cid {stagedtrees}

Context specific interventional discrepancy

Description

Compute the context specific interventional discrepeancy of a staged tree with respect to a reference staged tree.

Usage

cid(object1, object2, FUN = mean)

Arguments

object1

an object of class sevt.

object2

an object of class sevt.

FUN

a function that is used to aggregate CID for each variable. The default mean will obtain the CID as defined in Leonelli and Varando (2023).

Value

A list with components:

• wrong a stages-like structure which record where object2 wrongly infer the interventional distance with respect to object1.

• cid the value of the computed CID.

References

Leonelli M., Varando G. Context-Specific Causal Discovery for Categorical Data Using Staged Trees, The 26th International Conference on Artificial Intelligence and Statistics (AISTATS), 2023, https://arxiv.org/abs/2106.04416

Examples

model1 <- stages_bhc(full(Titanic))
model2 <- stages_bhc(full(Titanic,
  order = c("Survived", "Sex", "Age", "Class")
))
cid(model1, model2)$cid
cid(model1, model2)$wrong