baby.dat {Rfwdmv}R Documentation

Babyfood data

Description

The responses are the density in centipoises at the time of manufacture (y1) and when measured 3 (y2), 6 (y3) and 9 months later (y4). There are five explanatory variables.

Usage

data(baby.dat)

Format

A data frame with 27 observations on the following 9 variables.

x1
a numeric vector.
x2
a numeric vector.
x3
a numeric vector.
x4
a numeric vector.
x5
a numeric vector.
y1
a numeric vector, the initial viscosity of the babyfood.
y2
a numeric vector, the viscosity of the babyfood after three months storage.
y3
a numeric vector, the viscosity of the babyfood after six months storage.
y4
a numeric vector, the viscosity of the babyfood after nine months storage.

Details

Box and draper (1987, p. 572) find a linear model with terms x2, x3 and x5, as well as surprisingly the interaction x3:x4 in the absence of x4. This model was suggested for all four responses. It is generally agreed that such models, violating a marginality constraint, are undesirable: if the variables in this model are rescaled, the model will apparently change, a term in x3 appearing.

Source

Box and Draper (1987), p. 265 present part of a larger data set on the storage of a babyfood.

References

Atkinson, Riani and Cerioli (2004), p. 567; http://www.riani.it/arc.


[Package Rfwdmv version 0.72-2 Index]