Wrong results when evaluating an influence diagram

[Francisco Javier Díez]

When evaluating this network, the probability values for negative and not done are swapped.

[Francisco Javier Díez]

First: the results are not swapped, as I initially thought. Second, I have seen that in the network Ejerc-3.4.pgmx, the probability for the node Test result is given by a tree. After converting manually the tree into a table (see the file Ejerc-3.4-table.pgmx), OpenMarkov gives the right result. It seems that the problem is not in the inference algorithm but in the conversion of the tree into a table.

[maryebra]

Resolved