Arithmetic operations

Overview

All arithmetic operations are element-wise and return a new NDArray. Both array-array and array-scalar forms are supported. In-place variants (+=, -=, *=, /=) modify the array directly.

Supported element types: Byte, Short, Int, Long, Float, Double, ComplexFloat, ComplexDouble.

Addition

// Array + Array
operator fun <T, D : Dimension> MultiArray<T, D>.plus(other: MultiArray<T, D>): NDArray<T, D>

// Array + Scalar
operator fun <T, D : Dimension> MultiArray<T, D>.plus(other: T): NDArray<T, D>

// Scalar + Array
operator fun <D : Dimension> Int.plus(other: MultiArray<Int, D>): NDArray<Int, D>
// (also for Byte, Short, Long, Float, Double, ComplexFloat, ComplexDouble)

// In-place
operator fun <T, D : Dimension> MutableMultiArray<T, D>.plusAssign(other: MultiArray<T, D>)
operator fun <T, D : Dimension> MutableMultiArray<T, D>.plusAssign(other: T)

Example

val a = mk.ndarray(mk[1, 2, 3])
val b = mk.ndarray(mk[10, 20, 30])

val c = a + b        // [11, 22, 33]
val d = a + 100      // [101, 102, 103]
val e = 5 + a        // [6, 7, 8]

Subtraction

operator fun <T, D : Dimension> MultiArray<T, D>.minus(other: MultiArray<T, D>): NDArray<T, D>
operator fun <T, D : Dimension> MultiArray<T, D>.minus(other: T): NDArray<T, D>
operator fun <D : Dimension> Int.minus(other: MultiArray<Int, D>): NDArray<Int, D>

// In-place
operator fun <T, D : Dimension> MutableMultiArray<T, D>.minusAssign(other: MultiArray<T, D>)
operator fun <T, D : Dimension> MutableMultiArray<T, D>.minusAssign(other: T)

Example

val a = mk.ndarray(mk[10, 20, 30])
val b = mk.ndarray(mk[1, 2, 3])

val c = a - b        // [9, 18, 27]
val d = a - 5        // [5, 15, 25]
val e = 100 - a      // [90, 80, 70]

Multiplication

operator fun <T, D : Dimension> MultiArray<T, D>.times(other: MultiArray<T, D>): NDArray<T, D>
operator fun <T, D : Dimension> MultiArray<T, D>.times(other: T): NDArray<T, D>
operator fun <D : Dimension> Int.times(other: MultiArray<Int, D>): NDArray<Int, D>

// In-place
operator fun <T, D : Dimension> MutableMultiArray<T, D>.timesAssign(other: MultiArray<T, D>)
operator fun <T, D : Dimension> MutableMultiArray<T, D>.timesAssign(other: T)

Example

val a = mk.ndarray(mk[2, 3, 4])
val b = mk.ndarray(mk[10, 20, 30])

val c = a * b        // [20, 60, 120]
val d = a * 3        // [6, 9, 12]

Division

operator fun <T, D : Dimension> MultiArray<T, D>.div(other: MultiArray<T, D>): NDArray<T, D>
operator fun <T, D : Dimension> MultiArray<T, D>.div(other: T): NDArray<T, D>
operator fun <D : Dimension> Int.div(other: MultiArray<Int, D>): NDArray<Int, D>

// In-place
operator fun <T, D : Dimension> MutableMultiArray<T, D>.divAssign(other: MultiArray<T, D>)
operator fun <T, D : Dimension> MutableMultiArray<T, D>.divAssign(other: T)

Example

val a = mk.ndarray(mk[10.0, 20.0, 30.0])

val b = a / 2.0      // [5.0, 10.0, 15.0]

Unary negation

operator fun <T, D : Dimension> MultiArray<T, D>.unaryMinus(): NDArray<T, D>

Example

val a = mk.ndarray(mk[1, -2, 3])
val b = -a           // [-1, 2, -3]

In-place operations

Use the assign variants to modify an array without creating a copy:

val a = mk.ndarray(mk[1, 2, 3])
a += 10              // a is now [11, 12, 13]
a *= 2               // a is now [22, 24, 26]

Pitfalls

Both operands in array-array operations must have the same shape. Broadcasting is not supported.
Integer division truncates — mk.ndarray(mk[5]) / 2 yields [2], not [2.5].

28 February 2026