The maximum entropy method, or MAXENT, is variational approach for computing probability
distributions given a list of moment, or expected value, constraints.
Here are some links for background info.
A good overview of applications:
http://cmm.cit.nih.gov/maxent/letsgo.html
On the idea of maximum entropy in general:
http://en.wikipedia.org/wiki/Principle_of_maximum_entropy
Use this package to compute discrete maximum entropy distributions over a list of values and
list of constraints.
Here is a the example from Probability the Logic of Science
> maxent 0.00001 [1,2,3] [average 1.5]
Right [0.61, 0.26, 0.11]
The classic dice example
> maxent 0.00001 [1,2,3,4,5,6] [average 4.5]
Right [.05, .07, 0.11, 0.16, 0.23, 0.34]
One can use different constraints besides the average value there.
