Skip to contents

Matrix Multiplication

Source: R/api.R, R/api-generics.R
nv_matmul.Rd

Matrix multiplication of two arrays. You can also use the %*% operator. Supports batched matrix multiplication when inputs have more than 2 axes.

Usage

nv_matmul(lhs, rhs, precision = "highest")

Arguments

lhs, rhs

(arrayish)
Numeric arrays with at least 2 axes. Can be any numeric data type; the two are promoted to a common data type. An R value assumes the data type of the other operand, and materializes at its default data type when that has none either.

precision

(character(1))
Controls the trade-off between speed and numerical accuracy of the operation. One of "highest" (default), "high" or "default". See prim_dot_general() for details.

Value

(arrayish)
Has the operands' common data type and the shape given under Shapes.

Shapes

  • lhs: (b1, ..., bk, m, n)

  • rhs: (b1, ..., bk, n, p)

  • output: (b1, ..., bk, m, p)

See also

prim_dot_general() for the underlying primitive.

Examples

# a 2x3 times a 3x2 gives a 2x2 at the operands' common data type
x <- nv_matrix(1:6, nrow = 2)
y <- nv_matrix(1:6, nrow = 3)
nv_matmul(x, y)
#> AnvlArray
#>  22 49
#>  28 64
#> [ CPUi32{2,2} ] 
x %*% y
#> AnvlArray
#>  22 49
#>  28 64
#> [ CPUi32{2,2} ]