Major dependencies upgrades (#18)

* Major dependencies upgrades

* Move from gym to gymnasium since gym is no longer maintained
* Remove support for stable-baselines and ray rllib with tensorflow
* Update all environments and tests to reflect dependencies changes
* Update all workflow files to use latest actions versions
* Remove unnecessary debug actions in the integration workflow (workflow disabled manually through the UI)

.github/workflows/createrelease.yml

@@ -11,10 +11,10 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@v2
+        uses: actions/checkout@v4.1.6
       - name: Create Release
         id: create_release
-        uses: actions/create-release@v1.0.1
+        uses: actions/create-release@v1.14.0
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:

.github/workflows/deploydocs.yml

@@ -10,13 +10,13 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@v2
+        uses: actions/checkout@v4.1.6
         with:
           persist-credentials: false
-      - name: Set up Python 3.8
-        uses: actions/setup-python@v1
+      - name: Set up Python 3.11
+        uses: actions/setup-python@v5.1.0
         with:
-          python-version: 3.8
+          python-version: "3.11"
       - name: Upgrade pip
         run: |
           python -m pip install --upgrade pip
@@ -27,7 +27,8 @@ jobs:
         run: |
           make html
       - name: Deploy to GitHub Pages
-        uses: peaceiris/actions-gh-pages@v3
+        uses: peaceiris/actions-gh-pages@v4.0.0
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           publish_dir: ./_build/html/
+        if: ${{ !env.ACT }}

.github/workflows/integration.yml

@@ -7,13 +7,12 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        rl-implementation: [rllib, stable_baselines_v3]
-        python-version: [3.6, 3.7, 3.8]
+        python-version: ["3.8", "3.9", "3.10", "3.11"]
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v4.1.6
       name: Checkout code
     - name: Set up Python ${{ matrix.python-version }}
-      uses: actions/setup-python@v2
+      uses: actions/setup-python@v5.1.0
       with:
         python-version: ${{ matrix.python-version }}
     - name: Install integration tests apt dependencies
@@ -22,31 +21,13 @@ jobs:
         sudo apt-get install -y cmake libopenmpi-dev python3-dev zlib1g-dev
     - name: Upgrade pip
       run: |
-        python -m pip install --upgrade pip
+        pip install --upgrade pip
     - name: Install package
       run: |
         pip install .[tests]
-
     - name: Install integration tests dependencies
       run: |
-        pip install tensorflow torch
-    - name: Install rllib
-      env:
-        PY_VER: ${{ matrix.python-version }}
-      run: |
-        if [ $PY_VER = "3.6" ]; then
-          pip install "https://s3-us-west-2.amazonaws.com/ray-wheels/latest/ray-2.0.0.dev0-cp36-cp36m-manylinux2014_x86_64.whl"
-        elif [ $PY_VER = "3.7" ]; then
-          pip install "https://s3-us-west-2.amazonaws.com/ray-wheels/latest/ray-2.0.0.dev0-cp37-cp37m-manylinux2014_x86_64.whl"
-        else
-          pip install "https://s3-us-west-2.amazonaws.com/ray-wheels/latest/ray-2.0.0.dev0-cp38-cp38-manylinux2014_x86_64.whl"
-        fi
-        pip install "ray[rllib]"
-      if: matrix.rl-implementation == 'rllib'
-    - name: Install stable-baselines-v3
-      run: |
-        pip install stable-baselines3
-      if: matrix.rl-implementation == 'stable_baselines_v3'
+        pip install torch "ray[rllib]"
     - name: Run integration tests
       run: |
-        pytest -v tests -k "integration and ${{ matrix.rl-implementation }}" -W ignore::UserWarning -W ignore::DeprecationWarning -W ignore::RuntimeWarning
+        pytest -vvv -s tests -k "integration"

.github/workflows/pypipublish.yml

@@ -10,18 +10,18 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@v2
-      - name: Set up Python 3.8
-        uses: actions/setup-python@v1
+        uses: actions/checkout@v4.1.6
+      - name: Set up Python 3.11
+        uses: actions/setup-python@v5.1.0
         with:
-          python-version: 3.8
+          python-version: 3.11
       - name: Install setuptools
         run: |
           pip install -U setuptools
       - name: Build source archive
         run: |
           python setup.py sdist
       - name: Publish 📦 PyPi
-        uses: pypa/gh-action-pypi-publish@v1.1.0
+        uses: pypa/gh-action-pypi-publish@v1.8.14
         with:
           password: ${{ secrets.pypi_password }}

.github/workflows/testpackage.yml

@@ -3,22 +3,18 @@ name: build
 on: [push]
 
 jobs:
-  build:
+  test:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.6, 3.7, 3.8, 3.9]
+        python-version: ["3.8", "3.9", "3.10", "3.11", "3.12"]
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v4.1.6
       name: Checkout code
     - name: Set up Python ${{ matrix.python-version }}
-      uses: actions/setup-python@v2
+      uses: actions/setup-python@v5.1.0
       with:
         python-version: ${{ matrix.python-version }}
-    - name: Install integration tests apt dependencies
-      run: |
-        sudo apt-get update
-        sudo apt-get install -y cmake libopenmpi-dev python3-dev zlib1g-dev
     - name: Upgrade pip
       run: |
         python -m pip install --upgrade pip
@@ -38,5 +34,6 @@ jobs:
     - name: Post to Coveralls.io
       env:
         COVERALLS_REPO_TOKEN: ${{ secrets.COVERALLS_REPO_TOKEN }}
+      if: ${{ !env.ACT }}
       run: |
-        coveralls
+        coveralls

examples/3-Way Soccer example.ipynb

@@ -535,7 +535,7 @@
     "env.reset()\n",
     "for _ in range(1, max_steps_limit):\n",
     "    print(env.render())\n",
-    "    obs, reward, done, info = env.step(1)\n",
+    "    obs, reward, done, truncated, info = env.step(1)\n",
     "    if done:\n",
     "        break"
    ]
@@ -596,7 +596,7 @@
     "env.reset()\n",
     "for _ in range(1, max_steps_limit):\n",
     "    print(env.render())\n",
-    "    obs, reward, done, info = env.step(env.action_space.sample())\n",
+    "    obs, reward, done, truncated, info = env.step(env.action_space.sample())\n",
     "    if done:\n",
     "        break"
    ]

oddsgym/envs/__init__.py

@@ -1,4 +1,4 @@
-import gym
+import gymnasium as gym
 
 from .base import BaseOddsEnv
 from .base_percentage import BasePercentageOddsEnv
@@ -30,11 +30,11 @@
 
 # register the *-data.co.uk environments in ray, only if ray is installed
 try:
-    from ray.tune.registry import register_env
+    from ray import tune
 except ImportError:  # pragma: no cover
     pass
 else:  # pragma: no cover
-    register_env('FootballDataDaily-ray-v0', lambda env_config: FootballDataDailyEnv(**env_config))
-    register_env('FootballDataDailyPercent-ray-v0', lambda env_config: FootballDataDailyPercentageEnv(**env_config))
-    register_env('TennisDataDaily-ray-v0', lambda env_config: TennisDataDailyEnv(**env_config))
-    register_env('TennisDataDailyPercent-ray-v0', lambda env_config: TennisDataDailyPercentageEnv(**env_config))
+    tune.register_env('FootballDataDaily-ray-v0', lambda env_config: gym.wrappers.FlattenObservation(FootballDataDailyEnv(env_config)))
+    tune.register_env('FootballDataDailyPercent-ray-v0', lambda env_config: gym.wrappers.FlattenObservation(FootballDataDailyPercentageEnv(env_config)))
+    tune.register_env('TennisDataDaily-ray-v0', lambda env_config: gym.wrappers.FlattenObservation(TennisDataDailyEnv(env_config)))
+    tune.register_env('TennisDataDailyPercent-ray-v0', lambda env_config: gym.wrappers.FlattenObservation(TennisDataDailyPercentageEnv(env_config)))

oddsgym/envs/base.py

@@ -1,10 +1,11 @@
-import gym
+import gymnasium as gym
 import numpy
 import numexpr
 from itertools import compress
 from pandas import DataFrame
 from tabulate import tabulate
-from infi.traceback import pretty_traceback_and_exit_decorator
+from typing import Any
+# from infi.traceback import pretty_traceback_and_exit_decorator
 
 
 class ActionsDict(dict):
@@ -56,7 +57,7 @@ class BaseOddsEnv(gym.Env):
     metadata = {'render.modes': ['human']}
     HEADERS = ['Current Step', 'Odds', 'Verbose Action', 'Action', 'Balance', 'Reward', 'Results', 'Done']
 
-    def __init__(self, odds, odds_column_names, results=None, starting_bank=10):
+    def __init__(self, odds, odds_column_names, results=None, starting_bank=300):
         """Initializes a new environment
 
         Parameters
@@ -83,6 +84,12 @@ def __init__(self, odds, odds_column_names, results=None, starting_bank=10):
 
     def _get_current_index(self):
         return self.current_step % self._odds.shape[0]
+
+    def _get_maximum_reward(self):
+        """Returns the maximum reward possible for this environment"""
+        results = numpy.zeros(shape=self._odds.shape)
+        results[numpy.arange(self._odds.shape[0]), numpy.array(self._results)] = 1
+        return (results * self._odds).sum().sum()
 
     def get_odds(self):
         """Returns the odds for the current step.
@@ -110,7 +117,7 @@ def get_bet(self, action):
         """
         return numpy.array([[bit for bit in numpy.binary_repr(action, width=self._odds.shape[1])]]).astype(int)
 
-    @pretty_traceback_and_exit_decorator
+    # @pretty_traceback_and_exit_decorator
     def step(self, action):
         """Run one timestep of the environment's dynamics. When end of episode is reached,
         you are responsible for calling reset() to reset this environment's state.
@@ -137,6 +144,7 @@ def step(self, action):
         odds = self.get_odds()
         reward = 0
         done = False
+        truncated = False
         info = self.create_info(action)
         if self.balance < 1:  # no more money :-(
             done = True
@@ -158,7 +166,7 @@ def step(self, action):
             else:
                 odds = self.get_odds()
         info.update(done=done)
-        return odds, reward, done, info
+        return odds, reward, done, truncated, info
 
     def get_reward(self, bet, odds, results):
         """ Calculates the reward
@@ -180,7 +188,7 @@ def get_reward(self, bet, odds, results):
         expense = numexpr.evaluate('sum(bet * bet_size_matrix)')
         return reward - expense
 
-    def reset(self):
+    def reset(self, *, seed=None, options=None):
         """Resets the state of the environment and returns an initial observation.
 
         Returns
@@ -190,7 +198,7 @@ def reset(self):
         """
         self.balance = self.starting_bank
         self.current_step = 0
-        return self.get_odds()
+        return self.get_odds(), {}
 
     def render(self, mode='human'):
         """Outputs the current balance and the current step.
@@ -265,7 +273,7 @@ def create_info(self, action):
         info : dict
             The info dictionary.
         """
-        return {'current_step': self.current_step, 'odds': self.get_odds(), 'verbose_action': self._verbose_actions[action],
+        return {'current_step': self.current_step, 'odds': self.get_odds().tolist(), 'verbose_action': self._verbose_actions[action],
                 'action': action, 'balance': self.balance, 'reward': 0,
                 'legal_bet': False, 'results': None, 'done': False}
 

oddsgym/envs/base_percentage.py

@@ -1,5 +1,5 @@
 import numpy
-from gym import spaces
+from gymnasium import spaces
 from .base import BaseOddsEnv
 
 

oddsgym/envs/daily_bets.py

@@ -1,7 +1,7 @@
-import gym
+import gymnasium as gym
 import numpy
 import numexpr
-from pandas import DataFrame
+from pandas import DataFrame, concat as concat_dataframes
 from .base import BaseOddsEnv
 
 
@@ -77,7 +77,7 @@ def __init__(self, odds, odds_column_names, results=None, max_number_of_games='a
         super().__init__(odds.drop('date', axis='columns'), odds_column_names, results, *args, **kwargs)
         self._odds_with_dates = odds.copy()
         self.days = odds['date'].unique()
-        self.days.sort()
+        self.days = self.days[self.days.argsort()]
         self._max_number_of_games = None
         if max_number_of_games == 'auto':
             self.max_number_of_games = odds.set_index('date').groupby(by='date').size().max()
@@ -114,7 +114,7 @@ def get_odds(self):
         filler_odds = DataFrame(numpy.zeros(numpy.array([self.max_number_of_games, self._odds.shape[1]]) -
                                             numpy.array([current_odds.shape[0], 0])),
                                 columns=self._odds_columns_names)
-        return current_odds.append(filler_odds, ignore_index=True).values
+        return concat_dataframes([current_odds, filler_odds], ignore_index=True).values
 
     def get_bet(self, action):
         """Returns the betting matrix for the action provided.