INNER CODE UNIT · Python
value_iteration
rlcode/reinforcement-learning · 1-grid-world/2-value_iteration.py:12
def value_iteration(self):
next_value_table = [[0.0] * self.env.width
for _ in range(self.env.height)]
for state in self.env.get_all_states():
if state == [2, 2]:
next_value_table[state[0]][state[1]] = 0.0
continue
value_list = []
for action in self.env.possible_actions:
next_state = self.env.state_after_action(state, action)
reward = self.env.get_reward(state, action)
next_value = self.get_value(next_state)
value_list.append((reward + self.discount_factor * next_value))
# return the maximum value(it is the optimality equation!!)
next_value_table[state[0]][state[1]] = round(max(value_list), 2)
self.value_table = next_value_table