47 lines
2.0 KiB
Python
47 lines
2.0 KiB
Python
|
from DataCollectingScraper.DataScrapers.OtoDomDataScraperImpl import OtoDomDataScraperImpl
|
||
|
|
from DataCollectingScraper.DataCollectingScraper import DataCollectingScraper
|
||
|
|
|
||
|
|
from DataPreprocessor.helpers.OffersCSVReader import OffersCSVReader
|
||
|
|
from DataPreprocessor.DataPreprocessor import DataPreprocessor
|
||
|
|
|
||
|
|
from Prediction.Trainer.PredictionModelTrainer import PredictionModelTrainer
|
||
|
|
|
||
|
|
from sklearn.neural_network import MLPRegressor
|
||
|
|
from pandas.core.frame import DataFrame
|
||
|
|
import pandas as pd
|
||
|
|
import joblib
|
||
|
|
|
||
|
|
download_data = False
|
||
|
|
train_model = False
|
||
|
|
|
||
|
|
# Downloading raw data
|
||
|
|
if download_data:
|
||
|
|
offers_sublink = "wyniki/sprzedaz/mieszkanie/dolnoslaskie/wroclaw/wroclaw/wroclaw?viewType=listing"
|
||
|
|
scraper = DataCollectingScraper(OtoDomDataScraperImpl(offers_sublink))
|
||
|
|
scraper()
|
||
|
|
|
||
|
|
# 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()
|
||
|
|
|
||
|
|
if train_model:
|
||
|
|
preprocessed_data : DataFrame = data_preprocessor.get_preprocessed_data()
|
||
|
|
# Train neural network with preprocessed data
|
||
|
|
trainer = PredictionModelTrainer(preprocessed_data)
|
||
|
|
trainer.train()
|
||
|
|
trained_model : MLPRegressor = trainer.get_trained_model()
|
||
|
|
joblib.dump(trained_model, 'trained_model.pkl')
|
||
|
|
|
||
|
|
trained_model = joblib.load('trained_model.pkl')
|
||
|
|
scaled_area = data_preprocessor.get_value('Area', pd.DataFrame({'Area': [56.0]}))
|
||
|
|
scaled_construction_year = data_preprocessor.get_value('Construction year', pd.DataFrame({'Construction year': [1980]}))
|
||
|
|
encoded_location = data_preprocessor.get_value("Location", ['Krzyki'])
|
||
|
|
sample_data = [[scaled_area, 3, 8, 0, 2, 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)
|
||
|
|
print('Predicted price: ', round(float(prediction),0), 'zł')
|