xy list solves not working

26 views
Skip to first unread message

Jason Eastman

unread,
Nov 1, 2021, 12:56:17 PMNov 1
to astrometry
Hi,

I'm using Revision 0.84-25-gc1782b82, date Thu_Jan_7_07:49:26_2021_-0500.

I have an image that solves as a fits file with the following command:

solve-field --scale-units arcsecperpix --scale-low 0.15641275000000002 --scale-high 0.17287725 --parity pos --downsample 8 --ra 88.98636 --dec 34.265383 --radius 0.5 --quad-size-min 0.45 --quad-size-max 0.55 --cpulimit 30 --no-verify --crpix-center --no-plots --overwrite test.fits

I can't attach it here, but I've separately uploaded to the online solver:

However, I'd like to do my own source extraction (test.xy) and solve it based on that. 

When I use text2fits, it gives me test.xyls, attached. But then when I try to solve it,
it complains that the required columns 'X' and "Y' are missing, tries to treat it as an image, and fails (space limits won't allow me to include the verbatim message). 

When I write my own fits table (which is my preferred method, test.xyls.mine) with the following code (where x and y are the brightness sorted x/y positions):
                                                                                                                                                              colx = fits.Column(name='X', format='10E', array=x)
coly = fits.Column(name='Y', format='10E', array=y)
coldefs = fits.ColDefs([colx,coly])
hdu = fits.BinTableHDU.from_columns([colx,coly])
hdu.writeto('test.xyls.mine',overwrite=True)

It fails in the same way. Any idea what I'm doing wrong?

Thanks,
Jason

test.xy
test.xyls.mine
test.xyls

Dustin Lang

unread,
Nov 1, 2021, 1:18:31 PMNov 1
to Jason Eastman, astrometry
Hi,

For text2fits, it's generating string format columns; if you add command-line argument "-f ff" it will make x,y floating-point columns.

For your astropy snippet, you are creating array-valued columns.  You don't want "10E", you just want "E".  "10E" makes X and Y 2-dimensional arrays (size 24 rows x 10 values per row).

I use the "liststruc" program to find this out.

cheers,
--dustin





--
You received this message because you are subscribed to the Google Groups "astrometry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to astrometry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/astrometry/125682ef-7305-453c-9b40-2d75cc982188n%40googlegroups.com.

Jason Eastman

unread,
Nov 1, 2021, 1:28:53 PMNov 1
to astrometry
That did it for both methods -- thank you for the quick turnaround!
Reply all
Reply to author
Forward
0 new messages