test-edges: Test edges in graphical models with p-value/AIC value

In hojsgaard/gRim: Graphical Interaction Models

Test edges in graphical models with p-value/AIC value

Description

Test edges in graphical models with p-value/AIC value. The models must be iModels.

Usage

testEdges(
  object,
  edgeMAT = NULL,
  ingraph = TRUE,
  criterion = "aic",
  k = 2,
  alpha = NULL,
  headlong = FALSE,
  details = 1,
  ...
)

testInEdges(
  object,
  edgeMAT = NULL,
  criterion = "aic",
  k = 2,
  alpha = NULL,
  headlong = FALSE,
  details = 1,
  ...
)

testOutEdges(
  object,
  edgeMAT = NULL,
  criterion = "aic",
  k = 2,
  alpha = NULL,
  headlong = FALSE,
  details = 1,
  ...
)

Arguments

object	An iModel model object
edgeMAT	A p * 2 matrix with edges
ingraph	If TRUE, edges in graph are tested; if FALSE, edges not in graph are tested.
criterion	Either "aic" or "test" (for significance test)
k	Penalty term when criterion="aic". Only k=2 gives genuine AIC.
alpha	Critical value for deeming an edge to be significant/ insignificant. When criterion="aic", alpha defaults to 0; when criterion="test", alpha defaults to 0.05.
headlong	If TRUE then testing will stop once a model improvement has been found.
details	Controls the level of printing on the screen.
...	Further arguments to be passed on to testdelete (for testInEdges) and testadd (for testOutEdges).

Details

• testIn: Function which tests whether each edge in "edgeList" can be delete from model "object"

• testOut: Is similar but in the other direction.

Value

A dataframe with test statistics (p-value or change in AIC), edges and logical telling if the edge can be deleted.

Author(s)

Søren Højsgaard, [e-mail]

See Also

getEdges, testadd, testdelete

Examples

data(math)
cm1 <- cmod(~me:ve + ve:al + al:an, data=math)
testEdges(cm1, ingraph=TRUE)
testEdges(cm1, ingraph=FALSE)
## Same as
# testInEdges(cm1)
# testOutEdges(cm) 

hojsgaard/gRim documentation built on Oct. 19, 2024, 10:46 p.m.