Specifying the Seed value guarantees that the result is repeatable on the same cluster. However, the result can differ between clusters, as the sample graph is not deterministic.
SELECT * FROM randomwalksample ( ON citvertices AS "vertices" PARTITION BY id ON citedges AS "edges" PARTITION BY from_id TargetKey ('to_id') FlyBackRate (0.15) SampleRate (0.15) Seed (1000) OutputTables ('rw_vertices_15', 'rw_edges_15') );