Method _state

1-grid-world/env.py:203–210

(self)

        return self._state(), reward, done

    def _state(self):
        ax, ay = self.agent
        s = []
        for o in self.obstacles:
            ox, oy = o["state"]
            s += [ox - ax, oy - ay, -1, o["direction"]]
        s += [self.goal[0] - ax, self.goal[1] - ay, 1]
        return s

    def render(self):
        if self.render_mode is None: return