portes: An R Package for Portmanteau Tests in Time Series Models

Abstract: In this article, we introduce the R package portes with extensive illustrative applications. The asymptotic distributions and the Monte Carlo procedures of the most popular univariate and multivariate portmanteau test statistics, including a new generalized variance statistic, for time series models using the powerful parallel computing framework facility, are implemented in this package. The proposed package has a general mechanism where the user can compute the test statistic for the diagnostic checking of any time series models. This package is also useful for simulating univariate and multivariate seasonal and nonseasonal ARIMA/VARIMA time series with finite and infinite variances, testing for stationarity and invertibility, and estimating parameters from stable distributions.