staplot

Statistical analysis and publication-quality plots for biologists: a free, open-source alternative to GraphPad Prism, with a desktop app and a small Python API.

Features

• Data tables Prism-style (one column per group) or long format; CSV and Excel import (semicolon / decimal-comma CSVs included).
• Tests: t-tests (Welch, paired), Mann-Whitney, Wilcoxon, one- and two-way ANOVA, repeated measures, Kruskal-Wallis, post-hoc tests (Tukey, Dunn, ...), chi-square, correlation, linear and nonlinear regression (curve fitting), normality tests, outliers, ROC, survival, power analysis.
• Plots: bar, box, violin, scatter, grouped bar, before/after, histogram, QQ, residuals, Bland-Altman, heatmap, volcano, ROC and survival curves, with significance brackets added automatically.
• A test wizard that picks the right test for your data and explains why.

Install

pip install staplot            # add [excel] for .xlsx import

Python 3.10 or newer.

Use

Desktop app:

staplot

From Python (no GUI needed, safe to run headless):

from staplot import api

table = api.table_from_csv("counts.csv")       # one column per group
stats = api.auto_test(table)                   # picks the test, returns p-values
fig, record = api.plot(table, kind="box")      # matplotlib Figure with brackets
fig.savefig("counts.pdf")

auto_test checks normality per group (Shapiro-Wilk) and chooses Welch's t-test or Mann-Whitney for two groups, and one-way ANOVA + Tukey or Kruskal-Wallis + Dunn for more.

staplot draws the plot panels of PyFigures.

Tests

pip install pytest pytest-qt
python -m pytest tests

License

BSD 2-Clause, see LICENSE.