Xdimsum Makemask
dcrain wrote on Jul 18, 2007
Hello,
I am trying to run the makemask procedure and I am geting a floating point error. I am not sure why this error is occuring so I need some help to figure it out. Here is the error and my settings for the run.
xd> epar makemask
timeout - terminal type set wrong? (`stty termtype' to reset)
I R A F
Image Reduction and Analysis Facility
PACKAGE = xdimsum,
TASK = makemask
inlist = @skytest_k.lst The input images used to compute the masks
outlist = @skytest_k.lst The output masks
(hinlist= ) The list of images to add BPM keywords to
(subsamp= 1) Block averaging factor before median filtering
(checkli= yes) Check data limits before filtering?
(zmin = -1000.) Minimum data value for fmedian
(zmax = 3000.) Maximum data value for fmedian
(filtsiz= 15) Median filter size for local sky evaluation
(nsmooth= 3) Boxcar smoothing size before thresholding
(statsec= [50:1998,50:1998]) Image region for computing sky statistics
(nsigrej= 4.) The nsigma sky statistics rejection limit
(maxiter= 20) The maximum number of sky statistics iterations
(thresht= nsigma) Thresholding type: nsigma or constant
(nsigthr= 2.) Threshold for masking in sky sigma
(constth= 4.) Constant threshold above sky for masking
(negthre= no) Set negative as well as positive thresholds ?
(ngrow = 1) Half-width of box to grow around masked objects
(verbose= yes) Verbose output?
Working on image SKY_061103_0175.sub
Subtracting local sky
Median filtering sky
Data minimum = -32606658. maximum = 554779520.
Truncating data range -1000. to 3000.
Computing sky statistics
Mean=0.05885456 Rms=67.27022 Med=-3.201796 Mode=-3.005975
Smoothing image before thresholding
Thresholding image at level 41.645017333333
Saving mask as SKY_061103_0175.sub.pl
ERROR on line 358: floating point invalid operation
makemask (mode=h)
I am trying to run the makemask procedure and I am geting a floating point error. I am not sure why this error is occuring so I need some help to figure it out. Here is the error and my settings for the run.
xd> epar makemask
timeout - terminal type set wrong? (`stty termtype' to reset)
I R A F
Image Reduction and Analysis Facility
PACKAGE = xdimsum,
TASK = makemask
inlist = @skytest_k.lst The input images used to compute the masks
outlist = @skytest_k.lst The output masks
(hinlist= ) The list of images to add BPM keywords to
(subsamp= 1) Block averaging factor before median filtering
(checkli= yes) Check data limits before filtering?
(zmin = -1000.) Minimum data value for fmedian
(zmax = 3000.) Maximum data value for fmedian
(filtsiz= 15) Median filter size for local sky evaluation
(nsmooth= 3) Boxcar smoothing size before thresholding
(statsec= [50:1998,50:1998]) Image region for computing sky statistics
(nsigrej= 4.) The nsigma sky statistics rejection limit
(maxiter= 20) The maximum number of sky statistics iterations
(thresht= nsigma) Thresholding type: nsigma or constant
(nsigthr= 2.) Threshold for masking in sky sigma
(constth= 4.) Constant threshold above sky for masking
(negthre= no) Set negative as well as positive thresholds ?
(ngrow = 1) Half-width of box to grow around masked objects
(verbose= yes) Verbose output?
Working on image SKY_061103_0175.sub
Subtracting local sky
Median filtering sky
Data minimum = -32606658. maximum = 554779520.
Truncating data range -1000. to 3000.
Computing sky statistics
Mean=0.05885456 Rms=67.27022 Med=-3.201796 Mode=-3.005975
Smoothing image before thresholding
Thresholding image at level 41.645017333333
Saving mask as SKY_061103_0175.sub.pl
ERROR on line 358: floating point invalid operation
makemask (mode=h)
Mike Fitzpatrick wrote on Jul 18, 2007
If this is a MacIntel system then you may be seeing a known problem with an early set of binaries in which random FPE errors would be thrown. One way to check is to see if the binaries date from before Nov06 using the command
If so, then you'll need to update to the most recent binaries from the Downloads link above.
MAKEMASK is a script task but I can't tell from your output where the crash is happening. If you type 'd_trace' before running the task it will trace the script execution making that much simpler to figure out. It would also help to know which platform and IRAF version you are using, as well as whether there is anything peculiar about the data (e.g. values normalized to some very small number, lots of zeroes, etc) We would need to be able to reproduce the problem to track down a new bug, write back with more information if you still have problems.
Cheers,
-Mike
cl> dir bin$x_images.e long+If so, then you'll need to update to the most recent binaries from the Downloads link above.
MAKEMASK is a script task but I can't tell from your output where the crash is happening. If you type 'd_trace' before running the task it will trace the script execution making that much simpler to figure out. It would also help to know which platform and IRAF version you are using, as well as whether there is anything peculiar about the data (e.g. values normalized to some very small number, lots of zeroes, etc) We would need to be able to reproduce the problem to track down a new bug, write back with more information if you still have problems.
Cheers,
-Mike
dcrain wrote on Jul 18, 2007
I checked the binaries they were 2004. Then I decide to try the d_trace to see what it would show me. Fianally I just reran the script and it worked. The only difference was that I did not run it useing epar.
Thanks for your help.
Dan
Thanks for your help.
Dan
Last post on Jul 18, 2007