Class RandomMimicGraphGenerator

java.lang.Object

edu.cmu.tetrad.search.mimic.RandomMimicGraphGenerator

public final class RandomMimicGraphGenerator
extends Object

Generates random MIMIC models.

In singly-connected mode, the latent subgraph is generated as a directed tree, and each measured input and output is attached to exactly one latent. This guarantees that the underlying undirected graph is singly connected.

In general mode, the latent graph is a random DAG, and measured inputs and outputs may be attached to multiple latents.

The generator recognizes the following parameter names if the generate(Parameters) method is used:

• mimicNumInputs
• mimicNumLatents
• mimicNumOutputs
• mimicSinglyConnected
• mimicLatentEdgeProb
• mimicInputAttachProb
• mimicOutputAttachProb

Author:

josephramsey

Constructor Summary

Constructors

Constructor	Description
RandomMimicGraphGenerator()	Default constructor for the RandomMimicGraphGenerator class.

Method Summary

Modifier and Type	Method	Description
MimicModel	generate(int numInputs, int numLatents, int numOutputs, boolean singlyConnected, double latentEdgeProb, double inputAttachProb, double outputAttachProb)	Generates a random MIMIC model.
MimicModel	generate(Parameters parameters)	Generates a random MIMIC model using values read from the given parameters object.

Methods inherited from class Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

RandomMimicGraphGenerator

public RandomMimicGraphGenerator()

Default constructor for the RandomMimicGraphGenerator class. Initializes an instance of the generator, which is capable of generating random MIMIC models through its provided methods.

Method Details

generate

public MimicModel generate(Parameters parameters)

Generates a random MIMIC model using values read from the given parameters object.

Parameters:

parameters - the parameters

Returns:

the generated model

generate

public MimicModel generate(int numInputs,
 int numLatents,
 int numOutputs,
 boolean singlyConnected,
 double latentEdgeProb,
 double inputAttachProb,
 double outputAttachProb)

Generates a random MIMIC model.

Parameters:

numInputs - number of measured inputs

numLatents - number of latents

numOutputs - number of measured outputs

singlyConnected - whether singly-connected structure should be enforced

latentEdgeProb - latent-edge probability in general mode

inputAttachProb - input attachment probability in general mode

outputAttachProb - output attachment probability in general mode

Returns:

the generated model