cp_to_dp(param) dp_to_cp(param)
param
|
a vector of length al least three. If lenght(param) is m+2 , then the first m components refer to the regression coefficients (or the location parameter, in case m is 1), and the remaining two components refer to scale and shape, respectively; their role is preserved across parametrizations. |
Convert direct to centred parameters of the skew-normal distribution, and viceversa
The advantages of using the centred parametrization, rather than the direct one, are discussed in the reference below.
a vector of the same length of param
, representing param
in the alternative parametrization; cp_to_dp
converts centred to direct parameters, dp_to_cp
converts direct to centred parameters.
Azzalini, A. and Capitanio, A. (1999). Statistical applications of the multivariate skew-normal distribution. J.Roy.Statist.Soc. B 61, part 3.
cp = dp_to_cp([30,30,2,4]) dp = cp_to_dp(cp)