This PR updates the setup script to use setuptools in place of the older distutils, which on my test machine helped smooth out compilation on the new ARM-based Macs.
Together with PR #158, this brings full support to the dev branch for compilation on these machines. For initial writeup see issue #152.
(Note that it's not a problem to run older x86 Python packages, including Pandana, on ARM Macs. These updates are just to support native compilation, which will improve performance.)
Install all of Pandana's build- and runtime requirements from Conda. Check the install messages to confirm you're getting osx-arm64 binaries. (If it's the normal osx-64 binaries instead, you're probably running an x86 copy of Conda.)
This PR updates the setup script to use
setuptools
in place of the olderdistutils
, which on my test machine helped smooth out compilation on the new ARM-based Macs.Together with PR #158, this brings full support to the
dev
branch for compilation on these machines. For initial writeup see issue #152.(Note that it's not a problem to run older x86 Python packages, including Pandana, on ARM Macs. These updates are just to support native compilation, which will improve performance.)
How to compile natively on ARM Macs
Install MiniForge for
osx-arm64
(read, download).Install all of Pandana's build- and runtime requirements from Conda. Check the install messages to confirm you're getting
osx-arm64
binaries. (If it's the normalosx-64
binaries instead, you're probably running an x86 copy of Conda.)conda install clang llvm-openmp cython numpy pandas scikit-learn pytables
Compile and install Pandana.
python setup.py develop
Next steps to provide ARM binaries on Conda Forge
After that's completed, you'll be able to install Pandana for
osx-arm64
directly from Conda.