fix label column incorrectly added by feature extractors in RAI Visio…

…n dashboard for automl models (microsoft#2532)

* fix label column incorrectly added by feature extractors in RAI Vision dashboard for automl models

* fixup

responsibleai_vision/responsibleai_vision/utils/feature_extractors.py

@@ -12,13 +12,15 @@
 from tqdm import tqdm
 
 from responsibleai.feature_metadata import FeatureMetadata
-from responsibleai_vision.common.constants import ExtractedFeatures
+from responsibleai_vision.common.constants import (ExtractedFeatures,
+                                                   ImageColumns)
 from responsibleai_vision.utils.image_reader import (
     get_all_exif_feature_names, get_image_from_path,
     get_image_pointer_from_path)
 
 MEAN_PIXEL_VALUE = ExtractedFeatures.MEAN_PIXEL_VALUE.value
 MAX_CUSTOM_LEN = 100
+IMAGE_DETAILS = ImageColumns.IMAGE_DETAILS.value
 
 
 def extract_features(image_dataset: pd.DataFrame,
@@ -58,6 +60,8 @@ def extract_features(image_dataset: pd.DataFrame,
     start_meta_index = 2
     if isinstance(target_column, list):
         start_meta_index = len(target_column) + 1
+    if IMAGE_DETAILS in column_names:
+        start_meta_index += 1
     for j in range(start_meta_index, image_dataset.shape[1]):
         if has_dropped_features and column_names[j] in dropped_features:
             continue

responsibleai_vision/tests/rai_vision_insights_validator.py

@@ -31,6 +31,13 @@ def validate_rai_vision_insights(
         pd.testing.assert_frame_equal(rai_vision_test, test_data)
     assert rai_vision_insights.target_column == target_column
     assert rai_vision_insights.task_type == task_type
+    # make sure label column not in _ext_test extracted features data
+    assert target_column not in rai_vision_insights._ext_features
+    # also not in last column of _ext_test, which is prone to happen
+    # if incorrect number of metadata columns specified in
+    # feature_extractors call
+    first_row = rai_vision_insights._ext_test[0]
+    assert not isinstance(first_row[len(first_row) - 1], list)
 
 
 def run_and_validate_serialization(

responsibleai_vision/tests/test_feature_extractors.py

@@ -1,6 +1,7 @@
 # Copyright (c) Microsoft Corporation
 # Licensed under the MIT License.
 
+import pytest
 from common_vision_utils import (load_flowers_dataset, load_fridge_dataset,
                                  load_fridge_object_detection_dataset,
                                  load_imagenet_dataset)
@@ -39,8 +40,10 @@ def extract_dataset_features(data, feature_metadata=None):
 
 
 class TestFeatureExtractors(object):
-    def test_extract_features_fridge_object_detection(self):
-        data = load_fridge_object_detection_dataset(automl_format=False)
+    @pytest.mark.parametrize("automl_format", [True, False])
+    def test_extract_features_fridge_object_detection(self, automl_format):
+        data = load_fridge_object_detection_dataset(
+            automl_format=automl_format)
         extracted_features, feature_names = extract_dataset_features(data)
         expected_feature_names = [MEAN_PIXEL_VALUE] + FRIDGE_METADATA_FEATURES
         validate_extracted_features(extracted_features, feature_names,