box_behnken


box_behnken, a MATLAB code which returns a set of sample argument values to be used to analyze the behavior of a function.

Given DIM_NUM parameters, each of which is prescribed to lie between a minimum and maximum range, and a function taking these parameters as arguments, this program returns a table of sets of argument values that may be used to sample the function's behavior.

One sample point is the average. All other sample points are generated by setting a single spatial coordinate to its average value, and all other spatial coordinates to either the minimum or maximum. In this way, 1 + DIM_NUM*2**(DIM_NUM-1) points are generated.

Licensing:

The computer code and data files described and made available on this web page are distributed under the MIT license

Languages:

box_behnken is available in a C version and a C++ version and a FORTRAN90 version and a MATLAB version.

Related Data and Programs:

box_behnken_test

cvt, a MATLAB code which computes elements of a Centroidal Voronoi Tessellation.

faure, a MATLAB code which computes elements of a Faure sequence.

halton, a MATLAB code which computes elements of a Halton sequence.

hammersley, a MATLAB code which computes elements of a Hammersley sequence.

latin_center, a MATLAB code which computes elements of a Latin Hypercube dataset, choosing center points.

latin_edge, a MATLAB code which computes elements of a Latin Hypercube dataset, choosing edge points.

latin_random, a MATLAB code which computes elements of a Latin Hypercube dataset, choosing points at random.

latinize, a MATLAB code which can "latinize" a dataset.

niederreiter2, a MATLAB code which computes elements of a Niederreiter sequence using base 2.

sobol, a MATLAB code which computes elements of a Sobol sequence.

uniform, a MATLAB code which computes elements of a uniform pseudorandom sequence.

van_der_corput, a MATLAB code which computes elements of a (1 dimensional) van der Corput sequence.

Reference:

  1. George Box, Donald Behnken,
    Some new three level designs for the study of quantitative variables,
    Technometrics,
    Volume 2, pages 455-475, 1960.

Source Code:


Last revised on 02 December 2018.