On Thu, Sep 4, 2008 at 1:33 AM, david alis <[EMAIL PROTECTED]> wrote:
> Something doesn't seem quite right with the last example.
>
>   A =: 3 : 0
> 'A-y' (1!:2)2
> 'A-y';<y
> :
> 'x-A-y'(1!:2)2
> x;'x-A-y';<y
> )
>
>   B =: 3 : 0
> 'B-y' (1!:2)2
> 'B-y';<y
> :
> 'x-B-y'(1!:2)2
> x;'x-B-y';<y
> )
>
>   C =: (A :. B) ">    NB. C has the same rank as >
...
>   100 (C^:_1) 200
> |domain error
> |   100    (C^:_1)200

According to http://www.jsoftware.com/help/dictionary/d202n.htm,
100 (C^:_1) 200 should be equivalent to (100 & C^:_1) 200

This seems to be true, in that both produce domain errors.

I would guess that the problem is that the obverse mechanism
does not provide an inverse for the monad which gets formed when
the pre-obverse dyad gets bonded by a left argument.

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

Reply via email to