[R-sig-finance] VAR, VECM, Kalman, ... non-R software recomme ndations?

[Patrick Brandt]

On Aug 23, 2004, at 5:19 AM, Pfaff, Bernhard wrote:

> Hello to all and in particular to Dirk and Patrick,
>
> stepping in a little bit late into this thread, I am just wondering why
> nobody has mentioned GAUSS so far???
>

Umm, frustration with doing time series in Gauss got me to use R in the 
first place (back around R ~= 0.6)?  Gauss has a high license cost?  
Graphics were not very good in GAUSS?  GAUSS was my high level 
programming language of preference for about 5 years, until I realized 
I could do all of the same things faster in other stuff (RATS, R, OX -- 
things with objects that understand time series).

> As far as RATS is concerned, Dirk you want to have a look at:
>
> http://www.estima.com/catsinfo.shtml
>
> I worked with the early versions of CATS in the mid/end nineties. It 
> was a
> **must have** at that time for conducting VECM.
>

CATS is a must -- I would argue that you really cannot go wrong doing 
VECM and VARs in RATS because that is the MAIN GOAL of the software.

Patrick

[SNIP earlier parts of thread]

Patrick T. Brandt
Assistant Professor
Department of Political Science
University of North Texas
http://www.psci.unt.edu/~brandt