Release Notes

Installation

You can install PyKrige with conda:

conda install -c conda-forge pykrige

or with pip:

pip install pykrige

The documentation can be found at: https://pykrige.readthedocs.io/

Changes

add wheels for Python 3.13 and 3.14 as well as for Linux aarch64
pin the earliest supported scikit-learn release alongside NumPy/SciPy in the sdist CI
allow the housing-based classification/regression tests to attempt loading the California housing dataset and skip cleanly when the download fails

Bug fixes

prevent ValueError from NumPy when pseudo inverse kriging statistics encounter duplicate coordinates by keeping distance vectors 1-D