Wynn finds an optimal design using a sensitivity function and a Wynn-algorithm.
Usage
Wynn(sensF, tol, maxIter = 10000)
Arguments
sensF
function(x=NULL, desw=NULL, desx=NULL, mod=NULL), a sensitivity function.
It's attribute "defaults" shall contain identical x and desx, and sensF(desw=w) shall return sensitivities corresponding to each point in x.
tol
the tolerance level regarding the sensitivities.
maxIter
the maximum number of iterations.
Value
Wynn returns an object of class"desigh".
See design for its structural definition.
Details
See Dsensitivity and it's return value for a reference implementation of a function complying with the requirements for sensF.
The algorithm starts from a uniform weight design.
In each iteration weight is redistributed to the point which has the highest sensitivity.
Sequence: 1/i.
The algorithm stops when all sensitivities are below a specified tolerance level or the maximum number of iterations is reached.
References
Wynn, Henry P. (1970) The Sequential Generation of D-Optimum Experimental Designs.
The Annals of Mathematical Statistics, 41(5):1655-1664.