From octave-maintainers-request at bevo dot che dot wisc dot edu  Thu Aug  3 20:41:07 2000
Subject: Re: bug? in create_set
From: pkienzle at kienzle dot powernet dot co dot uk (Paul Kienzle)
To: octave-maintainers at bevo dot che dot wisc dot edu, ripper@cs.sandia.gov
Date: Mon, 31 Jul 2000 19:09:51 +0100 (BST)

In unique(), I use:

  if (nelx > 1)
    y (find (y(1:nelx-1) == y(2:nelx))) = [];
  endif

Paul Kienzle
pkienzle at kienzle dot powernet dot co dot uk

From: "Ross A. Lippert" <ripper at cs dot sandia dot gov>
>> create_set([1 1])
>issues an "empty matrix found" warning.  This is in 2.1.31, but
>not in 2.0.14 which used a different implementation of create_set.
>
>The source of the problem is the line
>
>y = y( [1, find (y(1:nelx-1) != y(2:nelx)) +1]);
>
>I'm wondering if that could be replaced with
>
>y = y ( find( y != [NaN y(2:nelx)] ) );
>
>or would that be too much of a hack?
>
>
>-r
>
>
>