Arithmetic operators for GPU-aware tensors

cudatensor objects support element-wise +, -, *, /, and ^. Operands follow trailing-dimension broadcasting. Mixed dtypes are promoted without silently truncating fractional values; integer arithmetic is promoted to float64 to avoid R integer overflow. Compatible dimension labels are retained. When both operands label the same non-broadcast dimension, their labels must be identical.

Usage

# S3 method for class 'cudatensor'
Ops(e1, e2)

# S3 method for class 'cudatensor'
x %*% y

Arguments

e1, e2

A cudatensor or numeric object for element-wise arithmetic.

x, y

A cudatensor or numeric matrix for matrix multiplication.

Value

A cudatensor on the device of the tensor operand on the left (or the tensor operand on the right when the left operand is a base object).

Details

Use %*% for matrix multiplication.

Examples

x <- cuda_tensor(matrix(1:6, 2, 3), device = "cpu")
to_cpu(x + c(0.5, 1, 1.5))
#>      [,1] [,2] [,3]
#> [1,]  1.5    4  6.5
#> [2,]  2.5    5  7.5

y <- cuda_tensor(matrix(1:6, 3, 2), device = "cpu")
to_cpu(x %*% y)
#>      [,1] [,2]
#> [1,]   22   49
#> [2,]   28   64