app.py

import streamlit as st
import joblib
import tensorflow as tf
import numpy as np
import pandas as pd
from sklearn.preprocessing import StandardScaler

# Load the models
svm_model = joblib.load('svm_model.pkl')
logreg_model = joblib.load('logistic_regression_model.pkl')
neural_network_model = tf.keras.models.load_model('neural_network_model.h5')

# Load the scaler used during training
scaler = joblib.load('scaler.pkl')

# Function to make predictions using the chosen model
def make_prediction(model, data):
    if model == 'SVM':
        return svm_model.predict(data)

    elif model == 'Logistic Regression':
        return logreg_model.predict(data)
    elif model == 'Neural Network':
        return (neural_network_model.predict(data) > 0.5).astype(int)

# Streamlit app
st.set_page_config(page_title='Kidney Disease Classification', layout='wide')
st.markdown(
    """
    <style>
    .main {
        background-color: #f0f8ff;
    }
    .header {
        font-size: 2em;
        color: #4f4f4f;
    }
    .red {
        color: red;
        font-weight: bold;
    }
    .green {
        color: green;
        font-weight: bold;
    }
    </style>
    """, unsafe_allow_html=True
)

st.title('Kidney Disease Classification')

# Sidebar for model selection
st.sidebar.header('Model Selection')
model_choice = st.sidebar.radio('Choose the model for prediction:', 
                                  ['SVM', 'Logistic Regression', 'Neural Network'])

# Input fields for features arranged in rows of 5
st.subheader('Input Features')

# First row
cols = st.columns(5)

with cols[0]:
    age = st.number_input('Age', min_value=0, max_value=120, value=30, step=1)
with cols[1]:
    bp = st.number_input('Blood Pressure (BP)', min_value=50, max_value=200, value=80, step=1)
with cols[2]:
    sg = st.number_input('Specific Gravity (SG)', min_value=1.000, max_value=1.050, value=1.020, step=0.001)
with cols[3]:
    al = st.number_input('Albumin (AL)', min_value=0, max_value=5, value=1, step=1)
with cols[4]:
    su = st.number_input('Sugar (SU)', min_value=0, max_value=5, value=0, step=1)

# Second row
cols = st.columns(5)

with cols[0]:
    pcc = st.selectbox('Pus Cell Count (PCC)', ['present', 'notpresent'])
with cols[1]:
    ba = st.selectbox('Bacteria (BA)', ['present', 'notpresent'])
with cols[2]:
    bgr = st.number_input('Blood Glucose Random (BGR)', min_value=0, value=120, step=1)
with cols[3]:
    bu = st.number_input('Blood Urea (BU)', min_value=0, value=10, step=1)
with cols[4]:
    sc = st.number_input('Serum Creatinine (SC)', min_value=0.0, value=0.8, step=0.1)

# Third row
cols = st.columns(5)

with cols[0]:
    sod = st.number_input('Sodium (SOD)', min_value=0.0, value=135.0, step=0.1)
with cols[1]:
    pot = st.number_input('Potassium (POT)', min_value=0.0, value=4.0, step=0.1)
with cols[2]:
    hemo = st.number_input('Hemoglobin (HEMO)', min_value=0.0, value=15.0, step=0.1)
with cols[3]:
    pcv = st.number_input('Packed Cell Volume (PCV)', min_value=0.0, value=45.0, step=0.1)
with cols[4]:
    wc = st.number_input('White Cell Count (WC)', min_value=0, value=7000, step=1)

# Fourth row
cols = st.columns(5)

with cols[0]:
    rc = st.number_input('Red Cell Count (RC)', min_value=0.0, value=5.0, step=0.1)
with cols[1]:
    htn = st.selectbox('Hypertension (HTN)', ['yes', 'no'])
with cols[2]:
    dm = st.selectbox('Diabetes Mellitus (DM)', ['yes', 'no'])
with cols[3]:
    cad = st.selectbox('Coronary Artery Disease (CAD)', ['yes', 'no'])
with cols[4]:
    appet = st.selectbox('Appetite', ['good', 'poor'])

# Fifth row
cols = st.columns(2)

with cols[0]:
    pe = st.selectbox('Pedal Edema (PE)', ['yes', 'no'])
with cols[1]:
    ane = st.selectbox('Anemia (ANE)', ['yes', 'no'])

# Convert categorical inputs into numerical format
categorical_mapping = {
    'present': 1, 'notpresent': 0,
    'yes': 1, 'no': 0,
    'good': 0, 'poor': 1
}

# Create input data dictionary
input_data = {
    'age': age,
    'bp': bp,
    'sg': sg,
    'al': al,
    'su': su,
    'pcc': categorical_mapping[pcc],
    'ba': categorical_mapping[ba],
    'bgr': bgr,
    'bu': bu,
    'sc': sc,
    'sod': sod,
    'pot': pot,
    'hemo': hemo,
    'pcv': pcv,
    'wc': wc,
    'rc': rc,
    'htn': categorical_mapping[htn],
    'dm': categorical_mapping[dm],
    'cad': categorical_mapping[cad],
    'appet': categorical_mapping[appet],
    'pe': categorical_mapping[pe],
    'ane': categorical_mapping[ane]
}

# Convert input data to DataFrame for prediction
input_df = pd.DataFrame([input_data])

# Apply the scaler to standardize the input data
input_df_scaled = scaler.transform(input_df)

# Make prediction when the button is clicked
if st.button('Predict'):
    prediction = make_prediction(model_choice, input_df_scaled)
  
    st.subheader('Prediction')
    if prediction[0] == 1:
        st.markdown('<p class="green">No Kidney Disease Found</p>', unsafe_allow_html=True)
    else:
        st.markdown('<p class="red">Kidney Disease Found</p>', unsafe_allow_html=True)