opt_rot_vec: Finds the optimal foldover plans for 3 level designs

In vinny-paris/optrotvec: Find Optimal Foldover Plans for 3 Level Factorial Designs

Description

For a given design matrix encoded with 0,1,2 this function will return the values of Omega, the determinant of the inverse of X_1'X_1, and

Usage

opt_rot_vec(design, return_n = 5, opt = "Omega")

Arguments

design	The encoded design matrix using 0,1,2 notation that is to be expanded
return_n	How many of the top rotation vectors (with regards to Omega) should be returned? Default is 5.
opt	What optimality criterion should be used? Options is Omega (default), Det, Run_Size, Min_Inc (see below for details).

Details

There are four unique optimality criterion's that the foldover design can be chosen for. They are Omega, which returns the highest Omega-values (ties broken by Det.), Det which returns designs ranked by the determinant value, Run_Size, which ranks designs in the minimum number of runs to compelte a foldover without replication and finally Min_Incident which ranks designs by the maximum of the minimum of the number of pairs of any column at any level coincide with each other.

Value

Returns a data frame that is return_n by f + 2. Omega-value. This will be the first column Determenant of the information matrix Det. Ratio. This is the ratio of the deteminant column over the the detemenant of the information matrix for a design that is just a tripling of the original. Rotation vectors

Examples

## Not run: 
d <- ran_D(25, 4)
opt_rot_vec(d, return_n) 

## End(Not run)

vinny-paris/optrotvec documentation built on April 9, 2021, 4:34 a.m.