Hello, I am having a problem trying to initialize the distributed property map in parallel code. I get a memory access violation when running this part of code: BGL_FORALL_EDGES(e, g, Graph) { remote_key k(rank, index); put(values, k, 0.1); index++; } I followed the distributed_property_map_test, but I cannot figure where is the problem. It fails at the first edge. The distributed map is defined as follows: typedef boost::parallel::distributed_property_map ValueMap; ValueMap values(g.process_group(), remote_key_to_global(), eindex); values.set_reduce(rank_accumulate_reducer<double>()); Where eindex is a local map with the local indices of the edges. Please, could someone help me? -- View this message in context: http://boost.2283326.n4.nabble.com/PBGL-distributed-property-map-problem-tp4... Sent from the Boost - Users mailing list archive at Nabble.com.