[AstroPy] ximgfit Python application

ptak at cmu.edu ptak at cmu.edu
Tue Apr 4 14:17:06 CDT 2000


Hello,
I have a python astronomy program that is fairly usable right now.  Its
"version" is 0.96 because there are still some features I would like to
add, and most of the debugging has occured through my use of the
program rather than a real attempt to break it (and hence the most robust
features are the ones I use often).  Its main purpose is to fit FITS
images with spatial models with emphasis on X-ray images where the pixels
represent low numbers of counts, i.e., in the Poisson statistical regime,
optionally also convolving the model with a PSF image (using the fftw
library).  Its default fit statistic is chi^2 so it is usuable for
optical, etc. work also.  The bells and wistles that help out with X-ray
data are:
- Use of the Cash's C (Poissonian) fit statistic
- Automated stepping of the fit parameters to get errors, including
starting over when a new minimum is found
- Genetic algorithm fitting (implimented entirely in Python) as a
global-minimization fit technique
- XSPEC-style syntax (XSPEC is the defacto standard X-ray spectral fitting
package).

Since it combines Numeric, fftw, and my C++ fits routines, I haven't
gotten around to a source release but Linux and Solaris binaries (as well
as an online manual) are available at 
http://snooker.phys.cmu.edu/ximgfit/

I am using it daily (both interactively and in bgd. jobs) on Chandra and
ROSAT data.  If nothing else, it is an example of an actual
astronomy application written (mostly) in Python.

This program is the spatial component of a larger project to automate the
analysis of X-ray data (particularly the error search feature).

Cheers,
Andy Ptak
ptak at cmu.edu



_____________________________________________________
AstroPy mailing list  -  astropy at stsci.edu
http://lheawww.gsfc.nasa.gov/~bridgman/AstroPy/



More information about the AstroPy mailing list