from flask import Flask, request, jsonify
from flask_cors import CORS
import joblib
import pandas as pd

from DataPreprocessor.helpers.OffersCSVReader import OffersCSVReader
from DataPreprocessor.DataPreprocessor import DataPreprocessor
from pandas.core.frame import DataFrame

app = Flask(__name__)

CORS(app)  # This will enable CORS for all routes

# Reading downloaded data
data_frame : DataFrame = OffersCSVReader.read_from_file("output.csv")

# Prepare data for neural network (data preprocessing)
data_preprocessor = DataPreprocessor(data_frame)
data_preprocessor.preprocess_data()
trained_model = joblib.load('trained_model.pkl')

@app.route('/calculate_price', methods=['POST'])
def calculate_price():
    input_data = request.json
    
    scaled_area = data_preprocessor.get_value('Area', pd.DataFrame({'Area': [input_data["powierzchnia"]]}))
    scaled_construction_year = data_preprocessor.get_value('Construction year', pd.DataFrame({'Construction year': [input_data["rok_budowy"]]}))
    encoded_location = data_preprocessor.get_value("Location", [input_data["dzielnica"]])
    encoded_state = data_preprocessor.get_value("State", [input_data["stan_nieruchomosci"]])
    encoded_property_form = data_preprocessor.get_value("Property form", [input_data["forma_wlasnosci"]])
    floor = input_data['numer_pietra']
    rooms = input_data['ilosc_pokoi']

    sample_data = [[scaled_area, rooms, floor, encoded_property_form, encoded_state, encoded_location, scaled_construction_year]]
    sample = pd.DataFrame(sample_data, columns=['Area', 'Rooms', 'Floor', 'Property form' , 'State', 'Location', 'Construction year'])

    prediction = trained_model.predict(sample)

    calculated_price = {
        'estimated_price': round(float(prediction),0)
    }
    return jsonify(calculated_price)


if __name__ == '__main__':
    app.run(debug=True, port=8081)