16e9d8fe38d7d3b7edaf20fb5fd3413e7f865c93,slm_lab/env/__init__.py,UnityEnv,reset,#UnityEnv#,205

Before Change

            env_info_a = self.get_env_info(env_info_dict, a)
            self.check_u_agent_to_body(env_info_a, a)
            state_e[(a, b)] = env_info_a.states[b]
        return state_e

    def step(self, action_e):
        action_e = util.flatten_nonan(action_e)
        env_info_dict = self.u_env.step(action_e)

After Change

        return env_info_a

    def reset(self):
        self.done = False
        env_info_dict = self.u_env.reset(
            train_mode=self.train_mode, config=self.spec.get("unity"))
        _reward_e = self.data_spaces["reward"].init_data_s(e=self.e)
        state_e = self.data_spaces["state"].init_data_s(e=self.e)
        _done_e = self.data_spaces["done"].init_data_s(e=self.e)
        for (a, b), body in util.ndenumerate_nonan(self.body_e):
            env_info_a = self.get_env_info(env_info_dict, a)
            self.check_u_agent_to_body(env_info_a, a)
            state_e[(a, b)] = env_info_a.states[b]
        return _reward_e, state_e, _done_e

    def step(self, action_e):
        // TODO implement clock_speed: step only if self.clock.to_step()
        if self.done:

Instances