[Clp] New Pivot selection rule in Clp
John Forrest
john.forrest at fastercoin.com
Thu Jun 25 16:23:14 EDT 2015
At the last INFORMS I learnt about a very interesting idea on pivot
selection for degenerate problems. Using donated code, this is now in
Clp trunk. The idea is to do a bit more work to identify which
variables are likely to lead to a degenerate move and avoid them.
I would be grateful if you can give me feedback - or the authors who so
generously contributed the code which I tinkered with.
The easiest way to use the stand-alone version of Clp to test if this
helps your problem is to add
-psi 0.5 (for example) - 1.0 has no effect , 0.1 has much more.
If you do
clp
psi??
you will get a bit more information.
The references are -
Towhidi, M., Desrosiers, J., Soumis, F., 2014. The positive edge
criterion within COIN-OR’s CLP. Computers & Operations Research 49, 41–46.
Omer, J., Towhidi, M., Soumis, F., 2015. The positive edge pricing rule
for the dual simplex. Computers & Operations Research,
http://dx.doi.org/10.1016/j.cor.2015.03.009
Enjoy
John Forrest
More information about the Clp
mailing list