par2nMm {norMmix}

Transform Parameter Vector to Multivariate Normal Mixture

Description

Transforms the (numeric) parameter vector of our MLE parametrization of a multivariate normal mixture model into the corresponding list of components determining the model. Additionally (partly redundantly), the dimension p and number of components k need to be specified as well.

Usage

par2nMm(par, p, k, model = c("EII","VII","EEI","VEI","EVI",
                             "VVI","EEE","VEE","EVV","VVV")
      , name = sprintf("model = %s , components = %s", model, k)
       )

Arguments

par

the model parameter numeric vector.

p

dimension of data space, i.e., number of variables (aka “features”).

k

the number of mixture components, a positive integer.

model

a character string, one of those listed; see nMm2par()'s documentation.

name

a character string naming the norMmix return value.

Value

returns a list with components

weight

..

mu

..

Sigma

..

k

..

dim

..

See Also

This is the inverse function of nMm2par().

Examples

## TODO: Show to get the list, and then how to get a  norMmix() object from the list
str(MW213)
# List of 6
#  $ model : chr "VVV"
#  $ mu    : num [1:2, 1:2] 0 0 30 30
#  $ Sigma : num [1:2, 1:2, 1:2] 1 3 3 11 3 6 6 13
#  $ weight: num [1:2] 0.5 0.5
#  $ k     : int 2
#  $ dim   : int 2
#  - attr(*, "name")= chr "#13 test VVV"
#  - attr(*, "class")= chr "norMmix"

para <- nMm2par(MW213, model="EEE")
par2nMm(para, 2, 2, model="EEE")