Binary Interface

calcPhiASE is the standalone C++ executable. Most users call it through PhiASE.run(...); use the binary directly when you already have a HASEonGPU openPMD input series.

Build

Manual compilation is required for direct command-line use. See CMake Build Options.

./build/calcPhiASE

Usage

The binary reads one openPMD input series and writes one openPMD result series:

./build/calcPhiASE \
    --input-path=<openPMD-input-series> \
    --output-path=<openPMD-output-series>

The input series contains mesh topology, material fields, dynamic beta fields, backend settings, parallel mode, sample range, and optional RNG seed. This is the same transport layout written by the Python interface; see openPMD Transport.

Examples

Single process:

./build/calcPhiASE --input-path=./input.bp --output-path=./output.bp

MPI launch:

mpiexec -npernode 4 ./build/calcPhiASE \
    --input-path=./input.bp \
    --output-path=./output.bp

Arguments

--input-path: Path to the HASEonGPU openPMD input series.
--output-path: Path for the result series. Results are written as core_result_phi_ase, core_result_mse, core_result_total_rays, and core_result_dndt_ase records.