Q-Learning Frozen Lake.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Reinforcement Learning:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gym\n",
    "import collections\n",
    "from tensorboardX import SummaryWriter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "ENV_NAME = \"FrozenLake-v0\"\n",
    "GAMMA = 0.9\n",
    "ALPHA = 0.2\n",
    "TEST_EPISODES = 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Agent:\n",
    "    def __init__(self):\n",
    "        self.env = gym.make(ENV_NAME) # Make environment.\n",
    "        self.state = self.env.reset() # Get start state.\n",
    "        self.values = collections.defaultdict(float) # Q table.\n",
    "\n",
    "    def sample_env(self): # Gets the pair: state, action, reward, next_state.\n",
    "        action = self.env.action_space.sample()\n",
    "        old_state = self.state\n",
    "        new_state, reward, is_done, _ = self.env.step(action)\n",
    "        self.state = self.env.reset() if is_done else new_state\n",
    "        return (old_state, action, reward, new_state)\n",
    "\n",
    "    def best_value_and_action(self, state): # Chooses best action among many possible.\n",
    "        best_value, best_action = None, None\n",
    "        for action in range(self.env.action_space.n): # For all action.\n",
    "            action_value = self.values[(state, action)] # Get the Q value for state-action.\n",
    "            if best_value is None or best_value < action_value:\n",
    "                best_value = action_value\n",
    "                best_action = action\n",
    "        return best_value, best_action\n",
    "\n",
    "    def value_update(self, s, a, r, next_s): # Bellman update with Blending.\n",
    "        best_v, _ = self.best_value_and_action(next_s)\n",
    "        new_val = r + GAMMA * best_v\n",
    "        old_val = self.values[(s, a)]\n",
    "        self.values[(s, a)] = old_val * (1-ALPHA) + new_val * ALPHA\n",
    "\n",
    "    def play_episode(self, env): # Run simulation.\n",
    "        total_reward = 0.0\n",
    "        state = env.reset()\n",
    "        while True:\n",
    "            _, action = self.best_value_and_action(state) # Get best action for current state.\n",
    "            new_state, reward, is_done, _ = env.step(action) # Perform action.\n",
    "            total_reward += reward # Store reward.\n",
    "            if is_done:\n",
    "                break\n",
    "            state = new_state\n",
    "        return total_reward"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best reward updated 0.000 -> 0.250\n",
      "Best reward updated 0.250 -> 0.350\n",
      "Best reward updated 0.350 -> 0.400\n",
      "Best reward updated 0.400 -> 0.450\n",
      "Best reward updated 0.450 -> 0.500\n",
      "Best reward updated 0.500 -> 0.550\n",
      "Best reward updated 0.550 -> 0.600\n",
      "Best reward updated 0.600 -> 0.650\n",
      "Best reward updated 0.650 -> 0.800\n",
      "Best reward updated 0.800 -> 0.900\n",
      "Solved in 2836 iterations!\n"
     ]
    }
   ],
   "source": [
    "if __name__ == \"__main__\":\n",
    "    test_env = gym.make(ENV_NAME)\n",
    "    agent = Agent()\n",
    "    writer = SummaryWriter(comment=\"-q-learning\")\n",
    "\n",
    "    iter_no = 0\n",
    "    best_reward = 0.0\n",
    "    while True:\n",
    "        iter_no += 1\n",
    "        s, a, r, next_s = agent.sample_env()\n",
    "        agent.value_update(s, a, r, next_s)\n",
    "\n",
    "        reward = 0.0\n",
    "        for _ in range(TEST_EPISODES):\n",
    "            reward += agent.play_episode(test_env)\n",
    "        reward /= TEST_EPISODES\n",
    "        writer.add_scalar(\"reward\", reward, iter_no)\n",
    "        if reward > best_reward:\n",
    "            print(\"Best reward updated %.3f -> %.3f\" % (best_reward, reward))\n",
    "            best_reward = reward\n",
    "        if reward > 0.80:\n",
    "            print(\"Solved in %d iterations!\" % iter_no)\n",
    "            break\n",
    "    writer.close()"
   ]
  },
  {
   "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}