c++ - User Defined Reduction on vector of varying size -

i'm trying define own reduction vectors of complex<float>, following this answer question reducing on array in openmp.

but size of vectors aren't fixed @ compile time, i'm not sure how define initializer vector in declare reduction pragma. is, can't have

initializer( omp_priv=tcomplexvector(10,0) ) 

but initializer needed vectors.

how can pass initializer clause size of vector need @ run time? have far below:

typedef std::vector<complex<float>> tcmplxvec;  void complexadd(tcmplxvec & x,tcmplxvec & y){   (int i=0;i<x.size();i++)    {       x.real()+= y.real();       //... same imaginary part , other operations   }  }  #pragma omp declare reduction(addcmplx: tcmplxvec: \ complexadd(&omp_out, &omp_in)) initializer( \ omp_priv={tcmplxvec(**here want variable length**,0} )  void dosomeoperation () {     //tcmplxvec vec empty , anothervec not      //so each thread runs inner loop serially   #pragma omp parallel reduction(addcmplx: vec)    ( n=0 ; n<10 ; ++n )     {       (m=0; m<=somelength; ++m){         vec[m] += anothervec[m+someoffset dependend on n , else];       }     } } 

you have dig little bit find online right now, in section 2.15 of openmp standard, user-declared reductions discussed, you'll find "the special identifier omp_orig can appear in initializer-clause , refer storage of original variable reduced."

so use initializer (omp_priv=tcmplxvec(omp_orig.size(),0)), or initalizer ( omp_priv(omp_orig) ) initialize vector in reduction.

so following works (note don't need write own routine; can use std::transform , std::plus add vectors; use std::valarray rather vectors, depending on how use them, has operator+ defined):

#include <complex> #include <vector> #include <algorithm> #include <functional> #include <iostream> #include <omp.h>  typedef std::vector< std::complex<float> > tcmplxvec;  #pragma omp declare reduction( + : tcmplxvec : \         std::transform(omp_in.begin( ),  omp_in.end( ), \                        omp_out.begin( ), omp_out.begin( ), \                        std::plus< std::complex<float> >( )) ) \                        initializer (omp_priv(omp_orig))  int main(int argc, char *argv[]) {      int size;      if (argc < 2)         size = 10;     else         size = atoi(argv[1]);      tcmplxvec result(size,0);      #pragma omp parallel reduction( + : result )     {         int tid=omp_get_thread_num();          (int i=0; i<std::min(tid+1,size); i++)              result[i] += tid;     }      (int i=0; i<size; i++)          std::cout << << "\t" << result[i] << std::endl;      return 0; } 

running gives

$ omp_num_threads=1 ./reduction 8 0   (0,0) 1   (0,0) 2   (0,0) 3   (0,0) 4   (0,0) 5   (0,0) 6   (0,0) 7   (0,0)  $ omp_num_threads=4 ./reduction 8 0   (6,0) 1   (6,0) 2   (5,0) 3   (3,0) 4   (0,0) 5   (0,0) 6   (0,0) 7   (0,0)  $ omp_num_threads=8 ./reduction 8 0   (28,0) 1   (28,0) 2   (27,0) 3   (25,0) 4   (22,0) 5   (18,0) 6   (13,0) 7   (7,0)