Skip to contents

rsmDesign: Generate a response surface design.

Source: R/3.3_Factorial_designs_Functions.R
rsmDesign.Rd

Generates a response surface design containing a cube, centerCube, star, and centerStar portion.

Usage

rsmDesign(
  k = 3,
  p = 0,
  alpha = "rotatable",
  blocks = 1,
  cc = 1,
  cs = 1,
  fp = 1,
  sp = 1,
  faceCentered = FALSE
)

Arguments

k

Integer value giving the number of factors. By default, k is set to `3`.

p

Integer value giving the number of additional factors in the response surface design by aliasing effects. Default is `0`.

alpha

Character string indicating the type of star points to generate. Should be `rotatable`(default), `orthogonal`, or `both`. If `both`, values for cc and cs will be discarded.

blocks

Integer value specifying the number of blocks in the response surface design. Default is `1`.

cc

Integer value giving the number of centerpoints (per block) in the cube portion (i.e., the factorial 2^k design) of the response surface design. Default is `1`.

cs

Integer value specifying the number of centerpoints in the star portion. Default is `1`.

fp

Integer value giving the number of replications per factorial point (i.e., corner points). Default is `1`.

sp

Integer value specifying the number of replications per star point. Default is `1`.

faceCentered

Logical value indicating whether to use a faceCentered response surface design (i.e., alpha = `1`). Default is FALSE.

Value

The function returns an object of class facDesign.c.

Details

Generated designs consist of a cube, centerCube, star, and centerStar portion. The replication structure can be set with the parameters cc (centerCube), cs (centerStar), fp (factorialPoints), and sp (starPoints).

See also

facDesign, fracDesign, fracChoose, pbDesign, rsmChoose

Examples

# Example 1: Central composite design for 2 factors with 2 blocks, alpha = 1.41,
# 5 centerpoints in the cube portion and 3 centerpoints in the star portion:
rsmDesign(k = 2, blocks = 2, alpha = sqrt(2), cc = 5, cs = 3)
#>    StandOrd RunOrder Block      A      B  y
#> 6         6        1     1  0.000  0.000 NA
#> 5         5        2     1  0.000  0.000 NA
#> 8         8        3     1  0.000  0.000 NA
#> 4         4        4     1  1.000  1.000 NA
#> 9         9        5     1  0.000  0.000 NA
#> 1         1        6     1 -1.000 -1.000 NA
#> 7         7        7     1  0.000  0.000 NA
#> 2         2        8     1  1.000 -1.000 NA
#> 3         3        9     1 -1.000  1.000 NA
#> 10       10       10     2 -1.414  0.000 NA
#> 14       14       11     2  0.000  0.000 NA
#> 13       13       12     2  0.000  1.414 NA
#> 15       15       13     2  0.000  0.000 NA
#> 16       16       14     2  0.000  0.000 NA
#> 12       12       15     2  0.000 -1.414 NA
#> 11       11       16     2  1.414  0.000 NA

# Example 2: Central composite design with both, orthogonality and near rotatability
rsmDesign(k = 2, blocks = 2, alpha = "both")
#>    StandOrd RunOrder Block      A      B  y
#> 6         6        1     1  0.000  0.000 NA
#> 2         2        2     1  1.000 -1.000 NA
#> 7         7        3     1  0.000  0.000 NA
#> 5         5        4     1  0.000  0.000 NA
#> 3         3        5     1 -1.000  1.000 NA
#> 1         1        6     1 -1.000 -1.000 NA
#> 4         4        7     1  1.000  1.000 NA
#> 9         9        8     2  1.414  0.000 NA
#> 8         8        9     2 -1.414  0.000 NA
#> 13       13       10     2  0.000  0.000 NA
#> 12       12       11     2  0.000  0.000 NA
#> 11       11       12     2  0.000  1.414 NA
#> 10       10       13     2  0.000 -1.414 NA
#> 14       14       14     2  0.000  0.000 NA

# Example 3: Central composite design with:
# 2 centerpoints in the factorial portion of the design (i.e., 2)
# 1 centerpoint in the star portion of the design (i.e., 1)
# 2 replications per factorial point (i.e., 2^3*2 = 16)
# 3 replications per star point (i.e., 3*2*3 = 18)
# Makes a total of 37 factor combinations
rsdo = rsmDesign(k = 3, blocks = 1, alpha = 2, cc = 2, cs = 1, fp = 2, sp = 3)