[R] fractional factorial design in R

Martin Maechler maechler at stat.math.ethz.ch
Sat Jan 21 09:33:20 CET 2006


>>>>> "BertG" == Berton Gunter <gunter.berton at gene.com>
>>>>>     on Fri, 20 Jan 2006 14:15:49 -0800 writes:

    BertG> I'm not sure what you mean by a "fractional
    BertG> factorial" design here. 

it's a pretty well known term, in (at least some schools of)
classical analysis of variance, e.g. 
in the famous "Box, Hunter^2" (1987) book which had a 2nd
edition last year (Wiley), see also Doug Bates's nice account on
a talk about it,
  http://finzi.psych.upenn.edu/R/Rhelp02a/archive/29768.html
and the R package  "BHH2"  which is devoted to the book (2nd Ed).

BHH2 has a function  ffDesMatrix()  for fractional factorial
designs, *and* in it's see also section a link to the
conf.design() function in the package of the same name which
allows to construct ff designs (via different specifications).
see e.g, http://finzi.psych.upenn.edu/R/library/BHH2/html/ffDesMatrix.html

    BertG> In general, when dealing with complex designs of the sort you appear to be thinking
    BertG> about, small orthogonal designs don't exist. You
    BertG> might wish to look at the AlgDesign package to
    BertG> generate an efficient design to estimate only main
    BertG> effects.

    BertG> -- Bert Gunter Genentech Non-Clinical Statistics
    BertG> South San Francisco, CA
 
    BertG> "The business of the statistician is to catalyze the
    BertG> scientific learning process."  - George E. P. Box
 
now, since you have this nice signature, maybe you should
get the above book? ;-) :-)
 

    >> -----Original Message----- From:
    >> r-help-bounces at stat.math.ethz.ch
    >> [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of
    >> Roberto Furlan Sent: Friday, January 20, 2006 12:37 PM
    >> To: r-help at stat.math.ethz.ch Subject: [R] fractional
    >> factorial design in R
    >> 
    >> Hi, i need to create a fractional factorial design
    >> sufficient to estimate the main effects.  The factors may
    >> have any number of levels, let's say any number from 2 to
    >> 6.  I've tried to use the library conf.design , but i
    >> cannot figure out how to write the code.  For example,
    >> what is the code for a design with 5 factors (2x3x3x5x2)
    >> and only main effects not confounded?
    >> 
    >> thanks in advance!  Roberto Furlan University of Turin,
    >> Italy




More information about the R-help mailing list