Automatic Differentiation
 
Loading...
Searching...
No Matches

◆ reduce_combine()

template<int call_id, typename ReduceF , typename CombineF >
result_t stan::math::mpi_parallel_call< call_id, ReduceF, CombineF >::reduce_combine ( )
inline

Once all data is distributed and cached the reduce_combine evaluates all assigned function evaluations locally, transfers all results back to the root and finally combines on the root all results.

Definition at line 269 of file mpi_parallel_call.hpp.