It's actually the other way around: rank-0 is atoms or
columns of matrix, and rank-1 is vectors or rows of matrix.
The reason you might have thought of rank-1 as columns is
because it makes verb seem to operate on cells collectively aligned
into columns, as in {."1 A will return first column, whereas
it actually returns list of first cell in each row--so it
applies the verb to rows.
This confusion may have resulted from APL axes.
See J Forums for example,
http://www.jsoftware.com/pipermail/programming/2006-June/002509.html
--- On Tue, 7/1/08, Alex Rufon <[EMAIL PROTECTED]> wrote:
most likely I misunderstand it. I only use rank 1 when I try to apply a
verb on the column level of a two dimension matrix. I assume (from what
I've read so far) that rank zero is the rows of a two dimension matrix
and is assumed that its there and is omitted.
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
