File size: 27,951 Bytes
df12872
 
 
 
 
 
2406611
 
 
df12872
 
2382a4f
df12872
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2406611
 
df12872
2406611
df12872
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0615822
2406611
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0615822
df12872
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2406611
 
 
 
 
 
 
 
 
 
df12872
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2406611
df12872
2406611
df12872
2406611
 
df12872
2406611
 
 
 
 
 
 
 
 
 
 
 
 
 
0615822
2406611
 
 
0615822
2406611
0615822
 
 
2406611
0615822
df12872
2406611
df12872
0615822
df12872
0615822
 
 
df12872
0615822
2406611
 
 
0615822
2406611
0615822
 
 
2406611
0615822
2406611
 
 
0615822
2406611
0615822
 
 
2406611
0615822
2406611
 
 
0615822
2406611
0615822
 
 
2406611
0615822
2406611
 
 
0615822
2406611
0615822
 
 
2406611
0615822
2406611
 
 
0615822
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2406611
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
df12872
 
 
 
 
 
 
 
 
 
2406611
 
 
 
 
 
df12872
 
 
 
 
 
2406611
 
df12872
 
 
2406611
 
 
 
 
df12872
 
 
 
 
 
 
 
 
2406611
df12872
 
 
 
 
 
 
 
 
 
 
 
2406611
 
 
df12872
 
 
 
 
2406611
df12872
 
 
2406611
 
df12872
 
 
 
 
 
 
 
 
2406611
 
df12872
 
 
 
 
 
 
 
 
 
 
 
 
0615822
df12872
 
2406611
df12872
 
 
 
 
2406611
df12872
 
2406611
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
df12872
 
 
 
 
 
 
2406611
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
df12872
 
2406611
df12872
2406611
 
df12872
 
2406611
 
 
 
 
df12872
2406611
 
 
 
 
df12872
2406611
 
 
 
 
 
df12872
 
2406611
 
 
df12872
 
2406611
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
df12872
2406611
 
 
df12872
 
 
2406611
df12872
 
 
2382a4f
df12872
2406611
2382a4f
 
df12872
 
 
2406611
 
0615822
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
# -*- coding: utf-8 -*-
"""chem-sim.ipynb

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/1rpq0orE7c2E_K8SsmIeH6gxNjw8ucycA

# Chem simulation using scipy
"""

# !pip install tensorflow==2.15

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from scipy.integrate import solve_ivp
import random
import tensorflow as tf

"""# Dataset

$$
\displaystyle
k = A \cdot e^{-\frac{E_a}{RT}}
$$


k	: Rate constant (what we’re solving for)

A	: Pre-exponential factor (frequency factor)

Ea :	Activation energy (J/mol)

R	: Gas constant 8.314 J/molΒ·K

T :	Temperature (in Kelvin)

| temp: Kelvin | pH: 0–14 scale | Ea: in kJ/mol | A_factor: 1/s |

## Zero order
"""

def zero(t, y, k):
    A, B, C = y
    dA_dt = -k
    dB_dt = 0
    dC_dt = k
    return [dA_dt, dB_dt, dC_dt]

"""## First Order"""

def first(t, y, k):
    A, B, C = y
    dA_dt = -k * A
    dB_dt = 0
    dC_dt = +k * A
    return [dA_dt, dB_dt, dC_dt]

def decay_first(t, y, k):
    A, B, C = y
    dA_dt = -k * A
    dB_dt = 0
    dC_dt = 0
    return [dA_dt, dB_dt, dC_dt]

def reversible_first(t, y, k, k_1):
    A, B, C = y
    dA_dt = -k * A + k_1 * C
    dB_dt = 0
    dC_dt = k * A - k_1 * C
    return [dA_dt, dB_dt, dC_dt]

"""## Second Order"""

def second1(t, y, k):
    A, B, C = y
    dA_dt = -k * A * B
    dB_dt = -k * A * B
    dC_dt = +k * A * B
    return [dA_dt, dB_dt, dC_dt]

def second2(t, y, k):
    A, B, C = y
    dA_dt = -2 * k * A**2
    dB_dt = 0
    dC_dt = +k * A**2
    return [dA_dt, dB_dt, dC_dt]

def reversible_second1(t, y, k, k_1):
    A, B, C = y
    dA_dt = -k * A * B + k_1 * C
    dB_dt = -k * A * B + k_1 * C
    dC_dt = +k * A * B - k_1 * C
    return [dA_dt, dB_dt, dC_dt]

def reversible_second2(t, y, k, k_1):
    A, B, C = y
    dA_dt = -2 * k * A**2 + 2 * k_1 * C
    dB_dt = 0
    dC_dt = +k * A**2 - k_1 * C
    return [dA_dt, dB_dt, dC_dt]

"""## Third order"""

def third1(t, y, k):
    A, B, C = y
    dA_dt = -3 * k * A**3
    dB_dt = 0
    dC_dt = +k * A**3
    return [dA_dt, dB_dt, dC_dt]

def third2(t, y, k):
    A, B, C = y
    dA_dt = -2 * k * A**2 * B
    dB_dt = -1 * k * A**2 * B
    dC_dt = +k * A**2 * B
    return [dA_dt, dB_dt, dC_dt]

def reversible_third1(t, y, k, k_1):
    A, B, C = y
    dA_dt = -3 * k * A**3 + 3 * k_1 * C
    dB_dt = 0
    dC_dt = +k * A**3 - k_1 * C
    return [dA_dt, dB_dt, dC_dt]

def reversible_third2(t, y, k, k_1):
    A, B, C = y
    dA_dt = -2 * k * A**2 * B + 2 * k_1 * C
    dB_dt = -1 * k * A**2 * B + 1 * k_1 * C
    dC_dt = +k * A**2 * B - k_1 * C
    return [dA_dt, dB_dt, dC_dt]

"""## functions"""

def compute_k(temp, Ea, A_factor):
    R = 8.314
    Ea_J = Ea * 1000  # Convert Ea from kJ/mol to J/mol
    k = A_factor * np.exp(-Ea_J / (R * temp))
    return k

def ode1(A0, B0, C0, temp, Ea, A_factor):
  y0 = [A0, B0, C0]
  k = compute_k(temp, Ea, A_factor)
  k_1 = k * random.uniform(0.5, 0.9)

  t_span = (0, 8)
  t_eval = np.linspace(0, 8, 11)

  num = random.randint(0, 11)   # For choosing between different functions randomly

  match num:
    case 0:
      func_name = zero
      is_reversible = 0
      order = 'zero'
    case 1:
      func_name = first
      is_reversible = 0
      order = 'first'
    case 2:
      func_name = decay_first
      is_reversible = 0
      order = 'first'
    case 3:
      func_name = reversible_first
      is_reversible = 1
      order = 'first'
    case 4:
      func_name = second1
      is_reversible = 0
      order = 'second'
    case 5:
      func_name = second2
      is_reversible = 0
      order = 'second'
    case 6:
      func_name = reversible_second1
      is_reversible = 1
      order = 'second'
    case 7:
      func_name = reversible_second2
      is_reversible = 1
      order = 'second'
    case 8:
      func_name = third1
      is_reversible = 0
      order = 'third'
    case 9:
      func_name = third2
      is_reversible = 0
      order = 'third'
    case 10:
      func_name = reversible_third1
      is_reversible = 1
      order = 'third'
    case 11:
      func_name = reversible_third2
      is_reversible = 1
      order = 'third'


  if is_reversible == 1:
    solution = solve_ivp(
      func_name,
      t_span,
      y0,
      args=(k, k_1),
      t_eval=t_eval
    )
  elif is_reversible == 0:
    solution = solve_ivp(
      func_name,
      t_span,
      y0,
      args=(k,),
      t_eval=t_eval
      )


  return solution.t, solution.y[0], solution.y[1], solution.y[2], k, k_1, is_reversible, order

"""## dataframe"""

results = []

counter = 0
while counter < 100000:
    counter += 1

    A0 = round(random.uniform(1.0, 10.0), 2)
    B0 = round(random.uniform(0.0, 5.0), 2)
    C0 = round(random.uniform(0.0, 5.0), 2)
    temp = random.randint(270, 280)
    pH = round(random.uniform(1.0, 14.0), 2)
    Ea = random.randint(90, 100)
    A_factor = round(random.uniform(2e16, 5e17), 2)
    pressure = round(random.uniform(0.5, 5.0), 2)
    weight = round(random.uniform(20, 200), 1)
    structure = random.choice(['Linear', 'Ring', 'Branched', 'Unknown'])
    catalyst = random.choice(['None', 'Enzyme', 'Acid', 'Base'])
    time, A, B, C, k, k_1, is_reversible, order  = ode1(A0, B0, C0, temp, Ea, A_factor)

    row = {
        'order' : order,
        'temp': temp,
        'pH': pH,
        'Ea': Ea,
        'A_factor': A_factor,
        'pressure': pressure,
        'log_pressure' : np.log(pressure),
        'weight': weight,
        'structure': structure,
        'catalyst': catalyst,
        'is_reversible': is_reversible,
        'k' : k,
        'k_1' : k_1,
        'A0': A[0], 'A1': A[1], 'A2': A[2], 'A3': A[3], 'A4': A[4],
        'A5': A[5], 'A6': A[6], 'A7': A[7], 'A8': A[8], 'A9': A[9], 'A10': A[10],
        'B0': B[0], 'B1': B[1], 'B2': B[2], 'B3': B[3], 'B4': B[4],
        'B5': B[5], 'B6': B[6], 'B7': B[7], 'B8': B[8], 'B9': B[9], 'B10': B[10],
        'C0': C[0], 'C1': C[1], 'C2': C[2], 'C3': C[3], 'C4': C[4],
        'C5': C[5], 'C6': C[6], 'C7': C[7], 'C8': C[8], 'C9': C[9], 'C10': C[10]
    }
    results.append(row)

df_train = pd.DataFrame(results)

df_train.to_csv('chem_data_train.csv',index=False)
df_train

results = []

counter = 0
while counter < 20000:
    counter += 1

    A0 = round(random.uniform(1.0, 10.0), 2)
    B0 = round(random.uniform(0.0, 5.0), 2)
    C0 = round(random.uniform(0.0, 5.0), 2)
    temp = random.randint(270, 280)
    pH = round(random.uniform(1.0, 14.0), 2)
    Ea = random.randint(90, 100)
    A_factor = round(random.uniform(2e16, 5e17), 2)
    pressure = round(random.uniform(0.5, 5.0), 2)
    weight = round(random.uniform(20, 200), 1)
    structure = random.choice(['Linear', 'Ring', 'Branched', 'Unknown'])
    catalyst = random.choice(['None', 'Enzyme', 'Acid', 'Base'])
    time, A, B, C, k, k_1, is_reversible, order  = ode1(A0, B0, C0, temp, Ea, A_factor)

    row = {
        'order' : order,
        'temp': temp,
        'pH': pH,
        'Ea': Ea,
        'A_factor': A_factor,
        'pressure': pressure,
        'log_pressure' : np.log(pressure),
        'weight': weight,
        'structure': structure,
        'catalyst': catalyst,
        'is_reversible': is_reversible,
        'k' : k,
        'k_1' : k_1,
        'A0': A[0], 'A1': A[1], 'A2': A[2], 'A3': A[3], 'A4': A[4],
        'A5': A[5], 'A6': A[6], 'A7': A[7], 'A8': A[8], 'A9': A[9], 'A10': A[10],
        'B0': B[0], 'B1': B[1], 'B2': B[2], 'B3': B[3], 'B4': B[4],
        'B5': B[5], 'B6': B[6], 'B7': B[7], 'B8': B[8], 'B9': B[9], 'B10': B[10],
        'C0': C[0], 'C1': C[1], 'C2': C[2], 'C3': C[3], 'C4': C[4],
        'C5': C[5], 'C6': C[6], 'C7': C[7], 'C8': C[8], 'C9': C[9], 'C10': C[10]
    }
    results.append(row)

df_test = pd.DataFrame(results)

df_test.to_csv('chem_data_test.csv',index=False)
df_test

"""- To concatenate df_test and df_train into df"""

df = pd.concat([df_test, df_train])

df

"""# Machine learning

## Data preparation

- removing 'structure' and 'catalyst' from dataframe
- mapping 0 to zero , 1 to first, 2 to second and 3 to third in order column
- mapping structure and catalyst
"""

structure_map = {'Linear': 0, 'Ring': 1, 'Branched': 2, 'Unknown': 3}
catalyst_map = {'None': 0, 'Enzyme': 1, 'Acid': 2, 'Base': 3}
order_map = {'zero': 0, 'first': 1, 'second': 2, 'third' : 3}
df['structure'] = df['structure'].map(structure_map)
df['catalyst'] = df['catalyst'].map(catalyst_map)
df['order'] = df['order'].map(order_map)
df

"""- creating x and y datasets for train and test"""

X = df.drop(['order'], axis=1)
y = df['order']

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

"""- scaling dataset"""

from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)

"""## Models"""

# from sklearn.metrics import accuracy_score

"""### Logistic Regression"""

# from sklearn.linear_model import LogisticRegression

# lr = LogisticRegression(max_iter=1000, C=10, penalty='l2')
# lr.fit(X_train_scaled, y_train)
# lr_pred = lr.predict(X_test_scaled)

# print("Logistic Regression Accuracy:", accuracy_score(y_test, lr_pred))

"""### RandomForestClassifier"""

# from sklearn.ensemble import RandomForestClassifier

# rf = RandomForestClassifier(class_weight='balanced', random_state=42, n_estimators=200, max_depth=None)
# rf.fit(X_train, y_train)
# rf_pred = rf.predict(X_test)

# print("RandomForestClassifier Accuracy:", accuracy_score(y_test, rf_pred))

"""### Gradient Boosting Classifier"""

# from sklearn.ensemble import GradientBoostingClassifier

# gb = GradientBoostingClassifier(n_estimators=200, max_depth=5, random_state=42)
# gb.fit(X_train, y_train)
# gb_pred = gb.predict(X_test)

# print("Gradient Boosting Accuracy:", accuracy_score(y_test, gb_pred))

"""### Support Vector Classifier"""

# from sklearn.svm import SVC

# svc = SVC(C=10, kernel='rbf', class_weight='balanced')
# svc.fit(X_train_scaled, y_train)
# svc_pred = svc.predict(X_test_scaled)

# print("SVC Accuracy:", accuracy_score(y_test, svc_pred))

"""### K-Nearest Neighbors"""

# from sklearn.neighbors import KNeighborsClassifier

# knn = KNeighborsClassifier(n_neighbors=7, weights='uniform')
# knn.fit(X_train_scaled, y_train)
# knn_pred = knn.predict(X_test_scaled)

# print("KNN Accuracy:", accuracy_score(y_test, knn_pred))

"""### XG Boost"""

# from xgboost import XGBClassifier

# xgb_model = XGBClassifier(learning_rate=0.1, max_depth=7, n_estimators=200, eval_metric='mlogloss', random_state=42)
# xgb_model.fit(X_train, y_train)
# xgb_pred = xgb_model.predict(X_test)

# print("XGBoost Accuracy:", accuracy_score(y_test, xgb_pred))

"""### Hyperparameter tuning"""

# from sklearn.linear_model import LogisticRegression
# from sklearn.svm import SVC
# from sklearn.neighbors import KNeighborsClassifier
# from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
# import xgboost as xgb

# models = {
#     'LogisticRegression': LogisticRegression(class_weight='balanced', max_iter=1000),
#     'SVC': SVC(class_weight='balanced'),
#     'KNN': KNeighborsClassifier(),
#     'RandomForest': RandomForestClassifier(class_weight='balanced', random_state=42),
#     'GradientBoosting': GradientBoostingClassifier(random_state=42),
#     'XGBoost': xgb.XGBClassifier(eval_metric='mlogloss', random_state=42)
# }


# param_grids = {
#     'LogisticRegression': {
#         'C': [0.1, 1, 10],
#         'penalty': ['l2']
#     },
#     'SVC': {
#         'C': [0.1, 1, 10],
#         'kernel': ['linear', 'rbf']
#     },
#     'KNN': {
#         'n_neighbors': [3, 5, 7],
#         'weights': ['uniform', 'distance']
#     },
#     'RandomForest': {
#         'n_estimators': [100, 200],
#         'max_depth': [5, 10, None]
#     },
#     'GradientBoosting': {
#         'n_estimators': [100, 200],
#         'max_depth': [3, 5, 7]
#     },
#     'XGBoost': {
#         'n_estimators': [100, 200],
#         'max_depth': [3, 5, 7],
#         'learning_rate': [0.05, 0.1]
#     }
# }

# from sklearn.model_selection import GridSearchCV

# best_models = {}

# for name, model in models.items():
#     print(f"Running GridSearch for {name}...")
#     grid = GridSearchCV(model, param_grids[name], cv=5, scoring='accuracy')

#     if name in ['LogisticRegression', 'SVC', 'KNN']:
#         grid.fit(X_train_scaled, y_train)
#     else:
#         grid.fit(X_train, y_train)

#     best_models[name] = grid.best_estimator_
#     print(f"Best params for {name}:", grid.best_params_)
#     print("Best CV Score:", grid.best_score_)
#     print("=====================================")

"""### BEST PARAMS
==========================================================================
- LogisticRegression
==========================================================================
Best params for LogisticRegression: {'C': 10, 'penalty': 'l2'}
Best CV Score: 0.8008333333333335
==========================================================================
- SVC
==========================================================================
Best params for SVC: {'C': 10, 'kernel': 'rbf'}
Best CV Score: 0.8791666666666668
==========================================================================
- KNN
==========================================================================
Best params for KNN: {'n_neighbors': 7, 'weights': 'uniform'}
Best CV Score: 0.5670833333333334
==========================================================================
- RandomForest
==========================================================================
Best params for RandomForest: {'max_depth': None, 'n_estimators': 200}
Best CV Score: 0.8362499999999999
==========================================================================
- GradientBoosting
==========================================================================
Best params for GradientBoosting: {'max_depth': 5, 'n_estimators': 200}
Best CV Score: 0.8945833333333333
==========================================================================
- XGBOOST
==========================================================================
Best params for XGBOOST: {'learning_rate': 0.1, 'max_depth': 7, 'n_estimators': 200}
Best CV Score: 0.8950000000000001
==========================================================================

## DNN
"""

csv_columns = ['temp', 'pH', 'Ea', 'A_factor', 'pressure', 'log_pressure', 'weight', 'structure', 'catalyst', 'is_reversible', 'k', 'k_1']
classes = ['First_Order','Second_Order','Third_Order']

train_path = './chem_data_train.csv'
test_path = './chem_data_train.csv'

train = pd.read_csv(train_path)
test = pd.read_csv(test_path)

train.head()

"""- Fill missing values in the 'catalyst' column
- NaN values arenot accepted by classifier thats why convert every Nan values to none
- the species column is now gone
"""

if 'order' in train.columns:
    train_y = train.pop('order')
if 'order' in test.columns:
    test_y = test.pop('order')


train['catalyst'] = train['catalyst'].fillna('None')
test['catalyst'] = test['catalyst'].fillna('None')


train.head()

"""- Define categorical and numerical feature columns
- Assining each string a numerical uinque value because our dumb ahh model canot understand english
"""

CATEGORICAL_COLUMNS = ['structure', 'catalyst'] #columns that have strings
NUMERIC_COLUMNS = ['temp', 'pH', 'Ea', 'A_factor', 'pressure', 'log_pressure', 'weight',
                   'is_reversible', 'k', 'k_1', 'A0', 'A1', 'A2', 'A3', 'A4', 'A5', 'A6', 'A7', 'A8', 'A9', 'A10',
                   'B0', 'B1', 'B2', 'B3', 'B4', 'B5', 'B6', 'B7', 'B8', 'B9', 'B10',
                   'C0', 'C1', 'C2', 'C3', 'C4', 'C5', 'C6', 'C7', 'C8', 'C9', 'C10'] #columns that have numerical values

feature_columns = []
for feature_name in CATEGORICAL_COLUMNS:
  vocabulary = train[feature_name].unique()
  cat_column = tf.feature_column.categorical_column_with_vocabulary_list(feature_name, vocabulary)
  indicator_column = tf.feature_column.indicator_column(cat_column) #it creates binary coolumns that will be mapped in to feature columns and it will be steamlined to our DNN model
  feature_columns.append(indicator_column)

for feature_name in NUMERIC_COLUMNS:
  feature_columns.append(tf.feature_column.numeric_column(feature_name, dtype=tf.float32))

print(feature_columns)

import logging
tf.get_logger().setLevel(logging.INFO)

"""- setting up input function
- convert the inputs to a dataset
"""

def input_fn(features,labels,training=True,batch_size=500):
  dataset = tf.data.Dataset.from_tensor_slices((dict(features), labels)) #this cnonverts the dataset into tensorflow object

  if training:
    dataset = dataset.shuffle(3000).repeat()

  return dataset.batch(batch_size)

"""- Normalize the numerical features in the training data"""

from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
train_normalized = train.copy()
train_normalized[NUMERIC_COLUMNS] = scaler.fit_transform(train[NUMERIC_COLUMNS])

test_normalized = test.copy()
test_normalized[NUMERIC_COLUMNS] = scaler.transform(test[NUMERIC_COLUMNS])

"""- Convert the 'order' labels to numerical values"""

from sklearn.preprocessing import LabelEncoder

le = LabelEncoder()
train_y_encoded = le.fit_transform(train_y) #we used sckit label encoder to encode the values

classifier = tf.estimator.DNNClassifier(
    feature_columns=feature_columns,
    hidden_units=[50, 40],
    n_classes=4,  # We have 4 classes: zero, first, second, third
    optimizer=tf.keras.optimizers.legacy.RMSprop(learning_rate=0.001))

classifier.train(
    input_fn=lambda: input_fn(train_normalized, train_y_encoded, training=True),
    steps=3000
)

test_y_encoded = le.fit_transform(test_y) #we used sckit label encoder to encode the values better than 1 2 3 4 5

classifier.evaluate(input_fn=lambda: input_fn(test_normalized,test_y_encoded,training=False))

"""- accuracy = 0.99983335

## Interactive
"""

def predict_order(inputs):

  try:
    # Create a pandas DataFrame from the input dictionary
    input_df = pd.DataFrame(inputs, index=[0])

    # Normalize the numerical features
    input_df[NUMERIC_COLUMNS] = scaler.transform(input_df[NUMERIC_COLUMNS])

    # Make a prediction
    predictions = classifier.predict(input_fn=lambda: input_fn(input_df, labels=None, training=False))

    # Get the predicted class and probability
    for pred_dict in predictions:
      class_id = pred_dict['class_ids'][0]
      probability = pred_dict['probabilities'][class_id]
      # Get the class name from the label encoder
      class_name = le.inverse_transform([class_id])[0]
      print('Order is "{}" ({:.1f}%)'.format(class_name, 100 * probability))
      return class_name
  except Exception as e:
    print(f"An error occurred: {e}")
    return None

#example input data

example_inputs = {
    'temp': 277,
    'pH': 6.5,
    'Ea': 93,
    'A_factor': 4.2e17,
    'pressure': 3.0,
    'log_pressure': 1.1,
    'weight': 150,
    'structure': 'Ring',
    'catalyst': 'Acid',
    'is_reversible': 1,
    'k': 0.05,
    'k_1': 0.02,
    'A0': 5.0,
    'A1': 4.5,
    'A2': 4.0,
    'A3': 3.5,
    'A4': 3.0,
    'A5': 2.5,
    'A6': 2.0,
    'A7': 1.5,
    'A8': 1.0,
    'A9': 0.5,
    'A10': 0.0,
    'B0': 2.0,
    'B1': 1.8,
    'B2': 1.6,
    'B3': 1.4,
    'B4': 1.2,
    'B5': 1.0,
    'B6': 0.8,
    'B7': 0.6,
    'B8': 0.4,
    'B9': 0.2,
    'B10': 0.0,
    'C0': 1.0,
    'C1': 1.2,
    'C2': 1.4,
    'C3': 1.6,
    'C4': 1.8,
    'C5': 2.0,
    'C6': 2.2,
    'C7': 2.4,
    'C8': 2.6,
    'C9': 2.8,
    'C10': 3.0
}

predict_order(example_inputs)

"""- ode2"""

def ode2(A0, B0, C0, temp, Ea, A_factor, is_reversible, predicted_order):
    y0 = [A0, B0, C0]
    k = compute_k(temp, Ea, A_factor)
    k_1 = k * random.uniform(0.5, 0.9) # Assuming k_1 is related to k, similar to ode1

    t_span = (0, 8)
    t_eval = np.linspace(0, 8, 11)

    func_name = None
    if predicted_order == 'zero':
        func_name = zero
    elif predicted_order == 'first':
        if is_reversible:
            func_name = reversible_first
        else:
            # Assuming decay_first is not used for plotting based on predicted order
            func_name = first
    elif predicted_order == 'second':
        if is_reversible:
            # Assuming reversible_second1 or reversible_second2 based on A and B concentrations
            # For simplicity, let's use reversible_second1 if B0 > 0, otherwise reversible_second2
            if B0 > 0:
              func_name = reversible_second1
            else:
              func_name = reversible_second2
        else:
             # Assuming second1 or second2 based on A and B concentrations
             # For simplicity, let's use second1 if B0 > 0, otherwise second2
            if B0 > 0:
              func_name = second1
            else:
              func_name = second2
    elif predicted_order == 'third':
        if is_reversible:
          # Assuming reversible_third1 or reversible_third2 based on A and B concentrations
          # For simplicity, let's use reversible_third2 if B0 > 0, otherwise reversible_third1
          if B0 > 0:
            func_name = reversible_third2
          else:
            func_name = reversible_third1
        else:
           # Assuming third1 or third2 based on A and B concentrations
           # For simplicity, let's use third2 if B0 > 0, otherwise third1
          if B0 > 0:
            func_name = third2
          else:
            func_name = third1


    if func_name is None:
        raise ValueError(f"Could not determine ODE function for predicted order: {predicted_order}")

    if is_reversible and predicted_order != 'zero': # Add condition to exclude zero order
        solution = solve_ivp(
            func_name,
            t_span,
            y0,
            args=(k, k_1),
            t_eval=t_eval
        )
    else: # Handle zero order separately, regardless of is_reversible
        solution = solve_ivp(
            func_name,
            t_span,
            y0,
            args=(k,),
            t_eval=t_eval
        )

    return solution.t, solution.y[0], solution.y[1], solution.y[2], k, k_1

"""### Gradio"""



import gradio as gr
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from scipy.integrate import solve_ivp
import random
import tensorflow as tf
from sklearn.preprocessing import StandardScaler, LabelEncoder

# Assuming all the necessary functions (compute_k, ode2, predict_order, etc.) and models are defined and trained in the previous cells.

def run_simulation_and_plot(temp, Ea, A_factor_base, A_factor_exponent, A_factor_std_perc, pH, pressure, is_reversible, structure, catalyst, A0, B0, C0):
    # --- 1. Data Preparation for Prediction ---
    # Reconstruct A_factor from user-friendly inputs
    A_factor = A_factor_base * (10**A_factor_exponent)
    A_factor_std = A_factor * (A_factor_std_perc / 100)

    # Add randomness to A_factor using standard deviation
    A_factor_randomized = np.random.normal(A_factor, A_factor_std)

    k = compute_k(temp, Ea, A_factor_randomized)
    k_1 = k * 0.7  # Using a fixed ratio for k_1 for consistency

    # Simulate reaction to get concentration data for prediction
    time_sim, A_sim, B_sim, C_sim, _, _ = ode2(A0, B0, C0, temp, Ea, A_factor_randomized, int(is_reversible), "zero")

    inputs = {
        'temp': temp, 'pH': pH, 'Ea': Ea, 'A_factor': A_factor_randomized,
        'pressure': pressure, 'log_pressure': np.log(pressure), 'weight': 150,
        'structure': structure, 'catalyst': catalyst, 'is_reversible': int(is_reversible),
        'k': k, 'k_1': k_1,
        'A0': A_sim[0], 'A1': A_sim[1], 'A2': A_sim[2], 'A3': A_sim[3], 'A4': A_sim[4],
        'A5': A_sim[5], 'A6': A_sim[6], 'A7': A_sim[7], 'A8': A_sim[8], 'A9': A_sim[9], 'A10': A_sim[10],
        'B0': B_sim[0], 'B1': B_sim[1], 'B2': B_sim[2], 'B3': B_sim[3], 'B4': B_sim[4],
        'B5': B_sim[5], 'B6': B_sim[6], 'B7': B_sim[7], 'B8': B_sim[8], 'B9': B_sim[9], 'B10': B_sim[10],
        'C0': C_sim[0], 'C1': C_sim[1], 'C2': C_sim[2], 'C3': C_sim[3], 'C4': C_sim[4],
        'C5': C_sim[5], 'C6': C_sim[6], 'C7': C_sim[7], 'C8': C_sim[8], 'C9': C_sim[9], 'C10': C_sim[10],
    }

    # --- 2. Prediction ---
    predicted_order = predict_order(inputs)

    # --- 3. Final Simulation with Predicted Order ---
    time_final, A_final, B_final, C_final, _, _ = ode2(A0, B0, C0, temp, Ea, A_factor_randomized, int(is_reversible), predicted_order)

    # --- 4. Plotting ---
    plt.style.use('seaborn-v0_8-whitegrid')
    fig, ax = plt.subplots(figsize=(10, 6))
    ax.plot(time_final, A_final, 'o-', label='[A]', color='royalblue', markersize=5)
    ax.plot(time_final, B_final, 's--', label='[B]', color='forestgreen', markersize=5)
    ax.plot(time_final, C_final, '^-.', label='[C]', color='darkorange', markersize=5)

    ax.set_xlabel('Time (s)', fontsize=12)
    ax.set_ylabel('Concentration (M)', fontsize=12)
    ax.set_title(f'πŸ§ͺ Concentration vs. Time (Predicted Order: {predicted_order})', fontsize=14)
    ax.legend(loc='best', fontsize=10)
    ax.grid(True, which='both', linestyle='--', linewidth=0.5)

    # Add watermark
    fig.text(0.99, 0.01, 'pinl',
             fontsize=12, color='gray',
             ha='right', va='bottom', alpha=0.5)

    return f"Predicted Order: {predicted_order}", fig

# --- 5. Gradio Interface ---
with gr.Blocks(theme=gr.themes.Soft()) as iface:
    gr.Markdown("# Project E-11: πŸ§ͺ Chemical Reaction Simulator", elem_id="title" "made by Team PinlAI")
    gr.Markdown("An interactive tool to predict reaction orders and visualize concentration changes over time.", elem_id="subtitle")

    with gr.Row():
        with gr.Column(scale=1):
            gr.Markdown("### βš™οΈ Reaction Parameters")
            temp = gr.Slider(270, 280, value=277, label="🌑️ Temperature (K)")
            Ea = gr.Slider(90, 100, value=93, label="⚑ Activation Energy (kJ/mol)")
            A_factor_base = gr.Slider(1, 9, value=4, label="πŸ…°οΈ Pre-exponential Factor (Base)")
            A_factor_exponent = gr.Slider(16, 18, value=17, step=1, label="πŸ…°οΈ Pre-exponential Factor (Exponent)")
            A_factor_std_perc = gr.Slider(0, 50, value=10, label="πŸ“ˆ A Factor Std Dev (%)")
            pH = gr.Slider(1.0, 14.0, value=6.5, label="πŸ’§ pH")
            pressure = gr.Slider(0.5, 5.0, value=3.0, label="πŸ’¨ Pressure (atm)")
            is_reversible = gr.Checkbox(label="πŸ”„ Reversible Reaction")
            structure = gr.Dropdown(['Linear', 'Ring', 'Branched', 'Unknown'], label="🧬 Molecular Structure")
            catalyst = gr.Dropdown(['None', 'Enzyme', 'Acid', 'Base'], label="πŸ”¬ Catalyst")

        with gr.Column(scale=1):
            gr.Markdown("### βš›οΈ Initial Concentrations")
            A0 = gr.Slider(0.0, 10.0, value=5.0, label="[A]β‚€")
            B0 = gr.Slider(0.0, 10.0, value=2.0, label="[B]β‚€")
            C0 = gr.Slider(0.0, 10.0, value=1.0, label="[C]β‚€")

            with gr.Row():
                predict_button = gr.Button("πŸš€ Predict & Plot", variant="primary")

    with gr.Row():
        with gr.Column(scale=2):
            order_output = gr.Textbox(label="πŸ“Š Predicted Reaction Order")
            plot_output = gr.Plot(label="πŸ“ˆ Concentration vs. Time")

    predict_button.click(
        fn=run_simulation_and_plot,
        inputs=[temp, Ea, A_factor_base, A_factor_exponent, A_factor_std_perc, pH, pressure, is_reversible, structure, catalyst, A0, B0, C0],
        outputs=[order_output, plot_output]
    )

iface.launch(debug=True)

"""### Streamlit"""

# !npm install -g localtunnel

# !streamlit run /content/app.py &>/content/logs.txt &  #this starts the loca server

# get_ipython().run_line_magic('shell', 'curl https://loca.lt/mytunnelpassword') #getting ur home pass πŸ₯Ά

# !npx localtunnel --port 8501 #the tunnel