dodox/fancy_gym
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

update according to reviews opinion & fix bugs in box pushing IK

Browse Source
This commit is contained in:
Hongyi Zhou
2022-11-20 21:56:32 +01:00
parent fc3051bf57
commit 2674bf80fe
10 changed files with 94 additions and 185 deletions
Download Patch File Download Diff File Expand all files Collapse all files
test/test_black_box.py
+2 -30
View File
@@ -158,7 +158,7 @@ def test_context_space(mp_type: str, env_wrap: Tuple[str, Type[RawInterfaceWrapp
@pytest.mark.parametrize('mp_type', ['promp', 'dmp', 'prodmp'])
@pytest.mark.parametrize('num_dof', [0, 1, 2, 5])
@pytest.mark.parametrize('num_basis', [0, 2, 5]) # should add 1 back after the bug is fixed
@pytest.mark.parametrize('num_basis', [0, 1, 2, 5])
@pytest.mark.parametrize('learn_tau', [True, False])
@pytest.mark.parametrize('learn_delay', [True, False])
def test_action_space(mp_type: str, num_dof: int, num_basis: int, learn_tau: bool, learn_delay: bool):
@@ -343,32 +343,4 @@ def test_learn_tau_and_delay(mp_type: str, tau: float, delay: float):
active_pos = pos[delay_time_steps: joint_time_steps - 1]
active_vel = vel[delay_time_steps: joint_time_steps - 2]
assert np.all(active_pos != pos[-1]) and np.all(active_pos != pos[0])
assert np.all(active_vel != vel[-1]) and np.all(active_vel != vel[0])
@pytest.mark.parametrize('mp_type', ['promp', 'prodmp'])
@pytest.mark.parametrize('max_planning_times', [1, 2, 3, 4])
@pytest.mark.parametrize('sub_segment_steps', [5, 10])
def test_replanning_schedule(mp_type: str, max_planning_times: int, sub_segment_steps: int):
basis_generator_type = 'prodmp' if mp_type == 'prodmp' else 'rbf'
phase_generator_type = 'exp' if mp_type == 'prodmp' else 'linear'
env = fancy_gym.make_bb('toy-v0', [ToyWrapper],
{'max_planning_times': max_planning_times,
'replanning_schedule': lambda pos, vel, obs, action, t: t % sub_segment_steps == 0,
'verbose': 2},
{'trajectory_generator_type': mp_type,
},
{'controller_type': 'motor'},
{'phase_generator_type': phase_generator_type,
'learn_tau': False,
'learn_delay': False
},
{'basis_generator_type': basis_generator_type,
},
seed=SEED)
_ = env.reset()
d = False
for i in range(max_planning_times):
_, _, d, _ = env.step(env.action_space.sample())
assert d
assert np.all(active_vel != vel[-1]) and np.all(active_vel != vel[0])
test/test_replanning_envs.py
-82
View File
@@ -1,82 +0,0 @@
from itertools import chain
from typing import Tuple, Type, Union, Optional, Callable
import gym
import numpy as np
import pytest
from gym import register
from gym.core import ActType, ObsType
import fancy_gym
from fancy_gym.black_box.raw_interface_wrapper import RawInterfaceWrapper
import fancy_gym
from test.utils import run_env, run_env_determinism
Fancy_ProDMP_IDS = fancy_gym.ALL_FANCY_MOVEMENT_PRIMITIVE_ENVIRONMENTS['ProDMP']
All_ProDMP_IDS = fancy_gym.ALL_MOVEMENT_PRIMITIVE_ENVIRONMENTS['ProDMP']
class Object(object):
pass
class ToyEnv(gym.Env):
observation_space = gym.spaces.Box(low=-1, high=1, shape=(1,), dtype=np.float64)
action_space = gym.spaces.Box(low=-1, high=1, shape=(1,), dtype=np.float64)
dt = 0.02
def __init__(self, a: int = 0, b: float = 0.0, c: list = [], d: dict = {}, e: Object = Object()):
self.a, self.b, self.c, self.d, self.e = a, b, c, d, e
def reset(self, *, seed: Optional[int] = None, return_info: bool = False,
options: Optional[dict] = None) -> Union[ObsType, Tuple[ObsType, dict]]:
return np.array([-1])
def step(self, action: ActType) -> Tuple[ObsType, float, bool, dict]:
return np.array([-1]), 1, False, {}
def render(self, mode="human"):
pass
class ToyWrapper(RawInterfaceWrapper):
@property
def current_pos(self) -> Union[float, int, np.ndarray, Tuple]:
return np.ones(self.action_space.shape)
@property
def current_vel(self) -> Union[float, int, np.ndarray, Tuple]:
return np.zeros(self.action_space.shape)
@pytest.fixture(scope="session", autouse=True)
def setup():
register(
id=f'toy-v0',
entry_point='test.test_black_box:ToyEnv',
max_episode_steps=50,
)
# @pytest.mark.parametrize('env_id', All_ProDMP_IDS)
# def test_replanning_envs(env_id: str):
# """Tests that ProDMP environments run without errors using random actions."""
# run_env(env_id)
#
# @pytest.mark.parametrize('env_id', All_ProDMP_IDS)
# def test_replanning_determinism(env_id: str):
# """Tests that ProDMP environments are deterministic."""
# run_env_determinism(env_id, 0)
@pytest.mark.parametrize('mp_type', ['promp', 'dmp', 'prodmp'])
def test_missing_local_state(mp_type: str):
basis_generator_type = 'prodmp' if mp_type == 'prodmp' else 'rbf'
env = fancy_gym.make_bb('toy-v0', [RawInterfaceWrapper], {},
{'trajectory_generator_type': mp_type},
{'controller_type': 'motor'},
{'phase_generator_type': 'exp'},
{'basis_generator_type': basis_generator_type})
env.reset()
with pytest.raises(NotImplementedError):
env.step(env.action_space.sample())
test/test_replanning_sequencing.py
+27 -1
View File
@@ -304,4 +304,30 @@ def test_replanning_with_learn_delay_and_tau(mp_type: str, max_planning_times: i
planning_times += 1
assert planning_times == max_planning_times
assert planning_times == max_planning_times
@pytest.mark.parametrize('mp_type', ['promp', 'prodmp'])
@pytest.mark.parametrize('max_planning_times', [1, 2, 3, 4])
@pytest.mark.parametrize('sub_segment_steps', [5, 10])
def test_replanning_schedule(mp_type: str, max_planning_times: int, sub_segment_steps: int):
basis_generator_type = 'prodmp' if mp_type == 'prodmp' else 'rbf'
phase_generator_type = 'exp' if mp_type == 'prodmp' else 'linear'
env = fancy_gym.make_bb('toy-v0', [ToyWrapper],
{'max_planning_times': max_planning_times,
'replanning_schedule': lambda pos, vel, obs, action, t: t % sub_segment_steps == 0,
'verbose': 2},
{'trajectory_generator_type': mp_type,
},
{'controller_type': 'motor'},
{'phase_generator_type': phase_generator_type,
'learn_tau': False,
'learn_delay': False
},
{'basis_generator_type': basis_generator_type,
},
seed=SEED)
_ = env.reset()
d = False
for i in range(max_planning_times):
_, _, d, _ = env.step(env.action_space.sample())
assert d