inconsistency between assignments and the plots

[Srheft]

@tskisner @sbailey

This is the first time I'm seeing this:

fiberassign output (intermediate, merged and qa.json) don't show any unassigned fiber while the plot shows unassigned fibers with this warning while creating the plot:

WARNING: Positioner at location 7243 cannot move to target 576471653823808043 at (x, y) = (-256.77994301496796, 146.77159324479825). This should have been dected during assignment!

The assignment files show that fiber 7243 is assigned to the same TARGETID as reported in the warning (576471653823808043) but the coordinates for the target in the warning is incorrect. The target is a sky supplement location at TARGET_RA=21.88100106, TARGET_DEC=31.15930374.

the inputs and running commands are here.

any thought?

[tskisner]

Will try to investigate soon.

[tskisner]

Note that this "warning" has been promoted to an exception, and I have not seen it in my recent tests. I'll leave this open for now, but post any updates if you see it happen again.

[dstndstn]

This may have been fixed as part of #340.

Another possibility (that I've run into!) is that if the focal-plane model changes between fiberassign and plotting and you don't set --rundate consistently, then the fiber states can change underneath you.