RankingPermutation {GeneSelector} | R Documentation |
The function is a wrapper for mt.sample.teststat
from the package multtest
(Dudoit et al., 2003). The ranking is
based on permutation p-values first, followed by the
absolute value of the statistic.
RankingPermutation(x, y, type = "unpaired", B = 100, gene.names = NULL, ...)
x |
A |
y |
If |
type |
Only the two sample case, |
B |
The number of permutations to generate. Defaults to 100,
but should be increased if computing power admits. Taking
|
gene.names |
An optional vector of gene names. |
... |
Further arguments passed to |
An object of class GeneRanking
The p-values, on which the ranking is primarily based, suffer from
the discreteness of the procedure. They follow a step function
with jump heights 1/B
.
Martin Slawski
Anne-Laure Boulesteix
Dudoit, S., Shaffer, J.P., Boldrick, J.C. (2003).
Multiple Hypothesis Testing in Microarray Experiments
Statistical Science, 18, 71-103
RepeatRanking, RankingTstat, RankingFC, RankingWelchT, RankingWilcoxon, RankingBaldiLong, RankingFoxDimmic, RankingLimma, RankingEbam, RankingWilcEbam, RankingSam, RankingShrinkageT, RankingSoftthresholdT
### Load toy gene expression data data(toydata) ### class labels yy <- toydata[1,] ### gene expression xx <- toydata[-1,] ### run RankingPermutation (100 permutations) perm <- RankingPermutation(xx, yy, B=100, type="unpaired")