MCPcopy Create free account
hub / github.com/BirolLab/abyss / reduce

Method reduce

Parallel/CommLayer.cpp:106–114  ·  view source on GitHub ↗

Block until all processes have reached this routine. * @return the sum of count from all processors */

Source from the content-addressed store, hash-verified

104 * @return the sum of count from all processors
105 */
106long long unsigned CommLayer::reduce(long long unsigned count)
107{
108 logger(4) << "entering reduce: " << count << '\n';
109 long long unsigned sum;
110 MPI_Allreduce(&count, &sum, 1, MPI_UNSIGNED_LONG_LONG, MPI_SUM,
111 MPI_COMM_WORLD);
112 logger(4) << "left reduce: " << sum << '\n';
113 return sum;
114}
115
116/** Reduce the specified vector. */
117vector<unsigned> CommLayer::reduce(const vector<unsigned>& v)

Callers 4

runMethod · 0.80
controlErodeMethod · 0.80
controlCoverageMethod · 0.80
runControlMethod · 0.80

Calls 1

sizeMethod · 0.45

Tested by

no test coverage detected