twilight.combi {twilight}R Documentation

All permutations of a binary vector

Description

For a given binary input vector, the function completely enumerates all possible permutations.

Usage

twilight.combi(xin, pin, bin)

Arguments

xin Binary input vector, e.g. class labels.
pin Logical value. TRUE if samples are paired, FALSE if not.
bin Logical value. TRUE if permutations should be balanced, FALSE if not.

Details

Please note, that the resulting permutations are always as "balanced" as possible. The balancing is done for the smaller subsample. If its sample size is odd, say 5, twilight.combi computes all permutations with 2 or 3 samples unchanged. In the paired case, the output matrix contains only one half of all permutations. The second half is simply 1-output which leads to the same absolute test statistics in a paired test.

Value

Returns a matrix where each row contains one permuted vector. Note that even for balanced permutations, the first row always contains the original vector. If the number of rows exceeds 10000, NULL is returned.

Author(s)

Stefanie Scheid http://www.molgen.mpg.de/~scheid

References

Scheid S and Spang R (2004): A stochastic downhill search algorithm for estimating the local false discovery rate, IEEE TCBB 1(3), 98–108.

See Also

twilight.permute.pair, twilight.permute.unpair

Examples

x <- c(rep(0,4),rep(1,3))
y <- twilight.combi(x,pin=FALSE,bin=FALSE)

[Package twilight version 1.4.0 Index]