Merge pull request #214 from elladyr/dev

MyoChallenge Locomotion metrics

.github/workflows/python-app.yml

@@ -82,7 +82,7 @@ jobs:
 #       if: ${{ runner.os == 'macOS' }}
 #       run: |
 #         brew install --cask xquartz
-#         brew install hdf5        
+#         brew install hdf5
 #         export CPATH="/opt/homebrew/include/"
 #         export HDF5_DIR=/opt/homebrew/
 #         pip3 install h5py --only-binary h5py
@@ -97,9 +97,13 @@ jobs:
       if: ${{ runner.os == 'Linux' }}
       run: python3 -m mujoco.render_test
 
-    - name: Run Test environment
+    - name: Test myoapi
       run: |
         conda activate $CONDA_DEFAULT_ENV
+        python3 -m myosuite.tests.test_myoapi
+
+    - name: Run Test environment
+      run: |
         python3 -m myosuite.tests.test_myo
 
     - name: Install ffmpeg

README.md

@@ -46,7 +46,7 @@ Test your installation using the following command (this will return also a list
 ``` bash
 python -m myosuite.tests.test_myo
 ```
- 
+
 
 You can also visualize the environments with random controls using the command below:
 ``` bash
@@ -57,7 +57,10 @@ python -m myosuite.utils.examine_env --env_name myoElbowPose1D6MRandom-v0
 mjpython -m myosuite.utils.examine_env --env_name myoElbowPose1D6MRandom-v0
 ```
 
-
+It is possible to take advantage of the latest MyoSkeleton. Once added (follow the instructions prompted by `python -m myosuite_init`), run:
+``` bash
+python -m myosuite.utils.examine_sim -s myosuite/simhive/myo_model/myoskeleton/myoskeleton.xml
+```
 
 ## Examples
 It is possible to create and interface with MyoSuite environments just like any other OpenAI gym environments. For example, to use the `myoElbowPose1D6MRandom-v0` environment, it is possible simply to run: [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1zFuNLsrmx42vT4oV8RbnEWtkSJ1xajEo)
@@ -74,7 +77,7 @@ for _ in range(1000):
 env.close()
 ```
 
-You can find our [tutorials](https://github.com/myohub/myosuite/tree/main/docs/source/tutorials#tutorials) on the general features and the **ICRA2023 Colab Tutorial** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1KGqZgSYgKXF-vaYC33GR9llDsIW9Rp-q) **ICRA2024 Colab Tutorial** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1JwxE7o6Z3bqCT4ewELacJ-Z1SV8xFhKK#scrollTo=QDppGIzHB9Zu) 
+You can find our [tutorials](https://github.com/myohub/myosuite/tree/main/docs/source/tutorials#tutorials) on the general features and the **ICRA2023 Colab Tutorial** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1KGqZgSYgKXF-vaYC33GR9llDsIW9Rp-q) **ICRA2024 Colab Tutorial** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1JwxE7o6Z3bqCT4ewELacJ-Z1SV8xFhKK#scrollTo=QDppGIzHB9Zu)
 on how to load MyoSuite models/tasks, train them, and visualize their outcome. Also, you can find [baselines](https://github.com/myohub/myosuite/tree/main/myosuite/agents) to test some pre-trained policies.
 
 

myosuite/envs/myo/myochallenge/run_track_v0.py

@@ -51,6 +51,11 @@ class RunTrack(WalkEnvV0):
     OSL_PARAM_LIST = []
     OSL_PARAM_SELECT = 0
 
+    # Joint dict
+    pain_jnt = ['hip_adduction_l', 'hip_adduction_r', 'hip_flexion_l', 'hip_flexion_r', 'hip_rotation_l', 'hip_rotation_r',
+                'knee_angle_l', 'knee_angle_l_rotation2', 'knee_angle_l_rotation3', 
+                'mtp_angle_l', 'ankle_angle_l', 'subtalar_angle_l']
+
     def __init__(self, model_path, obsd_model_path=None, seed=None, **kwargs):
         # This flag needs to be here to prevent the simulation from starting in a done state
         # Before setting the key_frames, the model and opponent will be in the cartesian position,
@@ -82,13 +87,15 @@ def _setup(self,
                real_width=1,
                distance_thr = 10,
                init_pose_path=None,
+               remap_required=False,
                **kwargs,
                ):
 
         self.startFlag = False
         # Terrain type
         self.terrain_type = TrackTypes.FLAT.value
 
+        self.remap_required = remap_required
         # Env initialization with data
         if init_pose_path is not None:
             file_path = os.path.join(init_pose_path)
@@ -173,7 +180,9 @@ def get_reward_dict(self, obs_dict):
         DEFAULT_RWD_KEYS_AND_WEIGHTS dict, or when registering the environment
         with gym.register in myochallenge/__init__.py
         """
-        act_mag = np.linalg.norm(self.obs_dict['act'], axis=-1)/self.sim.model.na if self.sim.model.na !=0 else 0
+        # act_mag = np.linalg.norm(self.obs_dict['act'], axis=-1)/self.sim.model.na if self.sim.model.na !=0 else 0
+        act_mag = np.mean( np.square(self.obs_dict['act']) ) if self.sim.model.na !=0 else 0
+        pain = self.get_pain()
 
         # The task is entirely defined by these 3 lines
         score = self.get_score()
@@ -186,6 +195,7 @@ def get_reward_dict(self, obs_dict):
 
                 # Optional Keys
                 ('act_reg', act_mag.squeeze()),
+                ('pain', pain),
                 # Must keys
                 ('sparse',  score),
                 ('solved',  win_cdt),
@@ -199,10 +209,16 @@ def get_metrics(self, paths):
         Evaluate paths and report metrics
         """
         # average sucess over entire env horizon
+        times = np.mean([np.round(p['env_infos']['obs_dict']['time'][-1], 5) for p in paths])
         score = np.mean([np.sum(p['env_infos']['rwd_dict']['sparse']) for p in paths])
+        effort = np.mean([np.sum(p['env_infos']['rwd_dict']['act_reg']) for p in paths])
+        pain = np.mean([np.sum(p['env_infos']['rwd_dict']['pain']) for p in paths])
 
         metrics = {
-            'score': score,
+            'score': score, # Distance travelled
+            'Time': times,
+            'effort': effort,
+            'pain': pain,
             }
         return metrics
 
@@ -395,6 +411,15 @@ def get_score(self):
         score = (y_pos - 15) / (- 15 - 15)
         return score.squeeze()
 
+    def get_pain(self):
+        if not self.startFlag:
+            return -1
+
+        pain_score = 0
+        for joint in self.pain_jnt:
+            pain_score += np.clip(np.abs(self.get_limitfrc(joint).squeeze()), -1000, 1000)
+        return pain_score / len(self.pain_jnt)
+
     def _get_muscle_lengthRange(self):
         return self.sim.model.actuator_lengthrange.copy()
 
@@ -449,6 +474,14 @@ def _get_fallen_condition(self):
             else:
                 return 0
 
+    def get_limitfrc(self, joint_name):
+        non_joint_limit_efc_idxs = np.where(self.sim.data.efc_type != self.sim.lib.mjtConstraint.mjCNSTR_LIMIT_JOINT)[0]
+        only_jnt_lim_efc_force = self.sim.data.efc_force.copy()
+        only_jnt_lim_efc_force[non_joint_limit_efc_idxs] = 0.0
+        joint_force = np.zeros((self.sim.model.nv,))
+        self.sim.lib.mj_mulJacTVec(self.sim.model._model, self.sim.data._data, joint_force, only_jnt_lim_efc_force)
+        return joint_force[self.sim.model.joint(joint_name).dofadr]
+
     def get_internal_qpos(self):
         temp_qpos = self.sim.data.qpos.copy()
         to_remove = [self.sim.model.joint('osl_knee_angle_r').qposadr[0].copy(), self.sim.model.joint('osl_ankle_angle_r').qposadr[0].copy()]
@@ -637,6 +670,11 @@ def _prepareActions(self, mus_actions):
         Only considers the 54 muscles of the OSLMyoleg model
         """
 
+        if self.remap_required:
+            if np.any( (mus_actions < -1) | (mus_actions > 1) ):
+                raise ValueError("Input value should be between -1 and 1")
+            mus_actions = (mus_actions + 1) / 2
+
         full_actions = np.zeros(self.sim.model.nu,)
         full_actions[0:self.sim.model.na] = mus_actions[0:self.sim.model.na].copy()
 
@@ -659,6 +697,6 @@ def get_osl_sens(self):
         osl_sens_data['knee_vel'] = self.sim.data.joint('osl_knee_angle_r').qvel[0].copy()
         osl_sens_data['ankle_angle'] = self.sim.data.joint('osl_ankle_angle_r').qpos[0].copy()
         osl_sens_data['ankle_vel'] = self.sim.data.joint('osl_ankle_angle_r').qvel[0].copy()
-        osl_sens_data['load'] = self.sim.data.sensor('r_socket_load').data[1].copy() # Only vertical
+        osl_sens_data['load'] = -1*self.sim.data.sensor('r_osl_load').data[1].copy() # Only vertical
 
         return osl_sens_data

myosuite/tests/test_myoapi.py

@@ -0,0 +1,25 @@
+
+from pathlib import Path
+from unittest.mock import patch
+
+from myosuite_init import fetch_simhive
+
+my_file = Path("/path/to/file")
+
+@patch('builtins.input', side_effect=['no'])
+def test_no_myoapi(mock_input):
+    print("mock_input", mock_input.side_effect)
+    fetch_simhive()
+    assert not Path("myosuite/simhive/myo_model/myoskeleton/myoskeleton.xml").exists()
+
+@patch('builtins.input', side_effect=['yes'])
+def test_yes_myoapi(mock_input):
+    print("mock_input", mock_input.side_effect)
+    fetch_simhive()
+    assert Path("myosuite/simhive/myo_model/myoskeleton/myoskeleton.xml").exists()
+
+test_no_myoapi()
+test_yes_myoapi()
+
+
+

myosuite/version.py

@@ -1 +1 @@
-__version_tuple__ = (2, 5, 0)
+__version_tuple__ = (2, 7, 0)

myosuite_init.py

@@ -0,0 +1,113 @@
+import os
+import shutil
+from os.path import expanduser
+
+import git
+
+curr_dir = os.path.dirname(os.path.abspath(__file__))
+simhive_path = os.path.join(curr_dir, 'myosuite', 'simhive')
+
+
+# from myosuite.utils.import_utils import fetch_git
+
+def fetch_git(repo_url, commit_hash, clone_directory, clone_path=None):
+    """
+    fetch git repo using provided details
+    """
+    if clone_path is None:
+        clone_path = os.path.join(expanduser("~"), ".myosuite")
+    clone_directory = os.path.join(clone_path, clone_directory)
+
+    try:
+        # Create the clone directory if it doesn't exist
+        os.makedirs(clone_directory, exist_ok=True)
+
+        # Clone the repository to the specified path
+        if not os.path.exists(os.path.join(clone_directory,'.git')):
+            repo = git.Repo.clone_from(repo_url, clone_directory)
+            print(f"{repo_url} cloned at {clone_directory}")
+        else:
+            repo = git.Repo(clone_directory)
+            origin = repo.remote('origin')
+            origin.fetch()
+
+        # Check out the specific commit if not already
+        current_commit_hash = repo.head.commit.hexsha
+        if current_commit_hash != commit_hash:
+            repo.git.checkout(commit_hash)
+            print(f"{repo_url}@{commit_hash} fetched at {clone_directory}")
+    except git.GitCommandError as e:
+        print(f"Error: {e}")
+
+    return clone_directory
+
+
+def clean_simhive():
+    """
+    Remove cached simhive if it exists
+    """
+    print("MyoSuite:> Clearing SimHive ...")
+    api_path = os.path.join(simhive_path, 'myo_model')
+    if os.path.exists(api_path):
+        shutil.rmtree(api_path)
+    else:
+        print("MyoSuite:> SimHive/myo_model directory does not exist.")
+    print("MyoSuite:> SimHive cleared")
+
+
+def accept_license():
+    prompt = """
+A permissive license for non-commercial scientific research is available.
+You can review the license at: https://github.com/myolab/myo_model/blob/main/LICENSE
+
+Do you accept the terms of the license? (yes/no):
+"""
+    response = input(prompt).strip().lower()
+
+    if response == 'yes':
+        print("Thank you for accepting the license. You may proceed.")
+        return True
+    elif response == 'no':
+        print("You have rejected the license terms. Exiting...")
+        return False
+    else:
+        print("Invalid input. Please enter 'yes' or 'no'.")
+        return accept_license()  # Recursively prompt again for valid input
+
+
+def fetch_simhive():
+    """
+    fetch a copy of simhive
+    """
+    print("MyoSuite:> Initializing...")
+
+    # Mark the SimHive version (ToDo: Remove this when commits hashes are auto fetched from submodules)
+    __version__ = "2.5.0"
+
+    # Inform user about API
+    if accept_license():
+        # Proceed with the rest of the code
+        print("MyoSuite:> License accepted. Proceeding initialization ...")
+    else:
+        # Exit or handle the rejection case
+        print("MyoSuite:> License rejected. Exiting")
+        return
+
+    # Fetch SimHive
+    print("MyoSuite:> Downloading simulation assets (upto ~100MBs)")
+    fetch_git(repo_url="https://github.com/myolab/myo_model.git",
+                commit_hash="619b1a876113e91a302b9baeaad6c2341e12ac81",
+                clone_directory="myo_model",
+                clone_path=simhive_path)
+
+
+    # mark successful creation of simhive
+    filename = os.path.join(simhive_path, "simhive-version")
+    with open(filename, 'w') as file:
+        file.write(__version__)
+
+    print("MyoSuite:> Successfully Initialized.")
+
+
+if __name__ == "__main__":
+    fetch_simhive()

setup.py

@@ -1,7 +1,8 @@
 import os
-import sys
 import re
-from setuptools import setup, find_packages
+import sys
+
+from setuptools import find_packages, setup
 
 if sys.version_info.major != 3:
     print("This Python is only compatible with Python 3, but you are running "
@@ -52,8 +53,14 @@ def package_files(directory):
             "Topic :: Scientific/Engineering :: Artificial Intelligence ",
             "Operating System :: OS Independent",
         ],
-        package_data={'': mjc_models_files},
+        package_data={'': mjc_models_files+['../myosuite_init.py']},
         packages=find_packages(exclude=("myosuite.agents")),
         python_requires=">=3.8",
         install_requires=fetch_requirements(),
+        entry_points={
+            'console_scripts': [
+                'myoapi_init = myosuite_init:fetch_simhive',
+                'myoapi_clean = myosuite_init:clean_simhive',
+            ],
+        },
     )