subset.Neighborhood: Subsets based on neighborhood

In gremes: Estimation of Tail Dependence in Graphical Models

Subsets based on neighborhood

Description

Derives subsets based on neighborhood of a certain order. The scope is to obtain for every node with observed variable a neibourhood nodes and use these subsets for estimation.

Usage

## S3 method for class 'Neighborhood'
subset(
  obj,
  eta,
  g,
  U_bar = NULL,
  vertices = base::setdiff(get.vertex.attribute(g, "name", V(g)), U_bar),
  matrixForm = FALSE,
  ...
)

Arguments

obj	Object of class Neighborhood
eta	Order of the neighborhood
g	The graph, an igraph object
U_bar	The set of nodes for which data are missing. Default is NULL. If there are nodes for which the corresponding variables are not observed the set of these nodes should be provided here. This will avoid taking a variable with unobserved variable to be taken as a root.
vertices	Vertex(vertices) for which the neighboring set(s) are derived. Default is all vertices with observed variable.
matrixForm	a matrix of ones and zeros: one if the node is in the subset and zero otherwise.
...	additional arguments

Value

An object of class Neighborhood containing two slots: slot $value can be a matrix or list. If it is a matrix, the number of columns is the same as the number of vertices in the graph. The number of rows is the length of argument vertices. If it is a list, the list contains for each of the nodes in vertices the neighborhood of order eta. The slot $root contains the vertices for which the neighborhood sets are obtained. If there are nodes with missing data subsets will be created for roots with available data only. A subset however can contain a node with missing data.

Examples

seg<- graph(c(1,2, 2,3, 2,4, 4,5, 5,6, 5,7), directed = FALSE)
name_stat<- c("paris", "2", "meaux", "melun", "5", "nemours", "sens")
seg<- set.vertex.attribute(seg, "name", V(seg), name_stat)
rdsobj<- Neighborhood()
subset(rdsobj, 2, seg)
subset(rdsobj, 2, seg, U_bar = c("melun", "5"), matrixForm = TRUE)

gremes documentation built on Feb. 16, 2023, 8:06 p.m.