HMgmm-class: Homogeneous mixed graphical Markov model

In rcastelo/qpgraph: Estimation of genetic and molecular regulatory networks from high-throughput genomics data

Homogeneous mixed graphical Markov model

Description

The "HMgmm" class is the class of homogeneous mixed graphical Markov models defined within the qpgraph package to store simulate and manipulate this type of graphical Markov models (GMMs).

An homogeneous mixed GMM is a family of multivariate conditional Gaussian distributions on mixed discrete and continuous variables sharing a set of conditional independences encoded by means of a marked graph. Further details can be found in the book of Lauritzen (1996).

Objects from the Class

Objects can be created by calls of the form HMgmm(g, ...) corresponding to constructor methods or rHMgmm(n, g, ...) corresponding to random simulation methods.

Slots

pI:

Object of class "integer" storing the number of discrete random variables.

pY:

Object of class "integer" storing the number of continuous random variables.

g:

Object of class graphBAM-class storing the associated marked graph.

vtype:

Object of class "factor" storing the type (discrete or continuous) of each random variable.

dLevels:

Object of class "integer" storing the number of levels of each discrete random variable.

a:

Object of class "numeric" storing the vector of additive linear effects on continuous variables connected to discrete ones.

rho:

Object of class "numeric" storing the value of the marginal correlation between two continuous random variables.

sigma:

Object of class dspMatrix-class storing the covariance matrix.

mean:

Object of class "numeric" storing the mean vector.

eta2:

Object of class "numeric" storing for each continuous variable connected to a discrete one, the fraction of variance of the continuous variable explained by the discrete one.

Methods

HMgmm(g)

Constructor method where g can be either an adjacency matrix or a graphBAM-class object.

rHMgmm(n, g)

Constructor simulation method that allows one to simulate homogeneous mixed GMMs where n is the number of GMMs to simulate and g can be either a markedGraphParam object, an adjacency matrix or a graphBAM-class object.

names(x)

Accessor method to obtain the names of the elements in the object x that can be retrieved with the $ accessor operator.

$

Accessor operator to retrieve elements of the object in an analogous way to a list.

dim(x)

Dimension of the homogeneous mixed GMM corresponding to the number of discrete and continuous random variables.

dimnames(x)

Names of the discrete and continuous random variables in the homogeneous mixed GMM.

show(object)

Method to display some bits of information about the input homogeneous mixed GMM specified in object.

summary(object)

Method to display a sumamry of the main features of the input homogeneous mixed GMM specified in object.

plot(x, ...)

Method to plot the undirected graph associated to the the input homogeneous mixed GMM specified in x. It uses the plotting capabilities from the Rgraphviz library to which further arguments specified in ... are further passed.

Author(s)

R. Castelo

References

Lauritzen, S.L. Graphical models. Oxford University Press, 1996.

See Also

UGgmm

rcastelo/qpgraph documentation built on Oct. 28, 2024, 5:15 a.m.