powered by
For a multivariate (extended) skew-normal distribution, compute its conditional distribution for given values of some of its components.
conditionalSECdistr(object, fixed.comp, fixed.values, name, drop = TRUE)
an object of class SECdistrMv, except in the case when
SECdistrMv
drop=TRUE operates, leading to an object of class
drop=TRUE
SECdistrUv-class.
SECdistrUv-class
an object of class SECdistrMv with family="SN" or family="ESN".
family="SN"
family="ESN"
a vector containing a subset of 1:d which selects the components whose values are to be fixed, if d denotes the dimensionality of the distribution.
1:d
d
a numeric vector of values taken on by the components fixed.comp; it must be of the same length of fixed.comp.
fixed.comp
an optional character string with the name of the outcome distribution; if missing, one such string is constructed.
logical (default=TRUE), to indicate whether the returned object must be of class SECdistrUv when length(fixed.comp)+1=d.
TRUE
SECdistrUv
length(fixed.comp)+1=d
For background information, see Section 5.3.2 of the reference below.
Azzalini, A. and Capitanio, A. (2014). The Skew-normal and Related Families. Cambridge University Press, IMS Monographs series.
makeSECdistr, SECdistrMv-class, affineTransSECdistr
makeSECdistr
SECdistrMv-class
affineTransSECdistr
Omega <- diag(3) + outer(1:3,1:3) sn <- makeSECdistr(dp=list(xi=rep(0,3), Omega=Omega, alpha=1:3), family="SN") esn <- conditionalSECdistr(sn, fixed.comp=2, fixed.values=1.5) show(esn)
Run the code above in your browser using DataLab