Row Euclidean Norms Computed in Parallel with OpenMP

Computes the Euclidean (L2) norm of each row of a numeric matrix, with the rows divided across threads by OpenMP.

Usage

parallel_row_norms(x)

Arguments

x

A numeric matrix.

Value

A numeric vector with one entry per row of x, each the square root of the sum of squares of that row.

Details

The #pragma omp parallel for directive splits the outer loop over rows across the available OpenMP threads. The directive is only meaningful when the package is compiled with OpenMP support; without it the loop simply runs serially.

Only the input matrix is read and the pre-allocated output vector is written inside the parallel region, and each thread writes a distinct index. No R API is touched inside the loop, which is what keeps the routine thread-safe.

Examples

x = matrix(rnorm(20), nrow = 5)
parallel_row_norms(x)
#> [1] 2.656382 2.013038 3.204453 1.736271 1.250746