R-beta: DSE time series package

Paul Gilbert (pgilbert@bank-banque-canada.ca)
Fri, 16 Jan 1998 11:50:36 -0500

Message-Id: <98Jan16.115834est.13454@mailgate.bank-banque-canada.ca>
Date: Fri, 16 Jan 1998 11:50:36 -0500
From: Paul Gilbert <pgilbert@bank-banque-canada.ca>
To: r-announce@stat.math.ethz.ch
Subject: R-beta: DSE time series package




An R version of my multivariate time series package (DSE) is now available at
CRAN/src/contrib/devel/dse.v.98.1alphaR.tar.gz and it should be reflected
at the mirrors shortly. The User's Guide is available at


HTML help documentation is distributed with the code and is also available
at the above web site. (The web page is a bit out of date but I hope to fix
that soon.)

The tar file includes two other packages which may be of interest by themselves.

The first, called syskern, provides a kernel of functions for S/R programmers
which is intended to help write code which is independent of operating system
differences and some small differences between S and R.
The second, called tframe, provides a kernel of functions for programming time
series methods relatively independently of the representation of time.
This is intended to make it easier to write code which can use any new/better
time representations when they appear. It also provides plotting, windowing,
and some other utility functions which are specifically intended for time

The DSE library implements an object oriented approach to time series
models (using classes and methods in R/S). This means that different
model representations can be implemented with fairly simple extensions
to the library. The original library included multi-variate state space
and ARMA (including VAR) models. The current version implements Troll
models as another class. These are run by completely separate software
(Troll from Intex Inc.) as if they were an integral part of the
library. Models and data are passed back and forth to Troll and the
results can be analyzed with already existing functions in the
library.  This also serves as an example for including other classes
and methods. [The Troll interface is not yet functional in R.]

Methods for simulating, estimating, and converting among different
model representations form the basic part of the library. Methods for
studying estimation methods and for examining the forecasting properties
of models are also documented in the User Guide.

Relative to commercial packages the library is especially useful for
time series research (such as studying estimation methods) and for
teaching. For usual time series applications there may be commercial
packages which are preferable.

A list server for questions and discussion of problems has also been set up.

To subscribe to the list send a message to
with a line in the body:

subscribe boc_dse Your Name

Subsequent message submissions to the list should be sent to

Caveats for R

I have tested the package with R0.61.1 in Solaris and R0.61 in Linux. I would
appreciate feedback about other OSs. The install procedures are currently Unix
specific (I believe) but if anyone attempts installing in Windows I would be
especially interested in knowing how to do this.

The package has been put in the devel area of CRAN because there are still
some things I would like to fix (although it is mostly working well), and
because the documentation has not been revised. I would appreciate any early
feedback as I would like to fix as many problems as possible before moving
this out of the devel category.

The package is divided into several parts, some logical and some because
the current version of R does not support large libraries.

Beware that some documentation in the code, in the HTML based help system,
and in the Brief Users' Guide is still somewhat specific to S and is not
correct for R (particularly with respect to installation). See the READ.ME
file for instruction to install in R.

Paul Gilbert

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@stat.math.ethz.ch