Table 1.
Clustering results for the Gap Procedure on simulated data
| Data | Average | |||||
|---|---|---|---|---|---|---|
| Sim | N | G | Time (in sec) | # clusters | # singletons | ARI |
| 1 | 100 | 4 | 0.1108 | 4.25 | 0.04 | 0.9854 |
| 2 | 150 | 6 | 0.1370 | 6.39 | 0.04 | 0.9856 |
| 3 | 500 | 20 | 0.6073 | 22.49 | 0.13 | 0.9750 |
| 4 | 1250 | 50 | 6.6194 | 58.11 | 0.43 | 0.9694 |
The average clustering results (taken over 100 runs) obtained by the Gap Procedure when applied to the simulated data. The dissimilarity matrix was calculated using the aK80 distance formula and sequences (of length 800) were mutated according to a GTR + I + Γ model