cKxfvTo0ca9GFQ6zp7f3n77bTVu3NjsMGGynTt3qn///iVGgFz9jX9WVpaeeOIJ9e7dW5LUr18/TZw4Ud9//71iY2MVHBys2NhYJSYmKjo62tHuen36+flpypQp8vDg3j5VQUXkyW/l5uZq1qxZjs/KUVFRmjVrltasWaOhQ4equLhYX331lSIjI/XSSy85LqFs3Lix5s2bd92iUl5eng4dOqSHHnqoxGjMkSNHuvYmwHAVlWdr1qxxjLLt1q2bJKl///569tlny+w3Oztb48eP15133mnEy4ILOEOgyrHZbNq9e7c6d+5cYvht48aN1a5du1Lt27RpU6JQ0LJlS0m/fki/cpKUpFatWkkSt/fGNQUEBOiXX35Ramqqu0OBG10Z9XjFkCFDJEkpKSlKSUmRpFLf4t/Mt/pXJonfsWOHiouLKyJUVDHOnt9iYmIoKNUQAQEBOnr0qDIzM6/ZplatWurVq5dj2cvLSy1btiz35xp/f39dunRJe/bsKdf+MF9F50nv3r1LfFbu2rWrQkJCHOe648eP68KFC7rzzjtLzMnVp08f+fv7XzdWHx8feXl56cCBA8rNzb3p1wj3q6g8S0lJUUhIiLp27epY5+vrqwEDBpT5nN7e3urbt28FvAJUNIpKqHJycnJUVFSkBg0alNrWsGHDUuuuvu77yknu6m9PrqznxIZriY+PV35+vv785z/rL3/5iz7//PMbjj5B9XP1sad+/fqyWCxKS0tTenq6LBaL6tevX6JNWcemq8XExCguLk5JSUl69NFH9eabb+qHH36Q1Wqt0PhReTl7fmNUbc0xZswYnTx5UhMmTNALL7ygxMREnTt3rkSbevXqlbpjaUBAQLk/1wwaNEgNGjTQ66+/rscff1xz587Vrl27yvsSYIKKzpOrj0UWi0URERFKT0+XJMf/V7fz8vIqdR68mre3t8aMGaOUlBSNHz9eL730kpYsWaLs7Oybeq1wn4rKs/T0dEVERJRqd63PTHXr1uWGApUUPxVUe9cass1QbjgrJiZG7777rrZt26bdu3fr+++/1/LlyzV+/Hj179/f3eHBTa7+MOTK8/zlL3/R4cOHtWPHDu3evVvvvfeevvnmG02fPl21atWqkH5QfXAnpZqje/fuuvXWW7V161bt3r1by5Yt05IlSzR58mTHpSQV/bmmTp06euutt7Rr1y7t2rVLKSkpSk5OVu/evUvc2AKVhzvyxBV33XWXOnbs6PhclZCQoMWLF2vq1KmKiopyd3i4BnflGee8yqvyHFWAmxQUFCQfH58yL0E6c+aMGyJCVedMUSAwMFB9+/bVU089pffee0/NmjXTggULDIwOlc3Vx56zZ8/KbrcrPDxcYWFhstvtpb6xc+bYFB0drQceeEAzZszQk08+qV9++UUbN26skNhRuXF+w/WEhIRo0KBBeu655zR79mzVrl1bCxcudOo5nC2Ce3l5qVOnTho3bpzeffddDRgwQOvWrdPZs2edeh6YpyLy5Iqrj0V2u11nz55VWFiYJDn+v7pdcXHxTV92GRERoeHDh+vFF1/U22+/reLiYi1btqxc8cI8FZFnYWFhjs9Qv8X5ruqhqIQqx8PDQ+3atdO2bduUkZHhWH/q1Cnt3r3bjZGhqvL19ZUk5efnX7fdxYsXSyzXqlVL9evX5/KkGuZf//pXieVvv/1WktS+fXvHN3QrVqwo0ebq5bLk5uaW+mAVGRkpSeRYDcH5DWWx2Wylzk916tRRSEiI0/OvXTnf5eXl3bDt1ec8Dw8PNWvWTBLHpMqoIvPkinXr1qmgoMCxvHnzZmVlZTnOdc2bN1dQUJBWr15doo/k5OQb5tilS5dUVFRUYl39+vVVq1Yt5hWsxCoyz2JjY5WVlaXNmzc71l26dElr1qypkFhhHi5/Q5UUHx+vXbt2aerUqRo4cKBsNpu+/fZbNWnShDlu4LQrtz/96quv1KNHD3l6eqpjx46l2j3zzDOKiYlR8+bNFRgYqGPHjmnLli2lJm5G9ZaWlqaZM2eqffv2Onz4sNavX6+ePXs6CkA9evTQqlWrlJ+fr9atW2vv3r2lRi6VZe3atVq1apU6d+6siIgIFRQU6LvvvpOfn586dOhg8KtCZcH5DVcrKCjQ448/rq5du6pZs2aqVauW9u7dq2PHjpW4+9LNqF+/vgICArR69Wr5+fnJ19dXrVq1KnN+rvfff1+5ublq06aN6tWrp/T0dK1cuVKRkZFq1KhRRb08VJCKzJMrAgMDNXXqVPXp00cXLlzQ8uXLFRER4bjk38vLS6NHj9a8efM0bdo0de/eXWlpaUpOTr7hnEqpqal65ZVX1K1bNzVu3Fienp7aunWrLly4oO7du5crXhivIvOsf//+WrlypWbPnq3jx48rJCRE69atcxS/UXVQVEKV1KxZM02ZMkWfffaZEhMTVa9ePcXHxysrK4sP3XBay5Ytdf/992v16tXatWuX7Ha7Zs+eXardkCFDtH37du3Zs0dWq1VhYWG6//77S9wKF9XfU089pcTERM2fP18eHh4aPHiwHnroIcf2CRMmKCgoSBs2bNC2bdvUpk0b/ed//qcmTJhw3eeNiYnR0aNHtWnTJl24cEH+/v5q0aKFnnzySSZkrkE4v+Fqvr6+GjRokHbv3q2tW7fKZrMpIiJC48aN08CBA516Li8vL02aNEnz58/XBx98oMuXL2vixIllHmN69eql7777TqtWrVJeXp6Cg4PVrVs3xcfHV6p5efCrisyTK0aOHKmff/5ZixcvVkFBgW6//XaNGzeuxB/9AwYMkM1m09KlS/XFF1+oadOmeu6555SQkHDd565Xr5569Oihffv2ad26dfL09FSjRo309NNPl7gbGCqXiswzX19fTZ06VR9//LFWrlwpHx8f9erVS+3bt9frr79u0CuAESz2q8faAwCAUhITE5WUlKQPP/xQQUFB7g4HAABD7N+/X9OmTdMzzzxDgQfADfE1AwAAAAAAAJxGUQkAAAAAAABOo6gEAAAAAAAApzGnEgAAAAAAAJzGSCUAAAAAAAA4jaISAAAAAAAAnEZRCQAAAAAAAE6jqAQAAAAAAACnUVQCAAAAAACA0ygqAQAAAAAAwGkUlQAAAAAAAOA0ikoAAAAAAABwGkUlAAAAAAAAOI2iEgAAAAAAAJxGUQkAAKAKioyMlMVi0YkTJ9wdCgAAqKEoKgEAAAAAAMBpFJUAAAAAAADgNIpKAAAAAAAAcBpFJQAAUONZLBZZLBZJ0ieffKJu3bqpTp06JeYs+vnnnzVz5kz169dPTZs2la+vr4KDg9WzZ0/993//t2w2W6nnPXHihCwWiyIjI2W32zVv3jx17NhRAQEBqlOnjgYOHKgff/zxmnEdOHBAo0aNUmhoqPz8/NSmTRv9/e9/1+XLl6/7ejIzM/XXv/5Vt912m/z9/VW7dm117NhRb775pgoKCkq1T05OlsViUZ8+fXTp0iVNmzZN0dHRqlWrlpo2barnn39ehYWFkqQLFy5o8uTJat68uWrVqqXIyEi9/PLLKi4uvtm3GwAAVBMWu91ud3cQAAAA7nSloPTEE09o7ty56t69u5o0aaLjx48rISFBzZo102uvvaa//e1vioqKUvPmzRUWFqbU1FT9+OOPKioq0u9+9zslJSU5nkv6tagUFRWlZs2aqU+fPpo/f7569eql0NBQ7dq1S4cPH5avr6/Wrl2ruLi4EjFt2LBBgwcPVl5enpo3b64uXbooIyNDa9eu1YgRI7R9+3b9/PPP+umnnxQZGenY7/jx4+rXr59+/vlnhYWFqXfv3rJarfrhhx908eJFdejQQWvWrFFISIhjn+TkZPXt21fdunWTj4+PUlJS1KdPH9lsNq1fv14XLlzQsGHD9Omnn6p79+7KzMxU7969dfHiRa1bt06FhYV6/PHH9d577xn7gwIAAJUKRSUAAFDjXSkEBQUF6V//+pe6du1aqs22bdsco4V+68yZMxo6dKh2796txMREjRo1yrHtSlFJkpo1a6ZVq1YpOjpaknT58mX9x3/8hz7++GMNHDhQ//rXvxz7FRYWKjo6Wr/88oueeuop/f3vf5enp6ckac+ePerfv78yMjIkqVRRqWvXrtqyZYtGjBih+fPnKyAgQJKUnp6uwYMHa+fOnXrwwQf15ZdfOva5UlSSpC5dumjFihWqV6+epF9HaMXGxiorK0u33367IiMj9fXXX8vf31+StH37dnXr1k02m00//fSTmjZt6uzbDwAAqigufwMAAPj/Jk+eXGZBSZI6d+5cqqAkSQ0bNtSbb74pSVqwYME1n/vdd991FJQkydPTU9OnT5ckrV27Vlar1bHtf//3f/XLL7+oSZMmevPNNx0FJUlq27atpkyZUmYfGzZs0JYtW+Tv76958+Y5CkqSFBYWpnnz5kmSvv76a506darU/haLRR999JGjoCT9Wgz7/e9/L+nXAtaHH37oKChJUqdOnTRkyBDZbDYlJydf8/UDAIDqx8vdAQAAAFQW991333W3X7p0SatWrdK2bduUlpamS5cuyW636+LFi5KkQ4cOlbmfl5eXBg8eXGp9RESEQkJClJWVpfPnzysiIkKSHMWZ+Ph4eXt7l9rv4Ycf1tNPP11q/ZX9Bg8erPr165fa3rFjR7Vr1067d+/W2rVrNWbMmBLbmzZtWmbhrFWrVo79w8PDr7n9zJkzpbYBAIDqi6ISAADA//fby8iutnnzZt1///06efLkNdvk5OSUub5BgwZlFoekXy+5y8rKckyELckxiujKpXNXCwkJUZ06dXThwoUS60+fPn3d/SSpRYsW2r17t6Ptb13r0rXAwMDrbq9du7YklXgNAACg+uPyNwAAgP/Pz8+vzPX5+fm65557dPLkST3yyCPaunWrMjMzVVxcLLvd7hihdK2pKj08qsZHrhvFWVVeBwAAMAcjlQAAAG5g3bp1OnfunDp06KCPP/641PYjR45UaH+NGjWS9OtE32XJzs4uNUrpt/sdP378ms99ZduVtgAAAOXF100AAAA3kJmZKenal3998cUXFdrfHXfcIUlKTEwsMYH3FZ999lmZ+/Xp00eStHLlSp07d67U9pSUFO3atUseHh7q3bt3xQUMAABqJIpKAAAAN3DrrbdKkr777jsdOHCgxLZ58+YpISGhQvu777771KhRI508eVIvvPCCbDabY9u+ffv02muvlblfz549FRcXp4KCAj322GPKz893bMvIyNBjjz0mSRo9erSaNGlSoTEDAICah6ISAADADcTGxuruu+/WxYsXFRsbq0GDBumBBx7Qrbfeqscff1x//etfK7Q/Pz8/ffnll/L399fbb7+t6OhoPfDAAxo4cKA6dOigXr16qVmzZmXuO3/+fDVr1kxLlixRVFSURo0apXvuuUctWrTQtm3b1KFDB82ePbtC4wUAADUTRSUAAICbsGDBAr311ltq3bq1NmzYoFWrVqlp06b617/+pXHjxlV4f3fccYe2bNmi3/3ud8rKytKiRYt06tQpvfLKK9cdGdW8eXPt3LlTL7zwgurVq6dvvvlGq1evVosWLTRjxgxt2LBBISEhFR4vAACoeSz2a92mBAAAAAAAALgGRioBAAAAAADAaRSVAAAAAAAA4DSKSgAAAAAAAHAaRSUAAAAAAAA4jaISAAAAAAAAnEZRCQAAAAAAAE6jqAQAAAAAAACnUVQCAAAAAACA07zcHUBVlpWVpeLiYneHgZsQFham9PR0d4eBKowcgqvIIbiKHIKryCG4ihyCq8ihqsHLy0shISE319bgWKq14uJiWa1Wd4eBG7BYLJJ+/XnZ7XY3R4OqiByCq8ghuIocgqvIIbiKHIKryKHqicvfAAAAAAAA4DSKSgAAAAAAAHAaRSUAAAAAAAA4jaISAAAAAAAAnEZRCQAAAAAAAE6jqAQAAAAAAACnebk7gIpQWFiopUuX6siRIzp69Kjy8vI0ceJE9enTp1Rbm82mNWvWaPXq1Tpz5ox8fX3VrFkzPfzww4qMjDQ9dgAAAAAAgKqoWhSVcnJylJSUpNDQUEVGRmr//v3XbPvee+9pw4YN6t27twYPHqxLly7pp59+Uk5OjokRAwAAAAAAVG3VoqgUEhKiefPmKTg4WMeOHdMLL7xQZrtNmzZp7dq1mjx5srp06WJylAAAAAAAANVHtZhTydvbW8HBwTdst3z5crVs2VJdunSRzWZTYWGh8cEBAAAAAABUQ9VipNLNyM/P19GjRzVw4EDNnz9fK1euVGFhocLDw/Xggw+qe/fu7g4RAAAAAACgyqgxRaVz587Jbrdr06ZN8vDw0EMPPSR/f3+tWLFC77zzjvz9/dW+ffsy97VarbJarY5li8UiPz8/x2NUbld+RvysUF7kEFxFDsFV5BBcRQ7BVeQQXEUOVU81pqh05VK3ixcvavr06WrVqpUkqVOnTpo0aZL+93//95pFpUWLFikpKcmxHBUVpZkzZyosLMzwuFFxIiIi3B0CqjhyCK4ih+AqcgjlkZeXp8DAQElSbm6uAgIC3BwRqjKOQ3AVOVS91Jiiko+PjyQpPDzcUVCSpFq1aqljx45av369Ll++LE9Pz1L7jhw5UsOGDXMsX6mspqenq7i42ODI4SqLxaKIiAidPXtWdrvd3eGgCiKH4CpyCK4ih+CK/Px8x+Nz5845RtwDzuA4BFeRQ1WHl5fXTQ+iqTFFpZCQEElSnTp1Sm2rU6eOLl++rEuXLsnf37/Udm9vb3l7e5f5vPwyVB12u52fF1xCDsFV5BBcRQ6hPH6bM+QQXEUOwVXkUPViWlEpLS1N27dv18GDB3X69Gnl5OTIYrGodu3aatSokW655RZ16tRJ4eHhhvRft25dBQcHKzMzs9S2rKwseXt7q1atWob0DQAAAAAAUN0YXlTasWOHli1bpoMHD8putysiIkLh4eFq0qSJpF+v8f7555+1ZcsWffrpp7rllls0YsQIdezYscJj6d69u1asWKE9e/aobdu2kqScnBxt27ZNbdq0kYeHR4X3CQAAAAAAUB0ZWlSaMmWKTpw4oc6dO+vpp5/W7bffXublZdKv13rv2bNHmzdv1qxZs9SsWTNNnz79pvtauXKl8vLylJWVJUnavn27zp8/L0kaMmSI/P39dc899+jHH3/U22+/rbvuukv+/v5avXq1Ll++rAceeMD1FwwAAAAAAFBDGFpUuu222/Tss88qODj4hm39/f3VtWtXde3aVdnZ2VqxYoVTfS1btkzp6emO5a1bt2rr1q2SpF69esnf31/BwcF65ZVX9Pnnn2v58uW6fPmyoqOj9ac//UmRkZFO9QcAAAAAAFCTGVpUevDBB8u1X3BwsNP7zpkz56ba1a9fX5MnTy5PWAAAAAAAAPj/mEQIAAAAAAAATjN8ou7jx487vU/z5s0NiAQAAAAAAAAVxfCi0gsvvOD0PgkJCQZEAgAAAAAAgIpieFFpwoQJRncBAAAAAAAAkxleVOrTp4/RXQAAAAAAAMBkpkzUnZmZqczMzBu2ycrKMiMcAAAAAAAAuMjwotLx48c1ceJEbdq06brtNm3apIkTJ+rkyZNGhwQAAAAAAAAXGV5UWrlypRo0aKC77rrruu3uuusuNWzYUCtWrDA6JAAAAAAAALjI8KLS/v371a1bN1ksluu2s1gs6tq1q/bt22d0SAAAAAAAAHCR4UWl7OxshYeH31Tb0NBQ5lUCAAAAAACoAgwvKvn6+io3N/em2ubl5cnHx8fgiAAAAAAAAOAqw4tKzZo1044dO26q7Y4dO9SsWTODIwIAAAAAAICrDC8q9e7dWwcOHNC333573XYrV67UgQMHdMcddxgdEgAAAAAAAFzkZXQHd9xxh3788Uf9z//8j1JSUtSrVy81bdpUfn5+Kigo0MmTJ7V+/Xrt3r1bbdu2VZ8+fYwOCQAAAAAAAC4yvKjk4eGhyZMn6/PPP9eaNWu0e/fuMtvceeed+sMf/nDDu8QBAAAAAADA/QwvKkmSj4+PHn30UY0cOVIpKSk6ffq08vPz5e/vr4YNGyo2Nlb16tUzIxQAAAAAAABUAFOKSlfUrVtX/fv3N7NLAAAAAAAAGMDwiboBAAAAAABQ/VBUAgAAAAAAgNMoKgEAAAAAAMBpFJUAAAAAAADgNIpKAAAAAAAAcJqpd3+7ms1m05EjR5SZmang4GBFR0fL09PTnSEBAAAAAADgJritqHT69GnNnDlT58+fV2BgoHJyclS3bl09++yzioyMdFdYAAAAAAAAuAluKyp9+OGHat++vR566CH5+PgoJydH//jHP/Tf//3feuONN5x6rsLCQi1dulRHjhzR0aNHlZeXp4kTJ6pPnz7X3Ke4uFjPPvusTp8+rYceekgjRoxw8RUBAAAAAADUHIbPqTRv3jzl5uaWWp+amqo+ffrIx8dHkhQUFKQuXbooNTXV6T5ycnKUlJSk06dP3/Qop5UrVyojI8PpvgAAAAAAAGBCUSkrK0t/+tOftGLFCtlsNsf6mJgYff755/r3v/+ts2fPaseOHVq+fLliYmKc7iMkJETz5s3T3Llz9dBDD92w/YULF5SUlKS7777b6b4AAAAAAABgwuVvzz//vHbt2qXPPvtMq1ev1sMPP6z27dtr3Lhx+vDDD/Xqq6/q8uXL8vDwUOfOnTVu3Din+/D29lZwcPBNt//yyy/VsGFD9e7dW4mJiU73BwAAAAAAUNOZMqdS+/bt1bZtW3377bd65513FB0drbFjx+rJJ5/UE088oYsXL6p27dry8DB84JSOHj2qtWvX6tVXXzW8LwAAAAAAgOrKtIm6PTw8dNddd6lXr1766quv9Oyzz2rgwIEaNWqU6tSpY0oMdrtdH3/8sbp3767o6GilpaXd1H5Wq1VWq9WxbLFY5Ofn53iMyu3Kz4ifFcqLHIKryCG4ihyCK36bNxaLhTxCuXAcgqvIoerJtKJScXGxioqKFBQUpMcee0yDBg3SJ598oieffFL333+/+vfvb3hyJScn6+TJk3rmmWec2m/RokVKSkpyLEdFRWnmzJkKCwur6BBhoIiICHeHgCqOHIKryCG4ihxCeeTl5Tke169fXwEBAW6MBlUdxyG4ihyqXgwvKmVlZem9997Tnj17ZLfbFRERoccee0wxMTGaNm2aNm3apC+++EKrVq3S2LFjyzVR983Iz8/X/PnzNWLECIWGhjq178iRIzVs2DDH8pXiV3p6uoqLiys0TlQ8i8WiiIgInT17Vna73d3hoAoih+AqcgiuIofgivz8fMfjc+fOOUbcA87gOARXkUNVh5eX100PojG8qDRv3jylp6dr6tSpCggI0MKFC/X2229r7ty58vX1Vffu3dWpUyctXrxYb7zxhmJjY50eSXQzli1bpuLiYnXv3t1x2VtmZqakX7+9SUtLU926deXlVfot8fb2lre3d5nPyy9D1WG32/l5wSXkEFxFDsFV5BDK47c5Qw7BVeQQXEUOVS+GF5X+/e9/a8yYMY4RSGPGjNGf/vQnnTp1Si1atJAk+fj4KD4+Xv369dPnn39uSBwZGRnKy8srs2C1aNEiLVq0SG+++aYiIyMN6R8AAAAAAKA6MbyoFBISoiNHjujOO++UJB05ckSSFBwcXKptaGionn76aUPiGDJkiDp37lxiXU5OjubNm6c+ffqoU6dOCg8PN6RvAAAAAACA6sbwotIDDzygd955RwcPHlRAQICOHz+uIUOGqF69ehXaz8qVK5WXl6esrCxJ0vbt23X+/HlJvxaUmjdvrubNm5fY58plcI0bN1aXLl0qNB4AAAAAAIDqzPCiUpcuXTRr1izt3r1bRUVFevjhh3XLLbdUeD/Lli1Tenq6Y3nr1q3aunWrJKlXr17y9/ev8D4BAAAAAABqKsOLSpIUHh7uuPzNKHPmzHF6n/DwcCUmJhoQDQAAAAAAQPXm4c7OMzMzdfLkSWVnZ7szDAAAgEorPz9fDRs2lMViKXFreAAwC8chANdi+EilTZs2KTo6WqGhoY51KSkp+uyzz3TmzBnHusjISI0dO1a33nqr0SEBAAAAAADARYYXld555x396U9/Us+ePSVJe/bs0cyZMxUUFKS7775boaGhOnv2rJKTk/Xaa69p+vTpioyMNDqsCrHvXJ4u5F9ydxi4AYvFopA8b2VlXZTdbnd3OKiCyCG4ihyCKwoL/m9UwPbTF+Vbq9iN0aAqIofgKnIIFYHPQ1VHHX9f9Q0Lu6m2psyp9FtfffWV6tWrp5kzZyowMNCxftiwYXr22We1cOFCPfPMM2aHBQAAAAAAACeYOqdScXGxjh8/rmHDhpUoKElS3bp1NWDAAP373/82MyQAAAAAAACUg6lFpcuXL0uS6tevX+b28PBw5eXlmRkSAAAAAAAAysGUy9927tzpuMNbrVq1lJmZWWa7rKwsBQQEmBESAAAAAAAAXGBKUWnjxo3auHGjY3nr1q0aMGBAqXZ79+5V48aNzQgJAAAAAACYoLAgXw/06SBJ+nptinxr+bk5IlQUw4tKs2fPLrXOYrGUWpeTk6P69eurffv2RocEAAAAAAAAFxleVAq7ydvQBQUFaeLEiQZHAwAAAAAAgIpg6kTdAAAAAAAAqB5MmVNJ+nWy7o0bN+r48ePKzMxUUVGRatWqpYYNG+r222/XnXfeqXr16pkVDgAAAAAAAFxgeFHp0qVLevvtt7V79+4S6z09PdWwYUNlZGRo0aJFWrFihf7jP/5DPXv2NDokAAAAAAAAuMjwolJCQoL27t2rRx55RD169JCvr68OHDigDz74QDExMRozZoxOnTqlL774QrNnz1bdunUVExNjdFgAAAAAAABwgeFzKm3atEmDBg3S4MGDVbt2bfn4+Kh9+/YaN26cvvnmG124cEGNGzfW888/rxYtWmjhwoVGhwQAAAAAAAAXGV5Uunjxoho3blxqfePGjWWz2ZSamipJslgs6tmzp44ePWp0SAAAAAAAAHCR4Ze/hYeHa9++fRowYECJ9fv375ckBQcHO9Z5enrKbrcbHRIAAKhGli37xt0hGMpadMnxePnyFfL28XVjNMYbPnyYu0MAAAA3yfCi0qBBg/TJJ5/Ix8dH3bt3l4+Pjw4ePKglS5bolltuUUREhKPtiRMnSiwDAAAAAACgcjK8qDR48GBlZGRo+fLlWrt2rWN9y5Yt9dRTT5Vo6+PjoyFDhhgdUoXxsBbJo6jQ3WHgBiwWiyyXCuVRVMhIOJQLOQRXkUPG8r5c7O4QjGUrlp/nrzMWeNuK5X3Z080BGYvPVhXPo+iSI4c8ii7Jw8Pi5ohQ1ZBDcBU5VLV4eN/8z8diN+nTbXZ2tg4fPiyr1aqGDRsqKirKjG4NdfbJMbIeO+TuMAAAAAAAACqEd4vWivjnlzfV1vCRSlcEBwerS5cuZnUHAAAAAAAAA5lWVKqODj3yki7kFbg7DNyAxWJRcHBdZWdnctkJyoUcgqvIIWOtWLHS3SEYymq9pK9mTZMkPfD0S/L2rt4TdQ8dOtjdIVQ7hQUFGjukhyTp05Wb5FurlpsjQlVDDsFV5FDVUifATzc72zVFJRfYvH1k8+GPg8rOYrHI7ltLNp9a/DGHciGH4CpyyFhWz+r9ccZ6+bIKLtt+fezhJVXz12vz4Q+Nima7bHPkkM3Hl/cYTiOH4CpyqGqxefvcdNtq8amksLBQS5cu1ZEjR3T06FHl5eVp4sSJ6tOnj6ONzWbTunXrtGXLFp04cUK5ubkKDw9X9+7dNXz4cPn43PybBgAAAAAAUNN5uDuAipCTk6OkpCSdPn1akZGRZbYpKirS3LlzlZOTozvvvFMPP/ywWrRoocTERL3xxht8cwwAAAAAAOCEajFSKSQkRPPmzVNwcLCOHTumF154oVQbLy8vvfrqq2rdurVj3YABAxQeHq7ExETt3btXbdu2NTNsAAAAAABU96ef3R2CoQoKCx2Pg386Kb9qPqdSZlQzd4dgmmoxUsnb21vBwcHXbePl5VWioHTFlTvSnT592ojQAAAAAAAAqiXTRyoVFhbq448/1t13361GjRqV2Hb48GGtWbNGEydONC2e7OxsSVLt2rVN6xMAAACVR53Uz9wdgqF8Coscj2unfim/WtV7LtELDf5gep/rT84yvU8zWQutjscbT74r71rebozGeL2aPu3uEIAqw/SiktVq1dq1a9W7d+9SRaVz585p7dq1phaVlixZIj8/P8XGxl6zjdVqldX6fwdSi8UiPz8/WSwWWSwWM8KECyyy/N///LhQDuQQXEUOATePz1ZwFTkEV5FDcFVVzyFn4q8WcyqV18KFC7V3716NGzdOAQEB12y3aNEiJSUlOZajoqI0c+ZMBQUFycP3shmhGibjlZpRhc9wdwAmCZ1avb8lc7fgkBB3h4AqqCA/T4PbRUqSVu4+IT//a59vgLJ4+/jqjy/+3d1hmCYkpK7pfV4+Y3qXMJA7ckjVezqcGsctOaQTbugTRnFPDlWcQF/Pm25bY4tKmzZtUkJCgvr166eBAwdet+3IkSM1bNgwx/KVql1OTo4u5F8yNE7AGVlZme4OoVqyyKLgkBBlZ2XJLu4UCecUFuQ7Hl/IzlbhJc4bwPW441wWZHqPMBKfh+Aqd+QQX11WL1X9OGTz99XNZqVpRaWMjF/HiuTm5kqSLly44FgnSaGhoWaFoj179mj27NmKjY3V+PHjb9je29tb3t6lrxu22+2y2/kDE5UH+WiQ/z/60y5+5+G83+YMOQTcGL8jcBU5BFeRQ3BVVc8hZ+I3rag0adKkEsv//Oc/SywnJCSYEseRI0f01ltvqUWLFnrmmWfk6Xnzw7oAAAAAAADwK9OKSi+99JIkKS8vT3//+9/1+9//Xs2bNzere0nSqVOnNGPGDIWHh+v555+Xj0/1vvMFAAAAAACAUUwrKsXExEiSLl68KEmKjIx0rKsIK1euVF5enrKysiRJ27dv1/nz5yVJQ4YMkcVi0fTp05Wbm6vhw4dr586dJfaPiIhQdHR0hcUDAAAAAABQnVWbibqXLVum9PR0x/LWrVu1detWSVKvXr0kyVFkmj9/fqn977jjDopKAAAAAAAAN6naFJXmzJlzwzaJiYkV2meb+gGyWqv2JXSXz//b3SGgAnk2ru3uEKoli8WiBg1ClZpqrfKT7sF8+fn/N3dfp0a15efn58ZoqqfN1vQbN0KV0cUN57LwwlOm92mmvAKr43GP4NMK8Ct9A5rqJM0NOfTThSOm92mmSz6XHY8bBh2Tr1/1npfWHcehhhnV+zNmnvf/vb6ete0KqFW9X++ZKv53WVk3KrsWDwPjKJOnp6diYmIUEBBQaludOnUq9JI4AAAAAAAAGMP0kUr+/v6OSbuv1rZtW7Vt29bkiAAAAAAAAOAs00cqAQAAAAAAoOoztKh06dIlt+wLAAAAAAAAYxlaVJowYYKSkpKUlZV10/tkZmYqISFBEydONDAyAAAAAAAAuMLQOZXGjRunBQsWKCkpSa1bt9btt9+u5s2bKzw8XAEBAbLb7crLy1NaWpqOHTumvXv36siRI2rQoIEeffRRI0MDAAAAAACACwwtKnXv3l1du3bV9u3blZycrEWLFqm4uLjsQLy81LZtWz3zzDPq1KmTPDyY7gkAAAAAgKouoFYt2ZevdHcYMIDhd3/z8PBQly5d1KVLF1mtVh0/flynT59Wbm6uJCkwMFCNGjVS8+bN5e3tbXQ4AAAAAAAAqACGF5V+y9vbW61bt1br1q3N7BYAAAAAAAAVjGvMAAAAAAAA4DSKSgAAAAAAAHAaRSUAAAAAAAA4jaISAAAAAAAAnGbqRN0AAFxtWUK2u0Mw1KVL+Y7Hyxdkydf3khujMd7w+4PdHQIAAABMwkglAAAAAAAAOM30kUoZGRlauHCh9u/fr5ycHD377LOKiYlRTk6OkpKS1LdvX0VFRZkdFgAAAAAAAJxg6kilU6dO6bnnntOPP/6o8PBw5efny2azSZKCgoJ06NAhrVy50syQAAAAAAAAUA6mFpW++OILBQQE6J133tGf/vSnUttjY2N18OBBM0MCAAAAAABAOZhaVPr3v/+tO++8U0FBQbJYLKW2h4aGKjMz08yQAAAAAAAAUA6mFpVsNpt8fX2vuT0nJ0deXtyQDgAAAAAAoLIztajUvHlz7dy5s8xtly9f1qZNmxQdHW1mSAAAAAAAACgHU4cF3XPPPZoxY4Y++OAD9ejRQ5KUnZ2tPXv2aNGiRTp9+rT++Mc/mhkSAAAAAOA6fP089V+r+rk7DACVkKlFpdjYWE2aNEmffPKJ1qxZI0l69913JUl+fn6aNGmSYmJizAwJAAAAAAAA5WD6BEa9e/dWly5dtGfPHp09e1Y2m00RERFq166d/Pz8zA4HAAAAAAAA5eCWWbFr1aqlLl26VNjzFRYWaunSpTpy5IiOHj2qvLw8TZw4UX369CnV9tSpU/r000918OBBeXl5qUOHDnr44YcVFBRUYfEAAAAAAABUd6YWlTIyMm6qXWhoqFPPm5OTo6SkJIWGhioyMlL79+8vs9358+f10ksvyd/fXw888IAKCwu1bNkynTx5Um+88QZ3ngMAAAAAALhJplZRJk2adFPtEhISnHrekJAQzZs3T8HBwTp27JheeOGFMtstWrRIly5d0syZMx2Fq5YtW+q1115TcnKyBgwY4FS/AAAAAAAANZWpRaUJEyaUWmez2ZSenq5169YpKChIgwYNcvp5vb29FRwcfMN2W7ZsUYcOHUqMhGrbtq0aNGigH3/8kaISAAAAAADATTK1qFTWHEdX3H333ZoyZYry8/MN6TszM1MXLlxQixYtSm1r2bKlUlJSrrmv1WqV1Wp1LFssFsek4haLpeKDBcqJfDTGlfeV9xe4MX5P4CpyCK4ih+Aqcgiuqkk5VGkmEapVq5b69Omj5cuXa+jQoRX+/FlZWZJ+vVTuaiEhIcrNzZXVapW3t3ep7YsWLVJSUpJjOSoqSjNnzlRYWFiFx2m2X9wdACpUgwYN3B1CtRYREeHuEKqpLHcHgArEcQiuckcOFR8xvUsYyC3HoX3mdwnjuCWHUvaY3ycMU5M+D1WaopIk2e12ZWdnG/LcRUVFklTmZNxXCklFRUVlFpVGjhypYcOGOZavVB3T09NVXFxsRLhAuaSmpro7hGrJYrEoIiJCZ8+eld1ud3c4QKXGcQiuckcOVf2vCfFbHIfgKnfkUM0pQdQMVf045OXlddODaCpFUSk/P1///ve/tXTpUkVFRRnSh4+PjySVWQS6cmnblTZX8/b2LrPYJIk/MFGpkI/GstvtvMfADfA7AleRQ3AVOQRXkUNwVU3KIVOLSvfff/91t4eGhmrcuHGG9H3lsrcrl8H9VlZWlgIDA69ZOAIAAAAAAEBJphaV7r333lITVlksFgUEBKh+/fpq166dPD09Dem7bt26CgoK0rFjx0ptO3r0qCIjIw3pFwAAAAAAoDoytagUHx9vZnelxMXFae3atcrIyFBoaKgkae/evUpNTdVdd93l1tgAAAAAAACqkkoxp1JFWLlypfLy8hyXt23fvl3nz5+XJA0ZMkT+/v4aOXKkNm/erGnTpmno0KEqLCzU0qVL1bRpU/Xt29ed4QMAAAAAAFQphhaV5s6d6/Q+FotFEyZMcHq/ZcuWKT093bG8detWbd26VZLUq1cv+fv7KzQ0VC+//LI+++wzzZ8/X15eXoqNjdUf/vAH5lMCAAAAAABwgqFFpf379zu9z9VzLt2sOXPm3FS7Jk2aaMqUKeXqAwAAZ/n6+uv9/3L+fAgAAABUdoYWlW620AMAAAAAAICqxcPdAQAAAAAAAKDqoagEAAAAAAAAp5l+97eUlBR98803+umnn5Sfny+73V6qTUJCgtlhAQAAAAAAwAmmjlTavHmzZsyYoQsXLqh79+6y2+3q0aOHevToIR8fHzVr1kz33XefmSEBAAAAAACgHEwdqbR48WK1bNlSr776qnJzc7V69Wr169dPbdq0UVpamqZMmaLw8HAzQwIAAAAAAEA5mDpS6dSpU+rRo4c8PDzk6ekpSSouLpYkhYeHa9CgQVqyZImZIQEAAAAAAKAcTC0q+fr6ysvr18FRAQEB8vLyUnZ2tmN7nTp1lJaWZmZIAAAAAAAAKAdTi0oNGzbUqVOnHMuRkZFat26dLl++rKKiIm3YsEGhoaFmhgQAAAAAAIByMLWo1LlzZ23btk1Wq1WS9Lvf/U779+/X2LFjNW7cOB08eFD33HOPmSEBAAAAAACgHAyfqHvHjh2KjY2Vh4eHRowYoREjRji2dezYUS+//LK2bNkiDw8PdejQQW3atDE6JAAAAAAAALjI8KLSm2++qdq1a6tbt27q1auXoqOjS2y/9dZbdeuttxodBgAAAAAAACqQ4UWl8ePHa+PGjVq9erVWrVqlsLAw9ezZUz179lTjxo2N7h4AAAAAAAAGMLyoNGDAAA0YMECZmZnauHGjNm7cqEWLFmnRokWKjIxUr1691L17d9WtW9foUAAAAAAAAFBBDC8qXVG3bl0NHz5cw4cP15kzZ7RhwwZt3LhRn3/+ub744gvFxMSoV69eiouLk7+/v1lhAQAAAAAAoBxMKyr9VsOGDRUfH6/4+HgdPXpUGzdu1KZNm/T+++/r448/1ueff+6OsAAAAAAAAHCTPNwdQHh4uOrXr6969epJkoqKitwcEQAAAAAAAG7ELSOVCgsLtWXLFm3YsEH79u2TzWZTWFiYRo4cqZ49e7ojJAAAAAAAADjBtKJScXGxdu7cqQ0bNiglJUVFRUUKDAxU//791bNnT91yyy1mhQIAAAAAAAAXGV5U2rdvnzZs2KAtW7YoPz9f3t7e6tixo3r16qXY2Fh5enoaHQIAAAAAAAAqmOFFpVdffVUWi0W33Xab4+5ufn5+RncLAAAAAAAAAxleVPr973+vHj16KCQkxOiuAAAAAAAAYBLDi0rDhg0zugsAAAAAAACYzC13f3On1NRUJSQk6ODBg8rNzVVoaKh69uyp4cOHy9fX193hAQAAAAAAVAk1qqiUkZGhv/71r/L399fgwYMVGBiow4cPKzExUcePH9dzzz3n7hABAAAAAACqhBpVVFq3bp3y8vL0yiuvqEmTJpKkAQMGyGazad26dcrNzVVgYKCbowQAAAAAAKj8PNwdgJkKCgokSXXq1CmxPiQkRBaLRV5eNarGBgAAAAAAUG6mFpU2bdqkoqIiM7ss4bbbbpMkvf/++zpx4oQyMjK0adMmrVq1SkOGDFGtWrXcFhsAAAAAAEBVYurQnHfeeUd+fn6Ki4tTr1691KZNGzO7V/v27XX//fdr0aJF2r59u2P97373O40ePfqa+1mtVlmtVseyxWKRn5+f4zFQWZCPxrjyvvL+AjfG7wlcRQ5VvAA/b1nX/sHdYZiGHIKryCG4qiblkKlFpVdeeUXr16/X5s2blZycrLp166pnz57q1auXmjZtakoMYWFhuvXWWxUXF6fatWtr586dWrRokYKDgzV48OAy91m0aJGSkpIcy1FRUZo5c6bCwsJMidlIv7g7AFSoBg0auDuEai0iIsLdIVRTWe4OABWI4xBc5Y4cKj5iepcwkFuOQ/vM7xLGcUsOpewxv08YpiZ9HjK1qNS6dWu1bt1ajzzyiHbt2qX169dr5cqVWrp0qZo2barevXurZ8+eCgkJMaT/jRs3at68eXrnnXdUr149SVJcXJzsdru+/PJL9ejRQ7Vr1y6138iRIzVs2DDH8pWqY3p6uoqLiw2JFSiP1NRUd4dQLVksFkVEROjs2bOy2+3uDgeo1DgOwVXuyKGq/zUhfovjEFzljhyqOSWImqGqH4e8vLxuehCNW2am9vT0VMeOHdWxY0cVFhZq69atSk5O1hdffKH58+crJiZGd9xxh7p3716hk2evWrVKUVFRjoLSFZ06dVJycrJ++ukntW3bttR+3t7e8vb2LvM5+QMTlQn5aCy73c57DNwAvyNwFTkEV5FDcBU5BFfVpBxy+93fTp48qaNHj+rkyZOSpIYNGyo3N1dz5szRn/70Jx08eLDC+srOzpbNZiu1/spoo7K2AQAAAAAAoDS3jFQ6c+aMNmzYoA0bNujcuXMKCgpSz5491bt3bzVv3lySdOzYMb3//vv64IMP9Pbbb1dIvw0aNNCePXt05swZNWzY0LF+48aNslgsps3rBAAAAAAAUNWZWlRasWKF1q9fr+PHj8vb21sdO3bU2LFj1b59e3l4lBw01aJFC9111116//33K6z/ESNGaNeuXXrppZc0aNAgx0TdKSkp6tevn+rWrVthfQEAAAAAAFRnphaVPv30U7Vu3Vrjx49X9+7d5e/vf932LVq00L333lth/cfExOi1117TggULtGrVKl28eFHh4eEaPXq07r777grrBwAAAAAAoLoztaj07rvvKjw8/KbbN2nSRE2aNKnQGFq2bKkXXnihQp8TAAAAAACgpjF1ou569eopPz//mtvz8/N1+fJlEyMCAAAAAABAeZhaVPrkk0/0t7/97Zrb//a3v+mzzz4zMSIAAAAAAACUh6lFpV27dikuLu6a27t27aqUlBQTIwIAAAAAAEB5mFpUysrKuu4d1kJCQpSZmWliRAAAAAAAACgPU4tKgYGBOnPmzDW3nz59Wn5+fiZGBAAAAAAAgPIwtajUvn17rVmzRj/99FOpbcePH9eaNWsUGxtrZkgAAAAAAAAoBy8zO7v//vu1a9cu/fWvf1XHjh3VpEkTSdIvv/yiHTt2KCgoSPfff7+ZIQEAAAAAAKAcTC0q1a1bVzNmzNCXX36p7du3a9u2bZIkPz8/9ezZUw888MB151wCAAAAAABA5WBqUUn6dTLuJ554Qna7XTk5OZKkoKAgWSwWs0MBAAAAAABAOZleVLrCYrGoTp067uoeAAAAAAAALjC0qJSUlCRJ+t3vficPDw/H8o3cd999RoYFAAAAAAAAFxlaVFqwYIEk6Z577pGHh4dj+UYoKgEAAAAAAFRuhhaVEhISrrsMAAAAAACAqsnD3QEAAAAAAACg6qGoBAAAAAAAAKcZevnbpEmTZLFYnNrHYrHo3XffNSgiAAAAAAAAVARDi0oxMTFOF5UAAAAAAABQ+Rk+UgkAqrL8/Hy1atVKknT06FH5+fm5OSIAAAAAqByYUwkAAAAAAABOM3SkUllsNpvWrVunnTt3KiMjQ5IUGhqqjh07qlevXvLwoM4FAAAAAABQ2ZlaVMrPz9f06dMdl5DUr19fkrR3715t2bJFq1at0pQpU+Tv729mWAAAAAAAAHCSqUWlr776SsePH9cf//hH9e/fX15ev3ZfXFys77//Xp988om+/vpr/fGPfzQzLAAAAAAAADjJ1GvNtm7dqoEDB2rQoEGOgpIkeXl5aeDAgbrzzju1ZcsWM0MCAAAAAABAOZg6Uik3N1cNGza85vZGjRopNzfX8DiOHz+uBQsW6ODBg7Jarapfv7769++voUOHGt43AAAAAABAdWBqUSkiIkLbt2/XoEGDyty+fft2xzxLRtm9e7dmzpypqKgo3XvvvapVq5bOnTunzMxMQ/sFAAAAAACoTkwtKg0cOFAff/yx3njjDQ0dOlQNGjSQJJ05c0bffvut9uzZo0cffdSw/vPz8zV79mx16NBBzzzzDHeaAwAAAAAAKCdTi0qDBg3ShQsXtGTJEu3atatkIF5euu+++zRw4EDD+t+wYYMuXLig0aNHy8PDQ4WFhfLx8aG4BAAAAAAA4CRTi0qSFB8fr8GDB2vPnj3KyMiQJIWFhen2229XUFCQoX3v3btXfn5+yszM1FtvvaXU1FT5+vqqd+/eevjhh+Xj42No/wAAAAAAANWFKUWl7OxsJScnKy0tTUFBQYqLi1PPnj3N6LqEs2fPymaz6a233lLfvn314IMPav/+/Vq5cqXy8vL01FNPlbmf1WqV1Wp1LFssFvn5+TkeA5UF+VjxfvueWiwW3mPgBvgdgavIIbiKHIKryCG4qiblkOFFpbS0NL3wwgsl7uq2ePFiPfHEE6YXlgoLC3Xp0iXdeeed+uMf/yhJiouLU3FxsdasWaP777/fMc/Tby1atEhJSUmO5aioKM2cOVNhYWGmxW6UX9wdACpUWfkL1+Tl5Tke169fXwEBAW6MprrKcncAqEAch+Aqd+RQ8RHTu4SB3HIc2md+lzCOW3IoZY/5fcIwNenzkOFFpcTERBUWFuqRRx5RmzZtdPbsWX3yySf69NNP1b17d1PnM7pyeVuPHj1KrO/Zs6fWrFmjw4cPl/nDHzlypIYNG+ZYvlJ1TE9PV3FxsYERA85JTU11dwjVTn5+vuPxuXPnHKMUAZSN4xBc5Y4cqvpfE+K3OA7BVe7IoZpTgqgZqvpxyMvL66YH0RheVDp06JAGDBigwYMHS5IaN24sDw8PzZw5U6dPn1aTJk2MDsEhJCREv/zyi4KDg0usr1OnjqSSIxJ+y9vbW97e3mVus9vtFRoj4AryseL99j212+28x8AN8DsCV5FDcBU5BFeRQ3BVTcohw4cJZWRkKCoqqsS65s2bS5IuXrxodPdl9puZmVlifVbWr5deGD1ROAAAAAAAQHVheFHJZrPJy6vkgChPT0/HNjN169ZNkvT999+XWP/dd9/J09NTMTExpsYDAAAAAABQVZly97djx46VuHysoKBAknTw4MEyLzmLi4szJI6oqCj17dtXP/zwgy5fvqyYmBjt379fmzdv1j333KO6desa0i8AAAAAAEB1Y0pRacWKFVqxYkWp9QsWLCizfUJCgmGxjB8/XqGhoUpOTtbWrVsVFhamhx9+WHfddZdhfQIAAAAAAFQ3hheVXnrpJaO7cIqXl5dGjRqlUaNGuTsUAAAAAACAKsvwohLzFAEAAAAAAFQ/hk/UDQAAAAAAgOqHohIAAAAAAACcRlEJAAAAAAAATqOoBAAAAAAAAKdRVAIAAAAAAIDTKCoBAAAAAADAaV5GPvnatWvLtd8dd9xRwZEAAAAAAACgIhlaVJo7d2659qOoBAAAAAAAULkZWlSaPXt2ieW8vDzNmTNH/v7+GjJkiBo2bChJOn36tFauXKmCggJNmjTJyJAAAAAAAABQAQwtKoWFhZVYXrBggYKCgvTiiy/KYrE41jdt2lRxcXGaPn26li9frokTJxoZFgAAAAAAAFxkaFHpatu2bdPo0aNLFJSu8PDwUJcuXfT111+bGRIAF9395UF3h2Coy0UFjsejvj4oTx8/N0ZjvCVjbnF3CAAAAACqCFPv/ma323X69Olrbj916pSJ0QAAAAAAAKC8TC0qde7cWatXr9Y333yjS5cuOdZfunRJy5Yt05o1a9SpUyczQwIAAAAAAEA5mHr52yOPPKK0tDR9/vnnmj9/vkJCQiRJWVlZunz5slq3bq2xY8eaGRIAAAAAAADKwdSikr+/v6ZNm6Zt27YpJSVFGRkZkqR27dqpQ4cO6tixY5nzLQEAAAAAAKByMbWodEXnzp3VuXNnd3QNAAAAAACACuCWolJmZqYOHDignJwcxcXFqV69erLZbMrPz5e/v788PEyd6gkAAAAAAABOMrWoZLfb9dlnn2nlypWy2WySpKZNm6pevXoqLCzUpEmTFB8fr7vuusvMsAAAAAAAAOAkU4cELV26VCtWrNDw4cP14osvltjm7++vLl26aMuWLWaGBAAAAAAAgHIwtaj03Xff6Y477tCDDz6oyMjIUtubNWum1NRUM0MCAAAAAABAOZhaVDp//ryio6Ovud3X11f5+fkmRgQAAAAAAIDyMLWoFBQUpPPnz19z+/HjxxUaGmpiRAAAAAAAACgPU4tKcXFxWr16tc6dO1dq2+7du5WcnKxu3bqZGRIAAAAAAADKwdS7v8XHx2v//v167rnndMstt0iSlixZooSEBB0+fFhRUVEaOXKkmSFp4cKF+vrrr9WkSRO9/fbbpvYNAAAAAABQVZk6Usnf31/Tp0/XiBEjlJmZKR8fHx04cED5+fkaNWqUXnnlFfn6+poWz/nz57Vo0SJT+wQAAAAAAKgOTB2pJEk+Pj669957de+995rddSmff/65WrVqJZvNposXL7o7HAAAAAAAgCrD1JFK06ZN0969e6+5fd++fZo2bZopsRw4cECbN2/W2LFjTekPAAAAAACgOjG1qHTgwAFduHDhmttzcnJ04MABw+Ow2Wz65JNP1K9fPzVt2tTw/gAAAAAAAKob0y9/u56zZ8/Kz8/P8H5WrVql9PR0/e1vf7up9larVVar1bFssVgccVosFkNiBMqDfISryCG4ihyCq8ghuIocgqvIIbiqJuWQ4UWl5ORkrV271rG8cOFCfffdd6Xa5efn6+eff1ZsbKyh8Vy8eFGJiYm69957FRQUdFP7LFq0SElJSY7lqKgozZw5U2FhYUaFaZpf3B0AKlSDBg3c0Ou/3dAnjOKeHMpyQ58wintyCNWJO3Ko+IjpXcJAbjkO7TO/SxjHLTmUssf8PmGYmvR5yPCiUlFRkXJychzLBQUFpap2FotFvr6+uvPOO3XfffcZGs/XX3+twMBADRky5Kb3GTlypIYNG+ZYvhJ/enq6iouLKzxGoLxSU1PdHQKqOHIIriKH4Cp35FDV/5oQv8VxCK5yRw7VnBJEzVDVj0NeXl43PYjG8KLSwIEDNXDgQEnSpEmT9Mgjj6hTp05Gd1um1NRUrVmzRmPHjlVmZqZjvdVqVXFxsdLS0uTv76/AwMAS+3l7e8vb27vM57Tb7YbGDDiDfISryCG4ihyCq8ghuIocgqvIIbiqJuWQqXMqzZkzx8zuSsnMzJTdbtcnn3yiTz75pNT2J554QkOHDuWOcAAAAAAAADfgtom6CwoKlJ+fX2YFLzQ01JA+mzRposmTJ5dan5CQoIKCAo0dO1YRERGG9A0AAAAAAFCdmF5UWrVqlb755hudO3fumm0SEhIM6TsoKEhdunQptX7FihWSVOY2AAAAAAAAlOZhZmerVq3SRx99pIiICI0ePVqSdNddd+mee+5RcHCwIiMjNWHCBDNDAgAAAAAAQDmYOlJp5cqVateunf7617/q4sWL+vrrr9WhQwe1adNGd999t/7zP/9TFy9eNDMkSdLLL79sep8AAAAAAABVmakjlc6dO6eOHTtKkjw9PSVJxcXFkiR/f3/169dPq1atMjMkAAAAAAAAlIOpRSV/f39dvnzZ8djHx0cZGRmO7X5+fsrOzjYzJAAAAAAAAJSDqUWlJk2a6Oeff3YsR0dHa/Xq1crMzFRGRobWrFmjBg0amBkSAAAAAAAAysHUolKvXr30yy+/yGq1SpJGjRqlU6dOacKECZo0aZLOnDnjmMAbAAAAAAAAlZepE3X37dtXffv2dSzfcsst+q//+i/t2LFDHh4eatu2rRo2bGhmSABwXZ4+fur05nfuDgMAAAAAKh1Ti0plqV+/voYOHeruMAAAAAAAAOAEtxWVbDab8vPzy9wWGBhocjQAAAAAAABwhqlFpeLiYi1ZskQ//PCDzp8/L5vNVma7hIQEM8MCAAAAAACAk0wtKs2bN09r165VdHS0OnfuLH9/fzO7BwAAAAAAQAUxtai0efNm9e7dW5MmTTKzWwAAAAAAAFQwDzM78/X1VatWrczsEgAAAAAAAAYwtajUo0cP7dy508wuAQAAAAAAYABTL3976KGHNHfuXM2YMUN9+/ZVvXr15OFRuq7VvHlzM8MCAAAAAACAk0wtKlmtVtntdqWkpCglJeWa7bj7GwAAAAAAQOVmalHpvffe09atW9WjRw+1bNmSu78BAAAAAABUUaYWlXbv3q3Bgwdr7NixZnYLAAAAAACACmbqRN1+fn6KiIgws0sAAAAAAAAYwNSiUv/+/bVx40bZbDYzuwUAAAAAAEAFM/Xyt8aNG2v79u16/vnndccdd1zz7m9xcXFmhgUAAAAAAAAnmVpU+sc//uF4/Pnnn1+zHXd/AwAAAAAAqNxMLSq99NJLZnYHAAAAAAAAg5haVIqJiTGzOwAAAAAAABjE1Im6AQAAAAAAUD0YOlJp2rRpslgsmjJlijw9PTVt2rQb7mOxWDR16lRD4jl69KjWrl2r/fv3Kz09XYGBgWrVqpVGjx6thg0bGtInAAAAAABAdWToSCW73S673V5i+Wb2McqSJUu0ZcsWtWnTRmPHjtWAAQP073//W88//7xOnjxpWL8AAAAAAADVjaEjlV5++eXrLptt2LBh+vOf/ywvr/972d27d9fkyZO1ePFiPfnkk26MDgAAAAAAoOowdU6lAwcOKCcn55rbc3JydODAAcP6b926dYmCkiQ1aNBAjRs31unTpw3rFwAAAAAAoLoxtag0bdo07dmz55rb9+3bd1PzLlUku92uCxcuKCgoyNR+AQAAAAAAqjJDL39zltVqlYeHuTekW79+vTIzMxUfH3/NNlarVVar1bFssVjk5+fneAxUFuQjXEUOwVXkEFxFDsFV5BBcRQ7BVTUphwwvKmVkZCgtLc2xfPr06TIvccvPz9eaNWsUFhZmdEglYvnoo48UHR2tPn36XLPdokWLlJSU5FiOiorSzJkzTY3VKL+4OwBUqAYNGrih13+7oU8YxT05lOWGPmEU9+QQqhN35FDxEdO7hIHcchzaZ36XMI5bcijl2lf0oOqpSZ+HDC8q/fDDDyUKMgsXLtTChQvLbOvh4aHx48cbHZIkKTs7WzNmzJC/v7+eeeaZ646QGjlypIYNG+ZYvlJ1TE9PV3FxseGxAjcrNTXV3SGgiiOH4CpyCK5yRw5V/a8J8Vsch+Aqd+RQzSlB1AxV/Tjk5eV104NoDC8qdevWTU2aNJEkzZo1S0OGDNEtt9xSoo3FYpGvr68iIyMVHBxsdEjKz8/X66+/rry8PL3yyiuqW7fuddt7e3vL29u7zG12u92IEIFyIR/hKnIIriKH4CpyCK4ih+Aqcgiuqkk5ZHhRqXHjxmrcuLEkacKECYqJiVF4eLjR3V5TUVGRZs6cqdTUVP3tb39zxAYAAAAAAICbZ+pE3deat+jcuXOyWq2GF3hsNpv+8Y9/6PDhw3r22WcVHR1taH8AAAAAAADVlalFpW+//VaHDh3SU0895Vg3d+5crV27VtKvE2C/8MILqlOnjiH9f/bZZ9q+fbs6duyo3NxcrVu3rsT23r17G9IvAAAAAABAdWNqUem7777Tbbfd5ljetWuX1q5dqwEDBqhp06b6+uuvtWDBAo0bN86Q/k+cOCFJ2rFjh3bs2FFqO0UlAAAAAACAm2NqUSk9PV2NGjVyLP/4448KDw933PEtOzu71OihivTyyy8b9twAAAAAAAA1iYc7O9+zZ4/at2/vWA4LC1N2drbb4gEAAAAAAMDNMbWo1KBBA23btk3Sr5e+ZWZmKjY21rE9MzNTAQEBZoYEAAAAAACAcjD18rfhw4frn//8px555BEVFhaqcePGateunWP7vn37FBkZaWZIAAAAAAAAKAdTi0o9evRQ7dq1tXPnTgUEBGjQoEHy9PSUJOXm5iowMJDJsgEAAAAAAKoAU4tKktS2bVu1bdu21PrAwEBNnjzZ7HAAAAAAAABQDobPqXT06FHl5ubeVNu0tDStXbvW4IgAAAAAAADgKsOLSlOmTNGuXbscy7m5uXrooYd04MCBUm0PHTqkuXPnGh0SAAAAAAAAXGTq3d8kyW63y2q1ymazmd01AAAAAAAAKojpRSUAAAAAAABUfRSVAAAAAAAA4DSKSgAAAAAAAHCalxmdpKWl6fjx45Kk/Px8SVJqaqr8/f1LtQMAAAAAAEDlZ0pRKSEhQQkJCSXWffjhh2Z0DQAAAAAAAAMYXlSaMGGC0V0AAAAAAADAZIYXlfr06WN0FwAAAAAAADAZE3UDAAAAAADAaRSVAAAAAAAA4DSKSgAAAAAAAHAaRSUAAAAAAAA4jaISAAAAAAAAnEZRCQAAAAAAAE6jqAQAAAAAAACnUVQCAAAAAACA07zcHYDZrFarEhIStH79euXm5qpZs2YaPXq02rZt6+7QAAAAAAAAqowaN1Jpzpw5Wr58uXr27KlHHnlEHh4eeuONN3Tw4EF3hwYAAAAAAFBl1Kii0tGjR7Vp0yY98MAD+v3vf68BAwZo6tSpCg0N1RdffOHu8AAAAAAAAKqMGlVU2rx5szw8PDRgwADHOh8fH/Xr10+HDx9WRkaGG6MDAAAAAACoOmpUUemnn35SgwYN5O/vX2J9y5YtJUknTpxwQ1QAAAAAAABVT42aqDs7O1shISGl1l9Zl5WVVeZ+VqtVVqvVsWyxWOTn5ycvr6r/9nm0aO3uEFCBPL29Te+zdXig6X3CON5uyKHQ8Fqm9wnjuCOHGjZsaHqfMI47ckiBzczvE4ZxRw6F125lep8wjluOQ0G1ze8ThnFLDlUgZ2odVb8q4oSioqIyf7hX1hUVFZW536JFi5SUlORY7tGjh/785z+XWaCqcv75pbsjQBX3xcNh7g4BVdy9Y8ghuObJJ590dwio6sJecXcEqEDuOKuMCXvfDb2iWgnj81B1UpN+mjXq8jcfH58SI46uuLLOx8enzP1Gjhyp//mf/3H8Gz9+fJnPg8qpoKBAzz//vAoKCtwdCqoocgiuIofgKnIIriKH4CpyCK4ih6qnGjVSKTg4WJmZmaXWX7ns7Vojj7y9vav88LWazG6366effpLdbnd3KKiiyCG4ihyCq8ghuIocgqvIIbiKHKqeatRIpcjISKWmpio/P7/E+iNHjji2AwAAAAAA4MZqVFGpa9eustlsWrNmjWOd1WpVcnKyWrVqpdDQUDdGBwAAAAAAUHXUqMvfWrVqpa5du+qrr75STk6OIiIitHbtWqWnp+vxxx93d3gwiLe3t+677z4uYUS5kUNwFTkEV5FDcBU5BFeRQ3AVOVQ9Wew17ILGoqIiJSQkaP369crLy1PTpk11//33q3379u4ODQAAAAAAoMqocUUlAAAAAAAAuK5GzakEAAAAAACAikFRCQAAAAAAAE6jqIRqJTExUfHx8e4OA9XEpEmTNGfOHHeHgWoiPj5eiYmJjuXk5GTFx8crLS3NjVGhquD8BqPNmTNHkyZNcncYqCZefvllvfzyy47ltLQ0xcfHKzk52W0xoWrYv3+/4uPjtX//fneHgptEUQlAjXfo0CElJiYqLy/P3aEAAGCYzMxMJSYm6sSJE+4OBQBQTXi5OwAAcLdDhw4pKSlJffr0UUBAgGP9P/7xD1ksFjdGhuqsd+/e6t69O7fVBWCarKwsJSUlKTw8XJGRkSW2PfbYY+L+PTBKWFiYvvjiC3l58ecnUN3wWw0A18Af+zCSh4eHfHx83B0Gqgmbzabi4mJyCuXGH/swksVi4fiECnXp0iX5+vq6OwyIohKqsIMHD+rTTz/VyZMnVbduXY0YMaJUm/j4eA0aNEgxMTFKTExUWlqaIiMj9dhjj6lp06ZavXq1li5dqszMTLVq1UoTJ05UeHi4G14N3CUxMVFJSUmSpCeeeMKxfvbs2Zo2bZpiYmIcc0wUFxdr0aJFWr9+vc6fPy9fX181atRIo0aNUtu2bd0SP8xzJVdmzZqlhIQE7d69W56enurVq5fGjBnj+LBstVr15Zdfav369bJarbrttts0bty4Us+XnJysuXPnavbs2Y7jzrFjx/T111/r+PHjKiwsVHBwsG677TZNnDjR1NcK93Lm/BYdHa1FixYpNTVVTz/9tLp06eKGiGG0goICJSQkaNu2bcrKypK/v7+aNWumMWPGqHnz5nr55Zd18eJFPf300/roo4905MgRBQQEaOjQobr77rsl/TpPybRp0yRJc+fO1dy5cyVJEydOVJ8+fTRnzhwdOHCgxFyCGzdu1NKlS5WamiqLxaLQ0FD1799fQ4cONf9NwA1VRJ5I/5crTz31lE6cOKEffvhBhYWFatOmjR599FGFhoaW6HfNmjVasmSJMjMz1bRpU/3hD38oFVtaWpqeeOIJR75JUnZ2tubPn689e/YoJydHgYGBatmypcaOHcvn8UqsovJMks6fP6+PPvpIe/fula+vr3r27Kn27duX6vPKc06aNEmffvqpjh07pgEDBmjs2LHmvGhcF0UlVEknT57Ua6+9pqCgII0aNUqXL19WYmKigoODS7U9ePCgtm/frkGDBkmSFi9erBkzZmjEiBFatWqVBg0apNzcXC1dulTvvfeeXnrpJZNfDdwpLi5Oqamp2rhxox5++GHVrl1bkhQUFFSq7YIFC7R48WL169dPLVu2VEFBgY4dO6bjx49TVKpBZs2apbCwMD3wwAM6cuSIvv32W+Xl5TmKku+//77Wr1+vnj17Kjo6Wvv27dOMGTNu+LwXLlxwHNfuvvtuBQQEKD09XVu2bDH6JaESceb8tm/fPv34448aPHiwateuzR9h1dgHH3ygzZs3a/DgwWrcuLEuXryogwcP6vTp02revLkkKTc3V9OnT1dcXJy6deumzZs368svv1TTpk0VGxurRo0aOW4YMGDAAN1yyy2SpNatW5fZ5549e/TOO+/o9ttv15gxYyRJp0+f1sGDBykqVVIVkSe/tXDhQlksFt19993KycnR8uXL9eqrr+qtt95yfJHy/fffa968eWrdurWGDh2qc+fO6c0331RgYKDq1at33Xjffvtt/fLLLxo8eLDCw8N14cIF7dmzRxkZGRzPKrGKyrOioiK98sorysjI0JAhQ1S3bl2tW7fumhN0X7x4Ua+//rq6d++uXr16qU6dOqa9ZlwfRSVUSQkJCbLb7XrllVcc35bExcVp8uTJpdqeOXNGs2bNcpycAgMDNW/ePC1cuFDvvPOO/Pz8JP166cDixYuVlpbGiawGadasmaKiorRx40Z17tz5uj/7nTt3KjY2Vo899piJEaKyCQ8P13PPPSdJGjx4sPz8/LRq1SoNHz5cdrtd69ev18CBAx2jkwYPHqx//vOf+vnnn6/7vIcOHVJeXp5efPFFtWjRwrF+9OjRxr0YVDrOnt/efvttNW7c2OwwYbKdO3eqf//+JUaAXP2Nf1ZWlp544gn17t1bktSvXz9NnDhR33//vWJjYxUcHKzY2FglJiYqOjra0e56ffr5+WnKlCny8ODePlVBReTJb+Xm5mrWrFmOz8pRUVGaNWuW1qxZo6FDh6q4uFhfffWVIiMj9dJLLzkuoWzcuLHmzZt33aJSXl6eDh06pIceeqjEaMyRI0e69ibAcBWVZ2vWrHGMsu3WrZskqX///nr22WfL7Dc7O1vjx4/XnXfeacTLggs4Q6DKsdls2r17tzp37lxi+G3jxo3Vrl27Uu3btGlTolDQsmVLSb9+SL9ykpSkVq1aSRK398Y1BQQE6JdfflFqaqq7Q4EbXRn1eMWQIUMkSSkpKUpJSZGkUt/i38y3+lcmid+xY4eKi4srIlRUMc6e32JiYigo1RABAQE6evSoMjMzr9mmVq1a6tWrl2PZy8tLLVu2LPfnGn9/f126dEl79uwp1/4wX0XnSe/evUt8Vu7atatCQkIc57rjx4/rwoULuvPOO0vMydWnTx/5+/tfN1YfHx95eXnpwIEDys3NvenXCPerqDxLSUlRSEiIunbt6ljn6+urAQMGlPmc3t7e6tu3bwW8AlQ0ikqocnJyclRUVKQGDRqU2tawYcNS666+7vvKSe7qb0+urOfEhmuJj49Xfn6+/vznP+svf/mLPv/88xuOPkH1c/Wxp379+rJYLEpLS1N6erosFovq169fok1Zx6arxcTEKC4uTklJSXr00Uf15ptv6ocffpDVaq3Q+FF5OXt+Y1RtzTFmzBidPHlSEyZM0AsvvKDExESdO3euRJt69eqVumNpQEBAuT/XDBo0SA0aNNDrr7+uxx9/XHPnztWuXbvK+xJggorOk6uPRRaLRREREUpPT5ckx/9Xt/Py8ip1Hryat7e3xowZo5SUFI0fP14vvfSSlixZouzs7Jt6rXCfisqz9PR0RURElGp3rc9MdevW5YYClRQ/FVR71xqyzVBuOCsmJkbvvvuutm3bpt27d+v777/X8uXLNX78ePXv39/d4cFNrv4w5Mrz/OUvf9Hhw4e1Y8cO7d69W++9956++eYbTZ8+XbVq1aqQflB9cCelmqN79+669dZbtXXrVu3evVvLli3TkiVLNHnyZMelJBX9uaZOnTp66623tGvXLu3atUspKSlKTk5W7969S9zYApWHO/LEFXfddZc6duzo+FyVkJCgxYsXa+rUqYqKinJ3eLgGd+UZ57zKq/IcVYCbFBQUJB8fnzIvQTpz5owbIkJV50xRIDAwUH379tVTTz2l9957T82aNdOCBQsMjA6VzdXHnrNnz8putys8PFxhYWGy2+2lvrFz5tgUHR2tBx54QDNmzNCTTz6pX375RRs3bqyQ2FG5cX7D9YSEhGjQoEF67rnnNHv2bNWuXVsLFy506jmcLYJ7eXmpU6dOGjdunN59910NGDBA69at09mzZ516HpinIvLkiquPRXa7XWfPnlVYWJgkOf6/ul1xcfFNX3YZERGh4cOH68UXX9Tbb7+t4uJiLVu2rFzxwjwVkWdhYWGOz1C/xfmu6qGohCrHw8ND7dq107Zt25SRkeFYf+rUKe3evduNkaGq8vX1lSTl5+dft93FixdLLNeqVUv169fn8qQa5l//+leJ5W+//VaS1L59e8c3dCtWrCjR5urlsuTm5pb6YBUZGSlJ5FgNwfkNZbHZbKXOT3Xq1FFISIjT869dOd/l5eXdsO3V5zwPDw81a9ZMEsekyqgi8+SKdevWqaCgwLG8efNmZWVlOc51zZs3V1BQkFavXl2ij+Tk5Bvm2KVLl1RUVFRiXf369VWrVi3mFazEKjLPYmNjlZWVpc2bNzvWXbp0SWvWrKmQWGEeLn9DlRQfH69du3Zp6tSpGjhwoGw2m7799ls1adKEOW7gtCu3P/3qq6/Uo0cPeXp6qmPHjqXaPfPMM4qJiVHz5s0VGBioY8eOacuWLaUmbkb1lpaWppkzZ6p9+/Y6fPiw1q9fr549ezoKQD169NCqVauUn5+v1q1ba+/evaVGLpVl7dq1WrVqlTp37qyIiAgVFBTou+++k5+fnzp06GDwq0JlwfkNVysoKNDjjz+url27qlmzZqpVq5b27t2rY8eOlbj70s2oX7++AgICtHr1avn5+cnX11etWrUqc36u999/X7m5uWrTpo3q1aun9PR0rVy5UpGRkWrUqFFFvTxUkIrMkysCAwM1depU9enTRxcuXNDy5csVERHhuOTfy8tLo0eP1rx58zRt2jR1795daWlpSk5OvuGcSqmpqXrllVfUrVs3NW7cWJ6entq6dasuXLig7t27lyteGK8i86x///5auXKlZs+erePHjyskJETr1q1zFL9RdVBUQpXUrFkzTZkyRZ999pkSExNVr149xcfHKysriw/dcFrLli11//33a/Xq1dq1a5fsdrtmz55dqt2QIUO0fft27dmzR1arVWFhYbr//vtL3AoX1d9TTz2lxMREzZ8/Xx4eHho8eLAeeughx/YJEyYoKChIGzZs0LZt29SmTRv953/+pyZMmHDd542JidHRo0e1adMmXbhwQf7+/mrRooWefPJJJmSuQTi/4Wq+vr4aNGiQdu/era1bt8pmsykiIkLjxo3TwIEDnXouLy8vTZo0SfPnz9cHH3ygy5cva+LEiWUeY3r16qXvvvtOq1atUl5enoKDg9WtWzfFx8dXqnl58KuKzJMrRo4cqZ9//lmLFy9WQUGBbr/9do0bN67EH/0DBgyQzWbT0qVL9cUXX6hp06Z67rnnlJCQcN3nrlevnnr06KF9+/Zp3bp18vT0VKNGjfT000+XuBsYKpeKzDNfX19NnTpVH3/8sVauXCkfHx/16tVL7du31+uvv27QK4ARLParx9oDAIBSEhMTlZSUpA8//FBBQUHuDgcAAEPs379f06ZN0zPPPEOBB8AN8TUDAAAAAAAAnEZRCQAAAAAAAE6jqAQAAAAAAACnMacSAAAAAAAAnMZIJQAAAAAAADiNohIAAAAAAACcRlEJAAAAAAAATqOoBAAAAAAAAKdRVAIAAAAAAIDTKCoBAAAAAADAaRSVAAAAAAAA4DSKSgAAAAAAAHAaRSUAAAAAAAA47f8B0NGulSmQzrUAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1400x1200 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# opeの結果の比較・可視化\n",
    "# DMはバイアスを発生させやすいですが，分散はとても小さいです．一方（trajectory-wise)の重点サンプリングを利用している推定量はバイアスは小さいものの大きな分散を発生させます．\n",
    "# またself-normalized推定量がとてもよく推定できています．\n",
    "ope.visualize_off_policy_estimates(input_dict, random_state=random_state, sharey=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead 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>cql</th>\n",
       "      <th>ddqn</th>\n",
       "      <th>random</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>dm</th>\n",
       "      <td>4.580120</td>\n",
       "      <td>11.139377</td>\n",
       "      <td>0.312904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tis</th>\n",
       "      <td>245.750211</td>\n",
       "      <td>243.984400</td>\n",
       "      <td>59.481848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pdis</th>\n",
       "      <td>24.247704</td>\n",
       "      <td>8.605686</td>\n",
       "      <td>3.429670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dr</th>\n",
       "      <td>1.632919</td>\n",
       "      <td>1.464681</td>\n",
       "      <td>0.016577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sntis</th>\n",
       "      <td>0.476102</td>\n",
       "      <td>243.984400</td>\n",
       "      <td>0.703988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>snpdis</th>\n",
       "      <td>7.658852</td>\n",
       "      <td>0.011433</td>\n",
       "      <td>0.429814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sndr</th>\n",
       "      <td>3.044155</td>\n",
       "      <td>2.626613</td>\n",
       "      <td>0.128472</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               cql        ddqn     random\n",
       "dm        4.580120   11.139377   0.312904\n",
       "tis     245.750211  243.984400  59.481848\n",
       "pdis     24.247704    8.605686   3.429670\n",
       "dr        1.632919    1.464681   0.016577\n",
       "sntis     0.476102  243.984400   0.703988\n",
       "snpdis    7.658852    0.011433   0.429814\n",
       "sndr      3.044155    2.626613   0.128472"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# OPEの推定の正確さを測るために2乗誤差を利用します\n",
    "eval_metric_ope_df = ope.evaluate_performance_of_ope_estimators(\n",
    "    input_dict, \n",
    "    metric=\"se\",\n",
    "    return_by_dataframe=True,\n",
    ")\n",
    "eval_metric_ope_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "OPEのより発展的なトピック(累積分布関数の推定など) やオフ方策選択 (OPS) は，以下を参照してください．[examples/quickstart_ja/rtb/rtb_synthetic_discrete_advanced_ja.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart_ja/rtb/rtb_synthetic_discrete_advanced_ja.ipynb). \n",
    "\n",
    "OPEのより発展的な推定量 (state(-action) marginal estimators や double reinforcement learningなど) は，以下を参照してください． [examples/quickstart_ja/rtb/rtb_synthetic_discrete_zoo_ja.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart_ja/rtb/rtb_synthetic_discrete_zoo_ja.ipynb). \n",
    "\n",
    "連続的な行動空間に対する例は，以下を参照してください． [examples/quickstart_ja/rtb/rtb_synthetic_continuous_basic_ja.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart_ja/rtb/rtb_synthetic_continuous_basic_ja.ipynb)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 参考文献\n",
    "\n",
    "- Yuta Saito, Shunsuke Aihara, Megumi Matsutani, and Yusuke Narita. \\\n",
    "\"Open Bandit Dataset and Pipeline: Towards Realistic and Reproducible Off-Policy Evaluation.\", 2021.\n",
    "\n",
    "- Takuma Seno and Michita Imai. \\\n",
    "\"d3rlpy: An Offline Deep Reinforcement Library.\", 2021.\n",
    "\n",
    "- Sergey Levine, Aviral Kumar, George Tucker, and Justin Fu. \\\n",
    "\"Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems.\" 2020.\n",
    "\n",
    "- Aviral Kumar, Aurick Zhou, George Tucker, and Sergey Levine. \\\n",
    "\"Conservative Q-Learning for Offline Reinforcement Learning.\", 2020.\n",
    "\n",
    "- Nathan Kallus and Masatoshi Uehara. \\\n",
    "\"Intrinsically Efficient, Stable, and Bounded Off-Policy Evaluation for Reinforcement Learning.\", 2019.\n",
    "\n",
    "- Hoang Le, Cameron Voloshin, and Yisong Yue. \\\n",
    "\"Batch Policy Learning under Constraints.\", 2019.\n",
    "\n",
    "- Di Wu, Xiujun Chen, Xun Yang, Hao Wang, Qing Tan, Xiaoxun Zhang, Jian Xu, and Kun Gai. \\\n",
    "\"Budget Constrained Bidding by Model-free Reinforcement Learning in Display Advertising.\", 2018.\n",
    "\n",
    "- Jun Zhao, Guang Qiu, Ziyu Guan, Wei Zhao, and Xiaofei He. \\\n",
    "\"Deep Reinforcement Learning for Sponsored Search Real-time Bidding.\", 2018.\n",
    "\n",
    "- Nan Jiang and Lihong Li. \\\n",
    "\"Doubly Robust Off-policy Value Evaluation for Reinforcement Learning.\", 2016.\n",
    "\n",
    "- Philip S. Thomas and Emma Brunskill. \\\n",
    "\"Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning.\", 2016.\n",
    "\n",
    "- Greg Brockman, Vicki Cheung, Ludwig Pettersson, Jonas Schneider, John Schulman, Jie Tang, and Wojciech Zaremba. \\\n",
    "\"OpenAI Gym.\", 2016.\n",
    "\n",
    "- Adith Swaminathan and Thorsten Joachims. \\\n",
    "\"The Self-Normalized Estimator for Counterfactual Learning.\", 2015.\n",
    "\n",
    "- Hado van Hasselt, Arthur Guez, and David Silver. \\\n",
    "\"Deep Reinforcement Learning with Double Q-learning.\", 2015.\n",
    "\n",
    "- Miroslav Dudík, Dumitru Erhan, John Langford, and Lihong Li. \\\n",
    "\"Doubly Robust Policy Evaluation and Optimization.\", 2014.\n",
    "\n",
    "- Alex Strehl, John Langford, Sham Kakade, and Lihong Li. \\\n",
    "\"Learning from Logged Implicit Exploration Data.\", 2010.\n",
    "\n",
    "- Alina Beygelzimer and John Langford. \\\n",
    "\"The Offset Tree for Learning with Partial Labels.\", 2009.\n",
    "\n",
    "- Doina Precup, Richard S. Sutton, and Satinder P. Singh. \\\n",
    "\"Eligibility Traces for Off-Policy Policy Evaluation.\", 2000."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.6"
  },
  "vscode": {
   "interpreter": {
    "hash": "eeb0e296031d5f4c93e05211083c58e9e78a3e8af86402f626ea4098c0a5e991"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}