1Blurb::  Use the multilevel MCMC algorithm.
2
3Description::
4Selects the multilevel algorithm described in \cite Prudencio.
5
6<b> Default Behavior </b>
7
8Five MCMC algorithm variants are supported: \c dram, \c
9delayed_rejection, \c adaptive_metropolis, \c metropolis_hastings, and
10\c multilevel.  The default is \c dram.
11
12<b> Usage Tips </b>
13
14The multilevel algorithm is a more experimental algorithm than the
15other MCMC approaches mentioned above.  It works well in cases where
16the prior can be "evolved" to a posterior in a structured way.
17Currently, the multilevel option is not in production form.
18
19Topics::	bayesian_calibration
20Examples::
21\verbatim
22method,
23        bayes_calibration queso
24          multilevel
25          samples = 10000 seed = 348
26\endverbatim
27
28Theory::
29Faq::
30See_Also::
31