[Osi] dual bound on the primal

[Kyle Tarplee]

I'm sure this question has been asked before but I could not find it (nor could google), so here goes.

I am using OSI and CLP.  I create the minimization problem then do the initialSolve().  After that I modify some column bounds with setColLower() and setColUpper().  I only modify the bounds to make them tighter.  I want to find a tighter lower bound on the primal objective function (better then the old objective function value).  How do I get this information from the OSI interface?   I will eventually resolve() the problem and do this again and again.

Generally this is simply the dual solution vector dotted with the new (more constrained) right hand side vector (if I remember correctly).  In OSI I think this corresponds to getRowPrice() and getRightHandSide() but the length seem wrong to me.  I have both upper and lower bounds on the rows and both these vectors are getNumRows() in length so half the constraints are unaccounted for in the dual.  Also the column bounds are unaccounted for in the dual.  Seems to be the length of the dual in my case should be 2*numCols + 2*numRows, because I have upper and lower bounds on both the rows and columns.

Thanks,
Kyle