MAG.to.DAG: Function to convert a Mixed Acyclic Graph (MAG) into a...

In BillShipley/CauseAndCorrelation: Functions for Path Analysis and SEM

Function to convert a Mixed Acyclic Graph (MAG) into a Directed Acyclic Graph (DAG) by replacing the dependent error by latent variables.

Description

This function converts a Mixed Acyclic Graph (MAG) into a Directed Acyclic Graph (MAG), by adding a latent for each pair of dependent errors in the MAG. More details can be found in Shipley, B. & Douma, J. Testing piecewise structural equations models in the presence of latent variables, including correlated errors. It is typically used to obtain a DAG from which the d-sep equivalent MAG can be derived and its basis-set.

Usage

MAG.to.DAG(full.MAG)

Arguments

full.DAG

a named binary matrix encoding the MAG. This is typically obtained via the DAG function of the ggm library.

Details

This function inputs a Mixed Acyclic Graph (MAG) and outputs the Directed Acyclic Graph (MAG).

Value

The resulting MAG is output from this function as a matrix.

Author(s)

Bob Douma

References

Shipley, B. & Douma, J. Testing piecewise structural equations models in the presence of latent variables, including correlated errors.

See Also

dsep.test, basiSet.mag, DAG,DAG.to.MAG

Examples

library(ggm)
initial.mag <- makeMG(dg=DAG(X3~X1+X2,X4~X3,X5~X4),bg=UG(~X1*X2+~X3*X5))
plotGraph(initial.mag) # plotgraph
my.DAG <- MAG.to.DAG(initial.mag) # resulting DAG
my.mag <- DAG.to.MAG(my.dag,latents=c("L1","L2")) 
plotGraph(my.mag) 
basiSet.mag(my.mag) 

BillShipley/CauseAndCorrelation documentation built on Jan. 31, 2023, 4:20 a.m.