-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathimproved_cycle_solver.py
225 lines (165 loc) · 7.35 KB
/
improved_cycle_solver.py
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
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
import random
import os
import sys
from heapq import heappush, heappop
sys.path.append('..')
sys.path.append('../..')
import argparse
import utils
from student_utils_sp18 import *
import numpy as np
import pickle
def cost(input_file, output_file, dist_dict, adjacency_matrix):
input_data = utils.read_file(input_file)
output_data = utils.read_file(output_file)
number_of_kingdoms, list_of_kingdom_names, starting_kingdom, adjacency_matrix = data_parser(input_data)
kingdom_tour = output_data[0]
conquered_kingdoms = output_data[1]
kingdom_tour = [list_of_kingdom_names.index(name) for name in kingdom_tour]
conquered_kingdoms = [list_of_kingdom_names.index(name) for name in conquered_kingdoms]
return cylce_val(dist_dict, kingdom_tour) + dominating_set_value(adjacency_matrix, conquered_kingdoms)
def output_cost(file_num, dist_dict, adjacency_matrix):
input_file, output_file = "./inputs/" + file_num + ".in", "./outputs/" + file_num + ".out"
return cost(input_file, output_file, dist_dict, adjacency_matrix)
############### DOM SET #############################
def random_dominating_set(neighbor_dict, source_index, number_of_kingdoms, node_prob, temp):
all_nodes = set(range(number_of_kingdoms))
con = set()
#sur = set([source_index])
sur = set()
prob = softmax(node_prob, temp)
if (0 in prob):
prob = None
order = np.random.choice(number_of_kingdoms, number_of_kingdoms, replace=False, p =prob)
for i in order:
con.add(i)
sur.add(i)
sur.update(neighbor_dict[i])
if len(sur) == len(all_nodes):
return con
return con
def get_dom_prob(neighbor_dict, adjacency_matrix, number_of_kingdoms):
return [1*len(neighbor_dict[i])/(adjacency_matrix[i][i]) for i in range(number_of_kingdoms)]
def softmax(x, temp):
"""Compute softmax values for each sets of scores in x."""
e_x = (np.exp(x - np.max(x))) / temp
return e_x / e_x.sum(axis=0) # only difference
def dominating_set_value(adjacency_matrix, dom_set):
val = 0
for node in dom_set:
val += adjacency_matrix[node][node]
return val
def best_dominating_set(neighbor_dict, source_index, number_of_kingdoms, adjacency_matrix, temp):
node_prob = get_dom_prob(neighbor_dict, adjacency_matrix, number_of_kingdoms)
all_dom = []
rep_check = set()
for i in range(20000):
dom_set = random_dominating_set(neighbor_dict, source_index, number_of_kingdoms, node_prob, temp)
val = dominating_set_value(adjacency_matrix, dom_set)
if val not in rep_check:
rep_check.add(val)
heappush(all_dom, (val, dom_set))
top10 = []
for i in range(15):
if len(all_dom) == 0:
break
top10.append(heappop(all_dom))
return top10
######################################### Cycle ##############
def best_cycle(dist_dict, dom_set, source_index):
source_con = False
if source_index in dom_set:
dom_set.remove(source_index)
source_con = True
best_cycle = None
if not dom_set:
dom_set.add(source_index)
return (0, [source_index])
for i in range(3000000):
cycle = list(random_cycle(dom_set))
cycle = [source_index] + cycle + [source_index]
val = cylce_val(dist_dict, cycle)
if best_cycle is None or best_cycle[0] > val:
best_cycle = (val, cycle)
if source_con:
dom_set.add(source_index)
return best_cycle
def random_cycle(dom_set):
return np.random.choice(list(dom_set), len(dom_set), replace=False)
def cylce_val(dist_dict, cycle):
total_cost = 0
for i in range(len(cycle) - 1):
total_cost += dist_dict[cycle[i]][cycle[i + 1]]
return total_cost
def get_path(cycle_order, path_dict):
path = []
order_len = len(cycle_order)
for i in range(order_len - 2):
path += path_dict[cycle_order[i]][cycle_order[i + 1]]
path.pop()
path += path_dict[cycle_order[order_len - 2]][cycle_order[order_len - 1]]
return path
################# write solutions ##################
def write_output(file_num, solution, list_of_kingdom_names, path_dict, write_to):
file = open(write_to + file_num + ".out", "w")
cycle_order = solution[1]
conquer_set = solution[2]
path = get_path(cycle_order, path_dict)
# print(path)
for i in path:
file.write(list_of_kingdom_names[i])
file.write(" ")
file.write("\n")
for j in conquer_set:
file.write(list_of_kingdom_names[j])
file.write(" ")
file.close()
######################################## SOLVER ##################
def solver(curr_file, iter_file, beaten_file, write_to, poly2, range_start, range_end):
for j in range(10000):
with open(iter_file, "a") as file_iter:
file_iter.write(str(j) + "\n")
file_names = []
for i in range(range_start, range_end):
if i in [102, 103, 104, 210, 211, 212, 375, 376, 377, 705, 706, 707, 249, 250, 310, 521, 696, 697, 698, 711, 712, 713]:
continue
if i in [195, 207, 208, 209, 336, 337, 338, 528, 529, 594, 596, 642, 643, 644]:
continue
file_names.append(str(i) + ".in")
for file_name in file_names:
print("#########################")
print(file_name)
print("#########################")
input_data = utils.read_file("./inputs/" + file_name)
number_of_kingdoms, list_of_kingdom_names, starting_kingdom, adjacency_matrix = data_parser(input_data)
source_index = list_of_kingdom_names.index(starting_kingdom)
if number_of_kingdoms < 100:
continue
temp = 1
file_num = file_name.split(".")[0]
with open(curr_file, "a") as file_curr:
file_curr.write(file_num + "\n")
poly_path = "./"
if poly2:
poly_path = "./dict_poly2/"
neighbor_dict = pickle.Unpickler(open( poly_path + "neighbors_dict/" + file_num + "_neighbors_dict.p", "rb" )).load()
dist_dict = pickle.Unpickler( open( poly_path + "shortest_dist_dict/" + file_num + "_dist_dict.p", "rb" ) ).load()
path_dict = pickle.Unpickler( open( poly_path + "shortest_path_dict/" + file_num + "_path_dict.p", "rb" ) ).load()
curr_best = output_cost(file_num, dist_dict, adjacency_matrix)
top10_dom = best_dominating_set(neighbor_dict, source_index, number_of_kingdoms, adjacency_matrix, temp)
for dom_cost, dom_set in top10_dom:
if dom_cost >= curr_best:
continue
cycle_tup = best_cycle(dist_dict, dom_set, source_index)
cycle_cost = cycle_tup[0]
cycle_path = cycle_tup[1]
val = dom_cost + cycle_cost
if curr_best > val:
with open(beaten_file, "a") as file_beat:
file_beat.write(file_num + "\n")
file_beat.write("curr_best: " + str(curr_best) + "\n")
file_beat.write("new_best: "+ str(val) + "\n" + "\n")
print("write")
best_solution = (dom_cost+cycle_cost, cycle_path, dom_set)
write_output(file_num, best_solution, list_of_kingdom_names, path_dict, write_to)
break;