import unittest
import importlib
import seaborn as sns
import pandas as pd
import numpy as np
from statsmodels.tsa.arima.model import ARIMA
from sklearn.metrics import mean_absolute_error, mean_squared_error

def _dynamic_test(test_case, condition, success_msg, failure_msg):
    if condition:
        test_case._testMethodName = success_msg
        test_case.assertTrue(True, success_msg)
    else:
        test_case._testMethodName = failure_msg
        test_case.fail(failure_msg)

class TestARIMAForecast(unittest.TestCase):
    @classmethod
    def setUpClass(cls):
        cls.user = importlib.import_module("user_code")

        flights = sns.load_dataset("flights").copy()
        flights["month"] = flights["month"].astype(str)
        flights["date"] = pd.to_datetime(flights["year"].astype(str) + "-" + flights["month"])
        flights.set_index("date", inplace=True)
        series = flights["passengers"]

        train = series.iloc[:-12]
        test = series.iloc[-12:]

        model = ARIMA(train, order=(2, 1, 2))
        fitted_model = model.fit()
        forecast = fitted_model.forecast(steps=12)

        cls.true_mae = mean_absolute_error(test, forecast)
        cls.true_rmse = np.sqrt(mean_squared_error(test, forecast))

    def test_mae_close(self):
        cond = abs(self.user.mae - self.true_mae) < 1e-2
        _dynamic_test(self, cond, "MAE computed correctly.", "MAE value is incorrect.")

    def test_rmse_close(self):
        cond = abs(self.user.rmse - self.true_rmse) < 1e-2
        _dynamic_test(self, cond, "RMSE computed correctly.", "RMSE value is incorrect.")

    def test_forecast_length(self):
        cond = len(self.user.forecast) == 12
        _dynamic_test(self, cond, "Forecast length correct (12 steps).", "Forecast length is incorrect.")

if __name__ == "__main__":
    unittest.main(argv=["first-arg-is-ignored"], exit=False)

test_main.py

Master the theory and practical implementation of ARIMA models for time series forecasting. This course guides you from foundational time series concepts through advanced ARIMA techniques, including hands-on coding and real-world forecasting challenges.

Explore the essential building blocks of time series data, including trend, seasonality, and stationarity, and learn how to visualize and interpret these patterns.

Delve into the mathematical principles behind ARIMA models, including autoregressive and moving average processes, and learn how to interpret ARIMA parameters.

Learn how to build, fit, and evaluate ARIMA models in Python, and understand key metrics for assessing forecast accuracy.

Explore advanced ARIMA variants, automated parameter selection, and strategies for comparing and selecting the best forecasting model.

Challenge: ARIMA Forecasting and Evaluation

Solution