kscheck: Convergence assessment for the pump failure data
Description
This function produces graphs and one or several samples from
the Gibbs sampler applied to the benchmark of the pump failure dataset
(given within the function). It also provides convergence assessments
in connection with the coda package.
Usage
kscheck(T = 10^3, heidel = FALSE)
Arguments
T
Number of Gibbs iterations
heidel
Boolean indicating whether or not several parallel chains need to be simulated
Value
Depending on the value of heidel, the function returns two plots of Kolmogorov-Smirnov
statistics comparing two different scales or the outputs of heidel.diag and geweke.diag
functions. It also returns as a list the sequence(s) of $beta$ and $lambda$.
References
Chapter 8 of EnteR Monte Carlo Statistical Methods