Grouped inferencing statements together in apply methods for easier readability (feast-dev#1667)

mavysavydav · Mwad22 · commit d71e4c50b06b · 2021-07-07T17:27:47.000-04:00
* grouped inferencing statements together

Signed-off-by: David Y Liu &lt;davidyliuliu@gmail.com&gt;

* update in testing

Signed-off-by: David Y Liu &lt;davidyliuliu@gmail.com&gt;
Signed-off-by: Mwad22 &lt;51929507+Mwad22@users.noreply.github.com&gt;
diff --git a/sdk/python/feast/feature_store.py b/sdk/python/feast/feature_store.py
@@ -31,8 +31,8 @@
 )
 from feast.feature_view import FeatureView
 from feast.inference import (
-    infer_entity_value_type_from_feature_views,
     update_data_sources_with_inferred_event_timestamp_col,
+    update_entities_with_inferred_types_from_feature_views,
 )
 from feast.infra.provider import Provider, RetrievalJob, get_provider
 from feast.online_response import OnlineResponse, _infer_online_entity_rows
@@ -228,8 +228,11 @@ def apply(
         assert isinstance(objects, list)
 
         views_to_update = [ob for ob in objects if isinstance(ob, FeatureView)]
-        entities_to_update = infer_entity_value_type_from_feature_views(
-            [ob for ob in objects if isinstance(ob, Entity)], views_to_update
+        entities_to_update = [ob for ob in objects if isinstance(ob, Entity)]
+
+        # Make inferences
+        update_entities_with_inferred_types_from_feature_views(
+            entities_to_update, views_to_update
         )
         update_data_sources_with_inferred_event_timestamp_col(
             [view.input for view in views_to_update]
diff --git a/sdk/python/feast/inference.py b/sdk/python/feast/inference.py
@@ -8,9 +8,9 @@
 from feast.value_type import ValueType
 
 
-def infer_entity_value_type_from_feature_views(
+def update_entities_with_inferred_types_from_feature_views(
     entities: List[Entity], feature_views: List[FeatureView]
-) -> List[Entity]:
+) -> None:
     """
     Infer entity value type by examining schema of feature view input sources
     """
@@ -57,8 +57,6 @@ def infer_entity_value_type_from_feature_views(
 
                 entity.value_type = inferred_value_type
 
-    return entities
-
 
 def update_data_sources_with_inferred_event_timestamp_col(
     data_sources: List[Union[BigQuerySource, FileSource]],
diff --git a/sdk/python/feast/repo_operations.py b/sdk/python/feast/repo_operations.py
@@ -14,8 +14,8 @@
 from feast import Entity, FeatureTable
 from feast.feature_view import FeatureView
 from feast.inference import (
-    infer_entity_value_type_from_feature_views,
     update_data_sources_with_inferred_event_timestamp_col,
+    update_entities_with_inferred_types_from_feature_views,
 )
 from feast.infra.provider import get_provider
 from feast.names import adjectives, animals
@@ -131,13 +131,19 @@ def apply_total(repo_config: RepoConfig, repo_path: Path):
     registry._initialize_registry()
     sys.dont_write_bytecode = True
     repo = parse_repo(repo_path)
-    repo = ParsedRepo(
-        feature_tables=repo.feature_tables,
-        entities=infer_entity_value_type_from_feature_views(
-            repo.entities, repo.feature_views
-        ),
-        feature_views=repo.feature_views,
+    data_sources = [t.input for t in repo.feature_views]
+
+    # Make sure the data source used by this feature view is supported by Feast
+    for data_source in data_sources:
+        data_source.validate()
+
+    # Make inferences
+    update_entities_with_inferred_types_from_feature_views(
+        repo.entities, repo.feature_views
     )
+    update_data_sources_with_inferred_event_timestamp_col(data_sources)
+    for view in repo.feature_views:
+        view.infer_features_from_input_source()
 
     sys.dont_write_bytecode = False
     for entity in repo.entities:
@@ -151,16 +157,6 @@ def apply_total(repo_config: RepoConfig, repo_path: Path):
     for t in repo.feature_views:
         repo_table_names.add(t.name)
 
-    data_sources = [t.input for t in repo.feature_views]
-
-    # Make sure the data source used by this feature view is supported by Feast
-    for data_source in data_sources:
-        data_source.validate()
-
-    update_data_sources_with_inferred_event_timestamp_col(data_sources)
-    for view in repo.feature_views:
-        view.infer_features_from_input_source()
-
     tables_to_delete = []
     for registry_table in registry.list_feature_tables(project=project):
         if registry_table.name not in repo_table_names:
diff --git a/sdk/python/tests/test_inference.py b/sdk/python/tests/test_inference.py
@@ -9,12 +9,14 @@
 from feast.errors import RegistryInferenceFailure
 from feast.feature_view import FeatureView
 from feast.inference import (
-    infer_entity_value_type_from_feature_views,
     update_data_sources_with_inferred_event_timestamp_col,
+    update_entities_with_inferred_types_from_feature_views,
 )
 
 
-def test_infer_entity_value_type_from_feature_views(simple_dataset_1, simple_dataset_2):
+def test_update_entities_with_inferred_types_from_feature_views(
+    simple_dataset_1, simple_dataset_2
+):
     with prep_file_source(
         df=simple_dataset_1, event_timestamp_column="ts_1"
     ) as file_source, prep_file_source(
@@ -24,22 +26,23 @@ def test_infer_entity_value_type_from_feature_views(simple_dataset_1, simple_dat
         fv1 = FeatureView(name="fv1", entities=["id"], input=file_source, ttl=None,)
         fv2 = FeatureView(name="fv2", entities=["id"], input=file_source_2, ttl=None,)
 
-        actual_1 = infer_entity_value_type_from_feature_views(
-            [Entity(name="id")], [fv1]
-        )
-        actual_2 = infer_entity_value_type_from_feature_views(
-            [Entity(name="id")], [fv2]
-        )
-        assert actual_1 == [Entity(name="id", value_type=ValueType.INT64)]
-        assert actual_2 == [Entity(name="id", value_type=ValueType.STRING)]
+        actual_1 = Entity(name="id")
+        actual_2 = Entity(name="id")
+
+        update_entities_with_inferred_types_from_feature_views([actual_1], [fv1])
+        update_entities_with_inferred_types_from_feature_views([actual_2], [fv2])
+        assert actual_1 == Entity(name="id", value_type=ValueType.INT64)
+        assert actual_2 == Entity(name="id", value_type=ValueType.STRING)
 
         with pytest.raises(RegistryInferenceFailure):
             # two viable data types
-            infer_entity_value_type_from_feature_views([Entity(name="id")], [fv1, fv2])
+            update_entities_with_inferred_types_from_feature_views(
+                [Entity(name="id")], [fv1, fv2]
+            )
 
 
 @pytest.mark.integration
-def test_infer_event_timestamp_column_for_data_source(simple_dataset_1):
+def test_update_data_sources_with_inferred_event_timestamp_col(simple_dataset_1):
     df_with_two_viable_timestamp_cols = simple_dataset_1.copy(deep=True)
     df_with_two_viable_timestamp_cols["ts_2"] = simple_dataset_1["ts_1"]
 
