PyERFA is the Python wrapper for the ERFA library (Essential Routines for Fundamental Astronomy), a C library containing key algorithms for astronomy, which is based on the SOFA library published by the International Astronomical Union (IAU). All C routines are wrapped as Numpy universal functions, so that they can be called with scalar or array inputs.
The project is a split of astropy._erfa module, developed in the
context of Astropy project, into a standalone package. It contains
the ERFA C source code as a git submodule. The wrapping is done
with help of the Jinja2 template engine.
If you use this package in your research, please cita it via DOI 10.5281/zenodo.3940699.
The package can be installed from the package directory using a simple:
$ pip install .
and similarly a wheel can be created with:
$ pip wheel .
Note
If you already have the C library liberfa on your
system, you can use that by setting environment variable
PYERFA_USE_SYSTEM_LIBERFA=1.
The package can be obtained from PyPI or directly from the git repository:
$ git clone --recursive https://github.com/liberfa/pyerfa/
The package also has nightly wheel that can be obtained as follows:
$ pip install --upgrade --index-url https://pypi.anaconda.org/liberfa/simple pyerfa --pre
For testing, one can install the packages together with its testing dependencies and then test it with:
$ pip install .[test] $ pytest
Alternatively, one can use tox, which will set up a separate testing
environment for you, with:
$ tox -e test
The package can be imported as erfa which has all ERFA ufuncs wrapped with
python code that tallies errors and warnings. Also exposed are the constants
defined by ERFA in erfam.h, as well
as numpy.dtype corresponding to structures used by ERFA. Examples:
>>> import erfa
>>> erfa.jd2cal(2460000., [0, 1, 2, 3])
(array([2023, 2023, 2023, 2023], dtype=int32),
array([2, 2, 2, 2], dtype=int32),
array([24, 25, 26, 27], dtype=int32),
array([0.5, 0.5, 0.5, 0.5]))
>>> erfa.plan94(2460000., [0, 1, 2, 3], 1)
array([([ 0.09083713, -0.39041392, -0.21797389], [0.02192341, 0.00705449, 0.00149618]),
([ 0.11260694, -0.38275202, -0.21613731], [0.02160375, 0.00826891, 0.00217806]),
([ 0.13401992, -0.37387798, -0.21361622], [0.0212094 , 0.00947838, 0.00286503]),
([ 0.15500031, -0.36379788, -0.21040601], [0.02073822, 0.01068061, 0.0035561 ])],
dtype={'names': ['p', 'v'], 'formats': [('<f8', (3,)), ('<f8', (3,))], 'offsets': [0, 24], 'itemsize': 48, 'aligned': True})
>>> erfa.dt_pv
dtype([('p', '<f8', (3,)), ('v', '<f8', (3,))], align=True)
>>> erfa.dt_eraLDBODY
dtype([('bm', '<f8'), ('dl', '<f8'), ('pv', [('p', '<f8', (3,)), ('v', '<f8', (3,))])], align=True)
>>> erfa.DAYSEC
86400.0
It is also possible to use the ufuncs directly, though then one has to deal with the warning and error states explicitly. For instance, compare:
>>> erfa.jd2cal(-600000., [0, 1, 2, 3]) Traceback (most recent call last): ... ErfaError: ERFA function "jd2cal" yielded 4 of "unacceptable date (Note 1)" >>> erfa.ufunc.jd2cal(-600000., [0, 1, 2, 3]) (array([-1, -1, -1, -1], dtype=int32), ..., array([-1, -1, -1, -1], dtype=int32))
PyERFA is licensed under a 3-clause BSD style license - see the LICENSE.rst file.
