Packaging with pyproject.toml (#180)

.github/workflows/build-publish.yml

@@ -39,15 +39,15 @@ jobs:
     steps:
     - uses: actions/checkout@v3
     - name: Set up Python
-      uses: actions/setup-python@v2
+      uses: actions/setup-python@v4
       with:
         python-version: '3.x'
     - name: Install dependencies
-      run: python -m pip install --upgrade setuptools wheel
-    - name: Build wheels
-      run: python setup.py sdist bdist_wheel
+      run: python -m pip install --upgrade pip setuptools build
+    - name: Build sdist and wheels
+      run: python -m build
     - name: Store wheels
-      uses: actions/upload-artifact@v2
+      uses: actions/upload-artifact@v3
       with:
         path: dist
 
@@ -58,7 +58,7 @@ jobs:
     if: github.event_name == 'release' && github.event.action == 'published'
     steps:
     - name: Download dists
-      uses: actions/download-artifact@v2
+      uses: actions/download-artifact@v3
       with:
         name: artifact
         path: dist

.github/workflows/build.yml

@@ -11,7 +11,7 @@ jobs:
       matrix:
         python-version: ['3.7', '3.8', '3.9', '3.10']
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v3
       - run: |
            docker build -f bin/all-py.Dockerfile \
              --build-arg PYTHON_VERSION=${{ matrix.python-version }} \
@@ -23,7 +23,7 @@ jobs:
     runs-on:
       ubuntu-latest
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v3
       - run: |
            docker build -f bin/necessary-py.Dockerfile \
              --build-arg PYTHON_VERSION='3.10' \

.github/workflows/pre-commit.yml

@@ -13,9 +13,9 @@ jobs:
   pre-commit:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v2
-      - uses: actions/setup-python@v2
-      - run: pip install pre-commit
-      - run: pre-commit --version
-      - run: pre-commit install
-      - run: pre-commit run --all-files
+      - uses: actions/checkout@v3
+      - uses: actions/setup-python@v4
+      - run: python -m pip install pre-commit
+      - run: python -m pre_commit --version
+      - run: python -m pre_commit install
+      - run: python -m pre_commit run --all-files

gymnasium/envs/mujoco/ant_v4.py

@@ -169,9 +169,9 @@ class AntEnv(MujocoEnv, utils.EzPickle):
     | `exclude_current_positions_from_observation`| **bool** | `True`| Whether or not to omit the x- and y-coordinates from observations. Excluding the position can serve as an inductive bias to induce position-agnostic behavior in policies |
 
     ## Version History
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks. Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/half_cheetah_v4.py

@@ -128,9 +128,9 @@ class HalfCheetahEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks. Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/hopper_v4.py

@@ -134,9 +134,9 @@ class HopperEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks. Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/humanoid_v4.py

@@ -208,9 +208,9 @@ class HumanoidEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks. Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/humanoidstandup_v4.py

@@ -184,9 +184,9 @@ class HumanoidStandupEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks. Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/inverted_double_pendulum_v4.py

@@ -117,9 +117,9 @@ class InvertedDoublePendulumEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks (including inverted pendulum)
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/inverted_pendulum_v4.py

@@ -86,9 +86,9 @@ class InvertedPendulumEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks (including inverted pendulum)
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/pusher_v4.py

@@ -133,8 +133,8 @@ class PusherEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks (not including reacher, which has a max_time_steps of 50). Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/reacher_v4.py

@@ -115,8 +115,8 @@ class ReacherEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks (not including reacher, which has a max_time_steps of 50). Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/swimmer_v4.py

@@ -117,9 +117,9 @@ class SwimmerEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks. Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

gymnasium/envs/mujoco/walker2d_v4.py

@@ -139,9 +139,9 @@ class Walker2dEnv(MujocoEnv, utils.EzPickle):
 
     ## Version History
 
-    * v4: all mujoco environments now use the mujoco bindings in mujoco>=2.1.3
-    * v3: support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
-    * v2: All continuous control environments now use mujoco_py >= 1.50
+    * v4: All MuJoCo environments now use the MuJoCo bindings in mujoco >= 2.1.3
+    * v3: Support for `gymnasium.make` kwargs such as `xml_file`, `ctrl_cost_weight`, `reset_noise_scale`, etc. rgb rendering comes from tracking camera (so agent does not run away from screen)
+    * v2: All continuous control environments now use mujoco-py >= 1.50
     * v1: max_time_steps raised to 1000 for robot based tasks. Added reward_threshold to environments.
     * v0: Initial versions release (1.0.0)
     """

pyproject.toml

@@ -1,3 +1,113 @@
+# Package ######################################################################
+
+[build-system]
+requires = ["setuptools >= 61.0.0"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "gymnasium"
+description = "A standard API for reinforcement learning and a diverse set of reference environments (formerly Gym)."
+readme = "README.md"
+requires-python = ">= 3.7"
+authors = [{ name = "Farama Foundation", email = "[email protected]" }]
+license = { text = "MIT License" }
+keywords = ["Reinforcement Learning", "game", "RL", "AI", "gymnasium"]
+classifiers = [
+    "Development Status :: 4 - Beta",  # change to `5 - Production/Stable` when ready
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.7",
+    "Programming Language :: Python :: 3.8",
+    "Programming Language :: Python :: 3.9",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    'Intended Audience :: Science/Research',
+    'Topic :: Scientific/Engineering :: Artificial Intelligence',
+]
+dependencies = [
+    "numpy >=1.21.0",
+    "jax-jumpy >=0.2.0",
+    "cloudpickle >=1.2.0",
+    "importlib-metadata >=4.8.0; python_version < '3.10'",
+    "typing-extensions >=4.3.0; python_version == '3.7'",
+    "gymnasium-notices >=0.0.1",
+    "shimmy >=0.1.0,<1.0",
+]
+dynamic = ["version"]
+
+[project.optional-dependencies]
+# Update dependencies in `all` if any are added or removed
+atari = ["shimmy[atari] >=0.1.0,<1.0"]
+accept-rom-license = ["autorom[accept-rom-license] ~=0.4.2"]
+box2d = ["box2d-py ==2.3.5", "pygame ==2.1.3.dev8", "swig ==4.*"]
+classic-control = ["pygame ==2.1.3.dev8"]
+classic_control = ["pygame ==2.1.3.dev8"]  # kept for backward compatibility
+mujoco-py = ["mujoco-py >=2.1,<2.2"]
+mujoco_py = ["mujoco-py >=2.1,<2.2"]  # kept for backward compatibility
+mujoco = ["mujoco >=2.3.0", "imageio >=2.14.1"]
+toy-text = ["pygame ==2.1.3.dev8"]
+toy_text = ["pygame ==2.1.3.dev8"]         # kept for backward compatibility
+jax = ["jax ==0.3.20", "jaxlib ==0.3.20"]
+other = [
+    "lz4 >=3.1.0",
+    "opencv-python >=3.0",
+    "matplotlib >=3.0",
+    "moviepy >=1.0.0",
+    "tensorflow >=2.1.0",
+    "torch >=1.0.0",
+]
+all = [
+    # All dependencies above except accept-rom-license
+    # NOTE: No need to manually remove the duplicates, setuptools automatically does that.
+    # atari
+    "shimmy[atari] >=0.1.0,<1.0",
+    # box2d
+    "box2d-py ==2.3.5",
+    "pygame ==2.1.3.dev8",
+    "swig ==4.*",
+    # classic-control
+    "pygame ==2.1.3.dev8",
+    # mujoco-py
+    "mujoco-py >=2.1,<2.2",
+    # mujoco
+    "mujoco >=2.3.0",
+    "imageio >=2.14.1",
+    # toy-text
+    "pygame ==2.1.3.dev8",
+    # jax
+    "jax ==0.3.20",
+    "jaxlib ==0.3.20",
+    # other
+    "lz4 >=3.1.0",
+    "opencv-python >=3.0",
+    "matplotlib >=3.0",
+    "moviepy >=1.0.0",
+    "tensorflow >=2.1.0",
+    "torch >=1.0.0",
+]
+testing = ["pytest ==7.1.3"]
+
+[project.urls]
+Homepage = "https://farama.org"
+Repository = "https://github.com/Farama-Foundation/Gymnasium"
+Documentation = "https://gymnasium.farama.org"
+"Bug Report" = "https://github.com/Farama-Foundation/Gymnasium/issues"
+
+[tool.setuptools]
+include-package-data = true
+
+[tool.setuptools.packages.find]
+include = ["gymnasium", "gymnasium.*"]
+
+[tool.setuptools.package-data]
+gymnasium = [
+    "envs/mujoco/assets/*.xml",
+    "envs/classic_control/assets/*.png",
+    "envs/toy_text/font/*.ttf",
+    "envs/toy_text/img/*.png",
+    "py.typed",
+]
+
 # Linters and Test tools #######################################################
 
 [tool.black]
@@ -7,6 +117,7 @@ safe = true
 atomic = true
 profile = "black"
 src_paths = ["gymnasium", "tests", "docs/scripts"]
+extra_standard_library = ["typing_extensions"]
 indent = 4
 lines_after_imports = 2
 multi_line_output = 3