iraf.net

NUMBER:	322
MODULE:	daophot.nstar
SYSTEM:	V2.10.3 and later
DATE:	Fri Jan  5 11:48:10 MST 1996
FROM:	davis

BUG:	In rare circumstances nstar can fail with floating operand errors due
	to a singular fitting matrix. This circumstance is most likely to occur
	when the fitting radius is small, the sky is being refitted, and the
	initial guesses for the parameters to be fit are very poor.

	If a group is too large to be fit, e.g. the group size is greater
	than the value of the maxgroup parameter, nstar may fail with
	memory corruption errors. This occurs because the error code
	array is not being properly reallocated in this circumstance and
	the error code values run off the end of the error code array.

STATUS:	Fixed for iraf 2.11 and in the ftp archive version of digiphot. A
	workaround for the second problem is to increase the value of the
	maxgroup parameter. Tweaking the fitting parameters, e.g. increasing
	the value of fitrad, may workaround the first problem.