8_reinforcement_learning.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "8.reinforcement_learning.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "eXNnnap-mvj6",
        "colab_type": "code",
        "outputId": "24bc9974-5ec3-420a-b1b9-2270d521288e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "!apt-get update > /dev/null 2>&1\n",
        "!apt-get install cmake > /dev/null 2>&1\n",
        "!pip install --upgrade setuptools 2>&1\n",
        "!pip install ez_setup > /dev/null 2>&1\n",
        "!pip install gym[atari] > /dev/null 2>&1"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already up-to-date: setuptools in /usr/local/lib/python3.6/dist-packages (41.0.1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QEprgkonBwZc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import seaborn as sns\n",
        "from pylab import rcParams\n",
        "import matplotlib.pyplot as plt\n",
        "import gym\n",
        "import math\n",
        "from matplotlib import rc\n",
        "import sys\n",
        "from contextlib import closing\n",
        "from io import StringIO\n",
        "from gym import utils\n",
        "from gym.envs.toy_text import discrete\n",
        "from gym.envs.registration import register\n",
        "import time\n",
        "from IPython.display import clear_output\n",
        "from scipy.signal import savgol_filter\n",
        "\n",
        "%matplotlib inline\n",
        "\n",
        "sns.set(style='whitegrid', palette='muted', font_scale=1.3)\n",
        "\n",
        "rcParams['figure.figsize'] = 14, 8\n",
        "\n",
        "RANDOM_SEED = 42\n",
        "\n",
        "np.random.seed(RANDOM_SEED)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "L4_UapBGNtzT",
        "colab_type": "text"
      },
      "source": [
        "# Environment"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cg7W9NcITrPv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "MAP = [\n",
        "    \"+---------+\",\n",
        "    \"|R: :A: :G|\",\n",
        "    \"| : : : : |\",\n",
        "    \"| :A:A: : |\",\n",
        "    \"| : : : : |\",\n",
        "    \"|Y: :A:B: |\",\n",
        "    \"+---------+\",\n",
        "]\n",
        "\n",
        "\n",
        "class SafeCab(discrete.DiscreteEnv):\n",
        "    \"\"\"\n",
        "    Modified Taxi Problem\n",
        "    from \"Hierarchical Reinforcement Learning with the MAXQ Value Function Decomposition\"\n",
        "    by Tom Dietterich\n",
        "\n",
        "    Description:\n",
        "    There are four designated locations in the grid world indicated by R(ed), B(lue), G(reen), and Y(ellow). When the episode starts, the taxi starts off at a random square and the passenger is at a random location. The taxi drive to the passenger's location, pick up the passenger, drive to the passenger's destination (another one of the four specified locations), and then drop off the passenger. The episode ends if the taxi ends up on anomaly location. Once the passenger is dropped off, the episode ends.\n",
        "\n",
        "    Observations: \n",
        "    There are 500 discrete states since there are 25 taxi positions, 5 possible locations of the passenger (including the case when the passenger is the taxi), and 4 destination locations. \n",
        "    \n",
        "    Actions: \n",
        "    There are 6 discrete deterministic actions:\n",
        "    - 0: move south\n",
        "    - 1: move north\n",
        "    - 2: move east \n",
        "    - 3: move west \n",
        "    - 4: pickup passenger\n",
        "    - 5: dropoff passenger\n",
        "    \n",
        "    Rewards: \n",
        "    There is a reward of -1 for each action and an additional reward of +1000 for delievering the passenger. There is a reward of -10 for executing actions \"pickup\" and \"dropoff\" illegally. Interaction with an anomaly gives -1000 reward and end the episode.\n",
        "    \n",
        "\n",
        "    Rendering:\n",
        "    - blue: passenger\n",
        "    - magenta: destination\n",
        "    - yellow: empty taxi\n",
        "    - green: full taxi\n",
        "    - red: anomaly\n",
        "    - other letters (R, G, B and Y): locations for passengers and destinations\n",
        "\n",
        "    actions:\n",
        "    - 0: south\n",
        "    - 1: north\n",
        "    - 2: east\n",
        "    - 3: west\n",
        "    - 4: pickup\n",
        "    - 5: dropoff\n",
        "\n",
        "    state space is represented by:\n",
        "        (taxi_row, taxi_col, passenger_location, destination)\n",
        "    \"\"\"\n",
        "    metadata = {'render.modes': ['human', 'ansi']}\n",
        "\n",
        "    def __init__(self):\n",
        "        self.desc = np.asarray(MAP, dtype='c')\n",
        "\n",
        "        self.locs = locs = [(0,0), (0,4), (4,0), (4,3)]\n",
        "        self.anomaly_locs = [(0,2), (2,1), (2,2), (4,2)]\n",
        "\n",
        "        num_states = 500\n",
        "        num_rows = 5\n",
        "        num_columns = 5\n",
        "        max_row = num_rows - 1\n",
        "        max_col = num_columns - 1\n",
        "        initial_state_distrib = np.zeros(num_states)\n",
        "        num_actions = 6\n",
        "        P = {state: {action: []\n",
        "                     for action in range(num_actions)} for state in range(num_states)}\n",
        "        for row in range(num_rows):\n",
        "            for col in range(num_columns):\n",
        "                for pass_idx in range(len(locs) + 1):  # +1 for being inside taxi\n",
        "                    for dest_idx in range(len(locs)):\n",
        "                        state = self.encode(row, col, pass_idx, dest_idx)\n",
        "                        if pass_idx < 4 and pass_idx != dest_idx:\n",
        "                            initial_state_distrib[state] += 1\n",
        "                        for action in range(num_actions):\n",
        "                            # defaults\n",
        "                            new_row, new_col, new_pass_idx = row, col, pass_idx\n",
        "                            reward = -1 # default reward when there is no pickup/dropoff\n",
        "                            done = False\n",
        "                            taxi_loc = (row, col)\n",
        "\n",
        "                            if action == 0:\n",
        "                                new_row = min(row + 1, max_row)\n",
        "                            elif action == 1:\n",
        "                                new_row = max(row - 1, 0)\n",
        "                            if action == 2:\n",
        "                                new_col = min(col + 1, max_col)\n",
        "                            elif action == 3:\n",
        "                                new_col = max(col - 1, 0)\n",
        "                            elif action == 4:  # pickup\n",
        "                                if (pass_idx < 4 and taxi_loc == locs[pass_idx]):\n",
        "                                    new_pass_idx = 4\n",
        "                                else: # passenger not at location\n",
        "                                    reward = -10\n",
        "                            elif action == 5:  # dropoff\n",
        "                                if (taxi_loc == locs[dest_idx]) and pass_idx == 4:\n",
        "                                    new_pass_idx = dest_idx\n",
        "                                    done = True\n",
        "                                    reward = 1000\n",
        "                                elif (taxi_loc in locs) and pass_idx == 4:\n",
        "                                    new_pass_idx = locs.index(taxi_loc)\n",
        "                                else: # dropoff at wrong location\n",
        "                                    reward = -10\n",
        "                                    \n",
        "                            new_loc = (new_row, new_col)\n",
        "                            if new_loc in self.anomaly_locs:\n",
        "                              reward = -1000\n",
        "                              done = True\n",
        "                            new_state = self.encode(\n",
        "                                new_row, new_col, new_pass_idx, dest_idx)\n",
        "                            P[state][action].append(\n",
        "                                (1.0, new_state, reward, done))\n",
        "        initial_state_distrib /= initial_state_distrib.sum()\n",
        "        discrete.DiscreteEnv.__init__(\n",
        "            self, num_states, num_actions, P, initial_state_distrib)\n",
        "\n",
        "    def encode(self, taxi_row, taxi_col, pass_loc, dest_idx):\n",
        "        # (5) 5, 5, 4\n",
        "        i = taxi_row\n",
        "        i *= 5\n",
        "        i += taxi_col\n",
        "        i *= 5\n",
        "        i += pass_loc\n",
        "        i *= 4\n",
        "        i += dest_idx\n",
        "        return i\n",
        "\n",
        "    def decode(self, i):\n",
        "        out = []\n",
        "        out.append(i % 4)\n",
        "        i = i // 4\n",
        "        out.append(i % 5)\n",
        "        i = i // 5\n",
        "        out.append(i % 5)\n",
        "        i = i // 5\n",
        "        out.append(i)\n",
        "        assert 0 <= i < 5\n",
        "        return reversed(out)\n",
        "\n",
        "    def render(self, mode='human'):\n",
        "        outfile = StringIO() if mode == 'ansi' else sys.stdout\n",
        "\n",
        "        out = self.desc.copy().tolist()\n",
        "        out = [[c.decode('utf-8') for c in line] for line in out]\n",
        "        taxi_row, taxi_col, pass_idx, dest_idx = self.decode(self.s)\n",
        "        \n",
        "        def ul(x): return \"_\" if x == \" \" else x\n",
        "        \n",
        "        if pass_idx < 4:\n",
        "            out[1 + taxi_row][2 * taxi_col + 1] = utils.colorize(\n",
        "                out[1 + taxi_row][2 * taxi_col + 1], 'yellow', highlight=True)\n",
        "            pi, pj = self.locs[pass_idx]\n",
        "            out[1 + pi][2 * pj + 1] = utils.colorize(out[1 + pi][2 * pj + 1], 'blue', bold=True)\n",
        "        else:  # passenger in taxi\n",
        "            out[1 + taxi_row][2 * taxi_col + 1] = utils.colorize(\n",
        "                ul(out[1 + taxi_row][2 * taxi_col + 1]), 'green', highlight=True)\n",
        "\n",
        "        di, dj = self.locs[dest_idx]\n",
        "        out[1 + di][2 * dj + 1] = utils.colorize(out[1 + di][2 * dj + 1], 'magenta')\n",
        "        \n",
        "        for (zx, zy) in self.anomaly_locs:\n",
        "          out[1 + zx][2 * zy + 1] = utils.colorize(\n",
        "            out[1 + zx][2 * zy + 1], 'red', bold=True)\n",
        "        \n",
        "        outfile.write(\"\\n\".join([\"\".join(row) for row in out]) + \"\\n\")\n",
        "        if self.lastaction is not None:\n",
        "            outfile.write(\"  ({})\\n\".format([\"South\", \"North\", \"East\", \"West\", \"Pickup\", \"Dropoff\"][self.lastaction]))\n",
        "        else: outfile.write(\"\\n\")\n",
        "\n",
        "        # No need to return anything for human\n",
        "        if mode != 'human':\n",
        "            with closing(outfile):\n",
        "                return outfile.getvalue()\n",
        "              \n",
        "register(\n",
        "    id='SafeCab-v0',\n",
        "    entry_point=f\"{__name__}:SafeCab\",\n",
        "    timestep_limit=100,\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LpZP-vrmEFSo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def print_frames(frames):\n",
        "  for i, frame in enumerate(frames):\n",
        "    clear_output(wait=True)\n",
        "    print(frame['frame'])\n",
        "    print(f\"Episode: {frame['episode']}\")\n",
        "    print(f\"Timestep: {frame['step']}\")    \n",
        "    print(f\"State: {frame['state']}\")\n",
        "    print(f\"Reward: {frame['reward']}\")\n",
        "    time.sleep(.4)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ScyqDf17m-7b",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "env = gym.make('SafeCab-v0')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w-uiTLLwNlY-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        },
        "outputId": "64b00ab9-7172-415f-cd76-fb6d1345d77d"
      },
      "source": [
        "env.reset()\n",
        "env.render()\n",
        "\n",
        "print(\"Action Space {}\".format(env.action_space))\n",
        "print(\"State Space {}\".format(env.observation_space))"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "+---------+\n",
            "|R: :\u001b[31;1mA\u001b[0m: :\u001b[34;1mG\u001b[0m|\n",
            "| : : : : |\n",
            "| :\u001b[31;1mA\u001b[0m:\u001b[31;1mA\u001b[0m: : |\n",
            "| : : : : |\n",
            "|\u001b[35mY\u001b[0m: :\u001b[31;1mA\u001b[0m:\u001b[43mB\u001b[0m: |\n",
            "+---------+\n",
            "\n",
            "Action Space Discrete(6)\n",
            "State Space Discrete(500)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nkiWPxzlmz9p",
        "colab_type": "code",
        "outputId": "367ce4f5-0989-48d5-9508-1077093a17b7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "env.action_space"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Discrete(6)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YiicmIvWvPpb",
        "colab_type": "code",
        "outputId": "02950c7a-3e24-4283-ecbb-4fae90f0c2d7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "action_size = env.action_space.n\n",
        "print(\"Action size \", action_size)\n",
        "\n",
        "state_size = env.observation_space.n\n",
        "print(\"State size \", state_size)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Action size  6\n",
            "State size  500\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2Odt8hLyN2vD",
        "colab_type": "text"
      },
      "source": [
        "# Q-Learning agent"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eaFQeRv-yLrO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class Agent:\n",
        "  \n",
        "  def __init__(self, n_states, n_actions, decay_rate=0.0001, learning_rate=0.7, gamma=0.618):\n",
        "    self.n_actions = n_actions\n",
        "    self.q_table = np.zeros((n_states, n_actions))\n",
        "    self.epsilon = 1.0\n",
        "    self.max_epsilon = 1.0\n",
        "    self.min_epsilon = 0.01\n",
        "    self.decay_rate = decay_rate\n",
        "    self.learning_rate = learning_rate\n",
        "    self.gamma = gamma # discount rate\n",
        "    self.epsilons_ = []\n",
        "    \n",
        "  def choose_action(self, explore=True):\n",
        "    exploration_tradeoff = np.random.uniform(0, 1)\n",
        "    \n",
        "    if explore and exploration_tradeoff < self.epsilon:\n",
        "      # exploration\n",
        "      return np.random.randint(self.n_actions)    \n",
        "    else:\n",
        "      # exploitation (taking the biggest Q value for this state)\n",
        "      return np.argmax(self.q_table[state, :])\n",
        "  \n",
        "  def learn(self, state, action, reward, next_state, done, episode):\n",
        "    # Update Q(s,a):= Q(s,a) + lr [R(s,a) + gamma * max Q(s',a') - Q(s,a)]\n",
        "    self.q_table[state, action] = self.q_table[state, action] + \\\n",
        "      self.learning_rate * (reward + self.gamma * \\\n",
        "        np.max(self.q_table[new_state, :]) - self.q_table[state, action])\n",
        "    \n",
        "    if done:\n",
        "      # Reduce epsilon to decrease the exploration over time\n",
        "      self.epsilon = self.min_epsilon + (self.max_epsilon - self.min_epsilon) * \\\n",
        "        np.exp(-self.decay_rate * episode)\n",
        "      self.epsilons_.append(self.epsilon)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3pKVLCOJN5I4",
        "colab_type": "text"
      },
      "source": [
        "# Training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PwJCcJ3nBS2b",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "total_episodes = 60000\n",
        "total_test_episodes = 10\n",
        "\n",
        "agent = Agent(env.observation_space.n, env.action_space.n)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5WC5IrXCBXt3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "untrained_frames = []\n",
        "\n",
        "for episode in range(total_test_episodes):\n",
        "    state = env.reset()\n",
        "\n",
        "    step = 1\n",
        "    \n",
        "    while True:\n",
        "        action = agent.choose_action()\n",
        "        \n",
        "        new_state, reward, done, info = env.step(action)        \n",
        "        \n",
        "        untrained_frames.append({\n",
        "          'frame': env.render(mode='ansi'),\n",
        "          'state': state,\n",
        "          'episode': episode + 1,\n",
        "          'step': step,\n",
        "          'reward': reward\n",
        "        })\n",
        "        \n",
        "        if done:\n",
        "            step = 0\n",
        "            break\n",
        "        state = new_state\n",
        "        step += 1"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-UX5vKIQD4Nd",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "outputId": "c3a5cd97-6bb7-4f85-8d12-aefa432212c8"
      },
      "source": [
        "print_frames(untrained_frames)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "+---------+\n",
            "|R: :\u001b[31;1mA\u001b[0m: :\u001b[35mG\u001b[0m|\n",
            "| : : : : |\n",
            "| :\u001b[31;1m\u001b[43mA\u001b[0m\u001b[0m:\u001b[31;1mA\u001b[0m: : |\n",
            "| : : : : |\n",
            "|Y: :\u001b[31;1mA\u001b[0m:\u001b[34;1mB\u001b[0m: |\n",
            "+---------+\n",
            "  (Dropoff)\n",
            "\n",
            "Episode: 10\n",
            "Timestep: 1\n",
            "State: 233\n",
            "Reward: -1000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1DstwZEW8PQn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "rewards = []\n",
        "\n",
        "for episode in range(total_episodes):\n",
        "    state = env.reset()\n",
        "    episode_rewards = []\n",
        "    \n",
        "    while True:\n",
        "            \n",
        "        action = agent.choose_action()\n",
        "        \n",
        "        # Take the action (a) and observe the outcome state(s') and reward (r)\n",
        "        new_state, reward, done, info = env.step(action)\n",
        "        \n",
        "        agent.learn(state, action, reward, new_state, done, episode)\n",
        "\n",
        "        state = new_state\n",
        "        \n",
        "        episode_rewards.append(reward)\n",
        "        \n",
        "        if done == True:\n",
        "          break\n",
        "          \n",
        "    rewards.append(np.mean(episode_rewards))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vkxwgcSKFX1X",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 526
        },
        "outputId": "d331a45f-4324-4bd1-b5ad-3e59a632b95a"
      },
      "source": [
        "plt.plot(savgol_filter(rewards, 1001, 2))\n",
        "plt.title(\"Smoothened training reward per episode\")\n",
        "plt.xlabel('Episode');\n",
        "plt.ylabel('Total Reward');"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2UAAAH9CAYAAACa3NRUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAbEcoAAAAAwEb/D+8U9QDptFDGAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1008x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0jM6og93ck5a",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 526
        },
        "outputId": "85ac2704-05d7-41b0-d336-c244bde8da7e"
      },
      "source": [
        "plt.plot(agent.epsilons_)\n",
        "plt.title(\"Epsilon for episode\")\n",
        "plt.xlabel('Episode');\n",
        "plt.ylabel('Epsilon');"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1QAAAH9CAYAAAAd7T1SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ntRADFRERERERUQv9PzkeqURFdsB5AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1008x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TDQgo2Y5N7jW",
        "colab_type": "text"
      },
      "source": [
        "# Evaluation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qQFHVyicAw4_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "frames = []\n",
        "\n",
        "rewards = []\n",
        "\n",
        "for episode in range(total_test_episodes):\n",
        "    state = env.reset()\n",
        "    episode_rewards = []\n",
        "\n",
        "    step = 1\n",
        "    \n",
        "    while True:\n",
        "        action = agent.choose_action(explore=False)\n",
        "        \n",
        "        new_state, reward, done, info = env.step(action)        \n",
        "        \n",
        "        frames.append({\n",
        "          'frame': env.render(mode='ansi'),\n",
        "          'state': state,\n",
        "          'episode': episode + 1,\n",
        "          'step': step,\n",
        "          'reward': reward\n",
        "        })\n",
        "        \n",
        "        episode_rewards.append(reward)\n",
        "        \n",
        "        if done:\n",
        "            step = 0\n",
        "            break\n",
        "        state = new_state\n",
        "        step += 1\n",
        "    \n",
        "    rewards.append(np.mean(episode_rewards))\n",
        "\n",
        "env.close()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YBWtJzAjItdP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 526
        },
        "outputId": "2a3210bd-db24-4755-a12f-465af802580d"
      },
      "source": [
        "plt.plot(rewards)\n",
        "plt.title(\"Test reward per episode\")\n",
        "plt.ylim((0, 150))\n",
        "plt.xlabel('Episode');\n",
        "plt.ylabel('Total Reward');"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1kAAAH9CAYAAADoE7biAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAABqJkAQAAAICBKFkAAAAAYKD/BTTuPR5+4rrkAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1008x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "04A1wDm93LSm",
        "colab_type": "code",
        "outputId": "48c03f43-3c86-43a1-befb-96e7ac9c8476",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        }
      },
      "source": [
        "print_frames(frames)"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "+---------+\n",
            "|R: :\u001b[31;1mA\u001b[0m: :\u001b[35m\u001b[34;1m\u001b[43mG\u001b[0m\u001b[0m\u001b[0m|\n",
            "| : : : : |\n",
            "| :\u001b[31;1mA\u001b[0m:\u001b[31;1mA\u001b[0m: : |\n",
            "| : : : : |\n",
            "|Y: :\u001b[31;1mA\u001b[0m:B: |\n",
            "+---------+\n",
            "  (Dropoff)\n",
            "\n",
            "Episode: 10\n",
            "Timestep: 11\n",
            "State: 97\n",
            "Reward: 1000\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}