[AstroPy] Stable parametrisation for fitting elongated models to images

[Christoph Deil]

Hi,

apologies if this is considered off-topic for this mailing list, I didn’t know where else to ask for advice.

The question I have is how to best parametrise elongated 2D models so that model fits on images converge as much as possible, even in slightly crowded regions with overlapping sources?

Concretely Astropy uses width and length and rotation angle [1, 2] and Sherpa uses width, ellipticity and rotation angle [3].
Looking at the SExtractor manual (section 10.1) there’s also the option to use the “ellipse parameters” CXX, CYY, CXY [4].

I realise it’s a vague question, and there might be different "best parametrisations” for different data, fit statistic and optimiser, but still:
Has someone had good results (i.e. converging fits) for crowded regions with some parametrisation?
Or maybe there even exists a writeup or study comparing different parametrisations?

Thanks!
Christoph


[1] http://astropy.readthedocs.org/en/latest/api/astropy.modeling.functional_models.Gaussian2D.html <http://astropy.readthedocs.org/en/latest/api/astropy.modeling.functional_models.Gaussian2D.html>
[2] http://astropy.readthedocs.org/en/latest/api/astropy.modeling.functional_models.Ellipse2D.html <http://astropy.readthedocs.org/en/latest/api/astropy.modeling.functional_models.Ellipse2D.html>
[3] http://cxc.harvard.edu/sherpa/ahelp/normgauss2d.html <http://cxc.harvard.edu/sherpa/ahelp/normgauss2d.html>
[4] https://www.astromatic.net/pubsvn/software/sextractor/trunk/doc/sextractor.pdf <https://www.astromatic.net/pubsvn/software/sextractor/trunk/doc/sextractor.pdf>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/astropy/attachments/20150218/5c761867/attachment.html>