Added tolerance-based termination

src/attributes.py

@@ -49,6 +49,7 @@ class Attributes:
             current_fitness             = 0,
             generation_goal             = 15,
             fitness_goal                = None,
+            tolerance_goal              = 1e-3,
             chromosome_mutation_rate    = 0.15,
             gene_mutation_rate          = 0.03,
             initialization_impl         = Initialization_Methods.random_initialization,
@@ -93,6 +94,7 @@ class Attributes:
         self.current_fitness    = deepcopy(current_fitness)
         self.generation_goal    = deepcopy(generation_goal)
         self.fitness_goal       = deepcopy(fitness_goal)
+        self.tolerance_goal     = deepcopy(tolerance_goal)
 
         # Mutation variables
         self.chromosome_mutation_rate = deepcopy(chromosome_mutation_rate)

src/termination_point/termination_methods.py

@@ -12,12 +12,31 @@ class Termination_Methods:
         if ga.fitness_goal is not None:
 
             # If minimum fitness goal reached, stop ga.
-            if ga.target_fitness_type == 'min' and ga.get_chromosome_fitness(0) >= ga.convert_fitness(ga.fitness_goal):
+            if ga.target_fitness_type == 'min' and ga.population.get_chromosome(0).get_fitness() <= ga.fitness_goal:
                 return False
 
             # If maximum fitness goal reached, stop ga.
-            elif ga.target_fitness_type == 'max' and ga.get_chromosome_fitness(0) >= ga.convert_fitness(ga.fitness_goal):
+            elif ga.target_fitness_type == 'max' and ga.population.get_chromosome(0).get_fitness() >= ga.fitness_goal:
                 return False
 
         # If generation goal is set, check it.
-        return ga.generation_goal is None or ga.current_generation < ga.generation_goal
+        if ga.generation_goal is not None and ga.current_generation >= ga.generation_goal:
+            return False
+
+        # If tolerance is set, check it.
+        if ga.tolerance_goal is not None:
+
+            best_fitness = ga.population.get_chromosome(0).get_fitness()
+            convergence_count = 0
+            tol = ga.tolerance_goal * (1 + abs(best_fitness))
+
+            # Find out how many chromosomes have converged
+            for chromosome in ga.population.get_chromosome_list():
+                if abs(best_fitness - chromosome.get_fitness()) < tol:
+                    convergence_count += 1
+
+            # Terminate if 10% of the population has converged
+            if convergence_count > 0.1*ga.population.size():
+                return False
+
+        return True