REINFORCE.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "REINFORCE.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GUazkQBS2fqi",
        "colab_type": "text"
      },
      "source": [
        "## Implementing policy gradients algorithms with keras - REINFORCE agent\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v7IZNH3rJmKP",
        "colab_type": "code",
        "outputId": "b6008e7a-d9fd-4b9f-ad3d-556233f99987",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 71
        }
      },
      "source": [
        "import gym\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow.python.keras.layers import Dense, Dropout\n",
        "from tensorflow.python.keras import Sequential\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from tensorflow.python.keras.models import load_model\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8lsjUvEGn5kV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "tf.get_logger().setLevel('ERROR')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "06GmoK10ABZw",
        "colab_type": "code",
        "outputId": "4f9066ed-d529-4d3e-d322-a20770cc55e5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "## Config ##\n",
        "ENV=\"CartPole-v1\"\n",
        "RANDOM_SEED=1\n",
        "N_EPISODES=500\n",
        "\n",
        "# random seed (reproduciblity)\n",
        "np.random.seed(RANDOM_SEED)\n",
        "tf.random.set_seed(RANDOM_SEED)\n",
        "\n",
        "# set the env\n",
        "env=gym.make(ENV) # env to import\n",
        "env.seed(RANDOM_SEED)\n",
        "env.reset() # reset to env "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 0.03073904,  0.00145001, -0.03088818, -0.03131252])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZpCchoxTZxB-",
        "colab_type": "text"
      },
      "source": [
        "Implement the REINFORCE agent class"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uGtqZgGiLxS9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class REINFORCE:\n",
        "  def __init__(self, env, path=None):\n",
        "    self.env=env #import env\n",
        "    self.state_shape=env.observation_space.shape # the state space\n",
        "    self.action_shape=env.action_space.n # the action space\n",
        "    self.gamma=0.99 # decay rate of past observations\n",
        "    self.alpha=1e-4 # learning rate in the policy gradient\n",
        "    self.learning_rate=0.01 # learning rate in deep learning\n",
        "    \n",
        "    if not path:\n",
        "      self.model=self._create_model() #build model\n",
        "    else:\n",
        "      self.model=self.load_model(path) #import model\n",
        "\n",
        "    # record observations\n",
        "    self.states=[]\n",
        "    self.gradients=[] \n",
        "    self.rewards=[]\n",
        "    self.probs=[]\n",
        "    self.discounted_rewards=[]\n",
        "    self.total_rewards=[]\n",
        "  \n",
        "  def _create_model(self):\n",
        "    ''' builds the model using keras'''\n",
        "    model=Sequential()\n",
        "\n",
        "    # input shape is of observations\n",
        "    model.add(Dense(24, input_shape=self.state_shape, activation=\"relu\"))\n",
        "    #model.add(Dropout(0.5))\n",
        "    # introduce a relu layer \n",
        "    model.add(Dense(12, activation=\"relu\"))\n",
        "    #model.add(Dropout(0.5))    \n",
        "\n",
        "    # output shape is according to the number of action\n",
        "    # The softmax function outputs a probability distribution over the actions\n",
        "    model.add(Dense(self.action_shape, activation=\"softmax\")) \n",
        "    model.compile(loss=\"categorical_crossentropy\",\n",
        "            optimizer=Adam(lr=self.learning_rate))\n",
        "        \n",
        "    return model\n",
        "\n",
        "  def hot_encode_action(self, action):\n",
        "    '''encoding the actions into a binary list'''\n",
        "\n",
        "    action_encoded=np.zeros(self.action_shape, np.float32)\n",
        "    action_encoded[action]=1\n",
        "\n",
        "    return action_encoded\n",
        "  \n",
        "  def remember(self, state, action, action_prob, reward):\n",
        "    '''stores observations'''\n",
        "    encoded_action=self.hot_encode_action(action)\n",
        "    self.gradients.append(encoded_action-action_prob)\n",
        "    self.states.append(state)\n",
        "    self.rewards.append(reward)\n",
        "    self.probs.append(action_prob)\n",
        "\n",
        "  \n",
        "  def get_action(self, state):\n",
        "    '''samples the next action based on the policy probabilty distribution \n",
        "      of the actions'''\n",
        "\n",
        "    # transform state\n",
        "    state=state.reshape([1, state.shape[0]])\n",
        "    # get action probably\n",
        "    action_probability_distribution=self.model.predict(state).flatten()\n",
        "    # norm action probability distribution\n",
        "    action_probability_distribution/=np.sum(action_probability_distribution)\n",
        "    \n",
        "    # sample action\n",
        "    action=np.random.choice(self.action_shape,1,\n",
        "                            p=action_probability_distribution)[0]\n",
        "\n",
        "    return action, action_probability_distribution\n",
        "\n",
        "\n",
        "  def get_discounted_rewards(self, rewards): \n",
        "    '''Use gamma to calculate the total reward discounting for rewards\n",
        "    Following - \\gamma ^ t * Gt'''\n",
        "    \n",
        "    discounted_rewards=[]\n",
        "    cumulative_total_return=0\n",
        "    # iterate the rewards backwards and and calc the total return \n",
        "    for reward in rewards[::-1]:      \n",
        "      cumulative_total_return=(cumulative_total_return*self.gamma)+reward\n",
        "      discounted_rewards.insert(0, cumulative_total_return)\n",
        "\n",
        "    # normalize discounted rewards\n",
        "    mean_rewards=np.mean(discounted_rewards)\n",
        "    std_rewards=np.std(discounted_rewards)\n",
        "    norm_discounted_rewards=(discounted_rewards-\n",
        "                          mean_rewards)/(std_rewards+1e-7) # avoiding zero div\n",
        "    \n",
        "    return norm_discounted_rewards\n",
        "\n",
        "\n",
        "  def update_policy(self):\n",
        "    '''Updates the policy network using the NN model.\n",
        "    This function is used after the MC sampling is done - following\n",
        "    \\delta \\theta = \\alpha * gradient + log pi'''\n",
        "      \n",
        "    # get X\n",
        "    states=np.vstack(self.states)\n",
        "\n",
        "    # get Y\n",
        "    gradients=np.vstack(self.gradients)\n",
        "    rewards=np.vstack(self.rewards)\n",
        "    discounted_rewards=self.get_discounted_rewards(rewards)\n",
        "    gradients*=discounted_rewards\n",
        "    gradients=self.alpha*np.vstack([gradients])+self.probs\n",
        "\n",
        "    history=self.model.train_on_batch(states, gradients)\n",
        "    \n",
        "    self.states, self.probs, self.gradients, self.rewards=[], [], [], []\n",
        "\n",
        "    return history\n",
        "\n",
        "\n",
        "\n",
        "  def train(self, episodes, rollout_n=1, render_n=50):\n",
        "    '''train the model\n",
        "        episodes - number of training iterations \n",
        "        rollout_n- number of episodes between policy update\n",
        "        render_n - number of episodes between env rendering ''' \n",
        "    \n",
        "    env=self.env\n",
        "    total_rewards=np.zeros(episodes)\n",
        "\n",
        "    for episode in range(episodes):\n",
        "      # each episode is a new game env\n",
        "      state=env.reset()\n",
        "      done=False          \n",
        "      episode_reward=0 #record episode reward\n",
        "      \n",
        "      while not done:\n",
        "        # play an action and record the game state & reward per episode\n",
        "        action, prob=self.get_action(state)\n",
        "        next_state, reward, done, _=env.step(action)\n",
        "        self.remember(state, action, prob, reward)\n",
        "        state=next_state\n",
        "        episode_reward+=reward\n",
        "\n",
        "        #if episode%render_n==0: ## render env to visualize.\n",
        "          #env.render()\n",
        "        if done:\n",
        "          # update policy \n",
        "          if episode%rollout_n==0:\n",
        "            history=self.update_policy()\n",
        "\n",
        "      total_rewards[episode]=episode_reward\n",
        "      \n",
        "    self.total_rewards=total_rewards\n",
        "\n",
        "  def save_model(self):\n",
        "    '''saves the moodel // do after training'''\n",
        "    self.model.save('REINFORCE_model.h5')\n",
        "  \n",
        "  def load_model(self, path):\n",
        "    '''loads a trained model from path'''\n",
        "    return load_model(path)\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TVHZ6LZAgsIb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Create agent and train the model\n",
        "\n",
        "agent=REINFORCE(env)\n",
        "agent.train(500, 1)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vXe0D5dkqrBi",
        "colab_type": "code",
        "outputId": "2e6d9e79-a3ca-499d-a570-c20c0593032c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "plt.title('REINFORCE Reward')\n",
        "plt.xlabel('Episode')\n",
        "plt.ylabel('Average reward (Episode length)')\n",
        "plt.plot(agent.total_rewards)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nFtrKaaNx2yy",
        "colab_type": "text"
      },
      "source": [
        "Save the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yVr7tmqKxpyx",
        "colab_type": "code",
        "outputId": "dd034856-8a7a-4377-f5b9-1723996cdc62",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/gdrive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bfpkQePstOOi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "PATH=F\"/content/gdrive/My Drive/\" \n",
        "agent.model.save(PATH+'REINFORCE.h5')"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}