I think the MatLab excluded the time for the random number generation.
Look closely at the posted code.  Also, the J code "jumbo" did 
include the 1000 repetitions of the matrix multiply.



----- Original Message -----
From: Randy MacDonald <[EMAIL PROTECTED]>
Date: Thursday, November 8, 2007 12:00
Subject: Re: [Jprogramming] speed of matrix*vector operation.
To: Programming forum <[email protected]>

> Let's be fair; I don't think the Matlab guy excluded the random number 
> generation.  The 1000 (or was it 1001?) repetitions of the matrix 
> multiply should also be in the J expression, something I don't think 
> I've seen.  As far as I can see, the execution speedup of Matlab doesn't 
> cover the extra time it takes to type in the Matlab code, assuming there 
> are no typos in that long piece of Matlab.
> 
> Roger Hui wrote:
> > You should exclude the time needed to generate the random numbers.
> >
> >
> >
> > ----- Original Message -----
> > From: Devon McCormick <[EMAIL PROTECTED]>
> > Date: Thursday, November 8, 2007 10:30
> > Subject: Re: [Jprogramming] speed of matrix*vector operation.
> > To: Programming forum <[email protected]>
> >
> >   
> >>     6!:2 '(?256$0)*"1]?312 256$0'
> >> 0.121752
> >>
> >> On 11/8/07, Michael Dykman <[EMAIL PROTECTED]> wrote:
> >>     
> >>> On Nov 8, 2007 1:13 PM,  <[EMAIL PROTECTED]> wrote:
> >>>       
> >>>> I told my friend about how nice J was.  Am I wrong?
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to