Differential analysis & tests / Kruskal–Wallis with Dunn post-hoc
Your data stays in your browser: files are read and processed on your device, never uploaded to any server, and cleared when you close the page. Free, no sign-in.
The Kruskal–Wallis test is the non-parametric counterpart of one-way ANOVA: it ranks all values and compares mean ranks between groups, without assuming normality; it suits skewed data or data with outliers.
After a significant overall test, Dunn’s test (pooled ranks, tie-corrected) compares every pair, with the adjustment you choose. Stars: * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001 (adjusted).
Letters in brackets after each group name: the group with the highest mean rank gets “a”; groups sharing no letter differ significantly. Effect size ε² = H / (n − 1).
Batch mode: upload a feature × sample matrix and a sample sheet; each row gets its own Kruskal–Wallis test and the adjustment setting is applied across features.
Kruskal W. H. & Wallis W. A. (1952) J Am Stat Assoc 47:583–621; Dunn O. J. (1964) Technometrics 6:241–252. H and P match scipy.stats.kruskal; Dunn results match scikit-posthocs posthoc_dunn.
Up to 100,000 rows, or 100,000 features in batch mode.
Send us your data and research question and you will receive a written plan within 1 working day: analysis steps, parameter rationale, deliverables and timeline. Quoted per project.