BooksNúñez Ares and Goos · 2020
Enumeration and Multicriteria Selection of Orthogonal Minimally Aliased Response Surface Designs
- Matches
- Close
- Differs
Factorial gives some of the paper's answers. Where they differ, this page says why.
Examples
Each example, and how closely Factorial gives its answers.
Pages 1 to 99
- Central composite designs (CCD) without centre runs, for 3 to 7 factorsp. 22Matches
- Small central composite designs (SCCD) without centre runs, for 3 to 7 factorsp. 22Matches
- Box-Behnken designs (BBD) without centre runs, for 3 to 7 factorsp. 22Matches
- Definitive screening designs (DSD) without centre runs, for 3 to 7 factorsp. 22Matches
- Design 1: bd.5.22.6.10.154.0, a five-factor basic OMARS design in 22 runspp. 25-28Differs
- Design ξ1: foldover of the printed weighing matrix W (order 7, weight 4), basic design bd.7.14.6.10.1pp. 27-28Differs
- Design ξ2: foldover of the printed weighing matrix W (order 7, weight 4), basic design bd.7.14.6.10.1, with one centre runpp. 27-28Differs
- Design ξ3: foldover of the printed weighing matrix W (order 7, weight 4), basic design bd.7.14.6.10.1, with two centre runspp. 27-28Differs
- Design ξ4: the 7-factor definitive screening design, basic design bd.7.16.2.4.2: the foldover of a conference matrix of order 8 with one column droppedpp. 27-28Differs
- Design ξ5: the 7-factor definitive screening design, basic design bd.7.16.2.4.2: the foldover of a conference matrix of order 8 with one column dropped, with one centre runpp. 27-28Differs
- Design ξ6: the 7-factor definitive screening design, basic design bd.7.16.2.4.2: the foldover of a conference matrix of order 8 with one column dropped, with two centre runspp. 27-28Differs
- Design 2: bd.6.31.9.15.353.0, a six-factor basic OMARS design in 31 runspp. 28-29Matches
- Design 3: bd.7.42.10.18.28.1, a seven-factor foldover OMARS design in 43 runspp. 28-29Matches
- Design 4: bd.6.36.6.12.400.1, a six-factor non-foldover OMARS design in 37 runspp. 29-30, 33Close
- Design 5: bd.6.35.17.27.275.0, a six-factor non-foldover OMARS design in 35 runspp. 31-32Differs
- bd.6.35.17.27.275.1: Design 5 with one centre run added, 36 runspp. 31-32Differs
- Design 4 projected on its first four factorsp. 33Matches
- Design 4 projected on its first five factorsp. 33Matches
Not yet
Answers the paper prints that Factorial cannot compute yet, by what it lacks.
- The library does not enumerate the catalogue (generate_design('omars_ilp') searches the foldover family for one design)pp. 21, 24, 25 and more
- The pair (n0ME, n0IE): the number of zeros in a main-effect column and in an interaction column, centre runs left out; no check kind counts zeros, and generate_design makes no SCCD for 3 or 4 factors; a half-fraction cube has no zeros and each of the 2k axial runs has k - 1 factors at 0, which gives (4,6)p. 22
- The pair (n0ME, n0IE): the number of zeros in a main-effect column and in an interaction column, centre runs left out; no check kind counts zeros, and generate_design makes no SCCD for 3 or 4 factors; a half-fraction cube has no zeros and each of the 2k axial runs has k - 1 factors at 0, which gives (6,8)p. 22
- Generate_design('dsd') always includes the design's centre run: it makes 9 runs for 3 factors, the printed 8 plus that centre runp. 22
- Generate_design('dsd') always includes the design's centre run: it makes 9 runs for 4 factors, the printed 8 plus that centre runp. 22
Show the rest
- Generate_design('dsd') always includes the design's centre run: it makes 13 runs for 5 factors, the printed 12 plus that centre runp. 22
- Generate_design('dsd') always includes the design's centre run: it makes 13 runs for 6 factors, the printed 12 plus that centre runp. 22
- Generate_design('dsd') always includes the design's centre run: it makes 17 runs for 7 factors, the printed 16 plus that centre runp. 22
- No library function counts isomorphic copies; applying all 5! x 2^5 = 3840 permutations and sign switches to the printed matrix gives 240 distinct designs (numpy, while cataloguing)p. 25