Releases: trevismd/statannotations
Releases · trevismd/statannotations
v0.7.1
v0.7.1
Fixes
- Fix minimum Python version (3.8) in setup.py 🙏 @getzze
v0.7
v0.7
v0.7.0
Features
- Compatibility with pandas v2+ and seaborn v0.12+ (PR #155 by getzze) 🚀
v0.6
v0.6.0
Features
- Add option to skip annotation of non-significant results (PR #95 by sepro)
Fixes
- Fix keeping annotation with reduced ylim (PR #116 by amkorb)
- Fix pvalue legend (usually for NS range)
Additional testing and documentation:
v0.5
- Add scipy's Brunner-Munzel test
- Fix applying statannotations for non-string group labels (Issue #65)
- Get Zenodo DOI
v0.4.5 - More fixes
- Add MANIFEST.IN (PR #56 by Matt Chan)
- Limit supported Seaborn version to v.0.11.x
- Fix adding annotations with hue if data is passed as arrays (PR #64 by getzze)
v0.4.4 - Fixes and doc
- Fix custom annotations order with respect to the given pairs (Issue #45)
- The label for Kruskal-Wallis test explicitly states that it is run pairwise (PR #40 by sepro)
- Fix broken link in readme (PR #43 by orena1)
v0.4.3 Format correction
What's Changed
- Choose how to format the annotation when the p-value is no longer significant after correction by @trevismd in #31
- Set verbose flag from parameters by @mxposed in #37
New Contributors
Full Changelog: v0.4.2...v0.4.3
FacetGrid
- Support of
FacetGrid
with
- Support empty initialization only defining pairs
plot_and_annotate_facets
Horizontal orientation
Support for horizontally oriented plots !
API change - Robustness
- New API separating calls to generate the annotations and to annotate the plot.
- More robust support for
barplot
, violinplot
, swarmplot
, stripplot
.
- Decreased complexity and increased maintainability of code base.
- Fixes
- See the changelog for more details.