!pip install aicrowd-cli > /dev/null

import os

AICROWD_DATASET_PATH = os.getenv("DATASET_PATH", os.getcwd()+"/a5562c7d-55f0-4d06-841c-110655bb04ec_a2_gridworld_inputs.zip")
AICROWD_RESULTS_DIR = os.getenv("OUTPUTS_DIR", "results")

!unzip -q $AICROWD_DATASET_PATH

DATASET_DIR = 'inputs/'

import numpy as np
from copy import deepcopy
import matplotlib.pyplot as plt
import pandas as pd
class GridEnv_HW2:
    def __init__(self, 
                 goal_location, 
                 action_stochasticity,
                 non_terminal_reward,
                 terminal_reward,
                 grey_in,
                 brown_in,
                 grey_out,
                 brown_out
                ):

        # Do not edit this section 
        self.action_stochasticity = action_stochasticity
        self.non_terminal_reward = non_terminal_reward
        self.terminal_reward = terminal_reward
        self.grid_size = [10, 10]

        # Index of the actions 
        self.actions = {'N': (1, 0), 
                        'E': (0,1),
                        'S': (-1,0), 
                        'W': (0,-1)}
        
        self.perpendicular_order = ['N', 'E', 'S', 'W']
        
        l = ['normal' for _ in range(self.grid_size[0]) ]
        self.grid = np.array([l for _ in range(self.grid_size[1]) ], dtype=object)

        self.grid[goal_location[0], goal_location[1]] = 'goal'
        self.goal_location = goal_location

        for gi in grey_in:
            self.grid[gi[0],gi[1]] = 'grey_in'
        for bi in brown_in:    
            self.grid[bi[0], bi[1]] = 'brown_in'

        for go in grey_out:    
            self.grid[go[0], go[1]] = 'grey_out'
        for bo in brown_out:    
            self.grid[bo[0], bo[1]] = 'brown_out'

        self.grey_outs = grey_out
        self.brown_outs = brown_out

    def _out_of_grid(self, state):
        if state[0] < 0 or state[1] < 0:
            return True
        elif state[0] > self.grid_size[0] - 1:
            return True
        elif state[1] > self.grid_size[1] - 1:
            return True
        else:
            return False

    def _grid_state(self, state):
        return self.grid[state[0], state[1]]        
        
    def get_transition_probabilites_and_reward(self, state, action):
        """ 
        Returns the probabiltity of all possible transitions for the given action in the form:
        A list of tuples of (next_state, probability, reward)
        Note that based on number of state and action there can be many different next states
        Unless the state is All the probabilities of next states should add up to 1
        """

        grid_state = self._grid_state(state)
        
        if grid_state == 'goal':
            return [(self.goal_location, 1.0, 0.0)]
        elif grid_state == 'grey_in':
            npr = []
            for go in self.grey_outs:
                npr.append((go, 1/len(self.grey_outs), 
                            self.non_terminal_reward))
            return npr
        elif grid_state == 'brown_in':
            npr = []
            for bo in self.brown_outs:
                npr.append((bo, 1/len(self.brown_outs), 
                            self.non_terminal_reward))
            return npr
        
        direction = self.actions.get(action, None)
        if direction is None:
            raise ValueError("Invalid action %s , please select among" % action, list(self.actions.keys()))

        dir_index = self.perpendicular_order.index(action)
        wrap_acts = self.perpendicular_order[dir_index:] + self.perpendicular_order[:dir_index]
        next_state_probs = {}
        for prob, a in zip(self.action_stochasticity, wrap_acts):
            d = self.actions[a]
            next_state = (state[0] + d[0]), (state[1] + d[1])
            if self._out_of_grid(next_state):
                next_state = state
            next_state_probs.setdefault(next_state, 0.0)
            next_state_probs[next_state] += prob

        npr = []
        for ns, prob in next_state_probs.items():
            next_grid_state = self._grid_state(ns)
            reward = self.terminal_reward if next_grid_state == 'goal' else self.non_terminal_reward
            npr.append((ns, prob, reward))

        return npr

    def step(self, state, action):
        npr = self.get_transition_probabilites_and_reward(state, action)
        probs = [t[1] for t in npr]
        sampled_idx = np.random.choice(range(len(npr)), p=probs)
        sampled_npr = npr[sampled_idx]
        next_state = sampled_npr[0]
        reward = sampled_npr[2]
        is_terminal = next_state == tuple(self.goal_location)
        return next_state, reward, is_terminal

def get_base_kwargs():
    goal_location = (9,9)
    action_stochasticity = [0.8, 0.2/3, 0.2/3, 0.2/3]
    grey_out = [(3,2), (4,2), (5,2), (6,2)]
    brown_in = [(9,7)]
    grey_in = [(0,0)]
    brown_out = [(1,7)]
    non_terminal_reward = 0
    terminal_reward = 10

    base_kwargs =  {"goal_location": goal_location, 
            "action_stochasticity": action_stochasticity,
            "brown_in": brown_in, 
            "grey_in": grey_in, 
            "brown_out": brown_out,
            "non_terminal_reward": non_terminal_reward,
            "terminal_reward": terminal_reward,
            "grey_out": grey_out,}
    
    return base_kwargs

base_kwargs = get_base_kwargs()

def value_iteration(env, gamma):
    # Initial Values
    values = np.zeros((10, 10))

    # Initial policy
    policy = np.empty((10, 10), object)
    policy[:] = 'N' # Make all the policy values as 'N'

    # Begin code here
    
    # Put your extra information needed for plots etc in this dictionary
    extra_info = {}
    n = values.shape[0]
    
    #initializing cost vector
    J = np.zeros((n*n,1))
    #Policy Vector
    pi_s = np.zeros((n*n,1))
    #Invoking all possible actions
    action_space = env.actions
    #Transtition Probability Matrix for each action
    transition_probs = {act:np.zeros((n*n,n*n)) for act in action_space.keys()}
    #Reward Matrix for each action
    transition_rewards = {act:np.zeros((n*n,n*n)) for act in action_space.keys()}
    #Assigning probabilities and rewards
    for action in action_space:
      for i in range(values.shape[0]):
        for j in range(values.shape[1]):
          state = (i,j)
          pseudo_state = i*values.shape[0]+j
          next = env.get_transition_probabilites_and_reward(state,action)
          for ele in next:
            next_state = ele[0]
            pseudo_next = next_state[0]*values.shape[0]+next_state[1]
            transition_probs[action][pseudo_state,pseudo_next] = ele[1]
            transition_rewards[action][pseudo_state,pseudo_next] = ele[2]
    #Setting Tolerance Level
    tol = 1e-8
    ite = 0
    delta = 0
    extra_info["Expected Rewards"] = [deepcopy(values)]
    extra_info["Policies"] = [deepcopy(policy)]

    #Convergence Criteria
    while delta>tol or ite == 0:
      ite += 1
      delta = 0
      J_new = np.zeros((n*n,1))
      policy_new = np.empty((n*n, 1), object)
      for i in range(len(J)):
        state = (i//n,i%n)
        j = J[i]
        MAX = -np.inf
        for action in action_space:
          r_a = transition_rewards[action][i,:].T
          P_a = transition_probs[action][i,:].T
          #Value Iteration
          tmp = np.dot(P_a,r_a+gamma*J.ravel())
          # Maximum J(s) across all actions
          if tmp > MAX:
            MAX = tmp
            J_new[i] = tmp
            best_action = action
        policy_new[i] = best_action
        delta = max(delta,abs(j-MAX))
      #Updating J
      J = deepcopy(J_new)
      policy = policy_new.reshape((n,n))
      values = J.reshape((n,n))
      extra_info["Expected Rewards"].append(deepcopy(values))
      extra_info["Policies"].append(deepcopy(policy))
    # End code

    # Do not change the number of output values
    return {"Values": values, "Policy": policy}, extra_info

env = GridEnv_HW2(**base_kwargs)
res, extra_info = value_iteration(env, 0.7)

 # The rounding off is just for making print statement cleaner
print(np.flipud(np.round(res['Values'], decimals=2)))
print(np.flipud(res['Policy']))

def policy_iteration(env, gamma):
    # Initial Values
    values = np.zeros((10, 10))

    # Initial policy
    policy = np.empty((10, 10), object)
    policy[:] = 'N' # Make all the policy values as 'N'

    # Begin code here   
    
    # Put your extra information needed for plots etc in this dictionary
    extra_info = {}
    n = values.shape[0]

    #initializing cost vector
    J = np.zeros((n*n,1))
    #Policy Vector
    pi_s = np.zeros((n*n,1))
    #Invoking all possible actions
    action_space = env.actions
    #Transtition Probability Matrix for each action
    transition_probs = {act:np.zeros((n*n,n*n)) for act in action_space.keys()}
    #Reward Matrix for each action
    transition_rewards = {act:np.zeros((n*n,n*n)) for act in action_space.keys()}
    #Assigning probabilities and rewards
    for action in action_space:
      for i in range(values.shape[0]):
        for j in range(values.shape[1]):
          state = (i,j)
          pseudo_state = i*values.shape[0]+j
          next = env.get_transition_probabilites_and_reward(state,action)
          for ele in next:
            next_state = ele[0]
            pseudo_next = next_state[0]*values.shape[0]+next_state[1]
            transition_probs[action][pseudo_state,pseudo_next] = ele[1]
            transition_rewards[action][pseudo_state,pseudo_next] = ele[2]
    
    done = 0
    tol = 1e-8
    ite1 = 0
    extra_info["Expected Rewards"] = [deepcopy(values)]
    extra_info["Policies"] = [deepcopy(policy)]
    while done == 0:
      ite1 += 1
      ite2 = 0
      delta = 0
      while delta > tol or ite2 == 0:
        ite2 += 1
        delta = 0
        tmp = np.zeros((n*n,1))
        for i in range(len(J)):
          state = (i//n,i%n)
          j = J[i]
          act = policy[state[0],state[1]]
          r_a = transition_rewards[act][i,:].T
          P_a = transition_probs[act][i,:].T
          tmp[i] = np.dot(P_a,r_a+gamma*J.ravel())
          delta = max(delta,abs(j-tmp[i]))
        J = deepcopy(tmp)
        values = J.reshape((n,n))

      done = 1

      policy_old = deepcopy(policy)
      for i in range(len(J)):
          state = (i//n,i%n)
          MAX = -np.inf
          for act in action_space:
            r_a = transition_rewards[act][i,:].T
            P_a = transition_probs[act][i,:].T
            J_new = np.dot(P_a,r_a+gamma*J.ravel())
            if J_new > MAX:
              MAX = J_new
              best_act = act
            # J_vals.append(J_new)
          policy[state[0],state[1]] = best_act
          if best_act!= policy_old[state[0],state[1]]:
            done = 0
      extra_info["Expected Rewards"].append(deepcopy(values))
      extra_info["Policies"].append(deepcopy(policy))

    # End code

    # Do not change the number of output values
    return {"Values": values, "Policy": policy}, extra_info

env = GridEnv_HW2(**base_kwargs)
res, extra_info = policy_iteration(env, 0.7)

 # The rounding off is just for making print statement cleaner
print(np.flipud(np.round(res['Values'], decimals=2)))
print(np.flipud(res['Policy']))

# The policy mentioned in the pdf to be used for TD lambda, do not modify this
def heuristic_policy(env, state):
    goal = env.goal_location
    dx = goal[0] - state[0]
    dy = goal[1] - state[1]
    if abs(dx) >= abs(dy):
        direction = (np.sign(dx), 0)
    else:
        direction = (0, np.sign(dy))
    for action, dir_val in env.actions.items():
        if dir_val == direction:
            target_action = action
            break
    return target_action

def td_lambda(env, lamda, seeds):
    alpha = 0.5
    gamma = 0.7
    N = len(seeds)
    # Usage of input_policy
    # heuristic_policy(env, state) -> action
    example_action = heuristic_policy(env, (1,2)) # Returns 'N' if goal is (9,9)

    # Example of env.step
    # env.step(state, action) -> Returns next_state, reward, is_terminal

    # Initial values
    values = np.zeros((10, 10))
    es = np.zeros((10,10))
    n = values.shape[0]
    extra_info = {}
    extra_info['Reward'] = [deepcopy(values)]
    for episode_idx in range(N):
         # Do not change this else the results will not match due to environment stochas
        np.random.seed(seeds[episode_idx])
        grey_in_loc = np.where(env.grid == 'grey_in')
        state = grey_in_loc[0][0], grey_in_loc[1][0]
        done = False
        while not done:
            action = heuristic_policy(env, state)
            ns, rew, is_terminal = env.step(state, action) 
            # env.step is already taken inside the loop for you, 
            # Don't use env.step anywhere else in your code

            # Begin code here
            delta = rew + gamma*values[ns[0],ns[1]] - values[state[0],state[1]]
            es[state[0],state[1]] += 1 
            values = values + alpha*delta*es
            es = gamma*lamda*es
            state = deepcopy(ns)
            if is_terminal:
              done = True
            extra_info['Reward'].append(deepcopy(values))

    # End code

    # Do not change the number of output values
    return {"Values": values}, extra_info

env = GridEnv_HW2(**base_kwargs)
res, extra_info = td_lambda(env, lamda=0.5, seeds=np.arange(1000))

 # The rounding off is just for making print statement cleaner
print(np.flipud(np.round(res['Values'], decimals=2)))

# This cell is only for your subjective evaluation results, display the results as asked in the pdf
# You can change it as you require, this code should run TD lamdba by default for different values of lambda

lamda_values = np.arange(0, 100+5, 5)/100
td_lamda_results = {}
extra_info_TD = {}
for lamda in lamda_values:
    env = GridEnv_HW2(**base_kwargs)
    td_lamda_results[lamda], extra_info_TD[lamda] = td_lambda(env, lamda,
                                                           seeds=np.arange(1000))

def get_results(kwargs):

    gridenv = GridEnv_HW2(**kwargs)

    policy_iteration_results = policy_iteration(gridenv, 0.7)[0]
    value_iteration_results = value_iteration(gridenv, 0.7)[0]
    td_lambda_results = td_lambda(env, 0.5, np.arange(1000))[0]

    final_results = {}
    final_results["policy_iteration"] = policy_iteration_results
    final_results["value_iteration"] = value_iteration_results
    final_results["td_lambda"] = td_lambda_results

    return final_results

# Do not edit this cell, generate results with it as is
if not os.path.exists(AICROWD_RESULTS_DIR):
    os.mkdir(AICROWD_RESULTS_DIR)

for params_file in os.listdir(DATASET_DIR):
  kwargs = np.load(os.path.join(DATASET_DIR, params_file), allow_pickle=True).item()
  results = get_results(kwargs)
  idx = params_file.split('_')[-1][:-4]
  np.save(os.path.join(AICROWD_RESULTS_DIR, 'results_' + idx), results)

# Check your score on the given test cases (There are more private test cases not provided)
target_folder = 'targets'
result_folder = AICROWD_RESULTS_DIR

def check_algo_match(results, targets):
    if 'Policy' in results:
        policy_match = results['Policy'] == targets['Policy']
    else:
        policy_match = True
    # Reference https://numpy.org/doc/stable/reference/generated/numpy.allclose.html
    rewards_match = np.allclose(results['Values'], targets['Values'], rtol=3)
    equal = rewards_match and policy_match
    return equal

def check_score(target_folder, result_folder):
    match = []
    for out_file in os.listdir(result_folder):
        res_file = os.path.join(result_folder, out_file)
        results = np.load(res_file, allow_pickle=True).item()
        idx = out_file.split('_')[-1][:-4]  # Extract the file number
        target_file = os.path.join(target_folder, f"targets_{idx}.npy")
        targets = np.load(target_file, allow_pickle=True).item()
        algo_match = []
        for k in targets:
            algo_results = results[k]
            algo_targets = targets[k]
            algo_match.append(check_algo_match(algo_results, algo_targets))
        match.append(np.mean(algo_match))
    return np.mean(match)

if os.path.exists(target_folder):
    print("Shared data Score (normalized to 1):", check_score(target_folder, result_folder))

pd.set_option('display.max_colwidth', None)
pd.DataFrame(td_lamda_results).T

env = GridEnv_HW2(**base_kwargs)
val_res, extra_info_val = value_iteration(env, 0.7)
pol_res, extra_info_pol = policy_iteration(env, 0.7)

brown_in_loc = np.where(env.grid == 'brown_in')
state = brown_in_loc[0][0], brown_in_loc[1][0]
value_iter_vals = [extra_info_val['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_val['Expected Rewards']))]
x_val = range(len(extra_info_val['Expected Rewards']))
pol_iter_vals = [extra_info_pol['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_pol['Expected Rewards']))]
x_pol = range(len(extra_info_pol['Expected Rewards']))
plt.figure(figsize = (8,6))
plt.plot(x_val,value_iter_vals,label = "Value Iteration")
plt.plot(x_pol,pol_iter_vals,label = "Policy Iteration")
plt.legend(loc = "best")
plt.xlabel("iteration")
plt.ylabel("Value")
plt.title("Value Convergence of Value and Policy Iteration for state Brown In")

brown_out_loc = np.where(env.grid == 'brown_out')
state = brown_out_loc[0][0], brown_out_loc[1][0]
value_iter_vals = [extra_info_val['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_val['Expected Rewards']))]
x_val = range(len(extra_info_val['Expected Rewards']))
pol_iter_vals = [extra_info_pol['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_pol['Expected Rewards']))]
x_pol = range(len(extra_info_pol['Expected Rewards']))
plt.figure(figsize = (8,6))
plt.plot(x_val,value_iter_vals,label = "Value Iteration")
plt.plot(x_pol,pol_iter_vals,label = "Policy Iteration")
plt.legend(loc = "best")
plt.xlabel("iteration")
plt.ylabel("Value")
plt.title("Value Convergence of Value and Policy Iteration for state Brown Out")

grey_out_loc = np.where(env.grid == 'grey_out')
state = grey_out_loc[0][0], grey_out_loc[1][0]
value_iter_vals = [extra_info_val['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_val['Expected Rewards']))]
x_val = range(len(extra_info_val['Expected Rewards']))
pol_iter_vals = [extra_info_pol['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_pol['Expected Rewards']))]
x_pol = range(len(extra_info_pol['Expected Rewards']))
plt.figure(figsize = (8,6))
plt.plot(x_val,value_iter_vals,label = "Value Iteration")
plt.plot(x_pol,pol_iter_vals,label = "Policy Iteration")
plt.legend(loc = "best")
plt.xlabel("iteration")
plt.ylabel("Value")
plt.title("Value Convergence of Value and Policy Iteration for state Grey Out")

grey_in_loc = np.where(env.grid == 'grey_in')
state = grey_in_loc[0][0], grey_in_loc[1][0]
value_iter_vals = [extra_info_val['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_val['Expected Rewards']))]
x_val = range(len(extra_info_val['Expected Rewards']))
pol_iter_vals = [extra_info_pol['Expected Rewards'][i][state[0],state[1]] for i in range(len(extra_info_pol['Expected Rewards']))]
x_pol = range(len(extra_info_pol['Expected Rewards']))
plt.figure(figsize = (8,6))
plt.plot(x_val,value_iter_vals,label = "Value Iteration")
plt.plot(x_pol,pol_iter_vals,label = "Policy Iteration")
plt.legend(loc = "best")
plt.xlabel("iteration")
plt.ylabel("Value")
plt.title("Value Convergence of Value and Policy Iteration for state Grey In")

errors = []
for lamda in td_lamda_results:
  errors.append(np.sqrt(np.sum((td_lamda_results[lamda]['Values'] - val_res['Values'])**2)/100))
lamda_vals = np.sort(list(td_lamda_results.keys()))
plt.figure(figsize = (8,6))
plt.plot(lamda_vals,errors)
plt.xlabel("lambda")
plt.ylabel("Error Standard Deviation")
plt.title("Error vs Lambda of TD(lambda)")

errors = []
extra_info_pol['Expected Rewards']
for i in range(len(extra_info_pol['Expected Rewards'])):
  errors.append(np.sqrt(np.sum((extra_info_pol['Expected Rewards'][i] - val_res['Values'])**2)/100))
x_pol = range(len(extra_info_pol['Expected Rewards']))
plt.figure(figsize = (8,6))
plt.plot(x_pol,errors)
plt.xlabel("iteration")
plt.ylabel("Error Standard Deviation")
plt.title("Error vs Iteration of Policy Iteration")

for lamda in [0, 0.25, 0.5, 0.75, 1]:
  errors = []
  for i in range(len(extra_info_TD[lamda]["Reward"])):
    errors.append(np.sqrt(np.sum((extra_info_TD[lamda]["Reward"][i] - val_res['Values'])**2)/100))
  plt.figure(figsize = (8,6))
  plt.plot(range(len(extra_info_TD[lamda]["Reward"])),errors,label = "lambda = "+str(lamda))
  plt.xlabel("iteration")
  plt.ylabel("Error Standard Deviation")
  plt.legend(loc = "best")
  plt.title("Error vs Iteration of TD("+str(lamda)+") Iteration")

!DATASET_PATH=$AICROWD_DATASET_PATH aicrowd notebook submit --no-verify -c iit-m-rl-assignment-2-gridworld -a assets