app.py

import numpy as np
import pandas as pd
import datetime as dt
from flask import Flask, request, jsonify, render_template
import joblib

app = Flask(__name__)
model = joblib.load('model.pkl')
fet = pd.read_csv('all_features.csv')

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

@app.route('/predict',methods=['POST'])
def predict():

    features = [x for x in request.form.values()]

    if features[3]=='0':
        features[3]=False
    else:
        features[3]=True

    df=fet[(fet['Store']==int(features[0])) & (fet['IsHoliday']==features[3]) & (fet['Date']==features[2])]
    f_features=[]
    d=dt.datetime.strptime(features[2], '%Y-%m-%d')
    c=0
    if df['Type'][0]=='C':
        c=1
    else:
        c=0

    if features[3]==False:
        features[3]=0
    else:
        features[3]=1

    if df.shape[0]==1:
        f_features.append(df['CPI'])
        f_features.append(d.date().day)
        f_features.append(int(features[1]))
        f_features.append(df['Fuel_Price'])
        f_features.append(features[3])
        f_features.append(d.date().month)
        f_features.append(df['Size'])
        f_features.append(int(features[0]))
        f_features.append(df['Temperature'])
        f_features.append(c)
        f_features.append(df['Unemployment'])
        f_features.append(d.date().year)

    final_features = [np.array(f_features)]
    output = model.predict(final_features)[0]
    return render_template('index.html', output=output)


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