as.igraph.DAG: Transform a DAG object into an `igraph` object

Description

This function extends the as.igraph function from the igraph package to allow the input of a DAG object. The result is an igraph object that includes only the structure of the DAG, not any specifications. May be useful for plotting purposes.

Usage

# S3 method for DAG
as.igraph(x, include_root_nodes=TRUE,
          include_td_nodes=TRUE, ...)

Value

Returns a igraph object.

Arguments

x: A DAG object created using the empty_dag function with nodes added to it using the + syntax. See ?empty_dag or ?node for more details. Supports DAGs with time-dependent nodes added using the node_td function. However, including such DAGs may result in cyclic causal structures, because time is not represented in the output matrix.
include_root_nodes: Whether to include root nodes in the output matrix. Should usually be kept at TRUE (default).
include_td_nodes: Whether to include time-dependent nodes added to the dag using the node_td function or not. When including these types of nodes, it is possible for the adjacency matrix to contain cycles, e.g. that it is not a classic DAG anymore, due to the matrix not representing the passage of time.
...: Currently not used.

Author

Robin Denz

Examples

Run this code

library(simDAG)

# some example DAG
dag <- empty_dag() +
  node("death", type="binomial", parents=c("age", "sex"), betas=c(1, 2),
       intercept=-10) +
  node("age", type="rnorm", mean=10, sd=2) +
  node("sex", parents="", type="rbernoulli", p=0.5) +
  node("smoking", parents=c("sex", "age"), type="binomial",
       betas=c(0.6, 0.2), intercept=-2)

if (requireNamespace("igraph")) {
  g <- igraph::as.igraph(dag)
  plot(g)
}