This directory contains data files, command files, and generation
programs for the examples of Bayesian learning for finite and
countably-infinite mixture models and for Dirichlet diffusion tree
models.  To play with the data as in the manual, all you need to use
are the data and command files (eg, rdata and rcmds.inf).  If you want
to change the way the data was generated, you'll need to recompile the
programs, which is done with a command such as the following:

   cc rgen.c rand.c -DRAND_FILE=\"../util/randfile\" -lm -o rgen

The data in 'sdata' and commands in the scmds-* files are for the test
in my tech report on "Markov chain sampling methods for Dirichlet
process mixture models".
