Hi, Infer.net people,

From the example in the website, the probability of satisfying the constraint is computed by Math.Exp(logEvidence).

The result seems a oddRatio to me rather than a probability.

http://research.microsoft.com/en-us/um/cambridge/projects/infernet/docs/Computing%20model%20evidence%20for%20model%20selection.aspx

Anyone can help me to understand this?

Thank you very much.