Get started
279 commits | Last commit ≈ 2 weeks ago
calibr8 is a Python toolbox for likelihood-based calibration modeling and Bayesian inference. It supports uni- and multivariate problems, continuous & discrete distributions and includes ready to use models for many non-linear problems. It integrates with PyMC for advanced Bayesian inference.
calibr8 helps you to easily make sense of your non-linear calibration problem without deep-diving into error propagation.
It enables you to model non-linear trends and observation noise in your calibration dataset at the same time, and correctly accounts for both when performing inference from new data.
For a quick introduction start here: https://calibr8.readthedocs.io
And for a thorough understanding here: https://doi.org/10.1371/journal.pcbi.1009223