Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| import joblib | |
| import numpy as np | |
| import pandas as pd | |
| # --- 1. Load Model and Dataset for Feature Information --- | |
| def load_data_and_model(): | |
| """ | |
| Loads the saved model and the dataset from the Excel file. | |
| Using st.cache_data to avoid reloading on every interaction. | |
| """ | |
| try: | |
| # Load the pre-trained Voting Classifier model | |
| model = joblib.load('voting_classifier_model.joblib') | |
| except FileNotFoundError: | |
| st.error("The model file 'voting_classifier_model.joblib' was not found.") | |
| st.info("Please ensure the model file is in the same directory as this script.") | |
| st.stop() | |
| try: | |
| # Load your specific dataset to get feature names and default values | |
| df = pd.read_excel('breast-cancer.xls') | |
| # Assuming the first column is 'id' and the second is 'diagnosis' (the target) | |
| # The rest are the features. | |
| feature_names = df.columns[2:].tolist() | |
| # Store the dataframe for calculating min/max/mean values for sliders | |
| feature_data = df[feature_names] | |
| except FileNotFoundError: | |
| st.error("The dataset file 'breast-cancer.xls' was not found.") | |
| st.info("Please ensure your Excel file is in the same directory as this script.") | |
| st.stop() | |
| except Exception as e: | |
| st.error(f"Could not load or process the dataset file. Error: {e}") | |
| st.stop() | |
| return model, feature_names, feature_data | |
| model, feature_names, feature_data = load_data_and_model() | |
| # --- 2. Streamlit App Interface --- | |
| st.set_page_config(page_title="Breast Cancer Predictor", layout="wide") | |
| # Main Title | |
| st.title("🔬 Breast Cancer Prediction Interface") | |
| st.markdown(""" | |
| This application uses your pre-trained model to predict whether a breast tumor is **Malignant** or **Benign**. | |
| The input fields below are based on the columns from your `breast-cancer.xls` file. | |
| """) | |
| st.write("---") | |
| # --- 3. User Input via Sliders --- | |
| st.sidebar.header("Input Tumor Features") | |
| st.sidebar.markdown("Use the sliders to provide the feature values.") | |
| # Dictionary to hold the user's input | |
| input_features = {} | |
| # Create sliders for all features based on your Excel file | |
| for feature in feature_names: | |
| # Set min/max/default values from the actual data for better usability | |
| min_val = float(feature_data[feature].min()) | |
| max_val = float(feature_data[feature].max()) | |
| mean_val = float(feature_data[feature].mean()) | |
| # Create a slider for each feature | |
| input_features[feature] = st.sidebar.slider( | |
| label=f"{feature.replace('_', ' ').title()}", | |
| min_value=min_val, | |
| max_value=max_val, | |
| value=mean_val, | |
| key=f"slider_{feature}" | |
| ) | |
| st.sidebar.write("---") | |
| # --- 4. Prediction Logic --- | |
| # Convert the dictionary of input features into a NumPy array | |
| # The order of features must match the order in the feature_names list | |
| input_data = np.array([list(input_features.values())]) | |
| # Main section for displaying inputs and results | |
| st.header("Prediction Results") | |
| col1, col2 = st.columns([2, 1]) | |
| with col1: | |
| st.subheader("Current Input Values") | |
| st.json(input_features) | |
| # "Predict" button | |
| if st.button("✨ Predict Diagnosis", key="predict_button"): | |
| try: | |
| # Make prediction. This returns the string label directly (e.g., 'M' or 'B'). | |
| prediction_label = model.predict(input_data)[0] | |
| # Get prediction probabilities. The order corresponds to model.classes_ | |
| prediction_proba = model.predict_proba(input_data)[0] | |
| with col2: | |
| st.subheader("Diagnosis") | |
| # Display the predicted label directly | |
| # We check for 'M' or 'B' as is common in this dataset | |
| if prediction_label.upper() == 'M': | |
| st.error("Predicted Diagnosis: **Malignant**") | |
| else: | |
| st.success("Predicted Diagnosis: **Benign**") | |
| st.subheader("Prediction Confidence") | |
| # Get the class labels from the model itself to ensure correct order | |
| class_labels = list(model.classes_) | |
| # Display probabilities for each class using the model's class order | |
| for i, label in enumerate(class_labels): | |
| display_label = "Malignant" if label.upper() == 'M' else "Benign" | |
| st.write(f"Confidence for **{display_label}**: `{prediction_proba[i]:.2%}`") | |
| except Exception as e: | |
| st.error(f"An error occurred during prediction: {e}") | |
| st.write("---") |