harmonicmeanp: Harmonic mean p-values and model averaging by mean maximum...

In harmonicmeanp: Harmonic Mean p-Values and Model Averaging by Mean Maximum Likelihood

Harmonic mean p-values and model averaging by mean maximum likelihood

Description

The harmonic mean p-value (HMP) test combines p-values and corrects for multiple testing while controlling the strong-sense family-wise error rate. It is more powerful than common alternatives including Bonferroni and Simes procedures when combining large proportions of all the p-values, at the cost of slightly lower power when combining small proportions of all the p-values. It is more stringent than controlling the false discovery rate, and possesses theoretical robustness to positive correlations between tests and unequal weights. It is a multi-level test in the sense that a superset of one or more significant tests is certain to be significant and conversely when the superset is non-significant, the constituent tests are certain to be non-significant. It is based on MAMML (model averaging by mean maximum likelihood), a frequentist analogue to Bayesian model averaging, and is theoretically grounded in generalized central limit theorem.

Details

Package:	harmonicmeanp
Type:	Package
Version:	3.0
Date:	2019-08-17
License:	GPL-3

The key function is p.hmp for combining p-values using the HMP. Type vignette("harmonicmeanp") for detailed examples.

Author(s)

Daniel J. Wilson

Maintainer: Daniel Wilson <[e-mail]>

References

Daniel J. Wilson (2019) The harmonic mean p-value for combining dependent tests. Proceedings of the National Academy of Sciences USA 116: 1195-1200.

See Also

Package FMStable

Examples

# For detailed examples type vignette("harmonicmeanp")
# Example: simulate from a non-uniform distribution mildly enriched for small \emph{p}-values. 
# Compare the significance of the combined p-value for Bonferroni, Benjamini-Hochberg (i.e. Simes), 
# HMP and (equivalently) MAMML with 2 degrees of freedom.
p = rbeta(1000,1/1.5,1)
min(p.adjust(p,"bonferroni"))
min(p.adjust(p,"BH"))
p.hmp(p,L=1000)
p.mamml(1/p,2,L=1000)

harmonicmeanp documentation built on May 29, 2024, 1:25 a.m.