carpet(data,D,...)
cube(data,D,...)carpet objects and
of length 3 for cube objects, specifying the number of doses of the components drugs in the trial.carpet or cube, respectively, with
the following slots.
carpet creates objects of class carpet from the
specified data in the list that are used row-wise to fill up
the 2-factorial treatment groups, i.e. in the order
(0,0), (0,1),..., (0,D[2]), (1,0), ..., (1,D[2]), ..., (D[1],D[2]);
resulting in a (D[1]+1)x(D[2]+1) data array.
To represent trifactorial designs for the evaluation of a
three-compound combination, an object of class cube can be
created using the function cube. The data in the treatment
groups are then filled up in the order (0,0,0), ..., (0,0,D[3])
first, then (0,1,0), ..., (0,1,D[3]) and up to
(0,D[2],0), ..., (0,D[2],D[3]). This is the order also for
the values 0, ..., D[1] for the first component group, always
taking the data succesively from the list elements of data. The
result is a (D[1]+1)x(D[2]+1)x(D[3]+1) data array. Methods for
multiple inference and global tests can
be applied to carpet and cube objects.
Hung HMJ, Chi GYH, Lipicky RJ (1993): Testing for the existence of a desirable dose combination. Biometrics 49, pp. 85-94
Hung HMJ (2000): Evaluation of a combination drug with multiple doses in unbalanced factorial design clinical trials. Statistics in Medicine 19, pp. 2079-2087
bifactorial, mintest,
margint, avetest, maxtest#Hypertension example from Hung (2000)
data(sidbp)
x<-split(sidbp$ynrmhom,sidbp$cb)
bifactorial<-carpet(data=x,D=c(2,3))
Run the code above in your browser using DataLab