Something is wrong with the obtained likelihood scores. For example, if the sum of edges in a subtree is zero, the log likelihood becomes infinity. Also, there seems to be an inconsistency between the scores we see (normal numbers) and the scores we would expect in that case (infinity).
Something is wrong with the obtained likelihood scores. For example, if the sum of edges in a subtree is zero, the log likelihood becomes infinity. Also, there seems to be an inconsistency between the scores we see (normal numbers) and the scores we would expect in that case (infinity).