bootUR: An R Package for Bootstrap Unit Root Tests

TL;DR

bootUR is an R package that offers a comprehensive, reliable, and scalable framework for unit root testing in time series analysis, utilizing bootstrap methods for accurate inference.

Contribution

It introduces a unified R package that combines augmented Dickey-Fuller tests with bootstrap methods, supporting single, multiple, and panel time series with user-friendly and customizable options.

Findings

Provides accurate inference through bootstrap methods.

Supports scalable analysis of many time series.

Offers both novice-friendly and expert-level functionalities.

Abstract

Unit root tests form an essential part of any time series analysis. We provide practitioners with a single, unified framework for comprehensive and reliable unit root testing in the R package bootUR.The package's backbone is the popular augmented Dickey-Fuller test paired with a union of rejections principle, which can be performed directly on single time series or multiple (including panel) time series. Accurate inference is ensured through the use of bootstrap methods. The package addresses the needs of both novice users, by providing user-friendly and easy-to-implement functions with sensible default options, as well as expert users, by giving full user-control to adjust the tests to one's desired settings. Our parallelized C++ implementation ensures that all unit root tests are scalable to datasets containing many time series.