12:22 p.m.
Hi,
not sure what the current status is of corrgm and xcorrgm: xcorrgm is
missing in the function reference, but mentioned recently in the
changelog. OTOH, doesn't the third optional parameter of corrgm cover
the xcorrgm case? I know we had a discussion about that API, but I don't
remember the outcome.
W.r.t. pergm I have a feature request: would it be possible to allow a
matrix argument to get the periodograms/spectra of all columns of the
matrix at once?
thanks,
sven
1:41 p.m.
On Tue, 16 Feb 2010, Sven Schreiber wrote:
not sure what the current status is of corrgm and xcorrgm: xcorrgm
is
missing in the function reference, but mentioned recently in the
changelog. OTOH, doesn't the third optional parameter of corrgm cover
the xcorrgm case? I know we had a discussion about that API, but I don't
remember the outcome.
There are two distinct _commands_, corrgm and xcorrgm, but there's
only one _function_, corrgm(), that combines the two cases.
W.r.t. pergm I have a feature request: would it be possible to allow
a
matrix argument to get the periodograms/spectra of all columns of the
matrix at once?
You mean, allow a general matrix for the first argument to
pergm()? (Right now it has to be a series or column vector.)
Allin.
2:39 p.m.
Allin Cottrell schrieb:
On Tue, 16 Feb 2010, Sven Schreiber wrote:
> not sure what the current status is of corrgm and xcorrgm: xcorrgm is
> missing in the function reference, but mentioned recently in the
> changelog. OTOH, doesn't the third optional parameter of corrgm cover
> the xcorrgm case? I know we had a discussion about that API, but I don't
> remember the outcome.
There are two distinct _commands_, corrgm and xcorrgm, but there's
only one _function_, corrgm(), that combines the two cases.
whoops, sorry -- now it all makes sense of course...
> W.r.t. pergm I have a feature request: would it be possible to allow a
> matrix argument to get the periodograms/spectra of all columns of the
> matrix at once?
You mean, allow a general matrix for the first argument to
pergm()? (Right now it has to be a series or column vector.)
Yes that's exactly what I meant. The return matrix would have n+1
columns, the first holding the omegas, the remaining ones the estimates.
Now that I think about it, I guess it would be logical to also allow a
list argument, once the matrix step is taken.
For example plotting the spectra:
list interestingvars = var1 var2
matrix specs25 = pergm(interestingvars,25)
gnuplot 2 3 1 --matrix=specs25
That in turn would probably beg the question why multiple selection and
right-clicking in the GUI doesn't give you the spectrum option... Oh my,
I'm sorry, I didn't want to push it this far, I can see the
ramifications now...
thanks,
sven
4:26 a.m.
El mar, 16-02-2010 a las 13:41 -0500, Allin Cottrell escribió:
On Tue, 16 Feb 2010, Sven Schreiber wrote:
> not sure what the current status is of corrgm and xcorrgm: xcorrgm is
> missing in the function reference, but mentioned recently in the
> changelog. OTOH, doesn't the third optional parameter of corrgm cover
> the xcorrgm case? I know we had a discussion about that API, but I don't
> remember the outcome.
There are two distinct _commands_, corrgm and xcorrgm, but there's
only one _function_, corrgm(), that combines the two cases.
> W.r.t. pergm I have a feature request: would it be possible to allow a
> matrix argument to get the periodograms/spectra of all columns of the
> matrix at once?
But the periodogram/spectra of a multiple time series is, for each
frequency, a matrix, not just a vector. For example if we have a vector
with Xt and Yt, the periodogram at a given frequency is a matrix with
the univariate periodogram entry of Xt, the univariate periodogram entry
of Yt and the cross-periodogram.
I do not have at hand any reference about multivariate spectral
analysis. I remember the following book was talking about that in some
chapter, but I think some of you may have any other better reference.
Lütkepohl (1987) Forecasting aggregated vector ARMA processes.
Springer-verlag. Berlin.
--
Ignacio Diaz-Emparanza
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU
Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.ea3.ehu.es
7:47 a.m.
Ignacio Diaz-Emparanza schrieb:
But the periodogram/spectra of a multiple time series is, for each
frequency, a matrix, not just a vector. For example if we have a vector
with Xt and Yt, the periodogram at a given frequency is a matrix with
the univariate periodogram entry of Xt, the univariate periodogram entry
of Yt and the cross-periodogram.
Well it depends on what you want. I wanted to treat the inputs as
separate univariate series. But you're right that there is an ambiguity
there, and possibly in the future one may want to extend the pergm()
function to do the multivariate periodogram.
So I withdraw that request.
thanks,
sven
5382
days inactive
5383
days old
4 comments
3 participants
participants (3)
-
Allin Cottrell -
Ignacio Diaz-Emparanza -
Sven Schreiber