multifluo (version 1.1)

difftable: Table of multivariate differences

Description

Built a table crossing modalities of a trestmanet variable about multivariate differences

Usage

difftable(mat,vep=NULL,axes=c(1,2),var.col=NULL,trt="zone",test="hotelling")

Arguments

mat

the matrix must have one column named as trt, and

vep

when vep is not NULL, a preliminary operation multiplies mat [,var.col] by the vep matrix (useful to do the test only on the first principal components of mat).

axes

vep can contain all eigenvectors. If it is the case, axes allows the user to select only some eigenvectors (1:2 for the first two eigenvectors, 1:5 for the first five eigenvectors...)

var.col

indicates the number of columns of the numerical variables. If NULL, all columns but trt are selected.

trt

name of the column

test

"hotelling" or "sri". Indicates the type of test to detect multivariate differences. "hotelling" is the usual hotelling T2 test whereas "sri" corresponds to the Srivastava's Test, allowing a test of differences even if there is more variables than observations.

Value

A table containing the p-values is returned.

See Also

hotelling.test, sri.test

Examples

# NOT RUN {
data(df.scaled)
# }
# NOT RUN {
difftable(df.scaled[,-5],trt="zone")
# }
# NOT RUN {
# }