Table 2.
Benchmark of the Average Computational Time Required for the DAP and Exact Computation
| Method |
Running Time (s) |
||||
|---|---|---|---|---|---|
|
Number of True QTLs | |||||
| 1 | 2 | 3 | 4 | 5 | |
| DAP (λ = 0.01) | 0.097 (0.234) | 0.275 (1.180) | 0.733 (3.704) | 1.276 (7.140) | 2.527 (13.181) |
| DAP (λ = 0.02) | 0.093 (0.268) | 0.208 (0.776) | 0.663 (3.128) | 1.275 (6.816) | 2.368 (12.965) |
| DAP (λ = 0.03) | 0.087 (0.238) | 0.133 (0.408) | 0.252 (1.060) | 0.844 (4.644) | 1.422 (7.876) |
| DAP (λ = 0.04) | 0.063 (0.116) | 0.122 (0.312) | 0.230 (0.732) | 0.615 (3.064) | 0.571 (2.596) |
| DAP (λ = 0.05) | 0.050 (0.072) | 0.120 (0.280) | 0.139 (0.320) | 0.184 (0.448) | 0.180 (0.276) |
| Exact | 19.8 (121.4) | ||||
The running time is measured in seconds by the UNIX utility program “time.” In each cell, we show the actual running time (“real” time), which is greatly reduced by parallel processing with ten threads; in the parentheses, the “user” time is reported, which objectively reflects the actual computational cost, i.e., this measurement is not reduced by the parallelization.