Created
August 30, 2013 21:21
-
-
Save samklr/6394396 to your computer and use it in GitHub Desktop.
DotProduct matrix in scala and on spark
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
def dotProduct(vector: Array[Int], matrix: Array[Array[Int]]): Array[Int] = { | |
// ignore dimensionality checks for simplicity of example | |
(0 to (matrix(0).size - 1)).toArray.map( colIdx => { | |
val colVec: Array[Int] = matrix.map( rowVec => rowVec(colIdx) ) | |
val elemWiseProd: Array[Int] = (vector zip colVec).map( entryTuple => entryTuple._1 * entryTuple._2 ) | |
elemWiseProd.sum | |
} ) | |
} | |
val A = sc.parallelize(Array(Array(7, 5, 4), Array(0, 3, 2), Array(8, 0, 5), Array(-11, 7, -4), Array(-8, 2, -13), Array(5, 0, -2))) | |
val B = sc.broadcast(Array(Array(100, -80, 75, -105, 30, -50), Array(60, -60, 60, -60, 60, -60), Array(-50, 30, -105, 75, -80, 100))) | |
A.map( row => dotProduct(row, B.value) ).collect | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
If
B
is not ginormous, it's actually a good idea to broadcast it, the performance should linearly scale on the number of workers.