Kotlin: fold vs. reduce

Kotlin’s fold and reduce functions both iterate a collection and feature an accumulator. They differ in how their accumulators work:

• With fold, the accumulator can have any type, and you explicitly specify its initial value. The latter makes fold safe to use with empty collections.
• With reduce, the type of the accumulator matches that of the collection elements, and its initial value is the collection’s first element. For empty collections, an exception is thrown, as there’s no accumulator value to return.

fold is thus more powerful than reduce. It can handle all the same operations and many more. The use case for reduce is performing simple calculations more elegantly—consider this comparison:

listOf(8, 16, 24).fold(initial = 1) { product, integer ->
    product * integer
}

listOf(8, 16, 24).reduce { product, integer -> product * integer }

If your collection is guaranteed to be nonempty, and reduce is functionally sufficient, then that’s your best choice. Otherwise, use fold:

listOf(Person.John, Person.Mary, Person.Anne)
    .fold(initial = 0) { totalAge, person -> totalAge + person.age }