GCT.test.missing: Generalized component test for missing data

In highD2pop: Two-Sample Tests for Equality of Means in High Dimension

Description

Performs the generalized component test from Gregory et al. (2014) for the equality of two p by 1 population mean vectors given samples of sizes n and m when there are missing data.

Usage

GCT.test.missing(X, Y, r, smoother = "parzen", ntoorderminus = 2)

Arguments

X	the n by p data matrix for sample one.
Y	the m by p data matrix for sample two.
r	the lag window size for variance estimation.
smoother	the lag window used in the variance estimation. Possible values are "parzen" and "trapezoid".
ntoorderminus	a value of 0,1, or 2 such that the centering constant will retain terms of order n^(-ntoorderminus). Enter 0 for the moderate-p GCT, and enter 2 for the large-p GCT. A value of 1 may be entered to retain only terms which are O(1/n), appropriate for a size of p between moderate and large.

Value

TSvalue	the unstudentized test statistic.
center	the centering constant for studentizing the test statistic.
var	the estimated variance of the unstudentized test statistic.
T	the studentized test statistic.
smoother	the choice of smoother used.
pvalue	the pvalue.
overallpctmiss	the overall proportion of values that are missing.
pctmissperX	a vector of length p containing the proportion of missing values per component in sample one.
pctmissperY	a vector of length p containing the proportion of missing values per component in sample two.

Author(s)

Karl Gregory [e-mail], http://www.stat.tamu.edu/~kbgregory.

References

Gregory, K., Carroll, R. J., Baladandayuthapani, V. and Lahiri, S. (2015). A two-sample test for equality of means in high dimension. Journal of the American Statistician, to appear

See Also

GCT.test

Examples

	data(chr1qseg)

	X <- chr1qseg$X
	Y <- chr1qseg$Y
	
	GCT.test.missing(X,Y,r=20,smoother="parzen")

highD2pop documentation built on May 2, 2019, 5:11 a.m.