weight.matrix {nparcomp} | R Documentation |
The function weight.matrix computes the weight matrix for a given contrast with which the estimator for the relative contrast effect is comupted
weight.matrix(n, type = c("Tukey", "AVE", "Dunnett", "Sequen", "Changepoint", "Marcus", "McDermott", "Williams"), base = 1)
n |
n vector of sample sizes |
type |
type Character string defining the type of contrast. It should be one of
"Tukey", "Dunnett", "Sequen", "Williams", "Changepoint", "AVE", "McDermott", "Marcus" |
base |
base Number of the basline values in Dunnett Comparisons. By default it is base = 1 |
The function needs the R-function 'contrMat' from the R-package 'multcomp'.
The function returns the weight matrix which can be used to compute the estimator of the relative contrast effect
For the analysis, the R-package 'multcomp' is required.
Frank Konietschke
Konietschke, F., Brunner, E., Hothorn, L.A. (2008). Nonparametric Relative Contrast Effects: Asymptotic Theory and Small Sample Approximations, Konietschke, F., Brunner, E., Hothorn, L.A. (2008). Simultaneous Confidence Intervals for Relative Effects in Dunnett Comparsisons.
contrMat
in package multcomp for contrast matrices
n<-c(10,20,30,40) weight.matrix(n, "Tukey") # result: identity matrix by construction weight.matrix(n, "Dunnett", base = 3) # Weight matrix for Dunnett comparison - 3rd group is baseline weight.matrix(n, "Changepoint") # weight matrix for relative changepoint effect # For comparison, see the contrasts matrices # available in the multcomp package: library(multcomp) n<-c(10,10,20,30,40) contrMat(n, "Changepoint") # Contrast matrix for Changepoint comparisons