documentation: Minor updates to Clarify API documentation

Updated the docstring of run_bias method and state clearly that it
computes both the pretraining bias metrics and the posttraining bias
metrics.
Updated the docstring of the constructor to list the supported dataset
types, and add a validation at the entrypoint to fail fast if the user
provided an unsupported dataset type.

Author: xgchena

src/sagemaker/clarify.py

@@ -48,12 +48,17 @@ def __init__(
             headers (list[str]): A list of column names in the input dataset.
             features (str): JSONPath for locating the feature columns for bias metrics if the
                 dataset format is JSONLines.
-            dataset_type (str): Format of the dataset. Valid values are "text/csv" for CSV
-                and "application/jsonlines" for JSONLines.
+            dataset_type (str): Format of the dataset. Valid values are "text/csv" for CSV,
+                "application/jsonlines" for JSONLines, and "application/x-parquet" for Parquet.
             s3_data_distribution_type (str): Valid options are "FullyReplicated" or
                 "ShardedByS3Key".
             s3_compression_type (str): Valid options are "None" or "Gzip".
         """
+        if dataset_type not in ["text/csv", "application/jsonlines", "application/x-parquet"]:
+            raise ValueError(
+                f"Invalid dataset_type '{dataset_type}'."
+                f" Please check the API documentation for the supported dataset types."
+            )
         self.s3_data_input_path = s3_data_input_path
         self.s3_output_path = s3_output_path
         self.s3_data_distribution_type = s3_data_distribution_type
@@ -508,7 +513,7 @@ def run_pre_training_bias(
         kms_key=None,
         experiment_config=None,
     ):
-        """Runs a ProcessingJob to compute the requested bias 'methods' of the input data.
+        """Runs a ProcessingJob to compute the requested pre-training bias 'methods' of the input data.
 
         Computes the requested methods that compare 'methods' (e.g. fraction of examples) for the
         sensitive group vs the other examples.
@@ -562,7 +567,7 @@ def run_post_training_bias(
         kms_key=None,
         experiment_config=None,
     ):
-        """Runs a ProcessingJob to compute the requested bias 'methods' of the model predictions.
+        """Runs a ProcessingJob to compute the requested post-training bias 'methods' of the model predictions.
 
         Spins up a model endpoint, runs inference over the input example in the
         's3_data_input_path' to obtain predicted labels. Computes a the requested methods that
@@ -635,10 +640,9 @@ def run_bias(
     ):
         """Runs a ProcessingJob to compute the requested bias 'methods' of the model predictions.
 
-        Spins up a model endpoint, runs inference over the input example in the
-        's3_data_input_path' to obtain predicted labels. Computes a the requested methods that
-        compare 'methods' (e.g. accuracy, precision, recall) for the sensitive group vs the other
-        examples.
+        The job will compute the metrics of both the pre-training methods and the post-training method.
+        To calculate post-training methods, it needs to spin up a model endpoint, runs inference over
+        the input example in the 's3_data_input_path' to obtain predicted labels.
 
         Args:
             data_config (:class:`~sagemaker.clarify.DataConfig`): Config of the input/output data.

tests/unit/test_clarify.py

@@ -68,6 +68,15 @@ def test_data_config():
     assert "FullyReplicated" == data_config.s3_data_distribution_type
 
 
+def test_invalid_data_config():
+    with pytest.raises(ValueError, match=r"^Invalid dataset_type"):
+        DataConfig(
+            s3_data_input_path="s3://bucket/inputpath",
+            s3_output_path="s3://bucket/outputpath",
+            dataset_type="whatnot_type",
+        )
+
+
 def test_data_bias_config():
     label_values = [1]
     facet_name = "F1"