DIST-GEN:  Generate values for state variables from the prior.

Dist-gen can be used to sample from the prior distribution of a
Bayesian model.  This capability can be used to initialize a Markov
chain sampler.  Looking at a sample from the prior can also be of
interest in itself.  Dist-gen cannot be used if the distribution
specification does not have the form of a Bayesian prior and
likelihood.

Usage:

    dist-gen log-file [ max-index ]

Records of state variables with indexes from zero up to the indicated
index are generated by sampling from the prior.  The default is
max-index of zero.  If the log file already contains some such
records, only indexes greater than the last one in the log file are
generated.  

Note that if the -read-prior option was given to dist-spec, the points
generated will be read from standard input, in alphabetical order -
ie, "u0","u1", ..., "u9", "u", "v0", ... (but note that "u" follows
"u9", unlike normal alphabetical order).  If -read-prior was not
given, the prior must be of a form allowing for automatic sampling.
See dist-spec.doc and dist-mc.doc for details.

            Copyright (c) 1995-2004 by Radford M. Neal