Convert scala array to Map of item counts

Question

I'm looking for a neat way of converting a scala array to a map containing the frequency of the items occurring in the array.

For eg. :

Given an array like:

val arr = Array("one", "one", "two", "three", "one", "three")

I want a map:

Map("one" -> 3, "two" -> 1, "three" -> 2)

I can do this by writing a function such as

import scala.collection.mutable

def counter[T](arr: Array[T]) = {
  val temp = mutable.Map[T, Int]()
  for (i <- arr) {
    if (temp.contains(i)) temp(i) += 1
    else temp(i) = 1
  }
  temp
}

counter(arr)

I'd like to know if this can be done more efficiently.

Martin Milichovsky · Accepted Answer · 2016-09-23 07:43:53Z

6

I'd use groupBy(identity) and mapValues(_.length):

scala> val arr = Array("one", "one", "two", "three", "one", "three")
arr: Array[String] = Array(one, one, two, three, one, three)

scala> arr.groupBy(identity).mapValues(_.length)
res0: scala.collection.immutable.Map[String,Int] = Map(one -> 3, three -> 2, two -> 1)

Update: This isn't more efficient than your code (and I think it's not possible to beat), but it is definitely more readable. The groupBy method aggregates all the identical (because we used identity) values to one Array, that has some overhead compared to just incrementing count.

Update 2: As noted in the comment, the output is not serializable because the mapValues just wraps the map and evaluates the function on each get. You can materialize the map by map(identity) on the wrapped map:

arr.groupBy(identity).mapValues(_.length).map(identity)

but it's not very nice code.

edited Sep 23, 2016 at 7:43

answered Sep 23, 2016 at 7:27

Martin Milichovsky

7404 silver badges9 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

zenofsahil Over a year ago

Thanks! +1 for being more readable. This way throws a java.io.NotSerializableException error while doing the same thing in Spark though. Hence accepting the alternate answer.

Samar · Accepted Answer · 2016-09-23 07:26:41Z

6

arr.groupBy(identity).map{case (x,y) => x -> y.size}

answered Sep 23, 2016 at 7:26

Samar

2,1011 gold badge15 silver badges18 bronze badges

Collectives™ on Stack Overflow

Convert scala array to Map of item counts

2 Answers 2

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related