Installation¶

Calibrated Explanations is published on PyPI and conda-forge. Install the base package first, then add extras that unlock plotting, notebook examples, or contributor tooling.

PyPI¶

pip install calibrated-explanations

Extras are opt-in so you only pull the dependencies you need:

Extra	Purpose	Install command
`viz`	Matplotlib-based plotting and PlotSpec adapters.	`pip install "calibrated-explanations[viz]"`
`notebooks`	Jupyter notebook tutorials with pinned dependencies.	`pip install "calibrated-explanations[notebooks]"`
`dev`	Full development toolchain (linters, docs, tests).	`pip install "calibrated-explanations[dev]"`
`eval`	Benchmarking and evaluation tools.	`pip install "calibrated-explanations[eval]"`
`external-plugins`	Curated optional bundles (e.g., FAST explanations and intervals).	`pip install "calibrated-explanations[external-plugins]"`

conda-forge¶

conda install -c conda-forge calibrated-explanations

If you rely on extras from PyPI inside a conda environment, install the base package via conda and then add the relevant extras with pip.

Verifying your environment¶

python -c "import calibrated_explanations; print(calibrated_explanations.__version__)"

The command should echo 0.10.3 or later once your environment is ready.

Optional contributor fast path¶

For repository development, pip remains the canonical fallback:

python -m pip install -e .[dev] -c constraints.txt

If you already use uv, you can install the editable development environment with the same CI constraints:

uv pip install -e .[dev] -c constraints.txt

This uv command is optional and constraint-based. The project does not treat uv.lock as an authoritative lockfile for v0.11.3.