Problem : ‘ERROR: floating point invalid operation’
elgreco wrote on Apr 29, 2008
Hi,
I'm using IRAF (V. 2.14) to do some wavelenght calibration of spectra.
1) I encounter this error message, by running the task apall (noao -> imred -> kpnoslit -> apall):
ERROR: floating point invalid operation
I do not know what that exactly means, but it occurs only after trimming my images. If I don't trim them I do not get this message but I need to do it because if I don't, I cannot use afterwards my xenon.fits file with the task identify because the overscan of the xenon spectrum image make the plot going to values 10 to the 33!!! so that I cannot give the wavelenght values in order to calibrate (by giving the apertures, typing m) because I cannot see the spectrum even by zooming in I don't know how many times....
2) Another question :roll:
Is it possible to say to the apall task to read the spectrum in a way so that I don't have as output file a spectrum in the wrong way (bluewards=rightwards, redwards=leftwards)...
Thank you
I'm using IRAF (V. 2.14) to do some wavelenght calibration of spectra.
1) I encounter this error message, by running the task apall (noao -> imred -> kpnoslit -> apall):
ERROR: floating point invalid operation
I do not know what that exactly means, but it occurs only after trimming my images. If I don't trim them I do not get this message but I need to do it because if I don't, I cannot use afterwards my xenon.fits file with the task identify because the overscan of the xenon spectrum image make the plot going to values 10 to the 33!!! so that I cannot give the wavelenght values in order to calibrate (by giving the apertures, typing m) because I cannot see the spectrum even by zooming in I don't know how many times....
2) Another question :roll:
Is it possible to say to the apall task to read the spectrum in a way so that I don't have as output file a spectrum in the wrong way (bluewards=rightwards, redwards=leftwards)...
Thank you
Francisco Valdes wrote on Apr 29, 2008
Hi,
You refer to using trimming to remove the overscan. Typically this is done using CCDPROC. I can't say why you get the error but I suggest you try the following. Do your trimming as you have been doing. Do the same trim for both your spectrum and your arc. If you want flip the spectrum at the same time you can do it with an image section. For example:
The second line may help solve any confusions later. This tells IRAF to forget that the original raw image was trimmed and flipped.
Maybe this will solve your problems.
Yours,
Frank Valdes
You refer to using trimming to remove the overscan. Typically this is done using CCDPROC. I can't say why you get the error but I suggest you try the following. Do your trimming as you have been doing. Do the same trim for both your spectrum and your arc. If you want flip the spectrum at the same time you can do it with an image section. For example:
imcopy raw[1024:33,*] trimmed
wcsreset trimmed wcs=physical
The second line may help solve any confusions later. This tells IRAF to forget that the original raw image was trimmed and flipped.
Maybe this will solve your problems.
Yours,
Frank Valdes
Last post on Apr 29, 2008