Message-ID: <20050420001012.1781.qmail@web31302.mail.mud.yahoo.com>
Date: 2005-04-20T02:10:24Z
From: Vivek Rao
Subject: Granger causality test?
In-Reply-To: 6667
Andy,
The "ar" function in the stats packages fits either
univariate AR or multivariate AR (VAR) models,
depending on whether the input time series is
univariate or multivariate.
Regards,
Vivek Rao
--- Andy Bunn <abunn@whrc.org> wrote:
>
> Has anybody implemented the Granger causality test
> in R? The bivariate test
> is just an F test for 0 values of lagged regressors
> and can be implemented
> easily. A function that for multivariate Granger
> causes would require
> fitting a vector autoregressive model and I'm hoping
> that somebody before me
> has coded it!
>
> Thanks in advance, Andy
>
> _______________________________________________
> R-sig-finance@stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-finance
>