Merge branch 'main' of https://github.com/MLOps-essi-upc/MLOps-braint

Monitoring/P1.py

@@ -0,0 +1,90 @@
+import evidently
+from evidently.dashboard import Dashboard
+from evidently.dashboard.tabs import DataDriftTab, ClassificationPerformanceTab,CatTargetDriftTab
+from evidently.options import DataDriftOptions
+from evidently.pipeline.column_mapping import ColumnMapping
+import os
+from PIL import Image
+import pandas as pd
+import requests
+
+
+def parse_images_from_folder(folder_path):
+    image_data = []
+
+    for root, dirs, files in os.walk(folder_path):
+        for filename in files:
+            if filename.endswith(('.jpg', '.jpeg', '.png')):
+                image_path = os.path.join(root, filename)
+
+                try:
+                    with open(image_path, 'rb') as file:
+                        # Prepare the files dictionary with a key ('file' in this case) and the file object
+                        files = {'file': (image_path, file, 'image/jpeg')}
+
+                        # Make the POST request with the files parameter
+                        response = requests.post(url, files=files)
+                        response= response.json()
+                        response['data']['predicted_class']=response['data']['predicted_class'].lower()
+                        image_data.append(response)
+
+                except Exception as e:
+                    print(f"Error reading {filename}: {e}")
+
+    # Convert the list of dictionaries to a Pandas DataFrame
+    df = image_data
+
+    return df
+
+def read_images_from_folder(folder_path):
+    image_data = []
+
+    for root, dirs, files in os.walk(folder_path):
+        for filename in files:
+            if filename.endswith(('.jpg', '.jpeg', '.png')):
+                image_path = os.path.join(root, filename)
+
+                try:
+                    with Image.open(image_path) as img:
+                        image_data.append({'class':root[root.rfind("\\")+1:], 'img':img})
+
+                except Exception as e:
+                    print(f"Error reading {filename}: {e}")
+
+    # Convert the list of dictionaries to a Pandas DataFrame
+    df = image_data
+
+    return df
+
+if __name__ == "__main__":
+    # Specify the path to the folder containing images
+    folder_path_new = "C:/Users/anpag/PycharmProjects/pythonProject3/new"  # Change this to the actual path
+    folder_path_old = "C:/Users/anpag/PycharmProjects/pythonProject3/Training"
+    url = "http://10.4.41.39:8000/predict"
+
+
+    # Read images from the folder into a DataFrame
+    result_old = parse_images_from_folder(folder_path_old)
+    result_new = parse_images_from_folder(folder_path_new)
+    im_old = read_images_from_folder(folder_path_old)
+    im_new = read_images_from_folder(folder_path_new)
+
+    column_mapping = ColumnMapping()
+    column_mapping.target = 'class'
+    column_mapping.prediction = 'prediction'
+
+    ref_data= pd.DataFrame()
+    ref_data['class']=[x['class'] for x in im_old]
+    ref_data['prediction'] = [x['data']['predicted_class'] for x in result_old]
+
+    prod_data = pd.DataFrame()
+    prod_data['class'] = [x['class'] for x in im_new]
+    prod_data['prediction'] = [x['data']['predicted_class'] for x in result_new]
+
+    classification_performance_report = Dashboard(tabs=[ClassificationPerformanceTab()])
+    classification_performance_report.calculate(ref_data,prod_data,column_mapping=column_mapping)
+    classification_performance_report.save("model_performance.html")
+
+    target_drift_dashboard = Dashboard(tabs=[CatTargetDriftTab(verbose_level=1)])
+    target_drift_dashboard.calculate(ref_data, prod_data,column_mapping=column_mapping)
+    target_drift_dashboard.save("target_drift.html")