Daophot/PSF problem
jimbojones wrote on Sep 27, 2008
I have a quick search and can't seem to find anything like the odd problem I am having with the PSF task in daophot.
Initially I was trying to create a PSF from 500 stars which I had done fine at work using PYRAF. Having finally got IRAF working on my laptop I decided to try this again using the same image/daofind result and phot result files as I had used previously.
Using a moffat25 function I get an arithmetic exception error while running PSF. I tried to narrow the problem down by reducing to just 10 stars and fiddling with various daopars/datapars. Eventually I identified the problem to be the analytic function I'm using. The task completes fine with "gauss" set but not with moffat15/moffat25/penny1/penny2/lorentz. Aside from the functional form everything is the same, same settings, image, files, etc. Does anyone have any ideas?
Cheers
James
Initially I was trying to create a PSF from 500 stars which I had done fine at work using PYRAF. Having finally got IRAF working on my laptop I decided to try this again using the same image/daofind result and phot result files as I had used previously.
Using a moffat25 function I get an arithmetic exception error while running PSF. I tried to narrow the problem down by reducing to just 10 stars and fiddling with various daopars/datapars. Eventually I identified the problem to be the analytic function I'm using. The task completes fine with "gauss" set but not with moffat15/moffat25/penny1/penny2/lorentz. Aside from the functional form everything is the same, same settings, image, files, etc. Does anyone have any ideas?
Cheers
James
Mike Fitzpatrick wrote on Sep 27, 2008
There are no known problems with the task similar to what you report and I would need to be able to reproduce it in order to give you an exact cause. Arithmetic exceptions can be either data-related (e.g. zeroes in the data, values near zero, etc) or a coding error (e.g. not checking for invalid values) but it's platform dependent whether a more precise error message gets returned by the system.
If you can tar up your image, input files and the task's parameters (e.g. output of "dpar psf", "dpar daopars", etc) with instructions on how to reproduce the problem I'll have a look. You can put this in the anonftp area on ftp://iraf.noao.edu/pub. It is unusual that only the gauss function works, e.g. a coding error in the moffat fitting shouldn't affect the penny functions. Did you reset any other parameters or can you for instance use moffat with all the rest of the defaults??
-Mike
If you can tar up your image, input files and the task's parameters (e.g. output of "dpar psf", "dpar daopars", etc) with instructions on how to reproduce the problem I'll have a look. You can put this in the anonftp area on ftp://iraf.noao.edu/pub. It is unusual that only the gauss function works, e.g. a coding error in the moffat fitting shouldn't affect the penny functions. Did you reset any other parameters or can you for instance use moffat with all the rest of the defaults??
-Mike
Last post on Sep 27, 2008