-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathconverter.py
1406 lines (1222 loc) · 64.3 KB
/
converter.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
#!/usr/bin/env python
import modules.conditions as conditions
import modules.setting as setting
import numpy as np
import pandas as pd
import itertools
import scipy.sparse as sparse
from scipy.stats import norm
import collections
import sys
sys.dont_write_bytecode = True
# 多重リストを平滑化する関数。
def flatten(l):
for el in l:
if isinstance(el, collections.abc.Iterable) and not isinstance(el, (str, bytes)):
yield from flatten(el)
else:
yield el
# 任意の整数の約数一覧(昇順)を作成する。
def divisors_list(num):
divisors = []
for i in range(1, num+1):
if num % i == 0:
divisors.append(i)
return divisors
# ある整数(total_num)を、ある大きさ(size)以上で割り切れる最小の整数を計算する。
def num_jobs(total_num, divisors, size=int(1E6)):
catal_comb_size = []
for i in range(len(divisors)):
catal_comb_size.append(total_num/divisors[i])
if catal_comb_size[i] > size:
slice_size = int(catal_comb_size[i])
num_slice = int(total_num / slice_size)
if total_num < size:
slice_size = total_num
num_slice = 1
return slice_size, num_slice
# 計算に用いる元素リストを作成する。
def elem_list(condition, converted, desc_cols):
pgm_num, prm_num = condition['pgm_num'], condition['prm_num']
add_num, supp_num = condition['add_num'], condition['supp_num']
drop_pgms, drop_prms, drop_adds = condition['drop_pgms'], condition['drop_prms'], condition['drop_adds']
essential_supp, drop_supp_list = condition['essential_supp'], condition['drop_supp_list']
CalT_list, ReacT_list, WHSV_list = condition['CalT_list'], condition['ReacT_list'], condition['WHSV_list']
H2_ppm_list, CO_ppm_list = condition['H2_ppm_list'], condition['CO_ppm_list']
NH3_ppm_list, C3H6_ppm_list = condition['NH3_ppm_list'], condition['C3H6_ppm_list']
CH4_ppm_list = condition['CH4_ppm_list']
pgm_desc, prm_desc = converted['pgm_desc'], converted['prm_desc']
add_desc, supp_desc = converted['add_desc'], converted['supp_desc']
supp_name_dict = desc_cols['supp_name_dict']
if pgm_num != 0:
all_pgm_list = pgm_desc.index[1:]
if len(drop_pgms) == 0:
use_pgm_list = all_pgm_list
else:
use_pgm_list = list(set(all_pgm_list) - set(drop_pgms))
else:
all_pgm_list, use_pgm_list = [], []
if prm_num != 0:
all_prm_list = prm_desc.index[1:]
if len(drop_prms) == 0:
use_prm_list = all_prm_list
else:
use_prm_list = list(set(all_prm_list) - set(drop_prms))
else:
all_prm_list, use_prm_list = [], []
if add_num != 0:
all_add_list = add_desc.index[1:]
if len(drop_adds) == 0:
use_add_list = all_add_list
else:
use_add_list = list(set(all_add_list) - set(drop_adds))
else:
all_add_list, use_add_list = [], []
if supp_num != 0:
all_supp_list = supp_desc.index
if len(drop_supp_list) == 0:
use_supp_list = all_supp_list
else:
use_supp_list = list(set(all_supp_list) - set(drop_supp_list))
if len(essential_supp) != 0:
for i in range(len(essential_supp)):
essential_supp[i] = supp_name_dict[essential_supp[i]]
use_supp_list = essential_supp
else:
all_supp_list, use_supp_list = [], []
elem_lists_dict = {
'all_pgm': all_pgm_list, 'use_pgm': use_pgm_list,
'all_prm': all_prm_list, 'use_prm': use_prm_list,
'all_add': all_add_list, 'use_add': use_add_list,
'all_supp': all_supp_list, 'use_supp': use_supp_list,
'CalT': CalT_list, 'ReacT': ReacT_list, 'WHSV': WHSV_list,
'H2_ppm': H2_ppm_list, 'CO_ppm': CO_ppm_list,
'NH3_ppm': NH3_ppm_list, 'C3H6_ppm': C3H6_ppm_list,
'CH4_ppm': CH4_ppm_list,
}
return elem_lists_dict
# 探索する触媒組成に応じたdictを作成(後の関数内で呼び出す)
def cols_order(condition):
cols_order = {}
cand_pgm_num, cand_prm_num = condition['cand_pgm_num'], condition['cand_prm_num']
cand_add_num, cand_supp_num = condition['cand_add_num'], condition['cand_supp_num']
CalT_num, ReacT_num, WHSV_num = condition['CalT_num'], condition['ReacT_num'], condition['WHSV_num']
H2_ppm_num, CO_ppm_num = condition['H2_ppm_num'], condition['CO_ppm_num']
NH3_ppm_num, C3H6_ppm_num = condition['NH3_ppm_num'], condition['C3H6_ppm_num']
CH4_ppm_num = condition['CH4_ppm_num']
if cand_pgm_num != 0 and cand_prm_num != 0 and cand_add_num != 0 \
and cand_supp_num != 0 and CalT_num != 0 and ReacT_num != 0 and WHSV_num != 0:
cols_order = {'pgm': 0, 'prm': 1, 'add': 2,
'supp': 3, 'CalT': 4, 'ReacT': 5, 'WHSV': 6}
elif cand_pgm_num != 0 and cand_prm_num != 0 and cand_add_num != 0 \
and cand_supp_num != 0 and CalT_num != 0 and ReacT_num != 0 and WHSV_num == 0:
cols_order = {'pgm': 0, 'prm': 1, 'add': 2,
'supp': 3, 'CalT': 4, 'ReacT': 5}
elif cand_pgm_num != 0 and cand_prm_num != 0 and cand_add_num != 0 \
and cand_supp_num != 0 and CalT_num != 0 and ReacT_num == 0 and WHSV_num != 0:
cols_order = {'pgm': 0, 'prm': 1, 'add': 2,
'supp': 3, 'CalT': 4, 'WHSV': 5}
elif cand_pgm_num != 0 and cand_prm_num == 0 and cand_add_num != 0 \
and cand_supp_num != 0 and CalT_num != 0 and ReacT_num == 0 and WHSV_num == 0:
cols_order = {'pgm': 0, 'add': 1, 'supp': 2, 'CalT': 3}
elif cand_pgm_num != 0 and cand_prm_num == 0 and cand_add_num != 0 \
and cand_supp_num != 0 and CalT_num == 0 and ReacT_num == 0 and WHSV_num == 0:
if H2_ppm_num != 0 or CO_ppm_num != 0 or NH3_ppm_num != 0 or C3H6_ppm_num != 0 or CH4_ppm_num != 0:
cols_order = {'pgm': 0, 'add': 1, 'supp': 2,
'H2_ppm': 3, 'CO_ppm': 4, 'NH3_ppm': 5, 'C3H6_ppm': 6, 'CH4_ppm': 7}
else:
cols_order = {'pgm': 0, 'add': 1, 'supp': 2}
elif cand_pgm_num == 0 and cand_prm_num == 0 and cand_add_num != 0 \
and cand_supp_num != 0 and CalT_num != 0 and ReacT_num == 0 and WHSV_num == 0:
cols_order = {'add': 0, 'supp': 1, 'CalT': 2}
elif cand_pgm_num == 0 and cand_prm_num == 0 and cand_add_num != 0 \
and cand_supp_num == 0 and CalT_num == 0 and ReacT_num == 0 and WHSV_num == 0:
cols_order = {'add': 0}
return cols_order
# 実験データを読み込む関数
def read_data(condition, desc_cols, ReacT_range=[], read_sheet_name='data_sheet_name', idx=None):
date, Reaction = condition['date'], condition['Reaction']
data_sheet_name, skip_rows = condition[read_sheet_name], condition['skip_rows']
supp_num = condition['supp_num']
supp_name_dict = desc_cols['supp_name_dict']
data_file_name = f'data/{date}_{Reaction}_dataset.xlsx'
data = pd.read_excel(data_file_name, sheet_name=data_sheet_name, skiprows=skip_rows)
if Reaction == 'EtOH_CO2' or Reaction == 'EtOH_CO':
if len(ReacT_range) == 2:
data = data[data.loc[:, 'Reaction T'] >= ReacT_range[0]].reset_index()
data = data[data.loc[:, 'Reaction T'] <= ReacT_range[1]].reset_index()
elif len(ReacT_range) == 1:
data = data[data.loc[:, 'Reaction T'] >= ReacT_range[0]].reset_index()
data_cols = setting.data_columns(condition)
elem_cols, wt_cols = data_cols['elem'], data_cols['wt']
data.loc[:, elem_cols] = data.loc[:, elem_cols].fillna('H')
data.loc[:, wt_cols] = data.loc[:, wt_cols].fillna(0)
if supp_num != 0:
for i in range(len(data)):
data.loc[i, 'Support_name'] = supp_name_dict[data.loc[i, 'Support_name']]
if Reaction == 'N2O-SCR':
gas_cols = data_cols['gas_cols']
data.loc[:, gas_cols] = data.loc[:, gas_cols].fillna(0)
if idx is not None:
#print(Reaction, 'Iteration =', idx)
data_idx = data[data['Iteration'] <= idx]
data_idx = data_idx.reset_index()
data = data_idx
return data
# Descriptorファイルを読み込む
def read_desc(elem, condition, desc_cols, select_desc=True, local=False, standard=False):
Reaction = condition['Reaction']
pgm_num, prm_num = condition['pgm_num'], condition['prm_num']
add_num, supp_num = condition['add_num'], condition['supp_num']
elem_comp_list = [pgm_num, prm_num, add_num, supp_num]
desc_file_name = 'data/Descriptors.xlsx'
if elem == 'pgm' or elem == 'prm' or elem == 'add' or elem == 'all':
desc_sheet_name = 'Descriptors_elem'
use_desc = desc_cols['basic_desc_columns']
elem_desc_dict = desc_cols['elem_desc_dict']
for i, j in enumerate(use_desc):
if j in list(elem_desc_dict.keys()):
use_desc[i] = elem_desc_dict[j]
desc = pd.read_excel(desc_file_name, sheet_name=desc_sheet_name, index_col='Symbol')
desc = desc.rename(columns=elem_desc_dict)
elif elem == 'supp':
desc_sheet_name = 'Descriptors_supp'
use_desc = desc_cols['all_supp_desc']
supp_desc_dict = desc_cols['supp_desc_dict']
for i, j in enumerate(use_desc):
if j in list(supp_desc_dict.keys()):
use_desc[i] = supp_desc_dict[j]
desc = pd.read_excel(desc_file_name, sheet_name=desc_sheet_name, index_col='Support_name_python')
desc = desc.rename(columns=supp_desc_dict)
all_elems = list(desc.index)
noble_gas = desc_cols['noble_gas']
drop_elems = desc_cols['drop_elems']
if elem == 'pgm':
use_elem = desc_cols['pgm_plus_ReAu']
if Reaction == 'N2O':
desc = desc.loc[use_elem].drop('Os', axis=0)
else:
desc = desc.loc[use_elem].drop(['Re', 'Os'], axis=0)
desc = desc[use_desc].fillna(desc.mean())
if len(elem_comp_list) - elem_comp_list.count(0) != 1:
desc = desc.rename(columns=lambda s: s+' [PGMs]')
aw = desc.loc[:, 'AW [PGMs]']
else:
aw = desc.loc[:, 'AW']
elif elem == 'prm':
use_elem = ['H'] + desc_cols['prm_elems']
desc = desc.loc[use_elem]
desc = desc[use_desc].fillna(desc.mean())
if len(elem_comp_list) - elem_comp_list.count(0) != 1:
desc = desc.rename(columns=lambda s: s+' [Promoters]')
aw = desc.loc[:, 'AW [Promoters]']
else:
aw = desc.loc[:, 'AW']
elif elem == 'add':
use_elem = [i for i in all_elems if i not in noble_gas+drop_elems]
desc = desc.loc[use_elem]
desc = desc[use_desc].fillna(desc.mean())
if len(elem_comp_list) - elem_comp_list.count(0) != 1:
desc = desc.rename(columns=lambda s: s+' [Additives]')
aw = desc.loc[:, 'AW [Additives]']
else:
aw = desc.loc[:, 'AW']
elif elem == 'supp':
use_elem = desc.index[desc[Reaction] == 1]
desc = desc.loc[use_elem]
#desc = desc.set_index('Support_name_python', inplace=True)
desc = desc[use_desc].replace('-', 0)
desc = desc[use_desc].fillna(desc.mean())
if len(elem_comp_list) - elem_comp_list.count(0) != 1:
desc = desc.rename(columns=lambda s: s+' [Supports]')
aw = desc.loc[:, 'MW [Supports]']
else:
aw = desc.loc[:, 'MW']
elif elem == 'all':
desc = desc[use_desc].fillna(desc.mean())
aw = desc.loc[:, 'AW']
if local == False:
selec_desc = conditions.desc_select(Reaction)
# else:
#selec_desc = desc_select(Reaction)
if elem == 'pgm' or elem == 'prm' or elem == 'add':
for i, j in enumerate(selec_desc['use_'+elem]):
if j in list(elem_desc_dict.keys()):
selec_desc['use_'+elem][i] = elem_desc_dict[j]
for i, j in enumerate(selec_desc['drop_'+elem]):
if j in list(elem_desc_dict.keys()):
selec_desc['drop_'+elem][i] = elem_desc_dict[j]
elif elem == 'supp':
for i, j in enumerate(selec_desc['use_'+elem]):
if j in list(supp_desc_dict.keys()):
selec_desc['use_'+elem][i] = supp_desc_dict[j]
for i, j in enumerate(selec_desc['drop_'+elem]):
if j in list(supp_desc_dict.keys()):
selec_desc['drop_'+elem][i] = supp_desc_dict[j]
if select_desc == True:
if elem == 'pgm':
selec_desc['use_'+elem] = [s +' [PGMs]' for s in selec_desc['use_'+elem]]
if elem == 'prm':
selec_desc['use_'+elem] = [s +' [Promoters]' for s in selec_desc['use_'+elem]]
if elem == 'add':
selec_desc['use_'+elem] = [s +' [Additives]' for s in selec_desc['use_'+elem]]
if elem == 'supp':
selec_desc['use_'+elem] = [s +' [Supports]' for s in selec_desc['use_'+elem]]
if elem != 'all':
desc = desc.loc[:, selec_desc['use_'+elem]]
else:
if elem == 'pgm':
selec_desc['drop_'+elem] = [s +' [PGMs]' for s in selec_desc['drop_'+elem]]
if elem == 'prm':
selec_desc['drop_'+elem] = [s +' [Promoters]' for s in selec_desc['drop_'+elem]]
if elem == 'add':
selec_desc['drop_'+elem] = [s +' [Additives]' for s in selec_desc['drop_'+elem]]
if elem == 'supp':
selec_desc['drop_'+elem] = [s +' [Supports]' for s in selec_desc['drop_'+elem]]
if elem != 'all':
desc = desc.drop(selec_desc['drop_'+elem], axis=1)
desc = desc.dropna(how='any', axis=1)
if standard == True:
desc = (desc - desc.mean()) / desc.std(ddof=1)
return desc, aw
# 触媒候補データ(cand_sum)を読み込む関数
def read_cand(condition):
date, Reaction = condition['date'], condition['Reaction']
ML_model, Search_method = condition['ML_model'], condition['Search_method']
pgm_num, prm_num = condition['pgm_num'], condition['prm_num']
add_num, supp_num = condition['add_num'], condition['supp_num']
pgm_model, prm_model = condition['pgm_model'], condition['prm_model']
add_model, supp_model = condition['add_model'], condition['supp_model']
# results_csv_file_all
if pgm_num != 0 and prm_num != 0 and add_num != 0 and supp_num != 0:
cand_file_name = f'results/{date}_{Reaction}_cand_sum_{ML_model}_prop{pgm_model}{prm_model}{add_model}{supp_model}_{Search_method}.csv'
elif pgm_num != 0 and prm_num == 0 and add_num != 0 and supp_num != 0:
cand_file_name = f'results/{date}_{Reaction}_cand_sum_{ML_model}_prop{pgm_model}{add_model}{supp_model}_{Search_method}.csv'
elif pgm_num == 0 and prm_num == 0 and add_num != 0 and supp_num != 0:
cand_file_name = f'results/{date}_{Reaction}_cand_sum_{ML_model}_prop{add_model}{supp_model}_{Search_method}.csv'
elif pgm_num == 0 and prm_num == 0 and add_num != 0 and supp_num == 0:
cand_file_name = f'results/{date}_{Reaction}_cand_sum_{ML_model}_prop{add_model}_{Search_method}.csv'
data_cols = setting.data_columns(condition)
cand_elem, cand_wt = data_cols['cand_elem'], data_cols['cand_wt']
cand = cand_separation(cand_file_name, condition)
cand = cand_str(cand, condition, data_cols)
cand.drop_duplicates(subset=['Top catal.'], inplace=True)
cand.loc[:, cand_elem] = cand.loc[:, cand_elem].fillna('H')
cand.loc[:, cand_wt] = cand.loc[:, cand_wt].fillna(0)
cand = cand.reset_index()
return cand
# Feat.を定義する関数
def make_feat(elem, elem_num, model, data, desc, aw):
if elem == 'PGM' or elem == 'Promoter' or elem == 'Ad':
if elem_num != 0:
feat_comp = pd.DataFrame(
index=np.arange(len(data)), columns=desc.index)
for i in range(len(data)):
for j in range(elem_num):
feat_comp.loc[i, data.loc[i, f'{elem}{j+1}']] = data.loc[i, f'{elem}{j+1}_wt%']
feat_comp = feat_comp.fillna(0)
if 'H' in desc.index:
feat_comp = feat_comp.drop('H', axis=1)
feat_comp_cols = list(desc.index[1:])
else:
feat_comp_cols = list(desc.index)
feat_desc = sum([
np.multiply(np.array(desc.loc[data.loc[:, f"{elem}{i+1}"]]),
np.array(data[f"{elem}{i+1}_wt%"]).reshape(-1, 1)) /
np.array(aw.loc[data.loc[:, f"{elem}{i+1}"]]).reshape(-1, 1) for i in range(elem_num)
])
feat_desc_cols = list(desc.columns)
if model == 0:
feat = feat_comp
feat_cols = feat_comp_cols
elif model == 1:
feat = np.hstack((feat_comp, feat_desc))
feat_cols = list(feat_comp_cols) + list(feat_desc_cols)
elif model == 2:
feat = feat_desc
feat_cols = feat_desc_cols
else:
feat, feat_cols = None, None
elif elem == 'Supp':
if elem_num != 0:
feat_comp = pd.DataFrame(index=np.arange(len(data)), columns=desc.index)
for i in range(len(data)):
feat_comp.loc[i, data.loc[i, 'Support_name']] = 1
feat_comp = feat_comp.fillna(0)
feat_comp_cols = list(desc.index)
feat_desc = desc.loc[data.loc[:, 'Support_name']]
feat_desc = feat_desc.replace('-', 0)
feat_desc_cols = list(desc.columns)
if model == 0:
feat = feat_comp
feat_cols = feat_comp_cols
elif model == 1:
feat = np.hstack((feat_comp, feat_desc))
feat_cols = list(feat_comp_cols) + list(feat_desc_cols)
elif model == 2:
feat = feat_desc
feat_cols = feat_desc_cols
else:
feat, feat_cols = None, None
elif elem == 'Calc T' or elem == 'Reaction T' or elem == 'WHSV (mL g-1 h-1)' or \
elem == 'H2_ppm' or elem == 'CO_ppm' or elem == 'NH3_ppm' or elem == 'C3H6_ppm' or elem == 'CH4_ppm':
if elem_num != 0:
feat = np.array(data.loc[:, elem]).reshape(-1, 1)
feat_cols = [elem]
else:
feat, feat_cols = None, None
return feat, feat_cols
# Targetを定義する関数
def make_target(Reaction, data, name, temp):
if Reaction == 'rwgs_250' or Reaction == 'rwgs_250_1wt' or \
Reaction == 'rwgs_300' or Reaction == 'rwgs_TL' or Reaction == 'CH3OH' or \
Reaction == 'H2SCR' or Reaction == 'EtOH_CO2' or Reaction == 'EtOH_CO':
target = data.loc[:, name]
target = target.fillna(target.mean())
target = target.astype(float)
elif Reaction == 'N2O' or Reaction == 'N2O-SCR' or Reaction == 'CH4':
target = data.loc[:, f'T{temp} (℃)']
target = target.fillna(900) # sufficiently high temperature
target = target.astype(float)
elif Reaction == 'NH3SCR':
target = data.loc[:, f'T{temp}_N2 yield']
target = target.replace('-', 400)
target = target.astype(float)
return target
# SWEDを定義する関数
def make_swed(elem, elem_num, data, desc, aw, data_cols):
if elem_num != 0:
elem_swed_names = []
for i in range(elem_num):
for j in desc.columns:
elem_swed_names.append(f'{j} [{i+1}]')
swed_elem = pd.DataFrame(columns=elem_swed_names)
if elem != 'supp':
for i in range(len(data)):
swed_elem.loc[i] = np.array(
list(
flatten(
[(
np.array(desc.loc[data.loc[i, f"{elem}{j+1}"]]) *
np.array(data.loc[i, f"{elem}{j+1}_wt%"]) /
np.array(aw.loc[data.loc[i, f"{elem}{j+1}"]])
) for j in range(elem_num)]
)))
else:
elem_wt_sum = data.loc[:, data_cols['wt']].sum(axis=1)
supp_wt = 100 - elem_wt_sum
for i in range(len(data)):
swed_elem.loc[i] = np.array(desc.loc[data.loc[i, 'Support_name']]) * \
np.array(supp_wt[i]) / \
np.array(aw.loc[data.loc[i, 'Support_name']])
else:
swed_elem = None
return swed_elem
# datasetと DescriptorのExcelファイルを読み込み、feat, target等必要な形状に加工する。
def data_convert(condition, data_cols, desc_cols, sheet_name='data_sheet_name',
select_desc=True, use_models=[], local=False, idx=None):
converted = {}
data = read_data(condition, desc_cols, ReacT_range=[condition['ReacT_min'], condition['ReacT_max']],
read_sheet_name=sheet_name, idx=idx)
pgm_desc, pgm_aw = read_desc('pgm', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
prm_desc, prm_aw = read_desc('prm', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
add_desc, add_aw = read_desc('add', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
supp_desc, supp_mw = read_desc('supp', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
if len(use_models) == 0:
pgm_model, prm_model = condition['pgm_model'], condition['prm_model']
add_model, supp_model = condition['add_model'], condition['supp_model']
elif len(use_models) == 1:
add_model = use_models[0]
pgm_model, prm_model, supp_model = 2, 2, 2
elif len(use_models) == 2:
add_model, supp_model = use_models[0], use_models[1]
pgm_model, prm_model = 2, 2
elif len(use_models) == 3:
pgm_model, add_model, supp_model = use_models[0], use_models[1], use_models[2]
prm_model = 2
elif len(use_models) == 4:
pgm_model, prm_model = use_models[0], use_models[1]
add_model, supp_model = use_models[0], use_models[1]
feat_pgm, feat_pgm_cols = make_feat('PGM', condition['pgm_num'], pgm_model, data, pgm_desc, pgm_aw)
feat_prm, feat_prm_cols = make_feat('Promoter', condition['prm_num'], prm_model, data, prm_desc, prm_aw)
feat_add, feat_add_cols = make_feat('Ad', condition['add_num'], add_model, data, add_desc, add_aw)
feat_supp, feat_supp_cols = make_feat('Supp', condition['supp_num'], supp_model, data, supp_desc, supp_mw)
feat_CalT, feat_CalT_cols = make_feat('Calc T', condition['CalT_num'], 0, data, add_desc, add_aw)
feat_ReacT, feat_ReacT_cols = make_feat('Reaction T', condition['ReacT_num'], 0, data, add_desc, add_aw)
feat_WHSV, feat_WHSV_cols = make_feat('WHSV (mL g-1 h-1)', condition['WHSV_num'], 0, data, add_desc, add_aw)
feat_H2, feat_H2_cols = make_feat('H2_ppm', condition['H2_ppm_num'], 0, data, add_desc, add_aw)
feat_CO, feat_CO_cols = make_feat('CO_ppm', condition['CO_ppm_num'], 0, data, add_desc, add_aw)
feat_NH3, feat_NH3_cols = make_feat('NH3_ppm', condition['NH3_ppm_num'], 0, data, add_desc, add_aw)
feat_C3H6, feat_C3H6_cols = make_feat('C3H6_ppm', condition['C3H6_ppm_num'], 0, data, add_desc, add_aw)
feat_CH4, feat_CH4_cols = make_feat('CH4_ppm', condition['CH4_ppm_num'], 0, data, add_desc, add_aw)
feat_list = [
feat_pgm, feat_prm, feat_add,
feat_supp, feat_CalT, feat_ReacT, feat_WHSV,
feat_H2, feat_CO, feat_NH3, feat_C3H6, feat_CH4
]
feat_cols_list = [
feat_pgm_cols, feat_prm_cols, feat_add_cols, feat_supp_cols,
feat_CalT_cols, feat_ReacT_cols, feat_WHSV_cols,
feat_H2_cols, feat_CO_cols, feat_NH3_cols, feat_C3H6_cols, feat_CH4_cols
]
feat_list = [i for i in feat_list if i is not None]
feat = np.hstack(feat_list)
feat_cols_list = [i for i in feat_cols_list if i is not None]
feat_cols = list(itertools.chain.from_iterable(feat_cols_list))
#feat_cols = [item for row in feat_cols_list for item in row]
feat = pd.DataFrame(feat, columns=feat_cols)
swed_pgm = make_swed('PGM', condition['pgm_num'], data, pgm_desc, pgm_aw, data_cols)
swed_prm = make_swed('Promoter', condition['prm_num'], data, prm_desc, prm_aw, data_cols)
swed_add = make_swed('Ad', condition['add_num'], data, add_desc, add_aw, data_cols)
swed_supp = make_swed('supp', condition['supp_num'], data, supp_desc, supp_mw, data_cols)
target = make_target(condition['Reaction'], data, condition['target_name'], condition['target_temp'])
converted = {
'data': data,
'pgm_desc': pgm_desc, 'prm_desc': prm_desc, 'add_desc': add_desc, 'supp_desc': supp_desc,
'pgm_aw': pgm_aw, 'prm_aw': prm_aw, 'add_aw': add_aw, 'supp_mw': supp_mw,
'swed_pgm': swed_pgm, 'swed_prm': swed_prm, 'swed_add': swed_add, 'swed_supp': swed_supp,
'feat': feat, 'target': target
}
return converted
# datasetと DescriptorのExcelファイルを読み込み、feat, target等必要な形状に加工する。
def cand_convert(condition, converted, data_cols, desc_cols, select_desc=True, use_models=[], local=False):
cand_dict = {}
data = converted['data']
data = data_str(data, condition, data_cols)
cand = read_cand(condition)
dupl_cat = list(set(data['Catalyst']) & set(cand['Top catal.']))
if len(dupl_cat) != 0:
for i in range(len(dupl_cat)):
cand.drop(list(cand[cand['Top catal.'] == dupl_cat[i]].index), inplace=True)
cand.reset_index(drop=True, inplace=True)
for i in list(cand.columns[:-1]): # drop 'Top catal.'
if type(cand.loc[0, i]) == str and ',' in cand.loc[0, i]:
cand[i] = cand[i].str.strip(',')
if i != 'Support_name':
cand[i] = cand[i].astype(float)
pgm_desc, pgm_aw = read_desc('pgm', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
prm_desc, prm_aw = read_desc('prm', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
add_desc, add_aw = read_desc('add', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
supp_desc, supp_mw = read_desc('supp', condition, desc_cols, select_desc, local, standard=condition['desc_std'])
if len(use_models) == 0:
pgm_model, prm_model = condition['pgm_model'], condition['prm_model']
add_model, supp_model = condition['add_model'], condition['supp_model']
elif len(use_models) == 1:
add_model = use_models[0]
pgm_model, prm_model, supp_model = 2, 2, 2
elif len(use_models) == 2:
add_model, supp_model = use_models[0], use_models[1]
pgm_model, prm_model = 2, 2
elif len(use_models) == 3:
pgm_model, add_model, supp_model = use_models[0], use_models[1], use_models[2]
prm_model = 2
elif len(use_models) == 4:
pgm_model, prm_model = use_models[0], use_models[1]
add_model, supp_model = use_models[0], use_models[1]
feat_pgm, feat_pgm_cols = make_feat('PGM', condition['cand_pgm_num'], pgm_model, cand, pgm_desc, pgm_aw)
feat_prm, feat_prm_cols = make_feat('Promoter', condition['cand_prm_num'], prm_model, cand, prm_desc, prm_aw)
feat_add, feat_add_cols = make_feat('Ad', condition['cand_add_num'], add_model, cand, add_desc, add_aw)
feat_supp, feat_supp_cols = make_feat('Supp', condition['cand_supp_num'], supp_model, cand, supp_desc, supp_mw)
feat_CalT, feat_CalT_cols = make_feat('Calc T', condition['CalT_num'], 0, cand, add_desc, add_aw)
feat_ReacT, feat_ReacT_cols = make_feat('Reaction T', condition['ReacT_num'], 0, cand, add_desc, add_aw)
feat_WHSV, feat_WHSV_cols = make_feat('WHSV (mL g-1 h-1)', condition['WHSV_num'], 0, cand, add_desc, add_aw)
feat_H2, feat_H2_cols = make_feat('H2_ppm', condition['H2_ppm_num'], 0, cand, add_desc, add_aw)
feat_CO, feat_CO_cols = make_feat('CO_ppm', condition['CO_ppm_num'], 0, cand, add_desc, add_aw)
feat_NH3, feat_NH3_cols = make_feat('NH3_ppm', condition['NH3_ppm_num'], 0, cand, add_desc, add_aw)
feat_C3H6, feat_C3H6_cols = make_feat('C3H6_ppm', condition['C3H6_ppm_num'], 0, cand, add_desc, add_aw)
feat_CH4, feat_CH4_cols = make_feat('CH4_ppm', condition['CH4_ppm_num'], 0, cand, add_desc, add_aw)
feat_list = [
feat_pgm, feat_prm, feat_add,
feat_supp, feat_CalT, feat_ReacT, feat_WHSV,
feat_H2, feat_CO, feat_NH3, feat_C3H6, feat_CH4
]
feat_cols_list = [
feat_pgm_cols, feat_prm_cols, feat_add_cols, feat_supp_cols,
feat_CalT_cols, feat_ReacT_cols, feat_WHSV_cols,
feat_H2_cols, feat_CO_cols, feat_NH3_cols, feat_C3H6_cols, feat_CH4_cols
]
feat_list = [i for i in feat_list if i is not None]
feat = np.hstack(feat_list)
feat_cols_list = [i for i in feat_cols_list if i is not None]
feat_cols = list(itertools.chain.from_iterable(feat_cols_list))
#feat_cols = [item for row in feat_cols_list for item in row]
feat = pd.DataFrame(feat, columns=feat_cols)
target = cand.loc[:, 'ei']
cand_dict = {
'cand': cand,
'pgm_desc': pgm_desc, 'prm_desc': prm_desc, 'add_desc': add_desc, 'supp_desc': supp_desc,
'pgm_aw': pgm_aw, 'prm_aw': prm_aw, 'add_aw': add_aw, 'supp_mw': supp_mw,
'feat': feat, 'target': target
}
return cand_dict
# 各元素成分ごとに、wtの組み合わせを作る関数
def elem_wt_comb(elem_wt, ess_elem_wt, cand_elem_num, fix_elem_num, max_elem_wt):
if cand_elem_num != 0:
elem_wt_comb_full = itertools.product(elem_wt, repeat=cand_elem_num-fix_elem_num)
if fix_elem_num != 0:
elem_wt_comb_fix = itertools.product(ess_elem_wt, repeat=fix_elem_num)
elem_wt_comb = itertools.product(elem_wt_comb_fix, elem_wt_comb_full)
else:
elem_wt_comb = itertools.product(elem_wt, repeat=cand_elem_num)
elem_wt_comb = np.array(list(flatten(elem_wt_comb))).reshape(-1, cand_elem_num)
if max_elem_wt is not None:
elem_wt_comb = elem_wt_comb[elem_wt_comb.sum(axis=1) < max_elem_wt]
elem_wt_comb = iter(elem_wt_comb)
else:
elem_wt_comb = None
return elem_wt_comb
# 上で定義したelem_wt_comb関数を成分の数だけ結合し、wtの全ての組み合わせを作成する。
def all_wt_comb(condition):
pgm_wt, prm_wt, add_wt = condition['pgm_wt'], condition['prm_wt'], condition['add_wt']
ess_pgm_wt, ess_prm_wt, ess_add_wt = condition['essential_pgm_wt'], condition['essential_prm_wt'], condition['essential_add_wt']
cand_pgm_num, cand_prm_num = condition['cand_pgm_num'], condition['cand_prm_num'],
cand_add_num, cand_supp_num = condition['cand_add_num'], condition['cand_supp_num']
fix_pgm_num, fix_prm_num, fix_add_num = condition['fix_pgm_num'], condition['fix_prm_num'], condition['fix_add_num']
max_pgm_wt, max_prm_wt, max_add_wt = condition['max_pgm_wt'], condition['max_prm_wt'], condition['max_add_wt']
elem_wt_pgm = elem_wt_comb(pgm_wt, ess_pgm_wt, cand_pgm_num, fix_pgm_num, max_pgm_wt)
elem_wt_prm = elem_wt_comb(prm_wt, ess_prm_wt, cand_prm_num, fix_prm_num, max_prm_wt)
elem_wt_add = elem_wt_comb(add_wt, ess_add_wt, cand_add_num, fix_add_num, max_add_wt)
_lis_wt = [elem_wt_pgm, elem_wt_prm, elem_wt_add]
comb_wt = [z for z in _lis_wt if z]
all_comb_wt = itertools.product(*comb_wt)
contens = len(_lis_wt) - _lis_wt.count(None)
if contens == 1:
elem_wt = [x1[0] for x1 in all_comb_wt]
elif contens == 2:
elem_wt = [(*x1, *x2) for (x1, x2) in all_comb_wt]
elif contens == 3:
elem_wt = [(*x1, *x2, *x3) for (x1, x2, x3) in all_comb_wt]
if cand_supp_num != 0:
elem_wt = [elem_wt[i] + tuple([100-sum(elem_wt[i])]) for i in range(len(elem_wt))]
num_elem_wt = len(elem_wt)
return elem_wt, num_elem_wt
# 必須元素が登場するリストとその他の元素(未登場の必須元素も含まれ得る)が登場するリストをそれぞれ作成。
def gen_fix_and_res(use_elem_list, essential_elems, cand_elem_num, fix_elem_num, duplicate=True):
# duplicate: True: 重複を含む, False: 重複を含まない
residue_elems = list(set(use_elem_list) - set(essential_elems))
elem_comb_fix = list(itertools.combinations(essential_elems, fix_elem_num))
if duplicate == True:
elem_comb_res = list(itertools.combinations(
use_elem_list, cand_elem_num-fix_elem_num))
else:
elem_comb_res = list(itertools.combinations(
residue_elems, cand_elem_num-fix_elem_num))
return elem_comb_fix, elem_comb_res
# 上のgen_fix_and_res関数で作成した必須元素リストと、その他元素リストを結合する。
def fix_and_res_comb(fix_list, res_list):
_lis = [fix_list, res_list]
comb = [z for z in _lis if z is not None]
comb_ = itertools.product(*comb)
comb_list = [(*x1, *x2) for (x1, x2) in comb_]
return comb_list
# 上のfix_and_res_comb関数で作成したリストに含まれる重複(e.g. ['Pt', 'Pt'])成分を落とす。
def drop_duplicate(elem_comb, fix_elem_num):
elem_comb_dup = []
for i in range(len(elem_comb)):
if len(list(set(elem_comb[i][:fix_elem_num]) & set(elem_comb[i][fix_elem_num:]))) == 0:
elem_comb_dup.append(elem_comb[i])
return elem_comb_dup
# drop_duplicate関数で作成した重複を含まないリストを要素ごとに分解する。
def make_elem_comp(elem_comb, cand_elem_num):
elem_comp_dict = {}
for i in range(cand_elem_num):
elem_comp_dict[i] = [x[i] for x in elem_comb]
return elem_comp_dict
# 触媒組成 (+ 実験条件) の全組み合わせを計算する。
def catal_comb(condition, elem_lists):
cand_pgm_num, cand_prm_num = condition['cand_pgm_num'], condition['cand_prm_num']
cand_add_num, cand_supp_num = condition['cand_add_num'], condition['cand_supp_num']
CalT_num, ReacT_num, WHSV_num = condition['CalT_num'], condition['ReacT_num'], condition['WHSV_num']
H2_ppm_num, CO_ppm_num = condition['H2_ppm_num'], condition['CO_ppm_num']
NH3_ppm_num, C3H6_ppm_num = condition['NH3_ppm_num'], condition['C3H6_ppm_num']
CH4_ppm_num = condition['CH4_ppm_num']
fix_pgm_num, fix_prm_num, fix_add_num = condition['fix_pgm_num'], condition['fix_prm_num'], condition['fix_add_num']
essential_pgms, essential_prms, essential_adds = condition['essential_pgms'], condition['essential_prms'], condition['essential_adds']
use_pgm_list, use_prm_list = elem_lists['use_pgm'], elem_lists['use_prm']
use_add_list, use_supp_list = elem_lists['use_add'], elem_lists['use_supp']
CalT_list, ReacT_list, WHSV_list = elem_lists['CalT'], elem_lists['ReacT'], elem_lists['WHSV']
H2_ppm_list, CO_ppm_list = elem_lists['H2_ppm'], elem_lists['CO_ppm']
NH3_ppm_list, C3H6_ppm_list = elem_lists['NH3_ppm'], elem_lists['C3H6_ppm']
CH4_ppm_list = elem_lists['CH4_ppm']
if cand_pgm_num != 0:
if fix_pgm_num != 0:
elem_comb_pgm_fix, elem_comb_pgm_res = \
gen_fix_and_res(use_pgm_list, essential_pgms, cand_pgm_num, fix_pgm_num, duplicate=True)
elem_comb_pgm = fix_and_res_comb(elem_comb_pgm_fix, elem_comb_pgm_res)
del elem_comb_pgm_fix
del elem_comb_pgm_res
elem_comb_pgm = drop_duplicate(elem_comb_pgm, fix_pgm_num)
else:
elem_comb_pgm = itertools.combinations(use_pgm_list, cand_pgm_num)
elem_comb_pgm = list(elem_comb_pgm)
comb_pgm_num = len(elem_comb_pgm)
else:
elem_comb_pgm = None
comb_pgm_num = 1
if cand_prm_num != 0:
if fix_prm_num != 0:
elem_comb_prm_fix, elem_comb_prm_res = \
gen_fix_and_res(use_prm_list, essential_prms, cand_prm_num, fix_prm_num, duplicate=True)
elem_comb_prm = fix_and_res_comb(elem_comb_prm_fix, elem_comb_prm_res)
del elem_comb_prm_fix
del elem_comb_prm_res
elem_comb_prm = drop_duplicate(elem_comb_prm, fix_prm_num)
else:
elem_comb_prm = itertools.combinations(use_prm_list, cand_prm_num)
elem_comb_prm = list(elem_comb_prm)
comb_prm_num = len(elem_comb_prm)
else:
elem_comb_prm = None
comb_prm_num = 1
if cand_add_num != 0:
if fix_add_num != 0:
elem_comb_add_fix, elem_comb_add_res \
= gen_fix_and_res(use_add_list, essential_adds, cand_add_num, fix_add_num, duplicate=True)
elem_comb_add = fix_and_res_comb(elem_comb_add_fix, elem_comb_add_res)
del elem_comb_add_fix
del elem_comb_add_res
elem_comb_add = drop_duplicate(elem_comb_add, fix_add_num)
else:
elem_comb_add = itertools.combinations(use_add_list, cand_add_num)
elem_comb_add = list(elem_comb_add)
comb_add_num = len(elem_comb_add)
else:
elem_comb_add = None
comb_add_num = 1
if cand_supp_num != 0:
elem_comb_supp = use_supp_list
comb_supp_num = len(elem_comb_supp)
else:
elem_comb_supp = None
comb_supp_num = 1
if CalT_num != 0:
elem_comb_CalT = CalT_list
comb_CalT_num = len(elem_comb_CalT)
else:
elem_comb_CalT = None
comb_CalT_num = 1
if ReacT_num != 0:
elem_comb_ReacT = ReacT_list
comb_ReacT_num = len(elem_comb_ReacT)
else:
elem_comb_ReacT = None
comb_ReacT_num = 1
if WHSV_num != 0:
elem_comb_WHSV = WHSV_list
comb_WHSV_num = len(elem_comb_WHSV)
else:
elem_comb_WHSV = None
comb_WHSV_num = 1
if H2_ppm_num != 0:
elem_comb_H2 = H2_ppm_list
comb_H2_num = len(elem_comb_H2)
else:
elem_comb_H2 = None
comb_H2_num = 1
if CO_ppm_num != 0:
elem_comb_CO = CO_ppm_list
comb_CO_num = len(elem_comb_CO)
else:
elem_comb_CO = None
comb_CO_num = 1
if NH3_ppm_num != 0:
elem_comb_NH3 = NH3_ppm_list
comb_NH3_num = len(elem_comb_NH3)
else:
elem_comb_NH3 = None
comb_NH3_num = 1
if C3H6_ppm_num != 0:
elem_comb_C3H6 = C3H6_ppm_list
comb_C3H6_num = len(elem_comb_C3H6)
else:
elem_comb_C3H6 = None
comb_C3H6_num = 1
if CH4_ppm_num != 0:
elem_comb_CH4 = CH4_ppm_list
comb_CH4_num = len(elem_comb_CH4)
else:
elem_comb_CH4 = None
comb_CH4_num = 1
total_comb_num = comb_pgm_num * comb_prm_num * comb_add_num \
* comb_supp_num * comb_CalT_num * comb_ReacT_num * comb_WHSV_num \
* comb_H2_num * comb_CO_num * comb_NH3_num * comb_C3H6_num * comb_CH4_num
_lis = [
elem_comb_pgm, elem_comb_prm, elem_comb_add,
elem_comb_supp, elem_comb_CalT, elem_comb_ReacT, elem_comb_WHSV,
elem_comb_H2, elem_comb_CO, elem_comb_NH3, elem_comb_C3H6, elem_comb_CH4
]
comb = [z for z in _lis if z is not None]
all_comb = itertools.product(*comb)
return all_comb, total_comb_num
# 触媒組成を one-hot表現に変換する関数
def one_hot(read_col_name, elem_list, cand_elem_num, cand_cols_order, condition, elem_lists, slice_size, start, end):
elem_comp_dict = {}
elem_cols = {}
ones = np.ones(slice_size)
rows = np.arange(0, slice_size, 1)
int_code = {x: j for (j, x) in enumerate(elem_list)}
read_col = cand_cols_order[read_col_name]
if cand_elem_num == 1:
all_comb, total_comb_num = catal_comb(condition, elem_lists)
all_comb_slice = itertools.islice(all_comb, start, end)
elem_comp_dict = [u[read_col] for u in all_comb_slice]
elem_cols = np.asarray([int_code[v] for v in elem_comp_dict])
elem_oh = [sparse.csr_matrix(
(ones, (rows, elem_cols)), shape=(slice_size, len(elem_list)))]
else:
for n in range(cand_elem_num):
all_comb, total_comb_num = catal_comb(condition, elem_lists)
all_comb_slice = itertools.islice(all_comb, start, end)
elem_comp_dict[n] = [u[read_col][n] for u in all_comb_slice]
elem_cols[n] = np.asarray([int_code[v] for v in elem_comp_dict[n]])
elem_oh = [sparse.csr_matrix((ones, (rows, elem_cols[n])), shape=(
slice_size, len(elem_list))) for n in range(cand_elem_num)]
return elem_oh
# 実験条件を入力表現に変換する関数
def exp_rep(read_col_name, cand_cols_order, condition, elem_lists, start, end):
read_col = cand_cols_order[read_col_name]
all_comb, total_comb_num = catal_comb(condition, elem_lists)
all_comb_slice = itertools.islice(all_comb, start, end)
exp_dict = [u[read_col] for u in all_comb_slice]
exp_col = np.array(exp_dict).reshape(-1, 1)
exp_rep = sparse.csr_matrix(exp_col)
return exp_rep
# 元素記述子を入力表現に変換する関数
def elem_rep(read_col_name, desc, aw, cand_elem_num, cand_cols_order, condition, elem_lists, start, end):
elem_dict = {}
read_col = cand_cols_order[read_col_name]
all_comb, total_comb_num = catal_comb(condition, elem_lists)
all_comb_slice = itertools.islice(all_comb, start, end)
elem_comp_dict = [u[read_col] for u in all_comb_slice]
if cand_elem_num == 1:
elem_rep = [np.array(desc.loc[elem_comp_dict]) /
np.array(aw.loc[elem_comp_dict]).reshape(-1, 1)]
elem_rep = [sparse.csr_matrix(m) for m in elem_rep]
else:
elem_dict = [np.asarray([i[n] for i in elem_comp_dict]) for n in range(cand_elem_num)]
elem_rep = [np.array(desc.loc[elem_dict[i]]) /
np.array(aw.loc[elem_dict[i]]).reshape(-1, 1) for i in range(cand_elem_num)]
elem_rep = [sparse.csr_matrix(m) for m in elem_rep]
return elem_rep
# 上で定義した3種の関数(one_hot, exp_rep, elem_rep)を使って、EI値計算のための最終的な入力表現を作成する関数
def input_comps(condition, converted, elem_lists, cand_cols_order, slice_size, start, end):
pgm_model, prm_model = condition['pgm_model'], condition['prm_model']
add_model, supp_model = condition['add_model'], condition['supp_model']
cand_pgm_num, cand_prm_num = condition['cand_pgm_num'], condition['cand_prm_num']
cand_add_num, cand_supp_num = condition['cand_add_num'], condition['supp_num']
CalT_num, ReacT_num, WHSV_num = condition['CalT_num'], condition['ReacT_num'], condition['WHSV_num']
H2_ppm_num, CO_ppm_num = condition['H2_ppm_num'], condition['CO_ppm_num'],
NH3_ppm_num, C3H6_ppm_num = condition['NH3_ppm_num'], condition['C3H6_ppm_num']
CH4_ppm_num = condition['CH4_ppm_num']
pgm_desc, prm_desc = converted['pgm_desc'], converted['prm_desc'],
add_desc, supp_desc = converted['add_desc'], converted['supp_desc']
pgm_aw, prm_aw = converted['pgm_aw'], converted['prm_aw']
add_aw, supp_mw = converted['add_aw'], converted['supp_mw']
all_pgm_list, all_prm_list = elem_lists['all_pgm'], elem_lists['all_prm'],
all_add_list, all_supp_list = elem_lists['all_add'], elem_lists['all_supp']
all_comb, total_comb_num = catal_comb(condition, elem_lists)
comb_slice = itertools.islice(all_comb, start, end)
comb_slice_list = list(comb_slice)
if pgm_model == 0 and cand_pgm_num != 0:
pgm_rep = None
pgm_oh = one_hot('pgm', all_pgm_list, cand_pgm_num, cand_cols_order,
condition, elem_lists, slice_size, start, end)
elif pgm_model == 1 and cand_pgm_num != 0:
pgm_rep = elem_rep('pgm', pgm_desc, pgm_aw, cand_pgm_num,
cand_cols_order, condition, elem_lists, start, end)
pgm_oh = one_hot('pgm', all_pgm_list, cand_pgm_num, cand_cols_order,
condition, elem_lists, slice_size, start, end)
elif pgm_model == 2 and cand_pgm_num != 0:
pgm_rep = elem_rep('pgm', pgm_desc, pgm_aw, cand_pgm_num,
cand_cols_order, condition, elem_lists, start, end)
pgm_oh = None
else:
pgm_rep, pgm_oh = None, None
if prm_model == 0 and cand_prm_num != 0:
prm_rep = None
prm_oh = one_hot('prm', all_prm_list, cand_prm_num, cand_cols_order,
condition, elem_lists, slice_size, start, end)
elif prm_model == 1 and cand_prm_num != 0:
prm_rep = elem_rep('prm', prm_desc, prm_aw, cand_prm_num,
cand_cols_order, condition, elem_lists, start, end)
prm_oh = one_hot('prm', all_prm_list, cand_prm_num, cand_cols_order,
condition, elem_lists, slice_size, start, end)
elif prm_model == 2 and cand_prm_num != 0:
prm_rep = elem_rep('prm', prm_desc, prm_aw, cand_prm_num,
cand_cols_order, condition, elem_lists, start, end)
prm_oh = None
else:
prm_rep, prm_oh = None, None
if add_model == 0 and cand_add_num != 0:
add_rep = None
add_oh = one_hot('add', all_add_list, cand_add_num, cand_cols_order,
condition, elem_lists, slice_size, start, end)
elif add_model == 1 and cand_add_num != 0:
add_rep = elem_rep('add', add_desc, add_aw, cand_add_num,
cand_cols_order, condition, elem_lists, start, end)
add_oh = one_hot('add', all_add_list, cand_add_num, cand_cols_order,
condition, elem_lists, slice_size, start, end)
elif add_model == 2 and cand_add_num != 0:
add_rep = elem_rep('add', add_desc, add_aw, cand_add_num,
cand_cols_order, condition, elem_lists, start, end)
add_oh = None
else:
add_rep, add_oh = None, None
if supp_model == 0 and cand_supp_num != 0:
supp_rep = None
supp_oh = one_hot('supp', all_supp_list, cand_supp_num,
cand_cols_order, condition, elem_lists, slice_size, start, end)
elif supp_model == 1 and cand_supp_num != 0: