code simplification

Author: coverdrive

rl/finite_horizon.py

@@ -145,10 +145,9 @@ def finite_horizon_MDP(
     for time in range(0, limit):
         for s in process.non_terminal_states:
             s_time = WithTime(state=s.state, time=time)
-            actions_map = process.action_mapping(s)
             mapping[s_time] = {a: result.map(
                 lambda sr: (WithTime(state=sr[0].state, time=time + 1), sr[1])
-            ) for a, result in actions_map.items()}
+            ) for a, result in process.mapping[s].items()}
 
     return FiniteMarkovDecisionProcess(mapping)
 
@@ -182,7 +181,7 @@ def without_time(arg: ActionMapping[A, WithTime[S]]) -> \
                 for a, sr_distr in arg.items()}
 
     return [{NonTerminal(s.state): without_time(
-        process.action_mapping(NonTerminal(s))
+        process.mapping[NonTerminal(s)]
     ) for s in states} for _, states in groupby(
         sorted(
             (nt.state for nt in process.non_terminal_states),

rl/markov_decision_process.py

@@ -172,9 +172,6 @@ def apply_finite_policy(self, policy: FinitePolicy[S, A])\
 
         return FiniteMarkovRewardProcess(transition_mapping)
 
-    def action_mapping(self, state: NonTerminal[S]) -> ActionMapping[A, S]:
-        return self.mapping[state]
-
     def actions(self, state: NonTerminal[S]) -> Iterable[A]:
         '''All the actions allowed for the given state.
 

rl/test_finite_horizon.py

@@ -148,7 +148,7 @@ def test_finite_horizon_MDP(self):
             self.assertEqual(set(finite.actions(s)), {False, True})
 
         start = NonTerminal(WithTime(state=True, time=0))
-        result = finite.action_mapping(start)[False]
+        result = finite.mapping[start][False]
         expected_result = Categorical({
             (NonTerminal(WithTime(False, time=1)), 2.0): 0.7,
             (NonTerminal(WithTime(True, time=1)), 1.0): 0.3
@@ -186,7 +186,7 @@ def action_mapping_for(s: WithTime[bool]) -> \
                 for a in True, False:
                     distribution.assert_almost_equal(
                         self,
-                        finite.action_mapping(NonTerminal(s_time))[a],
+                        finite.mapping[NonTerminal(s_time)][a],
                         action_mapping_for(s_time)[a]
                     )
 
@@ -211,7 +211,7 @@ def action_mapping_for(s: WithTime[bool]) -> \
             for a in True, False:
                 distribution.assert_almost_equal(
                     self,
-                    finite.action_mapping(NonTerminal(s_time))[a],
+                    finite.mapping[NonTerminal(s_time)][a],
                     act_map[a]
                 )