Next: confound2()
Up: Design Macros Help File
Previous: choosedef2()
Contents
Usage:
choosegen2(k,p, all:T [,res:r]), positive integers k, p, r.
choosegen2(k,p, tries:m [,res:r]), positive integers k, p, m, and r.
|
Keywords:
aliasing, design, factorial
choosegen2(k,p,all:T) , where k, p and r are positive integers, searches
all potential sets of generators for a 2^(k-p) fractional factorial
design. It returns information on a generator with maximum resolution.
choosegen2(k,p, tries:m), where m > 0 is an integer, does the same,
except it searches only m randomly selected sets of generators.
choosegen2(k,p,all:T,res:r) and choosegen2(k,p,tries:m,res:r), where r
is a positive integer, do the same, except they stop the search at the
first design of resolution r. If none is found, they return information
on a set having the highest resolution found.
The value returned is
structure(resolution:bestres,generators:bestgen,aberration:aber,\
basis:genmat)
Integer bestres > 0 is the best resolution found.
bestgen is a CHARACTER vector of length p with the names such as "ABDF"
of the defining contrasts in the set chosen.
aber is a vector of k integers >= 0 with ab[i] containing the number
of contrast with i letters that are confounded with "I".
genmat is a p by k matrix with entries 0 and 1, one row for each element
of bestgen. genmat[i,j] is 1 if and only if factor j is bestgen[i].
For example, if bestgen[i] is "ABDF", genmat[i,] is vector(1,1,0,1,0,1
[,...])'. genmat can be used as an argument to aliases2() to determine
all the aliases of any contrast.
Examples:
Cmd> choosegen2(5,2,all:T) # find best resolution
component: resolution
(1) 3
component: generators
(1) "ABCD"
(2) "BCE"
component: aberration
(1) 0 0 2 1 0
component: basis
(1,1) 1 1 1 1 0
(2,1) 0 1 1 0 1
Cmd> choosegen2(5,2,all:T,res:4) # try to find 4 (we won't)
component: resolution
(1) 3
component: generators
(1) "ABCD"
(2) "BCE"
component: aberration
(1) 0 0 2 1 0
component: basis
(1,1) 1 1 1 1 0
(2,1) 0 1 1 0 1
Cmd> # look for 2^(9-4) resol. 4, just make 1000 tries, since\
there are lots of combinations to explore
Cmd> choosegen2(9,4,tries:1000,res:4)# look for 2^(9-4) resol. 4
component: resolution
(1) 4
component: generators
(1) "CDEF"
(2) "BDEG"
(3) "ABCEH"
(4) "BCDJ"
component: aberration
(1) 0 0 0 7 7
(6) 0 0 0 1
component: basis
(1,1) 0 0 1 1 1
(1,6) 1 0 0 0
(2,1) 0 1 0 1 1
(2,6) 0 1 0 0
(3,1) 1 1 1 0 1
(3,6) 0 0 1 0
(4,1) 0 1 1 1 0
(4,6) 0 0 0 1
See also aliases2(), choosedef2(), confound2().
Gary Oehlert
2003-01-15