[R] Nonlinear time series

Martin Maechler maechler at stat.math.ethz.ch
Tue Oct 29 15:50:51 CET 2002

>>>>> "Erin" == Erin Hodgess <hodgess at uhddx01.dt.uh.edu>
>>>>>     on Mon, 28 Oct 2002 08:14:32 -0600 (CST) writes:

    Erin> Dear R People:
    Erin> Is there code for nonlinear time series available, please?

"nonlinear" := "anything not linear"
(which is a lot).
Do you mean parametric / non-parametric?
What is the field of application?

The new StructTS() functionality in R's "ts" package can also a
be considered nonlinear. See also Prof Brian Ripley's article in
the latest R News.

    Erin> I'm looking for something that could also provide a model for
    Erin> forecasts.

    Erin> This is for R V1.5.1 on a PC.

There's our  VLMC  package for Variable Length Markov Chains.  
It models very general (non-linear) time series which
__however__ have to have values 
in a small finite set, i.e. X_t \in  {x_1, x_2, ..., x_K}.

For the current implementation, mostly K <= 26 since we are using the
alphabet in several places to represent the values.
The success stories where mostly for K=2 (binary time series) or
also K=4 for DNA-modeling.

Martin Maechler <maechler at stat.math.ethz.ch>	http://stat.ethz.ch/~maechler/
Seminar fuer Statistik, ETH-Zentrum  LEO C16	Leonhardstr. 27
ETH (Federal Inst. Technology)	8092 Zurich	SWITZERLAND
phone: x-41-1-632-3408		fax: ...-1228			<><
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list