Skip to content

Latest commit

 

History

History
121 lines (105 loc) · 4.38 KB

CHANGELOG.md

File metadata and controls

121 lines (105 loc) · 4.38 KB
Raw

v0.7

v0.7.1

Fixes

  • Fix minimum Python requirements in setup.py

v0.7.0

Features

  • Compatibility with pandas v2+ and seaborn above 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:

v0.5.0

  • Add scipy's Brunner-Munzel test
  • Fix applying statannotations for non-string group labels (Issue #65)
  • Get Zenodo DOI

v0.4.5

  • 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

  • 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)
  • Fix custom annotations order with respect to the given pairs (Issue #45)

v0.4.3

  • The correction_format parameter allows changing how the multiple comparisons correction method adjusts the annotation when changing a result to non-significant.
  • Fix the show_test_name configuration.
  • Fix the verbose parameter (PR #37 by mxposed)

v0.4.2

  • Support of FacetGrid with
    • Support empty initialization only defining pairs
    • plot_and_annotate_facets

v0.4.1

  • Support for horizontal orientation

v0.4.0

  • Major refactoring, change to an Annotator class to prepare and add annotations in separate function (now method) calls.
  • Support for violinplot
  • Fixes in rendering of non-significant tests results after multiple comparisons correction
  • Fix the printout of the star annotation legend
  • Fix the failure when providing a dataframe with categories as different columns.
  • Fix a bug where an incorrect xunits calculation resulted in wrong association of points within a box, leading to erroneous max y position for that box.
  • Reduced code complexity, more SOLID.
  • Many unit and integration tests

v0.3

v0.3.2

  • Fix simple format outputs
  • Fix ImportError when applying a multiple comparison correction without statsmodels.
  • Multiple comparison correction is None by default, as statsmodels is an additional dependency.

v0.3.1

  • Added support of functions returning more than the two expected values when used in StatTest
  • Fix version numbers in CHANGELOG

v0.3.0

(Explicitly Requires Python 3.6)

  • Refactoring with implementation of StatTest (#7), removing the stat_func parameter, and test_long_name.
  • Annotations y-positions based on plot coordinates instead of data coordinates (PR #5 by JosephLalli, fixes webermarcolivier/statannot#21).
  • Add this CHANGELOG

v0.2

v0.2.8