FCVAR

An R Package for the Fractionally Cointegrated VAR Model

Description

Estimation and inference using the Fractionally Cointegrated Vector Autoregressive (VAR) model. It includes functions for model specification, including lag selection and cointegration rank selection, as well as a comprehensive set of options for hypothesis testing, including tests of hypotheses on the cointegrating relations, the adjustment coefficients and the fractional differencing parameters. See the file FCVAR_README.pdf for examples and the Webpage https://sites.google.com/view/mortennielsen/software for more information about the FCVAR model.

How to Install FCVAR

Install the latest release using the install.packages() function:

install.packages("FCVAR")
library(FCVAR)

Alternatively, you can install the development version on GitHub using the devtools package:

library(devtools)
devtools::install_github("LeeMorinUCF/FCVAR")

However, the version on CRAN is recommended because that version is tested and vetted for submission to CRAN.

Currently, the development version has no new features that are not already available in the released version available on CRAN.

Downstream dependencies

There are currently no downstream dependencies for this package.

Test environments

Checks on Rhub:

• Fedora Linux, R-devel, clang, gfortran
• Ubuntu Linux 20.04.1 LTS, R-release, GCC
• Windows Server 2022, R-devel, 64 bit

Checks on local machine:

• Windows 10 Enterprise, version 20H2, OS build 19042.985, R 4.0.5

Checks on win-builder:

• devel and release

Checks on M1mac

• using R version 4.2.0 RC (2022-04-15 r82193)
• using platform: aarch64-apple-darwin20 (64-bit)

R CMD check results

There were no ERRORs or WARNINGs.

There was one NOTE from the check on Windows Server 2022, R-devel, 64 bit on Rhub:

• checking for detritus in the temp directory ... NOTE 'lastMiKTeXException'
• Apparently, this is a known issue with Rhub and does not suggest a problem with the package.

Release Notes

May 4, 2022: FCVAR v0.1.4

• Removed test case that gave false alarm

from tests run on M1mac platform.

May 4, 2022: FCVAR v0.1.3

• Adjustment to test case to prevent false alarm

from tests run on M1mac platform.

March 10, 2022: FCVAR v0.1.2

• Adjustment for issue with counting the number of free parameters in constrained model.
• Added warning message for rank test with rank above 12. In this case,

p-values are only available by simulation with the FCVARbootRank() function.

August 6, 2021: FCVAR v0.1.1

• For models with restricted estimation, skipped tests on CRAN

that compare printed output and require exact equality (i.e. without tolerance), but retained tests on same with default tolerance for numerical equality.

August 5, 2021: FCVAR v0.1.1

• Modified tolerance of bootstrap test results for Solaris platforms.

Test results are within acceptable tolerance.

August 4, 2021: Fourth submission (FCVAR v0.1.0).

• Added link to reference paper in the DESCRIPTION file.
• Changed examples with \dontrun to \donttest for examples

with run time than took longer than 5s.

• Removed example from demo that changed par() settings.
• For function plot.FCVAR_grid() that changes par() settings,

because it creates a figure with thinner margins, inserted command on.exit(par(oldpar)) to restore user's settings, immediately after the change to par().

July 30, 2021: Third submission (FCVAR v0.1.0).

• Excluded examples with \dontrun if run time took longer than 5s.

July 30, 2021: Second submission (FCVAR v0.1.0).

• Added cran-comments.md to .Rbuildignore.

July 30, 2021: First submission (FCVAR v0.1.0).

• First submission of the FCVAR package.
• Added a NEWS.md file to track changes to the package.