For some reason, edge values (representing the number of coauthorship relations) are being computed twice by the parsers/sucupira.py script.