From: Roberta Johnson <rjohnson@ipac.caltech.edu ([email]rjohnson@ipac.caltech.edu[/email])>
Date: Oct 14, 2005 1:56 PM
Subject: mccmatch and mkapfile
To: iraf@noao.edu ([email]iraf@noao.edu[/email])
Hello,
I'm a grad student using IRAF and I have a few problems.
First one, when I use msccmatch ( and I do a fine search) on a long list
of files, I get an error message: "ERROR : floating point divide by
zero^G " and the program stops.
However, if I try to run thatparticular file again, I have no
problems. What is happening, is that
msccmatch if crashing everyother file, so it's essentially forcing me
to run the files one by one.
(Which is rather annoying becausethere are a lot of files.)
Is there any way, to make msccmatch not crash so much.
My other problem is rather similar, but with the mkapfile program. When I
try to run a list, I get anerror messages that says : "cog model could
not fit". But again, running
the files through separatelyseems to work. How can I make it so that
it will run a list through?
And I suppose while I'm asking questions, I might as well ask about pdump.
I'm not really sure ofwhat this one does. I know it 'dumps' certain
fields, but where is kinda
my question. Does it changethe headers, or does it just read out
certain fields? And where does it
put these fields, does it make alist of any kind that you can use later?
Thank you for your time and your help.
Sincerely,
Roberta Johnson

[b:0c07738b39]From: [/b:0c07738b39]Frank Valdes <valdes@noao.edu ([email]valdes@noao.edu[/email])>
[b:0c07738b39]Date: [/b:0c07738b39]November 3, 2005 2:55:49 PM MST
[b:0c07738b39]To: [/b:0c07738b39]Roberta Johnson <rjohnson@ipac.caltech.edu ([email]rjohnson@ipac.caltech.edu[/email])>
[b:0c07738b39]Subject: [/b:0c07738b39][b:0c07738b39]Re: mccmatch and mkapfile[/b:0c07738b39]
Hello Roberta,
I'm sorry for the delay in replying.
I am not aware of the problem with msccmatch you report.  Could you send me an example showing the command you type, the parameters (i.e. lpar msccmatch), the list you input (I assume this is an @file),
and any other information?  Are you having the program automatically get the reference stars or are you supplying the coordinate files?  Also tell me the version of IRAF (the first two lines of the login banner) and the type of computer.
The problem with mkapfile is similar and I would want similar information.  I need to understand how you are calling it and if I can reproduce the error.
I'm not quite sure what you are asking about pdump.  It prints information to the standard output which can then be directed to a file.  It can be useful for extracting information to me used by programs outside of the photometry package that read certain information from a text file in a certain order.
Yours,
Frank Valdes
[b:0c07738b39]From: [/b:0c07738b39]rjohnson@ipac.caltech.edu ([email]rjohnson@ipac.caltech.edu[/email])
[b:0c07738b39]Date: [/b:0c07738b39]November 7, 2005 12:20:10 PM MST
[b:0c07738b39]To: [/b:0c07738b39]"Frank Valdes" <valdes@noao.edu ([email]valdes@noao.edu[/email])>
[b:0c07738b39]Subject: [/b:0c07738b39][b:0c07738b39]Re: mccmatch and mkapfile[/b:0c07738b39]
Hi,
Thanks for getting back to me. I was able to fix my problems with mkapfile
and pdump.
But msccmatch is still giving me grief. I have a script, where I run it 3
times, each time the search gets more refined. The input looks like this
unlearn mscgetcatalog
mscgetcatalog.magmin=10
mscgetcatalog.magmax=16
unlearn msccmatch
msccmatch.coords="!mscgetcatalog $I $C"
msccmatch.usebpm=no
msccmatch.verbose=yes
msccmatch.nsearch=50    (which in the next runs become 100, 400)
msccmatch.search=120    (I try to get it to 20 , 1)
msccmatch.rsearch=2      ( goes to .5 and .01)
msccmatch.cbox=15      (goes to 10 then 5)
msccmatch.maxshif=7   (goes to 5 ten 3)
msccmatch.csig=0.1
msccmatch.cfrac=0.5
msccmatch.listcoo=yes
msccmatch.nfit=4
msccmatch.rms=2
msccmatch.fitgeom="general"
msccmatch.reject=3
msccmatch.update=yes
msccmatch.interac=no
msccmatch.fit=no
msccmatch.accept=yes
msccmatch @object.list   (which are .fits files)
This works for the coarse run, but the second and third time, I run into
problems.
A lot of time the warning 'automatic search failed' comes up ut then the
program just skips to the next file. When I get the 'floating point error'
the program crashes. I've tried running it as a script and putting
everything in by hand.
I hope you can help.
Thanks for your time.
Sincerely,
Roberta
[quote:0c07738b39][quote:0c07738b39][/quote:0c07738b39][/quote:0c07738b39][b:0c07738b39]From: [/b:0c07738b39]Frank Valdes <valdes@noao.edu ([email]valdes@noao.edu[/email])>
[b:0c07738b39]Date: [/b:0c07738b39]November 14, 2005 1:42:53 PM MST
[b:0c07738b39]To: [/b:0c07738b39]rjohnson@ipac.caltech.edu ([email]rjohnson@ipac.caltech.edu[/email])
[b:0c07738b39]Subject: [/b:0c07738b39][b:0c07738b39]Re: mccmatch and mkapfile[/b:0c07738b39]
Hi Roberta,
Thanks for the information.  I think I understand what you are doing.  The first thing I would suggest is that you put a "flpr" before each msccmatch execution.  This forces the task to restart each time and would eliminate any history source of the problem.  If it still fails it would then be possible to just debug a specific circumstance.   Along those same lines you could also break the script down so that each exposure is handled as a separate execution.  For example:
list = "object.list"
while (fscan (list, s1) != EOF) {
    flpr
    printf ("Running %s:n", s1)
    msccmatch (s1)
}
list = ""
At this point any crash will be isolated to a specific set of parameters for a specific image and, if you want, I could get the image and try and reproduce the error.
Yours,
Frank