!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
import copy
import matplotlib.pyplot as plt

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'
    delta_matrix = np.full((10,10), 1e-8)

    # Begin code here
    H = np.zeros((10,10))
    t_policy = np.empty((10, 10), object)
    t_policy[:] = 'N' # Make all the policy values as 'N'
    loop_assertion = False
    iterations = 0
    brown_in_cost = []
    brown_out_cost = []
    grey_in_cost = []
    while loop_assertion == False:
      delta = np.zeros((10,10))
      for i in range(10):
        for j in range(10):
          state = (i,j)
          possible_actions = env.actions
          costs = np.zeros(len(possible_actions))
          k = 0
          for action in possible_actions:
            next_states_probs_rews = env.get_transition_probabilites_and_reward(state, action)
            # print(next_states_probs_rews)
            for next_state in next_states_probs_rews:
              ns = next_state[0]
              costs[k] = costs[k] + next_state[1] * (next_state[2] + gamma * H[ns[0]][ns[1]])
            k = k+1
          H[i][j] = np.max(costs)
          # print(costs)
          # print(env.perpendicular_order[np.argmax(costs)])
          t_policy[i][j] = env.perpendicular_order[np.argmax(costs)]
          delta[i][j] = np.max([delta[i][j],np.abs(values[i][j] - H[i][j])])
          if env.grid[i,j] == 'brown_in':
            brown_in_cost.append(H[i][j])
          if env.grid[i,j] == 'brown_out':
            brown_out_cost.append(H[i][j])
          if env.grid[i,j] == 'grey_in':
            grey_in_cost.append(H[i][j])

      for i in range(10):
        for j in range(10):
          values[i][j] = H[i][j]
          policy[i][j] = t_policy[i][j]
      less_assertion = delta < delta_matrix
      iterations = iterations + 1
      if np.all(less_assertion) == True:
        loop_assertion = True
        break  
    
    # Put your extra information needed for plots etc in this dictionary
    extra_info = {"iterations": iterations, "brown_in_cost": np.array(brown_in_cost), "brown_out_cost": np.array(brown_out_cost), "grey_in_cost":np.array(grey_in_cost)}

    # 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)

res1 = res

 # 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 
    delta_matrix = np.full((10,10),1e-8)
    loop_assertion_1 = False
    iterations = 0
    brown_in_cost = []
    brown_out_cost = []
    grey_in_cost = []
    values_list = []
    while loop_assertion_1 == False:
      loop_assertion_2 = False
      while loop_assertion_2 == False:
        delta = np.zeros((10,10))
        #print('entered while loop')
        for i in range(10):
          for j in range(10):
            state = (i,j)
            temp_j = values[i][j]
            possible_actions = env.actions
            action = policy[i][j]
            next_states_probs_rews = env.get_transition_probabilites_and_reward(state, action)
            #print(next_states_probs_rews)
            sum = 0
            for next_state in next_states_probs_rews:
              ns = next_state[0]
              sum = sum + next_state[1] * (next_state[2] + gamma * values[ns[0]][ns[1]])
            delta[i][j] = np.max([delta[i][j], np.abs(temp_j-sum)])
            values[i][j] = sum
        less_assertion = delta < delta_matrix
        if np.all(less_assertion) == True:
          loop_assertion_2 = True
          break
      values_list.append(copy.deepcopy(values))
      done = 1
      #print('first while done')
      for i in range(10):
        for j in range(10):
          state = (i,j)
          b = policy[i][j]
          possible_actions = env.actions
          costs = np.zeros(len(possible_actions))
          k = 0
          for action in possible_actions:
            next_states_probs_rews = env.get_transition_probabilites_and_reward(state, action)
            for next_state in next_states_probs_rews:
              ns = next_state[0]
              costs[k] = costs[k] + next_state[1] * (next_state[2] + gamma * values[ns[0]][ns[1]])
            k = k+1
          policy[i][j] = env.perpendicular_order[np.argmax(costs)]
          if b != policy[i][j]:
            done = 0

          if env.grid[i,j] == 'brown_in':
            brown_in_cost.append(values[i][j])
          if env.grid[i,j] == 'brown_out':
            brown_out_cost.append(values[i][j])
          if env.grid[i,j] == 'grey_in':
            grey_in_cost.append(values[i][j])
      iterations = iterations + 1
      if done == 1:
        loop_assertion_1 = True
        break
    
    # Put your extra information needed for plots etc in this dictionary
    extra_info = {"iterations": iterations, "brown_in_cost": np.array(brown_in_cost), "brown_out_cost": np.array(brown_out_cost), "grey_in_cost":np.array(grey_in_cost),"values_list": values_list}

    # 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))

    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]
      #print('state=',state)
      loop_assertion = False
      values_list = []
      iterations = 0
      while loop_assertion == False:
        action = heuristic_policy(env, state)
        ns, rew, is_terminal = env.step(state, action) 
        #print(rew)
        # 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]] = es[state[0]][state[1]] + 1
        #print(es[state[0]][state[1]])
        for i in range(10):
          for j in range(10):
            values[i][j] = values[i][j] + (alpha * delta * es[i][j])
            es[i][j] = gamma * lamda * es[i][j]
        state = ns[0],ns[1]
        if_terminal = state == tuple(env.goal_location) 
        values_list.append(copy.deepcopy(values))
        iterations = iterations + 1
        if if_terminal==True:
          loop_assertion = True
          break 
      #print(episode_idx)  

    # Put your extra information needed for plots etc in this dictionary
    extra_info = {"values_list":values_list, "iterations": iterations}
    # 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)))

# # Testing
# np_load_old = np.load
# # # modify the default parameters of np.load
# np.load = lambda *a,**k: np_load_old(*a, allow_pickle=True, **k)
# target_data = np.load('/content/targets/targets_2.npy')
# np.load = np_load_old
# print(target_data.item().get('td_lambda'))

# # Testing
# np_load_old = np.load
# # # modify the default parameters of np.load
# np.load = lambda *a,**k: np_load_old(*a, allow_pickle=True, **k)
# data = np.load('/content/inputs/inputs_0.npy')
# np.load = np_load_old
# env = GridEnv_HW2(data.item().get('goal_location'), 
#                  data.item().get('action_stochasticity'),
#                  data.item().get('non_terminal_reward'),
#                  data.item().get('terminal_reward'),
#                  data.item().get('grey_in'),
#                  data.item().get('brown_in'),
#                  data.item().get('grey_out'),
#                  data.item().get('brown_out')
#                 )
# result,extras = td_lambda(env, lamda=0.5, seeds=np.arange(1000))
# print(result)

# 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 = {}
for lamda in lamda_values:
    env = GridEnv_HW2(**base_kwargs)
    td_lamda_results[lamda], extra_info[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))

gridenv = GridEnv_HW2(**base_kwargs)

extra_info_2 = policy_iteration(gridenv, 0.7)[1]
extra_info_1 = value_iteration(gridenv, 0.7)[1]

iterations_1 = range(extra_info_1['iterations'])
grey_in_cost_1 = extra_info_1['grey_in_cost']
brown_out_cost_1 = extra_info_1['brown_out_cost']
brown_in_cost_1 = extra_info_1['brown_in_cost']
plt.plot(iterations_1, grey_in_cost_1)
plt.xlabel('iterations ')
plt.ylabel('state grey_in cost values')
plt.title('Value iteration grey_in cost values')
plt.show()

plt.plot(iterations_1, brown_in_cost_1)
plt.xlabel('iterations')
plt.ylabel('state brown_in cost values')
plt.title('Value iteration brown_in cost values')
plt.show()

plt.plot(iterations_1, brown_out_cost_1)
plt.xlabel('iterations')
plt.ylabel('state brown_out cost values')
plt.title('Value iteration brown_out cost values')
plt.show()

iterations_2 = range(extra_info_2['iterations'])
grey_in_cost_2 = extra_info_2['grey_in_cost']
brown_out_cost_2 = extra_info_2['brown_out_cost']
brown_in_cost_2 = extra_info_2['brown_in_cost']


plt.plot(iterations_2, grey_in_cost_2)
plt.xlabel('iterations')
plt.ylabel('state grey_in cost values')
plt.title('Policy iteration grey_in cost values')
plt.show()

plt.plot(iterations_2, brown_in_cost_2)
plt.xlabel('iterations')
plt.ylabel('state brown_in cost values')
plt.title('Policy iteration brown_in cost values')
plt.show()

plt.plot(iterations_2, brown_out_cost_2)
plt.xlabel('iterations')
plt.ylabel('state brown_out cost values')
plt.title('Policy iteration brown_out cost values')
plt.show()

gridenv = GridEnv_HW2(**base_kwargs)
result_p, extras_p = policy_iteration(gridenv, 0.7)
result_v, extras_v = value_iteration(gridenv, 0.7)

vv = result_v['Values']
val_l = extras_p['values_list']
iterations = extras_p['iterations']
error = np.zeros(iterations)
#print(val_l)

for i in range(iterations):
  val_lp = val_l[i]  
  diff_square = 0
  for row in range(10):
    for column in range(10):
      diff_square = diff_square + (val_lp[row][column] - vv[row][column]) ** 2
  error[i] =   np.sqrt(diff_square/10)

plt.plot(np.arange(iterations), error)
plt.xlabel('number of iterations')
plt.ylabel('error')
plt.title('Error plot of policy iteration')

gridenv = GridEnv_HW2(**base_kwargs)

results_v, extras_v = value_iteration(gridenv, 0.7)

for i in [0, 0.25, 0.5, 0.75, 1]:
  results_t, extras_t = td_lambda(env, i, np.arange(1000))
  vv = results_v['Values']
  val_l = extras_t['values_list']
  iterations = extras_t['iterations']
  error = np.zeros(iterations)

  for it in range(iterations):
    val_lp = val_l[it]  
    diff_square = 0
    for row in range(10):
      for column in range(10):
        diff_square = diff_square + (val_lp[row][column] - vv[row][column]) ** 2
    error[it] =   np.sqrt(diff_square/10)

  plt.plot(np.arange(iterations), error)
  plt.xlabel('number of iterations')
  plt.ylabel('error')

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