mutoss (version 0.1-13)

storey_pi0_est: Storey-Taylor-Siegmund estimation of pi0 (finite sample version)

Description

The Storey-Taylor-Siegmund procedure for estimating pi0 is applied to pValues. The formula is equivalent to that in Schweder and Spjotvoll (1982), page 497, except the additional '+1' in the nominator that introduces a conservative bias which is proven to be sufficiently large for FDR control in finite families of hypotheses if the estimation is used for adjusting the nominal level of a linear step-up test.

Usage

storey_pi0_est(pValues, lambda)

Value

A list containing:

pi0

A numeric number containing the estimated value of pi0

lambda

A numeric number containing the tuning parameter for the estimation

Author

MarselScheer

Arguments

pValues

The raw p-values for the marginal test problems

lambda

A tuning parameter in the interval (0, 1)

References

Schweder, T. and Spjotvoll, E. (1982). Plots of P-values to evaluate many tests simultaneously. Biometrika 69, 3, 493-502.

Storey, J. D., Taylor, J. E. and Siegmund, D. (2004). Strong control, conservative point estimation and simultaneous conservative consistency of false discovery rates: a unified approach. JRSS B 66, 1, 187-205.

Examples

my.pvals <- c(runif(50), runif(50, 0, 0.01))
result <- storey_pi0_est(my.pvals, 0.5)