Menu.DoptTab1: Basic information for creation of D-optimal Designs

Description Brief statistical background Inputs on Tab Base Settings Author(s) See Also

Description

Basic information for creation of D-optimal designs; this menu has been created in August 2010 and is still experimental. Please report issues for helping improve the software.

Brief statistical background

D-optimal designs are useful for generating the design points such that they are efficient for estimation of the parameters of a pre-specified model.

For more information, look at the help file for function Dopt.design from package DoE.wrapper or at optFederov from package AlgDesign.

Inputs on Tab Base Settings

name of design

must be a valid name. The design itself is created under this name in the R workspace.

Determine the candidate design

An existing design can be selected, or a new design can be generated;
the candidate design is a large design from which the D-optimal design is selected.

Specify model

formula of the model for which an optimum design is sought

Specify constraints

constraint condition(s) on the design variables; for example, if several factors refer to sizes, it may not be possible to combine all settings (e.g. under space restrictions);
See Syntax and Logic for an explanation of the syntax of general and especially logical R expressions.

nRepeat

number of independent repeats for optimization;
increasing this number increases search time, but also the chances to find a global optimum

seed

can be provided for reproducing an existing design.

randomization

can be switched off in exceptional situations. Usually, one would not want to do that. Note that the checkbox here refers to the optimization process only. For the candidate set, this can also be decided on the respective separate menu.

Author(s)

Ulrike Groemping

See Also

See also function Dopt.design for the function from DoE.wrapper that creates the design, which in turn is based on function optFederov from package AlgDesign.


RcmdrPlugin.DoE documentation built on May 30, 2017, 6:42 a.m.