net.sansa_stack.rdf. spark.model.GraphLoader.constructGraph produces an incorrect graph (but constructHashedGraph is OK)

[glaisney]

Loading the taxonomy graph hereafter as an RDD[Triple], then converting it in GraphX format with myRdd.asGraph() or constructGraph(myRdd) leads to an incorrect result. https://github.com/obophenotype/ncbitaxon/releases/download/v2024-05-08/ncbitaxon.owl

Workaround : use myRdd.asHashedGraph or constructHashedGraph(myRdd) instead. Seems to work fine.