Make a design of experiment

makeDoe(
  levels,
  type = NULL,
  nTrials = NA,
  minTrials = NULL,
  search = FALSE,
  nRepeats = 5
)

Arguments

levels	A named list of vectors defining each attribute (the names) and each level for each attribute (the vectors). For example, a design with two attributes, "price" and "type", that each had three levels should be defined as `levels = list(price = c(1, 2, 3), type = c(1, 2, 3))`.
type	The type of design. Defaults to `NULL`, which returns a full factorial design. Set to `"D"`, `"A"`, and `"I"` to obtain a fractional factorial design with either D, A, or I optimality criteria.
nTrials	The number of trials to be used in a fractional factorial design. Defaults to `NA`, but must be a number less than the number of alternatives in the full factorial design if the `type` argument is anything other than `NULL`. If `search = TRUE`, then all feasible designs will be calculated up to `nTrials` or until a perfect D-efficiency is found.
minTrials	If `search == TRUE` and `minTrials` is a number less than or equal to `nTrials`, then the feasible designs are calculated over the range of `minTrials` to `nTrials`. Defaults to `NULL`, in which case the search starts at the lowest feasible number of levels.
search	If `TRUE`, all feasible designs are calculated up to `nTrials` or until a perfect D-efficiency is found, after which a summary of the search results is printed and the top-ranked d-efficient design is returned.
nRepeats	Number of times the design search using `AlgDesign::optFederov()` is repeated for any one instance of `nTrials`. Defaults to `5`.

Value

Returns a full factorial or fraction factorial design of experiment.

Examples

# Define the attributes and levels
levels <- list(
  price     = seq(1, 4, 0.5), # $ per pound
  type      = c('Fuji', 'Gala', 'Honeycrisp', 'Pink Lady', 'Red Delicious'),
  freshness = c('Excellent', 'Average', 'Poor')
)

# Make a full-factorial design of experiment
doe <- makeDoe(levels)

# Make a fraction-factorial design of experiment based on D-efficiency
doe <- makeDoe(levels, type = "D", nTrials = 100)
#>   nTrials     d balanced
#> 1     100 0.913    FALSE