paulin/EasyGA
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update stochastic methods

Browse Source 
- Added a stochastic rank selection and edited cleaner stochastic fitness selection using random.choices.
- Edited stochastic fitness selection to allow negative converted fitness values.
- Edited exception message for check_positive_fitness.
This commit is contained in:
SimpleArt
2020-12-02 17:55:57 -05:00
parent fb84ac40d4
commit 1761092abb
1 changed files with 40 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

57
src/parent_selection/parent_selection_methods.py
View File

@ -25,7 +25,7 @@ def check_positive_fitness(selection_method):
if ga.get_chromosome_fitness(0) > 0 and ga.get_chromosome_fitness(-1) >= 0:
selection_method(ga)
else:
raise Exception("Converted fitness values must be all positive. Consider using rank selection instead.")
raise Exception("Converted fitness values can't have negative values or be all 0. Consider using rank selection or stochastic selection instead.")
return new_method
@ -49,7 +49,7 @@ def compute_parent_amount(selection_method):
"""
def new_method(ga):
parent_amount = max(2, len(ga.population)*ga.parent_ratio)
parent_amount = max(2, round(len(ga.population)*ga.parent_ratio))
selection_method(ga, parent_amount)
return new_method
@ -65,6 +65,9 @@ class Parent_Selection:
class Rank:
"""Methods for selecting parents based on their rankings in the population
i.e. the n-th best chromosome has a fixed probability of being selected,
regardless of their chances"""
@check_selection_probability
@ensure_sorted
@ -107,6 +110,25 @@ class Parent_Selection:
return
@check_selection_probability
@ensure_sorted
@compute_parent_amount
def stochastic(ga, parent_amount):
"""
Selects parents using the same probability approach as tournament selection,
but doesn't create tournaments. Uses random.choices with weighted values to
select parents and may produce duplicate parents.
"""
weights = [
(1-ga.selection_probability) ** i
for i
in range(len(ga.population))
]
ga.population.mating_pool = random.choices(ga.population, weights, k = parent_amount)
class Fitness:
@check_selection_probability
@ -153,24 +175,25 @@ class Parent_Selection:
@check_selection_probability
@ensure_sorted
@check_positive_fitness
@compute_parent_amount
def stochastic(ga, parent_amount):
"""Stochastic roulette selection works based off of how strong the fitness is of the
chromosomes in the population. The stronger the fitness the higher the probability
that it will be selected. Instead of dividing the fitness by the sum of all fitnesses
and incrementally increasing the chance something is selected, the stochastic method
just divides by the highest fitness and selects randomly.
"""
Selects parents using the same probability approach as roulette selection,
but doesn't spin a roulette for every selection. Uses random.choices with
weighted values to select parents and may produce duplicate parents.
"""
max_fitness = ga.get_chromosome_fitness(0)
if ga.get_chromosome_fitness(-1) == ga.get_chromosome_fitness(0):
offset = 1-ga.get_chromosome_fitness(-1)
elif ga.get_chromosome_fitness(-1) < 0:
offset = -ga.get_chromosome_fitness(-1)
else:
offset = 0
# Loops until it reaches a desired mating pool size
while len(ga.population.mating_pool) < parent_amount:
weights = [
ga.get_chromosome_fitness(index) + offset
for index
in range(len(ga.population))
]
# Selected chromosome
index = random.randrange(len(ga.population))
# Probability of becoming a parent is fitness/max_fitness
if random.uniform(ga.selection_probability, 1) < ga.get_chromosome_fitness(index)/max_fitness:
ga.population.set_parent(index)
ga.population.mating_pool = random.choices(ga.population, weights, k = parent_amount)