Problem with residual fitting in ecidentify
conor wrote on May 19, 2006
Hi,
I'm relatively new to IRAF, but I'm having a problem while calbrating the dispersion of an echelle spectrum. When identifying lines with ecidentify, I type 'f' to get the residual vs wavelength window. However when I try to return to the spectrum I get a message saying 'Singular solution', and a segmentation violation error message on the command line.
Any ideas what I'm doing wrong?
Conor
I'm relatively new to IRAF, but I'm having a problem while calbrating the dispersion of an echelle spectrum. When identifying lines with ecidentify, I type 'f' to get the residual vs wavelength window. However when I try to return to the spectrum I get a message saying 'Singular solution', and a segmentation violation error message on the command line.
Any ideas what I'm doing wrong?
Conor
Francisco Valdes wrote on May 19, 2006
Hi Connor,
A singular solution means the mathematical fit for the echelle dispersion function, a 2D function, failed. This would usually be due to insufficient points to fit. The segmentation violation is a bug in the program since it should trap errors and handle things more gracefully. The main thing I can offer is that since the echelle dispersion function is 2D (a function of position along the order and the order number) that you make sure you have a couple of lines identified in at least three orders prior to doing the fit. You may have seen in the documentation that when you extract the orders you have to make sure the aperture numbers are related to the order numbers. I don't think this is your problem but a source of occassional error.
Yours,
Frank
A singular solution means the mathematical fit for the echelle dispersion function, a 2D function, failed. This would usually be due to insufficient points to fit. The segmentation violation is a bug in the program since it should trap errors and handle things more gracefully. The main thing I can offer is that since the echelle dispersion function is 2D (a function of position along the order and the order number) that you make sure you have a couple of lines identified in at least three orders prior to doing the fit. You may have seen in the documentation that when you extract the orders you have to make sure the aperture numbers are related to the order numbers. I don't think this is your problem but a source of occassional error.
Yours,
Frank
Last post on May 19, 2006