TABLE 1.
s/ca | True θ, λ | θ | σθ | % coverage | λ | σλ | % coverage | μ | σμ | % coverage |
---|---|---|---|---|---|---|---|---|---|---|
c | 0.05, 2 | 0.05124 | 0.008339 | 92 | 2.278 | 1.579 | 100 | — | — | — |
c | 0.05, 200 | 0.04976 | 0.007293 | 100 | 248.7 | 135.44 | 100 | — | — | — |
s | 0.05, 2 | 0.04937 | 0.008226 | 92 | 1.832 | 1.270 | 100 | 0.9723 | 0.06115 | 100 |
s | 0.05, 200 | 0.04832 | 0.006364 | 92 | 194.7 | 98.9 | 96 | 0.972 | 0.05112 | 96 |
N1, λ1 → 2, s | 0.1, 1 | 0.1006 | 0.02805 | 90 | 1.078 | 1.039 | 100 | 1.013 | 0.4470 | 100 |
N2, λ2 → 1, s | 0.2, 10 | 0.2054 | 0.04405 | 90 | 11.04 | 3.343 | 100 | |||
N1, λ1 → 2, s | 0.2, 1 | 0.2111 | 0.05082 | 92 | 1.509 | 1.939 | 96 | 0.9935 | 0.05762 | 92 |
N2, λ2 → 1, s | 0.1, 10 | 0.1001 | 0.02510 | 100 | 12.13 | 4.559 | 100 |
Where serial sequences are used, μ is included. All run lengths were three million states while some runs were longer; see the text for details. The last four rows show the asymmetric simulations.
Serial samples (s); contemporary samples (c).