-
Notifications
You must be signed in to change notification settings - Fork 2
/
Selecting.py
1071 lines (1015 loc) · 47.5 KB
/
Selecting.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
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
"""
基于NGAWest2地震动数据库的选波程序
开发者：Vincent
日期：2024年2月4日
更新：2024.03.13 增加输出缩放后反应谱
更新：2024.03.31 增加输出反应谱对比图、各个匹配规则的误差值、选波参数的记录文档
更新：2024.04.11 优化了梯度下降法的初值计算方法
更新：2024.05.07 更新Info.hdh5文件格式（2.0）
更新: 2024.06.23 选波结果写入pickle与origin
更新：2024.06.28 可调用类方法进行地震动直接提取
"""
import os
import sys
import shutil
from pathlib import Path
from PIL import Image
from tkinter import messagebox
from typing import Literal
import h5py
import originpro
import originpro as op
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from scipy.interpolate import interp1d
from SeismicUtils.Records import Records
class Selecting:
version = '2.2'
RSN_expected = set([i for i in range(1, 21541)]) # 官网宣称有的RSN（但实际不全）
df_info_columns = ['No.', 'RSN', 'earthquake_name', 'component', 'Rjb (km)', 'R_rup (km)',
'Tp-pluse (s)', 'arias Intensity (m/s)','5-75% Duration (s)',
'5-95% Duration (s)', 'duration (s)', 'magnitude', 'mechanism', 'station','Vs30 (m/s)',
'year', 'PGA (g)', 'PGV (mm/s)', 'PGD (mm)', 'dt', 'NPTS', 'scale factor', 'Norm. error']
file_accec = None
file_vel = None
file_disp = None
file_spec = None
file_info = None
def __init__(self, output_dir: Path | str):
"""基于NGA West2地震动数据库的选波程序
Args:
output_dir (str | Path): 输出文件夹路径
"""
self.T_spec = np.arange(0, 10.01, 0.01)
self.T_targ = None
self.Sa_targ = None
self.approach = None
self.para_scaling = None
self.rules = None
self.para_match = None
self.range_scale_factor = None
self.range_PGA = None
self.range_magnitude = None
self.range_Rjb = None
self.range_Rrup = None
self.range_vs30 = None
self.range_D5_95 = None
self.range_strike_slip = 'all'
self.range_pulse = 'all'
self.range_N_events = None
self.range_RSN = None
self.range_component = ['H1', 'H2', 'V']
self.norm_weight = None
self.selecting_text = ''
self.records = Records() # 导出的波库
# 打开文件
output_dir = Path(output_dir)
self._create_folder(output_dir)
self.output_dir = output_dir
@staticmethod
def _create_folder(folder: Path):
if not folder.exists():
os.makedirs(folder)
else:
res = messagebox.askokcancel('警告', f'"{folder.absolute()}"已存在，是否删除？', icon='warning')
if res:
shutil.rmtree(folder)
os.makedirs(folder)
print('已删除')
else:
res1 = messagebox.askokcancel('警告', '是否覆盖？', icon='warning')
if res1:
print('将覆盖数据')
else:
print('退出选波')
return
@classmethod
def import_files(cls,
file_acc: str | Path,
file_vel: str | Path,
file_disp: str | Path,
file_spec: str | Path,
file_info: str | Path,
):
"""类方法，无需实例化对象来导入波库"""
files = [file_acc, file_vel, file_disp, file_spec, file_info]
cls.file_accec = file_acc
cls.file_vel = file_vel
cls.file_disp = file_disp
cls.file_spec = file_spec
cls.file_info = file_info
for file in files:
print(f'正在校验文件 - {file}')
cls._check_file(file)
def check_database(self):
"""进行数据库统计"""
print('正在统计地震动本地数据库...')
if not self:
self = Selecting
if not all([self.file_accec, self.file_vel, self.file_disp,
self.file_spec, self.file_info]):
print('【Warining】请先导入五个hdf5文件')
return
# 版本号
f_info = h5py.File(self.file_info, 'r')
version = f_info['VERSION'][()].decode('utf-8')
print(f'数据库版本：{version}')
# 检查RSN数量
RSN_exists = set()
for item in f_info:
if item == 'VERSION':
continue
ds = f_info[item]
RSN_exists.add(ds.attrs['RSN'])
f_info.close()
# 检查地震波分量总数量
f_accec = h5py.File(self.file_accec, 'r')
n = 0
for item in f_accec:
n += 1
f_accec.close()
# 缺失地震波
RSN_missing = list(self.RSN_expected - RSN_exists)
RSN_missing = sorted(RSN_missing)
print(f'库存地震动：{len(RSN_exists)}组，共{n}条')
print(f'缺失地震动：{len(RSN_missing)}组')
print('缺失地震动RSN：')
for i in range(0, 999999, 100):
print(RSN_missing[i: i + 100])
if i + 100 >= len(RSN_missing):
break
def target_spectra(self, file: str | Path, plot_spectrum: bool=False, scale: float=1):
"""定义目标谱（两列数据，周期(s)-谱值(g)）
Args:
file (str | Path): 文件路径
plot_spectrum (bool, optional): 是否绘制规范谱，默认False
scale (float, optional): 将目标谱进行缩放，缩放系数默认为1
"""
data = np.loadtxt(file)
self.T_targ0 = data[:, 0] # 目标谱周期
self.Sa_targ0 = data[:, 1] * scale # 目标谱加速度
if max(self.T_targ0) > 10:
raise ValueError('【Error】目标谱周期范围不应超过10s')
self.T_targ = np.arange(self.T_targ0[0], self.T_targ0[-1], 0.01)
linear_interp = interp1d(self.T_targ0, self.Sa_targ0, kind='linear', fill_value=0, bounds_error=False)
self.Sa_targ = linear_interp(self.T_targ) # 将目标谱转换为0.01步长
if plot_spectrum:
plt.plot(self.T_targ0, self.Sa_targ0)
plt.show()
def scaling_approach(self, approach: str, para: float | tuple=None):
"""定义地震动缩放方法
Args:
approach (str): 缩放方法
* [a] 按Sa(0)(即PGA)匹配反应谱, para=None or Sa
* [b] 按Sa(Ta)匹配反应谱, para=Ta or (Ta, Sa)
* [c] 按Sa(Ta~Tb)匹配反应谱(几何平均数), 最小化RMSE, para=(Ta, Tb)
* [d] 按Sa,avg(Ta~Tb)匹配反应谱, para=(Ta, Tb) or (Ta, Tb, Sa)
* [e] 指定缩放系数, para=SF\n
para (float | tuple): 缩放参数，与`approach`的取值有关
"""
self.approach = approach
self.para_scaling = para
self._write('反应谱缩放方式：', end='')
match approach:
case 'a':
self._write('按PGA缩放')
if para:
self._write(f'(已指定PGA={para:.4f}g)')
case 'b':
if type(para) is not tuple:
self._write(f'按Sa({round(para, 6)})缩放')
else:
self._write(f'按Sa({round(para[0], 6)})缩放')
self._write(f'(已指定a({round(para[0], 6)})={round(para[1], 6)}g)')
case 'c':
self._write(f'按{round(para[0], 6)}~{round(para[1], 6)}周期范围进行缩放(令RSME最小)')
case 'd':
self._write(f'按{round(para[0], 6)}~{round(para[1], 6)}周期范围内的Sa_avg进行缩放(几何平均数)')
case 'e':
self._write(f'指定缩放系数({para})')
def matching_rules(self, rules: list[str], para: list[float | tuple], weight: list=None):
"""定义地震动匹配规则，可选多种，依次判断
Args:
rules (list[str]): 匹配规则
* [full] 按照给定的反应谱的最大周期范围完全匹配(归一化军方法误差NRSME最小), para=[None]
* [a] 按Sa(0)(即PGA)匹配反应谱, para=[None]
* [b] 按Sa(Ta)匹配反应谱, para=[Ta]
* [c] 按Sa(Ta~Tb)匹配反应谱(几何平均数), 最小化RMSE, para=[(Ta, Tb)]
* [d] 按Sa,avg(Ta~Tb)匹配反应谱, para=[(Ta, Tb)]\n
para (list[float | tuple]): 缩放参数，与`approach`的取值有关
weight (list[tuple]): 各匹配规则的权重系数
Example:
>>> matching_rules(rules=['full', 'b', 'c'], para=[None, 1.5, (1, 2)], weight=[1, 1.2, 0.8])
表示在匹配反应谱（计算拟合误差）时综合考虑全周期范围、Sa(1.5)处谱值和Sa(1~2)范围的匹配程度，
对应的权重依次为1、1.2、和0.8
Note:
`scaling_approach`方法中选用的`approach`参数不得作为该方法的`rules`参数
（这表示之前定义的缩放方法不得用于判断反应谱匹配程度）
"""
if (self.approach in rules) and (self.approach not in ['c', 'd']):
raise ValueError(f'【Error】地震动缩放方法("{self.approach}")不得再次作为匹配规则！')
self.rules = rules
self.para_match = para
if not weight:
weight = [1] * len(rules)
self.norm_weight = [i / sum(weight) for i in weight]
if len(rules) != len(para):
raise ValueError('【Error】参数rules和para的长度应一致')
self._write('反应谱的匹配规则：')
for i, rule in enumerate(rules):
match rule:
case 'full':
self._write(f'({i+1}) 按目标谱全周期范围的RSME值进行匹配，权重={weight[i]}')
case 'a':
self._write(f'({i+1}) 按PGA匹配，权重={weight[i]}')
case 'b':
self._write(f'({i+1}) 按Sa({round(para[i], 6)})匹配，权重={weight[i]}')
case 'c':
self._write(f'({i+1}) 按{round(para[i][0], 6)}~{round(para[i][1], 6)}周期范围的RSME值进行匹配，权重={weight[i]}')
case 'd':
self._write(f'({i+1}) 按{round(para[i][0], 6)}~{round(para[i][1], 6)}周期范围的Sa_avg值(几何平均数)进行匹配，权重={weight[i]}')
def constrain_range(self, scale_factor: tuple=None, PGA: tuple=None, magnitude: tuple=None, Rjb: tuple=None, Rrup: tuple=None,
vs30: tuple=None, D5_95: tuple=None, duration: tuple=None, strike_slip: str='all', pulse: str | bool='all',
N_events: int=None, RSN: tuple=None, component: list=['H1', 'H2', 'V']):
"""定义约束范围
Args:
scale_factor (tuple, optional): 缩放系数，默认None
PGA (tuple, optional): PGA，默认None
magnitude (tuple, optional): 震级，默认None
Rjb (tuple, optional): 默认None
Rrup (tuple, optional): 默认None
vs30 (tuple, optional): 剪切波速，默认None
D5_95 (tuple, optional): 有效持时，默认None
duration (tuple, optional): 持时，默认None
strike_slip (str, optional): 振源机制，默认'all'
* [all] all types
* [a] strike slip
* [b] normal/oblique
* [c] reverse/oblique
* [d] strike slip + normal/oblique
* [e] strike slip + reverse/oblique
* [f] normal/oblique + reverse/oblique\n
pulse (str | bool, optional): 脉冲型地震，默认'all'
* [all] 不限定范围
* [True] 仅脉冲型
* [False] 仅非脉冲型\n
N_events (int, optional): 相同地震事件所允许的最大出现次数，默认None
RSN (tuple, optional): RSN，默认None
component (list, optional): 地震动分量，默认['H1', 'H2', 'V']，可根据需要删减列表元素
"""
self.range_scale_factor = scale_factor
self.range_PGA = PGA
self.range_magnitude = magnitude
self.range_Rjb = Rjb
self.range_Rrup = Rrup
self.range_vs30 = vs30
self.range_D5_95 = D5_95
self.duration = duration
self.range_strike_slip = strike_slip
self.range_pulse = pulse
self.range_N_events = N_events
if RSN:
self.range_RSN = (int(RSN[0]), int(RSN[1]))
self.range_component = component
def run(self, number: int) -> tuple[list[str], dict]:
"""选波计算
Args:
number (int): 需要的地震动数量
Returns:
list[str]: 包括所有选波结果地震动名（无后缀）的列表
dict: {地震动名: (缩放系数, 匹配误差)}
"""
records = self.records
print('正在进行初步筛选...')
files_within_range = [] # 约束范围内（除了PGA,scale_factor,N_events）的备选波
f_spec = h5py.File(self.file_spec, 'r')
f_info = h5py.File(self.file_info, 'r')
# 1 初步筛选
for i, item in enumerate(f_info):
# if i == 1000:
# print(f' --------------- 调试模式，只考虑数据库中前{i}条地震波 --------------- ')
# break # NOTE: for test
if item == 'VERSION':
continue
print(f' {int(i/len(f_info)*100)}% \r', end='')
ds = f_info[item]
H1_file = ds.attrs['H1_file']
H2_file = ds.attrs['H2_file']
V_file = ds.attrs['V_file']
RSN = int(ds.attrs['RSN'])
Rjb = float(ds.attrs['Rjb'])
try:
Rrup = float(ds.attrs['Rrup'])
except KeyError:
Rrup = float(ds.attrs['Rrub'])
Tp = ds.attrs['Tp']
Tp = float(Tp) if type(Tp)==np.float64 else str(Tp)
arias = ds.attrs['arias']
arias = float(arias) if type(arias)!=str else ''
duration_5_75 = ds.attrs['duration_5_75']
duration_5_75 = float(duration_5_75) if type(duration_5_75)!=str else ''
duration_5_95 = ds.attrs['duration_5_95']
duration_5_95 = float(duration_5_95) if type(duration_5_95)!=str else ''
duration = float(ds.attrs['duration'])
earthquake_name = ds.attrs['earthquake_name']
magnitude = float(ds.attrs['magnitude'])
mechanism = ds.attrs['mechanism']
station = ds.attrs['station']
try:
vs30 = float(ds.attrs['vs30'])
except KeyError:
vs30 = float(ds.attrs['v30'])
year = int(ds.attrs['year'])
if self.range_magnitude and not self.range_magnitude[0] <= magnitude <= self.range_magnitude[1]:
continue
if self.range_Rjb and not self.range_Rjb[0] <= Rjb <= self.range_Rjb[1]:
continue
if self.range_Rrup and not self.range_Rrup[0] <= Rrup <= self.range_Rrup[1]:
continue
if self.range_vs30 and not self.range_vs30[0] <= vs30 <= self.range_vs30[1]:
continue
if self.range_D5_95:
if type(duration_5_95) is not float:
continue
if not self.range_D5_95[0] <= duration_5_95 <= self.range_D5_95[1]:
continue
if self.duration:
if not self.duration[0] <= duration <= self.duration[1]:
continue
if self.range_strike_slip != 'all':
if self.range_strike_slip == 'a' and mechanism != 'strike slip':
continue
elif self.range_strike_slip == 'b' and mechanism not in ['Normal Oblique', 'Normal']:
continue
elif self.range_strike_slip == 'c' and mechanism != 'Reverse Oblique':
continue
elif self.range_strike_slip == 'd' and mechanism not in ['strike slip', 'Normal Oblique']:
continue
elif self.range_strike_slip == 'e' and mechanism not in ['strike slip', 'Reverse Oblique']:
continue
elif self.range_strike_slip == 'f' and mechanism not in ['Normal Oblique', 'Reverse Oblique']:
continue
if self.range_pulse != 'all':
if self.range_pulse and type(Tp) is str:
continue
elif not self.range_pulse and type(Tp) is float:
continue
if self.range_RSN and not self.range_RSN[0] <= RSN <= self.range_RSN[1]:
continue
if 'H1' in self.range_component:
files_within_range.append(H1_file)
if 'H2' in self.range_component:
files_within_range.append(H2_file)
if 'V' in self.range_component and V_file != '-':
files_within_range.append(V_file)
# 2 二次筛选
print('正在进行选波计算...')
T = self.T_spec # 0-10s
file_SF = {} # {地震动: 缩放系数}
file_error = {} # {地震动: 匹配误差}
file_event = {} # {地震动: 事件}
file_PGA = {} # {地震动: PGA}
for i, file in enumerate(files_within_range):
# print(f'i = {i} \r', end='')
print(f' {int(i/len(files_within_range)*100)}% \r', end='')
ds = f_spec[file]
RSN = str(ds.attrs['RSN'])
event = f_info['RSN'+RSN].attrs['earthquake_name']
file_event[file] = event
PGA = ds.attrs['PGA']
file_PGA[file] = PGA
Sa = ds[:] # 当前地震动反应谱谱值
if self.approach == 'a':
if self.para_scaling:
Sa0 = self.para_scaling # 目标值
else:
if min(self.T_targ > 0):
raise ValueError('【Error】反应谱缺少T=0的谱值')
Sa0 = self._get_y(self.T_targ, self.Sa_targ, 0)
SF = Sa0 / Sa[0]
elif self.approach == 'b':
if type(self.para_scaling) is tuple and len(self.para_scaling) == 2:
Sa_a, Ta = self.para_scaling # 目标值
else:
Ta = self.para_scaling
Sa_a = self._get_y(self.T_targ, self.Sa_targ, Ta)
Sa_spec = self._get_y(T, Sa, Ta) # 当前值
SF = Sa_a / Sa_spec
elif self.approach == 'c':
Ta, Tb = self.para_scaling
learning_rate = 0.01 # 学习率
num_iterations = 1000 # 迭代次数
Sa_spec_list = Sa[(Ta<=T) & (T<=Tb)] # 当前值
Sa_targ_list = self.Sa_targ[(Ta<=self.T_targ) & (self.T_targ<=Tb)] # 目标值
init_SF = np.mean(Sa_targ_list) / np.mean(Sa_spec_list) # 初始缩放系数
SF = self._gradient_descent(Sa_spec_list, Sa_targ_list, init_SF, learning_rate, num_iterations)
elif self.approach == 'd':
if len(self.para_scaling) == 2:
Ta, Tb = self.para_scaling
Sa_list_targ = self.Sa_targ[(Ta<=self.T_targ) & (self.T_targ<=Tb)]
Sa_avg_targ = self._geometric_mean(Sa_list_targ) # 目标值
elif len(self.para_scaling) == 3:
Ta, Tb, Sa_avg_targ = self.para_scaling # 目标值
Sa_list_spec = Sa[(Ta<=T) & (T<=Tb)]
Sa_avg_spec = self._geometric_mean(Sa_list_spec) # 当前值
SF = Sa_avg_targ / Sa_avg_spec
elif self.approach == 'e':
SF = self.para_scaling
else:
raise ValueError('【Error】参数approach错误')
# 3 计算匹配分数 (缩放后值-目标谱值)/(目标谱值)
error = 0 # 误差
error_ls = []
Sa *= SF # 当前地震动缩放后的反应谱
for j, rule in enumerate(self.rules):
weight = self.norm_weight[j]
para = self.para_match[j]
if rule == 'full':
Ta = min(self.T_targ)
Tb = max(self.T_targ)
Sa_spec_list = Sa[(Ta<=T) & (T<=Tb)]
Sa_targ_list = self.Sa_targ[(Ta<=self.T_targ) & (self.T_targ<=Tb)]
NRMSE = self._RMSE(Sa_spec_list, Sa_targ_list) / np.mean(Sa_targ_list)
error += NRMSE * weight
error_ls.append(NRMSE)
elif rule == 'a':
Ta = 0
Sa_spec_0 = self._get_y(T, Sa, Ta)
Sa_targ_0 = self._get_y(self.T_targ, self.Sa_targ, Ta)
error += abs(Sa_spec_0 - Sa_targ_0) / Sa_targ_0 * weight
error_ls.append((Sa_spec_0 - Sa_targ_0) / Sa_targ_0)
elif rule == 'b':
Ta = para
Sa_spec_0 = self._get_y(T, Sa, Ta)
Sa_targ_0 = self._get_y(self.T_targ, self.Sa_targ, Ta)
error += abs(Sa_spec_0 - Sa_targ_0) / Sa_targ_0 * weight
error_ls.append((Sa_spec_0 - Sa_targ_0) / Sa_targ_0)
elif rule == 'c':
Ta, Tb = para
Sa_spec_list = Sa[(Ta<=T) & (T<=Tb)]
Sa_targ_list = self.Sa_targ[(Ta<=self.T_targ) & (self.T_targ<=Tb)]
NRMSE = self._RMSE(Sa_spec_list, Sa_targ_list) / np.mean(Sa_targ_list)
error += NRMSE * weight
error_ls.append(NRMSE)
elif rule == 'd':
Ta, Tb = para
Sa_spec_list = Sa[(Ta<=T) & (T<=Tb)]
Sa_targ_list = self.Sa_targ[(Ta<=self.T_targ) & (self.T_targ<=Tb)]
Sa_spec_avg = self._geometric_mean(Sa_spec_list)
Sa_targ_avg = self._geometric_mean(Sa_targ_list)
error += abs(Sa_spec_avg - Sa_targ_avg) / Sa_targ_avg * weight
error_ls.append((Sa_spec_avg - Sa_targ_avg) / Sa_targ_avg)
else:
raise ValueError('【Error】参数rule错误')
file_SF[file] = SF
file_error[file] = (error, error_ls)
# 4 筛选缩放系数，地震动事件数量，PGA
if self.range_scale_factor:
SF_a, SF_b = self.range_scale_factor
for file, SF in file_SF.copy().items():
if not SF_a <= SF <= SF_b:
del file_SF[file]
del file_error[file]
if self.range_N_events:
event_number = {} # {地震动: 已出现的次数}
for file in file_SF.copy().keys():
event = file_event[file]
if event not in event_number.keys():
event_number[event] = 1
continue
if event_number[event] < self.range_N_events:
event_number[event] += 1
continue
del file_SF[file]
del file_error[file]
if self.range_PGA:
PGA_a, PGA_b = self.range_PGA
for file in file_SF.copy().keys():
PGA = file_PGA[file]
if not PGA_a <= PGA <= PGA_b:
del file_SF[file]
del file_error[file]
if len(file_SF) < number:
self._write(f'【Warning】符合条件的地震动数量({len(file_SF)})小于期望值({number})')
number = len(file_SF)
print(f'符合条件的地震动数量：{len(file_SF)}')
files_selection = sorted(file_error, key=lambda k: file_error[k][0])[:number] # 选波结果（带排列）
# 绘制反应谱
f_accec = h5py.File(self.file_accec, 'r')
Sa_sum = np.zeros(len(T))
label = 'Individual'
individual_spec = T
for file in files_selection:
Sa = f_spec[file][:] * file_SF[file]
Sa_sum += Sa
plt.plot(T, Sa, color='#A6A6A6', label=label)
individual_spec = np.column_stack((individual_spec, Sa))
if label:
label = None
plt.plot(self.T_targ0, self.Sa_targ0, label='Target', color='black', lw=3)
plt.plot(T, Sa_sum / len(files_selection), color='red', label='Mean', lw=3)
plt.xlim(min(self.T_targ0), max(self.T_targ0))
plt.title('Selected records')
plt.xlabel('T [s]')
plt.ylabel('Sa [g]')
plt.legend()
f_info.close()
f_spec.close()
f_accec.close()
records.target_spec = np.column_stack((self.T_targ0, self.Sa_targ0))
records.individual_spec = individual_spec
records.mean_spec = np.column_stack((T, Sa_sum / len(files_selection)))
file_SF_error = {} # {地震名: (缩放系数, 匹配误差)}
for file in files_selection:
SF = file_SF[file]
error, error_ls = file_error[file][0], file_error[file][1]
file_SF_error[file] = (SF, error, error_ls)
return files_selection, file_SF_error
def get_results(self, files: list=[], file_SF_error: dict[str, tuple[float, float, list]]={},
write_unscaled_record: bool=True, write_norm_record: bool=True, write_scaled_records: bool=True):
"""提取地震动数据
Args:
RSN (int, optional): 按给定的单个RSN序号提取，默认None
RSN_list (list, optional): 按RSN列表提取，默认None
RSN_range (list, optional): 按RSN范围提取，默认None
files (list, optional): 按地震动文件名提取，默认[]
file_SF_error (dict, optional): 地震文件-缩放系数字典，默认{}
write_unscaled_record (bool, optional): 是否写入未缩放地震动，默认True
write_norm_record (bool, optional): 是否写入归一化地震动，默认True
write_scaled_records (bool, optional): 是否写入缩放后地震动，默认True
"""
records = self.records
output_dir = self.output_dir
plt.savefig(output_dir/'反应谱-规范谱对比.jpg', dpi=600)
img = Image.open(output_dir/'反应谱-规范谱对比.jpg')
records.img = img
print('选波完成，请查看反应谱曲线')
plt.show()
print('正在提取地震动...')
if write_unscaled_record:
self._new_folder(output_dir/'未缩放地震动')
if write_norm_record:
self._new_folder(output_dir/'归一化地震动')
if write_scaled_records:
self._new_folder(output_dir/'缩放后地震动')
f_info = h5py.File(self.file_info, 'r')
f_spec = h5py.File(self.file_spec, 'r')
f_A = h5py.File(self.file_accec, 'r')
f_V = h5py.File(self.file_vel, 'r')
f_D = h5py.File(self.file_disp, 'r')
# 读取数据
df_info_columns = self.df_info_columns
N = len(files) # 地震动数量
N_T = len(self.T_spec) # 周期点数量
df_info_columns += [f'error_{i}' for i in self.rules]
df_info = pd.DataFrame(pd.NA, columns=df_info_columns, index=range(N))
df_spec = pd.DataFrame(pd.NA, columns=['T (s)']+[f'No. {i}' for i in range(1, N + 1)]+['Mean'], index=range(N_T)) # 未缩放反应谱
df_spec['T (s)'] = self.T_spec
df_scaled_spec = pd.DataFrame(pd.NA, columns=['T (s)']+[f'No. {i}' for i in range(1, N + 1)]+['Mean'], index=range(N_T)) # 缩放后反应谱
df_scaled_spec['T (s)'] = self.T_spec
data_spec_sum = np.zeros(len(self.T_spec))
data_scaled_spec_sum = np.zeros(len(self.T_spec))
for i, file_stem in enumerate(files):
print(f'正在写入txt... ({i+1}/{N})\r', end='')
ds_A = f_A[file_stem + '.AT2']
ds_V = f_V[file_stem + '.VT2']
ds_D = f_D[file_stem + '.DT2']
RSN = ds_A.attrs['RSN']
PGA = ds_A.attrs['PGA']
PGV = ds_V.attrs['PGV']
PGD = ds_D.attrs['PGD']
data = ds_A[:]
dt = ds_A.attrs['dt']
NPTS = ds_A.attrs['NPTS']
ds_info = f_info[f'RSN{RSN}']
Rjb = ds_info.attrs['Rjb']
try:
Rrup = ds_info.attrs['Rrup']
except KeyError:
Rrup = ds_info.attrs['Rrub']
Tp = ds_info.attrs['Tp']
arias = ds_info.attrs['arias']
D_5_75 = ds_info.attrs['duration_5_75']
D_5_95 = ds_info.attrs['duration_5_95']
duration = ds_info.attrs['duration']
earthquake_name = ds_info.attrs['earthquake_name']
magnitude = ds_info.attrs['magnitude']
mechanism = ds_info.attrs['mechanism']
station = ds_info.attrs['station']
try:
vs30 = ds_info.attrs['vs30']
except KeyError:
vs30 = ds_info.attrs['v30']
year = ds_info.attrs['year']
if file_stem in file_SF_error.keys():
SF, error, error_ls = file_SF_error[file_stem]
data_scaled = data * SF
else:
SF, error, error_ls = '-', '-', []
data_scaled = data
if ds_info.attrs['H1_file'] == file_stem:
component = 'H1'
elif ds_info.attrs['H2_file'] == file_stem:
component = 'H2'
elif ds_info.attrs['V_file'] == file_stem:
component = 'V'
else:
raise ValueError('【Error】1')
line = [i+1, RSN, earthquake_name, component, Rjb, Rrup, Tp, arias,
D_5_75, D_5_95, duration, magnitude, mechanism, station, vs30,
year, PGA, PGV, PGD, dt, NPTS, SF, error, *error_ls]
df_info.loc[len(df_info.index)] = line
data_spec = f_spec[file_stem][:]
data_scaled_spec = f_spec[file_stem][:] * SF
data_spec_sum += data_spec
data_scaled_spec_sum += data_scaled_spec
df_spec[f'No. {i+1}'] = data_spec
df_scaled_spec[f'No. {i+1}'] = data_scaled_spec
earthquake_name_to_file = earthquake_name.replace('/', '_') # 文件名不得出现"/"、"\"
earthquake_name_to_file = earthquake_name_to_file.replace('\\', '_')
if write_unscaled_record:
np.savetxt(output_dir/'未缩放地震动'/f'No{i+1}_RSN{RSN}_{earthquake_name_to_file}_{NPTS}_{dt}.txt', data)
if write_norm_record:
np.savetxt(output_dir/'归一化地震动'/f'No{i+1}_RSN{RSN}_{earthquake_name_to_file}_{NPTS}_{dt}.txt', self._normalize(data))
if write_scaled_records:
np.savetxt(output_dir/'缩放后地震动'/f'No{i+1}_RSN{RSN}_{earthquake_name_to_file}_{NPTS}_{dt}.txt', data_scaled)
records._add_record(data, data_spec, SF, dt, 'A')
print()
records.info = df_info
data_spec_mean = data_spec_sum / len(files)
data_scaled_spec_mean = data_scaled_spec_sum / len(files)
df_spec['Mean'] = data_spec_mean
df_scaled_spec['Mean'] = data_scaled_spec_mean
df_info.to_csv(output_dir/'地震动信息.csv', index=None)
df_spec.to_csv(output_dir/'未缩放反应谱.csv', index=False)
df_scaled_spec.to_csv(output_dir/'缩放后反应谱.csv', index=False)
file_path = output_dir / f'records.opju'
with WriteOrigin(op, file_path, 'results') as f_op:
print('正在写入origin文件...\r', end='')
f_op.delete_obj('Book1')
wb1 = op.new_book('w')
wb1.lname = '选波信息'
ws: originpro.WSheet = wb1[0]
ws.from_df(df_info) # 写入选波信息
wb2 = op.new_book('w')
wb2.lname = '反应谱'
ws: originpro.WSheet = wb2[0]
ws.from_list(0, records.individual_spec[:, 0], 'T', 's', axis='X')
for col in range(1, records.individual_spec.shape[1]):
ws.from_list(col, records.individual_spec[:, col], 'Sa', 'g', axis='Y')
ws.from_list(col + 1, records.target_spec[:, 0], 'T', 's', axis='X')
ws.from_list(col + 2, records.target_spec[:, 1], 'Sa', 'g', 'Target', axis='Y') # 写入目标谱
ws.from_list(col + 3, records.mean_spec[:, 0], 'T', 's', axis='X')
ws.from_list(col + 4, records.mean_spec[:, 1], 'Sa', 'g', 'Mean', axis='Y') # 写入平均谱
print('已导出origin文件 ')
with open(output_dir/'选波参数设置.txt', 'w') as f:
f.write(self.selecting_text)
records.selecting_text = self.selecting_text # 记录选波设置文本
f_info.close()
f_spec.close()
f_A.close()
f_V.close()
f_D.close()
records._to_pkl('records', output_dir)
print('完成！')
@classmethod
def extract_records(cls,
output_dir: str | Path,
RSN: int=None,
RSN_list: list[int]=[],
RSN_range: list[int, int]=[],
components: list[str]=['H1', 'H2', 'V'],
type_: Literal['A', 'V', 'D']='A'
) -> Records:
"""提取地震动数据的类方法
Args:
output_dir (str | Path): 输出文件夹路径
RSN (int, optional): 按给定的单个RSN序号提取，默认None
RSN_list (list, optional): 按RSN列表提取，默认None
RSN_range (list, optional): 按RSN范围提取，默认None
component (list[str], optional): 地震动分量，默认包含三个分量
type_ (str, optional): 数据类型(加速度、速度或位移)，默认加速度
Returns (Records): Records对象
"""
output_dir = Path(output_dir)
cls._create_folder(output_dir)
RSNs = []
records = Records()
if RSN:
RSNs.append(RSN)
if RSN_list:
RSNs += RSN_list
if RSN_range:
RSNs += list(range(RSN_range[0], RSN_range[1]))
[cls.check_RSN_exists(RSN_) for RSN_ in RSNs]
f_info = h5py.File(cls.file_info, 'r')
f_spec = h5py.File(cls.file_spec, 'r')
f_A = h5py.File(cls.file_accec, 'r')
f_V = h5py.File(cls.file_vel, 'r')
f_D = h5py.File(cls.file_disp, 'r')
file_name: list[str] = []
for _, RSN_ in enumerate(RSNs):
file_name += cls._extract_one_RSN(None, RSN_, f_info, components)
T_spec = np.arange(0, 10.01, 0.01)
N = len(file_name)
N_T = len(T_spec)
df_info = pd.DataFrame(pd.NA, columns=cls.df_info_columns, index=range(N))
df_spec = pd.DataFrame(pd.NA, columns=['T (s)']+[f'No. {i}' for i in range(1, N + 1)]+['Mean'], index=range(N_T)) # 未缩放反应谱
df_spec['T (s)'] = T_spec
data_spec_sum = np.zeros((N_T))
data_scaled_spec_sum = np.zeros((N_T))
records.target_spec = np.column_stack((T_spec, np.zeros(N_T)))
individual_spec = T_spec
cls._new_folder(output_dir/'未缩放地震动')
cls._new_folder(output_dir/'归一化地震动')
cls._new_folder(output_dir/'缩放后地震动')
for i, file_stem in enumerate(file_name):
print(f'正在写入txt... ({i+1}/{N})\r', end='')
ds_A = f_A[file_stem + '.AT2']
ds_V = f_V[file_stem + '.VT2']
ds_D = f_D[file_stem + '.DT2']
RSN = ds_A.attrs['RSN']
PGA = ds_A.attrs['PGA']
PGV = ds_V.attrs['PGV']
PGD = ds_D.attrs['PGD']
if type_ == 'A':
data = ds_A[:]
elif type_ == 'V':
data = ds_V[:]
elif type_ == 'D':
data = ds_D[:]
else:
raise ValueError(f'参数`type_`错误: {type_}')
dt = ds_A.attrs['dt']
NPTS = ds_A.attrs['NPTS']
ds_info = f_info[f'RSN{RSN}']
Rjb = ds_info.attrs['Rjb']
try:
Rrup = ds_info.attrs['Rrup']
except KeyError:
Rrup = ds_info.attrs['Rrub']
Tp = ds_info.attrs['Tp']
arias = ds_info.attrs['arias']
D_5_75 = ds_info.attrs['duration_5_75']
D_5_95 = ds_info.attrs['duration_5_95']
duration = ds_info.attrs['duration']
earthquake_name = ds_info.attrs['earthquake_name']
magnitude = ds_info.attrs['magnitude']
mechanism = ds_info.attrs['mechanism']
station = ds_info.attrs['station']
try:
vs30 = ds_info.attrs['vs30']
except KeyError:
vs30 = ds_info.attrs['v30']
year = ds_info.attrs['year']
SF, error, error_ls = 1, 1, []
data_scaled = data
if ds_info.attrs['H1_file'] == file_stem:
component = 'H1'
elif ds_info.attrs['H2_file'] == file_stem:
component = 'H2'
elif ds_info.attrs['V_file'] == file_stem:
component = 'V'
else:
raise ValueError('【Error】1')
line = [i+1, RSN, earthquake_name, component, Rjb, Rrup, Tp, arias,
D_5_75, D_5_95, duration, magnitude, mechanism, station, vs30,
year, PGA, PGV, PGD, dt, NPTS, SF, error, *error_ls]
# df_info.loc[len(df_info.index)] = line
df_info.iloc[i] = line
data_spec = f_spec[file_stem][:]
individual_spec = np.column_stack((individual_spec, data_spec))
data_scaled_spec = f_spec[file_stem][:] * SF
data_spec_sum += data_spec
data_scaled_spec_sum += data_scaled_spec
df_spec[f'No. {i+1}'] = data_spec
earthquake_name_to_file = earthquake_name.replace('/', '_') # 文件名不得出现"/"、"\"
earthquake_name_to_file = earthquake_name_to_file.replace('\\', '_')
np.savetxt(output_dir/'未缩放地震动'/f'No{i+1}_RSN{RSN}_{earthquake_name_to_file}_{NPTS}_{dt}.txt', data)
np.savetxt(output_dir/'归一化地震动'/f'No{i+1}_RSN{RSN}_{earthquake_name_to_file}_{NPTS}_{dt}.txt', cls._normalize(data))
np.savetxt(output_dir/'缩放后地震动'/f'No{i+1}_RSN{RSN}_{earthquake_name_to_file}_{NPTS}_{dt}.txt', data_scaled)
records.individual_spec = individual_spec
records._add_record(data, data_spec, SF, dt, type_)
print()
records.info = df_info
records.mean_spec = np.column_stack((T_spec, np.mean(records.individual_spec[:, 1:], axis=1)))
df_spec['Mean'] = records.mean_spec[:, 1]
df_info.to_csv(output_dir/'地震动信息.csv', index=None)
df_spec.to_csv(output_dir/'反应谱.csv', index=False)
file_path = output_dir / f'records.opju'
with WriteOrigin(op, file_path, 'results') as f_op:
print('正在写入origin文件...\r', end='')
f_op.delete_obj('Book1')
wb1 = op.new_book('w')
wb1.lname = '选波信息'
ws: originpro.WSheet = wb1[0]
ws.from_df(df_info) # 写入选波信息
wb2 = op.new_book('w')
wb2.lname = '反应谱'
ws: originpro.WSheet = wb2[0]
ws.from_list(0, records.individual_spec[:, 0], 'T', 's', axis='X')
for col in range(1, records.individual_spec.shape[1]):
ws.from_list(col, records.individual_spec[:, col], 'Sa', 'g', axis='Y')
ws.from_list(col + 1, records.target_spec[:, 0], 'T', 's', axis='X')
ws.from_list(col + 2, records.target_spec[:, 1], 'Sa', 'g', 'Target', axis='Y') # 写入目标谱
ws.from_list(col + 3, records.mean_spec[:, 0], 'T', 's', axis='X')
ws.from_list(col + 4, records.mean_spec[:, 1], 'Sa', 'g', 'Mean', axis='Y') # 写入平均谱
print('已导出origin文件 ')
selecting_text = f'通过类方法提取地震动\nRSN: {RSN}\nRSN_list: {RSN_list}\nRSN_range: {RSN_range}\ncomponents: {components}\ntype_: {type_}'
with open(output_dir/'选波参数设置.txt', 'w') as f:
f.write(selecting_text)
records.selecting_text = selecting_text # 记录选波设置文本
label = 'Individual'
for i in range(N):
Sa = individual_spec[:, i + 1]
plt.plot(T_spec, Sa, color='#A6A6A6', label=label)
if label:
label = None
plt.plot(T_spec, records.mean_spec[:, 1], color='red', label='Mean', lw=3)
plt.xlim(0, 6)
plt.title('Selected records')
plt.xlabel('T [s]')
plt.ylabel('Sa [g]')
plt.legend()
plt.savefig(output_dir/'反应谱.jpg', dpi=600)
plt.show()
img = Image.open(output_dir/'反应谱.jpg')
records.img = img
records._to_pkl('records', output_dir)
records.img = None
f_info.close()
f_spec.close()
f_A.close()
f_V.close()
f_D.close()
print('完成！')
return records
@classmethod
def check_RSN_exists(cls, RSN: int):
if not RSN in cls.RSN_expected:
raise ValueError(f'RSN-{RSN}不存在')
def _extract_one_RSN(self,
RSN: int,
f_info: h5py.File,
component: list[str]=['H1']
) -> list[str]:
"""提取一组RSN"""
ds_name = f'RSN{RSN}'
RSN_files = []
if ds_name not in f_info:
if self is not None:
self._write(f'【Warning】数据库缺少RSN{RSN}')
else:
print(f'【Warning】数据库缺少RSN{RSN}')
return []
if 'H1' in component:
if f_info[ds_name].attrs['H1_file'] != '-':
RSN_files.append(f_info[ds_name].attrs['H1_file'])
else:
print(f'【Warning】RSN{RSN}缺少H1分量')
if 'H2' in component:
if f_info[ds_name].attrs['H2_file'] != '-':
RSN_files.append(f_info[ds_name].attrs['H2_file'])
else:
print(f'【Warning】RSN{RSN}缺少H2分量')
if 'V' in component:
if f_info[ds_name].attrs['V_file'] != '-':
RSN_files.append(f_info[ds_name].attrs['V_file'])
else:
print(f'【Warning】RSN{RSN}缺少V分量')
return RSN_files
def _write(self, text: str, end='\n'):
print(text)
self.selecting_text += text + end
@classmethod
def _check_file(cls, file_path: str | Path):
"""检查文件是否存在且为最新版本"""
file_path = Path(file_path)
if not file_path.exists():
raise FileExistsError(f'无法找到文件：{file_path}')
with h5py.File(file_path, 'r') as f:
if 'VERSION' not in f:
version = '1.0'
else:
version = f['VERSION'][()].decode('utf-8')
if not version == cls.version:
raise FileExistsError(f'数据库文件版本过旧（{version} < {cls.version}），请使用update.py进行升级')
@staticmethod
def _get_y(x: np.ndarray, y: np.ndarray, x0: float | int):
"""求曲线在某点处的值"""
for i in range(1, len(x)):
if not x[i] > x[i - 1]:
raise ValueError('【Error】x序列不是单调递增的')
if x0 < x[0] or x0 > x[-1]:
raise ValueError(f'【Error】x0超出范围\nx0 = {x0}, range=[{x[0]}, {x[-1]}]')
for i in range(1, len(x)):
xim1 = x[i - 1]
xi = x[i]
yim1 = y[i - 1]
yi = y[i]
if xim1 <= x0 <= xi:
k = (yi - yim1) / (xi - xim1)
y0 = k * (x0 - xi) + yi
break
return y0
@staticmethod
def _geometric_mean(data):
"""计算几何平均数"""
total = 1
n = len(data)
for i in data:
total *= pow(i, 1 / n)
return total
@staticmethod
def _gradient_descent(a, b, init_SF, learning_rate, num_iterations):
"""梯度下降"""
f = init_SF
for _ in range(num_iterations):
error = a * f - b
gradient = 2 * np.dot(error, a) / len(a)
f -= learning_rate * gradient
return f
@staticmethod
def _RMSE(y1, y2):
"""计算均方根值"""
result = np.sqrt(sum((y1 - y2) ** 2) / len(y1))
return result
@staticmethod
def _new_folder(folder_path: str | Path):
"""新建文件夹"""
folder_path = Path(folder_path)
if not folder_path.exists():
os.makedirs(folder_path)
@staticmethod
def _normalize(data: np.ndarray):
"""归一化数组"""
peak = max(abs(data))
data_norm = data / peak
return data_norm
class WriteOrigin():
def __init__(self, op: originpro, opju_file: Path, folder_name:str, set_show: bool=False) -> None:
"""写入origin文件
Args:
op (originpro): originpro对象
opju_file (Path): opju的路径，数据将保存到这个文件，若该opju文件不存在则将创建，若存在则将打开并写入数据
set_show (bool, optional): 是否显示origin窗口, 默认False