[Coin-lpsolver] OsiClp/ClpSimplex/clp performance disparity
John J Forrest
jjforre at us.ibm.com
Tue May 2 16:23:56 EDT 2006
Ojas,
If a feature is only available in one OsixxxSolverInterface (and especially
if it is of doubtful heritage) then I can not see the point of doing any
work which duplicates what is available by using getModelPtr() and also
encourages people to use the interface in untested ways. For example I
regret giving in to some request and writing OsiCbcSolverInterface as it
has caused problems, both with builds and with expectations.
John Forrest
Ojas Parekh
<ojas at mathcs.emor
y.edu> To
Sent by: John J Forrest/Watson/IBM at IBMUS
coin-lpsolver-bou cc
nces at list.coin-or coin-lpsolver at list.coin-or.org
.org Subject
Re: [Coin-lpsolver]
OsiClp/ClpSimplex/clp performance
05/02/06 03:09 PM disparity
John, would you mind telling us a little more about this crash and why it
may not be a good idea to use it from Osi land?
-Ojas
The reason qap10 is faster is because the non Osi interface allows
itself to use a very dubious crash which works fantastically well on
many problems. I am not sure about encouraging people to use it
directly from Osi land.
_______________________________________________
Coin-lpsolver mailing list
Coin-lpsolver at list.coin-or.org
http://list.coin-or.org/mailman/listinfo/coin-lpsolver
More information about the Clp
mailing list