added data types in _AnalysisConfigGenerator methods

Author: aws-byeldos

src/sagemaker/clarify.py

@@ -25,6 +25,8 @@
 
 import tempfile
 from abc import ABC, abstractmethod
+from typing import List, Union
+
 from sagemaker import image_uris, s3, utils
 from sagemaker.processing import ProcessingInput, ProcessingOutput, Processor
 
@@ -971,7 +973,7 @@ def _run(
     def run_pre_training_bias(
         self,
         data_config,
-        bias_config,
+        data_bias_config,
         methods="all",
         wait=True,
         logs=True,
@@ -986,7 +988,7 @@ def run_pre_training_bias(
 
         Args:
             data_config (:class:`~sagemaker.clarify.DataConfig`): Config of the input/output data.
-            bias_config (:class:`~sagemaker.clarify.BiasConfig`): Config of sensitive groups.
+            data_bias_config (:class:`~sagemaker.clarify.BiasConfig`): Config of sensitive groups.
             methods (str or list[str]): Selects a subset of potential metrics:
                 ["`CI <https://docs.aws.amazon.com/sagemaker/latest/dg/clarify-bias-metric-class-imbalance.html>`_",
                 "`DPL <https://docs.aws.amazon.com/sagemaker/latest/dg/clarify-data-bias-metric-true-label-imbalance.html>`_",
@@ -1022,7 +1024,7 @@ def run_pre_training_bias(
         """  # noqa E501  # pylint: disable=c0301
         analysis_config = _AnalysisConfigGenerator.bias_pre_training(
             data_config,
-            bias_config,
+            data_bias_config,
             methods
         )
         # when name is either not provided (is None) or an empty string ("")
@@ -1040,7 +1042,7 @@ def run_pre_training_bias(
     def run_post_training_bias(
         self,
         data_config,
-        bias_config,
+        data_bias_config,
         model_config,
         model_predicted_label_config,
         methods="all",
@@ -1060,7 +1062,7 @@ def run_post_training_bias(
 
         Args:
             data_config (:class:`~sagemaker.clarify.DataConfig`): Config of the input/output data.
-            bias_config (:class:`~sagemaker.clarify.BiasConfig`): Config of sensitive groups.
+            data_bias_config (:class:`~sagemaker.clarify.BiasConfig`): Config of sensitive groups.
             model_config (:class:`~sagemaker.clarify.ModelConfig`): Config of the model and its
                 endpoint to be created.
             model_predicted_label_config (:class:`~sagemaker.clarify.ModelPredictedLabelConfig`):
@@ -1103,7 +1105,7 @@ def run_post_training_bias(
         """  # noqa E501  # pylint: disable=c0301
         analysis_config = _AnalysisConfigGenerator.bias_post_training(
             data_config,
-            bias_config,
+            data_bias_config,
             model_predicted_label_config,
             methods,
             model_config
@@ -1314,10 +1316,10 @@ class _AnalysisConfigGenerator:
     @classmethod
     def explainability(
         cls,
-        data_config,
-        model_config,
-        model_scores,
-        explainability_config
+        data_config: DataConfig,
+        model_config: ModelConfig,
+        model_scores: ModelPredictedLabelConfig,
+        explainability_config: ExplainabilityConfig,
     ):
         analysis_config = data_config.get_config()
         predictor_config = model_config.get_predictor_config()
@@ -1358,7 +1360,7 @@ def explainability(
         return cls._common(analysis_config)
 
     @classmethod
-    def bias_pre_training(cls, data_config, bias_config, methods):
+    def bias_pre_training(cls, data_config: DataConfig, bias_config: BiasConfig, methods: Union[str, List[str]]):
         analysis_config = {
             **data_config.get_config(),
             **bias_config.get_config(),
@@ -1369,11 +1371,11 @@ def bias_pre_training(cls, data_config, bias_config, methods):
     @classmethod
     def bias_post_training(
         cls,
-        data_config,
-        bias_config,
-        model_predicted_label_config,
-        methods,
-        model_config
+        data_config: DataConfig,
+        bias_config: BiasConfig,
+        model_predicted_label_config: ModelPredictedLabelConfig,
+        methods: Union[str, List[str]],
+        model_config: ModelConfig,
     ):
         analysis_config = {
             **data_config.get_config(),
@@ -1391,12 +1393,12 @@ def bias_post_training(
     @classmethod
     def bias(
         cls,
-        data_config,
-        bias_config,
-        model_config,
-        model_predicted_label_config,
-        pre_training_methods="all",
-        post_training_methods="all",
+        data_config: DataConfig,
+        bias_config: BiasConfig,
+        model_config: ModelConfig,
+        model_predicted_label_config: ModelPredictedLabelConfig,
+        pre_training_methods: Union[str, List[str]] = "all",
+        post_training_methods: Union[str, List[str]] = "all",
     ):
         analysis_config = {
             **data_config.get_config(),

tests/unit/test_clarify.py

@@ -1382,18 +1382,18 @@ def test_analysis_config_generator_for_bias(data_config, data_bias_config, model
     )
     expected = {'dataset_type': 'text/csv',
                 'facet': [{'name_or_index': 'F1'}],
-                           'group_variable': 'F2',
-                           'headers': ['Label', 'F1', 'F2', 'F3', 'F4'],
-                           'joinsource_name_or_index': 'F4',
-                           'label': 'Label',
-                           'label_values_or_threshold': [1],
-                           'methods': {
-                               'report': {'name': 'report', 'title': 'Analysis Report'},
-                               'post_training_bias': {'methods': 'all'},
-                               'pre_training_bias': {'methods': 'all'}},
-                           'predictor': {'initial_instance_count': 1,
-                                         'instance_type': 'ml.c5.xlarge',
-                                         'label_headers': ['success'],
-                                         'model_name': 'xgboost-model',
-                                         'probability': 'pr'}}
+                'group_variable': 'F2',
+                'headers': ['Label', 'F1', 'F2', 'F3', 'F4'],
+                'joinsource_name_or_index': 'F4',
+                'label': 'Label',
+                'label_values_or_threshold': [1],
+                'methods': {
+                    'report': {'name': 'report', 'title': 'Analysis Report'},
+                    'post_training_bias': {'methods': 'all'},
+                    'pre_training_bias': {'methods': 'all'}},
+                'predictor': {'initial_instance_count': 1,
+                              'instance_type': 'ml.c5.xlarge',
+                              'label_headers': ['success'],
+                              'model_name': 'xgboost-model',
+                              'probability': 'pr'}}
     assert actual == expected