[R] TS code anyone?

Brian Scholl brianscholl1973 at yahoo.com
Fri Dec 28 16:45:14 CET 2001


I'm looking for some time series code that I imagine
someone has already coded at some point - I looked
through the available packages on CRAN and couldn't
find any of these (so I apologize in advance if I've
overlooked something).  Even if you might have some
code that isn't necessarily package-ready, it could be
helpful to have.

I'm looking for the following:

1. Bispectrum, and tests for linearity and
non-centrality as per Subba Rao and Gabr (1984,
p120-130) (or some way of doing an analysis of series
linearity, along with some measure of uncertainty).

2. Pre-whitened Autoregressive spectrum estimator as
per Brillinger (1981, p266).  I'm looking to do this
in the case of multiple series and I'm hoping to get
confidence intervals or some sort of measure of

3. Computing and plotting gain of a filter.   I tried
altering the code of spec.pgram and plot.spec.phase to
compute and plot the gain of a filter, and it seems to
work correctly.  But my confidence intervals (from
Koopmans 1974) don't seem to be working properly so
I'm pretty sure I've coded something wrong (the CI's
are basically the same as the gain estimate).  This is
my code for the CI's 

coh <-x$coh[,ind]

where spans is the width of the smoothing window and
pgram is the cross periodogram estimate, both being
passed (along with the gain) in my altered code as
part of the list in my version of spec.pgram.  

Any help would be appreciated on these matters.  



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