multivariateSkewNormal: Multivariate Skew Normal
In rockclimber112358/MCLE: Package to robustly fit distributions

Functions defining the multivariate skew normal distribution. These should be used to construct a distribution class as defined in this package.

devMSN(x, params)

gradDevMSN(x, params)

paramVec2ListMSN(paramVec)

paramList2VecMSN(paramList)

`x`	A matrix of observations with one row per observation.
`params`	A vector, typically as created by paramList2Vec called on a list object with three elements: xi (a numeric vector) Omega (a matrix), and alpha (a numeric vector).
`paramVec`	A vector of the parameters of the multivariate distribution.
`paramList`	A list with three elements: xi (a numeric vector), Omega (a numeric matrix) and alpha (a numeric vector).

The deviance function returns the deviance of all the observations with the given parameters. The gradDev function computes the gradient of this deviance. The other two functions are useful for converting parameter vectors to lists and vice versa, and thus they return the list or vector.

Note: the multivariate skew normal distribution is implemented as a multivariate skew t-distribution with nu parameter equal to 1e8. This leads to minimal differences between the two distributions.

rockclimber112358/MCLE documentation built on May 27, 2019, 12:15 p.m.