The preferred method of making contributions is to fork + pull request from the main repo.
After cloning your fork, do:
cd surfinBH
git submodule init
git submodule updateBefore doing a pull request, you should check that your changes don't break
anything by running py.test from the root directory of your check-out. Every
pull request will be automatically tested by github.
The fits in this package have the naming format: NRSur*Remnant
Let's say your
fancy new fit has fit_name = '23dModGR', the name to load and evaluate the
package would be 'NRSur23dModGRRemnant'.
You need to do the following to add this fit to this package:
- Add
fit_23dModGR.pyinsurfinBH/_fit_evaluators/; seesurfinBH/_fit_evaluators/fit_3dq8.pyfor an example. - Add
from fit_23dModGR import Fit23dModGRtosurfinBH/_fit_evaluators/__init__.py. - Add
'NRSur23dModGRRemnant'key tofits_collectioninsurfinBH/_loadFits.py. See example forsurfinBH3dq8in the same file (surfinBH*was the old format). - Add
example_23dModGR.ipynbinexamples. Seeexamples/example_3dq8.ipynbfor an (cough, cough) example. - Generate regression data for testing using
test/generate_regression_data.py. Run it from the root directory of your check-out. - Test (see above), commit and push all of the above changes.
Note: Do not push the fit data itself, but push the regression data generated
in test/regression_data/.
Note: This is currently under Vijay's account, so only he can do this. First make sure :
-
There are no data files in surfinBH/data. The required data files will get downloaded automatically.
-
surfinBH/_eval_pysuris not empty. If empty, dogit submodule update --init ..
Then:
python setup.py sdist bdist_wheel
twine upload dist/*NOTE for the uploader: For the last step, you should have the following in your ~/.pypirc:
[distutils]
index-servers =
pypi
[pypi]
username = __token__
When asked for the password, get a fresh API token https://pypi.org/manage/account/.