x=: i.10 2 3
   y=: i.4 5 2 4
   x -. y
|length error
|   x    -.y

After the change, x -. y would be just x itself.

The dictionary defines the dyad -. thus:
   x-.y includes all items of x except for
   those that are cells of y .
The dyad i. is used to determine "except for
those that are cells of y".



----- Original Message -----
From: bill lam <[EMAIL PROTECTED]>
Date: Tuesday, May 15, 2007 5:37 pm
Subject: Re: [Jgeneral] length error in i.

> In what way will be the proposed change affect dyad -. ?
> 
> Roger Hui wrote:
> > Currently,  x i. y  signals length error when the
> > rank r cells of y do not have the same shape as the 
> > the rank r cells of x, where  r=:0>.(#$x)-1  is the
> > rank of an item of x .  For example:
> > 
> >    x=: i.10 2 3
> >    y=: i.4 5 2 4
> >    x i. y
> > |length error
> > |   x     i.y
> > 
> > I propose to change this to return  #x  for each
> > rank r cell.  Thus:
> > 
> >    x i. y
> > 10 10 10 10 10
> > 10 10 10 10 10
> > 10 10 10 10 10
> > 10 10 10 10 10
> > 
> > One jusification for this change is that  x i. y
> > already returns  #x  for cells having a wrong type,
> > a situation no less egregious than having a wrong 
> > shape.
> > 
> >    x i. 4 2 3$'deipnosophist'
> > 10 10 10 10
> > 
> > i: e. -.  will also be changed correspondingly.
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to