You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We have seen a use-case where we have agreed it is necessary to store information per target as the experiment continues in such a way that we can access it quickly using the target ID only.
There are several ways of doing this:
Store per-target data in a new butler collection (we want to avoid storing features here since these could be to big for Mongo, so this would at minimum have to be some separate collection storing auxiliary info about targets)
Store target data in algorithms dictionary (the use-case considered requires this information to be accessible across algorithms, so this is not right)
Store target data in the experiment dictionary (It is possible in this case that the target set is very large and will not fit in one dictionary--we want to have one record for each target--not a big record for all of them)
The only place where the targets are currently treated on a "one target per record" basis is the target manager, and so it seems fitting to add updates here. (If there is concern about misunderstanding/misuse, we can restrict where in the target dictionary updates can be made.)
The text was updated successfully, but these errors were encountered:
We have seen a use-case where we have agreed it is necessary to store information per target as the experiment continues in such a way that we can access it quickly using the target ID only.
There are several ways of doing this:
The only place where the targets are currently treated on a "one target per record" basis is the target manager, and so it seems fitting to add updates here. (If there is concern about misunderstanding/misuse, we can restrict where in the target dictionary updates can be made.)
The text was updated successfully, but these errors were encountered: