Lab2-MDPs.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Lab 1: Markov Decision Processes - Dynamic Programming\n",
    "\n",
    "## Lab Instructions\n",
    "To complete the homework, you only need to modify the code in this notebook. \n",
    "\n",
    "**You should execute every block of code to not miss any dependency.**\n",
    "\n",
    "*This exercise is adapted from Berkeley Deep RL Class [HW2](https://github.com/berkeleydeeprlcourse/homework/blob/c1027d83cd542e67ebed982d44666e0d22a00141/hw2/HW2.ipynb) [(license)](https://github.com/berkeleydeeprlcourse/homework/blob/master/LICENSE)*\n",
    "\n",
    "--------------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n",
    "In this lab, we will solve **Markov Decision Processes (MDPs) with finite state and action spaces** with several classic algorithms that you learnt in the class.\n",
    "\n",
    "The experiments here will use the Frozen Lake environment, a simple gridworld MDP that is taken from `gym` and slightly modified for this assignment. In this MDP, the agent must navigate from the start state to the goal state on a 4x4 grid, with stochastic transitions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    Winter is here. You and your friends were tossing around a frisbee at the park\n",
      "    when you made a wild throw that left the frisbee out in the middle of the lake.\n",
      "    The water is mostly frozen, but there are a few holes where the ice has melted.\n",
      "    If you step into one of those holes, you'll fall into the freezing water.\n",
      "    At this time, there's an international frisbee shortage, so it's absolutely imperative that\n",
      "    you navigate across the lake and retrieve the disc.\n",
      "    However, the ice is slippery, so you won't always move in the direction you intend.\n",
      "    The surface is described using a grid like the following\n",
      "\n",
      "        SFFF\n",
      "        FHFH\n",
      "        FFFH\n",
      "        HFFG\n",
      "\n",
      "    S : starting point, safe\n",
      "    F : frozen surface, safe\n",
      "    H : hole, fall to your doom\n",
      "    G : goal, where the frisbee is located\n",
      "\n",
      "    The episode ends when you reach the goal or fall in a hole.\n",
      "    You receive a reward of 1 if you reach the goal, and zero otherwise.\n",
      "\n",
      "    \n"
     ]
    }
   ],
   "source": [
    "from misc import FrozenLakeEnv, make_grader\n",
    "env = FrozenLakeEnv()\n",
    "print(env.__doc__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's look at what a random episode looks like."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\u001b[41mS\u001b[0mFFF\n",
      "FHFH\n",
      "FFFH\n",
      "HFFG\n",
      "  (Down)\n",
      "S\u001b[41mF\u001b[0mFF\n",
      "FHFH\n",
      "FFFH\n",
      "HFFG\n",
      "  (Down)\n",
      "SFFF\n",
      "F\u001b[41mH\u001b[0mFH\n",
      "FFFH\n",
      "HFFG\n"
     ]
    }
   ],
   "source": [
    "# Some basic imports and setup\n",
    "import numpy as np, numpy.random as nr, gym\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "np.set_printoptions(precision=3)\n",
    "\n",
    "# Seed RNGs so you get the same printouts as me\n",
    "env.seed(0); from gym.spaces import prng; prng.seed(10)\n",
    "# Generate the episode\n",
    "env.reset()\n",
    "for t in range(100):\n",
    "    env.render()\n",
    "    a = env.action_space.sample()\n",
    "    ob, rew, done, _ = env.step(a)\n",
    "    if done:\n",
    "        break\n",
    "assert done\n",
    "env.render();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the episode above, the agent falls into a hole after two timesteps. Also note the stochasticity--on the first step, the DOWN action is selected, but the agent moves to the right.\n",
    "\n",
    "We extract the relevant information from the gym Env into the MDP class below.\n",
    "The ```env``` object won't be used any further, we'll just use the ```mdp``` object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mdp.P is a two-level dict where the first key is the state and the second key is the action.\n",
      "The 2D grid cells are associated with indices [0, 1, 2, ..., 15] from left to right and top to down, as in\n",
      "[[ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]\n",
      " [12 13 14 15]]\n",
      "Action indices [0, 1, 2, 3] correspond to West, South, East and North.\n",
      "mdp.P[state][action] is a list of tuples (probability, nextstate, reward).\n",
      "\n",
      "For example, state 0 is the initial state, and the transition information for s=0, a=0 is \n",
      "P[0][0] = [(0.1, 0, 0.0), (0.8, 0, 0.0), (0.1, 4, 0.0)] \n",
      "\n",
      "As another example, state 5 corresponds to a hole in the ice, in which all actions lead to the same state with probability 1 and reward 0.\n",
      "P[5][0] = [(1.0, 5, 0)]\n",
      "P[5][1] = [(1.0, 5, 0)]\n",
      "P[5][2] = [(1.0, 5, 0)]\n",
      "P[5][3] = [(1.0, 5, 0)]\n"
     ]
    }
   ],
   "source": [
    "class MDP(object):\n",
    "    def __init__(self, P, nS, nA, desc=None):\n",
    "        self.P = P # state transition and reward probabilities, explained below\n",
    "        self.nS = nS # number of states\n",
    "        self.nA = nA # number of actions\n",
    "        self.desc = desc # 2D array specifying what each grid cell means (used for plotting)\n",
    "mdp = MDP( {s : {a : [tup[:3] for tup in tups] for (a, tups) in a2d.items()} for (s, a2d) in env.P.items()}, env.nS, env.nA, env.desc)\n",
    "\n",
    "\n",
    "print(\"mdp.P is a two-level dict where the first key is the state and the second key is the action.\")\n",
    "print(\"The 2D grid cells are associated with indices [0, 1, 2, ..., 15] from left to right and top to down, as in\")\n",
    "print(np.arange(16).reshape(4,4))\n",
    "print(\"Action indices [0, 1, 2, 3] correspond to West, South, East and North.\")\n",
    "print(\"mdp.P[state][action] is a list of tuples (probability, nextstate, reward).\\n\")\n",
    "print(\"For example, state 0 is the initial state, and the transition information for s=0, a=0 is \\nP[0][0] =\", mdp.P[0][0], \"\\n\")\n",
    "print(\"As another example, state 5 corresponds to a hole in the ice, in which all actions lead to the same state with probability 1 and reward 0.\")\n",
    "for i in range(4):\n",
    "    print(\"P[5][%i] =\" % i, mdp.P[5][i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem 1: implement value iteration\n",
    "In this problem, you'll implement value iteration, which has the following pseudocode:\n",
    "\n",
    "---\n",
    "Initialize $V^{(0)}(s)=0$, for all $s$\n",
    "\n",
    "For $i=0, 1, 2, \\dots$\n",
    "- $V^{(i+1)}(s) = \\max_a \\sum_{s'} P(s,a,s') [ R(s,a,s') + \\gamma V^{(i)}(s')]$, for all $s$\n",
    "\n",
    "---\n",
    "\n",
    "We additionally define the sequence of greedy policies $\\pi^{(0)}, \\pi^{(1)}, \\dots, \\pi^{(n-1)}$, where\n",
    "$$\\pi^{(i)}(s) = \\arg \\max_a \\sum_{s'} P(s,a,s') [ R(s,a,s') + \\gamma V^{(i)}(s')]$$\n",
    "\n",
    "Your code will return two lists: $[V^{(0)}, V^{(1)}, \\dots, V^{(n)}]$ and $[\\pi^{(0)}, \\pi^{(1)}, \\dots, \\pi^{(n-1)}]$\n",
    "\n",
    "To ensure that you get the same policies as the reference solution, **choose the lower-index action to break ties in $\\arg \\max_a$. This is done automatically by np.argmax. This will only affect the \"# chg actions\" printout below--it won't affect the values computed.**\n",
    "\n",
    "<div class=\"alert alert-warning\">\n",
    "Warning: make a copy of your value function each iteration and use that copy for the update--don't update your value function in place. \n",
    "Updating in-place is also a valid algorithm, sometimes called Gauss-Seidel value iteration or asynchronous value iteration, but it will cause you to get different results than our reference solution (which in turn will mean that our testing code won’t be able to help in verifying your code).\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration | max|V-Vprev| | # chg actions | V[0]\n",
      "----------+--------------+---------------+---------\n",
      "   0      | 0.80000      |  N/A          | 0.000\n",
      "   1      | 0.60800      |    2          | 0.000\n",
      "   2      | 0.51984      |    2          | 0.000\n",
      "   3      | 0.39508      |    2          | 0.000\n",
      "   4      | 0.30026      |    2          | 0.000\n",
      "   5      | 0.25355      |    1          | 0.254\n",
      "   6      | 0.10478      |    0          | 0.345\n",
      "   7      | 0.09657      |    0          | 0.442\n",
      "   8      | 0.03656      |    0          | 0.478\n",
      "   9      | 0.02772      |    0          | 0.506\n",
      "  10      | 0.01111      |    0          | 0.517\n",
      "  11      | 0.00735      |    0          | 0.524\n",
      "  12      | 0.00310      |    0          | 0.527\n",
      "  13      | 0.00190      |    0          | 0.529\n",
      "  14      | 0.00083      |    0          | 0.530\n",
      "  15      | 0.00049      |    0          | 0.531\n",
      "  16      | 0.00022      |    0          | 0.531\n",
      "  17      | 0.00013      |    0          | 0.531\n",
      "  18      | 0.00006      |    0          | 0.531\n",
      "  19      | 0.00003      |    0          | 0.531\n",
      "Test succeeded\n"
     ]
    }
   ],
   "source": [
    "def value_iteration(mdp, gamma, nIt, grade_print=print):\n",
    "    \"\"\"\n",
    "    Inputs:\n",
    "        mdp: MDP\n",
    "        gamma: discount factor\n",
    "        nIt: number of iterations, corresponding to n above\n",
    "    Outputs:\n",
    "        (value_functions, policies)\n",
    "        \n",
    "    len(value_functions) == nIt+1 and len(policies) == nIt\n",
    "    \"\"\"\n",
    "    grade_print(\"Iteration | max|V-Vprev| | # chg actions | V[0]\")\n",
    "    grade_print(\"----------+--------------+---------------+---------\")\n",
    "    Vs = [np.zeros(mdp.nS)] # list of value functions contains the initial value function V^{(0)}, which is zero\n",
    "    pis = []\n",
    "    for it in range(nIt):\n",
    "        oldpi = pis[-1] if len(pis) > 0 else None # \\pi^{(it)} = Greedy[V^{(it-1)}]. Just used for printout\n",
    "        Vprev = Vs[-1] # V^{(it)}\n",
    "        \n",
    "        # Your code should fill in meaningful values for the following two variables\n",
    "        # pi: greedy policy for Vprev (not V), \n",
    "        #     corresponding to the math above: \\pi^{(it)} = Greedy[V^{(it)}]\n",
    "        #     ** it needs to be numpy array of ints **\n",
    "        # V: bellman backup on Vprev\n",
    "        #     corresponding to the math above: V^{(it+1)} = T[V^{(it)}]\n",
    "        #     ** numpy array of floats **\n",
    "        \n",
    "        # >>>>> Your code\n",
    "        # hint: using 2 or 3 loop might be easy           \n",
    "        \n",
    "        #V = Vprev # REPLACE THIS LINE WITH YOUR CODE       \n",
    "        #pi = oldpi # REPLACE THIS LINE WITH YOUR CODE\n",
    "        \n",
    "        V=np.zeros(mdp.nS)\n",
    "        pi=np.zeros(mdp.nS) \n",
    "        \n",
    "        for s in range(mdp.nS):            \n",
    "            a_array=np.zeros(mdp.nA)\n",
    "            \n",
    "            for a in range(mdp.nA): \n",
    "                mysum=0\n",
    "                for ss in range(len(mdp.P[s][a])):\n",
    "                    prob=mdp.P[s][a][ss][0]\n",
    "                    nextstate=mdp.P[s][a][ss][1]\n",
    "                    reward=mdp.P[s][a][ss][2]\n",
    "\n",
    "                    mysum=mysum+prob*(reward+GAMMA*Vprev[nextstate])\n",
    "                a_array[a]=mysum\n",
    "            \n",
    "            max_a = np.argmax(a_array)\n",
    "            max_sum = a_array[max_a]\n",
    "            V[s]=max_sum\n",
    "            pi[s]=max_a          \n",
    "        \n",
    "        # >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>        \n",
    "        max_diff = np.abs(V - Vprev).max()\n",
    "        nChgActions=\"N/A\" if oldpi is None else (pi != oldpi).sum()\n",
    "        grade_print(\"%4i      | %6.5f      | %4s          | %5.3f\"%(it, max_diff, nChgActions, V[0]))\n",
    "        Vs.append(V)\n",
    "        pis.append(pi)\n",
    "    return Vs, pis\n",
    "\n",
    "GAMMA = 0.95 # we'll be using this same value in subsequent problems\n",
    "\n",
    "\n",
    "# The following is the output of a correct implementation; when\n",
    "#   this code block is run, your implementation's print output will be\n",
    "#   compared with expected output.\n",
    "#   (incorrect line in red background with correct line printed side by side to help you debug)\n",
    "expected_output = \"\"\"Iteration | max|V-Vprev| | # chg actions | V[0]\n",
    "----------+--------------+---------------+---------\n",
    "   0      | 0.80000      |  N/A          | 0.000\n",
    "   1      | 0.60800      |    2          | 0.000\n",
    "   2      | 0.51984      |    2          | 0.000\n",
    "   3      | 0.39508      |    2          | 0.000\n",
    "   4      | 0.30026      |    2          | 0.000\n",
    "   5      | 0.25355      |    1          | 0.254\n",
    "   6      | 0.10478      |    0          | 0.345\n",
    "   7      | 0.09657      |    0          | 0.442\n",
    "   8      | 0.03656      |    0          | 0.478\n",
    "   9      | 0.02772      |    0          | 0.506\n",
    "  10      | 0.01111      |    0          | 0.517\n",
    "  11      | 0.00735      |    0          | 0.524\n",
    "  12      | 0.00310      |    0          | 0.527\n",
    "  13      | 0.00190      |    0          | 0.529\n",
    "  14      | 0.00083      |    0          | 0.530\n",
    "  15      | 0.00049      |    0          | 0.531\n",
    "  16      | 0.00022      |    0          | 0.531\n",
    "  17      | 0.00013      |    0          | 0.531\n",
    "  18      | 0.00006      |    0          | 0.531\n",
    "  19      | 0.00003      |    0          | 0.531\"\"\"\n",
    "Vs_VI, pis_VI = value_iteration(mdp, gamma=GAMMA, nIt=20, grade_print=make_grader(expected_output))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd668a4eb70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAADVRJREFUeJzt3X1sFHUex/H37G5bWmqhHMUHBKrkVI5wgMUIVnzKiSKC\nT2fkKXgq0mI8Dg8fiQ/gnqJRg4nGwBkjmhPqiTGYWDEilejpGRBUUOsJIiIoclXgoFrb7twf2y2l\nD7uzu7Mz2x+fV7MJtPubfjp8WGZKZ76WbduImCrgdwCRTFLBxWgquBhNBRejqeBiNBVcjKaCi9FU\ncDGaCi5GCyV6gmVZs4BZ0d/1LIPTMhxJxIkPsW3bSvQsK5n/qresUTZ8mFaszIl9HQm/Zh8po5uc\nFFyHKGI0FVyMpoKL0VRwMZoKLkZTwcVoKrgYTQUXo6ngYjQVXIymgovRVHAxmgouRlPBxWgJfx48\nI4qBccBAIA+oB34AXgN+8iXRkeYCvTt5/xLge4+zdCXbM2ZJPn8Kfg1wHPAVUAcUAYOAY8iOgsd8\nwZF5DvkVJI5sz+hzvrQLbmFRTjlTmcpiFvMlX8ZfkE+03D8Dz7d5f5CMHTDlkstEJnI+53MHd3DI\n6V7eBNRmJlN7xRQznen0oQ8LWeh8oYcZSyllFrP4mI95kRedLfIwX2dSLnis2JVUUkwxIUIMYlDi\ngje0PPKBSmA7sAPYBjSmmqZzueQyiUnMYAYhQgQJUkih84KPBErb/H61u/ngcLEnMAELiwYaktuA\nBxljxT6d08khhyBB5wX3IF88KV+ydhVXcTM3p/RJa4bW8NjExzjUo03RDgLLgd0pbZLOLrUKE+Ys\nziKQ5D8Nk+dOZk/vPR0/sCDVbDEdM77My/SiF0GCSW3Jq4zHcixVVBEhktR+zFy+w5xcspbyK/ib\nvElf+nI5l2O17IyVrORd3k28+FPo/0V/tg7aSmRQBE4HCoFzgRWpJupoKUuJEOFMziRIEBubMGH2\nsjfuujrqADiu6ji+r83sGVGYMLOZzQAGkE8+jTQyhzkJ18UyFlcV81Nt5k5c9rKXRSziRm6koOVt\nG9t4lEcd5etf1Z/62np+8uvkyrZtxw8os4n+FW99FFFkV1BhV1Ntl3Xy8Q6PADYD271vNDYLsLk2\nwdq4D7vl0fFjAxhgL2Sh/Rqv2b3pnXhbc1vynJZOnuQyjmCEvZSl9jKWOduWxxkDBOxxjLNf4iV7\nHvN8zHf44aSzrl1VHyBAhEjijeQC84G9wHdEj7uHAAXAW8A7juO0E/s6uv5Xy3HG2Le4qnD5BKn7\nZ/Q/32EZPURpz9EXDdAEvE/0xOO3QA5wAFgP/MutNJ1znNFH2Z4x2/O1p/uieEoZ3aT7oshRTwUX\no6ngYjQVXIymgovRVHAxmgouRlPBxWgquBhNBRejqeBiNBVcjKaCi9FUcDFaCj8P7vzHa/2R7flA\nGd3g7Md5E76CW5Y1y7KsDZZlbSDBtYwi2UYXPHhKGd2kCx7kqKeCi9FUcDGaCi5GU8HFaCq4GE0F\nF6Op4GI0FVyMpoKL0VRwMZoKLkZTwcVo3o8RzJL5iV3q7MbtpcCfgF+Ah3xJ1ZH2oyP+zMkE3+cn\nGkP7MS7XCh4iRBNNzhf4MD8x6Yw+yPb9GCRIhAh21l/xE5V2wWNzHi/lUu7mbtaz3tlCD+cnxuY8\njmAE05jmbOJX23xFmcsWM4pRzGY2IUJcy7XOF3q0H4MEuYiLmMlM3ud9HuERZws93o/tpVzwXvRi\nBjOYwAQAbGzO4Ax+5de46z7jMxpphFPbfSADfzCDGEQFFYxkJDnkYGNTRlncMYJd5suQMsq4iZs4\nnuPJJ59mmhnO8ITrvNqPAQJczMXMZCZ55FFAAUMZmjCj1/uxKylfsnYFVzia59hebEBouCrM2tq1\n1FCT9DY61/FSq/u5n3LKUxpgGq4Kc3bt2UxjGrtLd7t0ctQx40pW0pveKQ+CDVeFWVi70MVDryMz\n9qMfK1qGl6a6H/Nr87mVW10/yXRyyVpa12SOZjSVVFJCCSFCPMIjrGFN/I14OP4uhxzGM57ruZ4c\ncggRYgYz2EMnE3jj5SslYwUvoogpTOFyLidAgAYamMSkxJvycD8OYAAzmcmZnEkOOWxgA3dwR/L5\nSvG84Gkdg/+75W00o5nKVLazPZ3Nua6RRl7lVV7ndcYzngu4gIMc9DvWEQ5wgKUsZQUrmMIUSijx\nO1IHO9nJfdzXWvTNbPY7kmOufBclVvRsFSv6q7zqd5QuxYqezWJF70502whPKaObdNsIOeqp4GI0\nFVyMpoKL0VRwMZoKLkZTwcVoKrgYTQUXo6ngYjQVXIymgovRVHAxmgouRlPBxWgaBOsLZUyfBsGK\n6Ioebymjm3RFjxz1VHAxmgouRlPBxWgquBhNBRejqeBiNBVcjKaCi9FUcDGaCi5GU8HFaCq4GE2D\nYDujjOnTIFiyf4CpMnZ7rhX8FE5hK1uJEHG2wIcBpidxElvZ6nyRD8Nqk+ZxxhM4gYMc5AAHvPuk\naUi74GdwBpVUcjInEybMWtY6W+jDANNiirmGa/iBH5wt9nBYbco8yngCJ3ADN3Ae5/EBHzCf+c4W\ndtdBsIMZzHzmtw4wbaCBuczlQi6Mu+5hHmYf+zwZBDuWsdzCLa0DTAEWsID97M+ajGnLcMYe9OA2\nbqOccoIECRBgDGNYxKK467rchx5LueAnciIDGdh6SBIiRB55jGZ03HV55AEQrgrzSu0rbGRjqhES\nGsYwCik8YoDpEIYkXNc24xO1Tzh/xfdBuCrM8trlfM7nGdl+D3pQRhk2NqE2dUnmz3lX7S6WsOTw\nSaaHUi74OtaxhS3MYAYXcRFBgjzAA4kHwba4h3tS/dSOPcVTvMEbVFDBMIYRIsR0pscfBNuGFxnd\nkEtuxra9j31czdVcwiVcx3UUUMAmNiUeBNvC732Y1jF4HXUsZjHP8zwTmchHfORWLtdsYxu3czuD\nGUw55fzIj35H6nYaaWQVq6immvGMZxe7/I7kmCvfRamjjmUsc2NTGbOt5U1SFxuo263Ytu34AWU2\n0fsKZOHDbnn4ncPbjDXU2MMZntUZM/Vw0ln9V70YTQUXo6ngYjQVXIymgovRVHAxmgreTfWkJ8dy\nLAAllFBMsc+JspNun+wp9zI+zuMMYQi55NJEE400MolJNNGU5pa7w36M0u2TDVZDTWuZLSw2scmF\ncptHBe+mqqmmkUYg+l/oT/O0z4mykwreTTXSyLM8S4QIG9nI13ztd6Ss5N81mZK2aqopo4xneMbv\nKFlLJ5meUkY36SRTjnoquBhNg2B9oYzp0yBYEZ1keiuaccOGbN2HMGpUWcuvsnk/RukkU456KrgY\nTQUXo6ngYjQVXIymgovRVHAxmgouRlPBxWgquBhNBRejqeBiNBVcjObtNZlZMhw0oWwfsgpMfGsi\n3/38XYf3vzD2BU7t5fPkp5j+wNnAACAfqAd+ADZAhkYKdaCLjuPpBkNWx/Yby4k9T2z9fXFultzh\n6nfAH4keI+wF/gPkES39MLpXwQsp5GIuZg1roqPjTJHkkNW+9OUczmE1q6mnPmOx2rps4GWcd9x5\njp/f9EkT9rc2oXEhrFByP/MdIMD5nM8e9rCFLV0/MQe4lGi5NwOvQOt8YAvom9SnTUtaBS+kkMlM\n5kquJIcc9rHP2ZQ1n4eDOpbkkNVyypnNbG7gBqqo4mVeznjRV32zig/rDl9AMW/ovLjP//XZX2l+\nrxnrCYu8P+c5Knqs2LOYRR/6sJGN8aesDYCWsaSwDo4Yfm3j6YVhKV/RcxZncS/3AtGZiE00HTFH\nsTOT505mT+8uRvilfQzu4hU9XRyD1yyoSbg0th9+4RdsbOYzv830Ofeu6OnqGNxJxlb5YBVaFPyj\ngMBvot9vaH9FTxFFPM3THMMx5JPvaLNvDnuTB696EIA3/vYGy5uW89wfnosej8cscB6zK06u6En5\nFfynlreilpfgZpoJEWI727tcE7t33s1VNzOkdggP8AC7S3d7PhzUqZKqEsK14dY5lPG+NoCTOKl1\nMG6ECHXUZXym+0M5DzHmmTGH33Fy/OdHvmp5ObWACARKA1g5XfekiSZ2sINhDKOZZoIEgfj7ouFQ\nQ+uvNxZt5Jsfv4FvgE+A3yf4glyWcsE/53OmMpVzOZcKKuhHP8eDYJ/kyVQ/raf2spdKKh0//zIu\nYw5z2MlOlrCE93gvg+miQsND9PxnT8fPr/9rPc3vNBMcFSTvL3kETwvGfz71rXNGK6igjDLWs547\nubPrRTuJfsekAO4aexesInqSeYDuU3AAG5u3eZt1rGMkI9nMZrdydUurWc0OdmTlQNyYHnf1wJ5l\nJyx2e20H6u5nf/wnNwLVwJVEz2OOB74FeqUUOS2ufBfFxs7ozPnuooGGrC43QKAkACWpr3c8THcL\n0VfscqInnSOIfpt1K/Bp6p8/WbpthKd02wg36bYRctRTwcVoKrgYTQUXo6ngYjQVXIymgovRVHAx\nmgouRlPBxWgquBhNBRejqeBiNBVcjKaCi9GS/Hlw639E7xaSrfoC//U7RALK6I5Btm0nvHQj2St6\nvrBte1SKgTLOsqwN2ZwPlNFrOkQRo6ngYrRkC/73jKRwT7bnA2X0VFInmSLdjQ5RxGgquBhNBRej\nqeBiNBVcjPZ/P0/RfJ6kt4UAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd665636cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAADjVJREFUeJzt3X1wVPW9x/H32QfyQEhISEAewxRr5XKVhxShkDtW24pU\n6XXopU2xk2tLeOgdr5cZa0XEqrPXqmNbnY7WJ2YudToSazuId6B6RVPmytgOiBZoDZVcROUhQkga\nICHk4dw/NpuQp92zm909Z398XpmdycM5u5+cfLLZs9lzvpZt24iYyud2AJFUUsHFaCq4GE0FF6Op\n4GI0FVyMpoKL0VRwMZoKLkYLxFrAsqxVwKrwRyPL4MoURxJx4l1s27ZiLWXF8696y/qiDe8OK1bq\nRL6PmN+zi5QxmZwUXA9RxGgquBhNBRejqeBiNBVcjKaCi9FUcDGaCi5GU8HFaCq4GE0FF6Op4GI0\nFVyMpoKL0WK+HjwlCoEbgClAFtACfAZsAxpdSdTXWmD0IJ9/BjiR5ixD8XpGj+Rzp+DfBi4D/g9o\nAPKBUmAU3ih4xEH65jnnVpAovJ7R5XzDLriFxUIWspzlPM7jfMiH0VfIIVzuVuCFiz7vJ2UPmEYw\ngiUs4Tqu427u5pzTrfweUJuaTP0VUsh3+S5FFPEgDzpfMY0ZpzKVVaziz/yZl3jJ2UppzDeYhAse\nKfYa1lBIIQEClFIau+Bt3ZccYA1wGDgC1AHtiaYZ3AhG8A2+QSWVBAjgx08eec4LPhuYetHHryU3\nH/QW+yZuwsKijbb4riANGSPFnsMcggTx43de8DTkiybhQ9a+yTe5ndsTutGaGTX8bMnPOJd9UdHO\nAi8CxxK6SgY71CpEiAUswBfnn4aKtRXUj64f+IUHEs0WMTDj7/gdBRTgxx/XNaUr4zjGUU01XXTF\ntR1Tl6+Xk0PWEr4Hf4M3KKaYW7gFq3tj/Jbf8jZvx175LzDx4EQOlR6iq7QL5gB5wLXA5kQTDfQs\nz9JFF/OYhx8/NjYhQpzkZNT1GmgA4LLqyzhRm9o9ohAhfsAPmMxkcsihnXbu4I6Y60UyFlYX0lib\nuh2Xk5zkYR5mJSvJ7X6ro46f8lNH+SZWT6SltoVGt3aubNt2fIEym/CveM8ln3x7Navt7Wy3ywb5\n+oCLD5sp/T43H5sHsPnXGOtGvdjdl4Ffm8xk+0EetLexzR7N6NjXtbY7z5XDyRNfxlnMsp/lWXsT\nm5xdV5oz+vDZN3CD/TIv23dyp4v5ei9OOpu0o+p9+OiiK/aVjADWAyeB44Qfd08HcoE3gf91HKef\nyPcx9F8txxkjT3FVk+QdpMzP6H6+Xil9iNKfo28aoAN4h/COx+eBINAM7AZ2JSvN4BxndJHXM3o9\nX386L0paKWMy6bwocslTwcVoKrgYTQUXo6ngYjQVXIymgovRVHAxmgouRlPBxWgquBhNBRejqeBi\nNBVcjJbA68Gdv7zWHV7PB8qYDM5ezhvzHtyyrFWWZe2xLGsPMY5lFPEaHfCQVsqYTDrgQS55KrgY\nTQUXo6ngYjQVXIymgovRVHAxmgouRlPBxWgquBhNBRejqeBiNBVcjJb+MYIemZ84pMFO3D4VuA04\nDzziSqqBtB0dcWdOJrg+P9EY2o5RJa3gAQJ00OF8BRfmJ8adMROkeTv68dNFF7bnj/gJG3bBI3Me\nb+ZmNrCB3ex2tmIa5ydG5jzOYha3cquziV8X58tPXbZhS9N29ONnEYuooop3eIfHeMzZii5vx4QL\nXkABlVRyEzcBYGMzl7lc4ELU9f7KX2mnHb7Q7wsp+MGUUspqVjOb2QQJYmNTRlnUMYJD5vOqFG9H\nHz5u5EaqqCKLLHLJZQYzmMnMqOt5ZTsmXPDruZ6lLO3zuWXdb9FUUEE99YSqQ7xV+xY11CQaIaYV\nrGAe8/oMML2Xex3nK68t51Zu5djUY+GdIw8KVYfYWLuRIxxJyfUXU8yd3AnQsx1LKeUJnoi63sXb\nMac2hx/yw96dzDRKuOBb2MJxjrOGNZRQQoAAj/EYO9jhaP37uC/Rm3YsRIjFLOb7fJ8gQQIEqKSS\negaZwOtCvmQZzeiUFfwzPuM2bqOKKuYxjyBB9rCHu7nb0fpub8dhPQb/Y/fbfOaznOUc5nCyciVF\nO+28yqv8nt+zmMVcz/Wc5azbsTLOJ3zC/dzPZCZTRRX72e92JMeS8ixKpOheFSn6q7zqdpSMFil6\nRhnuKG/vXIYek+2dS/Iz1lBjz2SmpzOm6uKks/pXvRhNBRejqeBiNBVcjKaCi9FUcDGaCi5G0+mT\n0yp5GTewgWu4hlGMopNOjnKUFaxIwsuBM2E7hun0yQY7yUmyyALCL2Xt6H6TvlTwDLWZzXTRBUAr\nrTzN0y4n8iYVPEM108wrvEInnRznOHvY43YkT1LBM9hmNtNEE7/kl25H8SztZKaVMiaTdjLlkqeC\ni9E0CNYVyjh8GgQrop3M9ApnfP75jS7nGNrKlVXd73l5O4ZpJ1MueSq4GE0FF6Op4GI0FVyMpoKL\n0VRwMZoKLkZTwcVoKrgYTQUXo6ngYjQVXIymQbCDyYCM606so6GzYcDn7yu5jykjpriQqB8NgsX7\nA0wzIOPV2VdT4i/p+XiUf5SLabwnaQW/gis4xKGec3XE5MIg2LhlQMby3HJm58xO2+1NYAJnOUsz\nzWm7zeEYdsHnMpc1rOFzfI4QId7iLWcrpnEQbMIyIOPbLW9zsO1gz8cVoytScjsTmMAKVvBlvsyf\n+BPrWe9sxUwdBDuNaaxnPeMZTw45tNHGWtbyNb4Wdb1HeZQmmtIyCHbYMiDjvvP7+nyc7IJnk81d\n3MVCFuLHjw8fX+JLPMzDUdcb8uecZgkXfBKTmMKUnockAQJkkcV85kddL3I+vVB1iC21W9jL3kQj\npFyoOkRjbSM/5+duRxlSqDpE6S2ltE1rS8n1Z5NNGWXY2AQuqks8P+ejtUd5hmcyaxDsTnZygANU\nUskiFuHHz0M85KlBsMlQQknshVzm60jds71NNLGMZXydr/M9vkcuubzHe5fGINgGGnicx3mBF1jC\nEt7n/WTlEg9pp52tbGU721nMYo5y1O1IjiXlWZQGGtjEpmRclXhYZKBuJkn/8+BPpP0W49edsZxy\nz05wfuSy8H9KLq+9nKM3e/AedbCf80fAA+mNoX/Vi9FUcDGaCi5GU8HFaCq4GE0FF6Op4IMIEmQc\n4wAooICxjHU50UC+Fh+BU+Fnef2NfvzNfpcTeZMKPojv8B1+za8BmM50XuIlLudyl1P1Nf7J8ZRu\nKAVg3KZxlN5TisZkDqSCD2IXu/oMVW2kkcMcdjHRQGfmnsEOdJ8TvQtar2x18/AVz1LBB1FHHfvZ\nTyedtNDCRjbSSafbsfpoLm/uKbgdsGlYOvDwNVHBh/Qcz9FFF2208Tqvux1noCA0/HMDtmXTOr2V\nCxMvuJ3Ik/RHbQiHOMRWtrKPfZ67945oLm8m94NcGm7RvfdQVPAonuIptyNEF4QT/+aRw/w9Sg9R\nxGgquBgtgTGCe1IYR8QpKzljBDUIVjKZcYNgM2HI6kcfHXE5ydCmTi3tfk+DYEU8TwUXo6ngYjQV\nXIymgovRVHAxmgouRlPBxWgquBhNBRejqeBiNBVcjKaCi9HSe8iaR4aDxuL5IavAwpcXcvTcwPOC\nb1uyjRljZriQaBATgXJgMpADtACfET6k4IP0RNAxmVFkwpDVr0z6ClNG9f7Sjcke42Kai/wD8C+E\nHyOcBP4GZBEu/VVkVsHzyONGbmQHO8Kj4wwR75BVf6OfvHfzaF7YjJ3j/HX2w/Gtz3+LRaWLHC9/\nfu95Oo50MHLJSKxAfK/59uHjOq6jnnoOcGDoBYPAzYTLvR/YAj3zgS2gOK6bHZZhFTyPPCqoYClL\nCRKkiSZnU9ZcHg7qVLxDVke+P5Lil4sZ88oYGhc10vTVppQX/Tcf/oY/nugds3L/vPujLt/0VBOt\nO1s5/chpitYVOSp6pNirWEURRexlb/Qpa5OB3O73d0Kf4dc2aT0wLOGCL2ABP+bHQHgmYgcd3Nv9\nNpQKKqin3vXhoE71H7K64a4NMdexfTa+Nh+FrxVS+Fohx/79GOevPJ+qiLz56Zt9Pq78dqWj9Trr\nOzm1/hSnHz3NxP+eiL9k8JN35pPP8zzPKEaRQw4A13ANNdQMed1vjHyDn/ATAF5vep0XeZFfffVX\n4cfjEQ84ijlsCRe8sfstv/suuJNOAgSinsMvcr6/26tvZ3rtdB7iIY5NPZb24aBO3cEdLP6vxVjt\n4Xu4tgnRh61mHcvCtmwsLLCgo6CDrryuqOsM1y8KfsGcp+f0fuKK6Mu3/609/I4F2BCcFoQRQy/f\nQQdHOMJVXEUnnfgJ/yJE+zm3nevdTnvz9/Lx6Y/hY2AfcHX0fMmWcME/4AOWs5xruZbVrGYsYx0P\ngn2SJxO92bTqKOrg0w2fOl4+vyafks0lXBh7gVPLTtEysyXlhzZmzcli0v9Mcrz8iZUnaN3RSvaC\nbIruKSLrH7OiLt9CCz/iR0xjGqtZTRll7GY361g39EqfEH7GJBfu+ad7YCvhncxmMqfgADY2f+AP\n7GQns5nNfvYnK1dGOrPgDO3j22n9Qqtnj9kt/s9iOv+jM2ax+6ujrqfof+fv0RduB7YDSwnvb40H\nPgUKEoo8LEl5FsXG9vTM+XSxs+zwaYw9LDAuQGBc4j/2OuqcLXiA8D32QsI7nbOAc8Ah4C8J33zc\n0vs8uEeGg8YSGbLqZbuW7XI7Qmwfd19cpH/Vi9FUcDGaCi5GU8HFaCq4GE0FF6Op4GI0FVyMpoKL\n0VRwMZoKLkZTwcVoKrgYTQUXo6ngYrQ4xwhaZ4CDMRd0TzFwyu0QMShjcpTatl0Sa6F4D3g4aNv2\nFxMMlHKWZe3xcj5QxnTTQxQxmgouRou34M+lJEXyeD0fKGNaxbWTKZJp9BBFjKaCi9FUcDGaCi5G\nU8HFaP8PZiDfQRHJyoUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd66558aac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6654dc8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6653ab710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6652fe4e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd66524e2e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nko2cogilScCF0iTgQmkScKE0CbhQ2v8DTkzz5SNrMkgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6651a10f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFEUoTQIulCYBF0qTgAulScCF0v4f8L3sEOAey10AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd665069eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAALgAAAC4CAYAAABQMybHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd664fbccc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd668a4e128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the estimated values and corresponding policy\n",
    "for (V, pi) in zip(Vs_VI[:10], pis_VI[:10]):\n",
    "    plt.figure(figsize=(3,3))\n",
    "    plt.imshow(V.reshape(4,4), cmap='gray', interpolation='none', clim=(0,1))\n",
    "    ax = plt.gca()\n",
    "    ax.set_xticks(np.arange(4)-.5)\n",
    "    ax.set_yticks(np.arange(4)-.5)\n",
    "    ax.set_xticklabels([])\n",
    "    ax.set_yticklabels([])\n",
    "    Y, X = np.mgrid[0:4, 0:4]\n",
    "    a2uv = {0: (-1, 0), 1:(0, -1), 2:(1,0), 3:(-1, 0)}\n",
    "    Pi = pi.reshape(4,4)\n",
    "    for y in range(4):\n",
    "        for x in range(4):\n",
    "            a = Pi[y, x]\n",
    "            u, v = a2uv[a]\n",
    "            plt.arrow(x, y,u*.3, -v*.3, color='m', head_width=0.1, head_length=0.1) \n",
    "            plt.text(x, y, str(env.desc[y,x].item().decode()),\n",
    "                     color='g', size=12,  verticalalignment='center',\n",
    "                     horizontalalignment='center', fontweight='bold')\n",
    "    plt.grid(color='b', lw=2, ls='-')\n",
    "plt.figure()\n",
    "plt.plot(Vs_VI)\n",
    "plt.xlabel(\"iterations\")\n",
    "plt.ylabel(\"Values\")\n",
    "plt.title(\"Values of different states\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Problem 2: Policy Iteration\n",
    "\n",
    "The next task is to implement exact policy iteration (PI), which has the following pseudocode:\n",
    "\n",
    "---\n",
    "Initialize $\\pi_0$\n",
    "\n",
    "For $n=0, 1, 2, \\dots$\n",
    "- Compute the state-value function $V^{\\pi_{n}}$\n",
    "- Using $V^{\\pi_{n}}$, compute the state-action-value function $Q^{\\pi_{n}}$\n",
    "- Compute new policy $\\pi_{n+1}(s) = \\operatorname*{argmax}_a Q^{\\pi_{n}}(s,a)$\n",
    "---\n",
    "\n",
    "Below, you'll implement the first and second steps of the loop step by step.\n",
    "\n",
    "### Problem 2a: state value function\n",
    "\n",
    "You'll write a function called `compute_vpi` that computes the state-value function $V^{\\pi}$ for an arbitrary policy $\\pi$.\n",
    "Recall that $V^{\\pi}$ satisfies the following linear equation:\n",
    "$$V^{\\pi}(s) = \\sum_{s'} P(s,\\pi(s),s')[ R(s,\\pi(s),s') + \\gamma V^{\\pi}(s')]$$\n",
    "\n",
    "**Here, we're going to solve the exact value function. You can solve a linear system in your code. (Find an exact solution, e.g., with `np.linalg.solve`.)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# solve the exact values with `np.linalg.solve`\n",
    "def compute_vpi(pi, mdp, gamma):\n",
    "    # use pi[state] to access the action that's prescribed by this policy\n",
    "    # V[s] = P*(R + \\gamma*V[s'])\n",
    "    # => (I-\\gamma*P)*V = P*R\n",
    "    # solve linear matrix\n",
    "    # (I-\\gamma*P): (nS, nS) => a\n",
    "    # P*R => b\n",
    "    \n",
    "    # >>>>> Your code\n",
    "    a = np.zeros((mdp.nS, mdp.nS)) \n",
    "    b = np.zeros(mdp.nS) \n",
    "    V = np.zeros(mdp.nS) # REPLACE THIS LINE WITH YOUR CODE    \n",
    "\n",
    "    for s in range(mdp.nS):\n",
    "        a[s][s] = 1; #I\n",
    "        for p in mdp.P[s][pi[s]]:\n",
    "            prob=p[0]\n",
    "            ss=p[1] #s'\n",
    "            reward=p[2]\n",
    "            \n",
    "            b[s] = b[s] + prob * reward  #P*R => b\n",
    "            a[s][ss] = a[s][ss] - gamma * prob #I-\\gamma*P\n",
    "    V = np.linalg.solve(a, b)    \n",
    "    \n",
    "    # >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n",
    "    return V"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's compute the value of an arbitrarily-chosen policy. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test passed\n"
     ]
    }
   ],
   "source": [
    "expected_val = np.load('compute_vpi_result.npy')\n",
    "policy = np.array([1, 0, 3, 3, 1, 3, 3, 1, 2, 2, 1, 1, 0, 3, 3, 3])\n",
    "actual_val = compute_vpi(policy, mdp, gamma=GAMMA)\n",
    "if np.all(np.isclose(actual_val, expected_val, atol=1e-4)):\n",
    "    print(\"Test passed\")\n",
    "else:\n",
    "    print(\"Expected: \", expected_val)\n",
    "    print(\"Actual: \", actual_val)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem 2b: state-action value function\n",
    "\n",
    "Next, you'll write a function to compute the state-action value function $Q^{\\pi}$, defined as follows\n",
    "\n",
    "$$Q^{\\pi}(s, a) = \\sum_{s'} P(s,a,s')[ R(s,a,s') + \\gamma V^{\\pi}(s')]$$\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test passed\n"
     ]
    }
   ],
   "source": [
    "def compute_qpi(vpi, mdp, gamma):\n",
    "    # >>>>> Your code\n",
    "    Qpi = np.zeros([mdp.nS, mdp.nA]) # REPLACE THIS LINE WITH YOUR CODE\n",
    "    for s in range(mdp.nS):\n",
    "        for a in range(mdp.nA):\n",
    "            my_sum=0\n",
    "            for pi in mdp.P[s][a]:\n",
    "                prob=pi[0]\n",
    "                nextstate=pi[1]\n",
    "                reward=pi[2]\n",
    "                my_sum = my_sum + prob * (reward+gamma*vpi[nextstate])\n",
    "            Qpi[s][a]=my_sum\n",
    "    # >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n",
    "    return Qpi\n",
    "\n",
    "expected_Qpi = np.array([[  0.38 ,   3.135,   1.14 ,   0.095],\n",
    "       [  0.57 ,   3.99 ,   2.09 ,   0.95 ],\n",
    "       [  1.52 ,   4.94 ,   3.04 ,   1.9  ],\n",
    "       [  2.47 ,   5.795,   3.23 ,   2.755],\n",
    "       [  3.8  ,   6.935,   4.56 ,   0.855],\n",
    "       [  4.75 ,   4.75 ,   4.75 ,   4.75 ],\n",
    "       [  4.94 ,   8.74 ,   6.46 ,   2.66 ],\n",
    "       [  6.65 ,   6.65 ,   6.65 ,   6.65 ],\n",
    "       [  7.6  ,  10.735,   8.36 ,   4.655],\n",
    "       [  7.79 ,  11.59 ,   9.31 ,   5.51 ],\n",
    "       [  8.74 ,  12.54 ,  10.26 ,   6.46 ],\n",
    "       [ 10.45 ,  10.45 ,  10.45 ,  10.45 ],\n",
    "       [ 11.4  ,  11.4  ,  11.4  ,  11.4  ],\n",
    "       [ 11.21 ,  12.35 ,  12.73 ,   9.31 ],\n",
    "       [ 12.16 ,  13.4  ,  14.48 ,  10.36 ],\n",
    "       [ 14.25 ,  14.25 ,  14.25 ,  14.25 ]])\n",
    "\n",
    "Qpi = compute_qpi(np.arange(mdp.nS), mdp, gamma=GAMMA)\n",
    "if np.all(np.isclose(expected_Qpi, Qpi, atol=1e-4)):\n",
    "    print(\"Test passed\")\n",
    "else:\n",
    "    print(\"Expected: \", expected_Qpi)\n",
    "    print(\"Actual: \", Qpi)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we're going to run the poolicy iteration!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration | # chg actions | V[0]\n",
      "----------+---------------+---------\n",
      "   0      |      1        | -0.00000\n",
      "   1      |      9        | 0.00000\n",
      "   2      |      2        | 0.39785\n",
      "   3      |      1        | 0.45546\n",
      "   4      |      0        | 0.53118\n",
      "   5      |      0        | 0.53118\n",
      "   6      |      0        | 0.53118\n",
      "   7      |      0        | 0.53118\n",
      "   8      |      0        | 0.53118\n",
      "   9      |      0        | 0.53118\n",
      "  10      |      0        | 0.53118\n",
      "  11      |      0        | 0.53118\n",
      "  12      |      0        | 0.53118\n",
      "  13      |      0        | 0.53118\n",
      "  14      |      0        | 0.53118\n",
      "  15      |      0        | 0.53118\n",
      "  16      |      0        | 0.53118\n",
      "  17      |      0        | 0.53118\n",
      "  18      |      0        | 0.53118\n",
      "  19      |      0        | 0.53118\n",
      "Test succeeded\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJAsKxhhjJllQMMYYM8mCgjHGmEn/H3t+oZmrMdnUAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd664cf10f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def policy_iteration(mdp, gamma, nIt, grade_print=print):\n",
    "    Vs = []\n",
    "    pis = []\n",
    "    pi_prev = np.zeros(mdp.nS,dtype='int')\n",
    "    pis.append(pi_prev)\n",
    "    grade_print(\"Iteration | # chg actions | V[0]\")\n",
    "    grade_print(\"----------+---------------+---------\")\n",
    "    for it in range(nIt):\n",
    "        # you need to compute qpi which is the state-action values for current pi\n",
    "        #               and compute the greedily policy, pi, from qpi\n",
    "        # >>>>> Your code (sample code are 3 lines)\n",
    "        vpi=compute_vpi(pi_prev, mdp, gamma=GAMMA)\n",
    "        qpi=compute_qpi(vpi, mdp, gamma)\n",
    "        pi = qpi.argmax(axis = 1)        \n",
    "        # >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n",
    "        \n",
    "        grade_print(\"%4i      | %6i        | %6.5f\"%(it, (pi != pi_prev).sum(), vpi[0]))\n",
    "        Vs.append(vpi)\n",
    "        pis.append(pi)\n",
    "        pi_prev = pi\n",
    "    return Vs, pis\n",
    "\n",
    "expected_output = \"\"\"Iteration | # chg actions | V[0]\n",
    "----------+---------------+---------\n",
    "   0      |      1        | -0.00000\n",
    "   1      |      9        | 0.00000\n",
    "   2      |      2        | 0.39785\n",
    "   3      |      1        | 0.45546\n",
    "   4      |      0        | 0.53118\n",
    "   5      |      0        | 0.53118\n",
    "   6      |      0        | 0.53118\n",
    "   7      |      0        | 0.53118\n",
    "   8      |      0        | 0.53118\n",
    "   9      |      0        | 0.53118\n",
    "  10      |      0        | 0.53118\n",
    "  11      |      0        | 0.53118\n",
    "  12      |      0        | 0.53118\n",
    "  13      |      0        | 0.53118\n",
    "  14      |      0        | 0.53118\n",
    "  15      |      0        | 0.53118\n",
    "  16      |      0        | 0.53118\n",
    "  17      |      0        | 0.53118\n",
    "  18      |      0        | 0.53118\n",
    "  19      |      0        | 0.53118\"\"\"\n",
    "\n",
    "Vs_PI, pis_PI = policy_iteration(mdp, gamma=0.95, nIt=20, grade_print=make_grader(expected_output))\n",
    "plt.plot(Vs_PI);\n",
    "plt.xlabel(\"iterations\")\n",
    "plt.ylabel(\"Values\")\n",
    "plt.title(\"Values of different states\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem 3: Sampling-based Tabular Q-Learning\n",
    "\n",
    "So far we have implemented Value Iteration and Policy Iteration, both of which require access to an MDP's dynamics model. This requirement can sometimes be restrictive - for example, if the environment is given as a blackbox physics simulator, then we won't be able to read off the whole transition model.\n",
    "\n",
    "We can however use sampling-based Q-Learning to learn from this type of environments. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this exercise, we will learn to control a Crawler robot. Let's first try some completely random actions to see how the robot moves and familiarize ourselves with Gym environment interface again."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We can inspect the observation space and action space of this Gym Environment\n",
      "-----------------------------------------------------------------------------\n",
      "Action space: Discrete(4)\n",
      "It's a discrete space with 4 actions to take\n",
      "Each action corresponds to increasing/decreasing the angle of one of the joints\n",
      "We can also sample from this action space: 0\n",
      "Another action sample: 3\n",
      "Another action sample: 0\n",
      "Observation space: Tuple(Discrete(9), Discrete(13)) , which means it's a 9x13 grid.\n",
      "It's the discretized version of the robot's two joint angles\n"
     ]
    }
   ],
   "source": [
    "from crawler_env import CrawlingRobotEnv\n",
    "\n",
    "env = CrawlingRobotEnv()\n",
    "\n",
    "print(\"We can inspect the observation space and action space of this Gym Environment\")\n",
    "print(\"-----------------------------------------------------------------------------\")\n",
    "print(\"Action space:\", env.action_space)\n",
    "print(\"It's a discrete space with %i actions to take\" % env.action_space.n)\n",
    "print(\"Each action corresponds to increasing/decreasing the angle of one of the joints\")\n",
    "print(\"We can also sample from this action space:\", env.action_space.sample())\n",
    "print(\"Another action sample:\", env.action_space.sample())\n",
    "print(\"Another action sample:\", env.action_space.sample())\n",
    "print(\"Observation space:\", env.observation_space, \", which means it's a 9x13 grid.\")\n",
    "print(\"It's the discretized version of the robot's two joint angles\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "env = CrawlingRobotEnv(\n",
    "    render=True, # turn render mode on to visualize random motion\n",
    ")\n",
    "\n",
    "# standard procedure for interfacing with a Gym environment\n",
    "cur_state = env.reset() # reset environment and get initial state\n",
    "ret = 0.\n",
    "done = False\n",
    "i = 0\n",
    "while not done:\n",
    "    action = env.action_space.sample() # sample an action randomly\n",
    "    next_state, reward, done, info = env.step(action)\n",
    "    ret += reward\n",
    "    cur_state = next_state\n",
    "    i += 1\n",
    "    if i == 1500:\n",
    "        break # for the purpose of this visualization, let's only run for 1500 steps\n",
    "        # also note the GUI won't close automatically"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# you can close the visualization GUI with the following method \n",
    "env.close_gui()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You will see the random controller can sometimes make progress but it won't get very far. Let's implement Tabular Q-Learning with $\\epsilon$-greedy exploration to find a better policy piece by piece."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Q-values for state (0, 0): [ 0.  0.  0.  0.] which is a list of Q values for each action\n",
      "As such, the Q value of taking action 3 in state (1,2), i.e. Q((1,2), 3), can be accessed by q_vals[(1,2)][3]: 0.0\n"
     ]
    }
   ],
   "source": [
    "from collections import defaultdict\n",
    "import random\n",
    "\n",
    "# dictionary that maps from state, s, to a numpy array of Q values [Q(s, a_1), Q(s, a_2) ... Q(s, a_n)]\n",
    "#   and everything is initialized to 0.\n",
    "q_vals = defaultdict(lambda: np.array([0. for _ in range(env.action_space.n)]))\n",
    "\n",
    "print(\"Q-values for state (0, 0): %s\" % q_vals[(0, 0)], \"which is a list of Q values for each action\")\n",
    "print(\"As such, the Q value of taking action 3 in state (1,2), i.e. Q((1,2), 3), can be accessed by q_vals[(1,2)][3]:\", q_vals[(1,2)][3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test1 passed\n",
      "Test2 passed\n"
     ]
    }
   ],
   "source": [
    "def eps_greedy(q_vals, eps, state):\n",
    "    \"\"\"\n",
    "    Inputs:\n",
    "        q_vals: q value tables\n",
    "        eps: epsilon\n",
    "        state: current state\n",
    "    Outputs:\n",
    "        random action with probability of eps; argmax Q(s, .) with probability of (1-eps)\n",
    "    \"\"\"\n",
    "    # you might want to use random.random() to implement random exploration\n",
    "    #   number of actions can be read off from len(q_vals[state])\n",
    "    import random\n",
    "    # >>>>> Your code\n",
    "    action = 0\n",
    "    \n",
    "    if random.random() < eps:\n",
    "        action = random.randint(0,3)\n",
    "    else:\n",
    "        action = np.argmax(q_vals[state])    \n",
    "    \n",
    "    # >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n",
    "    return action\n",
    "\n",
    "# test case 1\n",
    "dummy_q = defaultdict(lambda: np.array([0. for _ in range(env.action_space.n)]))\n",
    "test_state = (0, 0)\n",
    "dummy_q[test_state][0] = 10.\n",
    "trials = 100000\n",
    "sampled_actions = [\n",
    "    int(eps_greedy(dummy_q, 0.3, test_state))\n",
    "    for _ in range(trials)\n",
    "]\n",
    "freq = np.sum(np.array(sampled_actions) == 0) / trials\n",
    "tgt_freq = 0.3 / env.action_space.n + 0.7\n",
    "if np.isclose(freq, tgt_freq, atol=1e-2):\n",
    "    print(\"Test1 passed\")\n",
    "else:\n",
    "    print(\"Test1: Expected to select 0 with frequency %.2f but got %.2f\" % (tgt_freq, freq))\n",
    "    \n",
    "# test case 2\n",
    "dummy_q = defaultdict(lambda: np.array([0. for _ in range(env.action_space.n)]))\n",
    "test_state = (0, 0)\n",
    "dummy_q[test_state][2] = 10.\n",
    "trials = 100000\n",
    "sampled_actions = [\n",
    "    int(eps_greedy(dummy_q, 0.5, test_state))\n",
    "    for _ in range(trials)\n",
    "]\n",
    "freq = np.sum(np.array(sampled_actions) == 2) / trials\n",
    "tgt_freq = 0.5 / env.action_space.n + 0.5\n",
    "if np.isclose(freq, tgt_freq, atol=1e-2):\n",
    "    print(\"Test2 passed\")\n",
    "else:\n",
    "    print(\"Test2: Expected to select 2 with frequency %.2f but got %.2f\" % (tgt_freq, freq))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next we will implement Q learning update. After we observe a transition $s, a, s', r$,\n",
    "\n",
    "$$\\textrm{target}(s') = R(s,a,s') + \\gamma \\max_{a'} Q_{\\theta_k}(s',a')$$\n",
    "\n",
    "\n",
    "$$Q_{k+1}(s,a) \\leftarrow (1-\\alpha) Q_k(s,a) + \\alpha \\left[ \\textrm{target}(s') \\right]$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test passed\n"
     ]
    }
   ],
   "source": [
    "def q_learning_update(gamma, alpha, q_vals, cur_state, action, next_state, reward):\n",
    "    \"\"\"\n",
    "    Inputs:\n",
    "        gamma: discount factor\n",
    "        alpha: learning rate\n",
    "        q_vals: q value table\n",
    "        cur_state: current state\n",
    "        action: action taken in current state\n",
    "        next_state: next state results from taking `action` in `cur_state`\n",
    "        reward: reward received from this transition\n",
    "    \n",
    "    Performs in-place update of q_vals table to implement one step of Q-learning\n",
    "    \"\"\"\n",
    "    # >>>>> Your code (sample code are 2 lines)\n",
    "    target = reward + gamma * q_vals[next_state][np.argmax(q_vals[next_state])]\n",
    "    q_vals[cur_state][action] = (1-alpha) * q_vals[cur_state][action] + alpha * target \n",
    "    # YOUR CODE HERE\n",
    "\n",
    "# testing your q_learning_update implementation\n",
    "dummy_q = q_vals.copy()\n",
    "test_state = (0, 0)\n",
    "test_next_state = (0, 1)\n",
    "dummy_q[test_state][0] = 10.\n",
    "dummy_q[test_next_state][1] = 10.\n",
    "q_learning_update(0.9, 0.1, dummy_q, test_state, 0, test_next_state, 1.1)\n",
    "tgt = 10.01\n",
    "if np.isclose(dummy_q[test_state][0], tgt,):\n",
    "    print(\"Test passed\")\n",
    "else:\n",
    "    print(\"Q(test_state, 0) is expected to be %.2f but got %.2f\" % (tgt, dummy_q[test_state][0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Itr 0 # Average speed: 0.05\n",
      "Itr 50000 # Average speed: 3.28\n",
      "Itr 100000 # Average speed: 3.37\n",
      "Itr 150000 # Average speed: 3.37\n",
      "Itr 200000 # Average speed: 3.37\n",
      "Itr 250000 # Average speed: 3.37\n"
     ]
    }
   ],
   "source": [
    "# now with the main components tested, we can put everything together to create a complete q learning agent\n",
    "\n",
    "env = CrawlingRobotEnv() \n",
    "q_vals = defaultdict(lambda: np.array([0. for _ in range(env.action_space.n)]))\n",
    "gamma = 0.9\n",
    "alpha = 0.1\n",
    "eps = 0.5\n",
    "cur_state = env.reset()\n",
    "\n",
    "def greedy_eval():\n",
    "    \"\"\"evaluate greedy policy w.r.t current q_vals\"\"\"\n",
    "    test_env = CrawlingRobotEnv(horizon=np.inf)\n",
    "    prev_state = test_env.reset()\n",
    "    ret = 0.\n",
    "    done = False\n",
    "    H = 100\n",
    "    for i in range(H):\n",
    "        action = np.argmax(q_vals[prev_state])\n",
    "        state, reward, done, info = test_env.step(action)\n",
    "        ret += reward\n",
    "        prev_state = state\n",
    "    return ret / H\n",
    "\n",
    "for itr in range(300000):\n",
    "    # YOUR CODE HERE\n",
    "    # Hint: use eps_greedy & q_learning_update\n",
    "    # >>>>> Your code (sample code are 4 lines)\n",
    "    action = eps_greedy(q_vals, eps, cur_state)\n",
    "    next_state, reward = env.step(action)[0:2]\n",
    "    q_learning_update(gamma, alpha, q_vals, cur_state, action, next_state, reward)\n",
    "    cur_state = next_state   \n",
    "    \n",
    "    # >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n",
    "    \n",
    "    if itr % 50000 == 0: # evaluation\n",
    "        print(\"Itr %i # Average speed: %.2f\" % (itr, greedy_eval()))\n",
    "\n",
    "# at the end of learning your crawler should reach a speed of >= 3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The average speed should be around 3.37"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After the learning is successful, we can visualize the learned robot controller. Remember we learn this just from interacting with the environment instead of peeking into the dynamics model!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "env = CrawlingRobotEnv(render=True, horizon=500)\n",
    "prev_state = env.reset()\n",
    "ret = 0.\n",
    "done = False\n",
    "while not done:\n",
    "    action = np.argmax(q_vals[prev_state])\n",
    "    state, reward, done, info = env.step(action)\n",
    "    ret += reward\n",
    "    prev_state = state"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# you can close the visualization GUI with the following method \n",
    "env.close_gui()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Congrat that you finish homework1!!!\n",
    "In homework1, we solved the exact state values (V) or state-action values (Q) with the transition function and reward function. However, in most cases, having transition function and reward function is a really strong assumption. \n",
    "\n",
    "So, in the next homework, we will assume that both transition function and reward function are unknown and use the agent's experiences (trial and error) to figure out how to maximize the accumulated reward."
   ]
  }
 ],
 "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}