[Ipopt] Problem while running NLP using IpOpt

[Nandini Chakravorti]

Hi,

            I have the Ipopt.lib(generated on Cygwin using Intel F90
compiler as I am using MUMPS) and then I have written a C++ Dll wrapper in
Visual Studio 2005 - which basically generates a dll for interfacing with
C#.  I have a problem which has 5 variables and 6 constraints.

            Well, now when the execution comes to " eval_grad_f " <*bool
eval_grad_f(int n, const Number* x, bool new_x, double* grad_f)*>  -  The
value of n is 5 as expected. But, if I try to see the value of x[2] in my
debugger watch window I see the following   <   x[2] error: index '2' out of
bound for pointer/array 'x'  >.  And even if I try to copy the values to
another array - I get some junk values in the copied array. Any idea what
might be going wrong?


            And another thing - while I build my Dll I get the following
linker warnings
                          warning LNK4248: unresolved typeref token
(01000017) for 'Ipopt.IteratesVector'; image may not run
                          warning LNK4248: unresolved typeref token
(01000018) for 'Ipopt.IpoptData'; image may not run
                          warning LNK4248: unresolved typeref token
(01000019) for 'Ipopt.IpoptCalculatedQuantities'; image may not run

Regards,
Nandini
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.coin-or.org/pipermail/ipopt/attachments/20081010/1db7a507/attachment.html