Learn R Programming

MConjoint (version 0.1)

mc.get.one.design: mc.get.one.design

Description

Get a single "good" design from a list of cards

Usage

mc.get.one.design(all.cards, cards, slack = 1, tol = 0.2, max.tries = 1e+06)

Arguments

all.cards
The set of cards from which the design is to be drawn
cards
The number of cards in the design.
slack
How much the number of each factor level can vary in a "good" design
tol
The largest cross correlation in a "good" design
max.tries
The maximum number of designs to look at.

Value

A single design or NULL and an error message if no design is found

Details

Take samples of size cards from all.cards to form designs. Check each of the designs to see if it is "good". Return the first good design found. (This is one very simple way of getting a design to start with. Alternatively you could use a classical design (e.g. a fractional factorial design) or an "optimal" design as produced by the AlgDesign package).

Examples

Run this code
data(hire.candidates)

#get a nine card design
mc.get.one.design(hire.candidates,9)

#get a 15 card design with cross correlations less that .1
mc.get.one.design(hire.candidates,15,tol=.1)

Run the code above in your browser using DataLab