feat (logging) Improve logging (#226)

* new logger

* new logger and evaluation message

* sed to remove unused import everywhere

* misc eval changes and gym up level

* removed last unused import

* docstrings and cleanup of eval

* readd verbose

* bug gym logging

* add to api

* add colors

* new style log

* misc ameliorations

* add test on second line also

* add a comment to warn in logging to be careful when changing default message

docs/api.rst

@@ -176,6 +176,15 @@ Check Utilities
    utils.check_seeding_agent
    utils.check_agent_manager
 
+Logging Utilities
+-----------------
+
+.. autosummary::
+   :toctree: generated/
+   :template: function.rst
+
+   utils.logging.set_level
+
 
 Typing
 ------

examples/demo_bandits/plot_mirror_bandit.py

@@ -19,12 +19,14 @@
 from rlberry.agents.bandits import BanditWithSimplePolicy
 from rlberry.wrappers import WriterWrapper
 import rlberry.spaces as spaces
-import logging
+
 import requests
 import matplotlib.pyplot as plt
 
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 # Environment definition
 

rlberry/__init__.py

@@ -1,13 +1,17 @@
 from ._version import __version__
+import logging
+
+logger = logging.getLogger("rlberry_logger")
+
 from rlberry.utils.logging import configure_logging
 
 
 __path__ = __import__("pkgutil").extend_path(__path__, __name__)
 
 # Initialize logging level
-
 configure_logging(level="INFO")
 
+
 # define __version__
 
-__all__ = ["__version__"]
+__all__ = ["__version__", "logger"]

rlberry/agents/adaptiveql/adaptiveql.py

@@ -1,10 +1,11 @@
-import logging
 import gym.spaces as spaces
 import numpy as np
 from rlberry.agents import AgentWithSimplePolicy
 from rlberry.agents.adaptiveql.tree import MDPTreePartition
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class AdaptiveQLAgent(AgentWithSimplePolicy):

rlberry/agents/agent.py

@@ -1,7 +1,6 @@
 from abc import ABC, abstractmethod
 import dill
 import pickle
-import logging
 import numpy as np
 from inspect import signature
 from pathlib import Path
@@ -14,8 +13,9 @@
 from typing import Optional
 import inspect
 
+import rlberry
 
-logger = logging.getLogger(__name__)
+logger = rlberry.logger
 
 
 class Agent(ABC):

rlberry/agents/bandits/bandit_base.py

@@ -2,10 +2,12 @@
 from rlberry.agents import AgentWithSimplePolicy
 from .tools import BanditTracker
 import pickle
-import logging
+
 from pathlib import Path
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class BanditWithSimplePolicy(AgentWithSimplePolicy):

rlberry/agents/bandits/index_agents.py

@@ -1,8 +1,10 @@
 import numpy as np
 from rlberry.agents.bandits import BanditWithSimplePolicy
-import logging
 
-logger = logging.getLogger(__name__)
+
+import rlberry
+
+logger = rlberry.logger
 
 # TODO : fix bug when doing several fit, the fit do not resume. Should define
 #        self.rewards and self.action and resume training.

rlberry/agents/bandits/randomized_agents.py

@@ -1,8 +1,10 @@
 import numpy as np
 from rlberry.agents.bandits import BanditWithSimplePolicy
-import logging
 
-logger = logging.getLogger(__name__)
+
+import rlberry
+
+logger = rlberry.logger
 
 
 class RandomizedAgent(BanditWithSimplePolicy):

rlberry/agents/bandits/tools/tracker.py

@@ -1,8 +1,9 @@
-import logging
 from rlberry import metadata_utils
 from rlberry.utils.writers import DefaultWriter
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class BanditTracker(DefaultWriter):

rlberry/agents/bandits/ts_agents.py

@@ -1,8 +1,10 @@
 import numpy as np
 from rlberry.agents.bandits import BanditWithSimplePolicy
-import logging
 
-logger = logging.getLogger(__name__)
+
+import rlberry
+
+logger = rlberry.logger
 
 
 class TSAgent(BanditWithSimplePolicy):

rlberry/agents/experimental/jax/dqn/dqn.py

@@ -29,7 +29,7 @@
 import haiku as hk
 import jax
 import jax.numpy as jnp
-import logging
+
 import numpy as np
 import optax
 import dill
@@ -43,7 +43,9 @@
 from rlberry.agents.jax.utils.replay_buffer import ReplayBuffer
 from typing import Any, Callable, Mapping, Optional
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 @chex.dataclass

rlberry/agents/experimental/jax/utils/replay_buffer.py

@@ -5,10 +5,12 @@
 * For priority updates, see https://github.com/deepmind/reverb/issues/28
 """
 
-import logging
+
 import tensorflow as tf
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 try:
     import reverb

rlberry/agents/experimental/torch/avec/avec_ppo.py

@@ -1,5 +1,5 @@
 import torch
-import logging
+
 import torch.nn as nn
 import inspect
 
@@ -12,7 +12,9 @@
 from rlberry.utils.torch import choose_device
 from rlberry.wrappers.uncertainty_estimator_wrapper import UncertaintyEstimatorWrapper
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class AVECPPOAgent(AgentWithSimplePolicy):

rlberry/agents/experimental/torch/sac/sac.py

@@ -3,7 +3,7 @@
 import torch
 import torch.nn as nn
 from torch.nn.functional import one_hot
-import logging
+
 import gym.spaces as spaces
 
 from rlberry.agents import AgentWithSimplePolicy
@@ -14,7 +14,9 @@
 from rlberry.utils.torch import choose_device
 from rlberry.wrappers.uncertainty_estimator_wrapper import UncertaintyEstimatorWrapper
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class SACAgent(AgentWithSimplePolicy):

rlberry/agents/kernel_based/rs_kernel_ucbvi.py

@@ -1,5 +1,3 @@
-import logging
-
 import numpy as np
 from rlberry.utils.jit_setup import numba_jit
 
@@ -11,7 +9,9 @@
 from rlberry.agents.kernel_based.kernels import kernel_func
 from rlberry.agents.kernel_based.common import map_to_representative
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 @numba_jit

rlberry/agents/kernel_based/rs_ucbvi.py

@@ -1,4 +1,3 @@
-import logging
 from rlberry.agents.agent import AgentWithSimplePolicy
 import numpy as np
 
@@ -7,7 +6,9 @@
 from rlberry.agents.dynprog.utils import backward_induction_in_place
 from rlberry.agents.kernel_based.common import map_to_representative
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class RSUCBVIAgent(AgentWithSimplePolicy):

rlberry/agents/linear/lsvi_ucb.py

@@ -1,10 +1,11 @@
-import logging
 import numpy as np
 from rlberry.agents import AgentWithSimplePolicy
 from gym.spaces import Discrete
 from rlberry.utils.jit_setup import numba_jit
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 @numba_jit

rlberry/agents/mbqvi/mbqvi.py

@@ -1,11 +1,13 @@
 import numpy as np
-import logging
+
 
 from rlberry.agents import AgentWithSimplePolicy
 from rlberry.agents.dynprog.utils import backward_induction, value_iteration
 from gym.spaces import Discrete
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class MBQVIAgent(AgentWithSimplePolicy):

rlberry/agents/optql/optql.py

@@ -1,11 +1,12 @@
-import logging
 import numpy as np
 
 import gym.spaces as spaces
 from rlberry.agents import AgentWithSimplePolicy
 from rlberry.exploration_tools.discrete_counter import DiscreteCounter
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class OptQLAgent(AgentWithSimplePolicy):

rlberry/agents/psrl/psrl.py

@@ -1,4 +1,3 @@
-import logging
 import numpy as np
 
 import gym.spaces as spaces
@@ -9,7 +8,9 @@
     backward_induction_sd,
 )
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class PSRLAgent(AgentWithSimplePolicy):

rlberry/agents/rlsvi/rlsvi.py

@@ -1,4 +1,3 @@
-import logging
 import numpy as np
 
 import gym.spaces as spaces
@@ -10,7 +9,9 @@
     backward_induction_sd,
 )
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class RLSVIAgent(AgentWithSimplePolicy):

rlberry/agents/stable_baselines/stable_baselines.py

@@ -1,4 +1,3 @@
-import logging
 from pathlib import Path
 from typing import Any, Dict, Optional, Tuple, Type, Union
 
@@ -13,7 +12,9 @@
 from rlberry.agents import AgentWithSimplePolicy
 
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 def is_recordable(value: Any) -> bool:

rlberry/agents/torch/a2c/a2c.py

@@ -1,6 +1,5 @@
 import torch
 import torch.nn as nn
-import logging
 
 import gym.spaces as spaces
 import numpy as np
@@ -13,7 +12,9 @@
 from rlberry.utils.factory import load
 from typing import Optional
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 class A2CAgent(AgentWithSimplePolicy):

rlberry/agents/torch/dqn/dqn.py

@@ -1,4 +1,3 @@
-import logging
 import inspect
 
 import numpy as np
@@ -19,7 +18,9 @@
 from typing import Callable, Optional, Union
 
 
-logger = logging.getLogger(__name__)
+import rlberry
+
+logger = rlberry.logger
 
 
 def default_q_net_fn(env, **kwargs):