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

refractoring of DMP environmets to fit gym interface better.

Browse Source
This commit is contained in:
ottofabian
2021-03-26 14:05:16 +01:00
parent 6233c85904
commit 7ceadeff0a
20 changed files with 661 additions and 568 deletions
Download Patch File Download Diff File Expand all files Collapse all files
alr_envs/mujoco/ball_in_a_cup/utils.py
+40 -40
View File
@@ -1,4 +1,4 @@
from alr_envs.utils.detpmp_env_wrapper import DetPMPEnvWrapper
from alr_envs.utils.wrapper.detpmp_wrapper import DetPMPWrapper
from alr_envs.mujoco.ball_in_a_cup.ball_in_a_cup import ALRBallInACupEnv
from alr_envs.mujoco.ball_in_a_cup.ball_in_a_cup_simple import ALRBallInACupEnv as ALRBallInACupEnvSimple
@@ -17,19 +17,19 @@ def make_contextual_env(rank, seed=0):
def _init():
env = ALRBallInACupEnv()
env = DetPMPEnvWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env = DetPMPWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env.seed(seed + rank)
return env
@@ -51,19 +51,19 @@ def make_env(rank, seed=0):
def _init():
env = ALRBallInACupEnvSimple()
env = DetPMPEnvWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.25,
zero_start=True,
zero_goal=True
)
env = DetPMPWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.25,
zero_start=True,
zero_goal=True
)
env.seed(seed + rank)
return env
@@ -85,19 +85,19 @@ def make_simple_env(rank, seed=0):
def _init():
env = ALRBallInACupEnvSimple()
env = DetPMPEnvWrapper(env,
num_dof=3,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos[1::2],
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env = DetPMPWrapper(env,
num_dof=3,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos[1::2],
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env.seed(seed + rank)
return env
alr_envs/mujoco/beerpong/utils.py
+40 -40
View File
@@ -1,4 +1,4 @@
from alr_envs.utils.detpmp_env_wrapper import DetPMPEnvWrapper
from alr_envs.utils.wrapper.detpmp_wrapper import DetPMPWrapper
from alr_envs.mujoco.beerpong.beerpong import ALRBeerpongEnv
from alr_envs.mujoco.beerpong.beerpong_simple import ALRBeerpongEnv as ALRBeerpongEnvSimple
@@ -17,19 +17,19 @@ def make_contextual_env(rank, seed=0):
def _init():
env = ALRBeerpongEnv()
env = DetPMPEnvWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env = DetPMPWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env.seed(seed + rank)
return env
@@ -51,19 +51,19 @@ def make_env(rank, seed=0):
def _init():
env = ALRBeerpongEnvSimple()
env = DetPMPEnvWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.25,
zero_start=True,
zero_goal=True
)
env = DetPMPWrapper(env,
num_dof=7,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos,
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.25,
zero_start=True,
zero_goal=True
)
env.seed(seed + rank)
return env
@@ -85,19 +85,19 @@ def make_simple_env(rank, seed=0):
def _init():
env = ALRBeerpongEnvSimple()
env = DetPMPEnvWrapper(env,
num_dof=3,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos[1::2],
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env = DetPMPWrapper(env,
num_dof=3,
num_basis=5,
width=0.005,
policy_type="motor",
start_pos=env.start_pos[1::2],
duration=3.5,
post_traj_time=4.5,
dt=env.dt,
weights_scale=0.5,
zero_start=True,
zero_goal=True
)
env.seed(seed + rank)
return env