Re: SPAM-LOW: Re: Sundials CVODE: y vector values

[Radu Serban-3]

Brian,

The problem is that you are not loading the derivatives into the
provided N_Vector ydot, but rather create a new one!

Your RHS function should be something like this (of course, you could
just use memcpy instead of the for loop):

//==============================================================
static int RHSfunction(realtype t, N_Vector y, N_Vector ydot, void* user_data)
{
        DiffusionModel* dmod;
        dmod =(DiffusionModel*)user_data;

        double *qdot;
        double *q;
        q = NV_DATA_S(y);
        qdot = NV_DATA_S(ydot);

        dmod->solve_for_qdot(t,q);

        for (int i = 0; i < dmod->totalNumDof; i++)
            qdot[i] = dmod->udot[i];

        int flag=0;
        return flag;
}
//==============================================================


--Radu



bowens1 wrote:
 > "lost"/corrupted  after first call to RHSfunction
 > Sender: owner-sundials-users@...
 > Reply-To: sundials-users@...
 >
 > I have looked into the problem ore and determined this is more likely a
 > problem with how "ydot" is being handled. I checked my assignment to ydot
 > within the RHS function and the contents are correct. However, after adding
 > a print statement in the CVODE code after the call to
 >
 > retval = f(tn, zn[0], zn[1], user_data);
 >
 > The values of "ydot" or "zn[1]" are much different than those that should to
 > be output (based off what I am setting in the RHS function).
 >
 > Based on what I've read from the documentation it seems that my definition
 > of ydot in the RHS function should be adequate but perhaps I am missing
 > something.
 >
 > Thanks,
 > Brian
 >
 >
 >
 > bowens1 wrote:
 >  >
 >  > Hello,
 >  >
 >  > I am currently implementing CVODE into an in-house FEA code
written in C++
 >  > that makes thorough use of object oriented programming. Within a
 >  > particular
 >  > method of a class I am trying to implement CVODE. I have made my RHS
 >  > function a static function outside the class as I have seen done in the
 >  > examples. I am having some problem in that at the second call to the
 >  > RHSfunction the values passed in under "N_vector y" seem clearly
 >  > incorrect.
 >  > For example my IVP for  the problem is [0.79 0 0] and the second call to
 >  > RHS
 >  > function has the y vector as [-5.0784477968231915e+057
 >  > -5.0784477968231915e+057 -5.0784477968231915e+057]. It seems as
though the
 >  > y
 >  > vector is being corrupted some where in the main CVODE function.
 >  >
 >  >
 >  > To test out sundials I wrote a smaller matlab code that is not nearly as
 >  > robust as the C++ implementation. The matlab implementation worked with
 >  > great success. I am using the same settings in both C++ and matlab
 >  > implementations. Also the two implementations are giving identical output
 >  > during the first call to the RHS function, the  C++ code gives erroneous
 >  > results after this.
 >  >
 >  > If needed here is my RHS function:
 >  >
 >  > DiffusionModel is a class and user_data points to an object of this class
 >  > that I later assign to dmod. The method "solve_for_qdot()"
solves for qdot
 >  > and assigns it to some data object "udot" local to the DiffusionModel
 >  > class.
 >  > //==============================================================
 >  > static int RHSfunction(realtype t, N_Vector y, N_Vector ydot, void
 >  > *user_data)
 >  > {
 >  >        DiffusionModel* dmod;
 >  >        dmod =(DiffusionModel*)user_data;
 >  >
 >  >        double *qdot;
 >  >        double *q;
 >  >        q = NV_DATA_S(y);
 >  >        dmod->solve_for_qdot(t,q);
 >  >
 >  >        qdot = dmod->udot;
 >  >        ydot = N_VMake_Serial(dmod->totalNumDof,qdot);
 >  >
 >  >        int flag=0;
 >  >        return flag;
 >  > }
 >  > //==============================================================
 >  >
 >  > I greatly appreciate any help in this matter, and I'll be glad to provide
 >  > anymore details as needed.
 >  >
 >  > Thanks,
 >  > Brian
 >  >
 >
 >

----------------------------------------------------------------------------------------------------------------
Center for Applied Scientific Computing
Lawrence Livermore National Laboratory
P.O. Box 808, L-561
Livermore, CA  94551

[phone]  (Office)            cswoodward@...
[phone]  (Fax)               http://people.llnl.gov/woodward6