Updated all the selector methods to use the decorator increase_counter.

Author: vrettasm

pygenalgo/operators/selection/boltzmann_selector.py

@@ -1,6 +1,7 @@
 from math import fsum
 from numpy import exp as np_exp
 from pygenalgo.genome.chromosome import Chromosome
+from pygenalgo.operators.genetic_operator import increase_counter
 from pygenalgo.operators.selection.select_operator import SelectionOperator
 
 
@@ -29,6 +30,7 @@ def __init__(self, select_probability: float = 1.0, k: float = 100.0):
         self._items = max(float(k), 50.0)
     # _end_def_
 
+    @increase_counter
     def select(self, population: list[Chromosome]):
         """
         Select the individuals, from the input population, that will be passed on
@@ -44,8 +46,8 @@ def select(self, population: list[Chromosome]):
         # Compute the Temperature.
         T = max(0.1, np_exp(-self.iter/self._items))
 
-        # Get the length of the population list.
-        N = len(population)
+        # Get the population size.
+        pop_size = len(population)
 
         # Extract the fitness value of each chromosome.
         # This assumes that the fitness values are all positive.
@@ -58,11 +60,9 @@ def select(self, population: list[Chromosome]):
         # in the population (Boltzmann distribution).
         selection_probs = [f/sum_fitness for f in exp_fitness]
 
-        # Select 'N' new individuals (indexes).
-        index = self.rng.choice(N, size=N, p=selection_probs, replace=True, shuffle=False)
-
-        # Increase the selection counter.
-        self.inc_counter()
+        # Select the new individuals indexes.
+        index = self.rng.choice(pop_size, size=pop_size, p=selection_probs,
+                                replace=True, shuffle=False)
 
         # Return the new parents (individuals).
         return [population[i] for i in index]

pygenalgo/operators/selection/linear_rank_selector.py

@@ -1,5 +1,7 @@
 from operator import attrgetter
+
 from pygenalgo.genome.chromosome import Chromosome
+from pygenalgo.operators.genetic_operator import increase_counter
 from pygenalgo.operators.selection.select_operator import SelectionOperator
 
 
@@ -25,6 +27,7 @@ def __init__(self, select_probability: float = 1.0):
         super().__init__(select_probability)
     # _end_def_
 
+    @increase_counter
     def select(self, population: list[Chromosome]):
         """
         Select the individuals, from the input population, that will be passed on
@@ -39,20 +42,18 @@ def select(self, population: list[Chromosome]):
         # Sort the population in ascending order using their fitness value.
         sorted_population = sorted(population, key=attrgetter("fitness"))
 
-        # Get the length of the population list.
-        N = len(sorted_population)
+        # Get the population size.
+        pop_size = len(population)
 
         # Calculate sum of all the ranked fitness values: "1 + 2 + 3 + ... + N".
-        sum_ranked_values = float(0.5 * N * (N + 1))
+        sum_ranked_values = float(0.5 * pop_size * (pop_size + 1))
 
         # Calculate the "selection probabilities", of each member in the population.
-        selection_probs = [n / sum_ranked_values for n in range(1, N + 1)]
-
-        # Increase the selection counter.
-        self.inc_counter()
+        selection_probs = [n / sum_ranked_values for n in range(1, pop_size + 1)]
 
         # Select 'N' new individuals (indexes).
-        index = self.rng.choice(N, size=N, p=selection_probs, replace=True, shuffle=False)
+        index = self.rng.choice(pop_size, size=pop_size, p=selection_probs,
+                                replace=True, shuffle=False)
 
         # Return the new parents (individuals).
         return [sorted_population[i] for i in index]

pygenalgo/operators/selection/random_selector.py

@@ -1,4 +1,5 @@
 from pygenalgo.genome.chromosome import Chromosome
+from pygenalgo.operators.genetic_operator import increase_counter
 from pygenalgo.operators.selection.select_operator import SelectionOperator
 
 
@@ -24,6 +25,7 @@ def __init__(self, select_probability: float = 1.0):
         super().__init__(select_probability)
     # _end_def_
 
+    @increase_counter
     def select(self, population: list[Chromosome]):
         """
         Select the individuals, from the input population, that will be passed on
@@ -34,14 +36,11 @@ def select(self, population: list[Chromosome]):
 
         :return: the selected parents population (as list of chromosomes).
         """
-        # Get the length of the population list.
-        N = len(population)
+        # Get the population size.
+        pop_size = len(population)
 
-        # Increase the selection counter.
-        self.inc_counter()
-
-        # Select 'N' new individuals (indexes).
-        index = self.rng.choice(N, size=N, replace=True, shuffle=False)
+        # Select the new individuals indexes.
+        index = self.rng.choice(pop_size, size=pop_size, replace=True, shuffle=False)
 
         # Return the new parents (individuals).
         return [population[i] for i in index]

pygenalgo/operators/selection/roulette_wheel_selector.py

@@ -1,5 +1,6 @@
 from math import fsum
 from pygenalgo.genome.chromosome import Chromosome
+from pygenalgo.operators.genetic_operator import increase_counter
 from pygenalgo.operators.selection.select_operator import SelectionOperator
 
 
@@ -26,6 +27,7 @@ def __init__(self, select_probability: float = 1.0):
         super().__init__(select_probability)
     # _end_def_
 
+    @increase_counter
     def select(self, population: list[Chromosome]):
         """
         Select the individuals, from the input population, that will be passed on
@@ -37,8 +39,8 @@ def select(self, population: list[Chromosome]):
         :return: the selected parents population (as list of chromosomes).
         """
 
-        # Get the length of the population list.
-        N = len(population)
+        # Get the population size.
+        pop_size = len(population)
 
         # Extract the fitness value of each chromosome.
         # This assumes that the fitness values are all
@@ -52,11 +54,9 @@ def select(self, population: list[Chromosome]):
         # in the population.
         selection_probs = [f / sum_fitness for f in all_fitness]
 
-        # Select 'N' new individuals (indexes).
-        index = self.rng.choice(N, size=N, p=selection_probs, replace=True, shuffle=False)
-
-        # Increase the selection counter.
-        self.inc_counter()
+        # Select the new individuals (indexes).
+        index = self.rng.choice(pop_size, size=pop_size, p=selection_probs,
+                                replace=True, shuffle=False)
 
         # Return the new parents (individuals).
         return [population[i] for i in index]

pygenalgo/operators/selection/stochastic_universal_selector.py

@@ -1,6 +1,7 @@
 from math import fsum
 from itertools import accumulate
 from pygenalgo.genome.chromosome import Chromosome
+from pygenalgo.operators.genetic_operator import increase_counter
 from pygenalgo.operators.selection.select_operator import SelectionOperator
 
 
@@ -27,6 +28,7 @@ def __init__(self, select_probability: float = 1.0):
         super().__init__(select_probability)
     # _end_def_
 
+    @increase_counter
     def select(self, population: list[Chromosome]):
         """
         Select the individuals, from the input population, that will be passed on to the next
@@ -42,21 +44,21 @@ def select(self, population: list[Chromosome]):
         # positive.
         all_fitness = [p.fitness for p in population]
 
-        # Get the size of the population.
-        N = len(population)
+        # Get the population size.
+        pop_size = len(population)
 
         # Compute the distance between pointers.
-        dist_p = fsum(all_fitness) / N
+        dist_p = fsum(all_fitness) / pop_size
 
         # Get a random number between 0 and dist_p.
         start_0 = dist_p * self.rng.random()
 
         # Calculate the pointers at equal distances 'dist_p'
         # starting from 'start_0'.
-        pointers = (start_0 + i*dist_p for i in range(0, N))
+        pointers = (start_0 + i*dist_p for i in range(pop_size))
 
         # Create a list that will contain the new parents.
-        new_parents = N * [None]
+        new_parents = pop_size * [None]
 
         # Compute the cumulative sum of the fitness values.
         cum_sum_fit = list(accumulate(all_fitness))
@@ -76,9 +78,6 @@ def select(self, population: list[Chromosome]):
             new_parents[n] = population[i]
         # _end_for_
 
-        # Increase the selection counter.
-        self.inc_counter()
-
         # Return the new parents (individuals).
         return new_parents
     # _end_def_

pygenalgo/operators/selection/tournament_selector.py

@@ -1,5 +1,6 @@
 from operator import attrgetter
 from pygenalgo.genome.chromosome import Chromosome
+from pygenalgo.operators.genetic_operator import increase_counter
 from pygenalgo.operators.selection.select_operator import SelectionOperator
 
 
@@ -29,6 +30,7 @@ def __init__(self, select_probability: float = 1.0, k: int = 2):
         self._items = max(2, int(k))
     # _end_def_
 
+    @increase_counter
     def select(self, population: list[Chromosome]):
         """
         Select the individuals, from the input population that will be passed on
@@ -39,18 +41,19 @@ def select(self, population: list[Chromosome]):
 
         :return: the selected parents population (as list of chromosomes).
         """
-        # Get the length of the population list.
-        N = len(population)
+
+        # Get the population size.
+        pop_size = len(population)
 
         # Create a list that will contain the new parents.
-        new_parents = N * [None]
+        new_parents = pop_size * [None]
 
         # Repeat the 'tournament' N times.
-        for i in range(N):
+        for i in range(pop_size):
 
             # Select randomly 'k' individuals (indexes)
             # from the initial population.
-            index = self.rng.choice(N, size=self._items,
+            index = self.rng.choice(pop_size, size=self._items,
                                     replace=False, shuffle=False)
 
             # Find the individual with the highest fitness value.
@@ -61,9 +64,6 @@ def select(self, population: list[Chromosome]):
             new_parents[i] = winner
         # _end_for_
 
-        # Increase the selection counter.
-        self.inc_counter()
-
         # Return the new parents (individuals).
         return new_parents
     # _end_def_

pygenalgo/operators/selection/truncation_selector.py

@@ -1,5 +1,6 @@
 from operator import attrgetter
 from pygenalgo.genome.chromosome import Chromosome
+from pygenalgo.operators.genetic_operator import increase_counter
 from pygenalgo.operators.selection.select_operator import SelectionOperator
 
 class TruncationSelector(SelectionOperator):
@@ -28,6 +29,7 @@ def __init__(self, select_probability: float = 1.0, p: float = 0.3):
         self._items = max(min(float(p), 0.9), 0.1)
     # _end_def_
 
+    @increase_counter
     def select(self, population: list[Chromosome]):
         """
         Select the individuals, from the input population, that will be passed on to the next
@@ -38,17 +40,16 @@ def select(self, population: list[Chromosome]):
         :return: the selected parents population (as list of chromosomes).
         """
 
-        # Get the length of the population list.
-        N = len(population)
+        # Get the population size.
+        pop_size = len(population)
 
         # Sort the population in descending order using their fitness value.
         sorted_population = sorted(population, key=attrgetter("fitness"), reverse=True)
 
-        # Select 'N' using only the higher 'p%' of the (old) population (indexes).
-        index = self.rng.choice(int(N*self._items), size=N, replace=True, shuffle=True)
-
-        # Increase the selection counter.
-        self.inc_counter()
+        # Select tne new parents using only the higher percentage '%' of
+        # the old population (indexes).
+        index = self.rng.choice(int(pop_size*self._items), size=pop_size,
+                                replace=True, shuffle=True)
 
         # Return the new parents (individuals).
         return [sorted_population[i] for i in index]