SR-SARSA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Successor Representation - SARSA\n",
    "\n",
    "This notebook contains an implentation of an agent which uses the successor framework to learn Q-values. The agent learns using the SARSA algorithm, and the environment is a simple gridworld with obstacles. This notebook is meant to accompany the following article: [\"The present in terms of the future: Successor representations in Reinforcement Learning\"](https://medium.com/@awjuliani/the-present-in-terms-of-the-future-successor-representations-in-reinforcement-learning-316b78c5fa3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import utils\n",
    "from sklearn.decomposition import PCA\n",
    "from gridworld import SimpleGrid\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "cmap = plt.cm.viridis\n",
    "cmap.set_bad(color='white')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Instantiate the gridworld environment\n",
    "\n",
    "We will use a simple tabular gridworld environment. It has a number of different patterns for the walls. In this case, we will use the `four_rooms` pattern. Note: the agent position is marked in red, the walls are marked in blue, and the goal position is marked in green."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x102ebaa58>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPgAAAD8CAYAAABaQGkdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAACs1JREFUeJzt3UuoXeUZxvHn8RjReqkDUwmJVAsiiAOjh0BRxFqUiKE66EBBB1LIpIpSilihFEediQ5KQaKtpV4QLyBivUADKtToiZdqEi1BLCZYoljRdFBRnw7OChxtyF5xr8vOu/8/CNn7ZO39fmflPGdd9lrf6yQCUNNRYw8AQH8IOFAYAQcKI+BAYQQcKIyAA4URcKAwAg4URsCBwo7u401PsXN6H2/cwvaR6o7v/JHrz+maH2u1vyflo3jSYu7jUtVFO0udv2s7E7/jssa+5HhO1/xYq31RytLkgLOLDhRGwIHCCDhQGAEHCiPgQGEEHCiMgAOFEXCgMAIOFEbAgcJaBdz2Rtvv2N5t+9a+BwWgGxMDbntB0u8kXS7pbEnX2D6774EBmF6bLfgGSbuTvJvkc0kPSbqy32EB6EKbgK+V9P6K53uarwGYcZ2dZLO92faS7aUPu3pTAFNpE/C9kk5b8Xxd87WvSXJ3ksUki6u7Gh2AqbQJ+CuSzrR9hu1jJF0t6Yl+hwWgCxOnbEryhe0bJD0jaUHSvUl29D4yAFNrNSdbkqckPdXzWAB0jCvZgMIIOFAYAQcKI+BAYQQcKIyAA4URcKAwAg4URsCBwgg4UFgv3UVtj93qEigvobsoMNcIOFAYAQcKI+BAYQQcKIyAA4URcKAwAg4URsCBwgg4UBgBBwpr0130Xtv7bL81xIAAdKfNFvyPkjb2PA4APZgY8CTPS/p4gLEA6BjH4EBhrVoXtWF7s6TNXb0fgOm1mvDB9umSnkxyTqs3ZcIHoHdM+ADMuTYfkz0o6W+SzrK9x/bP+h8WgC4wJxtwhGIXHZhzBBwojIADhRFwoDACDhRGwIHCCDhQGAEHCiPgQGEEHCiss9tFMbaxrw6eeNUkRsAWHCiMgAOFEXCgMAIOFEbAgcIIOFAYAQcKI+BAYQQcKIyAA4URcKCwNvOin2Z7q+2dtnfYvmmIgQGY3sR50W2vkbQmyau2T5S0XdJVSXYe4jVj3/kwh8Ze5dxsMrRO5kVP8kGSV5vHn0naJWnt9MMD0LfDOgZvmhCul7Stj8EA6Fbr+8FtnyDpUUk3J/n0IP9O+2BgxrRtH7xK0pOSnklyR4vlxz4gnENjr3KOwYfW5hi8zUk2S7pP0sdJbm5TmICPYexVTsCH1lXAL5T0gqQ3JX3VfPm2JE8d4jVj/7TNobFXOQEfWicB/zYI+BjGXuUEfGi0DwbmHAEHCiPgQGEEHCiMgAOFEXCgMAIOFEbAgcIIOFAYAQcK66l98PmSlvp564nGvGRy7MtFxzTm9z6P/+eLrZZiCw4URsCBwgg4UBgBBwoj4EBhBBwojIADhRFwoDACDhRGwIHCCDhQWJv2wcfaftn2G0374NuHGBiA6bW52eS/ki5Jsr9pYfSi7b8keannsQGY0sSAZ7kzwv7m6armzzzfNgUcMVodg9tesP26pH2SnktC+2DgCNAq4Em+THKupHWSNtg+55vL2N5se8n2kvRh1+ME8C0c1ln0JJ9I2ipp40H+7e4ki0kWpdVdjQ/AFNqcRV9t++Tm8XGSLpX0dt8DAzC9NmfR10i6z/aCln8hPJzkyX6HBaALbc6i/13S+gHGAqBjXMkGFEbAgcIIOFAYAQcKI+BAYQQcKIyAA4URcKAwAg4URsCBwgg4UJiXJ2zp+E1tZnwZ3NirfMwe3fMpycSVzhYcKIyAA4URcKAwAg4URsCBwgg4UBgBBwoj4EBhBBwojIADhbUOeNOf7DXbzIkOHCEOZwt+k6RdfQ0EQPfadhddJ+kKSVv6HQ6ALrXdgt8p6RZJX/U4FgAda9N8cJOkfUm2T1huRftgALNg4v3gtn8r6TpJX0g6VtJJkh5Lcu0hXjP2zclzaOxVzv3gQ2tzP/hhTfhg+2JJv0yyacJyY/+0zaGxVzkBHxoTPgBzjimbyhh7lbMFHxpbcGDOEXCgMAIOFEbAgcIIOFAYAQcKI+BAYQQcKIyAA4URcKAwAg4UdvTYA0BXuBYc/48tOFAYAQcKI+BAYQQcKIyAA4URcKAwAg4URsCBwgg4UBgBBwoj4EBhra5Ft/2epM8kfSnpiySLfQ4KQDcO52aTHyX5qLeRAOgcu+hAYW0DHknP2t5ue/PBFqB9MDB7WvUms702yV7b35P0nKQbkzx/iOXHbpQFlNdZb7Ike5u/90l6XNKG6YYGYAgTA277eNsnHngs6TJJb/U9MADTa3MW/VRJj9s+sPwDSZ7udVQAOkF/cOAIRX9wYM4RcKAwAg4URsCBwgg4UBgBBwoj4EBhBBwojIADhRFwoDDaB5cx9tXBtC+eRWzBgcIIOFAYAQcKI+BAYQQcKIyAA4URcKAwAg4URsCBwgg4UBgBBwprFXDbJ9t+xPbbtnfZ/mHfAwMwvbY3m9wl6ekkP7V9jKTv9DgmAB2Z2PjA9nclvS7pB2nZJYHGB2MYe5VzN9nQump8cIakDyX9wfZrtrc0Pcq+hvbBwOxpswVflPSSpAuSbLN9l6RPk/z6EK8Ze3Myh8Ze5WzBh9bVFnyPpD1JtjXPH5F03jQDAzCMiQFP8i9J79s+q/nSjyXt7HVUADrRqruo7XMlbZF0jKR3JV2f5N+HWH7s/cU5NPYqZxd9aG120WkfXMbYq5yAD432wcCcI+BAYQQcKIyAA4URcKAwAg4URsCBwgg4UBgBBwoj4EBhfbUP/kjSP7/la09pXj+GI7j2VJeKHsHf99zW/n6bhXq5Fn0atpeSLFKb2tSeHrvoQGEEHChsFgN+N7WpTe1uzNwxOIDuzOIWHEBHZirgtjfafsf2btu3Dlj3Xtv7bL81VM0VtU+zvdX2Tts7bN80YO1jbb9s+42m9u1D1V4xhoVmOu4nB677nu03bb8+9FTfQ3YKmplddNsLkv4h6VItz+T6iqRrkvQ+waPtiyTtl/SnJOf0Xe8btddIWpPkVdsnStou6aqBvm9LOj7JfturJL0o6aYkL/Vde8UYfiFpUdJJSTYNWPc9SYtJBv8c3PZ9kl5IsuVAp6Akn/RRa5a24Bsk7U7ybpLPJT0k6cohCid5XtLHQ9Q6SO0PkrzaPP5M0i5JaweqnST7m6ermj+D/ca3vU7SFVqe0HMuNJ2CLpJ0jyQl+byvcEuzFfC1kt5f8XyPBvpBnxW2T5e0XtK2Qy/Zac0F269L2ifpuRXz3w/hTkm3SPpqwJoHRNKztrfb3jxg3VadgroySwGfa7ZPkPSopJuTfDpU3SRfJjlX0jpJG2wPcohie5OkfUm2D1HvIC5Mcp6kyyX9vDlMG8LRWm4c8vsk6yX9R1Jv55tmKeB7JZ224vm65mvlNce/j0q6P8ljY4yh2U3cKmnjQCUvkPST5lj4IUmX2P7zQLWVZG/z9z5Jj2v5EHEIg3YKmqWAvyLpTNtnNCcerpb0xMhj6l1zouseSbuS3DFw7dW2T24eH6flE5xvD1E7ya+SrEtyupb/r/+a5Nohats+vjmhqWb3+DJJg3yCMnSnoL7uJjtsSb6wfYOkZyQtSLo3yY4hatt+UNLFkk6xvUfSb5LcM0RtLW/JrpP0ZnMsLEm3JXlqgNprJN3XfIJxlKSHkwz6cdVITpX0+PLvVh0t6YEkTw9Y/0ZJ9zcbsnclXd9XoZn5mAxA92ZpFx1Axwg4UBgBBwoj4EBhBBwojIADhRFwoDACDhT2Py0b8qaUVvVnAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid_size = 7\n",
    "pattern=\"four_rooms\"\n",
    "env = SimpleGrid(grid_size, block_pattern=pattern, obs_mode=\"index\")\n",
    "env.reset(agent_pos=[0,0], goal_pos=[0, grid_size-1])\n",
    "plt.imshow(env.grid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting up the agent\n",
    "\n",
    "This is the agent class. The main variables in the class are `w`, which is the reward estimate vector, and `M`, which is the successor matrix. Since we are going to be using it to generate Q values rather than V values, it is a three-dimensional tensor, with the first dimension corresponding to the action."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "class TabularSuccessorAgent(object):\n",
    "    def __init__(self, state_size, action_size, learning_rate, gamma):\n",
    "        self.state_size = state_size\n",
    "        self.action_size = action_size\n",
    "        self.M = np.stack([np.identity(state_size) for i in range(action_size)])\n",
    "        self.w = np.zeros([state_size])\n",
    "        self.learning_rate = learning_rate\n",
    "        self.gamma = gamma\n",
    "        \n",
    "    def Q_estimates(self, state, goal=None):\n",
    "        # Generate Q values for all actions.\n",
    "        if goal == None:\n",
    "            goal = self.w\n",
    "        else:\n",
    "            goal = utils.onehot(goal, self.state_size)\n",
    "        return np.matmul(self.M[:,state,:],goal)\n",
    "    \n",
    "    def sample_action(self, state, goal=None, epsilon=0.0):\n",
    "        # Samples action using epsilon-greedy approach\n",
    "        if np.random.uniform(0, 1) < epsilon:\n",
    "            action = np.random.randint(self.action_size)\n",
    "        else:\n",
    "            Qs = self.Q_estimates(state, goal)\n",
    "            action = np.argmax(Qs)\n",
    "        return action\n",
    "    \n",
    "    def update_w(self, current_exp):\n",
    "        # A simple update rule\n",
    "        s_1 = current_exp[2]\n",
    "        r = current_exp[3]\n",
    "        error = r - self.w[s_1]\n",
    "        self.w[s_1] += self.learning_rate * error        \n",
    "        return error\n",
    "    \n",
    "    def update_sr(self, current_exp, next_exp):\n",
    "        # SARSA TD learning rule\n",
    "        s = current_exp[0]\n",
    "        s_a = current_exp[1]\n",
    "        s_1 = current_exp[2]\n",
    "        s_a_1 = next_exp[1]\n",
    "        r = current_exp[3]\n",
    "        d = current_exp[4]\n",
    "        I = utils.onehot(s, env.state_size)\n",
    "        if d:            \n",
    "            td_error = (I + self.gamma * utils.onehot(s_1, env.state_size) - self.M[s_a, s, :])\n",
    "        else:\n",
    "            td_error = (I + self.gamma * self.M[s_a_1, s_1, :] - self.M[s_a, s, :])\n",
    "        self.M[s_a, s, :] += self.learning_rate * td_error\n",
    "        return td_error"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Learning online from experience\n",
    "\n",
    "The learning procedure is online, with the agent updating using pairs of `<state, action, reward, next_state, next_action>`, i.e. SARSA. The agent takes actions using an epsilon-greedy policy.\n",
    "\n",
    "Halfway through learning the goal location is changed, and the agent has to re-learn it's `w` vector."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Episode 950/2000, TD Error: 0.06684845897353445, Test Lengths: 7.828\n",
      "Switched reward locations\n",
      "Episode 1950/2000, TD Error: 0.07828320700868957, Test Lengths: 12.54"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Episode Lengths')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_episode_length = 50\n",
    "test_episode_length = 50\n",
    "episodes = 2000\n",
    "gamma = 0.95\n",
    "lr = 5e-2\n",
    "train_epsilon = 1.0\n",
    "test_epsilon = 0.1\n",
    "\n",
    "agent = TabularSuccessorAgent(env.state_size, env.action_size, lr, gamma)\n",
    "\n",
    "experiences = []\n",
    "test_experiences = []\n",
    "test_lengths = []\n",
    "lifetime_td_errors = []\n",
    "\n",
    "for i in range(episodes):\n",
    "    # Train phase\n",
    "    agent_start = [0,0]\n",
    "    if i < episodes // 2:\n",
    "        goal_pos = [0, grid_size-1]\n",
    "    else:\n",
    "        if i == episodes // 2:\n",
    "            print(\"\\nSwitched reward locations\")\n",
    "        goal_pos = [grid_size-1,grid_size-1]\n",
    "    env.reset(agent_pos=agent_start, goal_pos=goal_pos)\n",
    "    state = env.observation\n",
    "    episodic_error = []\n",
    "    for j in range(train_episode_length):\n",
    "        action = agent.sample_action(state, epsilon=train_epsilon)\n",
    "        reward = env.step(action)\n",
    "        state_next = env.observation\n",
    "        done = env.done\n",
    "        experiences.append([state, action, state_next, reward, done])\n",
    "        state = state_next\n",
    "        if (j > 1):\n",
    "            td_sr = agent.update_sr(experiences[-2], experiences[-1])\n",
    "            td_w = agent.update_w(experiences[-1])\n",
    "            episodic_error.append(np.mean(np.abs(td_sr)))\n",
    "        if env.done:\n",
    "            td_sr = agent.update_sr(experiences[-1], experiences[-1])\n",
    "            episodic_error.append(np.mean(np.abs(td_sr)))\n",
    "            break\n",
    "    lifetime_td_errors.append(np.mean(episodic_error))\n",
    "    \n",
    "    # Test phase\n",
    "    env.reset(agent_pos=agent_start, goal_pos=goal_pos)\n",
    "    state = env.observation\n",
    "    for j in range(test_episode_length):\n",
    "        action = agent.sample_action(state, epsilon=test_epsilon)\n",
    "        reward = env.step(action)\n",
    "        state_next = env.observation\n",
    "        test_experiences.append([state, action, state_next, reward])\n",
    "        state = state_next\n",
    "        if env.done:\n",
    "            break\n",
    "    test_lengths.append(j)\n",
    "    \n",
    "    if i % 50 == 0:\n",
    "        print('\\rEpisode {}/{}, TD Error: {}, Test Lengths: {}'\n",
    "              .format(i, episodes, np.mean(lifetime_td_errors[-50:]), \n",
    "                      np.mean(test_lengths[-50:])), end='')\n",
    "\n",
    "fig = plt.figure(figsize=(10, 6))\n",
    "\n",
    "ax = fig.add_subplot(2, 2, 1)\n",
    "ax.plot(lifetime_td_errors)\n",
    "ax.set_title(\"TD Error\")\n",
    "ax = fig.add_subplot(2, 2, 2)\n",
    "ax.plot(test_lengths)\n",
    "ax.set_title(\"Episode Lengths\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Agent Representation Analysis\n",
    "\n",
    "Once the agent has solved the navigation task and learned `M` and `w`, we can examine these learned representations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def print_occupancy(experiences, grid_size):\n",
    "    occupancy_grid = np.zeros([grid_size, grid_size])\n",
    "    for experience in experiences:\n",
    "        occupancy_grid += env.state_to_grid(experience[0])\n",
    "    occupancy_grid = np.sqrt(occupancy_grid)\n",
    "    occupancy_grid = utils.mask_grid(occupancy_grid, env.blocks)\n",
    "    plt.imshow(occupancy_grid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize state occupancy\n",
    "\n",
    "First, we can take a look at what states the agent most frequently visited during training and testing."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### During training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_occupancy(experiences, grid_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### During evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print_occupancy(test_experiences, grid_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Raw successor matrix\n",
    "\n",
    "We can visualize the raw successor matrix. It may not be very interpretable though. Fortunately there are other more useful methods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x10a2819e8>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAHZVJREFUeJztnXmMXVd9x7+/92Yfz3hm7DhxPG6cjQTTlgRCFkJFGkihYYklUMtSFKRIoYVKodBCKBIqEpWSPwqk7GERZk2AQBJFVCGkYRPFsbOY2DFe4sTOODMeLzMee/aZ++sf8xzPOb8zfnfe3HfnTc73I438zvG59/zufff3zju/91tEVUEIiYvCYgtACMkfKj4hEULFJyRCqPiERAgVn5AIoeITEiFUfEIihIpPSIQsSPFF5M0islNE9ojIrVkJRQipLlKp556IFAHsAnAdgB4AmwG8W1WfnuuYlV1FXbe2/sX2UwNnmDHFMbddd2g4NLfTVgSuwesaP6fFDCmMuecpTviyTNvzTky60yRJYG538slVrXbuKbctib2GwuCIO8a7biBw7WnezpYmtz0yFh6XAVLw1paCdw2B6w7e04omt/fLUsH9q4SQKFWYawzDmNDxshdet4A5LgewR1X3AoCI3AXgBgBzKv66tfV49MG1L7bP+8kHzJjlO4tOe9WXfm/GFJrcB1enpswYnXaVdtenXm3GtD3d4LTb97vHtO05bo6Rfb1OOzlhP5h0yv1w6HvPVWZMS7/7cNeP2oe9+d5HnbZ/3TNzudfuX7f/IQQAsv4V7pDHtpsxqfAVKzBXocX90JOmRveQUfuhk4yMmL5KkHr3/TUfOgDg3a809y8VBfc5lsDc6n/oJYGFZp5s0odTjVvIV/01AJ6f1e4p9RFCapyqG/dE5GYR2SIiWw4dWfgnGiFk4Szkq/4BAGtntbtLfQ6qeieAOwGg8Zy1et7PTn29f9/rf2dO+p2Gq532qtDMRfdrFKZDe0L3Q6b9qQYzYnKZ2x5d4X4ONvfZY+rr3FsmRfvZqdPeeQ5b+cTrmmy252n2O/z9MgCI1yf+njXwYZvRFjoV5b4qZ7Wfr3HM1/qZzvwFKbGQFX8zgAtF5FwRaQDwLgD3ZyMWIaSaVLziq+qUiPwzgAcBFAF8S1UrtBIRQvJkIV/1oao/B/DzjGQhhOQEPfcIiZAFrfjzpTgKdGw79Vnz3enXmTHPvvNrTvtN/3iJGSOecU99bxjAGL1a+6whZaLN/W11ZLXbHg8Y9xoa3T4dKZox4snTeMzOrd5hE23lP4Olzr5d/u/Oop5zU8B+JJPeb9dlZ55LIE/moCHRE8D/nTxPQga2SOGKT0iEUPEJiRAqPiERUnGQTiW0S5deIW+Y1zHD77jC9LXesykrkQh5SbFJH8aQHi0bpMMVn5AIoeITEiFUfEIihIpPSITk6sADESc5QiiyzY+8CxnyLtzsJnPY/IVXmTFdTw44bd2x14rjzS+tXpaeVSvMMYOvdPtaDk6YMQ09g047ea7HjPGTQoScc5JhN8mHvOYv7GmOeVl6hked9tSBFwJze95DFSaAKHZ2Ou3pgYE5RlaBjK4hVrjiExIhVHxCIoSKT0iE5LvHhzpRI6GYjjTJDf/wVXdPP/gmm7Cx4cRyp936dCBQxp/fy6BbODEKn2X73b7xFY1mTF27lzsnFCnj53S0IwyFISuPoT7FW7qImV9IbcAVn5AIoeITEiFUfEIihIpPSITka9xTL2OMnwoaNoNMiJWPDznt+tE2Mya5ud+d+mcBS6I//4TrjKOBKjn1vV6FlGS5GTPS7VaPaX48NLfnPORb+wLIMVvZB/X1TlPbbKkwQ0YRmaEKRrlBA+WC4IpPSIRQ8QmJECo+IRGSawaexnO69axP3vJie/l2a2JoOeTu3dru+oMZI42e00wgc6ufffavtlrnlx/86FqnvWK7e8yyva4tAQBknxv0kgzb8/rVcnd/8XIzZvkO11YQKrPlX7u5bgA66e2z/b1v4P1NXn+p0y78+gkzJhUpquUW29vdQ1pc56Zk2FbGTY4HbBkVYKoLB0qQ+ffPVMutMPgnFHRl5q5StVxm4CGEBKHiExIhVHxCIoSKT0iE5FtCa0zQse3UlBPW7wajXm35wBBjOAnWHvccYu76wbVmRN2VbsaY40NuRpnGQVOhHg19roFNxsbNGPWuYfl2W2Yr8apzja60n8H+tQcNRt61m4jDYAgkiR2u+IRECBWfkAih4hMSIfnu8ccV7ftOOUyMdtm979B55YN0ZJkbBINAMI3vyNK1wwaUnBjscNpjb3QddgYDwT+r+tygHPGy9gCAeI4s7c/bucfb3c/cY+enKJPdEgrA8RxgvKS/GnAKKQ67MlfqwmXKlQeCdoxDzKR3v6pYNlunyzszWYenbIJ/fLuTFMo/13nCFZ+QCKHiExIhZRVfRL4lIv0ism1WX5eIPCQiu0v/dp7uHISQ2iLNHv/bAL4I4Duz+m4F8LCq3iYit5baHy93osL4NFr3Hnux3dxbb8Ys620yfT5T56122vUvHDVjkqNuNZvWZ2zATdMh9zf5o1Punn79TdvNMbtPrHfanVvt3q3Qf8Sde9Y1n6S5yb31TQPlE2gk3atMX/Gwe24/ecj0gK30U+x171el6TSk2fVz0EBwjY57fg7enr7iZB4pAoSMfSGY1TmFHaASvPNWmlG6WpRd8VX1NwB8zboBwMbS640ANmQsFyGkilS6xz9TVXtLr/sAnJmRPISQHFiwcU9nAvrn/M4iIjeLyBYR2TIxbWOvCSH5U6niHxSR1QBQ+rd/roGqeqeqXqaqlzUUUySCJIRUnUodeO4HcCOA20r/3pfqqMlJ4PneF5uFQLmn5h7X4Bcy/Qyd7xqVlqv9UaHOc6zR/bZcdF2DO9eqQ65zzt7Bl5tj+l/nthsHbZbdFt+pZ9Y1n6TgOXS0vmADgvxrH3qZdShqa3SdaOoOek5RgdLV04ePmL5KKHiOVKHMOb6BzRjcKjRwpXEeWtTS2YtouEtDmp/zfgjg/wBcJCI9InITZhT+OhHZDeCNpTYhZIlQdsVX1XfP8V9vyFgWQkhO0HOPkAjJNUhHE0UyOquk9bh1fvH3biHan3XLYo932uyzxePu3js5bJ18xKucI94+sX2rnbtxwLUn7L+uwYzp1pXuMfsO2BMZWWywj0/bszYYabLdnV8LbuCRPLvPHGMy81ZKikyymex1fWcdwFQiIvODd4+QCKHiExIhVHxCIoSKT0iE5FwmW10HjpB/RYqMLHUH3Yi0wsQyM2a023V2adhRvsyWnzmn4EX4AUCjN2bNr1eaMc9tcI1RL3swYLjzjVMpMr/UHbYOMoUJ1xtycL1bsqr994ETZeTYooEMw1UhTeQdmRdc8QmJECo+IRFCxSckQnItk92yaq1e+PcfebEdKg3dcNzduzX+fLMZ45eLDjqkePvY3V+4wgzp3OZV7elxz9PynM3a4wfcOA5JJ+XxymR/+bnfmjFv2fhvTrtjl30fln+vOmWyRze4Zbub733UjElFmjLZK1e4h7S6Ngk/YxAATB+xzlaVUGj1sjEHymT7WX8T35Gq0jLZ9Z5jVyjLblbZiGbBMtmEkDmh4hMSIVR8QiKEik9IhOTqwFOYBJa9cHpjiRZTlNDyjUoBw4mq29f1pP2Mm250x4wvdyMDm1ps5J3JGhSIMPSdc972jY+ZIcVLXCekoQmbycfvMdcNmGtPUxX7xFnuddrcP9kh9W6WI23yyoyP2/Tf2c3tvVeBiD5NEm/IlPf/Fc5d9OYKGBaNKTQD415auOITEiFUfEIihIpPSITkuseXRFE3emrTNN1k96zjy93PolBBLWnwss6EnJC8MsXNA3azNtHqznVijdtuHLCzt77g9knAgcd3zGh7zsp3YsrdwX/+pq+bMf/1mVe4czVYm4O5du+6Q5v+9v0Z7SVNoFEgEMoL5BEva4+OBe5fRqjvjBOwkfhOM35564rn9kt0++2ZyTKZqxK44hMSIVR8QiKEik9IhFDxCYmQXKPz2qVLrxDW4UjL8DvciMLWezYtkiRkqcDoPELInFDxCYkQKj4hEZJvll0RJzOJCWQAAK+EVjJsM7TIq13HluKAHaNH3fLQyfCoPY83vzR74Sqr3OwxAHB8vdvXfNA6oNT3utl5p3tsmWw/uCZUOszf07f91mb07fnaBU678yk3+CfZuiMwtzdXhVlmCn9+sXuabX+q6DxlCTjeFLxsRMl4IONvtexXKTIP1Tpc8QmJECo+IRFCxSckQqj4hETI4pbQCkUnhaKYPIqHvbTXAcOYX7s9WHLJn3/UNQDKUddQBgBtu9y5Js5oNWOSVtdIGJzb70pREqrvjvNN36HXuu2GE27psOatgRNlFBU2cq5brqtpWyanNUhdve3zDbF+JB6QLh1R2cnLZ1jKZJ6c4YpPSIRQ8QmJkLKKLyJrReQREXlaRLaLyC2l/i4ReUhEdpf+7ay+uISQLEizx58C8FFVfVxE2gA8JiIPAXg/gIdV9TYRuRXArQA+XvZss/aXmoQ+d8rvl0xWl0BmGlNuKrCvNfN79gWZtBlgxXMEKrbauYcvcPe+zdvL76nD98Kl5YB1Fjpzk7vXLXyw3x1wb2iybBxOWv902GlXa6frlyQDgGRkxOuo0uzBe7V4mXOyouzTpqq9qvp46fVxADsArAFwA4CNpWEbAWyolpCEkGyZl1VfRNYBuBTAJgBnqupJX9Q+AGfOcczNAG4GgCa0hIYQQnImtXFPRJYBuAfAh1XV+T1NZ4L6g98fVfVOVb1MVS+rh632SgjJn1SJOESkHsADAB5U1c+W+nYCuEZVe0VkNYBfqepFpzsPE3Hkj5/MA2BCj5cymSXikJm6Td8EsOOk0pe4H8CNpdc3ArivEkEJIfmTZo9/NYD3AXhKRJ4s9f07gNsA/EhEbgKwD8DfVUdEQkjWlFV8Vf0dgLm+OvB7OyFLEHruERIh+ZbQam5C4aKXn37MlOscMb19px2TKgDHNVrq1ZfYIXXuF5niCddhp9hvg3SSQ67TShIMDnGvYewtrzFDjq91g33a99myVo0/3+y0/esGAtfuXXfIkHfRFjfoZedlgWvIiOR17n0fOdstQdb2bCB70uanMpm7+DIvqCmQ8UmOu45AyYCbPck4CgGpHKDqzvJ+3Q6V75pwn7fpI0fLnjcruOITEiFUfEIihIpPSISwkg6hk89LCFbSIYTMCRWfkAih4hMSIVR8QiIk3yy7pCYJGfJYovulDVd8QiKEik9IhFDxCYkQ7vFJEH9PTyeflxZc8QmJECo+IRFCxSckQqj4hETI4hr3QiWIfQLRg4WmJq/Dfn752U10yma4MfN75Y+l3t6eQrM3d70toYVxt8TX9NCQHVNOFsBce/GCc82QibVuycLG3Qed9lTPgfJzpyBkyOu9182mtHrDjkzm8jHZbABMnnuW0y5u22vGJMePz38y730Ilmfz+tQrrw7M8bxVQplnFEBF5cO44hMSIVR8QiKEik9IhDADD8kMOvksPszAQwiZEyo+IRFCxSckQqj4hERIviW0CgUUmltOPyhxy08lY2OZzF1obQ10ep97Xjkq3wkIKF+yKkSxY7npE88JSUesE0gqx58K0Ne+0pXl91szOW/IkHf+Zvc6H/vCpU67/Vn7/hZ++0Qm8hz+wFVOu2OPfT8be7wyaf1HnOb0scB7kMJhZuraVzvthqP2/S0MnnCPeW5/2fNmBVd8QiKEik9IhFDxCYmQfIN0RCCNjacf45WYRmiP7wcupNhnyzK7xxfvPDpZvly0jo17HeX3e9LWZs/T5to6pFg0Y+Dv8VME8qThxNpmp22ly47NX3b39Ecv8eRVL+gJQMdvs5m7Y7dXhrrJrnNJmzt/YdAtIR56XzTFHr/hsFf+OxRIVhd4z3OCKz4hEULFJyRCyiq+iDSJyKMislVEtovIp0v954rIJhHZIyJ3i0ggMJ0QUouk2eOPA7hWVU+ISD2A34nI/wD4CIDPqepdIvJVADcB+MrpTqRJgmRkZMFCV7Kv1eMnbJ+/Z/Z/xw8lU6gg6UEyeMz0iZe8IRm3vzEbMgqoWr7TtR0kc4zLgs5d7nXWjbt76kv/xf5m/8x3vY4KbRuNB933fLKz2YwZOn+Z0+4Yde08csLbqyPgyxF4JgrH3OO0xdoytKmMvQsACp4doILnL3jacgN0hpN3sL70pwCuBfCTUv9GABsykYgQUnVS7fFFpCgiTwLoB/AQgGcADKrqySWxB8Ca6ohICMmaVIqvqtOqegmAbgCXA7g47QQicrOIbBGRLZOajfstIWRhzMuqr6qDAB4BcBWADhE5aSPoBhDM6qiqd6rqZap6Wb3YfQ4hJH/KZuARkTMATKrqoIg0A/gFgNsB3AjgnlnGvT+q6pdPdy5m4CFpYInuykmbgSeNVX81gI0iUsTMN4QfqeoDIvI0gLtE5DMAngDwzQVJTAjJjbKKr6p/BHBpoH8vZvb7hJAlBj33CIkQlsle6vgOHn6QUxWzKBc73So+0wMD8z9JwDnH39Pv/pLN3nvRVzynqH3Wtpyqko53/4qdbtIUXetW7AEA6XGrFSXH7Dw6mcIhKwV+Ahn1qjQBlVXt4YpPSIRQ8QmJECo+IRFCxSckQmjcmy9pyhZXycAmgZLcftnuZNR1i87KyBQiOd8Lz9hSgXEvxb256OvWeNZ3TZfTXv2LQPakNMY9L9ot8bLqFhIrn3gGwMK0jW+cHsjmvhtjXihT0+xowZSPGld8QiKEik9IhFDxCYkQ7vHni78nTZFlN7OpA/v16Slvb5tj2XPZ8aw7dbUm2r3PdK0edm0Zz7z/TDNm3SefmfdUfnadJJCBp1D01suzVtoTVeLMlEKeUPSN1M3KDDxZNj4HAFd8QqKEik9IhFDxCYkQKj4hEVI2A0+WtBe69Mq6N512jPoOExmlE5a6gB3Td77xHG9MGmWgIuNZqGyYL49OWAeUajnf1HW7jjdTPcGsaZlw6J/cUtUrt7rp1esGA+XBn96VydwHfvoKp935vWVmTNvOQbej75Ary4BNjZ7mmez96Gud9sqtNqquqddN/z29fWfZ85YjbQYerviERAgVn5AIoeITEiE5O/BIOKhl9oiCt8/Oqr5TKLjBJ/Fkk8B+vgKHHb8cNxAov1y059XyVbsrQpvyK3PYucu1U0y2uWWoQxl4TE+FJbTa73YLgPddYc/TcMzd9zcNuQ47wRJaE+WDsDr2uFlxJtutqtUNuwFWwY15BSXh08AVn5AIoeITEiFUfEIihIpPSITka9xTDTvF5EFo3jIOPFlZFjWQoQVeSuTgmCoho9aZpFo09rgOMEm7W6P++LoWc4xxswkahFMY2P54xGk3DSw3Y/Zf5xo6/0zdKL/GUetg5KfTDjlaLdvhzj11RpsZM7bKdexqNiO8yDsA6kdjAhUZ/LjiExIhVHxCIoSKT0iE5J+BJ6Ogm/kSti14fVUKWArty4w8mXkqlSc5cjS3ubSn12kXGtw9dccRu+/2C0L5mYQBm3029P7q/hecdvPAkBlzzoQbsNTzQfe9+rM71ppj6nY+77STQRvI41933RGbkaetxbVvhAphFTrc+xOcq4JgLq74hEQIFZ+QCKHiExIhVHxCIiSe9No5ZhpKNXeOabl9krGx8oPS4NWWDxlukxE34w68dij6zUfOWWM7n3eNZwg52gy75/bLiwFA/XE3C86629c57b6rrePN2f2uwU0C5/XnDsmHUHYfDz3bTd1dCKVYH5p131M+VlzxCYmQ1IovIkUReUJEHii1zxWRTSKyR0TuFpH8grwJIQtiPiv+LQB2zGrfDuBzqnoBgAEAN2UpGCGkeqTa44tIN4C3APhPAB+RmZQy1wJ4T2nIRgD/AeArVZCRnI4U++yqTd3qOqAkobLUZWwrqZxPevrs3Cs63fMcrzdjpn1HpZANwtuLF7a7ZbfWHFlljun7G9fmcNYvrfNVstfb41do55Fn3QzI0tVhxhTGTjkzyVi2JbQ+D+BjOBUStQLAoKqedDbqARCwwBBCapGyii8ibwXQr6qPVTKBiNwsIltEZMsk8gsHJYTMTZqv+lcDeLuIXA+gCUA7gDsAdIhIXWnV7wYQrMqgqncCuBMA2qVrEX9TI4ScpKziq+onAHwCAETkGgD/qqrvFZEfA3gngLsA3AjgvirKSeZikYKeACAZHik/KIt5fF8ABL6qrrZ7cVQQjJT4wT8HrH3hrAfd9atngw3kOeuO5+Y9d4hpz24i4/Zbc/GMWb/196VzzVnI7/gfx4yhbw9m9vzfXMC5CCE5Mi/PPVX9FYBflV7vBXB59iIRQqoNPfcIiRAqPiERkm+ZbOnSKwpvnN9BWcnnO7oE5yqfubUSgiW6vRJaOhnIv1Ilw11xRZfTNo4uGTL9169y2g19nrFqxAa4TO173vRVQv+H3FLVK562c/nyoLffaU4PuUE8AFK9Ly9/zH3Pf/fV15gxnbtceQq/fqLsecvBMtmEkDmh4hMSIVR8QiIk/0QcZcpk55ltVgruVkj9Mtl+tRagagk9fFlm5KnKVOGy01Wi/pDnfFNw77G22Ay6WdH1J9fZZbzTBvLIZKvTbhjwSlcP2wQammKP/5tvuHv6Y6+39oXiuDuXDb+pHlzxCYkQKj4hEULFJyRCqPiEREj+xr0cjXfl0ET9Dq9dRecmb24jSzXJsSR34YRr3NN695FLloWKQ3uEjJEp3puGg152nQk7l1+quuEFd4yESlZNefIEZFmxzXPOmbBGzOZ/8DIFf8cMsdee0TPJFZ+QCKHiExIhVHxCIiT/IB15Q27zEbKUGH7HFaav9Z5N8zoHg3QIIXNCxSckQqj4hEQIFZ+QCImnTDZZshRaW02fKXldabYiz0Gm0OKWBZPu1faYg4ecZqVZenxChrw9n73SaV/0tcN2qr37TzUmsy2hRQh5CUHFJyRCqPiERAj3+KTmCZXQEj9LMQJZlNPssz0HNn+uwvMvmEMKZ6xw2sVAgNX00FD5uVNw8X+78x94W7cZs+Ynp2wMcihFNmlwxSckSqj4hEQIFZ+QCKHiExIhNO6R2icQQapTXsmxNCXSKpgrZFhM9rspt4srV5oxWeGXEzv7B9ZZqOd9F7/4euL7jeb/Q3DFJyRCqPiERAgVn5AIyTUDj4gcArAPwEoANtqgdllq8gJLT+alJi9QmzKfo6pnlBuUq+K/OKnIFlW9LPeJK2SpyQssPZmXmrzA0pT5JPyqT0iEUPEJiZDFUvw7F2neSllq8gJLT+alJi+wNGUGsEh7fELI4sKv+oRESO6KLyJvFpGdIrJHRG7Ne/5yiMi3RKRfRLbN6usSkYdEZHfp387FlHE2IrJWRB4RkadFZLuI3FLqr2WZm0TkURHZWpL506X+c0VkU+nZuFtEGhZb1tmISFFEnhCRB0rtmpb3dOSq+CJSBPAlAH8LYD2Ad4vI+jxlSMG3AbzZ67sVwMOqeiGAh0vtWmEKwEdVdT2AKwF8qHRPa1nmcQDXquorAVwC4M0iciWA2wF8TlUvADAA4KZFlDHELQB2zGrXurxzkveKfzmAPaq6V1UnANwF4IacZTgtqvobAEe97hsAbCy93ghgQ65CnQZV7VXVx0uvj2PmwVyD2pZZVfVktEl96U8BXAvgJ6X+mpJZRLoBvAXAN0ptQQ3LW468FX8NgNnhRj2lvlrnTFU9Wcy8D8CZiynMXIjIOgCXAtiEGpe59LX5SQD9AB4C8AyAQVU9GXZXa8/G5wF8DEBSaq9Abct7Wmjcmyc68zNIzf0UIiLLANwD4MOq6iR8q0WZVXVaVS8B0I2Zb4IXlzlk0RCRtwLoV9XHFluWrMg7Hv8AgLWz2t2lvlrnoIisVtVeEVmNmVWqZhCReswo/fdV9ael7pqW+SSqOigijwC4CkCHiNSVVtFaejauBvB2EbkeQBOAdgB3oHblLUveK/5mABeWrKENAN4F4P6cZaiE+wHcWHp9I4D7FlEWh9Je85sAdqjqZ2f9Vy3LfIaIdJReNwO4DjO2iUcAvLM0rGZkVtVPqGq3qq7DzDP7v6r6XtSovKlQ1Vz/AFwPYBdm9nSfzHv+FPL9EEAvgEnM7Ntuwsx+7mEAuwH8EkDXYss5S97XYeZr/B8BPFn6u77GZf5LAE+UZN4G4FOl/vMAPApgD4AfA2hcbFkDsl8D4IGlIu9cf/TcIyRCaNwjJEKo+IRECBWfkAih4hMSIVR8QiKEik9IhFDxCYkQKj4hEfL/M95XVMpFlhsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "averaged_M = np.mean(agent.M, axis=0)\n",
    "plt.imshow(averaged_M)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analyze SR for each state\n",
    "\n",
    "Instead of looking at the full `M` matrix, we can plot the successor representation for each state onto the environment itself."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1512x1512 with 40 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "averaged_M = np.reshape(averaged_M, [env.state_size, grid_size, grid_size])\n",
    "\n",
    "cmap = plt.cm.viridis\n",
    "cmap.set_bad(color='white')\n",
    "\n",
    "plt.figure(1, figsize=(grid_size*3, grid_size*3))\n",
    "for i in range(env.state_size):\n",
    "    if env.state_to_point(i) not in env.blocks:\n",
    "        ax = plt.subplot(grid_size, grid_size, i + 1)\n",
    "        ax.imshow(utils.mask_grid(averaged_M[i,:,:], env.blocks), cmap=cmap)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize bottleneck states\n",
    "\n",
    "It is also possible to analyze the learned successor representation for all of the states via dimensionality reduction. This can tell us how the different states relate to one another. In the case of the `four_rooms` environment, we find that states within each room are all most similar to other states in that room. We also find that the four door states between rooms are represented uniquely. Because of this, they are potential candidates for bottleneck states that might be used in a hierarchical agent."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x10b643da0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "M_s = np.mean(agent.M[:,:,:], axis=0)\n",
    "colors = np.zeros([env.state_size])\n",
    "for bottleneck in env.bottlenecks:\n",
    "    grid = np.zeros([env.grid_size,env.grid_size])\n",
    "    grid[bottleneck[0],bottleneck[1]] = 1\n",
    "    grid = grid.flatten()\n",
    "    b_state = np.where(grid==1)[0][0]\n",
    "    colors[b_state] = 1\n",
    "pca = PCA(n_components=2)\n",
    "pca_result = pca.fit_transform(M_s[:])\n",
    "\n",
    "plt.scatter(pca_result[:,0], pca_result[:,1], c=colors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Value estimates\n",
    "\n",
    "We can also examine `w` by using it to generate Q values, and plotting them onto the environment itself. We find that a gradient of value flows from the rewarding state."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x10cc13630>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = np.zeros([env.state_size])\n",
    "for i in range(env.state_size):\n",
    "    Qs = agent.Q_estimates(i)\n",
    "    V = np.mean(Qs)\n",
    "    a[i] = V\n",
    "V_Map = np.reshape(a, [grid_size, grid_size])\n",
    "V_Map = np.sqrt(V_Map)\n",
    "\n",
    "V_Map = utils.mask_grid(V_Map, env.blocks)\n",
    "\n",
    "plt.imshow(V_Map)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Optional - Analyze SR for each state-action pair\n",
    "\n",
    "Lastly, we can examine the successor representation for each state/action pair. This shows us the way in which the successor map for each action is slightly biased in the direction of that action."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_srs(action, M):\n",
    "    M = np.reshape(M, [env.action_size, env.state_size, grid_size, grid_size])\n",
    "    M = np.sqrt(M)\n",
    "    plt.figure(1, figsize=(grid_size*3, grid_size*3))\n",
    "    for i in range(grid_size*grid_size):\n",
    "        if env.state_to_point(i) not in env.blocks:\n",
    "            ax = plt.subplot(grid_size, grid_size, i + 1)\n",
    "            ax.imshow(utils.mask_grid(M[action,i,:,:], env.blocks))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Up action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1512x1512 with 40 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_srs(0, agent.M)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Down action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1512x1512 with 40 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_srs(1, agent.M)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Left action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1512x1512 with 40 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_srs(2, agent.M)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Right action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1512x1512 with 40 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_srs(3, agent.M)"
   ]
  }
 ],
 "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": 4
}