Installation

Pre-built packages

Pre-built binaries of hdf5plugin are available from:

• pypi, to install run: pip install hdf5plugin [--user]

• conda-forge, to install run: conda install -c conda-forge hdf5plugin

To maximize compatibility, those binaries are built without optimization options (such as AVX2 and OpenMP). Installation from source can achieve better performances than pre-built binaries.

Installation from source

The build process enables compilation optimizations that are supported by the host machine.

To install from source and recompile the HDF5 plugins, run:

pip install hdf5plugin --no-binary hdf5plugin [--user]

To override the defaults that are probed from the machine, it is possible to specify build options by setting environment variables, for example:

• HDF5PLUGIN_OPENMP=False pip install hdf5plugin --no-binary hdf5plugin

• From the source directory: HDF5PLUGIN_OPENMP=False pip install .

Available options

Environment variable	Description
HDF5PLUGIN_HDF5_DIR	Custom path to HDF5 (as in h5py). It must contain a “include” subfolder and on Windows a “lib” subfolder with hdf5.lib.
HDF5PLUGIN_OPENMP	Whether or not to compile with OpenMP. Default: True if probed (always False on macOS).
HDF5PLUGIN_NATIVE	True to compile specifically for the host, False for generic support (For unix compilers only). Default: True on supported architectures, False otherwise
HDF5PLUGIN_SSE2	Whether or not to compile with SSE2 support. Default: True on ppc64le and when probed on x86, False otherwise
HDF5PLUGIN_SSSE3	Whether or not to compile with SSSE3 support. Default: True when probed on x86, False otherwise
HDF5PLUGIN_AVX2	Whether or not to compile with AVX2 support. It requires enabling SSE2 support. Default: True on x86 when probed, False otherwise
HDF5PLUGIN_AVX512	Whether or not to compile with AVX512 Foundation (F) and Byte and Word (BW) instruction sets support. It requires enabling SSE2 and AVX2 support. Default: True on x86 when probed, False otherwise
HDF5PLUGIN_BMI2	Whether or not to compile Zstandard with BMI2 support if available. Default: True on Linux and MacOS, False otherwise
HDF5PLUGIN_CPP11	Whether or not to compile C++11 code if available. Default: True if probed.
HDF5PLUGIN_CPP14	Whether or not to compile C++14 code if available. Default: True if probed.
HDF5PLUGIN_CPP20	Whether or not to compile C++20 code if available. Default: True if probed.
HDF5PLUGIN_INTEL_IPP_DIR	Experimental feature: Path to Intel IPP root directory. If provided, the LZ4 compression library and Blosc2 are set to use Intel IPP. Default: Do not use INTEL IPP libraries.

Note: Boolean options are passed as True or False.