A Python 3 client for AzureMl web services.


Keywords
Azure, Machine, Learning, AzureML, web, services, client, network, json, azure-machine-learning, batch, blob, request, storage
License
BSD-3-Clause
Install
pip install azmlclient==2.6.0

Documentation

python-azmlclient

An unofficial generic client stack for azureML web services, working with both python 2 and 3.

Python versions Build Status Tests Status codecov

Documentation PyPI Downloads Downloads per week GitHub stars

This is the readme for developers. The documentation for users is available here: https://smarie.github.io/python-azureml-client/

Want to contribute ?

Contributions are welcome ! Simply fork this project on github, commit your contributions, and create pull requests.

Here is a non-exhaustive list of interesting open topics: https://github.com/smarie/python-azureml-client/issues

Installing all requirements

In order to install all requirements, including those for tests and packaging, use the following command:

pip install -r ci_tools/requirements-pip.txt

Running the tests

This project uses pytest.

pytest -v azmlclient/tests/

Packaging

This project uses setuptools_scm to synchronise the version number. Therefore the following command should be used for development snapshots as well as official releases:

python setup.py egg_info bdist_wheel rotate -m.whl -k3

Generating the documentation page

This project uses mkdocs to generate its documentation page. Therefore building a local copy of the doc page may be done using:

mkdocs build -f docs/mkdocs.yml

Generating the test reports

The following commands generate the html test report and the associated badge.

pytest --junitxml=junit.xml -v azmlclient/tests/
ant -f ci_tools/generate-junit-html.xml
python ci_tools/generate-junit-badge.py

PyPI Releasing memo

This project is now automatically deployed to PyPI when a tag is created. Anyway, for manual deployment we can use:

twine upload dist/* -r pypitest
twine upload dist/*

Merging pull requests with edits - memo

Ax explained in github ('get commandline instructions'):

git checkout -b <git_name>-<feature_branch> master
git pull https://github.com/<git_name>/python-azureml-client.git <feature_branch> --no-commit --ff-only

if the second step does not work, do a normal auto-merge (do not use rebase!):

git pull https://github.com/<git_name>/python-azureml-client.git <feature_branch> --no-commit

Finally review the changes, possibly perform some modifications, and commit.