Scala 3, ADTs, combining elements in a collection

I'm new to Scala and trying to use Scala 3 enums as ADTs, and I have a collection I'm trying to "reduce" down in the following way:

enum Stats:
    case MaximumHealth(val x: Float)
    case HealthRegen(val x: Float)

def combineStats: Array[Stats] =
    val startWith = Array(Stats.MaximumHealth(20), Stats.MaximumHealth(10), Stats.HealthRegen(5))

    // how can I iterate/recurse over 'startWith' and "combine" multiple stats 
    // (such as MaximumHealth) entries into one?
    // (ie, producing endWith below)?

    val endWith = Array(Stats.MaximumHealth(30), Stats.HealthRegen(5))
    endWith

How can I do this? GroupBy is often suggested but I don't know how to group by just the enum case without the values of the parameters making Stats.MaximumHealth(20) and Stats.MaximumHealth(10) different groups.

Solution

This seems like a case for groupMapReduce, but first, I would prepare some methods on your enum to make life easier:

enum Stats:
  // override is needed because of def x below
  case MaximumHealth(override val x: Float)
  case HealthRegen(override val x: Float)

  // the common getter for x
  def x: Float

  // combinator function for reducing
  def +(that: Stats): Stats = this match
    case MaximumHealth(x) => MaximumHealth(x + that.x)
    case HealthRegen(x)   => HealthRegen(x + that.x)

Now, we can do this:

def combineStats: Array[Stats] =
  val startWith = Array(Stats.MaximumHealth(20), Stats.MaximumHealth(10), Stats.HealthRegen(5))

  val reduced: Map[Int, Stats] =
    startWith.groupMapReduce(_.ordinal)(identity)(_ + _)

  val endWith = reduced.values.toArray
  endWith

Alternative solution

In above solution, I tried to keep your data structur as unchanged as possible. If I were to face that task, I would probably change it like this:

enum StatsType:
  case MaximumHealth, HealthRegen

case class Stats(statsType: StatsType, x: Float)

Then one could do it like this:

def combineStats(stats: List[Stats]): List[Stats] =
  val reduced: Map[StatsType, Float] =
    stats.groupMapReduce(_.statsType)(_.x)(_ + _)
  reduced.map{case (statsType, x) => Stats(statsType, x)}.toList

Additional hint

Not directly related to your question, but: if you want to keep your data structure as an enum as in your question, I'd consider making the x a member of the enum itself (as a simpler solution for a "common getter"):

enum Stats(val value: Float):
  case MaximumHealth(x: Float) extends Stats(x)
  case HealthRegen(x: Float) extends Stats(x)