Create app.py

app.py

@@ -0,0 +1,129 @@
+import gradio as gr
+import pandas as pd
+import joblib
+import numpy as np
+
+# 1. Load the Model and Artifacts
+# Model and artifact paths - adjust as necessary
+model_path = 'your_model.pkl'
+scaler_path = 'scaler.pkl'  # If you used a scaler
+# encoder_path = 'encoder.pkl' # If you used an encoder
+
+try:
+    model = joblib.load(model_path)
+    scaler = joblib.load(scaler_path) if scaler_path else None # Load scaler
+    # encoder = joblib.load(encoder_path) if encoder_path else None  # Load encoder
+except Exception as e:
+    print(f"Error loading model/artifacts: {e}")
+    model = None
+
+# 2. Preprocessing Function
+def preprocess_data(data):
+    """
+    Preprocesses input data for the model.
+
+    Args:
+        data (dict): A dictionary containing the input data from the Gradio interface.
+
+    Returns:
+        pandas.DataFrame: A DataFrame ready for model prediction.  Returns empty if error.
+    """
+    try:
+        df = pd.DataFrame([data])
+
+        #  Expected fields.  Adapt to match your form inputs.
+        expected_fields = [
+            'Quality_Report__c',
+            'Delay_Days__c',
+            'Incident_Log__c',
+            'Vendor__c'
+        ]
+        # 1. Check for missing fields
+        missing_fields = [field for field in expected_fields if field not in df.columns]
+        if missing_fields:
+            raise ValueError(f"Missing required fields: {', '.join(missing_fields)}")
+        # 2. Data Transformations (Adapt to your needs)
+        # Example: Scaling
+        if 'Delay_Days__c' in df.columns and scaler:
+            df['Delay_Days__c'] = df['Delay_Days__c'].fillna(0)
+            df[['Delay_Days__c']] = scaler.transform(df[['Delay_Days__c']])
+
+        # 3.  One-Hot Encoding (Example)
+        # if 'Quality_Report__c' in df.columns and encoder:
+        #
+        #     possible_categories = ['Good', 'Bad', 'Excellent', 'Average']  # Replace
+        #     if df['Quality_Report__c'][0] not in possible_categories:
+        #         df['Quality_Report__c'] = 'Average'
+        #     encoded_quality = encoder.transform(df[['Quality_Report__c']])
+        #     df = pd.concat([df.drop('Quality_Report__c', axis=1), pd.DataFrame(encoded_quality)], axis=1)
+        df = df[expected_fields]
+        return df
+    except ValueError as ve:
+        print(f"Error in preprocess_data: {ve}")
+        return pd.DataFrame()  # Return empty DataFrame on error
+    except Exception as e:
+        print(f"Error in preprocess_data: {e}")
+        return pd.DataFrame()
+
+# 3. Prediction Function
+def predict_vendor_score(*args):
+    """
+    Predicts the vendor performance score based on the input data.
+
+    Args:
+        *args:  Input values from the Gradio interface.
+
+    Returns:
+        dict: A dictionary containing the prediction and potentially an error message.
+    """
+    if model is None:
+        return {'Score': 'Model not loaded.  Check server logs.', 'error': True}
+
+    # 1. Prepare input data as a dictionary.  Order MUST match expected_fields in preprocess.
+    input_data = {
+        'Quality_Report__c': args[0],
+        'Delay_Days__c': args[1],
+        'Incident_Log__c': args[2],
+        'Vendor__c': args[3],
+    }
+
+    # 2. Preprocess the data
+    processed_df = preprocess_data(input_data)
+
+    if processed_df.empty:
+        return {'Score': 'Error in input data. Check logs.', 'error': True}
+
+    # 3. Make Prediction
+    try:
+        prediction = model.predict(processed_df)[0]  # Get the first element
+        # prediction = prediction.tolist() # convert numpy to list
+        return {'Score': prediction, 'error': False}
+    except Exception as e:
+        print(f"Error during prediction: {e}")
+        return {'Score': f'Error during prediction: {e}', 'error': True}
+
+# 4. Gradio Interface
+iface = gr.Interface(
+    fn=predict_vendor_score,
+    inputs=[
+        gr.Dropdown(['Good', 'Bad', 'Excellent', 'Average'], label="Quality Report"),  # Example
+        gr.Number(label="Delay Days"),
+        gr.Number(label="Incident Log"),
+        gr.Textbox(label="Vendor ID"),  # Or a dropdown if you have vendor names
+    ],
+    outputs=gr.outputs.HighlightedText(
+        label="Prediction",
+        # displayed_value_interpretation=lambda x: {"Score": x}  #  Simplest approach
+        ),
+    # examples=[  #  Example data (optional)
+    #     ["Good", 5, 2, "V123"],
+    #     ["Excellent", 0, 0, "V456"],
+    #     ["Bad", 10, 5, "V789"],
+    # ],
+    title="Subcontractor Performance Score Predictor",
+    description="Enter subcontractor data to get a performance score.",
+)
+
+# 5. Launch the Interface
+if __name__ == "__main__":
+    iface.launch()