On Wed, 28 Sep 2005, Malcolm J. Currie wrote:
>>> So is it a type of surface-fitting avoiding genuine features?
>>
>> Nearly, from the ACSIS point of view it is to remove the background of
>> each spectrum. So you pick some velocity ranges that avoid features, fit a
>> load of polynomials and subtract them.
>
> Yes. That's the way I read it. Apologies if "genuine" was ambiguous.
> FITLINES seemed to be for background or continuum removal but only along
> one axis. Just wondering where to classify it and if there's a better
> list of related applications, and most importantly whether or not the
> name is misleading. FITVECTORS? There's a FITCONT in Figaro.
>
Yes. FITLINES is actually doing the reverse (and in the ACSIS sense trying
to avoid the lines). How about MFITCONT in the sense of MLINPLOT vs
LINPLOT - it's fitting multiple continuum signals (and they won't be
smoothly varying across the spatial plane necessarily as they are
essentially independent fits).
--
Tim Jenness
JAC software
http://www.jach.hawaii.edu/~timj
|