Mathematical Sciences Discipline, School of Science, RMIT University, Melbourne, Victoria, Australia.
PLoS One. 2020 Feb 21;15(2):e0228812. doi: 10.1371/journal.pone.0228812. eCollection 2020.
In this article, we introduce the R package dLagM for the implementation of distributed lag models and autoregressive distributed lag (ARDL) bounds testing to explore the short and long-run relationships between dependent and independent time series. Distributed lag models constitute a large class of time series regression models including the ARDL models used for cointegration analysis. The dLagM package provides a user-friendly and flexible environment for the implementation of the finite linear, polynomial, Koyck, and ARDL models and ARDL bounds cointegration test. Particularly, in this article, a new search algorithm to specify the orders of ARDL bounds testing is proposed and implemented by the dLagM package. Main features and input/output structures of the dLagM package and use of the proposed algorithm are illustrated over the datasets included in the package. Features of dLagM package are benchmarked with some mainstream software used to implement distributed lag models and ARDLs.
在本文中，我们介绍了 R 包 dLagM，用于实现分布式滞后模型和自回归分布式滞后 (ARDL) 边界检验，以探索依赖时间序列和独立时间序列之间的短期和长期关系。分布式滞后模型构成了一类大型时间序列回归模型，包括用于协整分析的 ARDL 模型。dLagM 包为实现有限线性、多项式、Koyck 和 ARDL 模型以及 ARDL 边界协整检验提供了一个用户友好且灵活的环境。特别是，在本文中，我们提出并实现了一种新的搜索算法来指定 ARDL 边界检验的阶数。通过包中包含的数据集，说明了 dLagM 包的主要功能和输入/输出结构以及所提出算法的使用。还将 dLagM 包的功能与一些用于实现分布式滞后模型和 ARDL 的主流软件进行了基准测试。
