「ぎょーむ日誌」目次に戻る | KuboWeb top に戻る | twilog | atom

ぎょーむ日誌 2007-09-08

苦情・お叱りは, たいへんお手数かけて恐縮ですが, 久保 (kubo@ees.hokudai.ac.jp) までお知らせください.

2007 年 09 月 08 日 (土)

mcmc08a

Inference for Bugs model at "/home/kubo/ecoasZP/winbugs/model.bug.txt", fit using winbugs,
 3 chains, each with 16000 iterations (first 10000 discarded), n.thin = 30
 n.sims = 600 iterations saved
                mean     sd    2.5%     25%     50%     75%   97.5% Rhat n.eff
beta.f[1]      -0.71   0.45   -1.53   -1.04   -0.73   -0.41    0.18 1.00   600
beta.f[2]       0.89   0.25    0.40    0.73    0.89    1.04    1.41 1.02   130
beta.f[3]      -0.16   0.15   -0.46   -0.24   -0.16   -0.06    0.11 1.01   450
beta.f[4]       1.34   0.13    1.10    1.26    1.34    1.43    1.57 1.01   600
beta.z[1]      -0.12   0.00   -0.12   -0.12   -0.12   -0.12   -0.12 1.00     1
beta.z[2]       2.43   0.00    2.43    2.43    2.43    2.43    2.43 1.07    64
beta.z[3]       0.50   0.36   -0.25    0.28    0.51    0.73    1.15 1.03   130
beta.z[4]       0.20   0.12   -0.02    0.12    0.21    0.28    0.43 1.03    94
beta.c[1]      -0.10   0.21   -0.59   -0.21   -0.08    0.04    0.21 1.02   170
beta.c[2]       0.64   0.54   -0.24    0.24    0.57    0.93    1.81 1.01   280
beta.c[3]       0.25   0.07    0.14    0.20    0.25    0.30    0.41 1.01   390
beta.c[4]       0.00   0.01   -0.02    0.00    0.00    0.01    0.02 1.04    66
beta.c[5]       2.72   0.72    1.33    2.29    2.75    3.14    4.18 1.01   190
tau.err[1]     30.19  33.70    1.13    6.32   17.88   41.49  127.30 1.00   600
tau.err[2]    941.65 829.95   93.97  341.90  683.70 1308.25 3069.92 1.00   600
tau.err[3]     54.72 175.42    0.05    0.24    2.75   19.75  665.88 1.02   110
tau.err[4]     54.34 223.75    0.00    0.00    0.11    8.58  537.04 1.21    14
tau.err[5]    150.44 334.49    3.28   14.22   41.55  129.35  910.75 1.00   600
tau.delta[1]    0.14   0.03    0.08    0.11    0.13    0.16    0.20 1.00   600
tau.delta[2]    0.33   0.15    0.17    0.24    0.29    0.37    0.64 1.02   420
tau.delta[3]    5.59  24.88    0.61    0.94    1.49    3.36   25.40 1.03   120
tau.delta[4]  108.36 258.86    0.51    1.19    5.65   75.45  907.25 1.03   110
deviance     -250.71 140.08 -495.42 -350.30 -263.80 -159.12    9.43 1.01   320

For each parameter, n.eff is a crude measure of effective sample size,
and Rhat is the potential scale reduction factor (at convergence, Rhat=1).

pD = 282.3 and DIC = 31.6 (using the rule, pD = Dbar-Dhat)
DIC is an estimate of expected predictive error (lower deviance is better).

KuboLog | KuboWeb