-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathbenchmark.py
209 lines (180 loc) · 8.48 KB
/
benchmark.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
import datetime
import os
import re
import shutil
import subprocess
import pandas as pd
import matplotlib.pyplot as plt
from pathlib import Path, PurePath
class Benchmark:
def __init__(self, image_int_dir,
image_seg_dir,
work_dir,
nyxus_executable,
feature_list = "*ALL*",
generate_missing_image=False) -> None:
self._image_int_dir = Path(image_int_dir)
self._image_seg_dir = Path(image_seg_dir)
self._work_dir = Path(work_dir)
self._nyxus_executable = Path(nyxus_executable)
self._feature_list = feature_list
self._generate_mising_image = generate_missing_image
self._image_collection = {}
self._processed_images = []
self._result_dir = None
self._merged_result_file = None
self._num_sample = 3
self.collect_image_pairs()
if os.path.exists( PurePath(self._work_dir, Path("results"))):
self._result_dir = PurePath(self._work_dir, Path("results"))
else:
try:
os.mkdir(PurePath(self._work_dir, Path("results")))
self._result_dir = PurePath(self._work_dir, Path("results"))
except:
print(f"Unable to create {self._work_dir}/results directory")
exit()
def create_nyxus_dirs(self, work_dir):
try:
os.mkdir(PurePath(self._work_dir, Path("int")))
except FileExistsError :
pass
try:
os.mkdir(PurePath(self._work_dir, Path("seg")))
except FileExistsError :
pass
try:
os.mkdir(PurePath(self._work_dir, Path("out")))
except FileExistsError :
pass
def prepare_workdir(self, work_dir, seg_dir, int_dir, base_file_name):
self.cleanup_workdir(work_dir)
self.create_nyxus_dirs(work_dir)
try:
shutil.copyfile(PurePath(int_dir, Path(base_file_name)), PurePath(work_dir,Path("int"), Path(base_file_name)))
except:
pass
try:
shutil.copyfile(PurePath(seg_dir, Path(base_file_name)), PurePath(work_dir,Path("seg"),Path(base_file_name)))
except:
pass
def collect_result(self, base_file_name, out_dir, result_dir, tag):
dest_file_name = None
if result_dir == None:
pass
base_file_name_wo_ext, tmp = os.path.splitext(base_file_name)
abs_result_file_name = PurePath(out_dir, Path(f"{base_file_name_wo_ext}_nyxustiming.csv"))
if(os.path.exists(abs_result_file_name)):
try:
dest_file_name = PurePath(result_dir, Path(f"{base_file_name_wo_ext}_nyxustiming.csv._{tag}"))
shutil.copyfile(abs_result_file_name, dest_file_name)
except:
print(f"Result not generated for {base_file_name}")
return dest_file_name
def cleanup_workdir(self, work_dir):
try:
shutil.rmtree(PurePath(work_dir, Path("int")))
except:
pass
try:
shutil.rmtree(PurePath(work_dir, Path("seg")))
except:
pass
# try:
# shutil.rmtree(PurePath(work_dir, Path("out")))
# except:
# pass
def run_nyxus(self, base_file_name, seg_dir, int_dir, out_dir, feature_list):
print(f"Running nyxus for {base_file_name}")
subprocess.run([
self._nyxus_executable,
f"--features={feature_list}",
f"--segDir={seg_dir}",
f"--intDir={int_dir}",
f"--outDir={out_dir}",
"--csvFile=singlecsv",
"--verbosity=3"
])
def merge_benchmark_suit_results(self):
input_csv_list = self._work_dir.glob("results/*.csv")
timestamp = datetime.datetime.now().strftime("%m_%d_%Y_%H_%M_%S")
out_file_name = PurePath(self._work_dir, Path(f"merged_result_{timestamp}.csv"))
self.merge_csv_files(input_csv_list, out_file_name)
self._merged_result_file = out_file_name
def merge_csv_files(self, input_csv_list, output_csv_name):
with open(output_csv_name, "w") as ofp:
first_csv = True
for in_file_name in input_csv_list:
with open(in_file_name, 'r') as ifp:
header_line = True
lines = ifp.readlines()
for line in lines:
if header_line and not first_csv:
pass
else:
ofp.write(line)
header_line = False
first_csv = False
def get_benchmark_data(self, roi_count, roi_area, feature_list):
base_file_name = f"synthetic_nrois={roi_count}_roiarea={roi_area}.tif"
seg_dir = PurePath(self._work_dir, Path("seg"))
int_dir = PurePath(self._work_dir, Path("int"))
out_dir = PurePath(self._work_dir, Path("out"))
result_dir = self._result_dir
if (roi_count, roi_area) in self._image_collection and \
self._image_collection[(roi_count, roi_area)] == base_file_name :
result_file_list = []
for i in range(self._num_sample):
self.prepare_workdir(self._work_dir, self._image_seg_dir, self._image_int_dir, base_file_name)
self.run_nyxus(base_file_name, seg_dir, int_dir, out_dir, feature_list)
result_file = self.collect_result(base_file_name, out_dir, result_dir, "run_"+str(i))
if result_file != None:
result_file_list.append(result_file)
self.cleanup_workdir(self._work_dir)
self.calculate_average(base_file_name, result_file_list, result_dir)
else:
print("weird stuff")
def calculate_average(self, base_file_name, result_file_list, result_dir):
dest_file_name = None
if result_dir == None:
pass
base_file_name_wo_ext, tmp = os.path.splitext(base_file_name)
primary_result_df = pd.read_csv(result_file_list[0])
data_column_list = ["rawtime"]
for count, result_file in enumerate(result_file_list[1:]):
result_data = pd.read_csv(result_file)
rawtime_col = result_data["rawtime"]
col_title = "rawtime" + str(count)
primary_result_df.insert(len(primary_result_df.columns), col_title, rawtime_col)
data_column_list.append(col_title)
primary_result_df["rawtime_avg"] = primary_result_df[data_column_list].mean(axis=1)
dest_file_name = PurePath(result_dir, Path(f"{base_file_name_wo_ext}_nyxustiming.csv"))
primary_result_df.to_csv(dest_file_name)
self._processed_images.append(base_file_name)
def collect_image_pairs(self):
#file_list = glob.glob(self._image_int_dir+"/*.tif")
file_list = self._image_int_dir.glob("*.tif")
for full_file_name in file_list:
base_file_name = full_file_name.name
roi_count, roi_size = re.findall("=(\d+)", base_file_name)
self._image_collection[(int(roi_count), int(roi_size))] = base_file_name
def run_benchmark_suit(self):
for roi_param in self._image_collection:
self.get_benchmark_data(roi_param[0], roi_param[1], self._feature_list)
self.merge_benchmark_suit_results()
def create_benchmark_plot(self, feature_l1, feature_l2, feature_l3, rerun_merge=False):
if self._merged_result_file == None or rerun_merge:
self.merge_benchmark_suit_results()
df = pd.read_csv(self._merged_result_file)
filtered_view = df[(df["h1"]==feature_l1) & (df["h2"]==feature_l2) & (df["h3"]==feature_l3)]
roi_area_list = filtered_view["roiarea"].unique()
for value in roi_area_list:
tmp_df = filtered_view[filtered_view["roiarea"] == value].sort_values("nrois")
plt.loglog(tmp_df.nrois, tmp_df.rawtime_avg, label=str(value), marker='o')
plt.title(f"Timing Data for {feature_l1}, {feature_l2}, {feature_l3}")
plt.xlabel("Number of ROIs")
plt.ylabel("Time (s)")
plt.legend()
merged_result_file_wo_ext, dummy = os.path.splitext(self._merged_result_file)
plot_file_name = f"{merged_result_file_wo_ext}_{feature_l1}_{feature_l2}_{feature_l3}.jpg"
plt.savefig(plot_file_name)