mraad / dbscan-spark Goto Github PK

Hello ,the implementation only support clustering the points with two dimensions ?

This seems like very nicely written code and might exactly fit the problem I am trying to solve.
What is the largest dataset have you run this under? What was the spark job parameters?
Thanks in advance,
Kagan

As my data lies in Hive (not directly in hdfs), It's difficult for me to call your api directly. Is there any alternative plan for this? Thank you!

Hi,

Thanks a lot for great job. This code really works. But i found when tested on my data that it does not respond to the minPoints parameter - the minimum number of points when we consider that this is a cluster. May even put one point in a separate cluster (gives the number, but should give 0 if it is noise).