-
Notifications
You must be signed in to change notification settings - Fork 0
/
convpoint_yaml_parser.py
211 lines (183 loc) · 8.79 KB
/
convpoint_yaml_parser.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
import yaml
import argparse
from pathlib import Path
import matplotlib.pyplot as mp
import numpy as np
from matplotlib.gridspec import GridSpec
def arg_parse():
parser = argparse.ArgumentParser()
parser.add_argument("--rootdir", default=f"D:/Travail/UdeS_maitrise/data/convpoint_test/results/CMM_2018_ERD_2018_aoi1")
args = parser.parse_args()
return args
def read_parameters(param_file):
"""Read and return parameters in .yaml file
Args:
param_file: Full file path of the parameters file
Returns:
yaml CommentedMap dict-like object
"""
with open(param_file) as yamlfile:
params = yaml.load(yamlfile, Loader=yaml.FullLoader)
return params
def open_logs(file, mode):
with open(file=file, mode='r') as f:
lines = f.read().splitlines()
if mode == 1:
last_lines = lines[-4:]
else:
last_lines = lines[-6:]
return last_lines
def set_plot():
fig = mp.figure(constrained_layout=True, figsize=(20, 20))
gs = GridSpec(9, 2, figure=fig)
ax1 = fig.add_subplot(gs[0:2, :], title='Training Loss - 4 Classes', xlim=(0, 50), ylim=(0, 0.00001))
ax2 = fig.add_subplot(gs[2:4, :], title='Validation Loss - 4 Classes', xlim=(0, 50), ylim=(0, 0.00001))
ax3 = fig.add_subplot(gs[4:6, :], title='Training Loss - 6 Classes', xlim=(0, 50), ylim=(0, 0.00001))
ax4 = fig.add_subplot(gs[6:8, :], title='Validation Loss - 6 Classes', xlim=(0, 50), ylim=(0, 0.00001))
return ax1, ax2, ax3, ax4
class Metrics():
def __init__(self, file_path):
self.acc_tables = {'1': {}, '2': {}}
self.fscore_tables = {'1': {}, '2': {}}
self.iou_tables = {'1': {}, '2': {}}
self.markdown_file = open(file_path, 'w')
def add_to_table(self, data, exp_name, metric, mode):
if mode == 1:
data_str = f'|{exp_name}|{float(data[0].split()[-1]):.3f}|{float(data[1].split()[-1]):.3f}|' \
f'{float(data[2].split()[-1]):.3f}|{float(data[3].split()[-1]):.3f}|'
elif mode == 2:
data_str = f'|{exp_name}|{float(data[0].split()[-1]):.3f}|{float(data[1].split()[-1]):.3f}|' \
f'{float(data[2].split()[-1]):.3f}|{float(data[3].split()[-1]):.3f}|{float(data[4].split()[-1]):.3f}|{float(data[5].split()[-1]):.3f}'
else:
raise ValueError(f'Provided mode is not implemented')
if metric == 'iou':
self.iou_tables[str(mode)][exp_name] = data_str
elif metric == 'fscore':
self.fscore_tables[str(mode)][exp_name] = data_str
elif metric == 'acc':
self.acc_tables[str(mode)][exp_name] = data_str
else:
raise Exception
def write_to_file(self, mode):
mode = str(mode)
if mode == '1':
template_table = {'title': f'|Name|Other|Building|Water|Ground',
'format': f'|----|----|----|----|----|'}
self.markdown_file.write(f'# Metrics for 4 classes \n ## IOU \n')
else:
template_table = {'title': f'|Name|Other|Building|Water|Ground|Low Vegetation|Medium-High Vegetation|',
'format': f'|----|----|----|----|----|----|----|'}
self.markdown_file.write(f'# Metrics for 6 classes \n ## IOU \n')
self.markdown_file.write(template_table['title'] + ' \n' + template_table['format'] + ' \n')
for k, v in self.iou_tables[mode].items():
self.markdown_file.write(f'{v} \n')
self.markdown_file.write(f'## F1-Score \n')
self.markdown_file.write(template_table['title'] + ' \n' + template_table['format'] + ' \n')
for k, v in self.fscore_tables[mode].items():
self.markdown_file.write(f'{v} \n')
self.markdown_file.write(f'## Accuracy \n')
self.markdown_file.write(template_table['title'] + ' \n' + template_table['format'] + ' \n')
for k, v in self.acc_tables[mode].items():
self.markdown_file.write(f'{v} \n')
class Configs():
def __init__(self, file_path):
tables = {}
for i in ['1', '2']:
tables[i] = {"name": f"|Name|",
"format": f"|----|",
"batchsize": f"|BatchSize|",
"npoints": f"|NPoints|",
"blocksize": f"|BlockSize|",
"iter": f"|Iterations|",
"features": f"|Features|",
"test_step": f"|Test Step|",
"nepochs": f"|Epoch|",
"lr": f"|LR|",
"model": f"|Model|",
"num_workers": f"|Num workers|",
"drop": f"|Drop|"}
self.tables = tables
self.markdown_file = open(file_path, 'w')
self.data = []
def exp_to_tables(self):
for exp in self.data:
data = exp['data']
exp_name = exp['exp_name']
mode = exp['mode']
self.tables[str(mode)]['name'] += f"{exp_name}|"
self.tables[str(mode)]['format'] += f"----|"
for k, v in self.tables.items():
if k == str(mode):
for key, value in v.items():
if key in data:
self.tables[k][key] += f"{data[key]} |"
elif key not in ['name', 'format']:
self.tables[k][key] += f" |"
def add_to_exp(self, data, exp_name):
self.data.append({'data': data, 'exp_name': exp_name, 'mode': data['mode']})
def write_to_file(self):
for k, v in self.tables.items():
if k == '1':
n_classes = 3
else:
n_classes = 5
self.markdown_file.write(f"# Experiments on {n_classes} classes\n ")
for key, value in v.items():
self.markdown_file.write(value + '\n')
self.markdown_file.write('\n')
self.markdown_file.write(f" \n # Loss figure \n ")
self.markdown_file.close()
def main():
args = arg_parse()
root_path = Path(args.rootdir)
project_name = root_path.parts[-1]
configs = Configs(file_path=root_path / f'{project_name}_config.md')
val_metrics = Metrics(file_path=root_path / f'{project_name}_val_metrics.md')
trn_metrics = Metrics(file_path=root_path / f'{project_name}_trn_metrics.md')
tst_metrics = Metrics(file_path=root_path / f'{project_name}_tst_metrics.md')
ax1, ax2, ax3, ax4 = set_plot()
for file_path in root_path.glob('**/*.yaml'):
yaml_dict = read_parameters(file_path)
exp_name = file_path.parent.parts[-1]
configs.add_to_exp(yaml_dict, exp_name)
mode = yaml_dict['mode']
exp_path = file_path.parent
for metrics_file in exp_path.glob('**/*.log*'):
split_name = metrics_file.stem.split('_')
if split_name[1] == 'classwise':
log_file = open_logs(metrics_file, mode)
metric_name = split_name[-1]
if split_name[2] == 'val':
val_metrics.add_to_table(log_file, exp_name, metric_name, mode)
elif split_name[2] == 'trn':
trn_metrics.add_to_table(log_file, exp_name, metric_name, mode)
elif split_name[2] == 'tst':
tst_metrics.add_to_table(log_file, exp_name, metric_name, mode)
else:
if split_name[-1] == 'loss':
with open(file=metrics_file, mode='r') as f:
lines = f.read().splitlines()
if len(lines) > 1:
new_list = [float(x.split()[1]) for x in lines]
np_log_data = np.asarray(new_list)
if split_name[1] == 'trn' and mode == 1:
ax1.plot(np.arange(0, len(np_log_data), 1), np_log_data, label=exp_name)
elif split_name[1] == 'val' and mode == 1:
ax2.plot(np.arange(0, len(np_log_data), 1), np_log_data, label=exp_name)
if split_name[1] == 'trn' and mode == 2:
ax3.plot(np.arange(0, len(np_log_data), 1), np_log_data, label=exp_name)
elif split_name[1] == 'val' and mode == 2:
ax4.plot(np.arange(0, len(np_log_data), 1), np_log_data, label=exp_name)
mp.legend(bbox_to_anchor=(0, -0.1), loc='upper left', ncol=2)
# mp.show()
mp.savefig(root_path / f'loss.png')
val_metrics.write_to_file(mode=1)
val_metrics.write_to_file(mode=2)
trn_metrics.write_to_file(mode=1)
trn_metrics.write_to_file(mode=2)
tst_metrics.write_to_file(mode=1)
tst_metrics.write_to_file(mode=2)
configs.exp_to_tables()
configs.write_to_file()
if __name__ == '__main__':
main()