reduce

Accumulates a value by applying the operation to each element from left to right, using the first element as the initial accumulator. Throws on empty arrays; use reduceOrNull for a safe variant.

Signatures

inline fun <S, D : Dimension, T : S> MultiArray<T, D>.reduce(
    operation: (acc: S, T) -> S
): S

inline fun <S, T : S> MultiArray<T, D1>.reduceIndexed(
    operation: (index: Int, acc: S, T) -> S
): S

inline fun <S, D : Dimension, T : S> MultiArray<T, D>.reduceMultiIndexed(
    operation: (index: IntArray, acc: S, T) -> S
): S

inline fun <S, D : Dimension, T : S> MultiArray<T, D>.reduceOrNull(
    operation: (acc: S, T) -> S
): S?

Parameters

Parameter	Type	Description
`operation`	`(acc: S, T) -> S`	Combines the accumulator with each element.
`operation`	`(index: Int, acc: S, T) -> S`	With flat index (1D arrays only).
`operation`	`(index: IntArray, acc: S, T) -> S`	With multi-dimensional index.

Returns: The final accumulated value. reduce throws UnsupportedOperationException on empty arrays; reduceOrNull returns null.

Example

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

a.reduce { acc, v -> acc + v }        // 10
a.reduce { acc, v -> acc * v }        // 24

a.reduceIndexed { i, acc, v -> acc + i * v }
// 1 + 1*2 + 2*3 + 3*4 = 21

28 February 2026