From octave-sources-request at bevo dot che dot wisc dot edu  Fri Jul 28 23:49:47 2000
Subject: Re: matlab-like functions
From: Edward Jason Riedy <ejr at EECS dot Berkeley dot EDU>
To: pkienzle at kienzle dot powernet dot co dot uk
cc: octave-sources at bevo dot che dot wisc dot edu, tpikonen@pcu.helsinki.fi
Date: Fri, 28 Jul 2000 21:49:41 -0700

And Paul Kienzle writes:
 - Slightly worse I would say.  Mod(x,0) is NaN even if matlab says
 - otherwise.

Saw this while checking the archives for a bug in x{max,min} (patch
on the way after testing)...

By IEEE, rem(x, 0) is NaN.  Matlab agrees, although only giving it
the weight of `convention.'  Matlab's `convention' for mod(x, 0)
is 0.  

The convention comes from integer arithmetic, I believe.  Consider
the definition of the mod n operator as a map from the integers to 
Z_n = { 0, 1, ... , n-1 }.  Then it makes complete sense for x mod 0
to be 0, as Z_0 = { 0 }.  I can't cite a source for this at the
moment, but it seems sensible.  I think Matlab is trying to give 
integer people more-or-less what they expect, at least until the
integers become too large.

Perhaps someone with a copy of Concrete Mathematics could see if the
topic is discussed.  Graham, et al. like to confront these issues.

Jason



-----------------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.che.wisc.edu/octave/octave.html
How to fund new projects:  http://www.che.wisc.edu/octave/funding.html
Subscription information:  http://www.che.wisc.edu/octave/archive.html
-----------------------------------------------------------------------