biocmask: data masks for S4 objects, especially SummarizedExperiment

[mikelove]

No description provided.

[mikelove]

Notes from meeting:

• User should be able to load both packages (biocmask, tidySummarizedExperiment), potentially tidySE could use biocmask as a backend. Software can have logic about when to use what method.
• There are some tasks that tidySE can do (returning a tibble) and some that biocmask can do (e.g. grouping and returning a smaller SE), need to enumerate these.
• We should have a smooth transition in the documentation between the different options.
• We should move toward a third package for messaging and printing, across all tidyomics. Modular, lightweight, e.g. when an S4 object is not possible to be returned.
• For single cell: only printing sample-wise information, this was an early design choice. So far it was pivot_samples, pivot_features. See e.g. anchor_5p() in plyranges, which just specifies a future preference of behavior: Anchoring modifies how arithmetic is performed ... To remove anchoring use unanchor()