I want to find not just the maximum value of a function applied to a list (for which I would just use List.maxBy) but also the value in the list this occurred at. This feels like a fairly common operation and given the richness of the F# libraries in general I wouldn’t be at all surprised to discover it was actually already available but I cannot seem to find it if it is!
To illustrate with an example, I want to be able to map a list domain and a function f
let domain = [0 .. 5]
let f x = -x * (x - 2)
to (1, 1) (since the function applied to an other element of the list is less than 1).
I first tried this:
let findMaximum domain f =
let candidates = [ for x in domain do
yield x, f x ]
let rec findMaximumHelper domain f currentMax =
match domain with
| [] -> currentMax
| head::tail ->
let cand = f head
match currentMax with
| None ->
let newMax = Some(head, cand)
findMaximumHelper tail f newMax
| Some(maxAt, possMax) ->
let newMax =
if cand > possMax then Some(head, cand)
else Some(maxAt, possMax)
findMaximumHelper tail f newMax
findMaximumHelper domain f None
let answer = findMaximum domain f
at which point I realised this is very close to a fold operation, and put together
let findMaximum2 domain f =
let findMaximumHelper f acc x =
let cand = f x
match acc with
| None -> Some(x, cand)
| Some(maxAt, possMax) ->
if cand > possMax then Some(x, cand)
else Some(maxAt, possMax)
List.fold (findMaximumHelper f) None domain
let answer2 = findMaximum2 domain f
instead.
My question is, are these idiomatic F# ways of solving this problem, or indeed, is there a better way of solving this?
Indeed, the F# library provides all the necessary higher order functions to express this succinctly:
Note: updated to use
Seq.mapandSeq.maxByinstead ofList.mapandList.maxByto address @ildjarn’s concern about creating an unnecessary intermediate list.