qch: Query Composed Hypotheses

Provides functions for the joint analysis of K sets of p-values obtained for a same list of items. This joint analysis is performed by querying a composed hypothesis, i.e. an arbitrary complex combination of simple hypotheses, as described in Mary-Huard et al. (2021) <arXiv:2104.14601>. The null distribution corresponding to the composed hypothesis of interest is obtained by fitting non-parametric mixtures models (one for each of the simple hypothesis of the complex combination). Type I error rate control is achieved through Bayesian False Discovery Rate control. The 3 main functions of the package GetHinfo(), qch.fit() and qch.test() correspond to the 3 steps for querying a composed hypothesis (composed H0/H1 formulation, inferring the null distribution and testing the null hypothesis).

Version: 1.0.0
Depends: R (≥ 2.10)
Imports: graphics, ks, mclust, stats
Published: 2021-05-07
Author: Tristan Mary-Huard ORCID iD [aut, cre]
Maintainer: Tristan Mary-Huard <tristan.mary-huard at agroparistech.fr>
License: GPL-3
NeedsCompilation: no
CRAN checks: qch results

Documentation:

Reference manual: qch.pdf

Downloads:

Package source: qch_1.0.0.tar.gz
Windows binaries: r-devel: qch_1.0.0.zip, r-release: qch_1.0.0.zip, r-oldrel: qch_1.0.0.zip
macOS binaries: r-release (arm64): qch_1.0.0.tgz, r-oldrel (arm64): qch_1.0.0.tgz, r-release (x86_64): qch_1.0.0.tgz

Linking:

Please use the canonical form https://CRAN.R-project.org/package=qch to link to this page.