An Effective Design Based Model Fitting Method for Definitive Screening Designs
This function performs fits a model to a Definitive Screeing Design by first restricting main effects to the smallest main effects and those significant at at least the .20 level in a main effects model. Next forward stepwise selection is used to enter 2 factor interactions and quadratic effects.
FitDefSc(y,design,alpha=.05)
y |
input - this is a vector containing a single numeric column of response data. |
design |
input - this is a data frame containing the numeric columns of the candidate independent variables created by the DefScreen function with only numerical factors i.e. c=0. The factor names or colnames(design) should always be of length 1 (for example letters of the alphabet "A", "B", etc.) |
alpha |
input - alpha to enter in the forward stepwise regression with second order candidates should be between 0.05 and 0.20 |
John Lawson
design<-DefScreen(m=5,c=0,randomize=FALSE) Smeso<-c(241,295,260,338,320,265,275,248,66,383,313) FitDefSc(Smeso,design,alpha=.12)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.