app.py

import streamlit as st
import joblib
import numpy as np
import pandas as pd

# Load the models
ph_model = joblib.load('random_forest_ph_model.pkl')
tds_model = joblib.load('random_forest_tds_model.pkl')

# Set up the main title and subtitle
st.title("🔍 TDS and pH Level Predictor")
st.markdown("Get predictions for hourly **pH** and **TDS** levels based on selected date. Perfect for water quality monitoring!")

# Sidebar for input parameters with an expander
with st.sidebar.expander("🔧 Input Parameters"):
    st.write("Enter the date details for which you want to predict the TDS and pH levels.")
    day = st.number_input("Day", min_value=1, max_value=31, value=1)
    month = st.number_input("Month", min_value=1, max_value=12, value=1)
    year = st.number_input("Year", min_value=2000, max_value=2100, value=2024)

# Create a "Predict" button
if st.button("🚀 Predict"):
    # Generate hourly features
    hours = list(range(6, 22))  # From 6 AM to 10 PM
    day_input = [day] * len(hours)
    month_input = [month] * len(hours)
    year_input = [year] * len(hours)

    # Create feature array for prediction
    features = np.array([hours, day_input, month_input, year_input]).T

    # Predict pH and TDS values
    ph_predictions = ph_model.predict(features)
    tds_predictions = tds_model.predict(features)

    # Compile results into a DataFrame for better visualization
    results_df = pd.DataFrame({
        'Hour': hours,
        'Predicted pH': ph_predictions,
        'Predicted TDS': tds_predictions
    })

    # Display predictions with a styled table
    st.markdown("### 📊 Prediction Results")
    st.dataframe(results_df.style.set_properties(**{
        'background-color': 'lavender',
        'font-size': '14px'
    }).highlight_max(axis=0, color='lightblue'))