create_matrix: Create the predictor matrix and the vector of observed node...

In ghosoya/ludwig: Simulating and Estimating Undirected Binary Networks Using the Pseudolikelihood Method

Description

The main purpose of this function is to create a predictor matrix (predict_matrix) and a vector of observed states (y) for each node so that the network parameters are estimable using simple multivariate logistic regression based on the conditional distributions of each node given the rest of the network. This method is also known as the pseudolikelihood method, supposedly first introduced by Besag (1975). Note that the speed of this function is improvable.

Usage

create_matrix(data)

Arguments

data	a binary data matrix containing observed network states. The rows represent the states, the columns represent the nodes.

Value

A list containing...

• data the original data matrix.

• predict_matrix the predictor matrix.

• y a vector of the observed node states.

• n the number of observed state vectors.

• k the number of nodes.

• n_links the number of links in the network.

• n_param the number of parameters to be estimated.

ghosoya/ludwig documentation built on Sept. 20, 2020, 3:48 p.m.