import pandas as pd
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split

slice_size = 30000
data = pd.read_csv('true_car_listings.csv', index_col='Vin')[:slice_size]

unique_numbers = list(set(data['Model']))
data['Model'] = data['Model'].apply(unique_numbers.index)

clf = DecisionTreeClassifier(random_state=341)

# Выбираем параметры
Y = data['Price']
X = data[['Mileage', 'Year', 'Model']]
print(X)

# Разделяем набор на тренировочные и тестовые данные
X_train, X_test, y_train, y_test = train_test_split(
    X, Y, test_size=0.2, random_state=42)

# Запуск на тренировочных данных
clf.fit(X_train, y_train)

# Точность модели
print(f'Score: {clf.score(X_test, y_test)}')

# Значимость параметров
importances = clf.feature_importances_
print(f'Means {importances}')