[gecode-users] Re: Waiting for stability

[Christian Schulte]

Please, don't use LinExpr as this is really something that is only to be
used for modelling. A much simpler interface is to first create arrays of
integer ceofficients and variables and then post a linear constraint on
them.

Christian

-----Original Message-----
From: users-bounces at gecode.org [mailto:users-bounces at gecode.org] On Behalf
Of Luis Quesada
Sent: Thursday, June 15, 2006 3:02 PM
To: users at gecode.org
Subject: [gecode-users] Re: Waiting for stability


Dear all,

I am trying to use LinExpr in order to implement a (naive) order 
propagator for graph variables.
Is the following use of LinExpr correct?

template <class SPACE>
LinExpr
graph_order(SPACE* space, OutAdjSetsGraphView g){
    OutAdjSetsGraphView::LubNodeIterator g_ub = g.iter_nodes_UB();
    LinExpr Sum(IntVar(space,0,0));
    for (;g_ub();++g_ub){
      int n=g_ub.val();
      BoolVar a(space,0,1);
      dom(space, g.nodes, SRT_SUP, n, a);
      LinExpr BoolVarLinExpr(a);
      Sum=LinExpr(Sum,BoolVarLinExpr,1);
    }
    return Sum;
}

In order to post the constraint, what I am doing is the following:

    LinExpr SumBoolVars(graph_order(this,fg));
    LinRel Order(SumBoolVars,IRT_LQ,fg_lubOrder);
    Order.post(this,true,ICL_DOM);

Where fg_lubOrde is an integer.

However, it seems I am doing something wrong since there is no 
propagation taking place.

Thanks in advance for your answer,

Luis

_______________________________________________
Gecode users mailing list
users at gecode.org https://www.gecode.org/mailman/listinfo/gecode-users