Fix numpy warning and update migration guide (#307)

docs/guide/migration.rst

@@ -49,6 +49,7 @@ Breaking Changes
 - We dropped GAIL support as we are focusing on model-free RL only, you can however take a look at the :ref:`imitation project <imitation>` which implements
   GAIL and other imitation learning algorithms on top of SB3.
 - ``action_probability`` is currently not implemented in the base class
+- ``pretrain()`` method for behavior cloning was removed (see `issue #27 <https://github.com/DLR-RM/stable-baselines3/issues/27>`_)
 
 You can take a look at the `issue about SB3 implementation design <https://github.com/hill-a/stable-baselines/issues/576>`_ for more details.
 

docs/misc/changelog.rst

@@ -3,7 +3,7 @@
 Changelog
 ==========
 
-Pre-Release 0.11.0a6 (WIP)
+Pre-Release 0.11.0a7 (WIP)
 -------------------------------
 
 Breaking Changes:
@@ -38,6 +38,7 @@ Bug Fixes:
 - Added informative ``PPO`` construction error in edge-case scenario where ``n_steps * n_envs = 1`` (size of rollout buffer),
   which otherwise causes downstream breaking errors in training (@decodyng)
 - Fixed discrete observation space support when using multiple envs with A2C/PPO (thanks @ardabbour)
+- Fixed numpy warning (replaced ``np.bool`` with ``bool``)
 
 Deprecations:
 ^^^^^^^^^^^^^
@@ -66,6 +67,7 @@ Documentation:
 - Fix docstring of classes in atari_wrappers.py which were inside the constructor (@LucasAlegre)
 - Added SB3-Contrib page
 - Fix bug in the example code of DQN (@AptX395)
+- Updated migration guide
 
 Pre-Release 0.10.0 (2020-10-28)
 -------------------------------

stable_baselines3/common/base_class.py

@@ -377,7 +377,7 @@ class BaseAlgorithm(ABC):
         # Avoid resetting the environment when calling ``.learn()`` consecutive times
         if reset_num_timesteps or self._last_obs is None:
             self._last_obs = self.env.reset()
-            self._last_dones = np.zeros((self.env.num_envs,), dtype=np.bool)
+            self._last_dones = np.zeros((self.env.num_envs,), dtype=bool)
             # Retrieve unnormalized observation for saving into the buffer
             if self._vec_normalize_env is not None:
                 self._last_original_obs = self._vec_normalize_env.get_original_obs()

stable_baselines3/common/callbacks.py

@@ -447,7 +447,7 @@ class StopTrainingOnRewardThreshold(BaseCallback):
 
     def _on_step(self) -> bool:
         assert self.parent is not None, "``StopTrainingOnMinimumReward`` callback must be used " "with an ``EvalCallback``"
-        # Convert np.bool to bool, otherwise callback() is False won't work
+        # Convert np.bool_ to bool, otherwise callback() is False won't work
         continue_training = bool(self.parent.best_mean_reward < self.reward_threshold)
         if self.verbose > 0 and not continue_training:
             print(

stable_baselines3/common/vec_env/dummy_vec_env.py

@@ -29,7 +29,7 @@ class DummyVecEnv(VecEnv):
         self.keys, shapes, dtypes = obs_space_info(obs_space)
 
         self.buf_obs = OrderedDict([(k, np.zeros((self.num_envs,) + tuple(shapes[k]), dtype=dtypes[k])) for k in self.keys])
-        self.buf_dones = np.zeros((self.num_envs,), dtype=np.bool)
+        self.buf_dones = np.zeros((self.num_envs,), dtype=bool)
         self.buf_rews = np.zeros((self.num_envs,), dtype=np.float32)
         self.buf_infos = [{} for _ in range(self.num_envs)]
         self.actions = None

stable_baselines3/version.txt

@@ -1 +1 @@
-0.11.0a6
+0.11.0a7