esmucler/odpc: One-Sided Dynamic Principal Components
Functions to compute the one-sided dynamic principal components ('odpc') introduced in Peña, Smucler and Yohai (2019) <DOI:10.1080/01621459.2018.1520117>. 'odpc' is a novel dimension reduction technique for multivariate time series, that is useful for forecasting. These dynamic principal components are defined as the linear combinations of the present and past values of the series that minimize the reconstruction mean squared error.
Getting started
Browse package contents
Package details |
|
|---|---|
| Author | Daniel Peña <daniel.pena@uc3m.es>, Ezequiel Smucler <ezequiels.90@gmail.com>, Victor Yohai <vyohai@dm.uba.ar> |
| Maintainer | Ezequiel Smucler <ezequiels.90@gmail.com> |
| License | GPL (>= 2) |
| Version | 2.0.5 |
| Package repository | View on GitHub |
| Installation |
Install the latest version of this package by entering the following in R:
|
