Spin-Spherical Harmonic Transforms jit-compiled with numba
There is a single external requirement that will not be installed automatically,
and that is FFTW. This can usually be installed via your package manager,
but take care to install the shared library. If compiling yourself, that means
using the option --enable-shared.
Then, installation should be as simple as:
pip install .
from the top-level directory, or:
pip install git+git://github.com/UPennEoR/ssht_numba
If you clone this repo to install manually, you must also grab the submodules with:
git submodule update --init --recursive
If you installed FFTW to a non-default location, then you can point to its
location using the environment variable FFTW_PATH, which should be the path
to the lib folder, eg.: FFTW_PATH=/usr/lib pip install ..
Other dependencies of ssht_numba are installed automatically. However, if you
are using conda and would like these packages to be installed with conda
rather than pip, you may want to do conda install numpy numba cffi prior
to installing the package.
To install a development environment (with all required packages for testing,
creating documentation and linting), install with pip install -e .[dev], or
if you would just like to run tests, use pip install -e .[testing]. At this
time, an extra package is also required for testing: pyssht, which at this
time must be installed by downloading https://github.com/UPennEoR/ssht and doing
python setup.py install (not pip).
This package has pre-commit hooks set up. If developing this package, please
pip install pre-commit and pre-commit install.
This package uses pyscaffold, which in turn uses setuptools_scm, and manages
versioning via Git tags. Versioning should use https://semver.org semantic versioning
rules.
This project has been set up using PyScaffold 3.2.1. For details and usage information on PyScaffold see https://pyscaffold.org/.