No reduction functions should be column-wise by default (update code and docs)

[Nosferican]

There still outdated documentation and functions that still have a column-wise by default behavior. These should be updated to taking the dims argument and the documentation should reflect the change.

[Nosferican]

Some cases identified,

• https://docs.julialang.org/en/latest/manual/noteworthy-differences/

In Julia, functions such as sort that operate column-wise by default
(sort(A) is equivalent to sort(A,1)) do not have special behavior for 1xN
arrays; the argument is returned unmodified since it still performs sort(A,1).
To sort a 1xN matrix like a vector, use sort(A,2).

[JeffBezanson]

sort does require the dims argument now. I think this is just a doc issue?

[Nosferican]

If all functions are have been updated, yes. Is there any reason for why sort API is different from other functions,

using LinearAlgebra
A = rand(3,3)
maximum(A) == maximum(A[:,:])
norm(A) == norm(A[:,:])
mean(A) == mean(A[:,:])

while sort

sort(A)
ERROR: UndefKeywordError: keyword argument dims not assigned

shouldn't it be,

sort(A) == sort(A[:,:])

for consistency?

[JeffBezanson]

Those functions are reductions that have always operated over the whole array, unlike sort.