ecc8b0b34138fa480e1b1876fd2695c07be9b159,simple_rl/tasks/navigation/NavigationMDP.py,NavigationMDP,reset_goals,#NavigationMDP#,179

Before Change

                // Assume all goals are same
                self.cells[goal_r, goal_c] = len(self.cell_types)
                self.cell_rewards[goal_r, goal_c] = self.goal_rewards[0]
                self.goal_xy_to_idx[(goal_loc[0], goal_loc[1])] = 0
            else:
                // Each goal is different in type and rewards
                self.cells[goal_r, goal_c] = len(self.cell_types) + goal_idx
                self.cell_rewards[goal_r, goal_c] = self.goal_rewards[goal_idx]
                self.goal_xy_to_idx[(goal_loc[0], goal_loc[1])] = goal_idx
        self._policy_invalidated = True

    def _reward_func(self, state, action):

After Change

        self.goal_cell_locs = goal_cell_locs
        self.goal_cell_rewards = goal_cell_rewards
        self.goal_cell_types = goal_types
        self.goal_cell_ids = list(range(self.living_cell_ids[-1] + 1,
                                        self.living_cell_ids[-1] + 1 + len(
                                            self.goal_cell_locs)))
        // Reset goal xy to idx dict
        self.goal_xy_to_idx = {}
        // Reset cell type and cell reward grid with no goals
        self.state_space = self.state_space_wo_goals.copy()
        self.state_rewards = self.state_rewards_wo_goals.copy()

        // Update goals and their rewards
        for idx, goal_loc in enumerate(self.goal_cell_locs):
            goal_r, goal_c = self._xy_to_rowcol(goal_loc[0], goal_loc[1])
            self.state_space[goal_r, goal_c] = self.goal_cell_ids[idx]
            self.state_rewards[goal_r, goal_c] = self.goal_cell_rewards[idx]
            self.goal_xy_to_idx[(goal_loc[0], goal_loc[1])] = idx
        self.cell_ids = self.living_cell_ids + self.goal_cell_ids
        self.cell_types = self.living_cell_types + self.goal_cell_types
        self.cell_type_rewards = self.living_cell_rewards + self.goal_cell_rewards
        self._policy_invalidated = True

Instances