lit_sb3.wrappers.delayed_reward
[docs]
module
lit_sb3.wrappers.delayed_reward
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 | from __future__ import annotations
from typing import Any
import gymnasium as gym
from lit_sb3.core.types import GymResetReturn, GymStepReturn
class DelayedRewardWrapper(gym.Wrapper):
"""Delay the reward by ``delay`` steps (accumulated until released).
Makes the task harder but more realistic.
:param env: Gym environment.
:param delay: Number of steps to delay the reward.
"""
def __init__(self, env: gym.Env, delay: int = 10) -> None:
super().__init__(env)
self.delay = delay
self.current_step = 0
self.accumulated_reward = 0.0
def reset(self, seed: int | None = None, options: dict | None = None) -> GymResetReturn:
self.current_step = 0
self.accumulated_reward = 0.0
assert options is None, "Options not supported for now"
return self.env.reset(seed=seed)
def step(self, action: Any) -> GymStepReturn:
obs, reward, terminated, truncated, info = self.env.step(action)
self.accumulated_reward += float(reward)
self.current_step += 1
if self.current_step % self.delay == 0 or terminated or truncated:
reward = self.accumulated_reward
self.accumulated_reward = 0.0
else:
reward = 0.0
return obs, reward, terminated, truncated, info
|