Am 01.07.2010 21:36, schrieb Allin Cottrell:
The reason why it _should_ have an adjuster is to govern the order
for the fractional integration tests, something I had forgotten
about (though it's in the manual). But perhaps the label for the
spinner should be different in the two cases (sample periodgram /
Bartlett thingy)?
I think the proper solution would be to separate the fractional
integration tests from the estimate of the periodogram. Sure they're
related, but so is the periodogram and the spectrum, and they're
separate. Right now the fractional integration tests are actually quite
hidden, for example I found out about them basically by chance.
cheers,
sven