!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 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'

    # Begin code here
    
    states = list(np.ndindex(10,10))

    Js = values
    pi_s = policy
    delta = 1

    Js_brown_in = [0]
    Js_brown_out = [0]
    Js_grey_in = [0]

    while(delta > 1e-8):
      delta = 0
      j = Js.copy()
      for s in states:
        eval_list = {}

        for action in env.perpendicular_order[::-1]:
          J_eval = 0
          for (next_state, probability, reward) in env.get_transition_probabilites_and_reward(s,action):
            J_eval += probability*(reward+gamma*j[next_state[0],next_state[1]])
          eval_list[action]=J_eval

        Js[s[0],s[1]] = max(eval_list.values())
        eval_list_r = dict(map(reversed, eval_list.items()))
        pi_s[s[0],s[1]] = eval_list_r[Js[s[0],s[1]]]
        delta = max(delta,abs(j[s[0],s[1]]-Js[s[0],s[1]]))

      Js_brown_in.append(Js[np.where(env.grid == 'brown_in')][0])
      Js_brown_out.append(Js[np.where(env.grid == 'brown_out')][0])
      Js_grey_in.append(Js[np.where(env.grid == 'grey_in')][0])
 
    # Put your extra information needed for plots etc in this dictionary

    extra_info = {'brown_in':Js_brown_in,'brown_out':Js_brown_out,'grey_in':Js_grey_in}

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

env = GridEnv_HW2(**base_kwargs)
res_vi, extra_info_vi = value_iteration(env, 0.7)

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

    states = list(np.ndindex(10,10))

    Js = values
    pi_s = policy
    done = 0
     
    Js_list=[]
    Js_list.append(Js.copy())

    Js_brown_in = [0]
    Js_brown_out = [0]
    Js_grey_in = [0]

    while(done!=1):
      
      flag = True
      while(flag):
        j = Js.copy()
        delta = 0
        for s in states:
          J_eval = 0
          for (next_state, probability, reward) in env.get_transition_probabilites_and_reward(s,pi_s[s]):
            J_eval += probability*(reward+gamma*j[next_state[0],next_state[1]])
          Js[s] = J_eval
          delta = max(delta,abs(j[s]-Js[s]))
        if delta < 1e-8:          
          flag = False

      done = 1

      j = Js.copy()
      for st in states:
        b = pi_s[st]
        eval_list = {}
        for action in env.perpendicular_order[::-1]:
          J_eval = 0
          for (next_state, probability, reward) in env.get_transition_probabilites_and_reward(st,action):
            J_eval += probability*(reward+gamma*j[next_state[0],next_state[1]])
          eval_list[action]=J_eval

        eval_list_r = dict(map(reversed, eval_list.items()))
        pi_s[st] = eval_list_r[max(eval_list.values())]

        if (b!=pi_s[st]):
          done = 0

      Js_brown_in.append(Js[np.where(env.grid == 'brown_in')][0])
      Js_brown_out.append(Js[np.where(env.grid == 'brown_out')][0])
      Js_grey_in.append(Js[np.where(env.grid == 'grey_in')][0])
      Js_list.append(Js.copy())

    # Put your extra information needed for plots etc in this dictionary
    extra_info = {'brown_in':Js_brown_in,'brown_out':Js_brown_out,'grey_in':Js_grey_in,'Values':Js_list}

    # End code

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

env = GridEnv_HW2(**base_kwargs)
res_pi, extra_info_pi = policy_iteration(env, 0.7)

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

    states = list(np.ndindex(10,10))

    extra_info = []

    Js = 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*Js[ns]-Js[state]
            es[state] = es[state]+1

            for s in states:
              Js[s] = Js[s] + alpha*delta*es[s]
              es[s] = gamma*lamda*es[s]

            state = ns

            if(is_terminal): break

        extra_info.append(Js.copy())

         
    # Put your extra information needed for plots etc in this dictionary
    # extra_info =  {}                                                

    # 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))

#print(td_lamda_results)

fig, ax = plt.subplots(nrows=5,ncols=4,figsize=(20,25))
count=0
cmap = plt.cm.get_cmap('Oranges', 100)

for row in ax:
  for col in row:
      v = td_lamda_results[lamda_values[count]]['Values']
      norm = plt.Normalize(np.min(v),np.max(v))
      rgba = cmap(norm(v))
      im = col.imshow(rgba, interpolation='nearest')

      for i in range(v.shape[0]):
          for j in range(v.shape[1]):
              c = 'black'
              text = col.text(j, i, np.round(v[i, j], 1), ha="center", va="center", color=c)

      col.set_title(f'Value plot J(s) lambda={lamda_values[count]}')
      count+=1

plt.show()

fig, ax = plt.subplots(nrows=1,ncols=2,figsize=(22,8))

ax[0].plot(extra_info_vi['brown_in'],label='brown_in')
ax[0].plot(extra_info_vi['brown_out'],label='brown_out')
ax[0].plot(extra_info_vi['grey_in'],label='grey_in')
ax[0].set_title("J (s) vs iterations for Value iteration")
ax[0].legend()

plt.xlabel("iterations")
plt.ylabel("J(s)")

ax[1].plot(extra_info_pi['brown_in'],label='brown_in')
ax[1].plot(extra_info_pi['brown_out'],label='brown_out')
ax[1].plot(extra_info_pi['grey_in'],label='grey_in')
ax[1].set_title("J (s) vs iterations for Policy iteration")
ax[1].legend()

plt.xlabel("iterations")
plt.ylabel("J(s)")
plt.show()

fig, ax = plt.subplots(nrows=4,ncols=5,figsize=(32,15))

count = 0
for rows in ax:
  for cols in rows:
    diff = np.array(extra_info_td[lamda_values[count]]-res_vi['Values'])
    pts = np.linalg.norm(diff/1000,axis=(1,2))
    cols.plot(pts)
    cols.set_title(f'error(J) vs iterations for TD lambda = {lamda_values[count]}.')
    count+=1


plt.xlabel("iterations")
plt.ylabel("error(J)")
plt.show()

# Plot of error(J) with changing lambda
pts=[]
for lam in lamda_values:
  diff = np.array(td_lamda_results[lam]['Values']-res_vi['Values'])
  pts.append(np.linalg.norm(diff/1000))
plt.plot(lamda_values,pts,marker='.')  
plt.xlabel("lambda")
plt.ylabel("error(J)")
plt.title("RMS error(J) vs lambda")
plt.show()
print(f'Minimum error is {min(pts)} at lambda = {lamda_values[np.argmin(pts)]}')

extra_info_pi['Values']
diff = np.array(extra_info_pi['Values']-res_vi['Values'])
pts = np.linalg.norm(diff/len(extra_info_pi['Values']),axis=(1,2))
plt.plot(pts,marker='.')
plt.xlabel("iterations")
plt.ylabel("error(J)")
plt.title("RMS error(J) vs iterations for Policy Iteration")
plt.show()

fig, ax = plt.subplots(nrows=1,ncols=5,figsize=(32,5))
la =[0,0.25,0.5,0.75,1]
count = 0
for rows in ax:
    diff = np.array(extra_info_td[la[count]]-res_vi['Values'])
    pts = np.linalg.norm(diff/1000,axis=(1,2))
    rows.plot(pts)
    rows.set_title(f'error(J) vs iterations for TD lambda = {la[count]}.')
    count+=1

plt.xlabel("iterations")
plt.ylabel("error(J)")
plt.show()

#!DATASET_PATH=$AICROWD_DATASET_PATH aicrowd notebook submit -c rl-iitm-assignment2-gridworld -a assets
!DATASET_PATH=$AICROWD_DATASET_PATH aicrowd notebook submit --no-verify -c iit-m-rl-assignment-2-gridworld -a assets