fitcoords: bus error
margaret wrote on Jul 06, 2010
Hi,
I've executed identify and reidentify on a 1D spectrum image, and the resulting database file looks normal in content/structure compared to previous successful runs.
However, when I run fitcoords on the image and database file, e.g.,
fitcoords image
where the corresponding database file is named 'imagename' and there also exists 'imagename.fits' in the directory where I run fitcoords, I get the following error message:
Longslit coordinate fit name is imagename.
Longslit database is database.
Features from images:
imagename
ERROR: bus error
Fitcoords continues to work with this installation of IRAF on different data sets, but there is no obvious difference in the file structure -- header and image are all intact.
I saw from the forums that someone had had this error several years ago, but there was no solution described. Has anyone had any experience with an error like this?
Many thanks for any help!
Margaret
I've executed identify and reidentify on a 1D spectrum image, and the resulting database file looks normal in content/structure compared to previous successful runs.
However, when I run fitcoords on the image and database file, e.g.,
fitcoords image
where the corresponding database file is named 'imagename' and there also exists 'imagename.fits' in the directory where I run fitcoords, I get the following error message:
Longslit coordinate fit name is imagename.
Longslit database is database.
Features from images:
imagename
ERROR: bus error
Fitcoords continues to work with this installation of IRAF on different data sets, but there is no obvious difference in the file structure -- header and image are all intact.
I saw from the forums that someone had had this error several years ago, but there was no solution described. Has anyone had any experience with an error like this?
Many thanks for any help!
Margaret
Mike Fitzpatrick wrote on Jul 06, 2010
A "bus error" is a fairly generic message, usually caused by a memory corruption of some kind. Unfortunately, there are no known issues with the task and not enough information to say specifically what the problem might be.
Could you please tell us what platform you are using (OS and IRAF versions) and upload your database file and FITS image to the anon ftp at ftp://iraf.noao.edu/pub so we can try to reproduce the problem?
It would also help to see the parameters you used, i.e. a "dpar fitscoords" output listing.
-Mike
Could you please tell us what platform you are using (OS and IRAF versions) and upload your database file and FITS image to the anon ftp at ftp://iraf.noao.edu/pub so we can try to reproduce the problem?
It would also help to see the parameters you used, i.e. a "dpar fitscoords" output listing.
-Mike
margaret wrote on Jul 06, 2010
Thanks for the quick reply.
I have included the parameters used below, and I will upload the fits and database file to the ftp server just after this.
fitcoords.images =
fitcoords.fitname = ""
fitcoords.interactive = yes
fitcoords.combine = no
fitcoords.database = "database"
fitcoords.deletions = "deletions.db"
fitcoords.function = "chebyshev"
fitcoords.xorder = 6
fitcoords.yorder = 6
fitcoords.logfiles = "STDOUT,logfile"
fitcoords.plotfile = "plotfile"
fitcoords.graphics = "stdgraph"
fitcoords.cursor = ""
fitcoords.mode = "ql"
In the course of trying to troubleshoot this, we've found that stripping down the contents of the header seems to allow fitcoords to run normally. Perhaps the header is too long?
I have included the parameters used below, and I will upload the fits and database file to the ftp server just after this.
fitcoords.images =
fitcoords.fitname = ""
fitcoords.interactive = yes
fitcoords.combine = no
fitcoords.database = "database"
fitcoords.deletions = "deletions.db"
fitcoords.function = "chebyshev"
fitcoords.xorder = 6
fitcoords.yorder = 6
fitcoords.logfiles = "STDOUT,logfile"
fitcoords.plotfile = "plotfile"
fitcoords.graphics = "stdgraph"
fitcoords.cursor = ""
fitcoords.mode = "ql"
In the course of trying to troubleshoot this, we've found that stripping down the contents of the header seems to allow fitcoords to run normally. Perhaps the header is too long?
margaret wrote on Jul 06, 2010
The system version is Mac OS 10.5.8 and IRAF version 2.14.
Mike Fitzpatrick wrote on Jul 06, 2010
Perhaps the header is too long?
Perhaps. One way to check is to increase the size of the 'min_lenuserarea' variable to see if it makes a difference, e.g.
cl> reset min_lenuserarea = 256000
cl> gflush
cl> fitcoords ........
OTOH, it could have been something you removed actually had meaning, e.g. a WCS or dispersion keyword of some kind?
margaret wrote on Jul 06, 2010
Apparently the latter is more likely, as changing the 'min_lenuserarea' variable did not seem to make a difference.
Alessandra Candian wrote on Jul 06, 2010
I use to have the same problem!
To solve it I just changed the value for field WCSDIM in the image header from 4 to 2: I readin some old cookbook that identify and reidentify change the WCS information on the header of the file and this makes fitcoord to produce "bus error"
Hope it can help!
Cheers
A.
To solve it I just changed the value for field WCSDIM in the image header from 4 to 2: I readin some old cookbook that identify and reidentify change the WCS information on the header of the file and this makes fitcoord to produce "bus error"
Hope it can help!
Cheers
A.
Last post on Jul 06, 2010