shootingstar (#18) - strange problem with shortest_path_shooting_star (#307) - Message List

Hello friends, I'm in trouble!

The last edge of the routing generated, has a very high cost. Is wrong ! My query is:

SELECT * FROM shortest_path_shooting_star( 'SELECT gid AS id, source, target, (length * (cost_roadclass)) AS cost, (reverse_cost * (reverse_cost_roadclass)) AS reverse_cost, x1, y1, x2, y2, rule, to_cost FROM ways ', 18924, 11133, true, true)

The Data Output is:

vertex_id;edge_id;cost 30493;18924;1112.64237131776 10643;18925;836.255275860839 10083;18926;1310.90131172293 10084;25217;443.444989584447 73647;25218;302.520551722766 73648;3925;1630.75290942325 37358;3926;708.301774052125 10642;11133;100060000

You can see, the cost is very high because it is the opposite direction. This is what I do not understand. Why the edge last takes wrong?

Thanks for your help. regards !!