SudokuSolver/AlgoritmoGenetico.py at master · matecomp/SudokuSolver · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
import numpy as np
import copy
from HillClimbing import *

class AG(HC):

	#Construtor
	def __init__(self, n_individuals, cross_rate, mut_rate, hill_object, elitism=0, roubar=True):
		self.__n_individuals = n_individuals
		self.__cross_rate = cross_rate
		self.__mut_rate = mut_rate
		self.__hc = hill_object
		self.__individual = np.array([])
		self.__DNA = np.array([])
		self.__chromosome = np.array([])
		self.population_generation(n_individuals)
		self.__elitism = elitism
		self.__roubar = roubar

	def softmax(self, x):
		e_x = np.exp(x)/np.exp(x).sum()
		return e_x

	def train(self, iter=50):
		individuals = self.get_individual()
		n_individuals = self.__n_individuals
		hc = self.__hc
		elitism = self.__elitism
		solution = None
		maximum = 0
		for i in xrange(iter):
			scores = np.array([])
			for idx, ind in enumerate(individuals):
				if self.__roubar:
					_, score = hc.climbing(ind)
				else:
					score = hc.score(ind)
				scores = np.append(scores, score)
			probs = scores/scores.sum()
			# probs = self.softmax(scores*1.0)
			print "Generation:", i+1
			print scores
			print probs.sum()
			if elitism > 0:
				sols = np.argsort(scores)[-elitism:]
			else:
				sols = []

			sol = np.argmax(scores)
			if scores[sol] >= maximum:
				solution = copy.copy(self.__individual[sol])
				maximum = scores[sol]
			idxs = np.random.choice(n_individuals,n_individuals-elitism,p=probs)
			idxs = np.append(sols,idxs).astype(int)
			print idxs.shape
			self.__individual = self.__individual[idxs]

			np.random.shuffle(self.__individual)
			for idx in xrange(0,n_individuals,2):
				if self.__cross_rate < random.random(): continue
				ind1 = self.get_individual(idx)
				ind2 = self.get_individual(idx+1)
				ind1, ind2 = self.crossover(ind1, ind2)
				self.set_individual(ind1, idx)
				self.set_individual(ind2, idx+1)
				ind1 = self.get_individual(idx)

			np.random.shuffle(self.__individual)
			for idx in xrange(0,n_individuals):
				if self.__mut_rate < random.random(): continue
				ind = self.get_individual(idx)
				for i in xrange(20):
					ind.random_swap()
			# self.crossover()
			# self.mutation()
		print maximum
		return solution


	def crossover(self, ind1, ind2):
		rate = random.random()
		rows, cols = ind1.get_dimensions()
		cut = int(rate * rows)
		i1 = ind1.get_board()
		i2 = ind2.get_board()
		b1 = np.append(i1[:cut,:],i2[cut:,:])
		b2 = np.append(i2[:cut,:],i1[cut:,:])
		b1 = b1.reshape(rows,cols)
		b2 = b2.reshape(rows,cols)
		ind1.set_board(b1)
		ind2.set_board(b2)
		return ind1, ind2


	def get_individual(self, idx=None):
		if idx is None:
			return self.__individual
		return self.__individual[idx]

	def get_chromosome(self, idx=None):
		if idx is None:
			return self.__chromosome
		return self.__chromosome[idx]

	def set_individual(self, individual, idx):
		self.__individual[idx] = individual

	def set_choromosome(self, chromosome, idx):
		self.__chromosome[idx] = chromosome

	def append_individual(self, individual):
		individuals = self.get_individual()
		self.__individual = np.append(individuals,individual)

	def append_chromosome(self, chromosome):
		chromosomes = self.get_chromosome()
		self.__chromosome = np.append(chromosomes, chromosome)

	def ind2chromo(self, individual):
		dna = individual.get_dna()
		board = individual.get_board()
		rows, cols = individual.get_dimensions()
		chromosome = []
		for i, line in enumerate(board):
			idxs = []
			for j, element in enumerate(line):
				if not(individual.is_fixed(i,j)):
					basenit = np.where(dna==element)[0][i]
					dec = np.where(idxs < basenit)[0].shape[0]
					idxs = np.append(idxs, basenit)
					chromosome.append(basenit-dec)
		chromosome = np.array(chromosome)
		return chromosome

	def chromo2ind(self, individual, chromosome):
		dna = individual.get_dna()
		rows, cols = individual.get_dimensions()
		board = []
		for i, line in enumerate(chromosome):
			for j, element in enumerate(line):
				if not(individual.is_fixed(i,j)):
					up = np.where(line[:j] <= element)[0].shape[0]
					# print element, up
					value = dna[i,element+up]
					board.append(value)
				else:
					board.append(individual.get_position(i,j))
		board = np.array(board).reshape(rows,cols)
		individual.set_board(board)
		return individual


	def create_individual(self):
		individual = self.__hc.create_board()
		individual.init_board()
		return individual

	def population_generation(self, n_individuals):
		self.__individual = np.array([])
		self.__chromosome = np.array([])
		for ind in xrange(n_individuals):
			individual = self.create_individual()
			self.append_individual(individual)
#			chromosome = self.ind2chromo(individual)
#			self.append_chromosome(chromosome)