And to quote KEI: "One man's anomaly is another man's feature"

/Pablo Landherr

On 10/22/06, Chris Burke <[EMAIL PROTECTED]> wrote:

Pascal Jasmin wrote:
> a useful counterexample where 0 n $ y or i. 0 n produces a result that
is conveniently different than i.0 0 would substantially contribute to the
impression that the costs of eliminating that behaviour are
high.  Especially, if there is no natural way of rewritting that
counterexample.

You can use i. 0,n to initialise a table that is built up by appending
rows, where you need the final result to have n columns. This gives a
different result from i.0 0. For example:

   $ (i.0 0),1 2 3
1 3
   $ (i.0 5),1 2 3
1 5

In any case, right now the treatment of rank and shape of nouns is
consistent. You are asking for special treatment where values are zero.
This will not happen, since it complicates the language considerably.

The R language is an array language which does not have consistent
treatment of rank and shape, and this gives more trouble than it is
worth. There is an amusing note in the R faq about such an issue, which
ended "... after much discussion this (behaviour) has been determined to
be a feature."

----------------------------------------------------------------------
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