On Friday, January 04, 2013 9:10 AM Linda Alvord wrote:

...

>    f=: 13 :'([:<"1#"0)">y'
> 
>    3 4 5 f i.3
> 
> ------T-------T---------?
> ?0 0 0?1 1 1 1?2 2 2 2 2?
> L-----+-------+----------

Yes, that's done it! So verb trains, performed with the appropriate rank are
just as good as composition. Thank you!

So the very simplest implementations of my earlier example are:

   3 4 5 <@#"0 i.3
+-----+-------+---------+
|0 0 0|1 1 1 1|2 2 2 2 2|
+-----+-------+---------+
   3 4 5 ([:<#)"0 i.3
+-----+-------+---------+
|0 0 0|1 1 1 1|2 2 2 2 2|
+-----+-------+---------+

Doing everything rank zero gets the boxing done before any artificial
rectangular array constraints appear.

On Friday, January 04, 2013 9:35 AM Linda Alvord wrote:

> I'm no longer puzzled

Great! The subject could be changed to 'atop now ceases to puzzle me'. And I
have been helped by seeing that verb composition and verb trains have such
similar attributes, even when it comes to ragged arrays.

Regards

Graham







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

Reply via email to