It is an anomaly only if you can produce _. on arguments 
not containing _. .

>From  http://www.jsoftware.com/help/dictionary/d031.htm :

... should be removed from such data as soon as possible.
... Other primitives on arguments containing _. may not 
produce correct or consistent results.

In short, when an argument contains _. , all bets are off.



----- Original Message -----
From: "Mark D. Niemiec" <[EMAIL PROTECTED]>
Date: Wednesday, April 9, 2008 17:30
Subject: [Jbeta] Some NaN anomalies in 6.02
To: [email protected]

> I noticed the following in J 6.02:
> 
>    0 +. _   NB. Should probably be _ ((0 +. 
> y) = (|y) for all real y)
> |NaN error
> |   0    +._
>    0 +. _.  NB. Correct 
> _.
>    1 +. _.  NB. Oops!
> <infinite loop>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to