pyfracval.environments#

Environment variables used by PyFracVAL.

This module centralizes the environment variables that influence runtime behavior in PyFracVAL.

Environment Variables#

PYFRACVAL_DISABLE_PARALLEL_SUBCLUSTERSstr

When set to "1", disables multiprocessing for PCA subcluster generation and forces sequential execution.

PYFRACVAL_INSTALLED_WHEELstr

Records the wheel filename installed on Dask workers during remote package registration.

PYFRACVAL_EXPECTED_VERSIONstr

Records the package version expected on Dask workers after installation.

OMP_NUM_THREADSstr

Controls the number of OpenMP threads used by NumPy/SciPy-backed code.

MKL_NUM_THREADSstr

Controls the number of Intel MKL threads used by NumPy/SciPy-backed code.

OPENBLAS_NUM_THREADSstr

Controls the number of OpenBLAS threads used by NumPy/SciPy-backed code.

NUMEXPR_NUM_THREADSstr

Controls the number of NumExpr threads used by NumPy/SciPy-backed code.

Module Contents#

class pyfracval.environments.EnvironmentConfig[source]#

Runtime environment configuration for PyFracVAL.

Parameters:
  • disable_parallel_subclusters (bool) – Whether PCA subcluster generation should run sequentially.

  • installed_wheel (str | None) – Wheel filename recorded when a package installation is performed on a Dask scheduler or worker.

  • expected_version (str | None) – Expected package version recorded during remote installation.

  • omp_num_threads (str | None) – Value of OMP_NUM_THREADS.

  • mkl_num_threads (str | None) – Value of MKL_NUM_THREADS.

  • openblas_num_threads (str | None) – Value of OPENBLAS_NUM_THREADS.

  • numexpr_num_threads (str | None) – Value of NUMEXPR_NUM_THREADS.

pyfracval.environments.get_env_config()[source]#

Return the PyFracVAL runtime environment configuration.

Returns:

Snapshot of the environment variables used by PyFracVAL.

Return type:

EnvironmentConfig