autoCheckGC: Granger-Causality test with automatic choice of lag structure

In Thie1e/convenience: A small collection of diverse functions for more convenient working in R

Description

This function takes an xts formatted bivariate time series as input, automatically determines the amount of lags to be used in the Granger-Causality test using an information criterion after estimating a VAR and then checks for Granger-Causality in both directions. Uses lmtest::grangertest() for the Granger-Causality testing.

Usage

autoCheckGC(x = NULL, convertToWeekly = F, lagCriterion = "AIC",
  useMinLags = F, ...)

Arguments

x	A bivariate time series of class xts
convertToWeekly	If TRUE converts the time series to weekly using xts::to.weekly(). Default FALSE
lagCriterion	The information criterion to be used when deciding on the optimal amount of lags based on the estimated VAR. "AIC" by default. Other possible values are "HQ", "SC", and "FPE". Uses vars::VARselect() to estimate the VAR.
useMinLags	If TRUE, uses the minimum amount of lags that any of the information criteria suggested. Overrides "lagCriterion". FALSE by default.
...	Other parameters to be passed on to other functions, for example to vars::VARselect().

Thie1e/convenience documentation built on May 9, 2019, 4:40 p.m.