Open hbanoscervantes opened 4 years ago
It seems that when correcting the search space the choice of epsilon on the 2D DBSCAN is not enough to get the expected results. For now, a choice of twice the current epsilon seems to work (i.e multiply by 4 instead of 2 from the 1D to the 2D DBSCAN).
The function rangeQuery2 is incorrectly looking at the neighbors using the manhattan distance. A corrected version has is attached (substitute it in ConsensusStems.pl) rangeQuery2.txt