v0.7.1

Important

We advise all users using v0.7.0 to upgrade to this release as it comes with critical bug fixes.

What's Changed

• Add quantile score by @frazane in #78
• Preserve autograd history in torch backend by @frazane in #79
• Refactor interval scores and improve testing by @frazane in #82
• Bugfix for numba CRPS estimators check by @frazane in #85

Full Changelog: v0.7.0...v0.7.1

v0.7.0

This release adds several new scores. Big thanks to @sallen12 !

What's Changed

• Add log score for laplace distributions by @sallen12 in #63
• Add log scores for beta, exponential and two-piece distributions by @sallen12 in #64
• Add log score for negative binomial distribution by @sallen12 in #65
• Add log score for gev and gpd distributions by @sallen12 in #66
• Add crps and log score for mixture of normal distributions by @sallen12 in #73
• Add log score for ensemble forecasts by @sallen12 in #74
• Add Gaussian kernel score for ensemble forecasts by @sallen12 in #70
• Add ranked probability score by @sallen12 in #71
• Add the log score and ranked logarithmic score by @sallen12 in #72
• Add censored and conditional likelihood scores by @sallen12 in #76
• Add weighted versions of the Gaussian kernel score by @sallen12 in #77

Full Changelog: v0.6.1...v0.7.0

v0.6.1

Fixed missing imports.

Full Changelog: v0.6.0...v0.6.1

v0.6.0

This release focuses on the implementation of closed forms of the CRPS for several new distributions.

What's Changed

• Brier score: Correct order of arguments in pass-through to core score function by @simon-hirsch in #31
• improved examples in crps docs by @nicholasloveday in #33
• Add CRPS for beta distribution by @frazane in #37
• Add CRPS for the exponentialM distribution by @frazane in #39
• Add CRPS for binomial distribution by @frazane in #38
• Add CRPS for the gamma distribution by @frazane in #40
• Add CRPS for GEV distribution by @frazane in #42
• Add CRPS for generalized pareto distribution by @frazane in #43
• Add CRPS for the hypergeometric distribution by @frazane in #44
• Add CRPS for the Laplace distribution by @frazane in #45
• Add CRPS for the log-Laplace distribution by @frazane in #46
• Add CRPS for the loglogistic distribution by @frazane in #47
• Add CRPS for the generalised truncated and censored normal distribution by @sallen12 in #49
• Add CRPS for the uniform distribution by @sallen12 in #50
• Add CRPS for the generalised truncated and censored logistic distribution by @sallen12 in #48
• Add CRPS for the Poisson distribution by @sallen12 in #51
• Add CRPS for Student's t distribution by @sallen12 in #52
• Add CRPS for the generalised truncated and censored t distribution by @sallen12 in #53
• Use uv package management tool by @frazane in #54
• Add CRPS for two piece exponential distribution by @sallen12 in #57
• Update CI workflow by @frazane in #58
• Fix coverage report upload to codecov by @frazane in #59
• Add CRPS for the negative binomial distribution by @sallen12 in #55
• add crps for twopiece normal distribution by @sallen12 in #56
• Add log scores for some parametric distributions by @sallen12 in #61
• Add log score for the truncated logistic, normal and t distributions by @sallen12 in #62
• Wrap numba imports in try-except blocks by @frazane in #69

New Contributors

• @nicholasloveday made their first contribution in #33

Full Changelog: v0.5.3...v0.6.0

v0.5.3

Added support for python 3.12

v0.5.2

This minor release adds another parametric form of the CRPS and a new type of scoring rule: the interval-based Winkler score. Additionally, we start to widen the scope of the library by introducing the first graphical verification tool, the reliability diagram. In the future, we plan to add other graphical tools to diagnose calibration as they are important to complement scoring rules.

New features

• Closed-form CRPS for the exponential distribution in #27 by @sallen12 and @frazane
• CORP reliability diagram in #23 by @frazane
• Interval/Winkler Score in #24 by @simon-hirsch

v0.5.1

New features

• Added quantile-based CRPS in #22 by @simon-hirsch

v0.5.0

This release brings an important breaking change, improvements to the documentation and some preliminary work to include many new distributions for the analytical continuous ranked probability score.

Breaking changes 💥

• Changed the order of positional arguments for all functions in #18 by @frazane. Starting with this release, observations always come as the first positional argument.

New features

• Added special module and several statistical functions to the backends in #17 by @sallen12 and @frazane. This is preparatory work for #13.

Documentation

• General improvements in #12 by @frazane

v0.4.4

What's changed

• Added Error Spread Score by @frazane in #10

v0.4.0

This release brings a new framework-agnostic backend system to scoringrules, with added support for pytorch! This required a major internal refactoring to the codebase but there are no significant changes to the API except for some weighted scoring rules.