Generate an Orthogonal Array Design
DESCRIPTION:
Returns an orthogonal array design with the given levels.
USAGE: 
oa.design(levels, factor.names, min.resid.df=<<see below>>)
REQUIRED ARGUMENTS:
levels:
vector of the number of levels for each factor in the desired design. Currently only factors with two or three levels are allowed. Missing values are not accepted.
OPTIONAL ARGUMENTS:
factor.names:
factor names attribute. This may be a character vector, giving the names of the factors, or a list. If it is a list, the names attribute of the list is the names of the factors, and the elements of the list (which need not be of mode character) give the levels of the corresponding factor.
min.resid.df:
minimum residual degrees of freedom requested for a main-effects-only model. The default value is 0, unless the number of levels in the factors are all equal in which case the default is 3.
VALUE:
an object of class "design" which is a design for the factors specified, Three attributes are special to orthogonal array designs: gives the name of the object that contains the complete orthogonal array design from which the result was generated; says which columns of this object were used to produce the result; gives the number of residual degrees of freedom in the design, when only the main effects are fitted (you may want to check this value to see how many more residual degrees of freedom than are needed in your application are available). oa.design may not be able to find a design as requested. If so, an error stop is made.
DETAILS:
The design is generated by selecting some of the columns from one of a stored catalog of orthogonal array designs. An error occurs if oa.design can not find a design as specified.

If factor names are not given, they default to std.factor.names(length(levels)), that is, "A", "B", etc. If a factor's levels are not named, the levels are set to the factor name (possibly abbreviated) followed by level numbers.

BACKGROUND:
Orthogonal Array designs are (typically) highly fractional designs that are used with main-effects-only models.
SEE ALSO:
fac.design , randomize , factor.names , design , fractionate , aov .
EXAMPLES: 
oa <- oa.design(c(2,3,3,3,3,3))
#produces an 18 run design with 6 degrees of freedom
#for error assuming only main effects are fit.