NiaPy.algorithms.basic.mke.MonkeyKingEvolutionV3.runIteration() - Code Metrics - Inspection of "BA and HBA" - NiaOrg/NiaPy - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Pull Request — master (#202)

by Grega

created 2019-04-18 06:38 UTC

MonkeyKingEvolutionV3.runIteration() A

↳ Parent: NiaPy.algorithms.basic.mke

Complexity

Conditions

Size

Total Lines	31
Code Lines	11

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	4
eloc	11
nop	9
dl	0
loc	31
rs	9.85
c	0
b	0
f	0

How to fix Many Parameters

# encoding=utf8
# pylint: disable=mixed-indentation, trailing-whitespace, multiple-statements, attribute-defined-outside-init, logging-not-lazy, no-self-use, len-as-condition, singleton-comparison, arguments-differ, bad-continuation
import logging
from math import ceil

from numpy import apply_along_axis, vectorize, argmin, argmax, full, tril

from NiaPy.algorithms.algorithm import Algorithm, Individual, defaultIndividualInit, defaultNumPyInit

logging.basicConfig()
logger = logging.getLogger('NiaPy.algorithms.basic')
logger.setLevel('INFO')

__all__ = ['MonkeyKingEvolutionV1', 'MonkeyKingEvolutionV2', 'MonkeyKingEvolutionV3']

class MkeSolution(Individual):
	r"""Implementation of Monkey King Evolution individual.

	Data:
		2018

	Authors:
		Klemen Berkovič

	License:
		MIT

	Attributes:
		x_pb (array of (float or int)): Personal best position of Monkey particle.
		f_pb (float): Personal best fitness/function value.
		MonkeyKing (bool): Boolean value indicating if particle is Monkey King particle.

	See Also:
		* :class:`NiaPy.algorithms.Individual`
	"""
	def __init__(self, **kwargs):
		r"""Initialize Monkey particle.

		Args:
			**kwargs: Additional arguments

		See Also:
			* :class:`NiaPy.algorithms.Individual.__init__()`
		"""
		Individual.__init__(self, **kwargs)
		self.f_pb, self.x_pb = self.f, self.x
		self.MonkeyKing = False

	def uPersonalBest(self):
		r"""Update presonal best position of particle."""
		if self.f < self.f_pb: self.x_pb, self.f_pb = self.x, self.f

class MonkeyKingEvolutionV1(Algorithm):
	r"""Implementation of monkey king evolution algorithm version 1.

	Algorithm:
		Monkey King Evolution version 1

	Date:
		2018

	Authors:
		Klemen Berkovič

	License:
		MIT

	Reference URL:
		https://www.sciencedirect.com/science/article/pii/S0950705116000198

	Reference paper:
		Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009.

	Attributes:
		Name (List[str]): List of strings representing algorithm names.
		F (float): Scale factor for normal particles.
		R (float): TODO.
		C (int): Number of new particles generated by Monkey King particle.
		FC (float): Scale factor for Monkey King particles.

	See Also:
		* :class:`NiaPy.algorithms.algorithm.Algorithm`
	"""
	Name = ['MonkeyKingEvolutionV1', 'MKEv1']

	@staticmethod
	def algorithmInfo():
		r"""Get basic information of algorithm.

		Returns:
			str: Basic information.

		See Also:
			* :func:`NiaPy.algorithms.Algorithm.algorithmInfo`
		"""
		return r"""Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009."""

	@staticmethod

	def typeParameters():
		r"""Get dictionary with functions for checking values of parameters.

		Returns:
			Dict[str, Callable]:
				* F (Callable[[int], bool])
				* R (Callable[[Union[int, float]], bool])
				* C (Callable[[Union[int, float]], bool])
				* FC (Callable[[Union[int, float]], bool])
		"""
		d = Algorithm.typeParameters()
		d.update({
			'NP': lambda x: isinstance(x, int) and x > 0,
			'F': lambda x: isinstance(x, (float, int)) and x > 0,
			'R': lambda x: isinstance(x, (float, int)) and x > 0,
			'C': lambda x: isinstance(x, int) and x > 0,
			'FC': lambda x: isinstance(x, (float, int)) and x > 0
		})
		return d

	def setParameters(self, NP=40, F=0.7, R=0.3, C=3, FC=0.5, **ukwargs):
		r"""Set Monkey King Evolution v1 algorithms static parameters.

		Args:
			NP (int): Population size.
			F (float): Scale factor for normal particle.
			R (float): Procentual value of now many new particle Monkey King particle creates. Value in rage [0, 1].
			C (int): Number of new particles generated by Monkey King particle.
			FC (float): Scale factor for Monkey King particles.
			**ukwargs (Dict[str, Any]): Additional arguments.

		See Also:
			* :func:`NiaPy.algorithms.algorithm.Algorithm.setParameters`
		"""
		Algorithm.setParameters(self, NP=NP, itype=ukwargs.pop('itype', MkeSolution), InitPopFunc=ukwargs.pop('InitPopFunc', defaultIndividualInit), **ukwargs)
		self.F, self.R, self.C, self.FC = F, R, C, FC
		if ukwargs: logger.info('Unused arguments: %s' % (ukwargs))

	def moveP(self, x, x_pb, x_b, task):
		r"""Move normal particle in search space.

		For moving particles algorithm uses next formula:
		:math:`\mathbf{x_{pb} - \mathit{F} \odot \mathbf{r} \odot (\mathbf{x_b} - \mathbf{x})`
		where
		:math:`\mathbf{r}` is one dimension array with `D` components. Components in this vector are in range [0, 1].

		Args:
			x (numpy.ndarray): Paticle position.
			x_pb (numpy.ndarray): Particle best position.
			x_b (numpy.ndarray): Best particle position.
			task (Task): Optimization task.

		Returns:
			numpy.ndarray: Particle new position.
		"""
		return x_pb + self.F * self.rand(task.D) * (x_b - x)

	def moveMK(self, x, task):
		r"""Move Mokey King paticle.

		For moving Monkey King particles algorithm uses next formula:
		:math:`\mathbf{x} + \mathit{FC} \odot \mathbf{R} \odot \mathbf{x}`
		where
		:math:`\mathbf{R}` is two dimensional array with shape `{C * D, D}`. Componentes of this array are in range [0, 1]

		Args:
			x (numpy.ndarray): Monkey King patricle position.
			task (Task): Optimization task.

		Returns:
			numpy.ndarray: New particles generated by Monkey King particle.
		"""
		return x + self.FC * self.rand([int(self.C * task.D), task.D]) * x

	def movePartice(self, p, p_b, task):
		r"""Move patricles.

		Args:
			p (MkeSolution): Monke particle.
			p_b (MkeSolution): Population best particle.
			task (Task): Optimization task.
		"""
		p.x = self.moveP(p.x, p.x_pb, p_b.x, task)
		p.evaluate(task, rnd=self.Rand)

	def moveMokeyKingPartice(self, p, task):
		r"""Move Monky King Particles.

		Args:
			p (MkeSolution): Monkey King particle to apply this function on.
			task (Task): Optimization task
		"""
		p.MonkeyKing = False
		A = apply_along_axis(task.repair, 1, self.moveMK(p.x, task), self.Rand)
		A_f = apply_along_axis(task.eval, 1, A)
		ib = argmin(A_f)
		p.x, p.f = A[ib], A_f[ib]

	def movePopulation(self, pop, xb, task):
		r"""Move population.

		Args:
			pop (numpy.ndarray[MkeSolution]): Current population.
			xb (MkeSolution): Current best solution.
			task (Task): Optimization task.

		Returns:
			numpy.ndarray[MkeSolution]: New particles.
		"""
		for p in pop:
			if p.MonkeyKing: self.moveMokeyKingPartice(p, task)
			else: self.movePartice(p, xb, task)
			p.uPersonalBest()
		return pop

	def initPopulation(self, task):
		r"""Init population.

		Args:
			task (Task): Optimization task

		Returns:
			Tuple(numpy.ndarray[MkeSolution], numpy.ndarray[float], Dict[str, Any]]:
				1. Initialized solutions
				2. Fitness/function values of solution
				3. Additional arguments
		"""
		pop, fpop, _ = Algorithm.initPopulation(self, task)
		for i in self.Rand.choice(self.NP, int(self.R * len(pop)), replace=False): pop[i].MonkeyKing = True
		return pop, fpop, {}

	def runIteration(self, task, pop, fpop, xb, fxb, **dparams):
		r"""Core function of Monkey King Evolution v1 algorithm.

		Args:
			task (Task): Optimization task
			pop (numpy.ndarray[MkeSolution]): Current population
			fpop (numpy.ndarray[float]): Current population fitness/function values
			xb (MkeSolution): Current best solution.
			fxb (float): Current best solutions function/fitness value.
			**dparams (Dict[str, Any]): Additional arguments.

		Returns:
			Tuple(numpy.ndarray[MkeSolution], numpy.ndarray[float], Dict[str, Any]]:
				1. Initialized solutions.
				2. Fitness/function values of solution.
				3. Additional arguments.
		"""
		pop = self.movePopulation(pop, xb, task)
		for i in self.Rand.choice(self.NP, int(self.R * len(pop)), replace=False): pop[i].MonkeyKing = True
		return pop, [m.f for m in pop], {}

class MonkeyKingEvolutionV2(MonkeyKingEvolutionV1):
	r"""Implementation of monkey king evolution algorithm version 2.

	Algorithm:
		Monkey King Evolution version 2

	Date:
		2018

	Authors:
		Klemen Berkovič

	License:
		MIT

	Reference URL:
		https://www.sciencedirect.com/science/article/pii/S0950705116000198

	Reference paper:
		Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009.

	Attributes:
		Name (List[str]): List of strings representing algorithm names.

	See Also:
		* :class:`NiaPy.algorithms.basic.mke.MonkeyKingEvolutionV1`
	"""
	Name = ['MonkeyKingEvolutionV2', 'MKEv2']

	@staticmethod
	def algorithmInfo():
		r"""Get basic information of algorithm.

		Returns:
			str: Basic information.

		See Also:
			* :func:`NiaPy.algorithms.Algorithm.algorithmInfo`
		"""
		return r"""Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009."""

	def moveMK(self, x, dx, task):
		r"""Move Monkey King particle.

		For movment of particles algorithm uses next formula:
		:math:`\mathbf{x} - \mathit{FC} \odot \mathbf{dx}`

		Args:
			x (numpy.ndarray): Particle to apply movment on.
			dx (numpy.ndarray): Difference between to random paricles in population.
			task (Task): Optimization task.

		Returns:
			numpy.ndarray: Moved particles.
		"""
		return x - self.FC * dx

	def moveMokeyKingPartice(self, p, pop, task):
		r"""Move Monkey King particles.

		Args:
			p (MkeSolution): Monkey King particle to move.
			pop (numpy.ndarray[MkeSolution]): Current population.
			task (Task): Optimization task.
		"""
		p.MonkeyKing = False
		p_b, p_f = p.x, p.f
		for _i in range(int(self.C * self.NP)):
			r = self.Rand.choice(self.NP, 2, replace=False)
			a = task.repair(self.moveMK(p.x, pop[r[0]].x - pop[r[1]].x, task), self.Rand)
			a_f = task.eval(a)
			if a_f < p_f: p_b, p_f = a, a_f
		p.x, p.f = p_b, p_f

	def movePopulation(self, pop, xb, task):
		r"""Move population.

		Args:
			pop (numpy.ndarray[MkeSolution]): Current population.
			xb (MkeSolution): Current best solution.
			task (Task): Optimization task.

		Returns:
			numpy.ndarray[MkeSolution]: Moved population.
		"""
		for p in pop:
			if p.MonkeyKing: self.moveMokeyKingPartice(p, pop, task)
			else: self.movePartice(p, xb, task)
			p.uPersonalBest()
		return pop

class MonkeyKingEvolutionV3(MonkeyKingEvolutionV1):
	r"""Implementation of monkey king evolution algorithm version 3.

	Algorithm:
		Monkey King Evolution version 3

	Date:
		2018

	Authors:
		Klemen Berkovič

	License:
		MIT

	Reference URL:
		https://www.sciencedirect.com/science/article/pii/S0950705116000198

	Reference paper:
		Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009.

	Attributes:
		Name (List[str]): List of strings that represent algorithm names.

	See Also:
		* :class:`NiaPy.algorithms.basic.mke.MonkeyKingEvolutionV1`
	"""
	Name = ['MonkeyKingEvolutionV3', 'MKEv3']

	@staticmethod
	def algorithmInfo():
		r"""Get basic information of algorithm.

		Returns:
			str: Basic information.

		See Also:
			* :func:`NiaPy.algorithms.Algorithm.algorithmInfo`
		"""
		return r"""Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009."""

	def setParameters(self, **ukwargs):
		r"""Set core parameters of MonkeyKingEvolutionV3 algorithm.

		Args:
			**ukwargs (Dict[str, Any]): Additional arguments.

		See Also:
			* :func:`NiaPy.algorithms.basic.MonkeyKingEvolutionV1.setParameters`
		"""
		MonkeyKingEvolutionV1.setParameters(self, itype=ukwargs.pop('itype', None), InitPopFunc=ukwargs.pop('InitPopFunc', defaultNumPyInit), **ukwargs)

	def neg(self, x):
		r"""Transform function.

		Args:
			x (Union[int, float]): Sould be 0 or 1.

		Returns:
			float: If 0 thet 1 else 1 then 0.
		"""
		return 0.0 if x == 1.0 else 1.0

	def initPopulation(self, task):
		r"""Initialize the population.

		Args:
			task (Task): Optimization task.

		Returns:
			Tuple[numpy.ndarray, numpy.ndarray[float], Dict[str, Any]]:
				1. Initialized population.
				2. Initialized population function/fitness values.
				3. Additional arguments:
					* k (int): TODO.
					* c (int): TODO.

		See Also:
			* :func:`NiaPy.algorithms.algorithm.Algorithm.initPopulation`
		"""
		X, X_f, d = Algorithm.initPopulation(self, task)
		k, c = int(ceil(self.NP / task.D)), int(ceil(self.C * task.D))
		d.update({'k': k, 'c': c})
		return X, X_f, d

	def runIteration(self, task, X, X_f, xb, fxb, k, c, **dparams):
		r"""Core funciton of Monkey King Evolution v3 algorithm.

		Args:
			task (Task): Optimization task
			X (numpy.ndarray): Current population
			X_f (numpy.ndarray[float]): Current population fitness/function values
			xb (numpy.ndarray): Current best individual
			fxb (float): Current best individual function/fitness value
			k (int): TODO
			c (int: TODO
			**dparams: Additional arguments

		Returns:
			Tuple[numpy.ndarray, numpy.ndarray[float], Dict[str, Any]]:
				1. Initialized population.
				2. Initialized population function/fitness values.
				3. Additional arguments:
					* k (int): TODO.
					* c (int): TODO.
		"""
		X_gb = apply_along_axis(task.repair, 1, xb + self.FC * X[self.Rand.choice(len(X), c)] - X[self.Rand.choice(len(X), c)], self.Rand)
		X_gb_f = apply_along_axis(task.eval, 1, X_gb)
		M = full([self.NP, task.D], 1.0)
		for i in range(k): M[i * task.D:(i + 1) * task.D] = tril(M[i * task.D:(i + 1) * task.D])
		for i in range(self.NP): self.Rand.shuffle(M[i])
		X = apply_along_axis(task.repair, 1, M * X + vectorize(self.neg)(M) * xb, self.Rand)
		X_f = apply_along_axis(task.eval, 1, X)
		iw, ib_gb = argmax(X_f), argmin(X_gb_f)
		if X_gb_f[ib_gb] <= X_f[iw]: X[iw], X_f[iw] = X_gb[ib_gb], X_gb_f[ib_gb]
		return X, X_f, {'k': k, 'c': c}

# vim: tabstop=3 noexpandtab shiftwidth=3 softtabstop=3


1		# encoding=utf8
2		# pylint: disable=mixed-indentation, trailing-whitespace, multiple-statements, attribute-defined-outside-init, logging-not-lazy, no-self-use, len-as-condition, singleton-comparison, arguments-differ, bad-continuation
3		import logging
4		from math import ceil
5
6		from numpy import apply_along_axis, vectorize, argmin, argmax, full, tril
7
8		from NiaPy.algorithms.algorithm import Algorithm, Individual, defaultIndividualInit, defaultNumPyInit
9
10		logging.basicConfig()
11		logger = logging.getLogger('NiaPy.algorithms.basic')
12		logger.setLevel('INFO')
13
14		__all__ = ['MonkeyKingEvolutionV1', 'MonkeyKingEvolutionV2', 'MonkeyKingEvolutionV3']
15
16		class MkeSolution(Individual):
17		r"""Implementation of Monkey King Evolution individual.
18
19		Data:
20		2018
21
22		Authors:
23		Klemen Berkovič
24
25		License:
26		MIT
27
28		Attributes:
29		x_pb (array of (float or int)): Personal best position of Monkey particle.
30		f_pb (float): Personal best fitness/function value.
31		MonkeyKing (bool): Boolean value indicating if particle is Monkey King particle.
32
33		See Also:
34		* :class:`NiaPy.algorithms.Individual`
35		"""
36		def __init__(self, **kwargs):
37		r"""Initialize Monkey particle.
38
39		Args:
40		**kwargs: Additional arguments
41
42		See Also:
43		* :class:`NiaPy.algorithms.Individual.__init__()`
44		"""
45		Individual.__init__(self, **kwargs)
46		self.f_pb, self.x_pb = self.f, self.x
47		self.MonkeyKing = False
48
49		def uPersonalBest(self):
50		r"""Update presonal best position of particle."""
51		if self.f < self.f_pb: self.x_pb, self.f_pb = self.x, self.f
52
53		class MonkeyKingEvolutionV1(Algorithm):
54		r"""Implementation of monkey king evolution algorithm version 1.
55
56		Algorithm:
57		Monkey King Evolution version 1
58
59		Date:
60		2018
61
62		Authors:
63		Klemen Berkovič
64
65		License:
66		MIT
67
68		Reference URL:
69		https://www.sciencedirect.com/science/article/pii/S0950705116000198
70
71		Reference paper:
72		Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009.
73
74		Attributes:
75		Name (List[str]): List of strings representing algorithm names.
76		F (float): Scale factor for normal particles.
77		R (float): TODO.
78		C (int): Number of new particles generated by Monkey King particle.
79		FC (float): Scale factor for Monkey King particles.
80
81		See Also:
82		* :class:`NiaPy.algorithms.algorithm.Algorithm`
83		"""
84		Name = ['MonkeyKingEvolutionV1', 'MKEv1']
85
86		@staticmethod
87		def algorithmInfo():
88		r"""Get basic information of algorithm.
89
90		Returns:
91		str: Basic information.
92
93		See Also:
94		* :func:`NiaPy.algorithms.Algorithm.algorithmInfo`
95		"""
96		return r"""Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009."""
97
98	View Code Duplication	@staticmethod
		0 ignored issues – show Duplication introduced 2019-04-16 21:06 UTC by Report Bug Copy Issue Report This code seems to be duplicated in your project. Loading history...
99		def typeParameters():
100		r"""Get dictionary with functions for checking values of parameters.
101
102		Returns:
103		Dict[str, Callable]:
104		* F (Callable[[int], bool])
105		* R (Callable[[Union[int, float]], bool])
106		* C (Callable[[Union[int, float]], bool])
107		* FC (Callable[[Union[int, float]], bool])
108		"""
109		d = Algorithm.typeParameters()
110		d.update({
111		'NP': lambda x: isinstance(x, int) and x > 0,
112		'F': lambda x: isinstance(x, (float, int)) and x > 0,
113		'R': lambda x: isinstance(x, (float, int)) and x > 0,
114		'C': lambda x: isinstance(x, int) and x > 0,
115		'FC': lambda x: isinstance(x, (float, int)) and x > 0
116		})
117		return d
118
119		def setParameters(self, NP=40, F=0.7, R=0.3, C=3, FC=0.5, **ukwargs):
120		r"""Set Monkey King Evolution v1 algorithms static parameters.
121
122		Args:
123		NP (int): Population size.
124		F (float): Scale factor for normal particle.
125		R (float): Procentual value of now many new particle Monkey King particle creates. Value in rage [0, 1].
126		C (int): Number of new particles generated by Monkey King particle.
127		FC (float): Scale factor for Monkey King particles.
128		**ukwargs (Dict[str, Any]): Additional arguments.
129
130		See Also:
131		* :func:`NiaPy.algorithms.algorithm.Algorithm.setParameters`
132		"""
133		Algorithm.setParameters(self, NP=NP, itype=ukwargs.pop('itype', MkeSolution), InitPopFunc=ukwargs.pop('InitPopFunc', defaultIndividualInit), **ukwargs)
134		self.F, self.R, self.C, self.FC = F, R, C, FC
135		if ukwargs: logger.info('Unused arguments: %s' % (ukwargs))
136
137		def moveP(self, x, x_pb, x_b, task):
138		r"""Move normal particle in search space.
139
140		For moving particles algorithm uses next formula:
141		:math:`\mathbf{x_{pb} - \mathit{F} \odot \mathbf{r} \odot (\mathbf{x_b} - \mathbf{x})`
142		where
143		:math:`\mathbf{r}` is one dimension array with `D` components. Components in this vector are in range [0, 1].
144
145		Args:
146		x (numpy.ndarray): Paticle position.
147		x_pb (numpy.ndarray): Particle best position.
148		x_b (numpy.ndarray): Best particle position.
149		task (Task): Optimization task.
150
151		Returns:
152		numpy.ndarray: Particle new position.
153		"""
154		return x_pb + self.F * self.rand(task.D) * (x_b - x)
155
156		def moveMK(self, x, task):
157		r"""Move Mokey King paticle.
158
159		For moving Monkey King particles algorithm uses next formula:
160		:math:`\mathbf{x} + \mathit{FC} \odot \mathbf{R} \odot \mathbf{x}`
161		where
162		:math:`\mathbf{R}` is two dimensional array with shape `{C * D, D}`. Componentes of this array are in range [0, 1]
163
164		Args:
165		x (numpy.ndarray): Monkey King patricle position.
166		task (Task): Optimization task.
167
168		Returns:
169		numpy.ndarray: New particles generated by Monkey King particle.
170		"""
171		return x + self.FC * self.rand([int(self.C * task.D), task.D]) * x
172
173		def movePartice(self, p, p_b, task):
174		r"""Move patricles.
175
176		Args:
177		p (MkeSolution): Monke particle.
178		p_b (MkeSolution): Population best particle.
179		task (Task): Optimization task.
180		"""
181		p.x = self.moveP(p.x, p.x_pb, p_b.x, task)
182		p.evaluate(task, rnd=self.Rand)
183
184		def moveMokeyKingPartice(self, p, task):
185		r"""Move Monky King Particles.
186
187		Args:
188		p (MkeSolution): Monkey King particle to apply this function on.
189		task (Task): Optimization task
190		"""
191		p.MonkeyKing = False
192		A = apply_along_axis(task.repair, 1, self.moveMK(p.x, task), self.Rand)
193		A_f = apply_along_axis(task.eval, 1, A)
194		ib = argmin(A_f)
195		p.x, p.f = A[ib], A_f[ib]
196
197		def movePopulation(self, pop, xb, task):
198		r"""Move population.
199
200		Args:
201		pop (numpy.ndarray[MkeSolution]): Current population.
202		xb (MkeSolution): Current best solution.
203		task (Task): Optimization task.
204
205		Returns:
206		numpy.ndarray[MkeSolution]: New particles.
207		"""
208		for p in pop:
209		if p.MonkeyKing: self.moveMokeyKingPartice(p, task)
210		else: self.movePartice(p, xb, task)
211		p.uPersonalBest()
212		return pop
213
214		def initPopulation(self, task):
215		r"""Init population.
216
217		Args:
218		task (Task): Optimization task
219
220		Returns:
221		Tuple(numpy.ndarray[MkeSolution], numpy.ndarray[float], Dict[str, Any]]:
222		1. Initialized solutions
223		2. Fitness/function values of solution
224		3. Additional arguments
225		"""
226		pop, fpop, _ = Algorithm.initPopulation(self, task)
227		for i in self.Rand.choice(self.NP, int(self.R * len(pop)), replace=False): pop[i].MonkeyKing = True
228		return pop, fpop, {}
229
230		def runIteration(self, task, pop, fpop, xb, fxb, **dparams):
231		r"""Core function of Monkey King Evolution v1 algorithm.
232
233		Args:
234		task (Task): Optimization task
235		pop (numpy.ndarray[MkeSolution]): Current population
236		fpop (numpy.ndarray[float]): Current population fitness/function values
237		xb (MkeSolution): Current best solution.
238		fxb (float): Current best solutions function/fitness value.
239		**dparams (Dict[str, Any]): Additional arguments.
240
241		Returns:
242		Tuple(numpy.ndarray[MkeSolution], numpy.ndarray[float], Dict[str, Any]]:
243		1. Initialized solutions.
244		2. Fitness/function values of solution.
245		3. Additional arguments.
246		"""
247		pop = self.movePopulation(pop, xb, task)
248		for i in self.Rand.choice(self.NP, int(self.R * len(pop)), replace=False): pop[i].MonkeyKing = True
249		return pop, [m.f for m in pop], {}
250
251		class MonkeyKingEvolutionV2(MonkeyKingEvolutionV1):
252		r"""Implementation of monkey king evolution algorithm version 2.
253
254		Algorithm:
255		Monkey King Evolution version 2
256
257		Date:
258		2018
259
260		Authors:
261		Klemen Berkovič
262
263		License:
264		MIT
265
266		Reference URL:
267		https://www.sciencedirect.com/science/article/pii/S0950705116000198
268
269		Reference paper:
270		Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009.
271
272		Attributes:
273		Name (List[str]): List of strings representing algorithm names.
274
275		See Also:
276		* :class:`NiaPy.algorithms.basic.mke.MonkeyKingEvolutionV1`
277		"""
278		Name = ['MonkeyKingEvolutionV2', 'MKEv2']
279
280		@staticmethod
281		def algorithmInfo():
282		r"""Get basic information of algorithm.
283
284		Returns:
285		str: Basic information.
286
287		See Also:
288		* :func:`NiaPy.algorithms.Algorithm.algorithmInfo`
289		"""
290		return r"""Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009."""
291
292		def moveMK(self, x, dx, task):
293		r"""Move Monkey King particle.
294
295		For movment of particles algorithm uses next formula:
296		:math:`\mathbf{x} - \mathit{FC} \odot \mathbf{dx}`
297
298		Args:
299		x (numpy.ndarray): Particle to apply movment on.
300		dx (numpy.ndarray): Difference between to random paricles in population.
301		task (Task): Optimization task.
302
303		Returns:
304		numpy.ndarray: Moved particles.
305		"""
306		return x - self.FC * dx
307
308		def moveMokeyKingPartice(self, p, pop, task):
309		r"""Move Monkey King particles.
310
311		Args:
312		p (MkeSolution): Monkey King particle to move.
313		pop (numpy.ndarray[MkeSolution]): Current population.
314		task (Task): Optimization task.
315		"""
316		p.MonkeyKing = False
317		p_b, p_f = p.x, p.f
318		for _i in range(int(self.C * self.NP)):
319		r = self.Rand.choice(self.NP, 2, replace=False)
320		a = task.repair(self.moveMK(p.x, pop[r[0]].x - pop[r[1]].x, task), self.Rand)
321		a_f = task.eval(a)
322		if a_f < p_f: p_b, p_f = a, a_f
323		p.x, p.f = p_b, p_f
324
325		def movePopulation(self, pop, xb, task):
326		r"""Move population.
327
328		Args:
329		pop (numpy.ndarray[MkeSolution]): Current population.
330		xb (MkeSolution): Current best solution.
331		task (Task): Optimization task.
332
333		Returns:
334		numpy.ndarray[MkeSolution]: Moved population.
335		"""
336		for p in pop:
337		if p.MonkeyKing: self.moveMokeyKingPartice(p, pop, task)
338		else: self.movePartice(p, xb, task)
339		p.uPersonalBest()
340		return pop
341
342		class MonkeyKingEvolutionV3(MonkeyKingEvolutionV1):
343		r"""Implementation of monkey king evolution algorithm version 3.
344
345		Algorithm:
346		Monkey King Evolution version 3
347
348		Date:
349		2018
350
351		Authors:
352		Klemen Berkovič
353
354		License:
355		MIT
356
357		Reference URL:
358		https://www.sciencedirect.com/science/article/pii/S0950705116000198
359
360		Reference paper:
361		Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009.
362
363		Attributes:
364		Name (List[str]): List of strings that represent algorithm names.
365
366		See Also:
367		* :class:`NiaPy.algorithms.basic.mke.MonkeyKingEvolutionV1`
368		"""
369		Name = ['MonkeyKingEvolutionV3', 'MKEv3']
370
371		@staticmethod
372		def algorithmInfo():
373		r"""Get basic information of algorithm.
374
375		Returns:
376		str: Basic information.
377
378		See Also:
379		* :func:`NiaPy.algorithms.Algorithm.algorithmInfo`
380		"""
381		return r"""Zhenyu Meng, Jeng-Shyang Pan, Monkey King Evolution: A new memetic evolutionary algorithm and its application in vehicle fuel consumption optimization, Knowledge-Based Systems, Volume 97, 2016, Pages 144-157, ISSN 0950-7051, https://doi.org/10.1016/j.knosys.2016.01.009."""
382
383		def setParameters(self, **ukwargs):
384		r"""Set core parameters of MonkeyKingEvolutionV3 algorithm.
385
386		Args:
387		**ukwargs (Dict[str, Any]): Additional arguments.
388
389		See Also:
390		* :func:`NiaPy.algorithms.basic.MonkeyKingEvolutionV1.setParameters`
391		"""
392		MonkeyKingEvolutionV1.setParameters(self, itype=ukwargs.pop('itype', None), InitPopFunc=ukwargs.pop('InitPopFunc', defaultNumPyInit), **ukwargs)
393
394		def neg(self, x):
395		r"""Transform function.
396
397		Args:
398		x (Union[int, float]): Sould be 0 or 1.
399
400		Returns:
401		float: If 0 thet 1 else 1 then 0.
402		"""
403		return 0.0 if x == 1.0 else 1.0
404
405		def initPopulation(self, task):
406		r"""Initialize the population.
407
408		Args:
409		task (Task): Optimization task.
410
411		Returns:
412		Tuple[numpy.ndarray, numpy.ndarray[float], Dict[str, Any]]:
413		1. Initialized population.
414		2. Initialized population function/fitness values.
415		3. Additional arguments:
416		* k (int): TODO.
417		* c (int): TODO.
418
419		See Also:
420		* :func:`NiaPy.algorithms.algorithm.Algorithm.initPopulation`
421		"""
422		X, X_f, d = Algorithm.initPopulation(self, task)
423		k, c = int(ceil(self.NP / task.D)), int(ceil(self.C * task.D))
424		d.update({'k': k, 'c': c})
425		return X, X_f, d
426
427		def runIteration(self, task, X, X_f, xb, fxb, k, c, **dparams):
428		r"""Core funciton of Monkey King Evolution v3 algorithm.
429
430		Args:
431		task (Task): Optimization task
432		X (numpy.ndarray): Current population
433		X_f (numpy.ndarray[float]): Current population fitness/function values
434		xb (numpy.ndarray): Current best individual
435		fxb (float): Current best individual function/fitness value
436		k (int): TODO
437		c (int: TODO
438		**dparams: Additional arguments
439
440		Returns:
441		Tuple[numpy.ndarray, numpy.ndarray[float], Dict[str, Any]]:
442		1. Initialized population.
443		2. Initialized population function/fitness values.
444		3. Additional arguments:
445		* k (int): TODO.
446		* c (int): TODO.
447		"""
448		X_gb = apply_along_axis(task.repair, 1, xb + self.FC * X[self.Rand.choice(len(X), c)] - X[self.Rand.choice(len(X), c)], self.Rand)
449		X_gb_f = apply_along_axis(task.eval, 1, X_gb)
450		M = full([self.NP, task.D], 1.0)
451		for i in range(k): M[i * task.D:(i + 1) * task.D] = tril(M[i * task.D:(i + 1) * task.D])
452		for i in range(self.NP): self.Rand.shuffle(M[i])
453		X = apply_along_axis(task.repair, 1, M * X + vectorize(self.neg)(M) * xb, self.Rand)
454		X_f = apply_along_axis(task.eval, 1, X)
455		iw, ib_gb = argmax(X_f), argmin(X_gb_f)
456		if X_gb_f[ib_gb] <= X_f[iw]: X[iw], X_f[iw] = X_gb[ib_gb], X_gb_f[ib_gb]
457		return X, X_f, {'k': k, 'c': c}
458
459		# vim: tabstop=3 noexpandtab shiftwidth=3 softtabstop=3
460

NiaOrg / NiaPy

Pull Request — master (#202)

MonkeyKingEvolutionV3.runIteration() A

Complexity

Size

Duplication

Importance

How to fix Many Parameters

Many Parameters

Duplication Side-by-Side

Filter issues like