recursion-to-reduce.md

Recursion to Reduce

Relates To

• Recursion to Filter
• Recursion to Map

Languages

All

Motivation

In functional programming map, reduce and filter (and their synonyms) become fundamental building blocks for your programs.

reduce is used when you want reduce the values in a list down to a single value via some calculation.

Using reduce instead of a loop or recursive function creates more declarative code. It describes the process which you are performing on the data, rather than how you are implementing it.

It also separates to the responsibility of iterating over the collection from the calcuation rule. This increases reusability and composability.

Description

Any time where you have a recursive function or a loop which iterates over a list and to calculate a single value, you might have a good case to use reduce instead.

To refactor you need to extract the transformation function and then pass that to reduce instead.

Examples

Clojure

Before

(defn total-amount
  [item-amounts]
  (if (empty? item-amounts)
      0
      (+ (head item-amounts) (total-amount (tail items)))))

After

(defn total-amount [item-amounts] (reduce + 0 item-amounts))

Haskell

Before

totalAmount :: [Int] -> Int
totalAmount []           = 0
totalAmount (item:items) = item + totalAmount items

After

totalAmount :: [Int] -> Int
totalAmount = foldr (+) 0