Further evidence for Raul's last point:

   f0 =: ;@:(<@i."0)
   f0a=: ; @: (i.&.>)
   f1 =: (#&1"0 #&, i."0)
   f2 =: i.@(+/) - ] # +/[EMAIL PROTECTED]:@(0&,)

   x=: 1e4 [EMAIL PROTECTED] 200
   (f0 -: f0a) x
1
   (f0 -: f1) x
1
   (f0 -: f2) x
1
   ts=: 6!:2 , 7!:[EMAIL PROTECTED]

   ts 'f0 x'
0.0474203 1.06316e7
   ts 'f0a x'
0.0341342 1.13781e7
   ts 'f1 x'
0.139537 2.51666e7
   ts 'f2 x'
0.0691909 1.26492e7



----- Original Message -----
From: Roger Hui <[EMAIL PROTECTED]>
Date: Thursday, November 9, 2006 11:13 am
Subject: Re: RE: [Jprogramming] How to avoid fills without boxing?

>   (i.+/x) - x#+/\}:0,x=: i.5
> 0 0 1 0 1 2 0 1 2 3
>   (i.+/x) - x#+/\}:0,x=: 3 1 4 1 5 9
> 0 1 2 0 0 1 2 3 0 0 1 2 3 4 0 1 2 3 4 5 6 7 8
> 
> 
> 
> ----- Original Message -----
> From: "Miller, Raul D" <[EMAIL PROTECTED]>
> Date: Thursday, November 9, 2006 11:09 am
> Subject: RE: [Jprogramming] How to avoid fills without boxing?
> 
> > R.E. Boss wrote:
> > > How do I realize without boxing (performance reasons)
> > >    ;@:(<@i."0) i.5
> > > 0 0 1 0 1 2 0 1 2 3
> > 
> > Alternate formulations, which do not use boxing, are
> > possible.  For example:
> >   (#&1"0 #&, i."0) i.5
> > 0 0 1 0 1 2 0 1 2 3
> > 
> > However, "without boxing" need not mean "better
> > performance".  (And does not, in this case.)


----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to