Pingouin review
A focused Python library for common statistical tests, models, and effect-size work.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
Pingouin is an open-source statistical package for Python 3, built around Pandas and NumPy. It is intended for researchers and Python users who want higher-level statistical procedures with tabular outputs inside Python, IPython, or Jupyter. Solo practitioners, research teams, and organizations that self-host Python tools can use it for hypothesis testing, modeling, effect-size work, and reporting support.
Its main strength is the range of procedures available in one library. Pingouin covers N-way, repeated-measures, mixed, and ANCOVA procedures; parametric and non-parametric pairwise post-hoc tests; several correlation methods; linear and binary logistic regression; mediation analysis with bootstrap methods; and Bayes-factor calculations for binomial tests, t-tests, and Pearson correlation. Effect-size, confidence-interval, and statistical-power calculations extend the analysis beyond significance testing. Circular statistics, chi-squared tests, reliability measures, multivariate tests, and plotting helpers such as Bland–Altman, Q-Q, and paired plots add useful supporting depth.
Pingouin fits teams that already work in Python and want focused statistical routines rather than a separate desktop or hosted environment. Installation is local through pip, uv, or conda, with community and documentation support. Data workflows can use CSV and XLSX files through Pandas, while scripting support makes repeated analyses practical. The tradeoff is scope and delivery model: Pingouin is a library, not an end-to-end analysis suite with its own hosted workspace or graphical interface. Choose it when Python integration and statistical breadth matter most; consider another product when you need a standalone application, managed deployment, or a broader visual analytics environment.
Pingouin pros and cons
- Where it wins
- Broad coverage of ANOVA, correlations, regression, mediation, and Bayes factors
- Effect sizes, confidence intervals, and power calculations accompany tests
- Plotting helpers and tabular outputs fit Pandas-based research workflows
- Where it doesn't
- A Python library rather than a hosted statistical application
- Requires local installation and a Python, IPython, or Jupyter workflow
- Narrower scope than a complete analysis suite with a graphical interface
Pingouin fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update