def step()

in mbrl/env/humanoid_truncated_obs.py [0:0]

• 24 lines of code
• 2 McCabe index (conditional complexity)

    def step(self, a):
        pos_before = mass_center(self.model, self.sim)
        self.do_simulation(a, self.frame_skip)
        pos_after = mass_center(self.model, self.sim)
        alive_bonus = 5.0
        data = self.sim.data
        lin_vel_cost = 0.25 * (pos_after - pos_before) / self.model.opt.timestep
        quad_ctrl_cost = 0.1 * np.square(data.ctrl).sum()
        quad_impact_cost = 0.5e-6 * np.square(data.cfrc_ext).sum()
        quad_impact_cost = min(quad_impact_cost, 10)
        reward = lin_vel_cost - quad_ctrl_cost - quad_impact_cost + alive_bonus
        qpos = self.sim.data.qpos
        done = bool((qpos[2] < 1.0) or (qpos[2] > 2.0))
        return (
            self._get_obs(),
            reward,
            done,
            dict(
                reward_linvel=lin_vel_cost,
                reward_quadctrl=-quad_ctrl_cost,
                reward_alive=alive_bonus,
                reward_impact=-quad_impact_cost,
            ),
        )