Uncollect#
Uncollect files#
Given a file with many different slots, create multipe different files with a dedicated subset of slots. This is particularly useful when you want to undo the collect module and compute downstream task on different slots (layers, obsm, etc) separately.
Parameters#
DATASETS:
test:
input:
uncollect: test/input/
uncollect:
new_file_ids:
- file_1
- file_2
- file_3
sep: --
sep: separator used for matchingnew_file_idin the slots. Note, the separator should be unique to the slots you want to match, to avoid unintended side effects. Default: βββnew_file_ids: list of new file ids to create. If the file name (new_file_id) uniquely matches a slot name, that slot will be included in the new file. For slots that donβt contain the unique separator, they will be included in all new files. Only in cases where the separator is in the slot name, but the slot does not match thenew_file_id, the slot will not be included.