definition of ellipticity
cody wrote on Feb 26, 2008
Hi-
I am measuring the ellipticity of some stellar psfs (to see how elongated they are), and I'm trying to see what iraf task to use. I picked a star and did imexam with cursor 'a' on it to get one measurement. I also used fitpsf in elgauss mode to get xsigma and ysigma, as well as in moments mode to get ellipticity that way. The issue is that all three methods give somewhat different measures of the ellipticity (ranging from 0.06-0.17). This is probably related to the fact that the star is not a perfect ellipse but rather a distorted weird shape. But I still want to decide on which task to use to measure just how distorted it is. I attempted to find out exactly what ellipticity means in these various tasks, but I can't locate any documentation. It can be defined in a number of different ways. Does anyone know what definition iraf uses, particularly with imexam?
Thanks,
Ann Marie
I am measuring the ellipticity of some stellar psfs (to see how elongated they are), and I'm trying to see what iraf task to use. I picked a star and did imexam with cursor 'a' on it to get one measurement. I also used fitpsf in elgauss mode to get xsigma and ysigma, as well as in moments mode to get ellipticity that way. The issue is that all three methods give somewhat different measures of the ellipticity (ranging from 0.06-0.17). This is probably related to the fact that the star is not a perfect ellipse but rather a distorted weird shape. But I still want to decide on which task to use to measure just how distorted it is. I attempted to find out exactly what ellipticity means in these various tasks, but I can't locate any documentation. It can be defined in a number of different ways. Does anyone know what definition iraf uses, particularly with imexam?
Thanks,
Ann Marie
Mike Fitzpatrick wrote on Feb 26, 2008
There is no one standard definition of ellipticity used by all tasks, it depends on the algorithm and task. In the case of IMEXAM and the 'a' keystroke, however, the equation used is spelled out in the help page description of the keystroke. Specifically,
For FITPSF in elgauss mode I believe it's a parameter of the fit, I'd have to look at the code to be more precise.
-Mike
In addition to the flux, the second intensity moments are used to compute an ellipticity and position angle. The equations defining the moments and related parameters are:
Mxx = sum (x * x * I) / sum (I)
Myy = sum (y * y * I) / sum (I)
Mxy = sum (x * y * I) / sum (I)
e = sqrt ((Mxx - Myy) ** 2 + (2 * Mxy) ** 2) / (Mxx + Myy)
pa = 0.5 * atan (2 * Mxy / (Mxx - Myy))
For FITPSF in elgauss mode I believe it's a parameter of the fit, I'd have to look at the code to be more precise.
-Mike
Last post on Feb 26, 2008