Why didn't we catch this earlier? Well, QUEST still seems to work fine when there's no centroid error (eg if you disabled centroid-algo cog above), and also seems to work okay when there are many stars visible (eg, set --generate-saturation-photons 50 in the above examples). But I recently changed the default image generation parameters to put less stars per photo, revealing the issue.
For example, (testing on commit ad863b291795619d7629ec50f5b50f839fe06c82)
./lost database --kvector --kvector-max-distance 15 --min-mag 5.5 --output bbb.dat./lost pipeline --generate 1 --centroid-algo cog --star-id-algo py --attitude-algo quest --database bbb.dat --print-attitude -shows a different (and incorrect) result compared to
./lost pipeline --generate 1 --centroid-algo cog --star-id-algo py --attitude-algo dqm --database bbb.dat --print-attitude -Why didn't we catch this earlier? Well, QUEST still seems to work fine when there's no centroid error (eg if you disabled centroid-algo cog above), and also seems to work okay when there are many stars visible (eg, set
--generate-saturation-photons 50in the above examples). But I recently changed the default image generation parameters to put less stars per photo, revealing the issue.Pinging @karenhaining