QUESTION: Why different arguments for colVars and colSds?

[PeteHaitch]

> args(matrixStats::colVars)
function (x, rows = NULL, cols = NULL, na.rm = FALSE, center = NULL, 
    dim. = dim(x), ...) 
NULL
> args(matrixStats::colSds)
function (x, rows = NULL, cols = NULL, ...) 
NULL

I understand that na.rm, center, and dim.. can be passed to colSds() via ..., but wondering why they don't appear as formal arguments? Same for rowVars()/rowSds().

Thanks

[HenrikBengtsson]

This must have been because I wanted to keep the redundancy down so any tweaks of arguments could be done in a single place and thereby simplify documentation and lower the risk for mistakes.

Since args are settled by now, maybe it doesn't hurt to make them explicit everywhere - it would make it easy for the user and print(), args() would be more informative. Do you know of other cases in the package?

[PeteHaitch]

I suspected that was the reason.

No, I didn't find any other cases. (colRanks() has preserveShape whereas rowRanks() does not, but my reading of the docs is that this is for good reason)

[HenrikBengtsson]

Fixed.