[Coin-ipopt] Problem with IPOPT_CHECK

Andreas Waechter andreasw at watson.ibm.com
Tue May 24 17:37:36 EDT 2005

Hi Don,

This is old code that hasn't been updated in a while (and is therefore not
mentioned in the documentation).

The original idea was that a user can simply replace the call to 'IPOPT'
by 'IPOPT_CHECK', but a number of changes have happened in the meantime,
e.g internal memory allocation in IPOPT which is not in IPOPT_CHECK, and
the argument lists of GET_F etc have changed (as you noticed).

If you want you can fix it and send me the modification so that I can add
it to the distribution for the use of other users.  (Remember, this is an
open source project that welcomes contributions... :)

Best regards,


On Tue, 24 May 2005 don.jones at gm.com wrote:

> Hello IPOP users!
> I have a problem with IPOPT_CHECK, the provided routine for checking
> derivatives.   It seems to call routines get_f and get_g with a different
> argument list and, as a result, the compiler generates an error.   Here is
> the call to GET_F inside ipopt_check.f
>       call GET_F(NORIG, RW(p_x+1), IW(p_ivar+1), NORIG, XORIG,
>      1     M, RW(p_cscale+1), f, EV_F, DAT, IDAT)
>          IERR = 3475
>          goto 9999
> Here is how subroutine GET_F is defined in get_f.f
>       subroutine GET_F(N, X, IVAR, NORIG, XORIG, M, CSCALE, NLB, ILB,
>      1     S_L, NUB, IUB, S_U, MU, F, LIW, IW, IERR, EVAL_F, DAT, IDAT)
> As you can see, the above has more arguments.
> Is this a known bug and has it been fixed?
> Thanks,
> Don Jones
> djones at gm.com
> _______________________________________________
> Coin-ipopt mailing list
> Coin-ipopt at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/coin-ipopt

More information about the Coin-ipopt mailing list