msn.affine {sn} | R Documentation |
Computes the parameters of an affine transformation a+ A Y of a multivariate skew-normal or skew-t variable Y
msn.affine(dp, a=0, A, drop=TRUE) mst.affine(dp, a=0, A, drop=TRUE)
dp |
a list containg the pamaters of the variable being
transformed; it must include components xi , Omega ,
alpha as described for dmsn ; for mst.affine ,
also a component df is expected
|
A |
a matrix with ncol(A) equal to nrow(dp$Omega)
|
a |
a vector wiht length(a) equal to nrow(dp$Omega)
|
drop |
a logical flag (default value is TRUE ) operating when
nrow(A) equals 1. If these conditions are met, the output
is provided in the form of parameters of a scalar distribution,
dsn or dst , depending in the case.
|
A list containing the same components of the input parameter dp
For background information about the skew-normal and skew-t distributions, their parameters and the properties of affine transformations, see the references below. The specific formulae implemented by this function are given in Appendix A.2 of Capitanio et al.(2003).
Azzalini, A. and Capitanio, A. (1999). Statistical applications of the multivariate skew-normal distribution. J.Roy.Statist.Soc. B 61, 579–602.
Azzalini, A. and Capitanio, A. (2003). Distributions generated by perturbation of symmetry with emphasis on a multivariate skew-t distribution. J.Roy. Statist. Soc. B 65, 367–389.
Capitanio, A. et al. (2003). Graphical models for skew-normal variates. Scand. J. Statist. 30, 129–144.
dp<- list(xi=c(1,1,2), Omega=toeplitz(1/1:3), alpha=c(3,-1,2)) A <- matrix(c(1,-1,1,3,0,-2), 2, 3, byrow=TRUE) dp1 <- msn.affine(dp, 1:2, A) # dp$df <- 5 dp2<- mst.affine(dp,,A[1,,drop=FALSE]) dp3<- mst.affine(dp,,A[1,,drop=FALSE], drop=FALSE) if(zapsmall(dp2$scale^2 - dp3$Omega)) print("something wrong here!")