[petsc-users] PetscOptionsGetRealArray command line format

[Satish Balay]

http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Sys/PetscOptionsGetRealArray.html


Input Parameters
      nmax      - maximum number of values to retrieve

Output Parameters
       nmax     - actual number of values retreived


So you need to do:

   double myarray[10];
   int nmax = 10;

I'm not sure if there is a way to detect this incorrect usage [without valgrind] ...

Satish

On Fri, 14 Feb 2014, Xiangdong wrote:

> I have a simple script (see below) by just
> calling PetscOptionsGetRealArray. I run the program like this: mpirun -np
> 1 runarraytest -myarray 1,2,3,4,5. However, the output is always "the
> length of input myarray is 0".
> 
> Can you produce the correct result? Thank you.
> 
> 
> Below is the script I use:
> 
> 
> #include <petsc.h>
> 
> int main(int argc, char **argv)
> {
>   PetscInitialize(&argc, &argv, PETSC_NULL, PETSC_NULL);
> 
>   double myarray[10];
>   int nmax;
>   PetscBool flg;
> 
>   PetscOptionsGetRealArray(NULL,"-myarray",myarray,&nmax,&flg);
> 
>   printf("the length of input myarray is %d \n",nmax);
> 
>   int i;
>   for(i=0;i<nmax;++i)
>     {
>       printf(" %g\n",myarray[i]);
>     }
> 
>   PetscFinalize();
>   return 0;
> }
> 
> 
> 
> On Fri, Feb 14, 2014 at 3:47 PM, Jed Brown <jed at jedbrown.org> wrote:
> 
> > Xiangdong <epscodes at gmail.com> writes:
> > > Now I see some random behavior. After compiling the code, for the same
> > > command line input, the output sometimes are correct (print values
> > > 1,2,3,4,5 with nmax=5), but sometime output all zeros (nmax is zero). I
> > did
> > > not change anything, but just running same input many times.
> > >
> > > Any hints for this random behavior? Thank you.
> >
> > Run with valgrind.  It is probably some other memory corruption
> > interfering.
> >
>