[Dip] Differences between GAP examples

Matthew Galati matthew.galati at gmail.com
Tue Jun 11 16:34:30 EDT 2019


The Makefile gives some hints... one case using the built-in MILP solver,
one case uses a KP solver.


#PC: oracle for KP solve is Pisinger, BranchEnforceInMaster
OBJS += GAP_DecompApp. at OBJEXT@
CFLAGS = -DVERSION1
#TODO: BranchEnforceInSubProb using GAP_DecompApp2 or by option
# CPM/PC: built-in MILP solver, dense model build
# BranchEnforceInMaster or BranchEnforceInSubProb by option
#OBJS += GAP_DecompApp3. at OBJEXT@
#CFLAGS += -DVERSION3
# CPM/PC: built-in MILP solver, sparse model build
# BranchEnforceInMaster or BranchEnforceInSubProb by option
#OBJS += GAP_DecompApp4. at OBJEXT@
#CFLAGS += -DVERSION4

On Tue, Jun 11, 2019 at 3:41 PM Florian Fontan <dev at florian-fontan.fr>
wrote:

> Dear Dip community,
>
> I am trying to get familiar with Dip and understand how it works. Could
> someone tell me the differences between the multiple DecompApp
> implementation examples for the Generalized Assignment Problem?
> https://github.com/coin-or/Dip/tree/master/Dip/examples/GAP
>
> For example, in GAP_DecompApp.cpp, I noticed that the setModelRelax call
> passes NULL as first argument:
>
>      setModelRelax(NULL, modelName, i);
>
> whereas in GAP_DecompApp3.cpp, an other model is defined:
>
>      status = createModelPartKP(modelRelax, i);
>      modelName = "KP" + UtilIntToStr(i);
>      setModelRelax(modelRelax, modelName, i);
>
> What does that mean? I haven't found any documentation about it
>
> Florian
> _______________________________________________
> Dip mailing list
> Dip at list.coin-or.org
> https://list.coin-or.org/mailman/listinfo/dip
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/dip/attachments/20190611/ee00fcf2/attachment.html>


More information about the Dip mailing list