Julia One Liners compared to Scala

A comparison of one liners in Scala and how they would look in Julia.

Posted on 5 mins read

This was inspired by the blog post 10 Scala One Liners to Impress Your Friends.

1. Multiple Each Item in a List by 2

Original Scala version

(1 to 10) map { _ * 2 }

I think the Julia version is clearer. The x->2x notation for anonymous function looks very similar to regular mathematical notation.

map(x->2x, 1:10)

2. Sum a List of Numbers

Sums numbers from 1 to 1000 in Scala either by using reduceLeft or sum.

(1 to 1000).reduceLeft( _ + _ )
(1 to 1000).sum

Very similar in Julia, but the reduce seems more elegant in that you don’t have to explicitly give the arguments to +. reduce takes a binary operator. So you just pass that. You don’t have to tell Julia how to add numbers with it.

reduce(+, 1:1000)
sum(1:1000)

3. Verify if Exists in a String

Check if a word in a word list is present in a string.

val wordList = List("scala", "akka", "play framework", "sbt", "typesafe")
val tweet = "This is an example tweet talking about scala and sbt."

(wordList.foldLeft(false)( _ || tweet.contains(_) ))
wordList.exists(tweet.contains)

Here I really think the Scala version is rather cryptic. I can not really guess how this works without learning more about Scala. In Julia I would take another approach. This treats the tweet and the wordlist as two sets of words and perform a set intersection ∩ on them:

!isempty(split(tweet) ∩ wordlist)

4. Read in a File

This is a line for reading a file as a string in Scala. While that is apparently impressive by Java standards it is rather verbose for other languages.

val fileText = io.Source.fromFile("data.txt").mkString

In Julia this would just be

filetext = read(“data.txt”, String)

There is another Scala example of reading the file and turing every line into an element in list.

val fileLines = io.Source.fromFile("data.txt").getLines.toList

This is also the sort of thing that is trivial in Julia. Julia’s file open command can take a function as an argument. open gives this function a stream object and then close the file after the function is finished. There are many functions like readline, readlines, read etc which takes streams as arguments. So you can write:

filelines = open(readlines, "data.txt")

5. Happy Birthday to You!

The Scala way of doing this is honestly a bit Rube Goldberg’ish. It is a convoluted way of solving something rather simple:

(1 to 4).map { i => "Happy Birthday " + (if (i == 3) "dear NAME" else "to You") }.foreach { println }

The most straighforward way of doing this in Julia would be:

for i in 1:4
   println("Happy Birtday ", i == 3 ? "dear NAME" : "to You")
end

However if one wants to solve the problem more in the Scala fashion, you could write:

lines = map(1:4) do i
           "Happy Birtday " * (i == 3 ? "dear NAME" : "to You")
       end
foreach(println,  lines)

You could do this with one line, but I split it in two lines for the sake of clarity.

One thing to notice is that you can write code in Julia with a lot less clutter of parenthesis and curly braces. In Scala need 4 parenthesis for the if statement, while Julia avoids this and use a single end keyword instead.

6. Filter list of numbers

Partition a list of students into two categories.

val (passed, failed) = List(49, 58, 76, 82, 88, 90) partition ( _ > 60 )

There is no function like partition in Julia so you would have to write this using multiple lines:

grades = [49, 58, 76, 82, 88, 90]
passed = filter(grade->grade > 60, grades)
failed = filter(grade->grade <= 60, grades)

But of course nothing prevents us from defining a partition function in Julia:

function partition{T}(p, xs::Vector{T})
    as = T[]
    bs = T[]
    for x in xs
        if p(x)
            push!(as, x)
        else
            push!(bs, x)
        end
    end
    (as, bs)
end

Which then would allow us to do like in Scala:

passed, failed = partition(grade->grade > 60, grades)

7. Fetch and Parse an XML web service

Here’s an example fetching the Twitter search feed in Scala.

val results = XML.load("http://search.twitter.com/search.atom?&q=scala")

I don’t know anything about XML stuff in Julia, so I will not attempt and alternative.

8. Find minimum (or maximum) in a List

Another couple of examples using reduceLeft to iterate through a list and apply a function. Added simpler examples of the method min/max on the list.

List(14, 35, -7, 46, 98).reduceLeft ( _ min _ )
List(14, 35, -7, 46, 98).min

List(14, 35, -7, 46, 98).reduceLeft ( _ max _ )
List(14, 35, -7, 46, 98).max

Again this is the sort of thing were Julia is very succinct.

reduce(min, [14, 35, -7, 46, 98])
min(14, 35, -7, 46, 98)

reduce(max, [14, 35, -7, 46, 98])
max(14, 35, -7, 46, 98)

9. Parallel Processing

The following one-liner would give you parallel processing over the list in Scala.

val result = dataList.par.map( line => processItem(line) )

Parallel processing works quite different in Julia, so for this particular example, there is no elegant solution. You would do something like this:

result = zeros(Int, length(dataList))
@threads for (i, line) in enumerate(dataList)
    result[i] = processItem(line)
end

10. Sieve of Eratosthenes

We got this crazy line from Scala, but no I am not going to figure out what this really does and rewrite it in Julia.

(n: Int) => (2 to n) |> (r => r.foldLeft(r.toSet)((ps, x) => if (ps(x)) ps -- (x * x to n by x) else ps))

Perhaps one day when I am very bored I’ll revisit this and update the blog. Here is an example from RIP tutorial on how to do it:

iscoprime(P, i) = !any(x -> i % x == 0, P)

function sieve(n)
    P = Int[]
    for i in 2:n
        if iscoprime(P, i)
            push!(P, i)
        end
    end
    P
end
comments powered by Disqus