groupNDArray

Groups array elements by a key function, returning a Map where each key maps to a 1D NDArray of matching elements. groupingNDArrayBy returns a Grouping for lazy aggregation with eachCount, fold, or reduce.

Signatures

inline fun <T, D : Dimension, K> MultiArray<T, D>.groupNDArrayBy(
    keySelector: (T) -> K
): Map<K, NDArray<T, D1>>

inline fun <T, D : Dimension, K, V : Number> MultiArray<T, D>.groupNDArrayBy(
    keySelector: (T) -> K,
    valueTransform: (T) -> V
): Map<K, NDArray<V, D1>>

inline fun <T, D : Dimension, K> MultiArray<T, D>.groupingNDArrayBy(
    keySelector: (T) -> K
): Grouping<T, K>

Parameters

Parameter	Type	Description
`keySelector`	`(T) -> K`	Extracts the group key from each element.
`valueTransform`	`(T) -> V`	Transforms the element before adding to the group.

Returns: Map<K, NDArray<T, D1>> — each key maps to a 1D array of matching elements.

Example

val a = mk.ndarray(mk[1, 2, 3, 4, 5, 6])

val groups = a.groupNDArrayBy { if (it % 2 == 0) "even" else "odd" }
// "odd"  -> [1, 3, 5]
// "even" -> [2, 4, 6]

val grouping = a.groupingNDArrayBy { it % 3 }
grouping.eachCount()  // {1=2, 2=2, 0=2}

28 February 2026