--- Terrence Brannon <[EMAIL PROTECTED]> wrote:
> Oleg Kobchenko <[EMAIL PROTECTED]> writes:
>
> >
> > The confusion in B is based on incorrect use of word "items"
> > which is reserved for _1 cells.
> >
> > I think simply using different word like "elements" would fix it.
>
> yes
>
> >
> > I am not sure innermost dimension is good wording.
>
> Yes, I think J uses the term "axis" for what others call dimension. So unless
> your qualm with with innermost (and I'm betting it is), then we just need to
> say innermost axis.
I don't like the idea of re-inventing definitions and
introducing new terms.
Terms like rank, shape, frame and cell are very well
defined in the Dictionary. Once understood they should
be sufficient to describe your scenarios to avoid
prolification of confusion.
The parts of a noun when decomposed by a frame
are called cells. In general, a cell has multiple
"axes" not just one innermost.
I also see a problem in "thinking" that k-cell is
an "element" of an array, because in J array elements
are atoms.
It is a cell, so it need be thought of as cell.
If you are looking for a truthful and simple analogy,
the decomposition can be modeled as boxing with rank k,
replacing contents of each box with result of verb,
and unboxing the result.
A side question: in your model, how do you replace
a cell in an array, if you are to model rank application
in J? Especially, if the result is different rank/shape
than the argument cell.
(i.3 4)0}i.3 4
|rank error
| (i.3 4) 0}i.3 4
(i.5)0}i.3 4
|length error
| (i.5) 0}i.3 4
____________________________________________________________________________________
Fussy? Opinionated? Impossible to please? Perfect. Join Yahoo!'s user panel
and lay it on us. http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
