import pandas as pd
import numpy as np
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error

names = ['price', 'mileage', 'year', 'brand', 'engineType', 'engineCapacity']
x_names = ['mileage', 'year', 'brand', 'engineType', 'engineCapacity']
names_without_brand = ['mileage', 'year', 'engineType', 'engineCapacity']


def main():
    train_x, train_y = get_train_data()

    model = LinearRegression()
    model.fit(train_x, train_y)

    dev_x, dev_y = get_dev_data()

    predicted_dev_y = model.predict(dev_x)
    save_csv(predicted_dev_y, 'dev-0/out.tsv')

    test_x = get_test_data()

    predicted_test_y = model.predict(test_x)
    save_csv(predicted_test_y, 'test-A/out.tsv')

    print(RMSE(dev_y, predicted_dev_y))


def get_train_data():
    raw_data = pd.read_csv('train/train.tsv', sep='\t', names=names)
    x = raw_data[names_without_brand]
    x = pd.get_dummies(x, columns=['engineType'])
    y = raw_data['price']

    return x, y


def get_dev_data():
    dev_raw_data = pd.read_csv('dev-0/in.tsv', sep='\t', names=x_names)
    x = dev_raw_data[names_without_brand]
    x = pd.get_dummies(x, columns=['engineType'])

    with open('dev-0/expected.tsv', 'r') as file:
        y = [float(line.strip('\n')) for line in file.readlines()]

    return x, y


def get_test_data():
    test_raw_data = pd.read_csv('test-A/in.tsv', sep='\t', names=x_names)
    x = test_raw_data[names_without_brand]
    x = pd.get_dummies(x, columns=['engineType'])

    return x


def save_csv(data, path):
    df = pd.DataFrame(data)
    df.to_csv(path, sep='\t', index=False, header=False)


def RMSE(dev_y, predicted_dev_y):
    return np.sqrt(mean_squared_error(dev_y, predicted_dev_y))


if __name__ == '__main__':
    main()