docs: add column_stack to ndarray-for-numpy-users guide

[chiliec]

Closes #650.

What

Adds a np.column_stack row to the "Array manipulation" table in the
ndarray_for_numpy_users guide, next to the existing concatenate/stack rows.

Why

column_stack is a common NumPy call and the guide had no direct entry for it,
so users had to work out the stack(Axis(1), …) idiom themselves (see issue).

Idiom

For 1-D arrays, np.column_stack((a, b)) builds a 2-D array with a and b
as columns — that's exactly stack(Axis(1), &[a.view(), b.view()]) (or the
stack![Axis(1), a, b] macro), matching the table's existing style.

Verification

Compiled both forms against this checkout:

let a = array![1., 2., 3.];
let b = array![4., 5., 6.];
stack(Axis(1), &[a.view(), b.view()]).unwrap()  // => [[1,4],[2,5],[3,6]]
stack![Axis(1), a, b]                            // => [[1,4],[2,5],[3,6]]

Both produce the numpy column_stack result. Docs build clean:

• RUSTDOCFLAGS="-D warnings" cargo doc --no-deps — exit 0 (intra-doc links resolve)
• cargo test --doc ndarray_for_numpy_users — ok

Happy to adjust the wording.