net.gen.from.P

Generates an adjacency matrix from a given probability matrix,
 according independent Bernoulli -- the so-called inhomogeneous
 Erdos-Renyi model. It is used to generate new networks from a given model.

Linear regression model with nonparametric network effects on network-linked observations. The model is proposed by Le and Li (2022) <arXiv:2007.00803> and is assumed on observations that are connected by a network or similar relational data structure. The model does not assume that the relational data or network structure to be precisely observed; thus, the method is provably robust to a certain level of perturbation of the network structure. The package contains the estimation and inference function for the model.

Tianxi Li 

NetworkReg

Regression Model on Network-Linked Data with Statistical
Inference

Can M. Le 

net.gen.from.P function

<dl> <dt>P</dt>
<dd>connection probability between nodes</dd> <dt>mode</dt>
<dd>"undirected" (default) if the network is undirected, so the
adjacency matrix will be symmetric with only upper diagonal entries
being generated as independent Bernoulli. Otherwise, the adjacency matrix gives independent Bernoulli everywhere.</dd></dl>

Arguments

Can M. Le and Tianxi Li. 
Maintainer: Tianxi Li &lt;tianxili@umn.edu&gt;

Author

generates a network from the given connection probability — net.gen.from.P

<dl>

 <dt>P</dt>
<dd>connection probability between nodes</dd>

 <dt>mode</dt>
<dd>"undirected" (default) if the network is undirected, so the
adjacency matrix will be symmetric with only upper diagonal entries
being generated as independent Bernoulli. Otherwise, the adjacency matrix gives independent Bernoulli everywhere.</dd>

</dl>

Can M. Le and Tianxi Li. 
Maintainer: Tianxi Li &lt;tianxili@umn.edu&gt;

net.gen.from.P: generates a network from the given connection probability

Description

Usage

Value

Arguments

Author