[Coin-discuss] OsiClp proper use of initialSolve and resolve, setObjective

[Sebastian Nowozin]

Hello,

just as an update, I filed three issues (#54, #55 and #56) to the 
COIN-OR Osi trac bugtracker,

    OsiClp returns wrong objective for empty constraint matrix
    https://projects.coin-or.org/Osi/ticket/54

    OsiClp crashes on enableFactorization with empty constraint system
    https://projects.coin-or.org/Osi/ticket/55

    OsiClp solves problem but getColSolution() returns NULL
    https://projects.coin-or.org/Osi/ticket/56

Thanks,
Sebastian

Lou Hafer wrote:

> 	At a glance, for any Osi I would expect this to work. In particular, if
> resolve() and isProvenOptimal() succeed, then there should be a valid column
> solution returned by getColSolution().