pi0est: Estimation of the percentage of Null p-values

In MXM: Feature Selection (Including Multiple Solutions) and Bayesian Networks

Estimation of the percentage of Null p-values

Description

Estimation of the percentage of Null p-values.

Usage

pi0est(p, lambda = seq(0.05, 0.95, by = 0.01), dof = 3) 

Arguments

p	A vector of p-values.
lambda	A vector of values of the tuning parameter lambda.
dof	Number of degrees of freedom to use when estimating pi_0 with smoothing splines.

Details

The estimated proporiton of null p-values is estimated the algorithm by Storey and Tibshirani (2003).

Value

The estimated proportion of non significant (null) p-values. In the paper Storey and Tibshirani mention that the estimate of pi0 is with lambda=1, but in their R code they use the highest value of lambda and thus we do the same here.

Author(s)

Michail Tsagris

R implementation and documentation: Michail Tsagris [e-mail]

References

Storey J.D. and Tibshirani R. (2003). Statistical significance for genome-wide experiments. Proceedings of the National Academy of Sciences, 100: 9440-9445.

See Also

conf.edge.lower, bn.skel.utils, mmhc.skel, pc.skel, corfs.network, local.mmhc.skel

Examples

## simulate a dataset with continuous data
y <- rdag2(1000, p = 20, nei = 3)
ind <- sample(1:20, 20)
x <- y$x[, ind]
mod <- pc.skel( x, method = "comb.fast", alpha = 0.01 ) 
pval <- exp(mod$pvalue)
pval <- lower.tri(pval)
pi0est(pval)

MXM documentation built on Aug. 25, 2022, 9:05 a.m.