orenbenkiki
commented
1 year ago Ugh, a documentation bug. Wasn’t fixed when we separated the concerns. The vignette documentation is correct though. I’ll fix it.
On Tue, 4 Jul 2023 at 11:14 Oren Milman @.***> wrote:
the current README.rst says: "Noisy genes are not only forbidden from being selected for computing cells similarity, but are also [...]"
But currently, as far as I can tell, marking a gene as noisy still allows it to be selected. So the README.rst should be rephrased, but in my opinion the intuitively expected behavior is actually as said in the README. i.e., I would expect you to automatically not allow selecting noisy genes (maybe have an option to explicitly allow a gene to be noisy but still be selected, but not as the default).
If you decide to keep the current behavior, then because it is pretty unintuitive (at least for me), I think the vignettes should also explicitly say something like: "so you probably want to specify noisy genes also as lateral, otherwise they might be selected for computing similarity"
— Reply to this email directly, view it on GitHub https://github.com/tanaylab/metacells/issues/49, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAQUXTOCIHXBD56A2MI3M3XOPGHJANCNFSM6AAAAAAZ5L2RMY . You are receiving this because you are subscribed to this thread.Message ID: @.***>
the current README.rst says: "Noisy genes are not only forbidden from being selected for computing cells similarity, but are also [...]"
But currently, as far as I can tell, marking a gene as noisy still allows it to be selected. So the README.rst should be rephrased, but in my opinion the intuitively expected behavior is actually as said in the README. i.e., I would expect you to automatically not allow selecting noisy genes (maybe have an option to explicitly allow a gene to be noisy but still be selected, but not as the default).
If you decide to keep the current behavior, then because it is pretty unintuitive (at least for me), I think the vignettes should also explicitly say something like: "so you probably want to specify noisy genes also as lateral, otherwise they might be selected for computing similarity"