PowerIterationClustering

[Monish R <mo...@gmail.com>]

Hi guys,
I am new to mlib and trying out PowerIterationClustering as per the example
mentioned below,

https://github.com/apache/spark/blob/master/examples/src/main/java/org/apache/spark/examples/mllib/JavaPowerIterationClusteringExample.java


I am having trouble in understanding how the output is created.
For instance if i change the input as shown below, i would like to
understand how the algorithm arrived at grouping 0 and 2 , while keeping
the rest in another cluster.

k = 2 .

Input :
      new Tuple3<>(0L, 1L, 0.9),
      new Tuple3<>(1L, 2L, 0.7),
      new Tuple3<>(2L, 3L, 0.3),
      new Tuple3<>(3L, 4L, 0.5),
      new Tuple3<>(4L, 5L, 0.2)));

Output :
4 -> 0
0 -> 1
1 -> 0
3 -> 0
5 -> 0
2 -> 1

Kindly guide if you have any info on using the algorithm / point to some
materials that are suitable for beginners on this context.



Regards.