scaledTMix.sat: Fits saturated model to t-statistics

In hisemi: Hierarchical Semiparametric Regression of Test Statistics

Description

Fit saturated model to t-statistics, i.e., a two-component mixture model (a central t and a scaled central t with scale greater than 1) to each t-statistics separately.

Usage

scaledTMix.sat(tstat, df)

Arguments

tstat	A numeric vector of t-statistics.
df	A numeric scalar or vector of the same length as tstat, giving the degrees of freedom for the tstat.

Details

This functions assumes each t-statistics coming from either a central t-distribution or a scaled central t-distribution. Each t-statistic has a different mixing proportion pi0, whose maximum likelihood estimate will be either 0 or 1. Each t-statistic has a different scale parameter. If pi0=1, the scale parameter will be 1; if pi0=0, the scale parameter will be greater than 1.

Value

A numeric vector of estimated scale parameters, with two attributes

pi0	A numeric vector of estimated pi0
logLik	A numeric vector of log likelihood

Note

Whenever the absolute value of the tstat is less than 1, pi0 will be esitmated to be 1 and the scale will also be 1. Otherwise, the pi0 will be estimated to be 0 and scale will be the absolute value of tstat.

Author(s)

Long Qu [e-mail]

See Also

scaledTMix.null, scaledTMix.psat

Examples

set.seed(99927220)
(tstat=rt(5,1))
scaledTMix.sat(tstat,1)

hisemi documentation built on July 9, 2017, 9:02 a.m.