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
In cases where merging is not required (e.g. if the files are so large that it would be 1-to-1 merging, of if it's simply not desired), it should be possible to skip it.
This is mean that subsequent tasks (UniteColumns, ProduceColumns, WriteHistograms, PrepareMLEvents, MLEvaluation) should have the option to directly retrieve their inputs from ReduceEvents.
This could be solved by having in intermediate wrapper-like task that provides the input with a clean interface, so that switches between different requires()/input() structures only exist once.
In cases where merging is not required (e.g. if the files are so large that it would be 1-to-1 merging, of if it's simply not desired), it should be possible to skip it.
This is mean that subsequent tasks (UniteColumns, ProduceColumns, WriteHistograms, PrepareMLEvents, MLEvaluation) should have the option to directly retrieve their inputs from ReduceEvents.
This could be solved by having in intermediate wrapper-like task that provides the input with a clean interface, so that switches between different
requires()
/input()
structures only exist once.Also see #383.
The text was updated successfully, but these errors were encountered: