Thanks Malcolm, Peter,
This script does exactly what I want.
I am fitting two Gaussians. Judging from the result of the fit, I
gather that the 'width' parameter refers to the mathematical sigma of
the Gaussian function (i.e., I multiply it by 2.35 to get FWHM). It
might be good to state the meaning of 'width' for the Gaussian first
guess (and output) explicitly in the prologue documentation of the
script.
Another minor note about the documentation:
Under "output files" it says that the fitted model parameters are
descibed in "result.config" and "result.log", while the script in fact
writes to files called "report.config" and "report.log".
Cheers,
Matthijs
--
M.H.D. van der Wiel
Kapteyn Astronomical Institute, University of Groningen
& SRON Netherlands Institute for Space Research
[log in to unmask] // http://www.astro.rug.nl/~wiel/
Kapteynborg, room 183 [Landleven 12]
tel: +31 (0)50 363 4083
On Sun, Jan 2, 2011 at 1:02 PM, Peter W. Draper <[log in to unmask]> wrote:
> On Fri, 31 Dec 2010, Malcolm J. Currie wrote:
>
>> As far as I know specification of blended fits is only available via the
>> command-line tool $SPLAT_DIR/deblend and a text configuration file,
>> rather than a graphical interface.
>
> Yes, that is still the case.
>
>> Read the prologue of that file for more details.
>
> Indeed,
>
> Cheers,
>
> Peter.
>
|