Skip to main content
Fig. 2 | BMC Bioinformatics

Fig. 2

From: Exact p-values for pairwise comparison of Friedman rank sums, with application to comparing classifiers

Fig. 2

Distribution of exact mass point probabilities and exact p-values. a Exact mass point probabilities, and exact p-values, for k = n = 5. (b) Exact p-values, and log10-transformed exact (blue line) and normal approximate p-values (red line), for k = n = 10. (c) Histogram of simulated p-values under the overall null hypothesis with expected null frequency superimposed, and cumulative distribution function of the simulated 1 − p-values with diagonal line overlay, for k = 50, n = 5.

Back to article page