-
Notifications
You must be signed in to change notification settings - Fork 1
/
analysis.py
245 lines (194 loc) · 9.74 KB
/
analysis.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
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
import os
from timeit import timeit
import matplotlib.pyplot as plt
import pandas as pd
from genetic_algorithm.strategies import *
from model import Model
class Analysis:
GENERATIONS = 100
TEST_ITERATIONS = 10
ALL_CROSSOVER_METHODS = [order_crossover, order_based_crossover, position_based_crossover,
partially_mapped_crossover, cycle_crossover, edge_recombination_crossover]
PC_IMPACT_SUBDIRECTORY = "pc_impact"
problem_seed = 1
def __init__(self, customer_count=15, vehicle_count=3, results_directory=None):
self.model = Model()
self.customer_count = customer_count
self.vehicle_count = vehicle_count
self.results_directory = results_directory
if results_directory is None:
self.results_directory = f"analysis_results_cus{customer_count}_veh{vehicle_count}"
def analyse_crossovers(self, crossover_methods, iterations=TEST_ITERATIONS, generations=GENERATIONS, **kwargs):
self._generate_problem()
method_scores = []
for method in crossover_methods:
results = self._get_results(iterations, generations=generations, crossover_method=method, **kwargs)
best_scores, mean_scores, std_scores = zip(
*(self._calculate_scores_statistics(results, gen) for gen in range(generations + 1)))
method_scores.append((pretty_name(method), mean_scores))
self._plot_scores(range(generations + 1), best_scores, mean_scores, std_scores, pretty_name(method))
self._save_to_file({
"Generation": range(generations + 1),
"Best Score": best_scores,
"Mean Score": mean_scores,
"Std. Score": std_scores},
method.__name__)
self._plot_method_comparison(method_scores)
def analyse_pc_impact(self, crossover_methods, iterations=TEST_ITERATIONS, generations=GENERATIONS, **kwargs):
self._generate_problem()
probabilities = np.arange(0.0, 1.1, 0.1)
method_final_scores = []
for method in crossover_methods:
best_final_scores = []
mean_final_scores = []
std_final_scores = []
for pc in probabilities:
results = self._get_results(iterations, generations=generations, crossover_method=method, pc=pc,
**kwargs)
best_score, mean_score, std_score = self._calculate_scores_statistics(results)
best_final_scores.append(best_score)
mean_final_scores.append(mean_score)
std_final_scores.append(std_score)
method_final_scores.append((pretty_name(method), mean_final_scores))
self._save_to_file({
"PC": probabilities,
"Best Score": best_final_scores,
"Mean Score": mean_final_scores,
"Std. Score": std_final_scores},
method.__name__,
Analysis.PC_IMPACT_SUBDIRECTORY)
self._plot_pc_method_comparison(probabilities, method_final_scores)
def _generate_problem(self):
np.random.seed(self.problem_seed)
self.model.generate_customers(self.customer_count)
def _get_results(self, iterations, output=False, **kwargs):
results = []
for i in range(iterations):
np.random.seed(i)
self.model.generate_routes(output=output, show_plot=output, vehicle_count=self.vehicle_count, **kwargs)
results.append(self.model.result)
return results
@staticmethod
def _calculate_scores_statistics(results, generation=-1):
scores = [result.best_scores_history[generation] for result in results]
return np.min(scores), np.mean(scores), np.std(scores)
@staticmethod
def analyse_execution_time(individual_size=100, iterations=1000):
method_exec_time = []
for crossover in Analysis.ALL_CROSSOVER_METHODS:
def test_method():
p1, p2 = np.random.permutation(individual_size), np.random.permutation(individual_size)
crossover(p1, p2)
execution_time = timeit(test_method, number=iterations)
method_exec_time.append((crossover.__name__, execution_time))
print(f"{pretty_name(crossover)}: {execution_time}")
Analysis._plot_execution_time_comparison(method_exec_time)
@staticmethod
def analyse_crossover_impact_on_offsprings(population_size=100, iterations=1000):
def impact(arr1, arr2):
return 1.0 - np.mean(arr1 == arr2)
print("Impact on the offsprings")
print("|".join(f"{header:^17}" for header in ["p1 - o1", "p1 - o2", "p2 - o1", "p2 - o2", "name"]))
for crossover in Analysis.ALL_CROSSOVER_METHODS:
total_impacts = []
for _ in range(iterations):
p1, p2 = np.random.permutation(population_size), np.random.permutation(population_size)
o1, o2 = crossover(p1, p2)
total_impacts.append([impact(p1, o1), impact(p1, o2), impact(p2, o1), impact(p2, o2)])
mean_std = [f'{np.mean(np.array(total_impacts)[:, i]):7.2%} ± {np.std(np.array(total_impacts)[:, i]):6.2%}'
for i in range(4)]
print(" |".join(mean_std), "|", pretty_name(crossover))
@staticmethod
def plot_scores_from_file(file):
generations, best_scores, mean_scores, std_scores = analysis._read_file(file)
Analysis._plot_scores(generations, best_scores, mean_scores, std_scores, file.replace("_", " "))
@staticmethod
def plot_method_comparison_from_files(files):
method_scores = []
for file in files:
generations, best_scores, mean_scores, std_scores = analysis._read_file(file)
method_scores.append((file, mean_scores))
if method_scores:
Analysis._plot_method_comparison(method_scores)
@staticmethod
def plot_pc_method_comparison_from_files(files):
method_final_scores = []
pc = []
for file in files:
pc, best_scores, mean_scores, std_scores = analysis._read_file(file, Analysis.PC_IMPACT_SUBDIRECTORY)
method_final_scores.append((file, mean_scores))
if method_final_scores:
Analysis._plot_pc_method_comparison(pc, method_final_scores)
@staticmethod
def _plot_scores(generations, best_scores, mean_scores, std_scores, method_name):
plt.figure(figsize=(10, 5))
plt.plot(generations, best_scores, label='Best Score')
plt.plot(generations, mean_scores, label='Mean Score')
n = max(len(generations) // 10, 1) # Display every nth point
plt.errorbar(generations[::n], mean_scores[::n], yerr=std_scores[::n], fmt='o', label='Std Score', color="C3")
plt.errorbar(generations[-1], mean_scores[-1], yerr=std_scores[-1], fmt='o', color="C3")
plt.xlabel('Generation')
plt.ylabel('Score')
plt.title(f'Scores vs. Generation ({method_name})')
plt.legend()
plt.grid(True)
plt.show()
@staticmethod
def _plot_method_comparison(method_scores):
plt.figure(figsize=(10, 5))
for method_name, scores in method_scores:
plt.plot(range(len(scores)), scores, label=method_name)
plt.xlabel('Generation')
plt.ylabel('Mean Score')
plt.title('Scores mean comparison')
plt.legend()
plt.grid(True)
plt.show()
@staticmethod
def _plot_pc_method_comparison(probabilities, method_final_scores):
plt.figure(figsize=(10, 5))
for method_name, scores in method_final_scores:
plt.plot(probabilities, scores, marker=".", label=method_name)
plt.xlabel('Crossover probability')
plt.ylabel('Mean score')
plt.title(f'Scores vs. Pc')
plt.legend()
plt.grid(True)
plt.show()
@staticmethod
def _plot_execution_time_comparison(method_exec_time):
plt.figure(figsize=(10, 5))
for method_name, execution_time in method_exec_time:
plt.bar(method_name, execution_time)
plt.xlabel('Crossover Methods')
plt.ylabel('Execution Time (seconds)')
plt.title('Execution Time of Crossover Methods')
plt.xticks(rotation=45, ha='right')
plt.tight_layout()
plt.show()
def _save_to_file(self, dictionary, file_name, subdirectory=""):
directory = os.path.join(self.results_directory, subdirectory)
os.makedirs(directory, exist_ok=True)
pd.DataFrame(dictionary).round(2).to_csv(os.path.join(directory, f"{file_name}.csv"), index=False)
def _read_file(self, file_name, subdirectory=""):
file_path = os.path.join(self.results_directory, subdirectory, f"{file_name}.csv")
if not os.path.exists(file_path):
raise FileNotFoundError(f"CSV file {file_path} not found.")
df = pd.read_csv(file_path)
return tuple(df[column].tolist() for column in df.columns)
def pretty_name(crossover_method):
return crossover_method.__name__.replace("_", " ")
if __name__ == "__main__":
analysis = Analysis(customer_count=15, vehicle_count=3)
analyzed_crossover_methods = [order_crossover, order_based_crossover, partially_mapped_crossover, cycle_crossover]
file_names = [method.__name__ for method in analyzed_crossover_methods]
# Example 1
# analysis.analyse_crossovers(analyzed_crossover_methods, generations=500, pc=0.5, enable_2_opt=False)
# analysis.plot_method_comparison_from_files(file_names)
# Example 2
# analysis.analyse_pc_impact(analyzed_crossover_methods, generations=500)
# analysis.plot_pc_method_comparison_from_files(file_names)
# Example 3
# analysis.analyse_crossover_impact_on_offsprings()
# Example 4
# analysis.analyse_execution_time(10, 10000)