Global state

NumPy has a few import-time, compile-time, or runtime options which change the global behaviour. Most of these are related to performance or for debugging purposes and will not be interesting to the vast majority of users.

Performance-related options

Number of threads used for Linear Algebra

NumPy itself is normally intentionally limited to a single thread during function calls, however it does support multiple Python threads running at the same time. Note that for performant linear algebra NumPy uses a BLAS backend such as OpenBLAS or MKL, which may use multiple threads that may be controlled by environment variables such as OMP_NUM_THREADS depending on what is used. One way to control the number of threads is the package threadpoolctl

Madvise Hugepage on Linux

When working with very large arrays on modern Linux kernels, you can experience a significant speedup when transparent hugepage is used. The current system policy for transparent hugepages can be seen by:

cat /sys/kernel/mm/transparent_hugepage/enabled

When set to madvise NumPy will typically use hugepages for a performance boost. This behaviour can be modified by setting the environment variable:

NUMPY_MADVISE_HUGEPAGE=0

or setting it to 1 to always enable it. When not set, the default is to use madvise on Kernels 4.6 and newer. These kernels presumably experience a large speedup with hugepage support. This flag is checked at import time.

SIMD feature selection

Setting NPY_DISABLE_CPU_FEATURES will exclude simd features at runtime. See Runtime dispatch.

Debugging-related options

Relaxed strides checking

The compile-time environment variable:

NPY_RELAXED_STRIDES_DEBUG=0

can be set to help debug code written in C which iteraters through arrays manually. When an array is contiguous and iterated in a contiguous manner, its strides should not be queried. This option can help find errors where the strides are incorrectly used. For details see the memory layout documentation.

Warn if no memory allocation policy when deallocating data

Some users might pass ownership of the data pointer to the ndarray by setting the OWNDATA flag. If they do this without setting (manually) a memory allocation policy, the default will be to call free. If NUMPY_WARN_IF_NO_MEM_POLICY is set to "1", a RuntimeWarning will be emitted. A better alternative is to use a PyCapsule with a deallocator and set the ndarray.base.

Testing planned future behavior

NumPy has some code paths which are planned to be activated in the future but are not yet the default behavior. You can try testing some of these which may be shipped with a new “major” release (NumPy 2.0) by setting an environment before importing NumPy:

NPY_NUMPY_2_BEHAVIOR=1

By default this will also activate the NEP 50 related setting NPY_PROMOTION_STATE (please see the NEP for details on this).