popmon v0.4.2

Release notes are available here.

popmon v0.4.1

Release notes are available here.

popmon v0.4.0

Release notes available here.

popmon v0.3.14

• Pin histogrammar version for backwards compatibility

popmon v0.3.13

• Spark 3.0 support (histogrammar update) (#87)
• Improved documentation
• Few minor package improvements

popmon v0.3.12

• Add proper check of matrix invertibility of covariance matrix in stats/numpy.py
• Add support for the Spark date type
• Install Spark on Github Actions to be able to include spark tests in our CI/CD pipeline
• Upgrade linting to use pre-commit (including pyupgrade for python3.6 syntax upgrades)
• Add documentation on how to run popmon using spark on Google Colab (minimal example from scratch)

popmon v0.3.11

Features:

• Traffic light overview (#62)

Documentation:

• Downloads badge readme
• List talks and articles in readme (#66)
• Add image to README.rst (#64)

Other improvements:

• Change notebook testing to pytest-notebook (previously these tests were skipped in CI). Add try-except ImportError for pyspark code. (#67)
• Fix a few typo's
• suppress "matplotlib backend" verbose warning
• click on "popmon report" also scrolls to top
• Update HTML reports using Github Actions (#63)
• Bugfix in hist.py that broke the advanced tutorial.

Notebooks:

• Add %%capture to pip install inside of notebooks.
• Make package install in notebooks work with paths with spaces.
• Pickle doesn't work with tests (not really a popmon-specific feature anyway). Changed the notebook to fix the issue, left the code for reference.

popmon v0.3.10

• Traffic light overview
• Add image to README
• Add building of examples to Github Actions CI
• Format notebooks (nbqa)
• Remove matplotlib backend warning
• Fix navigation in title of report

popmon v0.3.9

• Fix: refactorize Bin creation and fix scipy version for pytestDevelop
• Fix: dataset links in tutorial
• Lint: isort 5, latest black version
• Internal: simplification of weighted mean/std computation

popmon v0.3.8

• Fixing automated PyPi deployment.
• Removing enabling of unnecessary notebook extensions.