16. Deep Reinforcement Learning with Stable Baselines/16.04. Playing Atari games with DQN and its variants.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Playing Atari games with DQN\n",
    "\n",
    "\n",
    "Now, let's learn how to create a deep Q network to play Atari games with stable baselines.\n",
    "\n",
    "First, let's import the necessary modules:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from stable_baselines import DQN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since we are dealing with Atari games we can use a convolutional neural network instead\n",
    "of a vanilla neural network. So, we use `CnnPolicy`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from stable_baselines.deepq.policies import CnnPolicy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We learned that we preprocess the game screen before feeding it to the agent. With\n",
    "baselines, we don't have to preprocess manually, instead, we can make use of `make_atari`\n",
    "module which takes care of preprocessing the game screen:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from stable_baselines.common.atari_wrappers import make_atari"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's create an Atari game environment. Let's create the Ice Hockey game\n",
    "environment:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "env = make_atari('IceHockeyNoFrameskip-v4')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Instantiate the agent:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent = DQN(CnnPolicy, env, verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Train the agent:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<stable_baselines.deepq.dqn.DQN at 0x7f63b8095198>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.learn(total_timesteps=25000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After training the agent, we can have a look at how our trained agent performs in the\n",
    "environment:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "state = env.reset()\n",
    "while True:\n",
    "    action, _ = agent.predict(state)\n",
    "    next_state, reward, done, info = env.step(action)\n",
    "    state = next_state\n",
    "    env.render()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}