Create an Object of Class designfactors

Description

Create an object of class designfactors, from the factor names and their numbers of levels, or from a named list of factor levels. Both ways can be used in the same call. Additional information can be provided: they will be used during the design search or in the summary functions applied to the object.

Usage

1
2
3
planor.factors(factors = NULL, nlevels = NULL,
  block = NULL, ordered = NULL, hierarchy = NULL,
  dummy = FALSE)

Arguments

factors

a character vector of factor names, or possibly a scalar, a dataframe or a list (see DETAILS).

nlevels

a vector of level numbers for each factor name (see DETAILS).

block

an additive model formula to indicate the block factors.

ordered

an additive model formula to indicate the quantitative factors (not used at all in the present version).

hierarchy

a formula or a list of formulae to indicate hierarchy relationships between factors (see the planor vignette for details).

dummy

a logical to identify dummy factors created and deleted by planor functions for technical reasons.

Value

An object of class designfactors.

Note

The basic usage is to specify the names of the factors by a character vector of length n in argument factors and their numbers of levels by a numeric vector of length n in argument nlevels. Alternatively, the factors argument can be an integer n, in which case the first n capital letters of the alphabet are used as factor names. If nlevels is a scalar s, it is considered that all factors have s levels. There are two more possibilities which allow for alphanumeric factor levels. If factors is a dataframe, the factors in this dataframe are extracted together with their levels. Finally factors can be a named list of n vectors, with each vector containing the levels of one factor. Note that nlevels is ignored in these latter two cases. See the examples. The argument block allows to specify the block or nuisance factors. This information is used by the alias and summary functions but it has no effect on the design generation and randomization which depend on other arguments.

Author(s)

Monod, H. and Bouvier, A.

See Also

Class designfactors

Examples

1
2
3
4
5
6
7
8
9
planor.factors(c("A","B","C","P"),c(2,3,6,3))
planor.factors(LETTERS[1:12],2)
planor.factors(12,2)
planor.factors( c("A","B","Block"), 3, block=~Block)
zz <- planor.factors( c("A","B","Block"), c(2,3,5))
zz@levels$A <- c("plus","moins")
planor.factors(factors=list(A=c("plus","moins"), B=1:3, Block=1:5))
AB <- data.frame( A=c(rep(c("a","b"),3)), B=rep(c("z","zz","zzz"),rep(2,3)), C=1:6)
planor.factors(factors=AB)