app.py

'''
This module contains the main function for the training pipeline.

author: Saurabh Bhardwaj
date: October 2023
'''

# import necessary libraries

import pandas as pd
from src.pipeline.prediction_pipeline import PredictPipeline
from flask import Flask, request, render_template


app = Flask(__name__)

@app.route("/")
def index():
    return render_template("index.html")

@app.route("/predict", methods=["POST"])
def predict():
    # In a real-world scenario, you might check if the pipeline is running or handle errors here.
    
    # Assuming you receive the data from a form with a file input named 'data_file'
    uploaded_file = request.files['data_file']
    
    if not uploaded_file:
        return "No file uploaded."
    
    # Assuming the file is in CSV format
    data = pd.read_csv(uploaded_file)
    
    # Assuming you have a PredictPipeline class
    predictions_data = PredictPipeline().predict(data)
    predictions = predictions_data['predicted_column'].tolist()

    return render_template("predict.html", predictions=predictions)

if __name__ == "__main__":
    app.run(debug=True, port=8000)