refactor setting loggers

automatminer/automl/adaptors.py

@@ -78,7 +78,7 @@ def __init__(self, logger=True, **tpot_kwargs):
         self._fitted_target = None
         self._backend = None
         self._features = None
-        self._logger = self.get_logger(logger)
+        self.logger = logger
 
     @log_progress(AMM_LOG_FIT_STR)
     @set_fitted
@@ -245,7 +245,7 @@ class SinglePipelineAdaptor(DFMLAdaptor, LoggableMixin):
 
     def __init__(self, regressor, classifier, logger=True):
         self.mode = None
-        self._logger = self.get_logger(logger)
+        self.logger = logger
         self._regressor = regressor
         self._classifier = classifier
         self._features = None
@@ -295,4 +295,3 @@ def features(self):
     def fitted_target(self):
         return self._fitted_target
 
-

automatminer/base.py

@@ -19,34 +19,47 @@ def logger(self):
         If the logger is None, the logging calls will be redirected to a dummy
         logger that has no output.
         """
-        if hasattr(self, "_logger"):
-            return self._logger
-        else:
-            raise AttributeError("Loggable object has no _logger attribute!")
+        return self._logger
+
+    @logger.setter
+    def logger(self, new_logger):
+        """Set a new logger.
+
+        Args:
+            new_logger (Logger, bool): A boolean or custom logger object to use
+            for logging. Alternatively, if set to True, the default automatminer
+            logger will be used. If set to False, then no logging will occur.
+        """
+        new_logger = self.get_logger(new_logger)
+        assert isinstance(
+            new_logger, logging.Logger
+        ), "The new logger must be an instance of the logger class."
+        self._logger = new_logger
+        if hasattr(self, "autofeaturizer"):
+            for x in ["autofeaturizer", "cleaner", "reducer", "learner"]:
+                getattr(self, x)._logger = new_logger
 
     @staticmethod
-    def get_logger(logger, level=None):
-        """Set the class logger.
+    def get_logger(logger):
+        """Handle boolean logger.
         Args:
             logger (Logger, bool): A custom logger object to use for logging.
                 Alternatively, if set to True, the default automatminer logger
                 will be used. If set to False, then no logging will occur.
-            level (int): The log level. For example logging.DEBUG.
         """
         # need comparison to True and False to avoid overwriting Logger objects
         if logger is True:
             logger = logging.getLogger(AMM_LOGGER_BASENAME)
 
             if not logger.handlers:
-                initialize_logger(AMM_LOGGER_BASENAME, level=level)
+                initialize_logger(AMM_LOGGER_BASENAME)
 
         elif logger is False:
             logger = logging.getLogger(AMM_LOGGER_BASENAME + "_null")
 
             if not logger.handlers:
                 initialize_null_logger(AMM_LOGGER_BASENAME)
 
-        logger.setLevel(logging.INFO)
         return logger
 
     @property

automatminer/featurization/core.py

@@ -128,7 +128,7 @@ def __init__(self, cache_src=None, preset=None, featurizers=None,
 
         self.cache_src = cache_src
         self.preset = "express" if preset is None else preset
-        self._logger = self.get_logger(logger)
+        self.logger = logger
         self.featurizers = featurizers
         self.exclude = exclude if exclude else []
         self.functionalize = functionalize

automatminer/pipeline.py

@@ -92,15 +92,13 @@ def __init__(self, autofeaturizer=None, cleaner=None, reducer=None,
                 reducer = config["reducer"]
                 learner = config["learner"]
 
-        self._logger = self.get_logger(logger, level=log_level)
         self.autofeaturizer = autofeaturizer
         self.cleaner = cleaner
         self.reducer = reducer
         self.learner = learner
-        self.autofeaturizer._logger = self.get_logger(logger)
-        self.cleaner._logger = self.get_logger(logger)
-        self.reducer._logger = self.get_logger(logger)
-        self.learner._logger = self.get_logger(logger)
+        self.logger = logger
+        if log_level:
+            self.logger.setLevel(log_level)
         self.pre_fit_df = None
         self.post_fit_df = None
         self.is_fit = False
@@ -340,9 +338,7 @@ def load(cls, filename, logger=True):
         with open(filename, 'rb') as f:
             pipe = pickle.load(f)
 
-        for obj in [pipe, pipe.learner, pipe.reducer, pipe.cleaner,
-                    pipe.autofeaturizer]:
-            obj._logger = cls.get_logger(logger)
+        pipe.logger = logger
 
         pipe.logger.info("Loaded MatPipe from file {}.".format(filename))
         pipe.logger.warning("Only use this model to make predictions (do not "

automatminer/preprocessing/core.py

@@ -78,7 +78,7 @@ def __init__(self, max_na_frac=0.01, feature_na_method="drop",
                  encode_categories=True, encoder='one-hot',
                  drop_na_targets=True, na_method_fit="drop",
                  na_method_transform="fill", logger=True):
-        self._logger = self.get_logger(logger)
+        self.logger = logger
         self.max_na_frac = max_na_frac
         self.feature_na_method = feature_na_method
         self.encoder = encoder
@@ -482,7 +482,7 @@ def __init__(self, reducers=('pca',), corr_threshold=0.95,
         self.n_pca_features = n_pca_features
         self.tree_importance_percentile = tree_importance_percentile
         self.n_rebate_features = n_rebate_features
-        self._logger = self.get_logger(logger)
+        self.logger = logger
         self._keep_features = keep_features or []
         self._remove_features = remove_features or []
         self.removed_features = {}

automatminer/preprocessing/feature_selection.py

@@ -39,7 +39,7 @@ class TreeFeatureReducer(DFTransformer, LoggableMixin):
 
     def __init__(self, mode, importance_percentile=0.95,
                  logger=True, random_state=0):
-        self._logger = self.get_logger(logger)
+        self.logger = logger
         self.mode = mode
         self.importance_percentile = importance_percentile
         self.selected_features = None

automatminer/tests/test_base.py

@@ -77,7 +77,7 @@ class TestLoggableMixin(LoggableMixin):
     """
 
     def __init__(self, logger=True):
-        self._logger = self.get_logger(logger)
+        self.logger = logger
 
 
 class TestBaseTransformers(unittest.TestCase):