Alpha b

.gitignore

@@ -100,3 +100,5 @@ Samples/*
 
 # Exclude backup load
 /Backup_*_COPY.ipynb
+/ruff_check_results.txt
+/ruff_format_results.txt

Archive/API/Python/Example.py

@@ -3,6 +3,7 @@
 import numpy as np
 from pdai import *
 from PIL import Image
+
 pp = pprint.PrettyPrinter(indent=4, width=10)
 
 # Instantiate the PneumoniaModel class
@@ -12,18 +13,18 @@
 pdai_model.load_model()
 
 # Load an image for prediction
-img_path = 'API\\Python\\test sampels\\PNEUMONIA\\person1947_bacteria_4876.jpeg'
+img_path = "API\\Python\\test sampels\\PNEUMONIA\\person1947_bacteria_4876.jpeg"
 img = Image.open(img_path)
-img = img.convert('RGB')  # Convert grayscale to RGB
+img = img.convert("RGB")  # Convert grayscale to RGB
 img = img.resize((280, 300))
 x = np.array(img)
 x = np.expand_dims(x, axis=0)
 
-print('without CLAHE>>>')
+print("without CLAHE>>>")
 # Make a prediction without CLAHE
 result = pdai_model.predict(x)
 pp.pprint(result)
-print('with CLAHE>>>')
+print("with CLAHE>>>")
 # Make a prediction with CLAHE
 result = pdai_model.predict(x, clahe=True)
 pp.pprint(result)

Archive/API/Python/pdai.py

@@ -1,10 +1,12 @@
 import os
-os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
+
+os.environ["TF_CPP_MIN_LOG_LEVEL"] = "3"
 from keras.models import load_model
 from typing import Union, Dict
 import numpy as np
 import cv2
 
+
 class PneumoniaModel:
     def __init__(self, model_path: str, verbose: int = 0):
         """
@@ -18,17 +20,16 @@ def __init__(self, model_path: str, verbose: int = 0):
         self.model = None
         self.verbose = verbose
 
-
     def load_model(self) -> Dict[str, Union[str, None]]:
         """
         Loads the model from the path specified during initialization.
 
         Returns:
-            dict: A dictionary with a "status" key. If the model is loaded successfully, "status" is "success". 
+            dict: A dictionary with a "status" key. If the model is loaded successfully, "status" is "success".
                   If an error occurs, "status" is "error" and an additional "message" key contains the error message.
         """
         try:
-            self.model = None 
+            self.model = None
             self.model = load_model(self.model_path)
             if self.verbose == 1:
                 print("Model loaded successfully.")
@@ -39,7 +40,6 @@ def load_model(self) -> Dict[str, Union[str, None]]:
 
         return {"status": "success"}
 
-
     def predict(self, image: np.ndarray, clahe: bool = False) -> Dict[str, Union[str, float, None]]:
         """
         Makes a prediction using the loaded model on the given image.
@@ -49,30 +49,36 @@ def predict(self, image: np.ndarray, clahe: bool = False) -> Dict[str, Union[str
             clahe (bool, optional): Whether to apply CLAHE to the image before making a prediction. Defaults to False.
 
         Returns:
-            dict: A dictionary with a "status" key. If the prediction is made successfully, "status" is "success", 
-                  and additional "prediction" and "confidence" keys contain the prediction and confidence level. 
+            dict: A dictionary with a "status" key. If the prediction is made successfully, "status" is "success",
+                  and additional "prediction" and "confidence" keys contain the prediction and confidence level.
                   If an error occurs, "status" is "error" and an additional "message" key contains the error message.
         """
         if self.model is None:
             if self.verbose == 1:
                 print("Model not loaded. Call load_model() first.")
-            return {"status": "error", "message": "Model not loaded. Call load_model() first."}
-
+            return {
+                "status": "error",
+                "message": "Model not loaded. Call load_model() first.",
+            }
+
         if image.ndim != 4 or image.shape[3] != 3:
-            return {"status": "error", "message": f"Invalid image format. The image should have three color channels (RGB). Img shape = {image.shape}."}
+            return {
+                "status": "error",
+                "message": f"Invalid image format. The image should have three color channels (RGB). Img shape = {image.shape}.",
+            }
 
         try:
             if clahe:
                 # Create a CLAHE object
-                clahe = cv2.createCLAHE(clipLimit=2, tileGridSize=(8,8))
-                
+                clahe = cv2.createCLAHE(clipLimit=2, tileGridSize=(8, 8))
+
                 b, g, r = cv2.split(image[0])
-                
+
                 # Convert the channels to the appropriate format
                 b = cv2.convertScaleAbs(b)
                 g = cv2.convertScaleAbs(g)
                 r = cv2.convertScaleAbs(r)
-                
+
                 # Apply adaptive histogram equalization to each channel
                 equalized_b = clahe.apply(b)
                 equalized_g = clahe.apply(g)
@@ -93,11 +99,19 @@ def predict(self, image: np.ndarray, clahe: bool = False) -> Dict[str, Union[str
             if np.argmax(prediction) == 0:
                 if self.verbose == 1:
                     print("Prediction: Normal")
-                return {"status": "success", "prediction": "Normal", "confidence": np.max(prediction)}
+                return {
+                    "status": "success",
+                    "prediction": "Normal",
+                    "confidence": np.max(prediction),
+                }
             else:
                 if self.verbose == 1:
                     print("Prediction: Pneumonia")
-                return {"status": "success", "prediction": "Pneumonia", "confidence": np.max(prediction)}
+                return {
+                    "status": "success",
+                    "prediction": "Pneumonia",
+                    "confidence": np.max(prediction),
+                }
         except IndexError as e:
             if self.verbose == 1:
                 print(f"Error making prediction: {str(e)}")

Archive/keras_applications_mod/models/applications_load_weight_test.py

@@ -141,10 +141,7 @@
     ),
 }
 
-TEST_IMAGE_PATH = (
-    "https://storage.googleapis.com/tensorflow/"
-    "keras-applications/tests/elephant.jpg"
-)
+TEST_IMAGE_PATH = "https://storage.googleapis.com/tensorflow/" "keras-applications/tests/elephant.jpg"
 _IMAGENET_CLASSES = 1000
 
 # Add a flag to define which application module file is tested.
@@ -171,9 +168,7 @@ def _get_elephant(target_size):
 class ApplicationsLoadWeightTest(tf.test.TestCase, parameterized.TestCase):
     def assertShapeEqual(self, shape1, shape2):
         if len(shape1) != len(shape2):
-            raise AssertionError(
-                f"Shapes are different rank: {shape1} vs {shape2}"
-            )
+            raise AssertionError(f"Shapes are different rank: {shape1} vs {shape2}")
         if shape1 != shape2:
             raise AssertionError(f"Shapes differ: {shape1} vs {shape2}")
 

Archive/keras_applications_mod/models/applications_test.py

@@ -140,9 +140,7 @@
 class ApplicationsTest(tf.test.TestCase, parameterized.TestCase):
     def assertShapeEqual(self, shape1, shape2):
         if len(shape1) != len(shape2):
-            raise AssertionError(
-                f"Shapes are different rank: {shape1} vs {shape2}"
-            )
+            raise AssertionError(f"Shapes are different rank: {shape1} vs {shape2}")
         for v1, v2 in zip(shape1, shape2):
             if v1 != v2:
                 raise AssertionError(f"Shapes differ: {shape1} vs {shape2}")
@@ -168,9 +166,7 @@ def test_application_notop(self, app, last_dim):
             only_check_last_dim = True
         else:
             only_check_last_dim = False
-        output_shape = _get_output_shape(
-            lambda: app(weights=None, include_top=False)
-        )
+        output_shape = _get_output_shape(lambda: app(weights=None, include_top=False))
         if only_check_last_dim:
             self.assertEqual(output_shape[-1], last_dim)
         else:
@@ -179,28 +175,20 @@ def test_application_notop(self, app, last_dim):
 
     @parameterized.parameters(*MODEL_LIST)
     def test_application_notop_custom_input_shape(self, app, last_dim):
-        output_shape = _get_output_shape(
-            lambda: app(
-                weights="imagenet", include_top=False, input_shape=(224, 224, 3)
-            )
-        )
+        output_shape = _get_output_shape(lambda: app(weights="imagenet", include_top=False, input_shape=(224, 224, 3)))
 
         self.assertEqual(output_shape[-1], last_dim)
 
     @parameterized.parameters(MODEL_LIST)
     def test_application_pooling(self, app, last_dim):
-        output_shape = _get_output_shape(
-            lambda: app(weights=None, include_top=False, pooling="avg")
-        )
+        output_shape = _get_output_shape(lambda: app(weights=None, include_top=False, pooling="avg"))
         self.assertShapeEqual(output_shape, (None, last_dim))
 
     @parameterized.parameters(MODEL_LIST)
     def test_application_classifier_activation(self, app, _):
         if "RegNet" in app.__name__:
             self.skipTest("RegNet models do not support classifier activation")
-        model = app(
-            weights=None, include_top=True, classifier_activation="softmax"
-        )
+        model = app(weights=None, include_top=True, classifier_activation="softmax")
         last_layer_act = model.layers[-1].activation.__name__
         self.assertEqual(last_layer_act, "softmax")
 
@@ -210,30 +198,20 @@ def test_application_variable_input_channels(self, app, last_dim):
             input_shape = (1, None, None)
         else:
             input_shape = (None, None, 1)
-        output_shape = _get_output_shape(
-            lambda: app(
-                weights=None, include_top=False, input_shape=input_shape
-            )
-        )
+        output_shape = _get_output_shape(lambda: app(weights=None, include_top=False, input_shape=input_shape))
         self.assertShapeEqual(output_shape, (None, None, None, last_dim))
         backend.clear_session()
 
         if backend.image_data_format() == "channels_first":
             input_shape = (4, None, None)
         else:
             input_shape = (None, None, 4)
-        output_shape = _get_output_shape(
-            lambda: app(
-                weights=None, include_top=False, input_shape=input_shape
-            )
-        )
+        output_shape = _get_output_shape(lambda: app(weights=None, include_top=False, input_shape=input_shape))
         self.assertShapeEqual(output_shape, (None, None, None, last_dim))
         backend.clear_session()
 
     @parameterized.parameters(*MOBILENET_V3_FOR_WEIGHTS)
-    def test_mobilenet_v3_load_weights(
-        self, mobilenet_class, alpha, minimalistic, include_top
-    ):
+    def test_mobilenet_v3_load_weights(self, mobilenet_class, alpha, minimalistic, include_top):
         mobilenet_class(
             input_shape=(224, 224, 3),
             weights="imagenet",

Archive/keras_applications_mod/models/convnext.py

@@ -36,9 +36,7 @@
 # isort: off
 from tensorflow.python.util.tf_export import keras_export
 
-BASE_WEIGHTS_PATH = (
-    "https://storage.googleapis.com/tensorflow/keras-applications/convnext/"
-)
+BASE_WEIGHTS_PATH = "https://storage.googleapis.com/tensorflow/keras-applications/convnext/"
 
 WEIGHTS_HASHES = {
     "convnext_tiny": (
@@ -239,9 +237,7 @@ def get_config(self):
         return config
 
 
-def ConvNeXtBlock(
-    projection_dim, drop_path_rate=0.0, layer_scale_init_value=1e-6, name=None
-):
+def ConvNeXtBlock(projection_dim, drop_path_rate=0.0, layer_scale_init_value=1e-6, name=None):
     """ConvNeXt block.
 
     References:
@@ -290,9 +286,7 @@ def apply(inputs):
                 name=name + "_layer_scale",
             )(x)
         if drop_path_rate:
-            layer = StochasticDepth(
-                drop_path_rate, name=name + "_stochastic_depth"
-            )
+            layer = StochasticDepth(drop_path_rate, name=name + "_stochastic_depth")
         else:
             layer = layers.Activation("linear", name=name + "_identity")
 
@@ -344,9 +338,7 @@ def Head(num_classes=1000, classifier_activation=None, name=None):
 
     def apply(x):
         x = layers.GlobalAveragePooling2D(name=name + "_head_gap")(x)
-        x = layers.LayerNormalization(
-            epsilon=1e-6, name=name + "_head_layernorm"
-        )(x)
+        x = layers.LayerNormalization(epsilon=1e-6, name=name + "_head_layernorm")(x)
         x = layers.Dense(
             num_classes,
             activation=classifier_activation,
@@ -432,10 +424,7 @@ def ConvNeXt(
         )
 
     if weights == "imagenet" and include_top and classes != 1000:
-        raise ValueError(
-            "If using `weights` as `'imagenet'` with `include_top`"
-            " as true, `classes` should be 1000"
-        )
+        raise ValueError("If using `weights` as `'imagenet'` with `include_top`" " as true, `classes` should be 1000")
 
     # Determine proper input shape.
     input_shape = imagenet_utils.obtain_input_shape(
@@ -462,9 +451,7 @@ def ConvNeXt(
 
     x = inputs
     if include_preprocessing:
-        channel_axis = (
-            3 if backend.image_data_format() == "channels_last" else 1
-        )
+        channel_axis = 3 if backend.image_data_format() == "channels_last" else 1
         num_channels = input_shape[channel_axis - 1]
         if num_channels == 3:
             x = PreStem(name=model_name)(x)
@@ -478,9 +465,7 @@ def ConvNeXt(
                 strides=4,
                 name=model_name + "_stem_conv",
             ),
-            layers.LayerNormalization(
-                epsilon=1e-6, name=model_name + "_stem_layernorm"
-            ),
+            layers.LayerNormalization(epsilon=1e-6, name=model_name + "_stem_layernorm"),
         ],
         name=model_name + "_stem",
     )
@@ -511,9 +496,7 @@ def ConvNeXt(
     # Stochastic depth schedule.
     # This is referred from the original ConvNeXt codebase:
     # https://github.com/facebookresearch/ConvNeXt/blob/main/models/convnext.py#L86
-    depth_drop_rates = [
-        float(x) for x in np.linspace(0.0, drop_path_rate, sum(depths))
-    ]
+    depth_drop_rates = [float(x) for x in np.linspace(0.0, drop_path_rate, sum(depths))]
 
     # First apply downsampling blocks and then apply ConvNeXt stages.
     cur = 0