Performance problem with BlockMatrix.add()

[Andi <ar...@googlemail.com>]

Hello,

I'm implementing a pagerank-like iterative algorithm where in each iteration
a number of matrix operations is performed. One step is to add two matrices
that are both the result of several matrix multiplications. Unfortunately,
using the add() Operation of BlockMatrix Spark gets completely stuck for >
30 sec. I implemented an alternative (naive) addition myself, but it still
takes > 10 sec. The matrix consists of about 1500x1500 entries. I tested in
local mode + standalone mode on a laptop with 4 cores and 20gb ram. The code
that causes the problems can be found here  http://imgur.com/a/ZaomX
<http://imgur.com/a/ZaomX>  
Any help is highly appreciated!

Thanks
Andi







--
View this message in context: http://apache-spark-user-list.1001560.n3.nabble.com/Performance-problem-with-BlockMatrix-add-tp27825.html
Sent from the Apache Spark User List mailing list archive at Nabble.com.

---------------------------------------------------------------------
To unsubscribe e-mail: user-unsubscribe@spark.apache.org