PoDDyHePoModelSelection: Model Selection Function Developed for PoDDy-HePo Project...

In zauziii/PoDDyHePo-projection: Projection Method for PoDDy-HePo Project

Model Selection Function Developed for PoDDy-HePo Project (BIC based)

Description

PoDDyHePoModelSelection returns a list of 4 elements: Models (Records all selected models of each imputed data set), Mean BIC (Records Mean BIC of each round of imputation), Selected Models (Records a selected model/selected models over imputation of cases) Suggested Model (Records a suggested model with smaller BIC value)

Usage

PoDDyHePoModelSelection(
  imp,
  DV,
  NsVar = NULL,
  df = NULL,
  knots = NULL,
  b.knots = NULL,
  f.var = NULL
)

Arguments

imp	An object of class mids (produced by mice().
DV	Abbreviation of Dependent Variable.
NsVar	A specific variable to which would apply nature splines. The default is NULL.
df	Degrees of freedom. The default is df=NULL.
knots	Breakpoints that define the spline.
b.knots	Boundary points at which to impose the natural boundary conditions and anchor the B-spline basis (default the range of the data).
f.var	Fixed variables. Variables that are fixed in the model (will not be removed by stepwise process).

Details

Description of PoDDyHePoModelSelection()

When NsVar = NULL, there is only one BIC value in Mean BIC and one model in Selected Models, the model in Selected Models is the same as in Suggested Model. df or knots should be denied. When both knots and b.knots are specified, df will be ignored.

Value

Returns a list with 4 elements: Models, Mean BIC, Selected Models and Suggested Model.

zauziii/PoDDyHePo-projection documentation built on May 2, 2024, 3:48 a.m.