change: enable attribute-defined-outside-init Pylint check

.pylintrc

@@ -84,7 +84,6 @@ disable=
     line-too-long, # We let Flake8 take care of this # TODO: Fix these and stop relying on flake8
     len-as-condition, # TODO: Enable this check once pylint 2.4.0 is released and consumed due to the fix in https://github.com/PyCQA/pylint/issues/2684
     import-error, # Since we run Pylint before any of our builds in tox, this will always fail
-    attribute-defined-outside-init, # TODO: Fix scope
     protected-access, # TODO: Fix access
     abstract-method, # TODO: Fix abstract methods
     wrong-import-order, # TODO: Fix import order

src/sagemaker/amazon/amazon_estimator.py

@@ -55,7 +55,7 @@ def __init__(
         data_location = data_location or "s3://{}/sagemaker-record-sets/".format(
             self.sagemaker_session.default_bucket()
         )
-        self.data_location = data_location
+        self._data_location = data_location
 
     def train_image(self):
         return get_image_uri(

src/sagemaker/analytics.py

@@ -37,6 +37,9 @@ class AnalyticsMetricsBase(with_metaclass(ABCMeta, object)):
     Understands common functionality like persistence and caching.
     """
 
+    def __init__(self):
+        self._dataframe = None
+
     def export_csv(self, filename):
         """Persists the analytics dataframe to a file.
 
@@ -88,6 +91,9 @@ def __init__(self, hyperparameter_tuning_job_name, sagemaker_session=None):
         sagemaker_session = sagemaker_session or Session()
         self._sage_client = sagemaker_session.sagemaker_client
         self._tuning_job_name = hyperparameter_tuning_job_name
+        self._tuning_job_describe_result = None
+        self._training_job_summaries = None
+        super(HyperparameterTuningJobAnalytics, self).__init__()
         self.clear_cache()
 
     @property
@@ -240,6 +246,8 @@ def __init__(
             self._metric_names = metric_names
         else:
             self._metric_names = self._metric_names_for_training_job()
+
+        super(TrainingJobAnalytics, self).__init__()
         self.clear_cache()
 
     @property

src/sagemaker/estimator.py

@@ -166,6 +166,7 @@ def __init__(
         self.output_path = output_path
         self.output_kms_key = output_kms_key
         self.latest_training_job = None
+        self.deploy_instance_type = None
 
         self._compiled_models = {}
 
@@ -1086,6 +1087,8 @@ def __init__(
         self.image_name = image_name
         self._enable_network_isolation = enable_network_isolation
 
+        self.uploaded_code = None
+
         self._hyperparameters = hyperparameters or {}
 
     def enable_network_isolation(self):

src/sagemaker/local/entities.py

@@ -75,7 +75,7 @@ def start(self, input_data_config, output_data_config, hyperparameters, job_name
         self.model_artifacts = self.container.train(
             input_data_config, output_data_config, hyperparameters, job_name
         )
-        self.end = datetime.datetime.now()
+        self.end_time = datetime.datetime.now()
         self.state = self._COMPLETED
 
     def describe(self):
@@ -110,6 +110,9 @@ def __init__(self, transform_job_name, model_name, local_session=None):
         self.start_time = None
         self.end_time = None
         self.batch_strategy = None
+        self.transform_resources = None
+        self.input_data = None
+        self.output_data = None
         self.environment = {}
         self.state = _LocalTransformJob._CREATING
 

src/sagemaker/model.py

@@ -100,6 +100,7 @@ def __init__(
         self.vpc_config = vpc_config
         self.sagemaker_session = sagemaker_session
         self._model_name = None
+        self.endpoint_name = None
         self._is_compiled_model = False
         self._enable_network_isolation = enable_network_isolation
 

src/sagemaker/pipeline.py

@@ -51,6 +51,7 @@ def __init__(
         self.vpc_config = vpc_config
         self.sagemaker_session = sagemaker_session
         self._model_name = None
+        self.endpoint_name = None
 
     def pipeline_container_def(self, instance_type):
         """Return a dict created by ``sagemaker.pipeline_container_def()`` for deploying this model to a specified

src/sagemaker/tuner.py

@@ -247,6 +247,7 @@ def __init__(
         self.latest_tuning_job = None
         self.warm_start_config = warm_start_config
         self.early_stopping_type = early_stopping_type
+        self.static_hyperparameters = None
 
     def _prepare_for_training(self, job_name=None, include_cls_metadata=False):
         if job_name is not None: