Closed Michael-E-Rose closed 1 month ago
The requirement, that candidates have to publish in two periods is very effective in reducing the initial search group - even though the active_year requirement appears somewhat arbitrary and hard to control.
Instead, I propose this solution: we introduce about equal sized buckets of volumes, and candidates have to publish in all of them. The user gets to set the number of buckets (or their size). So, like it is the case right now, we look at multiple periods of the search sources, and the intersection of the authors of each of them constitutes the search group.
In
search_group_from_sources()
, we require the search group to publish in the same journals in two points in time:The second requirement doesn't make sense and we should drop it.