Enable more rules - Part 3 (#225)

* Enable more rules

Signed-off-by: Ashwin Vaidya <[email protected]>

* Enable some more flake8 rules

Signed-off-by: Ashwin Vaidya <[email protected]>

* Enable more rules

Signed-off-by: Ashwin Vaidya <[email protected]>

---------

Signed-off-by: Ashwin Vaidya <[email protected]>

Author: ashwinvaidya17

model_api/python/model_api/adapters/onnx_adapter.py

@@ -42,11 +42,11 @@ def __init__(self, model: str, ort_options: dict = {}):
         loaded_model = onnx.load(model)
 
         inferred_model = SymbolicShapeInference.infer_shapes(
-            loaded_model,
-            int(sys.maxsize / 2),
-            False,
-            False,
-            False,
+            in_mp=loaded_model,
+            int_max=int(sys.maxsize / 2),
+            auto_merge=False,
+            guess_output_rank=False,
+            verbose=False,
         )
 
         self.session = ort.InferenceSession(

model_api/python/model_api/adapters/openvino_adapter.py

@@ -233,8 +233,7 @@ def get_input_layers(self):
                 input_layout,
                 input.get_element_type().get_type_name(),
             )
-        inputs = self._get_meta_from_ngraph(inputs)
-        return inputs
+        return self._get_meta_from_ngraph(inputs)
 
     def get_layout_for_input(self, input, shape=None) -> str:
         input_layout = ""
@@ -263,8 +262,7 @@ def get_output_layers(self):
                 list(output_shape),
                 precision=output.get_element_type().get_type_name(),
             )
-        outputs = self._get_meta_from_ngraph(outputs)
-        return outputs
+        return self._get_meta_from_ngraph(outputs)
 
     def reshape_model(self, new_shape):
         new_shape = {

model_api/python/model_api/adapters/utils.py

@@ -273,15 +273,14 @@ def crop_resize_graph(input: Output, size):
 
     target_size = list(size)
     target_size.reverse()
-    resized_image = opset.interpolate(
+    return opset.interpolate(
         cropped_frame,
         target_size,
         scales=np.array([0.0, 0.0], dtype=np.float32),
         axes=[h_axis, w_axis],
         mode="linear",
         shape_calculation_mode="sizes",
     )
-    return resized_image
 
 
 def resize_image_graph(

model_api/python/model_api/models/anomaly.py

@@ -126,8 +126,7 @@ def parameters(cls) -> dict:
     def _normalize(self, tensor: np.ndarray, threshold: float) -> np.ndarray:
         """Currently supports only min-max normalization."""
         normalized = ((tensor - threshold) / self.normalization_scale) + 0.5
-        normalized = np.clip(normalized, 0, 1)
-        return normalized
+        return np.clip(normalized, 0, 1)
 
     @staticmethod
     def _get_boxes(mask: np.ndarray) -> np.ndarray:

model_api/python/model_api/models/classification.py

@@ -8,6 +8,7 @@
 import copy
 import json
 from collections import defaultdict
+from pathlib import Path
 
 import numpy as np
 from openvino.preprocess import PrePostProcessor
@@ -102,7 +103,7 @@ def __init__(self, inference_adapter: InferenceAdapter, configuration: dict = {}
             self.load()
 
     def _load_labels(self, labels_file):
-        with open(labels_file) as f:
+        with Path(labels_file).open() as f:
             labels = []
             for s in f:
                 begin_idx = s.find(" ")
@@ -417,8 +418,7 @@ def get_predecessors(lbl, candidates):
                 if new_lbl not in output_labels:
                     output_labels.append(new_lbl)
 
-        output_predictions = [(self.label_to_idx[lbl], lbl, label_to_prob[lbl]) for lbl in sorted(output_labels)]
-        return output_predictions
+        return [(self.label_to_idx[lbl], lbl, label_to_prob[lbl]) for lbl in sorted(output_labels)]
 
 
 class ProbabilisticLabelsResolver(GreedyLabelsResolver):

model_api/python/model_api/models/instance_segmentation.py

@@ -45,7 +45,7 @@ def parameters(cls):
         )
         return parameters
 
-    def _get_outputs(self):
+    def _get_outputs(self):  # noqa: C901 TODO: Fix this method to reduce complexity
         if self.is_segmentoly:
             return self._get_segmentoly_outputs()
         filtered_names = []
@@ -82,7 +82,7 @@ def _get_outputs(self):
         if len(outputs) == 3:
             _append_xai_names(self.outputs, outputs)
             return outputs
-        self.raise_error(f"Unexpected outputs: {self.outputs}")
+        return self.raise_error(f"Unexpected outputs: {self.outputs}")
 
     def _get_segmentoly_outputs(self):
         outputs = {}

model_api/python/model_api/models/model.py

@@ -105,11 +105,9 @@ def get_model_class(cls, name):
         for subclass in subclasses:
             if name.lower() == subclass.__model__.lower():
                 return subclass
-        cls.raise_error(
-            'There is no model with name "{}" in list: {}'.format(
-                name,
-                ", ".join([subclass.__model__ for subclass in subclasses]),
-            ),
+        return cls.raise_error(
+            f"There is no model with name {name} in list: "
+            f"{', '.join([subclass.__model__ for subclass in subclasses])}",
         )
 
     @classmethod
@@ -214,8 +212,7 @@ def parameters(cls):
         Returns:
             - the dictionary with defined wrapper data parameters
         """
-        parameters = {}
-        return parameters
+        return {}
 
     def _load_config(self, config):
         """Reads the configuration and creates data attributes

model_api/python/model_api/models/result_types.py

@@ -243,5 +243,4 @@ def __str__(self):
 def _array_shape_to_str(array: np.ndarray | None) -> str:
     if array is not None:
         return f"[{','.join(str(i) for i in array.shape)}]"
-    else:
-        return "[]"
+    return "[]"

model_api/python/model_api/models/segmentation.py

@@ -209,14 +209,13 @@ def postprocess(self, outputs, meta):
         input_image_width = meta["original_shape"][1]
         result = outputs[self.output_blob_name].squeeze()
         result = 1 / (1 + np.exp(-result))
-        result = cv2.resize(
+        return cv2.resize(
             result,
             (input_image_width, input_image_height),
             0,
             0,
             interpolation=cv2.INTER_NEAREST,
         )
-        return result
 
 
 _feature_vector_name = "feature_vector"

model_api/python/model_api/models/ssd.py

@@ -61,7 +61,8 @@ def _get_output_parser(
             return parser
         except ValueError:
             pass
-        self.raise_error("Unsupported model outputs")
+        msg = "Unsupported model outputs"
+        raise ValueError(msg)
 
     def _parse_outputs(self, outputs):
         return self.output_parser(outputs)
@@ -156,8 +157,7 @@ def __call__(self, outputs):
             labels = np.full(len(bboxes), self.default_label, dtype=bboxes.dtype)
         labels = labels.squeeze(0)
 
-        detections = [Detection(*bbox, score, label) for label, score, bbox in zip(labels, scores, bboxes)]
-        return detections
+        return [Detection(*bbox, score, label) for label, score, bbox in zip(labels, scores, bboxes)]
 
 
 _bbox_area_threshold = 1.0