New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Change assignment of leftover fibers #321
Conversation
targets with zero obs remaining. Also remove some stale debug print statements.
thanks a lot @tskisner. |
Thanks for testing @araichoor, I can definitely add a switch to those functions so that the old behavior can be enabled for comparison. |
I realize that I forgot to check if targets which were receiving a fiber before still do after this change: I notice that ~50/350 TARGET_STATE="CALIB" targets now get one observation less. |
Hello @araichoor , I am not familiar with this Are the CALIB targets you are referring to standards, or something else? |
I m not very familiar with those either, but with checking: all the assigned TARGET_STATE="CALIB" targets have FA_TYPE=3, so those should be fiberassign STANDARD, right? |
Everything in the MTL ledger that is labeled In short: Treat |
Sorry, I had thought we were already using this behavior. Could we merge this in and start using it? |
Yes, I was going to add the options requested by @araichoor and then merge. Working on it. |
@tskisner : have you had a chance to look at this STANDARD/CALIB minor thing? |
Recall that the old behavior for leftover fibers (i.e. unassigned fibers after 10 standards / 40 sky per petal) was:
Now we have changed step (1) for these leftover fibers to be allowed to observe any science target regardless of numobs_more. So I would expect fewer "extra" observations of standards (we should always have 10 per petal though). |
…rsubscribing done targets with leftover fibers.
Ok, merging this now that checks have passed. Next up is timestamps... |
ok great then, that explains why we get slightly fewer STANDARD fibers. with the master branch (col1 = passid , col2-11: nstd per petal): with the obs_zero branch (col1 = passid , col2-11: nstd per petal): |
Ok, thanks for that investigation. Working now on the UTC fixes in desimodel, and then will make a corresponding PR for fiberassign. |
awesome, thanks a lot! |
After assigning science targets which have remaining observations, and then ensuring that the per-petal quota of standards and sky are met, there may be some unassigned fibers. These unassigned fibers are then assigned to science, standards, and sky if possible.
In the current main branch, science targets with less than one remaining obs are never considered for assignment, even in this "last ditch" effort to assign unused fibers.
This PR changes that behavior to allow assignment to science targets with less than one obs remaining during this final attempt to assign the leftover positioners.