SPECFIT: Obtaining final errors

Andrew J. Walker wrote on May 31, 1999

I have reduced some multifibre spectra recently through dofibers. To
obtain an error estimate for the fluxes, I divide the frames by sqrt(n)
immediately
after using ccdproc on them, and then processed them in the same manner
as the
data frames but without sky subtraction.
	Uisng specfit I have been able to obtain fits for all the lines in
my spectra. However on quitting specfit I always get error messages
before
an error calculation has been produced. Can anyone suggest why this
might happen
and possible measures to work around it, or if perhaps I can use my
error spectra
with splot?

Andrew Walker
University of Wollongong

---

Andrew J. Walker wrote on May 31, 1999

Not long after writing the above message I solved my problem.
The version of IRAF we are using is V2.10.4.
I managed to find that specfit has a home page at
http://jhufos.pha.jhu.edu/~gak/specfit/specfit.html
This site has an updated version of specfit that includes
many changes from earlier versions. I don't know if this
has been incorporated in recent vesrions of IRAF. Needless
to say, I installed this in my own directory, ran it over
my spectra and it worked beautifully! Error estimates
were produced without any hiccups. Thanks must go to
G. A. Kriss for fixing this up.

Andrew Walker
University of Wollongong

---

Last post on May 31, 1999