mirror of
https://github.com/Farama-Foundation/Gymnasium.git
synced 2025-08-01 06:07:08 +00:00
e913bc81b8
* feat: add `isort` to `pre-commit` * ci: skip `__init__.py` file for `isort` * ci: make `isort` mandatory in lint pipeline * docs: add a section on Git hooks * ci: check isort diff * fix: isort from master branch * docs: add pre-commit badge * ci: update black + bandit versions * feat: add PR template * refactor: PR template * ci: remove bandit * docs: add Black badge * ci: try to remove all `|| true` statements * ci: remove lint_python job - Remove `lint_python` CI job - Move `pyupgrade` job to `pre-commit` workflow * fix: avoid messing with typing * docs: add a note on running `pre-cpmmit` manually * ci: apply `pre-commit` to the whole codebase
36 lines
1.2 KiB
Python
36 lines
1.2 KiB
Python
import numpy as np
|
|
|
|
from gym import utils
|
|
from gym.envs.mujoco import mujoco_env
|
|
|
|
|
|
class SwimmerEnv(mujoco_env.MujocoEnv, utils.EzPickle):
|
|
def __init__(self):
|
|
mujoco_env.MujocoEnv.__init__(self, "swimmer.xml", 4)
|
|
utils.EzPickle.__init__(self)
|
|
|
|
def step(self, a):
|
|
ctrl_cost_coeff = 0.0001
|
|
xposbefore = self.sim.data.qpos[0]
|
|
self.do_simulation(a, self.frame_skip)
|
|
xposafter = self.sim.data.qpos[0]
|
|
reward_fwd = (xposafter - xposbefore) / self.dt
|
|
reward_ctrl = -ctrl_cost_coeff * np.square(a).sum()
|
|
reward = reward_fwd + reward_ctrl
|
|
ob = self._get_obs()
|
|
return ob, reward, False, dict(reward_fwd=reward_fwd, reward_ctrl=reward_ctrl)
|
|
|
|
def _get_obs(self):
|
|
qpos = self.sim.data.qpos
|
|
qvel = self.sim.data.qvel
|
|
return np.concatenate([qpos.flat[2:], qvel.flat])
|
|
|
|
def reset_model(self):
|
|
self.set_state(
|
|
self.init_qpos
|
|
+ self.np_random.uniform(low=-0.1, high=0.1, size=self.model.nq),
|
|
self.init_qvel
|
|
+ self.np_random.uniform(low=-0.1, high=0.1, size=self.model.nv),
|
|
)
|
|
return self._get_obs()
|