#92 fitgauss fails to fit simple absorption line

[Anonymous]

fitgauss failed miserably when I tried to have it fit a
simple uncomplicated absorption line. It seems that
either fitgauss is assuming that the gaussian has to be
positive relative to the baseline or that the baseline
is near zero - both of which are bad assumptions for an
absorption spectra.

A postscript plot of the absorption spectrum and the
bad fit is attached.

A (keep) fits file with the spectrum is available in
Green Bank at /users/tminter/fitgauss-fail.fits (along
with the index file and a copy of the attached
postscript file - all which the same file name but
different extensions).

[Nobody/Anonymous]

postscript plot of fitgauss failure