jpt.inference.mpe
=================

.. py:module:: jpt.inference.mpe


Classes
-------

.. autoapisummary::

   jpt.inference.mpe.MPEState
   jpt.inference.mpe.BnBNode
   jpt.inference.mpe.BnBSolution
   jpt.inference.mpe.MPESolver


Module Contents
---------------

.. py:class:: MPEState(assignment: jpt.variables.VariableMap)

   .. py:attribute:: assignment


   .. py:attribute:: latest_variable
      :value: None



   .. py:method:: copy()


   .. py:method:: assign(variable: Variable or str, value: Any) -> MPEState


   .. py:method:: __repr__()


.. py:class:: BnBNode(solver, state, actual_cost)

   .. py:attribute:: solver
      :type:  MPESolver


   .. py:attribute:: old_state
      :type:  MPEState


   .. py:attribute:: variable
      :type:  jpt.variables.Variable
      :value: None



   .. py:attribute:: values
      :type:  Iterator[Any]


   .. py:attribute:: actual_cost


   .. py:attribute:: step_cost
      :value: 0



   .. py:attribute:: state
      :type:  MPEState
      :value: None



   .. py:method:: _generate_next()


   .. py:method:: pop()


   .. py:property:: cost


   .. py:property:: free_variables


.. py:class:: BnBSolution(state: MPEState, cost: float)

   .. py:attribute:: state


   .. py:attribute:: cost


.. py:class:: MPESolver(distributions: jpt.variables.VariableMap, likelihood_divisor: float = None)

   Solver for k-MPE inference about n independent variables

   This algorithm iteratively constructs all ``k`` most probable explanations
   in descending order by performing a branch-and-bound search in the space of
   atomic areas of the respective distributions. In constrast to classic BnB search,
   we do not throw away the pruned states but save them in a priority queue, where
   we can continue the search for the subsequent 2nd, 3rd, ..., k-th best solution.


   .. py:attribute:: distributions
      :type:  jpt.variables.VariableMap


   .. py:attribute:: domains


   .. py:attribute:: constraints


   .. py:attribute:: lb


   .. py:method:: likelihood(solution: BnBSolution)


   .. py:method:: is_goal_state(state: MPEState) -> bool


   .. py:method:: variable_order(state: MPEState = None) -> Iterator[jpt.variables.Variable]

      :param state: The state to start from
      :return: An iterator over free variables for a state sorted by number of different possible states
          (lowest amount of values first).



   .. py:method:: solve(k: int = 0) -> Iterator[List[jpt.variables.ValueAssignment]]

      Generate ``k`` MPE states with decreasing probability.

      :return: