The only requirement for pygram11 is NumPy. All installation methods
will ensure that
numpy is installed. If you install binaries from
conda-forge or PyPI, OpenMP acceleration is enabled.
Extras for Source Builds¶
When building from source, all you need is a C++ compiler with C++11
setup.py script will test to see if OpenMP is
available during compilation and linking of the backend extenstion
module. Most Linux distributions with relatively modern GCC versions
should provide OpenMP automatically (search the web to see how to
install OpenMP from your distribution’s package manager). On macOS
you’ll want to install
libomp from Homebrew to use OpenMP with the
Clang compiler shipped with macOS.
You can use pygram11 without OpenMP, but you might want to try fast-histogram if you just need to compute fixed bin histograms (see the benchmarks). If you’re here for variable width histograms or the sum-of-weights-squared first class citizenry or the multiple weight variation histograms - I think you’ll still find pygram11 useful.
$ pip install pygram11
Installations from conda-forge provide a build that used OpenMP.
$ conda install pygram11 -c conda-forge
On macOS the OpenMP libraries from LLVM (
libomp) and Intel
libiomp) can clash if your
conda environment includes the
Intel Math Kernel Library (MKL) package distributed by
Anaconda. You may need to install the
nomkl package to prevent
the clash (Intel MKL accelerates many linear algebra operations,
but does not impact pygram11):
$ pip install git+https://github.com/douglasdavis/pygram11.git@master
If installing from source on macOS 10.14 you might have to prepend
pip command with
may be required for Anaconda environments, because the Python
distribution from Anaconda for macOS is built with an older macOS
$ MACOSX_DEPLOYMENT_TARGET=10.14 pip install ...