app.py

import streamlit as st
import numpy as np
import tensorflow as tf
import joblib

# --- Load model and label encoder ---
model = tf.keras.models.load_model("lung_cancer_classifier.h5")  # Load H5 model
le_target = joblib.load("label_encoder.pkl")

# --- Page configuration ---
st.set_page_config(page_title="Lung Cancer Risk Classifier", layout="centered")

# --- Title and description ---
st.title("Lung Cancer Risk Prediction")
st.write("""
This application predicts the **level of lung cancer risk** for a patient.  
**Note:** The data used to train this model consists of patients already diagnosed with lung cancer.  
The target of this classification is the **Lung Cancer Level**.
""")

# --- Input fields ---
Age = st.number_input("Age", min_value=1, max_value=100, value=37)

# Gender dropdown fixed to show Male/Female but return encoded value
Gender = st.selectbox(
    "Gender",
    options=[("Male", 1), ("Female", 2)],
    format_func=lambda x: x[0]  # show only label
)[1]  # get encoded value

Air_Pollution = st.slider("Air Pollution Exposure", 1, 8, 4)
Alcohol_use = st.slider("Alcohol Use", 1, 8, 5)
Dust_Allergy = st.slider("Dust Allergy", 1, 8, 5)
Occupational_Hazards = st.slider("Occupational Hazards", 1, 8, 5)
Genetic_Risk = st.slider("Genetic Risk", 1, 7, 5)
Chronic_Lung_Disease = st.slider("Chronic Lung Disease", 1, 7, 4)
Balanced_Diet = st.slider("Balanced Diet", 1, 7, 4)
Obesity = st.slider("Obesity", 1, 7, 4)
Smoking = st.slider("Smoking", 1, 8, 4)
Passive_Smoker = st.slider("Passive Smoker", 1, 8, 4)
Chest_Pain = st.slider("Chest Pain", 1, 9, 4)
Coughing_of_Blood = st.slider("Coughing of Blood", 1, 9, 4)
Fatigue = st.slider("Fatigue", 1, 9, 4)
Weight_Loss = st.slider("Weight Loss", 1, 8, 4)
Shortness_of_Breath = st.slider("Shortness of Breath", 1, 9, 4)
Wheezing = st.slider("Wheezing", 1, 8, 4)
Swallowing_Difficulty = st.slider("Swallowing Difficulty", 1, 8, 4)
Clubbing_of_Finger_Nails = st.slider("Clubbing of Finger Nails", 1, 9, 4)
Frequent_Cold = st.slider("Frequent Cold", 1, 7, 4)
Dry_Cough = st.slider("Dry Cough", 1, 7, 4)
Snoring = st.slider("Snoring", 1, 7, 3)

# --- Combine inputs ---
input_data = np.array([[
    Age, Gender, Air_Pollution, Alcohol_use, Dust_Allergy,
    Occupational_Hazards, Genetic_Risk, Chronic_Lung_Disease,
    Balanced_Diet, Obesity, Smoking, Passive_Smoker,
    Chest_Pain, Coughing_of_Blood, Fatigue, Weight_Loss,
    Shortness_of_Breath, Wheezing, Swallowing_Difficulty,
    Clubbing_of_Finger_Nails, Frequent_Cold, Dry_Cough, Snoring
]])

# --- Prediction ---
if st.button("Predict"):
    prediction = model.predict(input_data)
    predicted_class = prediction.argmax(axis=1)[0]
    confidence = prediction.max(axis=1)[0]

    label = le_target.inverse_transform([predicted_class])[0]

    st.success(f"Predicted Lung Cancer Level: {label}")
    st.info(f"Prediction Confidence: {confidence:.2%}")

# --- Footer ---
st.markdown("---")
st.markdown("**Train By:** Edcel Bogay")