fastPLS_blas: Report the numerical library selected when fastPLS was...

In fastPLS: Fast Partial Least Squares for High-Dimensional Data

Report the numerical library selected when fastPLS was compiled

Description

This reports the CPU matrix library selected by the package configuration, rather than attempting to infer a library from the current R session. macOS builds normally report "Accelerate". Linux and Windows builds report "OpenBLAS" when OpenBLAS was found or explicitly requested, and "R BLAS/LAPACK" when the package used R's portable fallback.

Usage

fastPLS_blas(details = TRUE)

Arguments

details

Logical. Return a detailed named list when TRUE, or the former scalar backend name when FALSE.

Details

With details = TRUE, the result additionally reports the library version, configuration string, selected CPU core, parallel runtime, active thread count, and resolved library path when these are exposed by the linked library. OpenBLAS provides all fields except that a statically linked Windows build may not expose a separate library path. Accelerate and R BLAS/LAPACK do not expose the same runtime metadata, so unavailable fields are NA.

For reproducible performance benchmarks on Linux or Windows, install with FASTPLS_USE_OPENBLAS=1 and verify both backend and the detailed version and core fields before running the analysis.

Value

With details = TRUE, a named list containing backend, version, configuration, core, parallel, threads, and library. With details = FALSE, a single character string: "Accelerate", "OpenBLAS", or "R BLAS/LAPACK".

Examples

fastPLS_blas()
fastPLS_blas(details = FALSE)

fastPLS documentation built on Sept. 29, 2026, 1:06 a.m.