Greetings O-Matrix users,

Early this week version 2.1 of the Statistical Time Series Analysis
toolbox was released for O-Matrix. This release offers a couple of
major modeling enhancements and several other minor updates.
Specifically, the new version of STSA contains two new directories,
NONPAR and POD that extend modeling with STSA for nonlinear and
nonparametric time series (NONPAR) and for series that are
decomposable into trends and cyclical components (POD).

The NONPAR directory contains functions that are adapted from the
book of Fan and Yao "Nonlinear TIme Series: Parametric and
Nonparametric Models", Springer-Verlag, 2003, and are suitable for a
wide variety of problems were a parametric form for modeling is
either not appropriate or not desirable. Many of the functions in
NONPAR are not directly available in other software programs - in
addition, they offer automated bandwindth selection and forecasting
with ease. Most functions also offer automatic plotting options.

The POD directory contains functions for analyzing time series using
the method of Singular Spectrum Analysis (SSA) which is mainly
suitable for series that exhibit trends and cyclical components which
we wish to extract and possibly to forecast separately. SSA has been
proven successful in many modeling situations, especially in the
natural sciences.

If you have any comments or recommendations about this new release of
STSA, including bugs, or about potential future enhancements that you
may find useful in your line of work I would be happy to know about
them.

Happy computing!

Dimitrios D. Thomakos
STSA developer