Embedded Probabilistic Programming
		joint work with Chung-chieh Shan

embedpp.pdf		Paper presented at the 25th conference
    			on uncertainty in artificial intelligence (UAI 2009)

	The library code

ptypes.mli	Library type declarations

probM.mli	The interface of the library for probabilistic programming:
		the definition of the embedded probabilistic DSL
probM.ml	The implementation of the library/EDSL
		The `dist' function, evidence assertion, reification and
		reflection, call-by-need evaluation, sharing and memoization.

inference.mli	Exact and approximate inference procedures, explorers of
inference.ml	the lazy search tree, the result of reifying of the
		probabilistic program


		Tests and sample applications
exactInfM.ml		Exact inference, comparison with IBAL
samplingM.ml		Rejection sampling; importance sampling by
			evidence look-ahead (aka, evidence pulling)
slazy.ml		Lazy evaluation, sharing, probabilistic
			context-free grammars, memoization of stochastic
			functions

		The IBAL music evolution model
music1.ml		The warm-up: finite evolution, can be solved exactly
music1a.ml		The warm-up: the same with memoization
music2.ml		The full model, potentially infinite evolution, and
			very low probability of evidence

		Nested inference
nested.ml		Several examples and the regression tests

		The alternative implementation of the library,
		relying on a stream of samples
probN.mli		The interface
probN.ml		The implementation
exactInfN.ml		Tests of exact inference
samplingN.ml		Tests of importance sampling


Makefile	How to build it all

		Benchmarks, from:
		Comparative Study of Probabilistic Logic Languages and Systems
		http://www.cs.aau.dk/~jaeger/plsystems/models.html

bloodtype/		Blood type
HMM/			Hidden Markov model
noisy-or/		Noisy OR

	       Models with indefinite number of objects
blip/			The blip example

June 2009.