raphaelvallat/pingouin

📦 Summary

• Minor release with one new feature, several bugfixes, and internal improvements.

✨ New features

• `compute_effsize`: added `eftype='cohen_dz'` for paired-samples designs ($d_z = \bar{X-Y} / \sigma_{X-Y}$); `y` now also accepts a scalar for one-sample effect sizes (#508)

📦 Improvements

• `compute_bootci`: replaced custom bootstrap implementation with `scipy.stats.bootstrap`; default CI method upgraded to BCa; minimum SciPy bumped to 1.10 (#505)
• `intraclass_corr`: updated ICC type labels in output dataframe and documentation (#501)

🐛 Bugfixes

• `partial_corr` / `pcorr`: fixed numerical instability when variables differ by many orders of magnitude (#510)
• `partial_corr`: raise `ValueError` on identical covariates; warn on rank-deficient covariance matrix (#500)
• `bayesfactor_pearson`: fixed catastrophic float64 cancellation in one-sided tests for strongly negative `r` (#503)
• `logistic_regression`: fixed compatibility with scikit-learn >= 1.8 (#504)

💥 Breaking changes

• Removed `plot_shift` function (#502)
• ***

📋 What's Changed

• Fix broken links in doc; add uv install instructions; use uv in ruff workflow; do not trigger pytest/coverage workflow for doc-only PRs by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/496
• Use relative imports for intra-package dependencies by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/499
• fix(partial_corr): raise on identical covariates, warn on rank-defici… by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/500
• Update ICC types in output dataframe and documentation by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/501
• Fix catastrophic float64 cancellation in `bayesfactor_pearson` one-si… by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/503
• Fix test logistic regression by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/504
• Remove plot_shift function by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/502
• Add pre-commit hooks and consolidate dev dependencies by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/506
• + 5 more

📋 Changes

• Disable one-sided Bayes Factor for T-tests, which were ill-defined ([PR487](https://github.com/raphaelvallat/pingouin/pull/487))
• Update RBC calculation for Wilcoxon signed-rank test to be dependent on the alternative ([PR457](https://github.com/raphaelvallat/pingouin/pull/457))
• Sphericity fix with very low eigenvalues ([PR482](https://github.com/raphaelvallat/pingouin/pull/482))
• Fix divide-by-zero in internal _correl_pvalue when r == 1 ([PR474](https://github.com/raphaelvallat/pingouin/pull/474)
• Fix boxplot z-order in [pingouin.plot_paired()](https://pingouin-stats.org/generated/pingouin.plot_paired.html#pingouin.plot_paired) ([PR442](https://github.com/raphaelvallat/pingouin/pull/442))
• Column names update ([PR443](https://github.com/raphaelvallat/pingouin/pull/443)): removed characters that restrict column access to the bracket format (df[“p_val”]) rather than dot method (df.p_val). This includes:
• Replaced dashes with underscores in column names (e.g., p-val –> p_val)
• Replaced parentheses with underscores in column names (e.g., mean(A) –> mean_A)
• + 9 more

📋 What's Changed

• Fix boxplot z-order in `plot_paired` by @sbwiecko in https://github.com/raphaelvallat/pingouin/pull/442
• Fix the Github Action CI for Python tests by @remrama in https://github.com/raphaelvallat/pingouin/pull/445
• Update parametric.py - fix typo by @Petemir in https://github.com/raphaelvallat/pingouin/pull/448
• Clean dependencies; add extras feature by @getzze in https://github.com/raphaelvallat/pingouin/pull/451
• Pandas-friendly column names by @remrama in https://github.com/raphaelvallat/pingouin/pull/443
• replaces black/flake8 formatting/linting with ruff and ensures numpy 2.0 compatibility by @remrama in https://github.com/raphaelvallat/pingouin/pull/446
• Updated deprecated Seaborn function by @sjg2203 in https://github.com/raphaelvallat/pingouin/pull/459
• Fix ruff + CI by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/460
• + 11 more

✨ New Contributors

• @sbwiecko made their first contribution in https://github.com/raphaelvallat/pingouin/pull/442
• @Petemir made their first contribution in https://github.com/raphaelvallat/pingouin/pull/448
• @rhazn made their first contribution in https://github.com/raphaelvallat/pingouin/pull/457
• @AlexanderJCS made their first contribution in https://github.com/raphaelvallat/pingouin/pull/474
• Full Changelog: https://github.com/raphaelvallat/pingouin/compare/v0.5.5...v0.6.0

📋 What's Changed

• Fix penalty for LogisticRegression by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/403
• Switch to modern python packaging by @getzze in https://github.com/raphaelvallat/pingouin/pull/406
• Remove call to sns.despine by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/410
• Updated deprecated function by @sjg2203 in https://github.com/raphaelvallat/pingouin/pull/414
• Add errstate(divide="ignore") in Bayes Factor calculation by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/415
• Remove inplace on single column by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/423
• Fix RBC sign in mwu by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/424
• Overhaul documentation (pydata_sphinx_theme) by @yann1cks in https://github.com/raphaelvallat/pingouin/pull/432
• + 1 more

✨ New Contributors

• @getzze made their first contribution in https://github.com/raphaelvallat/pingouin/pull/406
• @sjg2203 made their first contribution in https://github.com/raphaelvallat/pingouin/pull/414
• @yann1cks made their first contribution in https://github.com/raphaelvallat/pingouin/pull/432
• Full Changelog: https://github.com/raphaelvallat/pingouin/compare/v0.5.4...v0.5.5

📋 What's Changed

• Minor typo fix in docs by @musicinmybrain in https://github.com/raphaelvallat/pingouin/pull/329
• clip r values by @remrama in https://github.com/raphaelvallat/pingouin/pull/342
• fix: deprecated parameter by @bitsnaps in https://github.com/raphaelvallat/pingouin/pull/341
• hotfix: CI crash in test_power_chi2 [WIP] by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/344
• hotfix: plot_rm_corr crash with specific column names by @remrama in https://github.com/raphaelvallat/pingouin/pull/351
• Add check for noncentrality parameters. by @agkphysics in https://github.com/raphaelvallat/pingouin/pull/347
• Use pyupgrade by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/364
• fix groupby.mean for only numeric values by @jajcayn in https://github.com/raphaelvallat/pingouin/pull/363
• + 11 more

✨ New Contributors

• @musicinmybrain made their first contribution in https://github.com/raphaelvallat/pingouin/pull/329
• @bitsnaps made their first contribution in https://github.com/raphaelvallat/pingouin/pull/341
• @agkphysics made their first contribution in https://github.com/raphaelvallat/pingouin/pull/347
• @jajcayn made their first contribution in https://github.com/raphaelvallat/pingouin/pull/363
• @kraktus made their first contribution in https://github.com/raphaelvallat/pingouin/pull/382
• Full Changelog: https://github.com/raphaelvallat/pingouin/compare/v0.5.3...v0.5.4

📋 What's Changed

• Fix numerical stability issue in multivariate_normality by @gkanwar in https://github.com/raphaelvallat/pingouin/pull/292
• Add new function for pairwise T-tests between columns of a dataframe (pingouin.ptests) by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/291
• Handle single-sample comparsion in pairwise_test by @George3d6 in https://github.com/raphaelvallat/pingouin/pull/299
• Change TestRegression class test methods to fix victim flakiness by @blazyy in https://github.com/raphaelvallat/pingouin/pull/303
• Add aesthetic flexibility to plot_rm_corr by @remrama in https://github.com/raphaelvallat/pingouin/pull/312
• Update distribution.py by @ALL-SPACE-Rob in https://github.com/raphaelvallat/pingouin/pull/310
• Plotting seaborn.FacetGrid compatibility by @remrama in https://github.com/raphaelvallat/pingouin/pull/314
• Use scikit-learn>=1.1.2 by @raphaelvallat in https://github.com/raphaelvallat/pingouin/pull/300
• + 5 more

✨ New Contributors

• @gkanwar made their first contribution in https://github.com/raphaelvallat/pingouin/pull/292
• @George3d6 made their first contribution in https://github.com/raphaelvallat/pingouin/pull/299
• @blazyy made their first contribution in https://github.com/raphaelvallat/pingouin/pull/303
• @remrama made their first contribution in https://github.com/raphaelvallat/pingouin/pull/312
• @ALL-SPACE-Rob made their first contribution in https://github.com/raphaelvallat/pingouin/pull/310
• @turkalpmd made their first contribution in https://github.com/raphaelvallat/pingouin/pull/320

📦 Pingouin 0.5.1

• This is a minor release, with several bugfixes and improvements. This release is compatible with SciPy 1.8 and Pandas 1.4.
• Bugfixes
• Added support for SciPy 1.8 and Pandas 1.4. https://github.com/raphaelvallat/pingouin/pull/234
• Fixed bug where [pingouin.rm_anova()](https://pingouin-stats.org/generated/pingouin.rm_anova.html#pingouin.rm_anova) and [pingouin.mixed_anova()](https://pingouin-stats.org/generated/pingouin.mixed_anova.html#pingouin.mixed_anova) changed the dtypes of categorical columns in-place https://github.com/raphaelvallat/pingouin/issues/224
• Enhancements
• Faster implementation of [pingouin.gzscore()](https://pingouin-stats.org/generated/pingouin.gzscore.html#pingouin.gzscore), adding all options available in zscore: axis, ddof and nan_policy. Warning: this function is deprecated and will be removed in the next version of Pingouin (use [scipy.stats.gzscore()](https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.gzscore.html#scipy.stats.gzscore) instead). https://github.com/raphaelvallat/pingouin/pull/210.
• Replace use of statsmodels’ studentized range distribution functions with more SciPy’s more accurate scipy.stats.studentized_range(). https://github.com/raphaelvallat/pingouin/pull/229.
• Add support for optional keywords argument in the [pingouin.homoscedasticity()](https://pingouin-stats.org/generated/pingouin.homoscedasticity.html#pingouin.homoscedasticity) function https://github.com/raphaelvallat/pingouin/issues/218
• + 1 more

📋 Changes

• Important bugfix in pingouin.ttest() in which the 95% confidence intervals for one-sample T-test with `y` != 0 were invalid.
• Added an "options" module to control global rounding/display behavior.
• Several enhancements / new features in existing functions.

📋 Changes

• Fixed a bug in pingouin.pairwise_corr caused by the deprecation of ``pandas.core.index`` in the new version of Pandas (1.0). For now, both Pandas 0.25 and Pandas 1.0 are supported.
• The standard deviation in pingouin.pairwise_ttests when using ``return_desc=True`` is now calculated with ``np.nanstd(ddof=1)`` to be consistent with Pingouin/Pandas default unbiased standard deviation.
• Added the pingouin.plot_circmean function to plot the circular mean and circular vector length of a set of angles (in radians) on the unit circle. Note that this function is still in beta and some parameters may change without warnings in the next releases.

📋 Changes

• MAJOR: Fixed a bug in [pingouin.pairwise_ttests()](https://pingouin-stats.org/generated/pingouin.pairwise_ttests.html#pingouin.pairwise_ttests) when using mixed or two-way repeated measures design. Specifically, the T-tests were performed without averaging over repeated measurements first (i.e. without calculating the marginal means). Note that for mixed design, this only impacts the between-subject T-test(s). Practically speaking, this led to higher degrees of freedom (because they were conflated with the number of repeated measurements) and ultimately incorrect T and p-values because the assumption of independence was violated. Pingouin now averages over repeated measurements in mixed and two-way repeated measures design, which is the same behavior as [JASP](https://jasp-stats.org/) or [JAMOVI](https://www.jamovi.org/). As a consequence, and when the data has only two groups, the between-subject p-value of the pairwise T-test should be (almost) equal to the p-value of the same factor in the [pingouin.mixed_anova()](https://pingouin-stats.org/generated/pingouin.mixed_anova.html#pingouin.mixed_anova) function. The old behavior of Pingouin can still be obtained using the ``marginal=False`` argument.
• Minor: Added a check in [pingouin.mixed_anova()](https://pingouin-stats.org/generated/pingouin.mixed_anova.html#pingouin.mixed_anova) to ensure that the ``subject`` variable has a unique set of values for each between-subject group defined in the ``between`` variable. For instance, the subject IDs for group1 are [1, 2, 3, 4, 5] and for group2 [6, 7, 8, 9, 10]. The function will throw an error if there are one or more overlapping subject IDs between groups (e.g. the subject IDs for group1 AND group2 are both [1, 2, 3, 4, 5]).
• Minor: Fixed a bug which caused the [pingouin.plot_rm_corr()](https://pingouin-stats.org/generated/pingouin.plot_rm_corr.html#pingouin.plot_rm_corr) and [pingouin.ancova()](https://pingouin-stats.org/generated/pingouin.ancova.html#pingouin.ancova) (with >1 covariates) to throw an error if any of the input variables started with a number (because of statsmodels / [Patsy formula formatting](https://patsy.readthedocs.io/en/latest/builtins-reference.html)).
• Upon loading, Pingouin will now use the [outdated](https://github.com/alexmojaki/outdated) package to check and warn the user if a newer stable version is available.
• Globally removed the ``export_filename`` parameter, which allowed to export the output table to a .csv file. This helps simplify the API and testing. As an alternative, one can simply use [pandas.to_csv()](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_csv.html) to export the output dataframe generated by Pingouin.
• Added the ``correction`` argument to [pingouin.pairwise_ttests()](https://pingouin-stats.org/generated/pingouin.pairwise_ttests.html#pingouin.pairwise_ttests) to enable or disable Welch’s correction for independent T-tests.

📋 Changes

• Fixed a bug in which missing values were removed from all columns in the dataframe in [pingouin.kruskal()](https://pingouin-stats.org/generated/pingouin.kruskal.html#pingouin.kruskal), even columns that were unrelated. See https://github.com/raphaelvallat/pingouin/issues/74.
• The [pingouin.power_corr()](https://pingouin-stats.org/generated/pingouin.power_corr.html#pingouin.power_corr) function now throws a warning and return a np.nan when the sample size is too low (and not an error like in previous version). This is to improve compatibility with the pingouin.pairwise_corr() function.
• Fixed quantile direction in the [pingouin.plot_shift()](https://pingouin-stats.org/generated/pingouin.plot_shift.html#pingouin.plot_shift) function. In v0.3.0, the quantile subplot was incorrectly labelled as Y - X, but it was in fact calculating X - Y. See https://github.com/raphaelvallat/pingouin/issues/73

📋 Changes

• Added [pingouin.plot_rm_corr()](https://pingouin-stats.org/generated/pingouin.plot_rm_corr.html#pingouin.plot_rm_corr) to plot a repeated measures correlation
• Added the `relimp` argument to [pingouin.linear_regression()](https://pingouin-stats.org/generated/pingouin.linear_regression.html#pingouin.linear_regression) to return the relative importance (= contribution) of each individual predictor to the R^2 of the full model.
• Complete refactoring of [pingouin.intraclass_corr()](https://pingouin-stats.org/generated/pingouin.intraclass_corr.html#pingouin.intraclass_corr) to closely match the R implementation in the [psych](https://cran.r-project.org/web/packages/psych/psych.pdf) package. Pingouin now returns the 6 types of ICC, together with F values, p-values, degrees of freedom and confidence intervals.
• The [pingouin.plot_shift()](https://pingouin-stats.org/generated/pingouin.plot_shift.html#pingouin.plot_shift) now 1) uses the Harrel-Davis robust quantile estimator in conjunction with a bias-corrected bootstrap confidence intervals, and 2) support paired samples.
• Added the axis argument to [pingouin.harrelldavis()](https://pingouin-stats.org/generated/pingouin.harrelldavis.html#pingouin.harrelldavis) to support 2D arrays.

📋 Changes

• Fixed ERROR in two-sided p-value for Wilcoxon test (`pingouin.wilcoxon()`), the p-values were accidentally squared, and therefore smaller. Make sure to always use the latest release of Pingouin.
• `pingouin.wilcoxon()` now uses the continuity correction by default (the documentation was saying that the correction was applied but it was not applied in the code.)
• The show_median argument of the `pingouin.plot_shift()` function was not working properly when the percentiles were different that the default parameters.
• The current release of statsmodels (0.9.0) is not compatible with the newest release of Scipy (1.3.0). In order to avoid compatibility issues in the `pingouin.ancova()` and `pingouin.anova()` functions (which rely on statsmodels for certain cases), Pingouin will require SciPy < 1.3.0 until a new stable version of statsmodels is released.
• Added `pingouin.chi2_independence()` tests.
• Added `pingouin.chi2_mcnemar()` tests.
• Added `pingouin.power_chi2()` function.
• Added `pingouin.bayesfactor_binom()` function.
• + 9 more

📋 Changes

• Fixed error in p-values for one-sample one-sided T-test (pingouin.ttest()), the two-sided p-value was divided by 4 and not by 2, resulting in inaccurate (smaller) one-sided p-values.
• Fixed global error for unbalanced two-way ANOVA (pingouin.anova()), the sums of squares were wrong, and as a consequence so were the F and p-values. In case of unbalanced design, Pingouin now computes a type II sums of squares via a call to the statsmodels package.
• The epsilon factor for the interaction term in two-way repeated measures ANOVA (pingouin.rm_anova()) is now computed using the lower bound approach. This is more conservative than the Greenhouse-Geisser approach and therefore give (slightly) higher p-values. The reason for choosing this is that the Greenhouse-Geisser values for the interaction term differ than the ones returned by R and JASP. This will be hopefully fixed in future releases.
• Added pingouin.multivariate_ttest() (Hotelling T-squared) test.
• Added pingouin.cronbach_alpha() function.
• Added pingouin.plot_shift() function.
• Several functions of pandas can now be directly used as pandas.DataFrame methods.
• Added pingouin.pcorr() method to compute the partial Pearson correlation matrix of a pandas.DataFrame (similar to the pcor function in the ppcor package).
• + 17 more

📋 Changes

• Added pingouin.distance_corr() (distance correlation) function.
• pingouin.rm_corr() now requires at least 3 unique subjects (same behavior as the original R package).
• The pingouin.pairwise_corr() is faster and returns the number of outlier if a robust correlation is used.
• Added support for 2D level in the pingouin.pairwise_corr(). See Jupyter notebooks for examples.
• Added support for partial correlation in the pingouin.pairwise_corr() function.
• Greatly improved execution speed of pingouin.correlation.skipped() function.
• Added default random state to compute the Min Covariance Determinant in the pingouin.correlation.skipped() function.
• The default number of bootstrap samples for the pingouin.correlation.shepherd() function is now set to 200 (previously 2000) to increase computation speed.
• + 18 more