[Octave-bug-tracker] [bug #31239] isequalwithnans(1, 2, 1) gives true instead of false (patch provided)

[anonymous]

URL:
  <http://savannah.gnu.org/bugs/?31239>

                 Summary: isequalwithnans(1,2,1) gives true instead of false
(patch provided)
                 Project: GNU Octave
            Submitted by: None
            Submitted on: Mon Oct  4 11:07:39 2010
                Category: Libraries
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: Incorrect Result
                  Status: None
             Assigned to: None
         Originator Name: Iain Murray
        Originator Email: address@hidden
             Open/Closed: Open
         Discussion Lock: Any
                 Release: dev
        Operating System: GNU/Linux

    _______________________________________________________

Details:

When checking several arrays isequalwithnans currently only checks that the
last array matches the first. For example:

isequalwithequalnans(1,2,1)

returns true but shouldn't because 1!=2.

The cause is a bug in __isequal__ -- when the first argument to __isequal__
is 1 (to allow nans to be present) the function no longer returns as soon as
it notices an inequality not involving nans. As long as the last array to be
checked matches, the function forgets about previous inequalities.

I have attached a patch against the version of __isequal__.m I got by
checking out the Hg repository just now. I've attached the full version of the
patched function too.



    _______________________________________________________

File Attachments:


-------------------------------------------------------
Date: Mon Oct  4 11:07:39 2010  Name: __isequal__.m.PATCH  Size: 443B   By:
None

<http://savannah.gnu.org/bugs/download.php?file_id=21606>
-------------------------------------------------------
Date: Mon Oct  4 11:07:39 2010  Name: __isequal__.m  Size: 5kB   By: None

<http://savannah.gnu.org/bugs/download.php?file_id=21607>

    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?31239>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/