examples/quickstart/rec/rec_synthetic_customize_env.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Quickstart Example with Synthetic Recommender Simulation (Customization)\n",
    "This notebook provides an example of customizing the Recommender environment.\n",
    "\n",
    "This example on the Synthetic Recommender Simulation consists of the following 3 steps:\n",
    "1. Setup Synthetic Recommender Simulation Environment and Interacting Online RL Agent\n",
    "2. Standardized Environment\n",
    "3. Customize Environmental Configuration\n",
    "\n",
    "\\* This library uses [d3rlpy](https://github.com/takuseno/d3rlpy)'s algorithm implementations.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import SCOPE-RL modules\n",
    "import scope_rl\n",
    "from scope_rl.policy import OnlineHead\n",
    "\n",
    "# import recgym modules\n",
    "from recgym import RECEnv\n",
    "from recgym import BaseUserModel\n",
    "from recgym.utils import sigmoid\n",
    "\n",
    "# import d3rlpy algorithms\n",
    "from d3rlpy.algos import DiscreteRandomPolicyConfig\n",
    "\n",
    "# import from other libraries\n",
    "import gym\n",
    "\n",
    "from sklearn.utils import check_random_state\n",
    "from sklearn.utils import check_scalar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from dataclasses import dataclass\n",
    "from typing import Tuple, Union, Optional\n",
    "\n",
    "import torch\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": [
      "0.1.2\n"
     ]
    }
   ],
   "source": [
    "# version\n",
    "print(scope_rl.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set random state\n",
    "random_state = 12345\n",
    "random_ = check_random_state(random_state)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = \"cuda:0\" if torch.cuda.is_available() else \"cpu\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ignore warnings\n",
    "import warnings\n",
    "warnings.simplefilter('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Setup Synthetic Recommender Simulation Environment and Interacting Online RL Agent\n",
    "To begin with, we briefly describe the basic usage of the environment.\n",
    "\n",
    "#### RL setup for Recommendation\n",
    "In recommendation, the objective of the RL agent is to maximize reward\n",
    "\n",
    "We often formulate this recommendation problem as the following (Partially Observable) Markov Decision Process ((PO)MDP):\n",
    "- `state`: \n",
    "   - A vector representing user preference.  The preference changes over time in an episode by the actions presented by the RL agent.\n",
    "   - When the true state is unobservable, the agent uses observation instead of state.\n",
    "- `action`: Indicating which item to present to the user.\n",
    "- `reward`: User engagement signal. Either binary or continuous.\n",
    "\n",
    "Let's see how it works!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# setup environment\n",
    "env = RECEnv(random_state=random_state)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define a random agent\n",
    "agent = OnlineHead(\n",
    "    DiscreteRandomPolicyConfig().create(device=device),\n",
    "    name=\"random\",\n",
    ")\n",
    "agent.build_with_env(env)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Discrete(5)\n",
      "Box(-1.0, 1.0, (5,), float64)\n"
     ]
    }
   ],
   "source": [
    "print(env.action_space)\n",
    "print(env.observation_space)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "59286ec2093e49188c36533f85cdb759",
       "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"
    }
   ],
   "source": [
    "import scope_rl\n",
    "from scope_rl.ope.online import calc_on_policy_policy_value\n",
    "# calculate on-policy policy value\n",
    "on_policy_performance = calc_on_policy_policy_value(\n",
    "  env,\n",
    "  agent,\n",
    "  n_trajectories=100,\n",
    "  random_state=12345\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-0.35236748239501536\n"
     ]
    }
   ],
   "source": [
    "print(on_policy_performance)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# interact agent with the environment\n",
    "# only 6 lines are needed for RL interaction\n",
    "for episode in range(10):\n",
    "    obs, info = env.reset()\n",
    "    done = False\n",
    "\n",
    "    while not done:\n",
    "        action = agent.sample_action_online(obs)\n",
    "        obs, reward, done, truncated, info = env.step(action)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5,)\n",
      "[-0.04552487  0.1208106   0.11597855  0.14365492 -0.15756327]\n"
     ]
    }
   ],
   "source": [
    "# state \n",
    "print(obs.shape)\n",
    "print(obs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# let's visualize the transition of reward\n",
    "def visualize_interaction_on_environment(env, agent):\n",
    "    obs, info = env.reset()\n",
    "    done = False\n",
    "    reward_list = []\n",
    "\n",
    "    while not done:\n",
    "        action = agent.sample_action_online(obs)\n",
    "        obs, reward, done, truncated, info = env.step(action)\n",
    "        reward_list.append(reward)\n",
    "\n",
    "    # plot\n",
    "    fig = plt.figure()\n",
    "    ax1 = fig.add_subplot(111)\n",
    "    ax1.plot(reward_list[:-1], label='reward', color='tab:orange')\n",
    "    ax1.set_xlabel('timestep')\n",
    "    ax1.set_ylabel('reward')\n",
    "    ax1.legend(loc='upper left')\n",
    "    plt.show()\n",
    "    \n",
    "visualize_interaction_on_environment(env, agent)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Standardized Environment\n",
    "\n",
    "We provide standardize recommender environment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "env = gym.make('RECEnv-v0')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Customize Environmental Configuration\n",
    "Now, we customize the recommender simulation environment.\n",
    "\n",
    "We have the following environmental configurations:\n",
    "- `step_per_episode`: Number of timesteps in an episode.\n",
    "- `n_items`: Number of items used in the recommender system.\n",
    "- `n_users`: Number of users used in the recommender system.\n",
    "- `item_feature_dim`: Dimension of the item feature vectors.\n",
    "- `user_feature_dim`: Dimension of the user feature vectors.\n",
    "- `item_feature_vector`: Feature vectors that characterize each item.\n",
    "- `user_feature_vector`: Feature vectors that characterize each user.\n",
    "- `reward_type`: Reward type.\n",
    "- `reward_std`: Noise level of the reward. Applicable only when reward_type is \"continuous\".\n",
    "- `obs_std`: Noise level of the state observation.\n",
    "- `UserModel`: User model that defines the user preference dynamics and reward function.\n",
    "- `random_state`: Random state."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Case 1. Customizing the Basic Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "env = RECEnv(\n",
    "    step_per_episode=10,\n",
    "    n_items=100,\n",
    "    n_users=100,\n",
    "    item_feature_dim=5,\n",
    "    user_feature_dim=5,\n",
    "    reward_type=\"continuous\",  # \"binary\"\n",
    "    reward_std=0.3,\n",
    "    obs_std=0.3,\n",
    "    random_state=12345,\n",
    ")\n",
    "visualize_interaction_on_environment(env, agent)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Case 2. Defining the Items and Users"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature1</th>\n",
       "      <th>feature2</th>\n",
       "      <th>feature3</th>\n",
       "      <th>feature4</th>\n",
       "      <th>feature5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ad1</th>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad3</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad4</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad5</th>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad96</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad97</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad98</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad99</th>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad100</th>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       feature1  feature2  feature3  feature4  feature5\n",
       "ad1          -1         1        -1        -1         1\n",
       "ad2           1         1         1         1         1\n",
       "ad3           1        -1         1         1         1\n",
       "ad4           1        -1        -1         1        -1\n",
       "ad5          -1         1         1        -1        -1\n",
       "...         ...       ...       ...       ...       ...\n",
       "ad96          1        -1        -1        -1         1\n",
       "ad97          1         1        -1         1         1\n",
       "ad98         -1        -1         1         1        -1\n",
       "ad99         -1         1        -1         1         1\n",
       "ad100        -1         1         1         1         1\n",
       "\n",
       "[100 rows x 5 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# we use the following items for recommendation\n",
    "item_features = np.sign(random_.normal(size=(100, 5))).astype(int)\n",
    "item_names, feature_names = [f\"ad{i+1}\" for i in range(100)], [f\"feature{i+1}\" for i in range(5)]\n",
    "item_df = pd.DataFrame(item_features, columns=feature_names, index=item_names)\n",
    "item_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature1</th>\n",
       "      <th>feature2</th>\n",
       "      <th>feature3</th>\n",
       "      <th>feature4</th>\n",
       "      <th>feature5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>user1</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user2</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user3</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user5</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user96</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user97</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user98</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user99</th>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user100</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         feature1  feature2  feature3  feature4  feature5\n",
       "user1          -1        -1        -1         1         1\n",
       "user2          -1        -1         1         1        -1\n",
       "user3          -1        -1         1         1        -1\n",
       "user4           1         1         1         1         1\n",
       "user5          -1        -1         1         1         1\n",
       "...           ...       ...       ...       ...       ...\n",
       "user96         -1        -1        -1        -1         1\n",
       "user97          1        -1         1         1        -1\n",
       "user98          1        -1         1        -1        -1\n",
       "user99         -1         1        -1         1        -1\n",
       "user100         1        -1        -1         1         1\n",
       "\n",
       "[100 rows x 5 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# we assume that the following 100 users \n",
    "user_features = np.sign(random_.normal(size=(100, 5))).astype(int)\n",
    "user_names, feature_names = [f\"user{i+1}\" for i in range(100)], [f\"feature{i+1}\" for i in range(5)]\n",
    "user_df = pd.DataFrame(user_features, columns=feature_names, index=user_names)\n",
    "user_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# using the above data, we can simulate an recommendation as follows\n",
    "env = RECEnv(\n",
    "        n_items=100,\n",
    "        n_users=100,\n",
    "        item_feature_dim=5,\n",
    "        user_feature_dim=5,\n",
    "        item_feature_vector=item_features,\n",
    "        user_feature_vector=user_features,\n",
    "        random_state = 12345,\n",
    ")\n",
    "visualize_interaction_on_environment(env, agent)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Case 3. Using Customized UserModel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "@dataclass\n",
    "class CustomizedUserModel(BaseUserModel):\n",
    "    user_feature_dim: int\n",
    "    item_feature_dim: int\n",
    "    reward_type: str = \"continuous\"  # \"binary\"\n",
    "    reward_std: float = 0.0\n",
    "    random_state: Optional[int] = None\n",
    "\n",
    "    def __post_init__(self):\n",
    "        self.random_ = check_random_state(self.random_state)\n",
    "        self.coef = self.random_.normal(size=(self.user_feature_dim, self.item_feature_dim))\n",
    "\n",
    "    def user_preference_dynamics(\n",
    "        self,\n",
    "        state: np.ndarray,\n",
    "        action: Union[int, np.ndarray],\n",
    "        item_feature_vector: np.ndarray,\n",
    "        alpha: float = 1.0,\n",
    "    ) -> np.ndarray:\n",
    "        coefficient = state.T @ self.coef @ item_feature_vector[action]\n",
    "        state = state + alpha * coefficient * item_feature_vector[action]\n",
    "        state = state / np.linalg.norm(state, ord=2)\n",
    "        return state\n",
    "\n",
    "    def reward_function(\n",
    "        self,\n",
    "        state: np.ndarray,\n",
    "        action: Union[int, np.ndarray],\n",
    "        item_feature_vector: np.ndarray,\n",
    "    ) -> float:\n",
    "        logit = state.T @ self.coef @ item_feature_vector[action]\n",
    "        reward = (\n",
    "            logit if self.reward_type == \"continuous\" else sigmoid(logit)\n",
    "        )\n",
    "\n",
    "        if self.reward_type == \"discrete\":\n",
    "            reward = self.random_.binominal(1, p=reward)\n",
    "\n",
    "        return reward"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# setup environment\n",
    "env = RECEnv(\n",
    "    UserModel=CustomizedUserModel,\n",
    "    random_state=12345\n",
    ")\n",
    "visualize_interaction_on_environment(env, agent)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For data collection and visualization, please refer to [examples/quickstart/rec/rec_synthetic_data_collection.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart/rec/rec_synthetic_data_collection.ipynb). \n",
    "\n",
    "For offline RL and OPE, please refer to [examples/quickstart/rec/rec_synthetic_discrete_basic.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart/rec/rec_synthetic_discrete_basic.ipynb).\n",
    "\n",
    "For advanced topics regarding OPE and OPS, please refer to [examples/quickstart/rec/rec_synthetic_discrete_advanced.ipynb](https://github.com/hakuhodo-technologies/scope-rl/blob/main/examples/quickstart/rec/rec_synthetic_discrete_advanced.ipynb)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reference\n",
    "- Sarah Dean and Jamie Morgenstern. \\\n",
    "\"Preference Dynamics Under Personalized Recommendations.\", 2022.\n",
    "\n",
    "- Takuma Seno and Michita Imai. \\\n",
    "\"d3rlpy: An Offline Deep Reinforcement Library.\", 2021.\n",
    "\n",
    "- David Rohde, Stephen Bonner, Travis Dunlop, Flavian Vasile, Alexandros Karatzoglou. \\\n",
    "\"RecoGym: A Reinforcement Learning Environment for the Problem of Product Recommendation in Online Advertising.\" 2018.\n",
    "\n",
    "- Greg Brockman, Vicki Cheung, Ludwig Pettersson, Jonas Schneider, John Schulman, Jie Tang, and Wojciech Zaremba. \\\n",
    "\"OpenAI Gym.\", 2016."
   ]
  },
  {
   "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": "70404ee114725fce8ed9e697d67827f8546c678889944e6d695790702cbfe1f5"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}