Bill Lam's analysis was correct. For a dyad,
once you have figured out the frame and cells,
it remains to apply something to corresponding
cells, and with respect to that dyad that
something is a "black box".
You have figured out that the cells are
1 0 1 0 0 0 and 3 1 4 1 5 9. (Since the left
rank is 1 and the right rank is _ , that makes
a lot of sense.) Now you apply something
(the kernel of the verb #) to these cells, and
in that application you have to depend on the
definition of # to know what to do. Consider
the following examples:
x=: 1 0 1 0 0 0
y=: 3 1 4 1 5 9
kernel=: #
x kernel"1 _ y
kernel=: ,
x kernel"1 _ y
kernel=: +
x kernel"1 _ y
kernel=: +//.
x kernel"1 _ y
In each case, because the verb ranks are the same
(1 _) and the arguments are the same, the analysis
in your "worksheet" would be the same, resulting in
the same frame and cells, but the application and
result of "kernel" would be different in each case.
----- Original Message -----
From: Terrence Brannon <[EMAIL PROTECTED]>
Date: Thursday, April 19, 2007 5:21 am
Subject: Re: [Jgeneral] How are isolated pair-wise actions accumulated into a
return value
> On 4/18/07, bill lam <[EMAIL PROTECTED]> wrote:
>
>
> > >
> > > The frames are the same length, so we just go pair-wise
> > >
> > > 1 # 3
> > > 0 # 1
> > > 1 # 4
> > > 0 # 1
> > > 0 # 5
> > > 0 # 9
> >
> > I think that this part of analysis is not needed,
>
> why do you think that? I'm just going step by step and leaving out
> nothing.
> >when the verb # got arguments
> > 1 0 1 0 0 0 and 3 1 4 1 5 9, how it come up with an answer is
> not related to
> > frame/cell/agreement.
>
> What substantiates this assertion? I think frame-cell agreement does
> relate and what happens when the left-frame and right frame are the
> same is the case that applies.
>
> > Just think it as an black box and gives 3 4 as the result.
>
> No thanks. Black boxes are great for systems processes like
> psychology, but here we need nuts and bolts.
>
> >
> > If # is rank-0 then it needs further analysis
>
> # b. 0
> _ 1 _
>
> Could you tell me why you are bringing up rank 0? The verb rank is
> shown above.
>
> > 1 0 1 0 0 0 #("0) 3 1 4 1 5 9
> > 3
> > 0
> > 4
> > 0
> > 0
> > 0
> >
> > 0#n will give null, but on assembling the final result fill-
> element is used.
>
> I learned about fills here:
> http://www.jsoftware.com/help/jforc/loopless_code_i_verbs_have_r.htm#_Toc141157989
>
> I cant quite figure out how to apply what you say about the
> fill-element to the final result of the expression 1 0 1 0 0 0 # 3
> 1 4
> 1 5 9 NB. the version without force rank
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
