ELKI DBSCAN for million files

Question

I am using dbscan for clustering points, as my points are more than 1 million I use r*-tree too.

I use ELKI in command line:

java -cp elki.jar
de.lmu.ifi.dbs.elki.application.KDDCLIApplication
-db.index tree.spatial.rstarvariants.rstar.RStarTreeFactory
-algorithm clustering.DBSCAN
-dbc.in points1.txt
-dbscan.epsilon 20
-dbscan.minpts 10
-out results3/DBSCANeps20min10

for small files its ok but for 4 million files the error occurred:

at de.lmu.ifi.dbs.elki.database.ids.integer.DoubleIntegerArrayQuickSort.quickSort(Unknown Source)

Answer 1

This is a known bug in an old version of ELKI, when there are many duplicate distances.

It can be resolved by updating to a current version.