Question about the ‘center’ task
Bosco Yung wrote on May 01, 2007
Hi everyone!
I was using the 'center' task in 'noao.digiphot.apphot' to find out the center coordinates of a star (point source). It was done by fitting a gaussian function. The output results not only include the fitted center coordinates (X,Y) but also with the errors of them (i.e. Xerr and Yerr). However, I do not know how the errors were computed. I could not find it on the help page. If anyone know it please advise me. Thank you!
Bosco
I was using the 'center' task in 'noao.digiphot.apphot' to find out the center coordinates of a star (point source). It was done by fitting a gaussian function. The output results not only include the fitted center coordinates (X,Y) but also with the errors of them (i.e. Xerr and Yerr). However, I do not know how the errors were computed. I could not find it on the help page. If anyone know it please advise me. Thank you!
Bosco
Mike Fitzpatrick wrote on May 01, 2007
Bosco,
The CENTER task simply does a least-squares fit to the 1-D marginal distributions of the centering box. The error reported is then just the error of the fitted gaussian center position in each axis. The program comments from the code which actually computes the errors might also help:
Cheers,
-Mike
The CENTER task simply does a least-squares fit to the 1-D marginal distributions of the centering box. The error reported is then just the error of the fitted gaussian center position in each axis. The program comments from the code which actually computes the errors might also help:
# NLERRORS -- Procedure to calculate the reduced chi-squared of the fit
# and the errors of the coefficients. First the variance
# and the reduced chi-squared of the fit are estimated. If these two
# quantities are identical the variance is used to scale the errors
# in the coefficients. The errors in the coefficients are proportional
# to the inverse diagonal elements of MATRIX.
Cheers,
-Mike
Bosco Yung wrote on May 01, 2007
Hi Mike,
Thank you so much for your help!
However, you have mentioned that 'the error reported is then just the error of the fitted gaussian center position in each axis'. Which method do they use to find out the fitting error?
Cheers,
Bosco
Thank you so much for your help!
However, you have mentioned that 'the error reported is then just the error of the fitted gaussian center position in each axis'. Which method do they use to find out the fitting error?
Cheers,
Bosco
Last post on May 01, 2007