DIST-DGEN:  Generate values for target variables using given parameters

Dist-dgen randomly samples from the distribution for targets values
given parameter values for a Bayesian model.  Target values are
generated for each set of input values in a data file, and are written
to another data file.  This may be of interest in itself, and can also
be used as part of Geweke's procedure for testing the correctness of
Markov chain sampling methods (see xxx-mc-test.doc).  Dist-dgen cannot
be used if the distribution specification does not have the form of a
Bayesian prior and likelihood.

Usage:

    dist-dgen log-file [ index ] / [ inputs ] targets-file

State variables (model parameters) are read from the record in the log
file with the specified index.  For each line of input (specified as
described in numin.doc), the input variables, i0 (=i), i1, ..., i9,
are set to the values from this line (up to the number present), and
values for all the target variables, t0 (=t), t1, ..., t9, that are
mentioned in the likelihood specification are then randomly generated,
and written to the targets file, whose previous contents are
overwritten.  The random number seed used is that saved for the index
specified.  The likelihood must have a form that allows for automatic
sampling (see dist-spec.doc for details).

The default if no inputs are specified is the file of inputs for
training cases as specified with data-spec.  The default if no index
is specified is the last index in the log file.  When the last index
is used by default, the random number state is appended to the file
after being updated during generation of the targets.

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