taxi_2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from agent import Agent\n",
    "from monitor import interact\n",
    "import gym\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode 27/20000 || Best average reward -inf"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/sungmok/udacity/gym/gym/__init__.py:22: UserWarning: DEPRECATION WARNING: to improve load times, gym no longer automatically loads gym.spaces. Please run \"import gym.spaces\" to load gym.spaces on your own. This warning will turn into an error in a future version of gym.\n",
      "  warnings.warn('DEPRECATION WARNING: to improve load times, gym no longer automatically loads gym.spaces. Please run \"import gym.spaces\" to load gym.spaces on your own. This warning will turn into an error in a future version of gym.')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode 20000/20000 || Best average reward 9.492\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1afa87a860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "env = gym.make('Taxi-v2')\n",
    "agent = Agent()\n",
    "avg_rewards, best_avg_reward = interact(env, agent)\n",
    "\n",
    "plt.figure(figsize=(20, 10))\n",
    "plt.plot(avg_rewards)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}