Class C(20, 6)

Conference designs C(20, 6)

This page contains information about conference designs. The results have been generated with the Orthogonal Array package. If you use these results, please cite the paper A Classification Criterion for Definitive Screening Designs, E.D. Schoen, P.T. Eendebak, P. Goos, Annals of Statistics, 2019.

Results

Statistic Results
Number of non-isomorphic designs 219
Minimum/Maximum rank of model matrix with 2FI and QE 13/20
Minimum/Maximum rank of model matrix for 2FI 12/15
Minimum B4 0.3600
Maximum B4 1.5600
Minimum F4 (0, 0, 0, 9, 6)
Maximum F4 (0, 3, 0, 12, 0)
Data all arrays

Note: 2FI: two-factor interaction; QE: quadratic effect


Pareto optimal designs

There are 6 Pareto optimal designs in 6 classes. Pareto optimality is according to r(2FI, QE), r(2FI), F4, B4, PEC4, PEC5, PIC4, PIC5, PPC4, and PPC5 (any other statistics are ignored).

All Pareto optimal designs.

Index Pareto file Index design file Rank 2FI and QE Rank 2FI F4 B4 PEC4 PEC5 PIC4 PIC5 PPC4 PPC5 Extensions Max. columns
3 195 19 15 (0, 0, 0, 9, 6) 0.36 1.000 1.000 17.174 16.880 1.798 1.277 Yes 20
1 159 20 15 (0, 0, 0, 10, 5) 0.40 1.000 1.000 17.158 16.665 1.792 1.208 Yes 20
0 112 20 15 (0, 0, 0, 11, 4) 0.44 1.000 1.000 17.143 16.770 1.786 1.221 Yes 20
2 178 20 15 (0, 0, 1, 9, 5) 0.52 1.000 1.000 17.109 16.717 1.794 1.276 Yes 20
4 199 19 15 (0, 0, 1, 9, 5) 0.52 1.000 1.000 17.109 16.735 1.794 1.294 Yes 20
5 208 19 15 (0, 0, 3, 9, 3) 0.84 1.000 1.000 16.980 16.375 1.786 1.307 Yes 20


Page generated on Tue Oct 22 20:35:41 2019.