Closed jasondk closed 3 years ago
I think 'p' in this model represents the prior probability of starting in each submodel. This should usually be 1 for the first model and 0 for the rest. But other values should be allowed (I think) as long as they sum to 1 (?)
(The current version allows the p's to sum to something else, and it seems to effect the results)
I think 'p' in this model represents the prior probability of starting in each submodel. This should usually be 1 for the first model and 0 for the rest. But other values should be allowed (I think) as long as they sum to 1 (?)