omxDefaultComputePlan: Construct default compute plan

View source: R/MxCompute.R

omxDefaultComputePlanR Documentation

Construct default compute plan

Description

This function generates a default compute plan, where "default compute plan" refers to an object of class MxComputeSequence which is appropriate for use in a wide variety of cases. The exact specification of the plan will depend upon the arguments provided to omxDefaultComputePlan().

Usage

omxDefaultComputePlan(modelName=NULL, intervals=FALSE,
  useOptimizer=TRUE, optionList=options()$mxOption, penaltySearch=FALSE)

Arguments

modelName

Optional (defaults to NULL) character string, providing the name of the MxModel the fitfunction of which is to be evaluated, and usually, optimized.

intervals

Logical; will confidence intervals be computed? Defaults to FALSE.

useOptimizer

Logical; will a fitfunction be minimized? Defaults to TRUE.

optionList

List of mxOptions. Defaults to the current list of global mxOptions.

penaltySearch

Logical; whether to wrap the optimizer step with mxComputePenaltySearch

Details

At minimum, argument optionList must include “Gradient algorithm”, “Gradient iterations”, “Gradient step size”, “Calculate Hessian”, and “Standard Errors”.

Value

Returns an object of class MxComputeSequence.

Examples

foo <- omxDefaultComputePlan(modelName="bar")
str(foo)

OpenMx/OpenMx documentation built on April 17, 2024, 3:32 p.m.