[Kstars-devel] Some observations regarding USNO NOMAD in KStars
Akarsh Simha
akarshsimha at gmail.com
Sun Apr 15 08:34:58 UTC 2012
Hi,
Here's a tarball containing a bunch of plots of the region around
trixel 12310. The x-axis is right ascension in hours and the y-axis is
declination in degrees.
http://www.ph.utexas.edu/~asimha/KStars-NOMAD-Tests.tar.bz2
Test 1: Shows that there are sizable differences between data dumps of
USNO NOMAD made with different criteria.
Test 2: Shows the effect of precession computed by KStars, between
J2000.0 and 2012.3
Test 3: The precessed coordinates of Test 2 were shifted by constant
RA and Dec shifts to make them coincide with the unprecessed
J2000.0 coordinates. Note the existence of spuriously
introduced precessed stars (green crosses that are not
coincident with a red plus). Note also the existence of
outliers that did not have precessed counterparts (red plusses
that are not coincident with green crosses). Note also that
these red outliers form a nearly right-angled triangle on the
top right of the image. Hmmm...
Test 4: The red outliers from the previous figure are all from Trixel
12310 for some strange reason, and not the neighbouring trixel
-- they all coincide with a blue asterisk from Trixel 12310.
Test 5: Note that some stars that were existant in trixel 12310 never
got JITUpdate()-ed despite making magnitude limits for trixel
12310 as high as 30 (red plusses without coinciding green
crosses). The reasons could be legitimate (i.e. some other
check failed; not checked yet). These could be the same stars
that were the red outliers in Test 3 (not checked).
Finally, attached is a diff containing the extremely ugly edits that I
used to do the debugging. I'm attaching them mostly for reference and
not so much for future re-use. They are applied against
29ef0d11f906963ce4d0d066df781b483a6a0ca0, which I will push into
master in a few minutes.
Regards
Akarsh
-------------- next part --------------
A non-text attachment was scrubbed...
Name: NOMADdebugpatch.diff
Type: text/x-diff
Size: 5127 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/kstars-devel/attachments/20120415/2c31332e/attachment-0001.bin>
More information about the Kstars-devel
mailing list