# -*- coding: utf-8 -*-
# File generated according to Generator/ClassesRef/Optimization/OptiGenAlg.csv
# WARNING! All changes made in this file will be lost!
"""Method code available at https://github.com/Eomys/pyleecan/tree/master/pyleecan/Methods/Optimization/OptiGenAlg
"""
from os import linesep
from sys import getsizeof
from logging import getLogger
from ._check import check_var, raise_
from ..Functions.get_logger import get_logger
from ..Functions.save import save
from ..Functions.load import load_init_dict
from ..Functions.Load.import_class import import_class
from copy import deepcopy
from .OptiSolver import OptiSolver
from ntpath import basename
from os.path import isfile
from ._check import CheckTypeError
import numpy as np
import random
from numpy import isnan
from ._check import InitUnKnowClassError
[docs]class OptiGenAlg(OptiSolver):
"""Genetic algorithm class"""
VERSION = 1
# generic save method is available in all object
save = save
# get_logger method is available in all object
get_logger = get_logger
def __init__(
self,
selector=None,
crossover=None,
mutator=None,
p_cross=0.9,
p_mutate=0.1,
size_pop=40,
nb_gen=100,
problem=-1,
xoutput=-1,
logger_name="Pyleecan.OptiSolver",
is_keep_all_output=False,
init_dict=None,
init_str=None,
):
"""Constructor of the class. Can be use in three ways :
- __init__ (arg1 = 1, arg3 = 5) every parameters have name and default values
for pyleecan type, -1 will call the default constructor
- __init__ (init_dict = d) d must be a dictionary with property names as keys
- __init__ (init_str = s) s must be a string
s is the file path to load
ndarray or list can be given for Vector and Matrix
object or dict can be given for pyleecan Object"""
if init_str is not None: # Load from a file
init_dict = load_init_dict(init_str)[1]
if init_dict is not None: # Initialisation by dict
assert type(init_dict) is dict
# Overwrite default value with init_dict content
if "selector" in list(init_dict.keys()):
selector = init_dict["selector"]
if "crossover" in list(init_dict.keys()):
crossover = init_dict["crossover"]
if "mutator" in list(init_dict.keys()):
mutator = init_dict["mutator"]
if "p_cross" in list(init_dict.keys()):
p_cross = init_dict["p_cross"]
if "p_mutate" in list(init_dict.keys()):
p_mutate = init_dict["p_mutate"]
if "size_pop" in list(init_dict.keys()):
size_pop = init_dict["size_pop"]
if "nb_gen" in list(init_dict.keys()):
nb_gen = init_dict["nb_gen"]
if "problem" in list(init_dict.keys()):
problem = init_dict["problem"]
if "xoutput" in list(init_dict.keys()):
xoutput = init_dict["xoutput"]
if "logger_name" in list(init_dict.keys()):
logger_name = init_dict["logger_name"]
if "is_keep_all_output" in list(init_dict.keys()):
is_keep_all_output = init_dict["is_keep_all_output"]
# Set the properties (value check and convertion are done in setter)
self.selector = selector
self.crossover = crossover
self.mutator = mutator
self.p_cross = p_cross
self.p_mutate = p_mutate
self.size_pop = size_pop
self.nb_gen = nb_gen
# Call OptiSolver init
super(OptiGenAlg, self).__init__(
problem=problem,
xoutput=xoutput,
logger_name=logger_name,
is_keep_all_output=is_keep_all_output,
)
# The class is frozen (in OptiSolver init), for now it's impossible to
# add new properties
def __str__(self):
"""Convert this object in a readeable string (for print)"""
OptiGenAlg_str = ""
# Get the properties inherited from OptiSolver
OptiGenAlg_str += super(OptiGenAlg, self).__str__()
if self._selector_str is not None:
OptiGenAlg_str += "selector = " + self._selector_str + linesep
elif self._selector_func is not None:
OptiGenAlg_str += "selector = " + str(self._selector_func) + linesep
else:
OptiGenAlg_str += "selector = None" + linesep + linesep
if self._crossover_str is not None:
OptiGenAlg_str += "crossover = " + self._crossover_str + linesep
elif self._crossover_func is not None:
OptiGenAlg_str += "crossover = " + str(self._crossover_func) + linesep
else:
OptiGenAlg_str += "crossover = None" + linesep + linesep
if self._mutator_str is not None:
OptiGenAlg_str += "mutator = " + self._mutator_str + linesep
elif self._mutator_func is not None:
OptiGenAlg_str += "mutator = " + str(self._mutator_func) + linesep
else:
OptiGenAlg_str += "mutator = None" + linesep + linesep
OptiGenAlg_str += "p_cross = " + str(self.p_cross) + linesep
OptiGenAlg_str += "p_mutate = " + str(self.p_mutate) + linesep
OptiGenAlg_str += "size_pop = " + str(self.size_pop) + linesep
OptiGenAlg_str += "nb_gen = " + str(self.nb_gen) + linesep
return OptiGenAlg_str
def __eq__(self, other):
"""Compare two objects (skip parent)"""
if type(other) != type(self):
return False
# Check the properties inherited from OptiSolver
if not super(OptiGenAlg, self).__eq__(other):
return False
if other._selector_str != self._selector_str:
return False
if other._crossover_str != self._crossover_str:
return False
if other._mutator_str != self._mutator_str:
return False
if other.p_cross != self.p_cross:
return False
if other.p_mutate != self.p_mutate:
return False
if other.size_pop != self.size_pop:
return False
if other.nb_gen != self.nb_gen:
return False
return True
[docs] def compare(self, other, name="self", ignore_list=None, is_add_value=False):
"""Compare two objects and return list of differences"""
if ignore_list is None:
ignore_list = list()
if type(other) != type(self):
return ["type(" + name + ")"]
diff_list = list()
# Check the properties inherited from OptiSolver
diff_list.extend(
super(OptiGenAlg, self).compare(
other, name=name, ignore_list=ignore_list, is_add_value=is_add_value
)
)
if other._selector_str != self._selector_str:
diff_list.append(name + ".selector")
if other._crossover_str != self._crossover_str:
diff_list.append(name + ".crossover")
if other._mutator_str != self._mutator_str:
diff_list.append(name + ".mutator")
if (
other._p_cross is not None
and self._p_cross is not None
and isnan(other._p_cross)
and isnan(self._p_cross)
):
pass
elif other._p_cross != self._p_cross:
if is_add_value:
val_str = (
" (self="
+ str(self._p_cross)
+ ", other="
+ str(other._p_cross)
+ ")"
)
diff_list.append(name + ".p_cross" + val_str)
else:
diff_list.append(name + ".p_cross")
if (
other._p_mutate is not None
and self._p_mutate is not None
and isnan(other._p_mutate)
and isnan(self._p_mutate)
):
pass
elif other._p_mutate != self._p_mutate:
if is_add_value:
val_str = (
" (self="
+ str(self._p_mutate)
+ ", other="
+ str(other._p_mutate)
+ ")"
)
diff_list.append(name + ".p_mutate" + val_str)
else:
diff_list.append(name + ".p_mutate")
if other._size_pop != self._size_pop:
if is_add_value:
val_str = (
" (self="
+ str(self._size_pop)
+ ", other="
+ str(other._size_pop)
+ ")"
)
diff_list.append(name + ".size_pop" + val_str)
else:
diff_list.append(name + ".size_pop")
if other._nb_gen != self._nb_gen:
if is_add_value:
val_str = (
" (self="
+ str(self._nb_gen)
+ ", other="
+ str(other._nb_gen)
+ ")"
)
diff_list.append(name + ".nb_gen" + val_str)
else:
diff_list.append(name + ".nb_gen")
# Filter ignore differences
diff_list = list(filter(lambda x: x not in ignore_list, diff_list))
return diff_list
def __sizeof__(self):
"""Return the size in memory of the object (including all subobject)"""
S = 0 # Full size of the object
# Get size of the properties inherited from OptiSolver
S += super(OptiGenAlg, self).__sizeof__()
S += getsizeof(self._selector_str)
S += getsizeof(self._crossover_str)
S += getsizeof(self._mutator_str)
S += getsizeof(self.p_cross)
S += getsizeof(self.p_mutate)
S += getsizeof(self.size_pop)
S += getsizeof(self.nb_gen)
return S
[docs] def as_dict(self, type_handle_ndarray=0, keep_function=False, **kwargs):
"""
Convert this object in a json serializable dict (can be use in __init__).
type_handle_ndarray: int
How to handle ndarray (0: tolist, 1: copy, 2: nothing)
keep_function : bool
True to keep the function object, else return str
Optional keyword input parameter is for internal use only
and may prevent json serializability.
"""
# Get the properties inherited from OptiSolver
OptiGenAlg_dict = super(OptiGenAlg, self).as_dict(
type_handle_ndarray=type_handle_ndarray,
keep_function=keep_function,
**kwargs,
)
if self._selector_str is not None:
OptiGenAlg_dict["selector"] = self._selector_str
elif keep_function:
OptiGenAlg_dict["selector"] = self.selector
else:
OptiGenAlg_dict["selector"] = None
if self.selector is not None:
self.get_logger().warning(
"OptiGenAlg.as_dict(): "
+ f"Function {self.selector.__name__} is not serializable "
+ "and will be converted to None."
)
if self._crossover_str is not None:
OptiGenAlg_dict["crossover"] = self._crossover_str
elif keep_function:
OptiGenAlg_dict["crossover"] = self.crossover
else:
OptiGenAlg_dict["crossover"] = None
if self.crossover is not None:
self.get_logger().warning(
"OptiGenAlg.as_dict(): "
+ f"Function {self.crossover.__name__} is not serializable "
+ "and will be converted to None."
)
if self._mutator_str is not None:
OptiGenAlg_dict["mutator"] = self._mutator_str
elif keep_function:
OptiGenAlg_dict["mutator"] = self.mutator
else:
OptiGenAlg_dict["mutator"] = None
if self.mutator is not None:
self.get_logger().warning(
"OptiGenAlg.as_dict(): "
+ f"Function {self.mutator.__name__} is not serializable "
+ "and will be converted to None."
)
OptiGenAlg_dict["p_cross"] = self.p_cross
OptiGenAlg_dict["p_mutate"] = self.p_mutate
OptiGenAlg_dict["size_pop"] = self.size_pop
OptiGenAlg_dict["nb_gen"] = self.nb_gen
# The class name is added to the dict for deserialisation purpose
# Overwrite the mother class name
OptiGenAlg_dict["__class__"] = "OptiGenAlg"
return OptiGenAlg_dict
[docs] def copy(self):
"""Creates a deepcopy of the object"""
# Handle deepcopy of all the properties
if self._selector_str is not None:
selector_val = self._selector_str
else:
selector_val = self._selector_func
if self._crossover_str is not None:
crossover_val = self._crossover_str
else:
crossover_val = self._crossover_func
if self._mutator_str is not None:
mutator_val = self._mutator_str
else:
mutator_val = self._mutator_func
p_cross_val = self.p_cross
p_mutate_val = self.p_mutate
size_pop_val = self.size_pop
nb_gen_val = self.nb_gen
if self.problem is None:
problem_val = None
else:
problem_val = self.problem.copy()
if self.xoutput is None:
xoutput_val = None
else:
xoutput_val = self.xoutput.copy()
logger_name_val = self.logger_name
is_keep_all_output_val = self.is_keep_all_output
# Creates new object of the same type with the copied properties
obj_copy = type(self)(
selector=selector_val,
crossover=crossover_val,
mutator=mutator_val,
p_cross=p_cross_val,
p_mutate=p_mutate_val,
size_pop=size_pop_val,
nb_gen=nb_gen_val,
problem=problem_val,
xoutput=xoutput_val,
logger_name=logger_name_val,
is_keep_all_output=is_keep_all_output_val,
)
return obj_copy
def _set_None(self):
"""Set all the properties to None (except pyleecan object)"""
self.selector = None
self.crossover = None
self.mutator = None
self.p_cross = None
self.p_mutate = None
self.size_pop = None
self.nb_gen = None
# Set to None the properties inherited from OptiSolver
super(OptiGenAlg, self)._set_None()
def _get_selector(self):
"""getter of selector"""
return self._selector_func
def _set_selector(self, value):
"""setter of selector"""
if value is None:
self._selector_str = None
self._selector_func = None
elif isinstance(value, str) and "lambda" in value:
self._selector_str = value
self._selector_func = eval(value)
elif isinstance(value, str) and isfile(value) and value[-3:] == ".py":
self._selector_str = value
f = open(value, "r")
exec(f.read(), globals())
self._selector_func = eval(basename(value[:-3]))
elif callable(value):
self._selector_str = None
self._selector_func = value
else:
raise CheckTypeError(
"For property selector Expected function or str (path to python file or lambda), got: "
+ str(type(value))
)
selector = property(
fget=_get_selector,
fset=_set_selector,
doc="""Selector of the genetic algorithm
:Type: function
""",
)
def _get_crossover(self):
"""getter of crossover"""
return self._crossover_func
def _set_crossover(self, value):
"""setter of crossover"""
if value is None:
self._crossover_str = None
self._crossover_func = None
elif isinstance(value, str) and "lambda" in value:
self._crossover_str = value
self._crossover_func = eval(value)
elif isinstance(value, str) and isfile(value) and value[-3:] == ".py":
self._crossover_str = value
f = open(value, "r")
exec(f.read(), globals())
self._crossover_func = eval(basename(value[:-3]))
elif callable(value):
self._crossover_str = None
self._crossover_func = value
else:
raise CheckTypeError(
"For property crossover Expected function or str (path to python file or lambda), got: "
+ str(type(value))
)
crossover = property(
fget=_get_crossover,
fset=_set_crossover,
doc="""Crossover of the genetic algorithm
:Type: function
""",
)
def _get_mutator(self):
"""getter of mutator"""
return self._mutator_func
def _set_mutator(self, value):
"""setter of mutator"""
if value is None:
self._mutator_str = None
self._mutator_func = None
elif isinstance(value, str) and "lambda" in value:
self._mutator_str = value
self._mutator_func = eval(value)
elif isinstance(value, str) and isfile(value) and value[-3:] == ".py":
self._mutator_str = value
f = open(value, "r")
exec(f.read(), globals())
self._mutator_func = eval(basename(value[:-3]))
elif callable(value):
self._mutator_str = None
self._mutator_func = value
else:
raise CheckTypeError(
"For property mutator Expected function or str (path to python file or lambda), got: "
+ str(type(value))
)
mutator = property(
fget=_get_mutator,
fset=_set_mutator,
doc="""Mutator of the genetic algorithm
:Type: function
""",
)
def _get_p_cross(self):
"""getter of p_cross"""
return self._p_cross
def _set_p_cross(self, value):
"""setter of p_cross"""
check_var("p_cross", value, "float", Vmin=0, Vmax=1)
self._p_cross = value
p_cross = property(
fget=_get_p_cross,
fset=_set_p_cross,
doc="""Probability of crossover
:Type: float
:min: 0
:max: 1
""",
)
def _get_p_mutate(self):
"""getter of p_mutate"""
return self._p_mutate
def _set_p_mutate(self, value):
"""setter of p_mutate"""
check_var("p_mutate", value, "float", Vmin=0, Vmax=1)
self._p_mutate = value
p_mutate = property(
fget=_get_p_mutate,
fset=_set_p_mutate,
doc="""Probability of mutation
:Type: float
:min: 0
:max: 1
""",
)
def _get_size_pop(self):
"""getter of size_pop"""
return self._size_pop
def _set_size_pop(self, value):
"""setter of size_pop"""
check_var("size_pop", value, "int", Vmin=1)
self._size_pop = value
size_pop = property(
fget=_get_size_pop,
fset=_set_size_pop,
doc="""Size of the population
:Type: int
:min: 1
""",
)
def _get_nb_gen(self):
"""getter of nb_gen"""
return self._nb_gen
def _set_nb_gen(self, value):
"""setter of nb_gen"""
check_var("nb_gen", value, "int", Vmin=1)
self._nb_gen = value
nb_gen = property(
fget=_get_nb_gen,
fset=_set_nb_gen,
doc="""Number of generations
:Type: int
:min: 1
""",
)
