app_3.py

import streamlit as st
import pandas as pd
import joblib
import matplotlib.pyplot as plt
from huggingface_hub import hf_hub_download
import time
import base64
from transformers import AutoTokenizer

# Load the pre-trained numerical imputer, scaler, and model using joblib

REPO_ID = "AlbieCofie/predict-customer-churn"

num_imputer = joblib.load(
    hf_hub_download(repo_id=REPO_ID, filename="numerical_imputer.joblib")
)

scaler = joblib.load(
    hf_hub_download(repo_id=REPO_ID, filename="scaler.joblib")
)

model = joblib.load(
    hf_hub_download(repo_id=REPO_ID, filename="Final_model.joblib")
)

# Define a function to preprocess the input data
def preprocess_input_data(input_data):
    input_data_df = pd.DataFrame(input_data, columns=['PRG', 'PL', 'PR', 'SK', 'TS', 'M11', 'BD2', 'Age', 'Insurance'])
    num_columns = input_data_df.select_dtypes(include='number').columns

    input_data_imputed_num = num_imputer.transform(input_data_df[num_columns])
    input_scaled_df = pd.DataFrame(scaler.transform(input_data_imputed_num), columns=num_columns)

    return input_scaled_df


# Define a function to make the sepsis prediction
def predict_sepsis(input_data):
    input_scaled_df = preprocess_input_data(input_data)
    prediction = model.predict(input_scaled_df)[0]
    probabilities = model.predict_proba(input_scaled_df)[0]
    sepsis_status = "Positive" if prediction == 1 else "Negative"
    
    status_icon = "✔" if prediction == 1 else "✘"  # Red 'X' icon for positive sepsis prediction, green checkmark icon for negative sepsis prediction
    sepsis_explanation = "Sepsis is a life-threatening condition caused by an infection. A positive prediction suggests that the patient might be exhibiting sepsis symptoms and requires immediate medical attention." if prediction == 1 else "Sepsis is a life-threatening condition caused by an infection. A negative prediction suggests that the patient is not currently exhibiting sepsis symptoms."

    output_df = pd.DataFrame(input_data, columns=['PRG', 'PL', 'PR', 'SK', 'TS', 'M11', 'BD2', 'Age', 'Insurance'])
    output_df['Prediction'] = sepsis_status
    output_df['Negative Probability'] = probabilities[0]
    output_df['Positive Probability'] = probabilities[1]

    return output_df, probabilities, status_icon, sepsis_explanation

# Create a Streamlit app
def main():
    st.title('Sepsis Prediction App')

    st.image("Strealit_.jpg")

    # How to use
    st.sidebar.title('How to Use')
    st.sidebar.markdown('1. Adjust the input parameters on the left sidebar.')
    st.sidebar.markdown('2. Click the "Predict" button to initiate the prediction.')
    st.sidebar.markdown('3. The app will simulate a prediction process with a progress bar.')
    st.sidebar.markdown('4. Once the prediction is complete, the results will be displayed below.')


    st.sidebar.title('Input Parameters')

    # Input parameter explanations
    st.sidebar.markdown('**PRG:** Plasma Glucose')
    PRG = st.sidebar.number_input('PRG', value=0.0)

    st.sidebar.markdown('**PL:** Blood Work Result 1')
    PL = st.sidebar.number_input('PL', value=0.0)

    st.sidebar.markdown('**PR:** Blood Pressure Measured')
    PR = st.sidebar.number_input('PR', value=0.0)

    st.sidebar.markdown('**SK:** Blood Work Result 2')
    SK = st.sidebar.number_input('SK', value=0.0)

    st.sidebar.markdown('**TS:** Blood Work Result 3')
    TS = st.sidebar.number_input('TS', value=0.0)

    st.sidebar.markdown('**M11:** BMI')
    M11 = st.sidebar.number_input('M11', value=0.0)

    st.sidebar.markdown('**BD2:** Blood Work Result 4')
    BD2 = st.sidebar.number_input('BD2', value=0.0)

    st.sidebar.markdown('**Age:** What is the Age of the Patient: ')
    Age = st.sidebar.number_input('Age', value=0.0)

    st.sidebar.markdown('**Insurance:** Does the patient have Insurance?')
    insurance_options = {0: 'NO', 1: 'YES'}
    Insurance = st.sidebar.radio('Insurance', list(insurance_options.keys()), format_func=lambda x: insurance_options[x])


    input_data = [[PRG, PL, PR, SK, TS, M11, BD2, Age, Insurance]]

    if st.sidebar.button('Predict'):
        with st.spinner("Predicting..."):
            # Simulate a long-running process
            progress_bar = st.progress(0)
            step = 20 # A big step will reduce the execution time
            for i in range(0, 100, step):
                time.sleep(0.1)
                progress_bar.progress(i + step)

            output_df, probabilities, status_icon, sepsis_explanation = predict_sepsis(input_data)

            st.subheader('Prediction Result')
            prediction_text = "Positive" if status_icon == "✔" else "Negative"
            st.markdown(f"Prediction: **{prediction_text}**")
            st.markdown(f"{status_icon} {sepsis_explanation}")
            st.write(output_df)

            # Add a download button for output_df
            csv = output_df.to_csv(index=False)
            b64 = base64.b64encode(csv.encode()).decode()
            href = f'<a href="data:file/csv;base64,{b64}" download="output.csv">Download Output CSV</a>'
            st.markdown(href, unsafe_allow_html=True)
            

            # Plot the probabilities
            fig, ax = plt.subplots()
            ax.bar(['Negative', 'Positive'], probabilities)
            ax.set_xlabel('Sepsis Status')
            ax.set_ylabel('Probability')
            ax.set_title('Sepsis Prediction Probabilities')
            st.pyplot(fig)

            # Print feature importance
            if hasattr(model, 'coef_'):
                feature_importances = model.coef_[0]
                feature_names = ['PRG', 'PL', 'PR', 'SK', 'TS', 'M11', 'BD2', 'Age', 'Insurance']

                importance_df = pd.DataFrame({'Feature': feature_names, 'Importance': feature_importances})
                importance_df = importance_df.sort_values('Importance', ascending=False)

                st.subheader('Feature Importance')
                fig, ax = plt.subplots()
                bars = ax.bar(importance_df['Feature'], importance_df['Importance'])
                ax.set_xlabel('Feature')
                ax.set_ylabel('Importance')
                ax.set_title('Feature Importance')
                ax.tick_params(axis='x', rotation=45)

                # Add data labels to the bars
                for bar in bars:
                    height = bar.get_height()
                    ax.annotate(f'{height:.2f}', xy=(bar.get_x() + bar.get_width() / 2, height),
                    xytext=(0, 3),  # 3 points vertical offset
                    textcoords="offset points",
                    ha='center', va='bottom')
                st.pyplot(fig)

            else:
                st.write('Feature importance is not available for this model.')

            #st.subheader('Sepsis Explanation')
            #st.markdown(f"{status_icon} {sepsis_explanation}")


if __name__ == '__main__':
    main()