MESON_BUILD.md

How to build SciPy with Meson

note: these instructions are for Linux, if you are on another OS things likely don't work yet! macOS support should be easy, please comment if we need to prioritize adding that.

quickstart from scratch

Clone the repo and check out the right branch:

git clone git@github.com:rgommers/scipy.git
git submodule update --init
git checkout meson

Create a conda development environment, build SciPy with Meson and run the test suite:

conda env create -f environment.yml
conda activate scipy-dev
python -m pip install git+https://github.com/rgommers/meson.git@scipy
./mesondev.sh build $PWD/install

Full details and explanation

Now to build SciPy, we need the meson branch from @rgommers's fork (this is the integration branch until Meson support is merged into SciPy master):

git clone git@github.com:rgommers/scipy.git
git submodule update --init
git checkout meson

We will use conda here, because it's the easiest way to get a fully reproducible environment. If you do not have a conda environment yet, the recommended installer is Mambaforge (mamba is basically a much faster conda).

To create a development environment:

conda env create -f environment.yml  # `mamba` works too for this command
conda activate scipy-dev

The one dependency that we now miss is meson. Support for Cython in Meson is very new, and we expect to also need some bug fixes and new features in the Meson master branch as well as fixes we may not even have submitted yet or are pending review. So install Meson from the scipy branch of @rgommers's fork:

python -m pip install git+https://github.com/rgommers/meson.git@scipy

Meson uses a configure and a build stage. To configure it for putting the build artifacts in build/ and a local install under installdir/ and then build:

meson setup build --prefix=$PWD/installdir
ninja -C build

In the command above, -C is followed by the name of the build directory. You can have multiple builds at the same time. Meson is fully out-of-place, so those builds will not interfere with each other. You can for example have a GCC build, a Clang build and a debug build in different directories.

To then install SciPy into the prefix (installdir/ here):

meson install -C build

It will then install to installdir/lib/python3.9/site-packages/scipy, which is not on your Python path, so to add it do:

export PYTHONPATH=$PWD/installdir/lib/python3.9/site-packages/

Now we should be able to import scipy and run the tests. Remembering that we need to move out of the root of the repo to ensure we pick up the package and not the local scipy/ source directory:

cd doc
python -c "from scipy import constants as s; s.test()"

The above runs the tests for a single module, constants. Other ways of running the tests should also work, for example:

pytest scipy

Current status (24 July) is that the full test suite passes on Linux with OpenBLAS. And there is CI on this fork to keep it that way. Other platforms are not tested yet. The current status is already good enough to work on both build related issues (e.g. build warnings, debugging some C/C++ extension) and on general SciPy development. It is already a much smoother than working with the default distutils-based build one gets with python setup.py develop.

The above configure-build-install-test docs are useful to understand and for working on build improvements (you don't need to install for that to, for example, see the build succeeds and a build warning has disappeared); if you want the "all-in-one" command for the above, run:

./mesondev.sh build --prefix=$PWD/installdir

It's worth pointing out that Meson has very good documentation; it's worth reading and is often the best source of answers for "how to do X".