-
Notifications
You must be signed in to change notification settings - Fork 2
/
run_analyze_results.py
280 lines (216 loc) · 9.78 KB
/
run_analyze_results.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
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
"""
This file is part of PRINS.
Copyright (C) 2021 University of Luxembourg
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.
SPDX-FileType: SOURCE
SPDX-FileCopyrightText: 2021 University of Luxembourg
SPDX-License-Identifier: GPL-3.0-or-later
"""
import math
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from matplotlib import rc
from natsort import natsorted
from scipy.stats import entropy
latex_figures = './'
latex_data = './'
SYSTEMS = ['Hadoop', 'HDFS', 'Linux', 'Zookeeper', 'CoreSync', 'NGLClient', 'Oobelib', 'PDApp']
# activate latex text rendering
rc('text', usetex=True)
plt.style.use('bmh')
plt.rcParams["font.family"] = "Times New Roman"
def message_distribution():
results = []
for system in SYSTEMS:
logs_df = pd.read_csv(f'dataset/{system}/{system}_preprocessed_logs.csv')
counts = natsorted(logs_df.groupby(by='component')['message'].count(), reverse=True)
dist = np.array(counts[:4])/sum(counts[:4])
results.append((system, f'{np.std(dist):.3f}', f'{np.mean(np.absolute(dist - np.mean(dist))):.3f}', f'{entropy(dist, base=2)/2:.3f}', dist))
df = pd.DataFrame(results, columns=['system', 'std', 'MAD', 'norm_entropy', 'dist'])
df.to_csv('message_distribution.csv', index=False)
def rq1_boxplot(ncols=4):
file = 'expr_output/summary_model_inference.csv'
df = pd.read_csv(file)
grouped = df.groupby(by=['system', 'technique'])
nrows = math.ceil(len(SYSTEMS) / ncols)
fig, axs = plt.subplots(nrows=nrows, ncols=ncols, figsize=(10, 5))
for system in SYSTEMS:
i, j = divmod(SYSTEMS.index(system), ncols)
labels = [1, 2, 3, 4]
data = []
for w in labels:
technique = f'PRINS-w{w}'
data.append(grouped.get_group((system, technique))['time_s'])
axs[i][j].boxplot(data, labels=labels, sym='.')
axs[i][j].set_title(system, size=14)
if j == 0:
axs[i][j].set_ylabel('Execution Time (s)', fontsize=13)
if i == nrows-1:
axs[i][j].set_xlabel('Workers', fontsize=13)
axs[i][j].tick_params(axis='both', which='major', labelsize=13)
plt.tight_layout()
plt.show()
fig.savefig(latex_figures + 'rq1-boxplot.pdf', dpi=300)
def rq2_boxplot(ncols=4):
file = 'expr_output/summary_model_inference.csv'
df = pd.read_csv(file)
grouped = df.groupby(by=['system', 'technique'])
nrows = math.ceil(len(SYSTEMS) / ncols)
fig, axs = plt.subplots(nrows=nrows, ncols=ncols, figsize=(10, 5))
for system in SYSTEMS:
i, j = divmod(SYSTEMS.index(system), ncols)
labels = [x for x in range(0, 11)]
data = []
for w in labels:
technique = f'hybrid-{w}'
data.append(grouped.get_group((system, technique))['time_s'])
axs[i][j].boxplot(data, labels=labels, sym='.')
axs[i][j].set_title(system, size=14)
if j == 0:
axs[i][j].set_ylabel('Execution time (s)', fontsize=13)
if i == nrows-1:
axs[i][j].set_xlabel('Parameter', fontsize=13)
axs[i][j].tick_params(axis='both', which='major', labelsize=13)
plt.tight_layout()
plt.show()
fig.savefig(latex_figures + 'rq2-boxplot.pdf', dpi=300)
def rq3_line():
file = 'expr_output/summary_k_folds_cv.csv'
df = pd.read_csv(file)
grouped = df.groupby(by=['system', 'technique'])
fig, axs = plt.subplots(nrows=6, ncols=4, figsize=(10, 12))
metrics = ['recall', 'specificity', 'BA']
for metric in metrics:
for system in SYSTEMS:
i, j = divmod(SYSTEMS.index(system), 4)
i += metrics.index(metric) * 2
labels = ['PRINS']
for u in range(1, 11):
labels.append(f'PRINS:hybrid-{u}')
y_values = []
for technique in labels:
if metric == 'BA':
value = (grouped.get_group((system, technique))['recall'].values[0] +
grouped.get_group((system, technique))['specificity'].values[0]) / 2
else:
value = grouped.get_group((system, technique))[metric].values[0]
y_values.append(value)
# axs[i][j].plot([x for x in range(11)], y_values, '.k', markersize=3)
axs[i][j].plot([x for x in range(11)], y_values, linestyle='-', color='gray', linewidth=1, zorder=1)
axs[i][j].scatter([x for x in range(11)], y_values, color='black', s=10, zorder=2)
axs[i][j].set_title(system, size=14)
if i == 5:
axs[i][j].set_xlabel('Parameter', fontsize=13)
if j == 0:
axs[i][j].set_ylabel(metric, fontsize=13)
axs[i][j].set_xticks([x for x in range(0, 11, 2)])
axs[i][j].set_yticks([y / 10 for y in range(0, 11, 2)])
axs[i][j].tick_params(axis='both', which='major', labelsize=13)
plt.tight_layout()
plt.show()
fig.savefig(latex_figures + 'rq3-boxplot.pdf', dpi=300)
def rq4_line(ncols=4):
file = 'expr_output/summary_model_inference_duplicated_logs.csv'
df = pd.read_csv(file, na_values=['timeout', 'crash'])
grouped = df.groupby(by=['system', 'technique', 'duplicated'])
techniques = ['MINT-SYS', 'PRINS-w1', 'PRINS-w4']
nrows = math.ceil(len(SYSTEMS) / ncols)
fig, axs = plt.subplots(nrows=nrows, ncols=ncols, figsize=(10, 5))
for system in SYSTEMS:
i, j = divmod(SYSTEMS.index(system), ncols)
X = sorted(df.loc[df['system'] == system]['duplicated'].unique())
Y = {}
for technique in techniques:
Y[technique] = []
for x in X:
try:
y = grouped.get_group((system, technique, x))['time_s'].mean()
except KeyError as e:
print(f'KeyError: {e}')
y = np.nan
if y is not np.nan and 'PRINS' in technique:
y += grouped.get_group((system, 'hybrid-1', x))['time_s'].mean()
Y[technique].append(y)
axs[i][j].plot(X, Y['MINT-SYS'], label='MINT', marker='x', ls='-', c='r', linewidth=1)
axs[i][j].plot(X, Y['PRINS-w1'], label=r'\textit{PRINS-N}', marker='^', ls='-', c='b', linewidth=1)
axs[i][j].plot(X, Y['PRINS-w4'], label=r'\textit{PRINS-P}', marker='v', ls='-', c='g', linewidth=1)
axs[i][j].legend()
axs[i][j].set_title(system, size=14)
if j == 0:
axs[i][j].set_ylabel('Execution Time (s)', fontsize=13)
if i == nrows-1:
axs[i][j].set_xlabel('Duplication Factor', fontsize=13)
plt.tight_layout()
plt.show()
fig.savefig(latex_figures + 'rq4-line.pdf', dpi=300)
def rq5_table(det: str = 'hybrid-1'):
file = 'expr_output/summary_k_folds_cv.csv'
df = pd.read_csv(file)
df = df.set_index(['system', 'technique'])
summary = []
for system in SYSTEMS:
if system == 'Spark':
continue
recall_m = df.loc[system, 'MINT-SYS']['recall']
recall_p = df.loc[system, f'PRINS:{det}']['recall']
recall_diff = (recall_p - recall_m) * 100
spec_m = df.loc[system, 'MINT-SYS']['specificity']
spec_p = df.loc[system, f'PRINS:{det}']['specificity']
spec_diff = (spec_p - spec_m) * 100
ba_m = (recall_m + spec_m) / 2
ba_p = (recall_p + spec_p) / 2
ba_diff = (ba_p - ba_m) * 100
summary.append((system, recall_m, recall_p, recall_diff, spec_m, spec_p, spec_diff, ba_m, ba_p, ba_diff))
df = pd.DataFrame(summary, columns=['system',
'r_mint', 'r_prins', 'r_diff',
's_mint', 's_prins', 's_diff',
'ba_mint', 'ba_prins', 'ba_diff'])
df = df.set_index('system')
# load log confidence scores
df2 = pd.read_csv('dataset/dataset_summary.csv', index_col='system')
df = df.join(df2['log_confidence'])
# calculate average and save the resulting data frame
df.loc['Average'] = df.mean()
df.round(9).to_csv(latex_data + 'rq5-table.csv')
def rq5_model_size(det: str = 'hybrid-1'):
file = 'expr_output/summary_model_inference.csv'
df = pd.read_csv(file)
grouped = df.groupby(by=['system', 'technique']).mean()
summary = []
for system in SYSTEMS:
if system == 'Spark':
continue
states_m = grouped.loc[system, 'MINT-SYS']['states']
states_p = grouped.loc[system, det]['states']
trans_m = grouped.loc[system, 'MINT-SYS']['transitions']
trans_p = grouped.loc[system, det]['transitions']
summary.append((system,
states_m, states_p, states_p/states_m,
trans_m, trans_p, trans_p/trans_m))
df = pd.DataFrame(summary, columns=['system',
'states_m', 'states_p', 'states_r',
'trans_m', 'trans_p', 'trans_r'])
df = df.set_index('system')
df.loc['Average'] = df.mean()
df.round(9).to_csv(latex_data + 'rq5-model-size.csv')
if __name__ == '__main__':
message_distribution()
# execution time
rq1_boxplot(ncols=4)
rq2_boxplot(ncols=4)
rq4_line(ncols=4)
# accuracy
rq3_line()
rq5_table()
# model size
rq5_model_size()