Probabilistic Sensitivity Analysis (PSA) is not working properly in networks with imposed policies and pre-resolution evidence

[Jorge Pérez]

PSA must work in the attached network. When removed all policies and evidence from the network, PSA works fine (at least it return some results) but, when imposing policies it crashes.

[Jorge Pérez]

PSA imposes its own scenario and it can collide with the imposed policies (OpenMarkov tries to be consistent with the evidence but not with the policies at this moment). I think this issue must be revised in September because it implies to well define how to combine the preresolution evidence, the imposed policies and the choosen scenario.