feat: add feature processor APIs to public doc (#4055)

* Adding feature processor APIs to public doc

* Minor fix for the path of feature processor

* Add spark dependencies to doc requirements

* Fix the path

* Remove iceberg data source to resolve build error

* Remove feature_processor doc for now

* Add feature_processor back

* Rename the feature_processor file

* Modify the feature processor path in test

* Fix the path of feature processor

* r

* Fix the doc string error by removing the example block

* r

* r

* Fix the black format validation

* Add newline to requirement txt file

* Fix the doc build failure

* Add feature processor file

* Fix the test

* Remove requirements

* mock pyspark import

* Add more more import mocks

* Add py4j for mock

* Fix black validation failure

Author: can-sun

doc/api/prep_data/feature_store.rst

@@ -120,3 +120,41 @@ Feature Store
 .. autoclass:: sagemaker.feature_store.feature_store.FeatureStore
     :members:
     :show-inheritance:
+
+
+@feature_processor Decorator
+****************************
+
+.. autodecorator:: sagemaker.feature_store.feature_processor.feature_processor
+
+
+Feature Processor Data Source
+*****************************
+
+.. autoclass:: sagemaker.feature_store.feature_processor.FeatureGroupDataSource
+    :members:
+    :show-inheritance:
+
+.. autoclass:: sagemaker.feature_store.feature_processor.CSVDataSource
+    :members:
+    :show-inheritance:
+
+.. autoclass:: sagemaker.feature_store.feature_processor.ParquetDataSource
+    :members:
+    :show-inheritance:
+
+
+Feature Processor Scheduler
+***************************
+
+.. automethod:: sagemaker.feature_store.feature_processor.to_pipeline
+
+.. automethod:: sagemaker.feature_store.feature_processor.schedule
+
+.. automethod:: sagemaker.feature_store.feature_processor.execute
+
+.. automethod:: sagemaker.feature_store.feature_processor.delete_schedule
+
+.. automethod:: sagemaker.feature_store.feature_processor.describe
+
+.. automethod:: sagemaker.feature_store.feature_processor.list_pipelines

doc/conf.py

@@ -109,6 +109,8 @@
 # autosectionlabel
 autosectionlabel_prefix_document = True
 
+autodoc_mock_imports = ["pyspark", "feature_store_pyspark", "py4j"]
+
 
 def setup(app):
     sys.stdout.write("Generating JumpStart model table...")

src/sagemaker/feature_store/feature_processor/feature_processor.py

@@ -49,7 +49,10 @@ def feature_processor(
     names nand types in [params: Dict[str, Any], spark: SparkSession]. Outputs: a single return
     value of type pyspark.sql.DataFrame. The function can have any name.
 
-    Example:
+    **Example:**
+
+    .. code-block:: python
+
         @feature_processor(
             inputs=[FeatureGroupDataSource("input-fg"), CSVDataSource("s3://bucket/prefix)],
             output='arn:aws:sagemaker:us-west-2:123456789012:feature-group/output-fg'
@@ -60,7 +63,10 @@ def transform(
         ) -> DataFrame:
             return ...
 
-    More concisely:
+    **More concisely:**
+
+    .. code-block:: python
+
         @feature_processor(
             inputs=[FeatureGroupDataSource("input-fg"), CSVDataSource("s3://bucket/prefix)],
             output='arn:aws:sagemaker:us-west-2:123456789012:feature-group/output-fg'