bootstrapCor {maigesPack} | R Documentation |
This function takes a numerical matrix (or two vectors) and calculates bootstrapped (by permutation) p-values to test if the correlation value is equal to zero. If the first argument is a matrix, the p-values are calculated between all pairs of rows of the matrix.
bootstrapCor(x, y=NULL, bRep, type="Rpearson", ret="p-value", alternative="two.sided")
x |
numerical matrix or vector to be analysed. If a vector, the
argument y must be informed. |
y |
numerical vector. Must be informed if x is a
vector. If x is a matrix, this argument is ignored. Defaults
to NULL. |
bRep |
number of permutation to be done in the test. |
type |
character string specifying the type of correlation statistic to be used. Possible values are 'Rpearson', 'pearson', 'spearman' or 'kendall'. |
ret |
character string with the value to return. Must be 'p-value' (default) for the usual p-value or 'max', to return the maximum absolute correlation value obtained by the permutation. |
alternative |
character specifying the type of test to do, must be 'two.sided' (default), 'less' or 'greater'. |
Pearson, spearman and kendall types of correlation values are
calculated by cor
function from package
stats. The method Rpearson was developed in this package and is a
generalisation of the jackniffe correlation proposed by Heyer
et al. (1999), it
is calculated using the function robustCorr
.
The result of this function is a square matrix (length equal to the
number of rows of x
) if x
is a matrix or a numerical
value if x
and y
are vectors. The result is the p-values
or maximum correlation values calculated by permutation tests.
Gustavo H. Esteves <gesteves@vision.ime.usp.br>
Heyer, L.J.; Kruglyak, S. and Yooseph, S. Exploring expression data: identification and analysis of coexpressed genes, Genome Research, 9, 1106-1115, 1999 (http://www.genome.org/cgi/content/full/9/11/1106)
x <- runif(50, 0, 1) y <- rbeta(50, 1, 2) bootstrapCor(x, y, bRep=100) z <- matrix(rnorm(100, 0, 1), 4, 25) bootstrapCor(z, bRep=100)