Astrometry ignoring --cpulimit

434 views
Skip to first unread message

Anshula Gandhi

unread,
Jul 20, 2014, 4:33:23 PM7/20/14
to
The "solve-field" command keeps running until it solves, even if it takes 1 hour.  It keeps outputting "CPU time limit reached!", but doesn't stop running.  It works on neither my Mac OS X Mavericks nor Ubuntu 14.04 operating systems.

Is there some additional flag I must include or code I must run in order to actually stop the solver once the time limit has been reached?

Example output:

Reading input file 1 of 1: "images-solved/1405880650/4211/apod3.jpg"...
Extracting sources...
simplexy: found 1389 sources.
Solving...
Reading file "images-solved/1405880650/4211/apod3.axy"...
Field 1 did not solve (index index-4211.fits, field objects 1-10).
Field 1 did not solve (index index-4211.fits, field objects 11-20).
CPU time limit reached!
Field 1 did not solve (index index-4211.fits, field objects 21-30).
Field 1 did not solve (index index-4211.fits, field objects 31-40).
Field 1 did not solve (index index-4211.fits, field objects 41-50).
CPU time limit reached!
Field 1 did not solve (index index-4211.fits, field objects 51-60).
CPU time limit reached!
Field 1 did not solve (index index-4211.fits, field objects 61-70).
CPU time limit reached!
...
Field 1 did not solve (index index-4211.fits, field objects 171-180).
  log-odds ratio 16.7247 (1.83426e+07), 3 match, 0 conflict, 3 distractors, 51 index.
  RA,Dec = (21.4954,-35.9019), pixel scale 17.3995 arcsec/pix.
  log-odds ratio 183.401 (4.46775e+79), 109 match, 5 conflict, 123 distractors, 237 index.
  RA,Dec = (130.268,-43.8692), pixel scale 38.0672 arcsec/pix.
  Hit/miss:   Hit/miss: +++-++-+++++-++-+-++++++-+++---++-++-+++++++-++++c-++c+++-+++--++---+++-----+c++--+++-+-+++------+--
CPU time limit reached!
Field 1: solved with index index-4211.fits.
Field 1 solved: writing to file images-solved/1405880650/4211/apod3.solved to indicate this.
Field: images-solved/1405880650/4211/apod3.jpg
Field center: (RA,Dec) = (130.3, -43.88) deg.
Field center: (RA H:M:S, Dec D:M:S) = (08:41:05.584, -43:52:30.244).
Field size: 8.44351 x 6.34257 degrees
Field rotation angle: up is 87.7028 degrees E of N
Creating new FITS file "images-solved/1405880650/4211/apod3.new"...
Creating index object overlay plot...
Creating annotation plot...




Dustin Lang

unread,
Jul 20, 2014, 7:17:54 PM7/20/14
to astro...@googlegroups.com
Yes, this is a known bug.  See how it says it's trying field objects 21-30, 31-40, and so on?  The time limit applies for each of those phases individually, not for the whole solve process.  Sorry for the trouble.  One change you can make to work around this is to set the "inparallel" option in the config file (eg, /usr/local/astrometry/etc/astrometry.cfg).  Then the time limit will work correctly.

cheers,
--dustin

Dustin Lang

unread,
Jul 20, 2014, 7:19:27 PM7/20/14
to astro...@googlegroups.com
PS, I've very surprised that is solves at object 171-180!  Usually it solves much earlier or not at all.  Could you post an example image here, or submit to nova.astrometry.net ?

Thanks,
--dustin

Anshula Gandhi

unread,
Jul 20, 2014, 10:23:54 PM7/20/14
to astro...@googlegroups.com
Oh okay thank you!  I didn't realize it worked that way (and I guess that explains why the CPU limit reached warning was showing up after every set of field objects).  I ended up working around it by using the timeout command before solve-field (ex: timeout 2m solve-field apod3.jpg).  But since the timeout command won't work on my Mac, and it isn't recommended to run "inparallel" on my 32 bit Ubuntu machine, I'm running "inparallel" on my Mac and using the timeout command on Ubuntu.  Thanks again for your prompt response.

The image I used was apod3.jpg from the astrometry/examples directory.  

Dustin Lang

unread,
Jul 21, 2014, 7:22:10 AM7/21/14
to astro...@googlegroups.com
Sorry for the trouble... which index files are you using?

Thanks,
--dustin

Anshula Gandhi

unread,
Jul 21, 2014, 9:15:23 AM7/21/14
to astro...@googlegroups.com
I'm using index 4211 from http://data.astrometry.net/4200/
Reply all
Reply to author
Forward
0 new messages