[Osi] issue in OsiCbc - corrupting bounds

[Matthew Galati]

I was using release 0.102.3.



On Mon, Aug 9, 2010 at 11:15 AM, Matthew Galati <Matthew.Galati at sas.com>wrote:

> Using OsiCbc.
>
> I have a 2 variable example where x[0] in [0,6] and x[1] in [0,6].
>
> Checking m_osi.getColLower and getColUpper, everything looks ok at the
> start.
>
> I do:
>  m_osi.setObjective(xxx);
>  m_osi.branchAndBound()
>
> I then check, m_osi.getColLower and getColUpper and the bounds have
> changed. This breaks my 2nd time thru.
>
> Presumably, they are pointing to the internal bounds which may be changing
> due to reduced cost fixing or something else. This does not seem like the
> correct behavior. This is not how OsiCpx works. The bounds are kept intact
> after a call to branchAndBound.
>
> Thanks,
> Matt
>
>
> _______________________________________________
> Osi mailing list
> Osi at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/osi
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/osi/attachments/20100809/3975ec08/attachment.html>