plot.bsgw {BSGW} | R Documentation |
Four sets of MCMC diagnostic plots are currently generated: 1) log-likelihood and log-posterior (including shrinkage effect) as a function of iteration number, 2) coefficient trace plots, 3) coefficient autocorrelation plots, 4) coefficient histograms.
## S3 method for class 'bsgw'
plot(x, pval=0.05, burnin=round(x$control$iter/2), nrow=2, ncol=3, ...)
x |
A |
pval |
The P-value at which lower/upper bounds on coefficients are calculated and overlaid on trace plots and historgrams. |
burnin |
Number of samples discarded from the beginning of an MCMC chain, after which parameter quantiles are calculated. |
nrow |
Number of rows of subplots within each figure, applied to plot sets 2-4. |
ncol |
Number of columns of subplots within each figure, applied to plot sets 2-4. |
... |
Further arguments to be passed to/from other methods. |
Alireza S. Mahani, Mansour T.A. Sharabiani
library("survival")
data(ovarian)
est <- bsgw(Surv(futime, fustat) ~ ecog.ps + rx, ovarian
, control=bsgw.control(iter=400, nskip=100))
plot(est)