splot

Guest wrote on Mar 20, 2006

I can't seem to get error calculations when I fit a gaussian profile with splot. I have set nerrsample=50 and both sigma0 and invgain > 0. I don't get any warning or error messages, I just don't get any errors with the output. I am using IRAF v2.12 with fedora. Any advice would be appreciated.

Thanks

---

Francisco Valdes wrote on Mar 20, 2006

I believe the usual cause of this behavior is data which has negative values. In order to calculate errors with the uncertainty model used by splot it assumes the data are in positive counts. It is true that there is no error message.

If this is not the cause try searching the forums on this site for similar questions.

Yours,
Frank Valdes

---

Guest wrote on Mar 20, 2006

Frank

There were indeed negative points in my spectrum. I thought that a previous update to splot meant that negative values were treated as being zero and the error calculation would proceed as normal. I did not get any error message listed at all when the error caculation was not done.

Thanks.

---

Last post on Mar 20, 2006