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")
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.

**Important:**  
The model was trained using data from patients **already diagnosed** with lung cancer.  
The output represents a **risk/severity level**, not a medical diagnosis.
""")

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

Gender = st.selectbox(
    "Gender",
    options=[("Male", 1), ("Female", 2)],
    format_func=lambda x: x[0]
)[1]

Air_Pollution = st.slider(
    "Air Pollution Exposure",
    1, 8, 4,
    help="""
    1 – Rural area
    2 – Small town
    3 – Urban residential
    4 – Heavy traffic area
    5 – Near industrial zone
    6 – Factory worker
    7 – High industrial exposure
    8 – Extremely polluted environment
    """
)

Alcohol_use = st.slider(
    "Alcohol Use",
    1, 8, 5,
    help="""
    1 – Never drinks
    2 – Rare
    3 – Occasional
    4 – Social drinking
    5 – Weekly
    6 – Several times a week
    7 – Daily
    8 – Heavy drinker
    """
)

Dust_Allergy = st.slider(
    "Dust Allergy",
    1, 8, 5,
    help="1=None | 3=Mild | 5=Moderate | 7=Severe | 8=Chronic"
)

Occupational_Hazards = st.slider(
    "Occupational Hazards",
    1, 8, 5,
    help="""
    1 – Office / home work
    3 – Outdoor non-industrial
    5 – Factory work
    6 – Chemical exposure
    8 – Extreme industrial hazard
    """
)

Genetic_Risk = st.slider(
    "Genetic Risk",
    1, 7, 5,
    help="1=None | 3=Distant family history | 5=Immediate family | 7=Strong genetic risk"
)

Chronic_Lung_Disease = st.slider(
    "Chronic Lung Disease",
    1, 7, 4,
    help="1=None | 3=Past condition | 5=Ongoing | 7=Severe"
)

Balanced_Diet = st.slider(
    "Balanced Diet",
    1, 7, 4,
    help="1=Very poor | 3=Poor | 5=Mostly balanced | 7=Highly balanced"
)

Obesity = st.slider(
    "Obesity",
    1, 7, 4,
    help="1=Normal | 3=Overweight | 5=Obese | 7=Severely obese"
)

Smoking = st.slider(
    "Smoking",
    1, 8, 4,
    help="""
    1 – Never smoked
    3 – Former smoker
    4 – Occasional smoker
    5 – Light daily smoker
    6 – Moderate smoker
    7 – Heavy smoker
    8 – Chain smoker
    """
)

Passive_Smoker = st.slider(
    "Passive Smoker",
    1, 8, 4,
    help="1=None | 3=Occasional | 5=Daily exposure | 8=Constant exposure"
)

Chest_Pain = st.slider(
    "Chest Pain",
    1, 9, 4,
    help="1=None | 3=Mild | 5=Moderate | 7=Severe | 9=Extreme"
)

Coughing_of_Blood = st.slider(
    "Coughing of Blood",
    1, 9, 4,
    help="1=None | 5=Occasional | 9=Frequent / severe"
)

Fatigue = st.slider(
    "Fatigue",
    1, 9, 4,
    help="1=None | 3=Mild | 5=Moderate | 7=Severe | 9=Extreme"
)

Weight_Loss = st.slider(
    "Weight Loss",
    1, 8, 4,
    help="1=None | 4=Noticeable | 6=Significant | 8=Severe"
)

Shortness_of_Breath = st.slider(
    "Shortness of Breath",
    1, 9, 4,
    help="1=None | 3=Mild | 5=Moderate | 7=Severe | 9=Extreme"
)

Wheezing = st.slider(
    "Wheezing",
    1, 8, 4,
    help="1=None | 4=Occasional | 6=Frequent | 8=Constant"
)

Swallowing_Difficulty = st.slider(
    "Swallowing Difficulty",
    1, 8, 4,
    help="1=None | 4=Occasional | 6=Frequent | 8=Severe"
)

Clubbing_of_Finger_Nails = st.slider(
    "Clubbing of Finger Nails",
    1, 9, 4,
    help="1=None | 5=Visible | 9=Severe"
)

Frequent_Cold = st.slider(
    "Frequent Cold",
    1, 7, 4,
    help="1=Rare | 3=Occasional | 5=Frequent | 7=Very frequent"
)

Dry_Cough = st.slider(
    "Dry Cough",
    1, 7, 4,
    help="1=None | 3=Mild | 5=Persistent | 7=Severe"
)

Snoring = st.slider(
    "Snoring",
    1, 7, 3,
    help="1=None | 3=Occasional | 5=Frequent | 7=Severe"
)

# --- 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("**Trained By:** Edcel Bogay")