add output of anomaly score and sorted according to spec, related to #21 and #22

tests/test_detection.py

@@ -67,9 +67,9 @@ def test_measure_length(single_img_path):
         pixels_per_metric_X,
         pixels_per_metric_Y,
     )
-    assert len(data) == 18
-    assert isinstance(data[0], str)
-    assert float(data[0].replace(",", ".")) == pytest.approx(1266.932)
+    assert len(data["sensor_sizes"]) == 18
+    assert isinstance(data["sensor_sizes"][0], str)
+    assert float(data["sensor_sizes"][0].replace(",", ".")) == pytest.approx(1266.932)
     img_left = imgs["left"]
     assert 235 < img_left.shape[0] < 260
     assert 910 < img_left.shape[1] < 960
@@ -89,20 +89,20 @@ def test_isolated_pipeline(results_folder, path_img_with_failure_TrainedModel):
     DETECTION_MODELS = dopt_sensor_anomalies._find_paths.get_detection_models(MODEL_FOLDER)
     assert DETECTION_MODELS["left"].exists()
     assert DETECTION_MODELS["right"].exists()
-    data_csv, sensor_images = detect.measure_length(
+    export_data, sensor_images = detect.measure_length(
         path_img_with_failure_TrainedModel,
         pixels_per_metric_X,
         pixels_per_metric_Y,
     )
 
     # measured sizes
-    assert len(data_csv) == 18
+    assert len(export_data["sensor_sizes"]) == 18
     assert sensor_images["left"] is not None
     assert sensor_images["right"] is not None
     detect.anomaly_detection(
         img_path=path_img_with_failure_TrainedModel,
         detection_models=DETECTION_MODELS,
-        data_csv=data_csv,
+        export_data=export_data,
         sensor_images=sensor_images,
         anomaly_threshold=constants.ANOMALY_THRESHOLD_DEFAULT,
     )