Yes that is correct; u"n M always divides up into cells formed by grouping the 
last n dimensions together. Another way to see it is that elements of the array 
M who have the same (#$M) - n first indices are grouped together.

A variant on u"n M is "(-n), which is roughly equivalent to u"((#$M)-n) M.

If you need to break M up according to axes which are not on the tail end, you 
can look into |: transpose.

Cheers,
Louis

> On 28 Feb 2018, at 07:18, Nick S <[email protected]> wrote:
> 
> Someone please tell me if I understand this:
> 
> If I have a list of 5x5 tables, that is, a thing with a shape of n 5 5
> where n is anything from 1 to 30, That is, the shape of the table might be
> (5 5 or 2 5 5 or 3 5 5 etc.) and I want to compare it against a single 5x5
> thing in 5x5 groups, the proper rank to use for the comparison is 2, as
> 
> a ="2 b or a -:"2 b or a *."2 b
> 
> depending on what I want to do.
> 
> When I use a rank 2, does it take the last two dimensions of whatever and
> use those?  The examples I see are all for the more common cases of taking
> a table and processing it by individual items, rows or columns.  My
> experimentation seems to indicate that my guess is true, but I am so damned
> confused at this point, I need to see if I am on the right track.
> 
> -- 
> Of course I can ride in the carpool lane, officer.  Jesus is my constant
> companion.
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to