Github user shahidki31 commented on the issue:
https://github.com/apache/spark/pull/22784
Hi @kiszk Maximum it can go upto the following limit.
https://github.com/apache/spark/blob/23cfda1547355a823a3b2b2d374e64608c9ce175/mllib/src/main/scala/org/apache/spark/mllib/linalg/EigenValueDecomposition.scala#L78-L79
where ncv = min(n, 2*k), normally k << n.
For eg: if n = 1 million features, we can compute top 100 principle
components.
Number of principle components to compute is configurable.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
