Make a scalar center argument for matrix functions defunct

[HenrikBengtsson]

We have had the following since matrixStats 0.58.0 (2021-01-26):

• Using a scalar value for argument center of colSds(),
  rowSds(), colVars(), rowVars(), colMads(), rowMads(),
  colWeightedMads(), and rowWeightedMads() is now deprecated.

We should move to make it defunct by default.

I've started revdep checks with R_MATRIXSTATS_CENTER_ONSCALAR=defunct to see if there are any negative side effects.

[HenrikBengtsson]

@PeteHaitch, passing a scalar to argument center will soon be defunct. I noticed that you explicitly allow for scalars in https://github.com/PeteHaitch/DelayedMatrixStats/blob/20c2e278d9ccf503a49dd489dc3597a21443c01d/R/rowVars.R#L16-L25.

[HenrikBengtsson]

@const-ae, FYI, passing a scalar to argument center will soon be defunct; currently deprecated. I noticed that sparseMatrixStats accepts scalar center values, e.g.

> sparseMatrixStats::colVars(sparse_mat, center = 4)
[1] 17.7777778  0.0000000  0.4444444  1.8888889  1.0000000  0.0000000

[const-ae]

Thanks for the heads up. I made scalar 'center' arguments defunct in the devel version of sparseMatrixStats. This was, anyways, not supported.

[PeteHaitch]

Thanks for the heads up. I'll take care of it before the upcoming BioC release.

[HenrikBengtsson]

Thank you. I've checked that there are no negative revdep side effects, so this will now be the default in the next release.

[PeteHaitch]

Is the new release imminent?
More exactly, is it likely to be before or after the upcoming BioC release?

[HenrikBengtsson]

I'm thinking before ... Would it matter, given that I don't find any revdep packages being affected?

[HenrikBengtsson]

... and in worst case scenario (for current Bioc release and likes), it can for now be reverted using:

R_MATRIXSTATS_CENTER_ONSCALAR=deprecated

or in R, by:

options(matrixStats.center.onScalar = "deprecated")

This will be documented in ?matrixStats::matrixStats.options.

[PeteHaitch]

Thanks for the extra info.
I don't expect it to matter but it's helpful to me in prioritising tasks as the BioC release approaches.

[PeteHaitch]

Made the change for current devel (BioC 3.19) but not touched current release (BioC 3.18).