CatBoostSelection Example
[1]:
import pandas as pd
import numpy as np
from raimitigations.utils import create_dummy_dataset
from raimitigations.dataprocessing import CatBoostSelection
1 - Dataset with Headers
[2]:
df = create_dummy_dataset(
samples=10000,
n_features=6,
n_num_num=2,
n_cat_num=2,
n_cat_cat=1,
num_num_noise=[0.01, 0.02],
pct_change=[0.03, 0.05]
)
label_col = "label"
df
[2]:
| num_0 | num_1 | num_2 | num_3 | num_4 | num_5 | label | num_c0_num_0 | num_c1_num_1 | CN_0_num_0 | CN_1_num_1 | CC_0_num_0 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | -3.086844 | -3.307366 | -1.562467 | 0.250221 | 0.019486 | 1.773897 | 1 | -3.088523 | -3.294641 | val0_1 | val1_4 | val0_1 |
| 1 | -2.817753 | -1.305013 | -2.649276 | 1.577143 | 0.808751 | 2.885842 | 1 | -2.791640 | -1.301471 | val0_1 | val1_0 | val0_1 |
| 2 | 0.838206 | 1.005391 | -1.835377 | 2.740660 | 2.297497 | 0.647299 | 1 | 0.838768 | 1.004481 | val0_2 | val1_3 | val0_2 |
| 3 | -4.779331 | -1.087113 | -2.335707 | 4.325008 | 0.183701 | 2.754548 | 1 | -4.788262 | -1.085760 | val0_0 | val1_4 | val0_1 |
| 4 | -0.054659 | -3.051933 | -2.188465 | 0.813331 | 3.929342 | 2.225584 | 1 | -0.033824 | -3.055057 | val0_0 | val1_2 | val0_2 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 9995 | -3.165588 | -3.506656 | -2.577650 | 3.412720 | 1.865103 | 3.026053 | 1 | -3.149584 | -3.511459 | val0_1 | val1_1 | val0_1 |
| 9996 | -2.533155 | -4.917804 | -2.034633 | 0.993088 | 2.458396 | 2.773911 | 1 | -2.553134 | -4.923671 | val0_1 | val1_1 | val0_1 |
| 9997 | -1.443761 | 2.141603 | -2.699636 | 4.187722 | 4.510588 | 3.026089 | 1 | -1.449267 | 2.138270 | val0_1 | val1_3 | val0_2 |
| 9998 | -0.732597 | 0.254555 | -2.290704 | 1.860502 | 4.571144 | 2.770080 | 1 | -0.724269 | 0.286252 | val0_1 | val1_3 | val0_2 |
| 9999 | -1.932632 | -3.802342 | -1.526755 | 1.456337 | 1.263657 | 1.237932 | 1 | -1.908979 | -3.789566 | val0_1 | val1_1 | val0_1 |
10000 rows × 12 columns
CatBoost is a tree boosting method capable of handling categorical features (hence the name CatBoost). The CatBoost library offers a select_features functionality that selects the best k features based on the importance of each feature. There are three ways to compute this feature importance:
RecursiveByPredictionValuesChange: according to CatBoost’s own documentation: “the fastest algorithm and the least accurate method (not recommended for ranking losses)” - “For each feature, PredictionValuesChange shows how much on average the prediction changes if the feature value changes. The bigger the value of the importance the bigger on average is the change to the prediction value, if this feature is changed.”;
RecursiveByLossFunctionChange: according to CatBoost’s own documentation: “the optimal option according to accuracy/speed balance” - “For each feature the value represents the difference between the loss value of the model with this feature and without it. The model without this feature is equivalent to the one that would have been trained if this feature was excluded from the dataset. Since it is computationally expensive to retrain the model without one of the features, this model is built approximately using the original model with this feature removed from all the trees in the ensemble. The calculation of this feature importance requires a dataset and, therefore, the calculated value is dataset-dependent.”;
RecursiveByShapValues: according to CatBoost’s own documentation: “the most accurate method.”. For this algorithm, CatBoost uses Shap Values to determine the importance of each feature;
The class CatBoostSelection simply encapsulates all the complexities associated to this functionality and makes it easier for the user. Here is an example of how to use our class to perform feature selection using the CatBoost method:
[3]:
feat_sel = CatBoostSelection()
feat_sel.fit(df=df, label_col=label_col)
feat_sel.get_summary()
No columns specified for imputation. These columns have been automatically identified:
[]
Step #1 out of 1
0: learn: 0.6262765 test: 0.6259290 best: 0.6259290 (0) total: 75.9ms remaining: 1m 15s
1: learn: 0.5638859 test: 0.5636973 best: 0.5636973 (1) total: 117ms remaining: 58.2s
2: learn: 0.5091594 test: 0.5093635 best: 0.5093635 (2) total: 156ms remaining: 52s
3: learn: 0.4597668 test: 0.4599395 best: 0.4599395 (3) total: 207ms remaining: 51.5s
4: learn: 0.4154522 test: 0.4159994 best: 0.4159994 (4) total: 236ms remaining: 46.9s
5: learn: 0.3767378 test: 0.3772350 best: 0.3772350 (5) total: 255ms remaining: 42.3s
6: learn: 0.3466999 test: 0.3468497 best: 0.3468497 (6) total: 271ms remaining: 38.4s
7: learn: 0.3152596 test: 0.3157513 best: 0.3157513 (7) total: 291ms remaining: 36.1s
8: learn: 0.2902757 test: 0.2907094 best: 0.2907094 (8) total: 310ms remaining: 34.1s
9: learn: 0.2652099 test: 0.2659559 best: 0.2659559 (9) total: 329ms remaining: 32.6s
10: learn: 0.2435918 test: 0.2444984 best: 0.2444984 (10) total: 348ms remaining: 31.3s
11: learn: 0.2230890 test: 0.2241702 best: 0.2241702 (11) total: 369ms remaining: 30.4s
12: learn: 0.2061223 test: 0.2072776 best: 0.2072776 (12) total: 384ms remaining: 29.2s
13: learn: 0.1907189 test: 0.1921295 best: 0.1921295 (13) total: 404ms remaining: 28.5s
14: learn: 0.1764498 test: 0.1778341 best: 0.1778341 (14) total: 433ms remaining: 28.5s
15: learn: 0.1634081 test: 0.1648774 best: 0.1648774 (15) total: 465ms remaining: 28.6s
16: learn: 0.1529162 test: 0.1543498 best: 0.1543498 (16) total: 494ms remaining: 28.6s
17: learn: 0.1434670 test: 0.1446691 best: 0.1446691 (17) total: 514ms remaining: 28s
18: learn: 0.1339607 test: 0.1352541 best: 0.1352541 (18) total: 534ms remaining: 27.6s
19: learn: 0.1257332 test: 0.1272102 best: 0.1272102 (19) total: 568ms remaining: 27.9s
20: learn: 0.1182353 test: 0.1198133 best: 0.1198133 (20) total: 601ms remaining: 28s
21: learn: 0.1117719 test: 0.1135357 best: 0.1135357 (21) total: 622ms remaining: 27.6s
22: learn: 0.1061024 test: 0.1079912 best: 0.1079912 (22) total: 646ms remaining: 27.4s
23: learn: 0.1002019 test: 0.1023139 best: 0.1023139 (23) total: 662ms remaining: 26.9s
24: learn: 0.0953247 test: 0.0976063 best: 0.0976063 (24) total: 699ms remaining: 27.3s
25: learn: 0.0907208 test: 0.0931501 best: 0.0931501 (25) total: 723ms remaining: 27.1s
26: learn: 0.0869117 test: 0.0892396 best: 0.0892396 (26) total: 742ms remaining: 26.8s
27: learn: 0.0829544 test: 0.0851959 best: 0.0851959 (27) total: 793ms remaining: 27.5s
28: learn: 0.0796043 test: 0.0819606 best: 0.0819606 (28) total: 840ms remaining: 28.1s
29: learn: 0.0763128 test: 0.0786785 best: 0.0786785 (29) total: 860ms remaining: 27.8s
30: learn: 0.0730505 test: 0.0756061 best: 0.0756061 (30) total: 903ms remaining: 28.2s
31: learn: 0.0705571 test: 0.0731937 best: 0.0731937 (31) total: 929ms remaining: 28.1s
32: learn: 0.0680423 test: 0.0707766 best: 0.0707766 (32) total: 955ms remaining: 28s
33: learn: 0.0658324 test: 0.0687334 best: 0.0687334 (33) total: 984ms remaining: 28s
34: learn: 0.0639789 test: 0.0669711 best: 0.0669711 (34) total: 1.01s remaining: 28s
35: learn: 0.0619305 test: 0.0650420 best: 0.0650420 (35) total: 1.04s remaining: 27.8s
36: learn: 0.0604671 test: 0.0636237 best: 0.0636237 (36) total: 1.06s remaining: 27.7s
37: learn: 0.0585985 test: 0.0619196 best: 0.0619196 (37) total: 1.09s remaining: 27.6s
38: learn: 0.0570130 test: 0.0604502 best: 0.0604502 (38) total: 1.12s remaining: 27.6s
39: learn: 0.0558069 test: 0.0593872 best: 0.0593872 (39) total: 1.15s remaining: 27.5s
40: learn: 0.0545196 test: 0.0581564 best: 0.0581564 (40) total: 1.17s remaining: 27.4s
41: learn: 0.0533724 test: 0.0570456 best: 0.0570456 (41) total: 1.19s remaining: 27.2s
42: learn: 0.0521728 test: 0.0559936 best: 0.0559936 (42) total: 1.22s remaining: 27.2s
43: learn: 0.0510634 test: 0.0549365 best: 0.0549365 (43) total: 1.24s remaining: 27s
44: learn: 0.0502664 test: 0.0540737 best: 0.0540737 (44) total: 1.27s remaining: 27s
45: learn: 0.0495233 test: 0.0533117 best: 0.0533117 (45) total: 1.29s remaining: 26.8s
46: learn: 0.0486353 test: 0.0525358 best: 0.0525358 (46) total: 1.32s remaining: 26.8s
47: learn: 0.0478350 test: 0.0517928 best: 0.0517928 (47) total: 1.34s remaining: 26.7s
48: learn: 0.0470791 test: 0.0512021 best: 0.0512021 (48) total: 1.46s remaining: 28.4s
49: learn: 0.0463986 test: 0.0506853 best: 0.0506853 (49) total: 1.53s remaining: 29s
50: learn: 0.0456364 test: 0.0499667 best: 0.0499667 (50) total: 1.6s remaining: 29.8s
51: learn: 0.0450289 test: 0.0493476 best: 0.0493476 (51) total: 1.64s remaining: 29.8s
52: learn: 0.0444011 test: 0.0489387 best: 0.0489387 (52) total: 1.71s remaining: 30.6s
53: learn: 0.0437853 test: 0.0485611 best: 0.0485611 (53) total: 1.77s remaining: 31s
54: learn: 0.0432912 test: 0.0480787 best: 0.0480787 (54) total: 1.82s remaining: 31.4s
55: learn: 0.0428045 test: 0.0476111 best: 0.0476111 (55) total: 1.85s remaining: 31.3s
56: learn: 0.0424227 test: 0.0473464 best: 0.0473464 (56) total: 1.92s remaining: 31.7s
57: learn: 0.0421444 test: 0.0471191 best: 0.0471191 (57) total: 2s remaining: 32.5s
58: learn: 0.0417806 test: 0.0468586 best: 0.0468586 (58) total: 2.05s remaining: 32.6s
59: learn: 0.0413020 test: 0.0464242 best: 0.0464242 (59) total: 2.08s remaining: 32.6s
60: learn: 0.0409189 test: 0.0460178 best: 0.0460178 (60) total: 2.13s remaining: 32.9s
61: learn: 0.0405308 test: 0.0457084 best: 0.0457084 (61) total: 2.2s remaining: 33.3s
62: learn: 0.0402788 test: 0.0455216 best: 0.0455216 (62) total: 2.24s remaining: 33.4s
63: learn: 0.0400807 test: 0.0453329 best: 0.0453329 (63) total: 2.27s remaining: 33.3s
64: learn: 0.0397318 test: 0.0450720 best: 0.0450720 (64) total: 2.32s remaining: 33.4s
65: learn: 0.0395446 test: 0.0449737 best: 0.0449737 (65) total: 2.4s remaining: 33.9s
66: learn: 0.0392332 test: 0.0448538 best: 0.0448538 (66) total: 2.44s remaining: 34s
67: learn: 0.0390014 test: 0.0447700 best: 0.0447700 (67) total: 2.48s remaining: 34s
68: learn: 0.0387706 test: 0.0445673 best: 0.0445673 (68) total: 2.54s remaining: 34.2s
69: learn: 0.0384443 test: 0.0443434 best: 0.0443434 (69) total: 2.6s remaining: 34.5s
70: learn: 0.0382938 test: 0.0442449 best: 0.0442449 (70) total: 2.65s remaining: 34.7s
71: learn: 0.0381452 test: 0.0441173 best: 0.0441173 (71) total: 2.69s remaining: 34.6s
72: learn: 0.0380754 test: 0.0440624 best: 0.0440624 (72) total: 2.71s remaining: 34.4s
73: learn: 0.0379158 test: 0.0439653 best: 0.0439653 (73) total: 2.74s remaining: 34.2s
74: learn: 0.0377931 test: 0.0438711 best: 0.0438711 (74) total: 2.78s remaining: 34.3s
75: learn: 0.0375055 test: 0.0436029 best: 0.0436029 (75) total: 2.82s remaining: 34.3s
76: learn: 0.0372847 test: 0.0435041 best: 0.0435041 (76) total: 2.86s remaining: 34.2s
77: learn: 0.0371941 test: 0.0434831 best: 0.0434831 (77) total: 2.89s remaining: 34.1s
78: learn: 0.0370210 test: 0.0434022 best: 0.0434022 (78) total: 2.92s remaining: 34.1s
79: learn: 0.0368264 test: 0.0433539 best: 0.0433539 (79) total: 2.96s remaining: 34s
80: learn: 0.0367551 test: 0.0433177 best: 0.0433177 (80) total: 2.99s remaining: 33.9s
81: learn: 0.0365249 test: 0.0431655 best: 0.0431655 (81) total: 3.01s remaining: 33.7s
82: learn: 0.0364907 test: 0.0431416 best: 0.0431416 (82) total: 3.04s remaining: 33.5s
83: learn: 0.0363693 test: 0.0430096 best: 0.0430096 (83) total: 3.06s remaining: 33.4s
84: learn: 0.0362952 test: 0.0429866 best: 0.0429866 (84) total: 3.09s remaining: 33.3s
85: learn: 0.0361773 test: 0.0429510 best: 0.0429510 (85) total: 3.12s remaining: 33.2s
86: learn: 0.0361264 test: 0.0429329 best: 0.0429329 (86) total: 3.14s remaining: 33s
87: learn: 0.0360874 test: 0.0429063 best: 0.0429063 (87) total: 3.16s remaining: 32.8s
88: learn: 0.0359659 test: 0.0428113 best: 0.0428113 (88) total: 3.18s remaining: 32.6s
89: learn: 0.0359200 test: 0.0427805 best: 0.0427805 (89) total: 3.2s remaining: 32.4s
90: learn: 0.0357636 test: 0.0426830 best: 0.0426830 (90) total: 3.22s remaining: 32.2s
91: learn: 0.0357229 test: 0.0426723 best: 0.0426723 (91) total: 3.24s remaining: 32s
92: learn: 0.0355230 test: 0.0425768 best: 0.0425768 (92) total: 3.28s remaining: 32s
93: learn: 0.0354443 test: 0.0425836 best: 0.0425768 (92) total: 3.31s remaining: 31.9s
94: learn: 0.0353648 test: 0.0425870 best: 0.0425768 (92) total: 3.32s remaining: 31.7s
95: learn: 0.0353449 test: 0.0425746 best: 0.0425746 (95) total: 3.34s remaining: 31.5s
96: learn: 0.0352742 test: 0.0425514 best: 0.0425514 (96) total: 3.36s remaining: 31.3s
97: learn: 0.0351718 test: 0.0424652 best: 0.0424652 (97) total: 3.4s remaining: 31.3s
98: learn: 0.0351450 test: 0.0424399 best: 0.0424399 (98) total: 3.46s remaining: 31.5s
99: learn: 0.0350642 test: 0.0424022 best: 0.0424022 (99) total: 3.5s remaining: 31.5s
100: learn: 0.0349449 test: 0.0423017 best: 0.0423017 (100) total: 3.53s remaining: 31.5s
101: learn: 0.0348812 test: 0.0422872 best: 0.0422872 (101) total: 3.57s remaining: 31.4s
102: learn: 0.0348313 test: 0.0422798 best: 0.0422798 (102) total: 3.61s remaining: 31.4s
103: learn: 0.0347022 test: 0.0422509 best: 0.0422509 (103) total: 3.65s remaining: 31.5s
104: learn: 0.0346160 test: 0.0421936 best: 0.0421936 (104) total: 3.69s remaining: 31.5s
105: learn: 0.0345950 test: 0.0421951 best: 0.0421936 (104) total: 3.72s remaining: 31.4s
106: learn: 0.0345669 test: 0.0422117 best: 0.0421936 (104) total: 3.76s remaining: 31.4s
107: learn: 0.0345097 test: 0.0422036 best: 0.0421936 (104) total: 3.8s remaining: 31.4s
108: learn: 0.0344821 test: 0.0421948 best: 0.0421936 (104) total: 3.85s remaining: 31.5s
109: learn: 0.0343047 test: 0.0421612 best: 0.0421612 (109) total: 3.88s remaining: 31.4s
110: learn: 0.0341973 test: 0.0421486 best: 0.0421486 (110) total: 3.91s remaining: 31.3s
111: learn: 0.0341850 test: 0.0421223 best: 0.0421223 (111) total: 3.94s remaining: 31.2s
112: learn: 0.0341222 test: 0.0421114 best: 0.0421114 (112) total: 4.16s remaining: 32.7s
113: learn: 0.0340218 test: 0.0420765 best: 0.0420765 (113) total: 4.19s remaining: 32.6s
114: learn: 0.0339637 test: 0.0420909 best: 0.0420765 (113) total: 4.22s remaining: 32.5s
115: learn: 0.0338358 test: 0.0420489 best: 0.0420489 (115) total: 4.25s remaining: 32.4s
116: learn: 0.0336609 test: 0.0419139 best: 0.0419139 (116) total: 4.3s remaining: 32.5s
117: learn: 0.0336609 test: 0.0419139 best: 0.0419139 (117) total: 4.32s remaining: 32.3s
118: learn: 0.0336122 test: 0.0418924 best: 0.0418924 (118) total: 4.37s remaining: 32.4s
119: learn: 0.0335176 test: 0.0418638 best: 0.0418638 (119) total: 4.44s remaining: 32.6s
120: learn: 0.0335176 test: 0.0418638 best: 0.0418638 (120) total: 4.46s remaining: 32.4s
121: learn: 0.0334385 test: 0.0417971 best: 0.0417971 (121) total: 4.54s remaining: 32.7s
122: learn: 0.0334042 test: 0.0417904 best: 0.0417904 (122) total: 4.57s remaining: 32.6s
123: learn: 0.0333446 test: 0.0418348 best: 0.0417904 (122) total: 4.61s remaining: 32.6s
124: learn: 0.0332989 test: 0.0418367 best: 0.0417904 (122) total: 4.68s remaining: 32.8s
125: learn: 0.0332730 test: 0.0418669 best: 0.0417904 (122) total: 4.71s remaining: 32.7s
126: learn: 0.0332194 test: 0.0418578 best: 0.0417904 (122) total: 4.77s remaining: 32.8s
127: learn: 0.0332031 test: 0.0418616 best: 0.0417904 (122) total: 4.8s remaining: 32.7s
128: learn: 0.0331921 test: 0.0418602 best: 0.0417904 (122) total: 4.83s remaining: 32.6s
129: learn: 0.0331191 test: 0.0418412 best: 0.0417904 (122) total: 4.85s remaining: 32.5s
130: learn: 0.0330914 test: 0.0418405 best: 0.0417904 (122) total: 4.9s remaining: 32.5s
131: learn: 0.0329783 test: 0.0417880 best: 0.0417880 (131) total: 4.95s remaining: 32.5s
132: learn: 0.0329162 test: 0.0417784 best: 0.0417784 (132) total: 4.96s remaining: 32.4s
133: learn: 0.0329042 test: 0.0417673 best: 0.0417673 (133) total: 5.03s remaining: 32.5s
134: learn: 0.0328855 test: 0.0417602 best: 0.0417602 (134) total: 5.08s remaining: 32.5s
135: learn: 0.0328722 test: 0.0417611 best: 0.0417602 (134) total: 5.12s remaining: 32.5s
136: learn: 0.0328114 test: 0.0417697 best: 0.0417602 (134) total: 5.18s remaining: 32.6s
137: learn: 0.0327836 test: 0.0417598 best: 0.0417598 (137) total: 5.26s remaining: 32.9s
138: learn: 0.0327313 test: 0.0417098 best: 0.0417098 (138) total: 5.29s remaining: 32.8s
139: learn: 0.0326356 test: 0.0416656 best: 0.0416656 (139) total: 5.34s remaining: 32.8s
140: learn: 0.0326031 test: 0.0416524 best: 0.0416524 (140) total: 5.37s remaining: 32.7s
141: learn: 0.0325819 test: 0.0416768 best: 0.0416524 (140) total: 5.4s remaining: 32.6s
142: learn: 0.0325735 test: 0.0416766 best: 0.0416524 (140) total: 5.42s remaining: 32.5s
143: learn: 0.0325617 test: 0.0416731 best: 0.0416524 (140) total: 5.43s remaining: 32.3s
144: learn: 0.0325617 test: 0.0416731 best: 0.0416524 (140) total: 5.53s remaining: 32.6s
145: learn: 0.0325457 test: 0.0416727 best: 0.0416524 (140) total: 5.56s remaining: 32.5s
146: learn: 0.0324796 test: 0.0416390 best: 0.0416390 (146) total: 5.67s remaining: 32.9s
147: learn: 0.0323724 test: 0.0415533 best: 0.0415533 (147) total: 5.71s remaining: 32.9s
148: learn: 0.0323126 test: 0.0415343 best: 0.0415343 (148) total: 5.76s remaining: 32.9s
149: learn: 0.0322991 test: 0.0415304 best: 0.0415304 (149) total: 5.8s remaining: 32.9s
150: learn: 0.0322720 test: 0.0415305 best: 0.0415304 (149) total: 5.84s remaining: 32.8s
151: learn: 0.0321710 test: 0.0415626 best: 0.0415304 (149) total: 5.86s remaining: 32.7s
152: learn: 0.0321489 test: 0.0415607 best: 0.0415304 (149) total: 5.9s remaining: 32.7s
153: learn: 0.0320937 test: 0.0415345 best: 0.0415304 (149) total: 5.93s remaining: 32.6s
154: learn: 0.0320721 test: 0.0415235 best: 0.0415235 (154) total: 5.97s remaining: 32.5s
155: learn: 0.0320381 test: 0.0415211 best: 0.0415211 (155) total: 6.02s remaining: 32.6s
156: learn: 0.0320282 test: 0.0415170 best: 0.0415170 (156) total: 6.07s remaining: 32.6s
157: learn: 0.0319831 test: 0.0414981 best: 0.0414981 (157) total: 6.1s remaining: 32.5s
158: learn: 0.0319179 test: 0.0414695 best: 0.0414695 (158) total: 6.13s remaining: 32.4s
159: learn: 0.0318965 test: 0.0414904 best: 0.0414695 (158) total: 6.17s remaining: 32.4s
160: learn: 0.0318744 test: 0.0414791 best: 0.0414695 (158) total: 6.21s remaining: 32.4s
161: learn: 0.0318744 test: 0.0414791 best: 0.0414695 (158) total: 6.22s remaining: 32.2s
162: learn: 0.0318611 test: 0.0414801 best: 0.0414695 (158) total: 6.25s remaining: 32.1s
163: learn: 0.0318547 test: 0.0414829 best: 0.0414695 (158) total: 6.26s remaining: 31.9s
164: learn: 0.0318010 test: 0.0414461 best: 0.0414461 (164) total: 6.31s remaining: 32s
165: learn: 0.0317640 test: 0.0414634 best: 0.0414461 (164) total: 6.37s remaining: 32s
166: learn: 0.0316557 test: 0.0414250 best: 0.0414250 (166) total: 6.42s remaining: 32s
167: learn: 0.0316557 test: 0.0414250 best: 0.0414250 (166) total: 6.44s remaining: 31.9s
168: learn: 0.0316439 test: 0.0414280 best: 0.0414250 (166) total: 6.5s remaining: 32s
169: learn: 0.0316284 test: 0.0414038 best: 0.0414038 (169) total: 6.53s remaining: 31.9s
170: learn: 0.0315929 test: 0.0413974 best: 0.0413974 (170) total: 6.58s remaining: 31.9s
171: learn: 0.0315094 test: 0.0414119 best: 0.0413974 (170) total: 6.63s remaining: 31.9s
172: learn: 0.0314990 test: 0.0414198 best: 0.0413974 (170) total: 6.69s remaining: 32s
173: learn: 0.0314300 test: 0.0413408 best: 0.0413408 (173) total: 6.73s remaining: 32s
174: learn: 0.0313656 test: 0.0413189 best: 0.0413189 (174) total: 6.82s remaining: 32.1s
175: learn: 0.0313453 test: 0.0413180 best: 0.0413180 (175) total: 6.85s remaining: 32.1s
176: learn: 0.0313340 test: 0.0413217 best: 0.0413180 (175) total: 6.88s remaining: 32s
177: learn: 0.0313028 test: 0.0413462 best: 0.0413180 (175) total: 6.91s remaining: 31.9s
178: learn: 0.0312781 test: 0.0413245 best: 0.0413180 (175) total: 6.98s remaining: 32s
179: learn: 0.0312535 test: 0.0413002 best: 0.0413002 (179) total: 7.03s remaining: 32s
180: learn: 0.0312017 test: 0.0412764 best: 0.0412764 (180) total: 7.06s remaining: 31.9s
181: learn: 0.0311958 test: 0.0412711 best: 0.0412711 (181) total: 7.1s remaining: 31.9s
182: learn: 0.0311593 test: 0.0412698 best: 0.0412698 (182) total: 7.12s remaining: 31.8s
183: learn: 0.0311176 test: 0.0412663 best: 0.0412663 (183) total: 7.19s remaining: 31.9s
184: learn: 0.0311095 test: 0.0412685 best: 0.0412663 (183) total: 7.21s remaining: 31.8s
185: learn: 0.0310862 test: 0.0412775 best: 0.0412663 (183) total: 7.25s remaining: 31.7s
186: learn: 0.0310450 test: 0.0412891 best: 0.0412663 (183) total: 7.29s remaining: 31.7s
187: learn: 0.0310393 test: 0.0412685 best: 0.0412663 (183) total: 7.32s remaining: 31.6s
188: learn: 0.0310308 test: 0.0412530 best: 0.0412530 (188) total: 7.37s remaining: 31.6s
189: learn: 0.0309388 test: 0.0412409 best: 0.0412409 (189) total: 7.45s remaining: 31.8s
190: learn: 0.0309332 test: 0.0412403 best: 0.0412403 (190) total: 7.52s remaining: 31.9s
191: learn: 0.0308994 test: 0.0412364 best: 0.0412364 (191) total: 7.56s remaining: 31.8s
192: learn: 0.0308994 test: 0.0412364 best: 0.0412364 (191) total: 7.58s remaining: 31.7s
193: learn: 0.0308759 test: 0.0412356 best: 0.0412356 (193) total: 7.65s remaining: 31.8s
194: learn: 0.0308613 test: 0.0412365 best: 0.0412356 (193) total: 7.69s remaining: 31.7s
195: learn: 0.0308474 test: 0.0412527 best: 0.0412356 (193) total: 7.74s remaining: 31.7s
196: learn: 0.0308312 test: 0.0412509 best: 0.0412356 (193) total: 7.77s remaining: 31.7s
197: learn: 0.0308136 test: 0.0412514 best: 0.0412356 (193) total: 7.82s remaining: 31.7s
198: learn: 0.0308004 test: 0.0412534 best: 0.0412356 (193) total: 7.89s remaining: 31.8s
199: learn: 0.0307877 test: 0.0412550 best: 0.0412356 (193) total: 7.93s remaining: 31.7s
200: learn: 0.0307734 test: 0.0412397 best: 0.0412356 (193) total: 7.98s remaining: 31.7s
201: learn: 0.0307520 test: 0.0412065 best: 0.0412065 (201) total: 8.02s remaining: 31.7s
202: learn: 0.0307346 test: 0.0412026 best: 0.0412026 (202) total: 8.06s remaining: 31.6s
203: learn: 0.0307216 test: 0.0411946 best: 0.0411946 (203) total: 8.1s remaining: 31.6s
204: learn: 0.0307068 test: 0.0411843 best: 0.0411843 (204) total: 8.14s remaining: 31.6s
205: learn: 0.0307002 test: 0.0411674 best: 0.0411674 (205) total: 8.24s remaining: 31.8s
206: learn: 0.0306986 test: 0.0411700 best: 0.0411674 (205) total: 8.27s remaining: 31.7s
207: learn: 0.0306305 test: 0.0411786 best: 0.0411674 (205) total: 8.37s remaining: 31.9s
208: learn: 0.0305743 test: 0.0411513 best: 0.0411513 (208) total: 8.4s remaining: 31.8s
209: learn: 0.0305423 test: 0.0411447 best: 0.0411447 (209) total: 8.47s remaining: 31.9s
210: learn: 0.0305297 test: 0.0411481 best: 0.0411447 (209) total: 8.5s remaining: 31.8s
211: learn: 0.0305198 test: 0.0411497 best: 0.0411447 (209) total: 8.57s remaining: 31.9s
212: learn: 0.0304894 test: 0.0411659 best: 0.0411447 (209) total: 8.62s remaining: 31.8s
213: learn: 0.0304726 test: 0.0411662 best: 0.0411447 (209) total: 8.67s remaining: 31.9s
214: learn: 0.0304563 test: 0.0411585 best: 0.0411447 (209) total: 8.72s remaining: 31.8s
215: learn: 0.0304274 test: 0.0411349 best: 0.0411349 (215) total: 8.75s remaining: 31.8s
216: learn: 0.0304274 test: 0.0411349 best: 0.0411349 (215) total: 8.78s remaining: 31.7s
217: learn: 0.0303995 test: 0.0411292 best: 0.0411292 (217) total: 8.83s remaining: 31.7s
218: learn: 0.0303482 test: 0.0410850 best: 0.0410850 (218) total: 8.86s remaining: 31.6s
219: learn: 0.0302319 test: 0.0411003 best: 0.0410850 (218) total: 8.92s remaining: 31.6s
220: learn: 0.0301332 test: 0.0411206 best: 0.0410850 (218) total: 8.97s remaining: 31.6s
221: learn: 0.0301219 test: 0.0411187 best: 0.0410850 (218) total: 9s remaining: 31.6s
222: learn: 0.0300818 test: 0.0411138 best: 0.0410850 (218) total: 9.03s remaining: 31.5s
223: learn: 0.0300818 test: 0.0411139 best: 0.0410850 (218) total: 9.04s remaining: 31.3s
224: learn: 0.0299988 test: 0.0411165 best: 0.0410850 (218) total: 9.07s remaining: 31.2s
225: learn: 0.0299889 test: 0.0411180 best: 0.0410850 (218) total: 9.1s remaining: 31.2s
226: learn: 0.0299472 test: 0.0411226 best: 0.0410850 (218) total: 9.14s remaining: 31.1s
227: learn: 0.0299279 test: 0.0411170 best: 0.0410850 (218) total: 9.2s remaining: 31.1s
228: learn: 0.0298818 test: 0.0410931 best: 0.0410850 (218) total: 9.25s remaining: 31.2s
229: learn: 0.0298587 test: 0.0410900 best: 0.0410850 (218) total: 9.35s remaining: 31.3s
230: learn: 0.0298313 test: 0.0410878 best: 0.0410850 (218) total: 9.44s remaining: 31.4s
231: learn: 0.0298000 test: 0.0410987 best: 0.0410850 (218) total: 9.49s remaining: 31.4s
232: learn: 0.0298000 test: 0.0410987 best: 0.0410850 (218) total: 9.5s remaining: 31.3s
233: learn: 0.0297900 test: 0.0411003 best: 0.0410850 (218) total: 9.55s remaining: 31.3s
234: learn: 0.0297623 test: 0.0410952 best: 0.0410850 (218) total: 9.62s remaining: 31.3s
235: learn: 0.0296987 test: 0.0410746 best: 0.0410746 (235) total: 9.68s remaining: 31.3s
236: learn: 0.0296657 test: 0.0410729 best: 0.0410729 (236) total: 9.77s remaining: 31.4s
237: learn: 0.0296147 test: 0.0410576 best: 0.0410576 (237) total: 9.81s remaining: 31.4s
238: learn: 0.0295930 test: 0.0410643 best: 0.0410576 (237) total: 9.83s remaining: 31.3s
239: learn: 0.0295466 test: 0.0410450 best: 0.0410450 (239) total: 9.86s remaining: 31.2s
240: learn: 0.0295257 test: 0.0410678 best: 0.0410450 (239) total: 9.89s remaining: 31.2s
241: learn: 0.0295257 test: 0.0410678 best: 0.0410450 (239) total: 9.9s remaining: 31s
242: learn: 0.0295075 test: 0.0410646 best: 0.0410450 (239) total: 9.95s remaining: 31s
243: learn: 0.0294378 test: 0.0410452 best: 0.0410450 (239) total: 10s remaining: 31.1s
244: learn: 0.0293946 test: 0.0410432 best: 0.0410432 (244) total: 10.1s remaining: 31.1s
245: learn: 0.0293847 test: 0.0410438 best: 0.0410432 (244) total: 10.2s remaining: 31.2s
246: learn: 0.0292624 test: 0.0410627 best: 0.0410432 (244) total: 10.2s remaining: 31.2s
247: learn: 0.0292507 test: 0.0410663 best: 0.0410432 (244) total: 10.3s remaining: 31.2s
248: learn: 0.0292394 test: 0.0410585 best: 0.0410432 (244) total: 10.4s remaining: 31.2s
249: learn: 0.0292242 test: 0.0410504 best: 0.0410432 (244) total: 10.4s remaining: 31.2s
250: learn: 0.0292242 test: 0.0410504 best: 0.0410432 (244) total: 10.4s remaining: 31.2s
251: learn: 0.0292199 test: 0.0410509 best: 0.0410432 (244) total: 10.5s remaining: 31s
252: learn: 0.0291927 test: 0.0410216 best: 0.0410216 (252) total: 10.5s remaining: 31s
253: learn: 0.0291548 test: 0.0410341 best: 0.0410216 (252) total: 10.6s remaining: 31s
254: learn: 0.0291170 test: 0.0410257 best: 0.0410216 (252) total: 10.6s remaining: 30.9s
255: learn: 0.0291056 test: 0.0410250 best: 0.0410216 (252) total: 10.6s remaining: 30.9s
256: learn: 0.0290949 test: 0.0410305 best: 0.0410216 (252) total: 10.7s remaining: 30.9s
257: learn: 0.0290568 test: 0.0410119 best: 0.0410119 (257) total: 10.7s remaining: 30.9s
258: learn: 0.0290483 test: 0.0410138 best: 0.0410119 (257) total: 10.8s remaining: 30.8s
259: learn: 0.0290412 test: 0.0410162 best: 0.0410119 (257) total: 10.8s remaining: 30.8s
260: learn: 0.0290411 test: 0.0410163 best: 0.0410119 (257) total: 10.8s remaining: 30.7s
261: learn: 0.0289844 test: 0.0410564 best: 0.0410119 (257) total: 10.9s remaining: 30.6s
262: learn: 0.0289844 test: 0.0410564 best: 0.0410119 (257) total: 10.9s remaining: 30.5s
263: learn: 0.0289844 test: 0.0410564 best: 0.0410119 (257) total: 10.9s remaining: 30.4s
264: learn: 0.0289723 test: 0.0410504 best: 0.0410119 (257) total: 11s remaining: 30.4s
265: learn: 0.0289594 test: 0.0410532 best: 0.0410119 (257) total: 11s remaining: 30.3s
266: learn: 0.0289506 test: 0.0410579 best: 0.0410119 (257) total: 11s remaining: 30.3s
267: learn: 0.0289288 test: 0.0411030 best: 0.0410119 (257) total: 11.1s remaining: 30.3s
268: learn: 0.0289147 test: 0.0410941 best: 0.0410119 (257) total: 11.1s remaining: 30.3s
269: learn: 0.0288757 test: 0.0411001 best: 0.0410119 (257) total: 11.2s remaining: 30.2s
270: learn: 0.0288216 test: 0.0411180 best: 0.0410119 (257) total: 11.2s remaining: 30.2s
271: learn: 0.0287975 test: 0.0411176 best: 0.0410119 (257) total: 11.3s remaining: 30.1s
272: learn: 0.0287767 test: 0.0411215 best: 0.0410119 (257) total: 11.3s remaining: 30.1s
273: learn: 0.0287128 test: 0.0410880 best: 0.0410119 (257) total: 11.3s remaining: 30s
274: learn: 0.0287128 test: 0.0410880 best: 0.0410119 (257) total: 11.3s remaining: 29.9s
275: learn: 0.0286911 test: 0.0410850 best: 0.0410119 (257) total: 11.4s remaining: 29.9s
276: learn: 0.0286816 test: 0.0410912 best: 0.0410119 (257) total: 11.4s remaining: 29.9s
277: learn: 0.0286637 test: 0.0410918 best: 0.0410119 (257) total: 11.5s remaining: 29.8s
278: learn: 0.0286458 test: 0.0410729 best: 0.0410119 (257) total: 11.5s remaining: 29.8s
279: learn: 0.0286440 test: 0.0410763 best: 0.0410119 (257) total: 11.6s remaining: 29.8s
280: learn: 0.0286309 test: 0.0410807 best: 0.0410119 (257) total: 11.7s remaining: 30s
281: learn: 0.0286309 test: 0.0410807 best: 0.0410119 (257) total: 11.7s remaining: 29.9s
282: learn: 0.0285967 test: 0.0410897 best: 0.0410119 (257) total: 11.8s remaining: 29.9s
283: learn: 0.0285687 test: 0.0410797 best: 0.0410119 (257) total: 11.8s remaining: 29.9s
284: learn: 0.0285419 test: 0.0410982 best: 0.0410119 (257) total: 11.9s remaining: 29.8s
285: learn: 0.0285150 test: 0.0410916 best: 0.0410119 (257) total: 11.9s remaining: 29.7s
286: learn: 0.0284837 test: 0.0411057 best: 0.0410119 (257) total: 11.9s remaining: 29.6s
287: learn: 0.0284062 test: 0.0410991 best: 0.0410119 (257) total: 12s remaining: 29.6s
288: learn: 0.0283965 test: 0.0411045 best: 0.0410119 (257) total: 12s remaining: 29.5s
289: learn: 0.0283129 test: 0.0410974 best: 0.0410119 (257) total: 12s remaining: 29.4s
290: learn: 0.0282831 test: 0.0410895 best: 0.0410119 (257) total: 12.1s remaining: 29.4s
291: learn: 0.0282743 test: 0.0410825 best: 0.0410119 (257) total: 12.1s remaining: 29.3s
292: learn: 0.0282549 test: 0.0410827 best: 0.0410119 (257) total: 12.1s remaining: 29.2s
293: learn: 0.0282226 test: 0.0410781 best: 0.0410119 (257) total: 12.2s remaining: 29.3s
294: learn: 0.0281561 test: 0.0410783 best: 0.0410119 (257) total: 12.2s remaining: 29.3s
295: learn: 0.0281453 test: 0.0410883 best: 0.0410119 (257) total: 12.3s remaining: 29.2s
296: learn: 0.0281321 test: 0.0410798 best: 0.0410119 (257) total: 12.3s remaining: 29.1s
297: learn: 0.0281321 test: 0.0410798 best: 0.0410119 (257) total: 12.3s remaining: 29s
298: learn: 0.0281252 test: 0.0410858 best: 0.0410119 (257) total: 12.4s remaining: 29s
299: learn: 0.0281010 test: 0.0410579 best: 0.0410119 (257) total: 12.4s remaining: 29s
300: learn: 0.0280891 test: 0.0410470 best: 0.0410119 (257) total: 12.5s remaining: 29s
301: learn: 0.0280774 test: 0.0410452 best: 0.0410119 (257) total: 12.5s remaining: 28.9s
302: learn: 0.0280774 test: 0.0410453 best: 0.0410119 (257) total: 12.5s remaining: 28.8s
303: learn: 0.0279906 test: 0.0410205 best: 0.0410119 (257) total: 12.6s remaining: 28.8s
304: learn: 0.0279845 test: 0.0410262 best: 0.0410119 (257) total: 12.6s remaining: 28.7s
305: learn: 0.0279661 test: 0.0410225 best: 0.0410119 (257) total: 12.6s remaining: 28.6s
306: learn: 0.0279163 test: 0.0409944 best: 0.0409944 (306) total: 12.6s remaining: 28.6s
307: learn: 0.0278373 test: 0.0409741 best: 0.0409741 (307) total: 12.7s remaining: 28.5s
308: learn: 0.0278103 test: 0.0409912 best: 0.0409741 (307) total: 12.7s remaining: 28.5s
309: learn: 0.0277229 test: 0.0410123 best: 0.0409741 (307) total: 12.8s remaining: 28.5s
310: learn: 0.0277109 test: 0.0410219 best: 0.0409741 (307) total: 12.9s remaining: 28.5s
311: learn: 0.0276891 test: 0.0410388 best: 0.0409741 (307) total: 12.9s remaining: 28.5s
312: learn: 0.0276754 test: 0.0410334 best: 0.0409741 (307) total: 12.9s remaining: 28.4s
313: learn: 0.0276470 test: 0.0410353 best: 0.0409741 (307) total: 13s remaining: 28.4s
314: learn: 0.0275930 test: 0.0410160 best: 0.0409741 (307) total: 13s remaining: 28.3s
315: learn: 0.0275930 test: 0.0410160 best: 0.0409741 (307) total: 13s remaining: 28.2s
316: learn: 0.0275602 test: 0.0409754 best: 0.0409741 (307) total: 13.1s remaining: 28.2s
317: learn: 0.0275270 test: 0.0409736 best: 0.0409736 (317) total: 13.1s remaining: 28.1s
318: learn: 0.0274983 test: 0.0409666 best: 0.0409666 (318) total: 13.2s remaining: 28.1s
319: learn: 0.0274630 test: 0.0409775 best: 0.0409666 (318) total: 13.2s remaining: 28s
320: learn: 0.0274357 test: 0.0409644 best: 0.0409644 (320) total: 13.2s remaining: 28s
321: learn: 0.0274061 test: 0.0409665 best: 0.0409644 (320) total: 13.3s remaining: 27.9s
322: learn: 0.0273915 test: 0.0409648 best: 0.0409644 (320) total: 13.3s remaining: 27.9s
323: learn: 0.0273592 test: 0.0409627 best: 0.0409627 (323) total: 13.3s remaining: 27.8s
324: learn: 0.0273508 test: 0.0409652 best: 0.0409627 (323) total: 13.4s remaining: 27.8s
325: learn: 0.0273378 test: 0.0409718 best: 0.0409627 (323) total: 13.4s remaining: 27.8s
326: learn: 0.0273132 test: 0.0409460 best: 0.0409460 (326) total: 13.5s remaining: 27.7s
327: learn: 0.0273045 test: 0.0409409 best: 0.0409409 (327) total: 13.5s remaining: 27.6s
328: learn: 0.0272780 test: 0.0409170 best: 0.0409170 (328) total: 13.5s remaining: 27.6s
329: learn: 0.0272535 test: 0.0409181 best: 0.0409170 (328) total: 13.6s remaining: 27.5s
330: learn: 0.0272398 test: 0.0409140 best: 0.0409140 (330) total: 13.6s remaining: 27.5s
331: learn: 0.0272300 test: 0.0409074 best: 0.0409074 (331) total: 13.7s remaining: 27.5s
332: learn: 0.0272063 test: 0.0409082 best: 0.0409074 (331) total: 13.7s remaining: 27.5s
333: learn: 0.0271809 test: 0.0409109 best: 0.0409074 (331) total: 13.8s remaining: 27.4s
334: learn: 0.0271681 test: 0.0409031 best: 0.0409031 (334) total: 13.8s remaining: 27.4s
335: learn: 0.0271277 test: 0.0409115 best: 0.0409031 (334) total: 13.8s remaining: 27.3s
336: learn: 0.0271205 test: 0.0409118 best: 0.0409031 (334) total: 13.9s remaining: 27.3s
337: learn: 0.0270963 test: 0.0409227 best: 0.0409031 (334) total: 13.9s remaining: 27.2s
338: learn: 0.0270885 test: 0.0409217 best: 0.0409031 (334) total: 13.9s remaining: 27.2s
339: learn: 0.0270758 test: 0.0409227 best: 0.0409031 (334) total: 14s remaining: 27.1s
340: learn: 0.0270013 test: 0.0409207 best: 0.0409031 (334) total: 14s remaining: 27.1s
341: learn: 0.0269885 test: 0.0409190 best: 0.0409031 (334) total: 14.1s remaining: 27.1s
342: learn: 0.0269744 test: 0.0409139 best: 0.0409031 (334) total: 14.1s remaining: 27s
343: learn: 0.0269511 test: 0.0409170 best: 0.0409031 (334) total: 14.2s remaining: 27s
344: learn: 0.0269229 test: 0.0409095 best: 0.0409031 (334) total: 14.2s remaining: 27s
345: learn: 0.0269039 test: 0.0409018 best: 0.0409018 (345) total: 14.3s remaining: 27s
346: learn: 0.0268768 test: 0.0408935 best: 0.0408935 (346) total: 14.3s remaining: 27s
347: learn: 0.0268697 test: 0.0408948 best: 0.0408935 (346) total: 14.4s remaining: 26.9s
348: learn: 0.0268575 test: 0.0408867 best: 0.0408867 (348) total: 14.4s remaining: 26.9s
349: learn: 0.0268506 test: 0.0408857 best: 0.0408857 (349) total: 14.5s remaining: 26.9s
350: learn: 0.0268325 test: 0.0408874 best: 0.0408857 (349) total: 14.5s remaining: 26.9s
351: learn: 0.0268102 test: 0.0408759 best: 0.0408759 (351) total: 14.6s remaining: 26.9s
352: learn: 0.0268008 test: 0.0408839 best: 0.0408759 (351) total: 14.6s remaining: 26.8s
353: learn: 0.0267919 test: 0.0408864 best: 0.0408759 (351) total: 14.7s remaining: 26.8s
354: learn: 0.0267840 test: 0.0408873 best: 0.0408759 (351) total: 14.7s remaining: 26.7s
355: learn: 0.0267661 test: 0.0408733 best: 0.0408733 (355) total: 14.7s remaining: 26.7s
356: learn: 0.0267498 test: 0.0408729 best: 0.0408729 (356) total: 14.8s remaining: 26.6s
357: learn: 0.0267398 test: 0.0408710 best: 0.0408710 (357) total: 14.8s remaining: 26.6s
358: learn: 0.0267186 test: 0.0408749 best: 0.0408710 (357) total: 14.8s remaining: 26.5s
359: learn: 0.0267019 test: 0.0408718 best: 0.0408710 (357) total: 14.9s remaining: 26.4s
360: learn: 0.0267019 test: 0.0408718 best: 0.0408710 (357) total: 14.9s remaining: 26.4s
361: learn: 0.0266949 test: 0.0408691 best: 0.0408691 (361) total: 15s remaining: 26.4s
362: learn: 0.0266574 test: 0.0408854 best: 0.0408691 (361) total: 15s remaining: 26.3s
363: learn: 0.0266219 test: 0.0409013 best: 0.0408691 (361) total: 15s remaining: 26.3s
364: learn: 0.0266111 test: 0.0409004 best: 0.0408691 (361) total: 15.1s remaining: 26.2s
365: learn: 0.0265948 test: 0.0409085 best: 0.0408691 (361) total: 15.1s remaining: 26.2s
366: learn: 0.0265757 test: 0.0408918 best: 0.0408691 (361) total: 15.2s remaining: 26.1s
367: learn: 0.0265612 test: 0.0408867 best: 0.0408691 (361) total: 15.2s remaining: 26.1s
368: learn: 0.0265511 test: 0.0408798 best: 0.0408691 (361) total: 15.3s remaining: 26.1s
369: learn: 0.0265373 test: 0.0408749 best: 0.0408691 (361) total: 15.3s remaining: 26.1s
370: learn: 0.0265077 test: 0.0408675 best: 0.0408675 (370) total: 15.4s remaining: 26.2s
371: learn: 0.0264723 test: 0.0408708 best: 0.0408675 (370) total: 15.5s remaining: 26.2s
372: learn: 0.0264468 test: 0.0408646 best: 0.0408646 (372) total: 15.5s remaining: 26.1s
373: learn: 0.0264342 test: 0.0408646 best: 0.0408646 (372) total: 15.6s remaining: 26.1s
374: learn: 0.0264340 test: 0.0408641 best: 0.0408641 (374) total: 15.6s remaining: 26s
375: learn: 0.0264222 test: 0.0408635 best: 0.0408635 (375) total: 15.6s remaining: 26s
376: learn: 0.0264096 test: 0.0408616 best: 0.0408616 (376) total: 15.7s remaining: 25.9s
377: learn: 0.0263980 test: 0.0408623 best: 0.0408616 (376) total: 15.7s remaining: 25.9s
378: learn: 0.0263937 test: 0.0408604 best: 0.0408604 (378) total: 15.7s remaining: 25.8s
379: learn: 0.0263701 test: 0.0408514 best: 0.0408514 (379) total: 15.8s remaining: 25.7s
380: learn: 0.0263578 test: 0.0408443 best: 0.0408443 (380) total: 15.8s remaining: 25.7s
381: learn: 0.0263503 test: 0.0408432 best: 0.0408432 (381) total: 15.9s remaining: 25.6s
382: learn: 0.0263419 test: 0.0408447 best: 0.0408432 (381) total: 15.9s remaining: 25.6s
383: learn: 0.0263223 test: 0.0408462 best: 0.0408432 (381) total: 15.9s remaining: 25.6s
384: learn: 0.0263008 test: 0.0408424 best: 0.0408424 (384) total: 16s remaining: 25.5s
385: learn: 0.0262912 test: 0.0408377 best: 0.0408377 (385) total: 16s remaining: 25.4s
386: learn: 0.0262771 test: 0.0408324 best: 0.0408324 (386) total: 16s remaining: 25.4s
387: learn: 0.0262705 test: 0.0408339 best: 0.0408324 (386) total: 16s remaining: 25.3s
388: learn: 0.0262510 test: 0.0408160 best: 0.0408160 (388) total: 16.1s remaining: 25.2s
389: learn: 0.0262391 test: 0.0408157 best: 0.0408157 (389) total: 16.1s remaining: 25.2s
390: learn: 0.0262265 test: 0.0408145 best: 0.0408145 (390) total: 16.1s remaining: 25.2s
391: learn: 0.0262133 test: 0.0408072 best: 0.0408072 (391) total: 16.2s remaining: 25.1s
392: learn: 0.0261629 test: 0.0408361 best: 0.0408072 (391) total: 16.2s remaining: 25s
393: learn: 0.0261480 test: 0.0408360 best: 0.0408072 (391) total: 16.2s remaining: 24.9s
394: learn: 0.0261356 test: 0.0408333 best: 0.0408072 (391) total: 16.2s remaining: 24.9s
395: learn: 0.0261322 test: 0.0408372 best: 0.0408072 (391) total: 16.3s remaining: 24.8s
396: learn: 0.0261010 test: 0.0408336 best: 0.0408072 (391) total: 16.3s remaining: 24.8s
397: learn: 0.0260849 test: 0.0408309 best: 0.0408072 (391) total: 16.3s remaining: 24.7s
398: learn: 0.0260711 test: 0.0408339 best: 0.0408072 (391) total: 16.4s remaining: 24.6s
399: learn: 0.0260387 test: 0.0408415 best: 0.0408072 (391) total: 16.4s remaining: 24.6s
400: learn: 0.0260256 test: 0.0408415 best: 0.0408072 (391) total: 16.4s remaining: 24.5s
401: learn: 0.0259923 test: 0.0408709 best: 0.0408072 (391) total: 16.4s remaining: 24.4s
402: learn: 0.0259744 test: 0.0408669 best: 0.0408072 (391) total: 16.5s remaining: 24.4s
403: learn: 0.0259546 test: 0.0408754 best: 0.0408072 (391) total: 16.5s remaining: 24.3s
404: learn: 0.0259119 test: 0.0408946 best: 0.0408072 (391) total: 16.5s remaining: 24.3s
405: learn: 0.0258874 test: 0.0408796 best: 0.0408072 (391) total: 16.5s remaining: 24.2s
406: learn: 0.0258746 test: 0.0408797 best: 0.0408072 (391) total: 16.6s remaining: 24.1s
407: learn: 0.0258608 test: 0.0408803 best: 0.0408072 (391) total: 16.6s remaining: 24.1s
408: learn: 0.0258521 test: 0.0408765 best: 0.0408072 (391) total: 16.6s remaining: 24s
409: learn: 0.0258394 test: 0.0408755 best: 0.0408072 (391) total: 16.6s remaining: 23.9s
410: learn: 0.0258310 test: 0.0408716 best: 0.0408072 (391) total: 16.6s remaining: 23.8s
411: learn: 0.0258229 test: 0.0408683 best: 0.0408072 (391) total: 16.6s remaining: 23.7s
412: learn: 0.0258074 test: 0.0408818 best: 0.0408072 (391) total: 16.6s remaining: 23.6s
413: learn: 0.0257943 test: 0.0408802 best: 0.0408072 (391) total: 16.6s remaining: 23.5s
414: learn: 0.0257819 test: 0.0408879 best: 0.0408072 (391) total: 16.6s remaining: 23.5s
415: learn: 0.0257733 test: 0.0408858 best: 0.0408072 (391) total: 16.7s remaining: 23.4s
416: learn: 0.0257570 test: 0.0408864 best: 0.0408072 (391) total: 16.7s remaining: 23.3s
417: learn: 0.0257488 test: 0.0408833 best: 0.0408072 (391) total: 16.7s remaining: 23.2s
418: learn: 0.0257333 test: 0.0408851 best: 0.0408072 (391) total: 16.7s remaining: 23.1s
419: learn: 0.0257206 test: 0.0408851 best: 0.0408072 (391) total: 16.7s remaining: 23.1s
420: learn: 0.0257055 test: 0.0408984 best: 0.0408072 (391) total: 16.7s remaining: 23s
421: learn: 0.0256785 test: 0.0409013 best: 0.0408072 (391) total: 16.7s remaining: 22.9s
422: learn: 0.0256578 test: 0.0409226 best: 0.0408072 (391) total: 16.7s remaining: 22.8s
423: learn: 0.0256126 test: 0.0409473 best: 0.0408072 (391) total: 16.8s remaining: 22.8s
424: learn: 0.0255995 test: 0.0409475 best: 0.0408072 (391) total: 16.8s remaining: 22.7s
425: learn: 0.0255629 test: 0.0409664 best: 0.0408072 (391) total: 16.8s remaining: 22.6s
426: learn: 0.0255628 test: 0.0409665 best: 0.0408072 (391) total: 16.8s remaining: 22.5s
427: learn: 0.0255500 test: 0.0409667 best: 0.0408072 (391) total: 16.8s remaining: 22.4s
428: learn: 0.0255391 test: 0.0409658 best: 0.0408072 (391) total: 16.8s remaining: 22.3s
429: learn: 0.0255275 test: 0.0409658 best: 0.0408072 (391) total: 16.8s remaining: 22.3s
430: learn: 0.0255275 test: 0.0409659 best: 0.0408072 (391) total: 16.8s remaining: 22.2s
431: learn: 0.0255161 test: 0.0409660 best: 0.0408072 (391) total: 16.8s remaining: 22.1s
432: learn: 0.0255047 test: 0.0409658 best: 0.0408072 (391) total: 16.8s remaining: 22s
433: learn: 0.0254937 test: 0.0409659 best: 0.0408072 (391) total: 16.8s remaining: 21.9s
434: learn: 0.0254898 test: 0.0409655 best: 0.0408072 (391) total: 16.8s remaining: 21.9s
435: learn: 0.0254830 test: 0.0409667 best: 0.0408072 (391) total: 16.8s remaining: 21.8s
436: learn: 0.0254763 test: 0.0409619 best: 0.0408072 (391) total: 16.8s remaining: 21.7s
437: learn: 0.0254622 test: 0.0409694 best: 0.0408072 (391) total: 16.9s remaining: 21.6s
438: learn: 0.0254333 test: 0.0409689 best: 0.0408072 (391) total: 16.9s remaining: 21.6s
439: learn: 0.0254258 test: 0.0409772 best: 0.0408072 (391) total: 16.9s remaining: 21.5s
440: learn: 0.0254027 test: 0.0409893 best: 0.0408072 (391) total: 16.9s remaining: 21.4s
441: learn: 0.0253821 test: 0.0409742 best: 0.0408072 (391) total: 16.9s remaining: 21.3s
442: learn: 0.0253608 test: 0.0409592 best: 0.0408072 (391) total: 16.9s remaining: 21.3s
443: learn: 0.0253482 test: 0.0409563 best: 0.0408072 (391) total: 16.9s remaining: 21.2s
444: learn: 0.0253341 test: 0.0409719 best: 0.0408072 (391) total: 16.9s remaining: 21.1s
445: learn: 0.0253235 test: 0.0409719 best: 0.0408072 (391) total: 16.9s remaining: 21s
446: learn: 0.0253131 test: 0.0409720 best: 0.0408072 (391) total: 17s remaining: 21s
447: learn: 0.0252998 test: 0.0409869 best: 0.0408072 (391) total: 17s remaining: 20.9s
448: learn: 0.0252873 test: 0.0409991 best: 0.0408072 (391) total: 17s remaining: 20.8s
449: learn: 0.0252872 test: 0.0409991 best: 0.0408072 (391) total: 17s remaining: 20.7s
450: learn: 0.0252517 test: 0.0410091 best: 0.0408072 (391) total: 17s remaining: 20.7s
451: learn: 0.0252220 test: 0.0410151 best: 0.0408072 (391) total: 17s remaining: 20.6s
452: learn: 0.0252030 test: 0.0410266 best: 0.0408072 (391) total: 17s remaining: 20.5s
453: learn: 0.0252030 test: 0.0410266 best: 0.0408072 (391) total: 17s remaining: 20.5s
454: learn: 0.0251922 test: 0.0410265 best: 0.0408072 (391) total: 17s remaining: 20.4s
455: learn: 0.0251922 test: 0.0410265 best: 0.0408072 (391) total: 17s remaining: 20.3s
456: learn: 0.0251812 test: 0.0410268 best: 0.0408072 (391) total: 17s remaining: 20.2s
457: learn: 0.0251686 test: 0.0410246 best: 0.0408072 (391) total: 17s remaining: 20.2s
458: learn: 0.0251567 test: 0.0410366 best: 0.0408072 (391) total: 17s remaining: 20.1s
459: learn: 0.0251567 test: 0.0410366 best: 0.0408072 (391) total: 17.1s remaining: 20s
460: learn: 0.0251464 test: 0.0410367 best: 0.0408072 (391) total: 17.1s remaining: 19.9s
461: learn: 0.0251274 test: 0.0410462 best: 0.0408072 (391) total: 17.1s remaining: 19.9s
462: learn: 0.0251172 test: 0.0410465 best: 0.0408072 (391) total: 17.1s remaining: 19.8s
463: learn: 0.0250840 test: 0.0410378 best: 0.0408072 (391) total: 17.1s remaining: 19.7s
464: learn: 0.0250706 test: 0.0410374 best: 0.0408072 (391) total: 17.1s remaining: 19.7s
465: learn: 0.0250478 test: 0.0410272 best: 0.0408072 (391) total: 17.1s remaining: 19.6s
466: learn: 0.0250146 test: 0.0410254 best: 0.0408072 (391) total: 17.1s remaining: 19.5s
467: learn: 0.0250082 test: 0.0410206 best: 0.0408072 (391) total: 17.1s remaining: 19.5s
468: learn: 0.0249970 test: 0.0410237 best: 0.0408072 (391) total: 17.1s remaining: 19.4s
469: learn: 0.0249931 test: 0.0410224 best: 0.0408072 (391) total: 17.2s remaining: 19.4s
470: learn: 0.0249930 test: 0.0410224 best: 0.0408072 (391) total: 17.2s remaining: 19.3s
471: learn: 0.0249930 test: 0.0410224 best: 0.0408072 (391) total: 17.2s remaining: 19.2s
472: learn: 0.0249722 test: 0.0410558 best: 0.0408072 (391) total: 17.2s remaining: 19.2s
473: learn: 0.0249545 test: 0.0410443 best: 0.0408072 (391) total: 17.2s remaining: 19.1s
474: learn: 0.0249443 test: 0.0410445 best: 0.0408072 (391) total: 17.2s remaining: 19s
475: learn: 0.0249053 test: 0.0410273 best: 0.0408072 (391) total: 17.2s remaining: 19s
476: learn: 0.0248703 test: 0.0410146 best: 0.0408072 (391) total: 17.2s remaining: 18.9s
477: learn: 0.0248604 test: 0.0410183 best: 0.0408072 (391) total: 17.3s remaining: 18.8s
478: learn: 0.0248471 test: 0.0410181 best: 0.0408072 (391) total: 17.3s remaining: 18.8s
479: learn: 0.0248396 test: 0.0410152 best: 0.0408072 (391) total: 17.3s remaining: 18.7s
480: learn: 0.0248363 test: 0.0410179 best: 0.0408072 (391) total: 17.3s remaining: 18.6s
481: learn: 0.0248186 test: 0.0410181 best: 0.0408072 (391) total: 17.3s remaining: 18.6s
482: learn: 0.0248116 test: 0.0410154 best: 0.0408072 (391) total: 17.3s remaining: 18.5s
483: learn: 0.0247985 test: 0.0410249 best: 0.0408072 (391) total: 17.3s remaining: 18.4s
484: learn: 0.0247773 test: 0.0410154 best: 0.0408072 (391) total: 17.3s remaining: 18.4s
485: learn: 0.0247689 test: 0.0410157 best: 0.0408072 (391) total: 17.3s remaining: 18.3s
486: learn: 0.0247499 test: 0.0410179 best: 0.0408072 (391) total: 17.3s remaining: 18.3s
487: learn: 0.0247370 test: 0.0410124 best: 0.0408072 (391) total: 17.3s remaining: 18.2s
488: learn: 0.0247133 test: 0.0410128 best: 0.0408072 (391) total: 17.4s remaining: 18.1s
489: learn: 0.0246881 test: 0.0410125 best: 0.0408072 (391) total: 17.4s remaining: 18.1s
490: learn: 0.0246573 test: 0.0410049 best: 0.0408072 (391) total: 17.4s remaining: 18s
491: learn: 0.0246448 test: 0.0410374 best: 0.0408072 (391) total: 17.4s remaining: 17.9s
492: learn: 0.0246365 test: 0.0410356 best: 0.0408072 (391) total: 17.4s remaining: 17.9s
493: learn: 0.0246269 test: 0.0410334 best: 0.0408072 (391) total: 17.4s remaining: 17.8s
494: learn: 0.0246101 test: 0.0410242 best: 0.0408072 (391) total: 17.4s remaining: 17.8s
495: learn: 0.0245932 test: 0.0410353 best: 0.0408072 (391) total: 17.4s remaining: 17.7s
496: learn: 0.0245811 test: 0.0410445 best: 0.0408072 (391) total: 17.4s remaining: 17.7s
497: learn: 0.0245494 test: 0.0410344 best: 0.0408072 (391) total: 17.4s remaining: 17.6s
498: learn: 0.0245359 test: 0.0410368 best: 0.0408072 (391) total: 17.5s remaining: 17.5s
499: learn: 0.0245234 test: 0.0410342 best: 0.0408072 (391) total: 17.5s remaining: 17.5s
500: learn: 0.0244797 test: 0.0410268 best: 0.0408072 (391) total: 17.5s remaining: 17.4s
501: learn: 0.0244612 test: 0.0410140 best: 0.0408072 (391) total: 17.5s remaining: 17.3s
502: learn: 0.0244489 test: 0.0410085 best: 0.0408072 (391) total: 17.5s remaining: 17.3s
503: learn: 0.0244227 test: 0.0410145 best: 0.0408072 (391) total: 17.5s remaining: 17.2s
504: learn: 0.0244047 test: 0.0410020 best: 0.0408072 (391) total: 17.5s remaining: 17.2s
505: learn: 0.0244047 test: 0.0410019 best: 0.0408072 (391) total: 17.5s remaining: 17.1s
506: learn: 0.0243811 test: 0.0410015 best: 0.0408072 (391) total: 17.5s remaining: 17s
507: learn: 0.0243775 test: 0.0410004 best: 0.0408072 (391) total: 17.5s remaining: 17s
508: learn: 0.0243708 test: 0.0409978 best: 0.0408072 (391) total: 17.5s remaining: 16.9s
509: learn: 0.0243642 test: 0.0409956 best: 0.0408072 (391) total: 17.6s remaining: 16.9s
510: learn: 0.0243566 test: 0.0409971 best: 0.0408072 (391) total: 17.6s remaining: 16.8s
511: learn: 0.0243465 test: 0.0409974 best: 0.0408072 (391) total: 17.6s remaining: 16.8s
512: learn: 0.0243401 test: 0.0409953 best: 0.0408072 (391) total: 17.6s remaining: 16.7s
513: learn: 0.0243287 test: 0.0409971 best: 0.0408072 (391) total: 17.6s remaining: 16.6s
514: learn: 0.0243237 test: 0.0409987 best: 0.0408072 (391) total: 17.6s remaining: 16.6s
515: learn: 0.0243076 test: 0.0410096 best: 0.0408072 (391) total: 17.6s remaining: 16.5s
516: learn: 0.0243023 test: 0.0410107 best: 0.0408072 (391) total: 17.6s remaining: 16.5s
517: learn: 0.0242782 test: 0.0410441 best: 0.0408072 (391) total: 17.6s remaining: 16.4s
518: learn: 0.0242782 test: 0.0410441 best: 0.0408072 (391) total: 17.6s remaining: 16.3s
519: learn: 0.0242395 test: 0.0410728 best: 0.0408072 (391) total: 17.6s remaining: 16.3s
520: learn: 0.0242282 test: 0.0410789 best: 0.0408072 (391) total: 17.6s remaining: 16.2s
521: learn: 0.0242243 test: 0.0410759 best: 0.0408072 (391) total: 17.7s remaining: 16.2s
522: learn: 0.0242091 test: 0.0410770 best: 0.0408072 (391) total: 17.7s remaining: 16.1s
523: learn: 0.0242042 test: 0.0410779 best: 0.0408072 (391) total: 17.7s remaining: 16.1s
524: learn: 0.0241999 test: 0.0410764 best: 0.0408072 (391) total: 17.7s remaining: 16s
525: learn: 0.0241999 test: 0.0410764 best: 0.0408072 (391) total: 17.7s remaining: 15.9s
526: learn: 0.0241798 test: 0.0410872 best: 0.0408072 (391) total: 17.7s remaining: 15.9s
527: learn: 0.0241680 test: 0.0410865 best: 0.0408072 (391) total: 17.7s remaining: 15.8s
528: learn: 0.0241679 test: 0.0410865 best: 0.0408072 (391) total: 17.7s remaining: 15.8s
529: learn: 0.0241679 test: 0.0410865 best: 0.0408072 (391) total: 17.7s remaining: 15.7s
530: learn: 0.0241630 test: 0.0410881 best: 0.0408072 (391) total: 17.7s remaining: 15.7s
531: learn: 0.0241378 test: 0.0410849 best: 0.0408072 (391) total: 17.8s remaining: 15.6s
532: learn: 0.0241223 test: 0.0410833 best: 0.0408072 (391) total: 17.8s remaining: 15.6s
533: learn: 0.0241149 test: 0.0410852 best: 0.0408072 (391) total: 17.8s remaining: 15.5s
534: learn: 0.0240952 test: 0.0410969 best: 0.0408072 (391) total: 17.8s remaining: 15.5s
535: learn: 0.0240891 test: 0.0410990 best: 0.0408072 (391) total: 17.8s remaining: 15.4s
536: learn: 0.0240668 test: 0.0410984 best: 0.0408072 (391) total: 17.8s remaining: 15.3s
537: learn: 0.0240630 test: 0.0411001 best: 0.0408072 (391) total: 17.8s remaining: 15.3s
538: learn: 0.0240455 test: 0.0410878 best: 0.0408072 (391) total: 17.8s remaining: 15.2s
539: learn: 0.0240394 test: 0.0410892 best: 0.0408072 (391) total: 17.8s remaining: 15.2s
540: learn: 0.0240275 test: 0.0410887 best: 0.0408072 (391) total: 17.8s remaining: 15.1s
541: learn: 0.0240233 test: 0.0410903 best: 0.0408072 (391) total: 17.8s remaining: 15.1s
542: learn: 0.0240133 test: 0.0410873 best: 0.0408072 (391) total: 17.9s remaining: 15s
543: learn: 0.0240133 test: 0.0410873 best: 0.0408072 (391) total: 17.9s remaining: 15s
544: learn: 0.0240020 test: 0.0410940 best: 0.0408072 (391) total: 17.9s remaining: 14.9s
545: learn: 0.0239907 test: 0.0410890 best: 0.0408072 (391) total: 17.9s remaining: 14.9s
546: learn: 0.0239846 test: 0.0410870 best: 0.0408072 (391) total: 17.9s remaining: 14.8s
547: learn: 0.0239764 test: 0.0410906 best: 0.0408072 (391) total: 17.9s remaining: 14.8s
548: learn: 0.0239673 test: 0.0410912 best: 0.0408072 (391) total: 17.9s remaining: 14.7s
549: learn: 0.0239462 test: 0.0410997 best: 0.0408072 (391) total: 17.9s remaining: 14.7s
550: learn: 0.0239234 test: 0.0411156 best: 0.0408072 (391) total: 17.9s remaining: 14.6s
551: learn: 0.0238986 test: 0.0411229 best: 0.0408072 (391) total: 17.9s remaining: 14.6s
552: learn: 0.0238835 test: 0.0411337 best: 0.0408072 (391) total: 18s remaining: 14.5s
553: learn: 0.0238729 test: 0.0411402 best: 0.0408072 (391) total: 18s remaining: 14.5s
554: learn: 0.0238620 test: 0.0411354 best: 0.0408072 (391) total: 18s remaining: 14.4s
555: learn: 0.0238213 test: 0.0411290 best: 0.0408072 (391) total: 18s remaining: 14.4s
556: learn: 0.0238175 test: 0.0411305 best: 0.0408072 (391) total: 18s remaining: 14.3s
557: learn: 0.0238029 test: 0.0411222 best: 0.0408072 (391) total: 18s remaining: 14.3s
558: learn: 0.0237962 test: 0.0411218 best: 0.0408072 (391) total: 18s remaining: 14.2s
559: learn: 0.0237799 test: 0.0411258 best: 0.0408072 (391) total: 18s remaining: 14.2s
560: learn: 0.0237702 test: 0.0411485 best: 0.0408072 (391) total: 18s remaining: 14.1s
561: learn: 0.0237648 test: 0.0411440 best: 0.0408072 (391) total: 18s remaining: 14.1s
562: learn: 0.0237516 test: 0.0411423 best: 0.0408072 (391) total: 18.1s remaining: 14s
563: learn: 0.0236773 test: 0.0411530 best: 0.0408072 (391) total: 18.1s remaining: 14s
564: learn: 0.0236506 test: 0.0411609 best: 0.0408072 (391) total: 18.1s remaining: 13.9s
565: learn: 0.0236468 test: 0.0411623 best: 0.0408072 (391) total: 18.1s remaining: 13.9s
566: learn: 0.0236102 test: 0.0411700 best: 0.0408072 (391) total: 18.1s remaining: 13.8s
567: learn: 0.0235974 test: 0.0411651 best: 0.0408072 (391) total: 18.1s remaining: 13.8s
568: learn: 0.0235850 test: 0.0411602 best: 0.0408072 (391) total: 18.1s remaining: 13.7s
569: learn: 0.0235721 test: 0.0411528 best: 0.0408072 (391) total: 18.1s remaining: 13.7s
570: learn: 0.0235690 test: 0.0411542 best: 0.0408072 (391) total: 18.1s remaining: 13.6s
571: learn: 0.0235185 test: 0.0411536 best: 0.0408072 (391) total: 18.2s remaining: 13.6s
572: learn: 0.0235083 test: 0.0411551 best: 0.0408072 (391) total: 18.2s remaining: 13.5s
573: learn: 0.0235034 test: 0.0411509 best: 0.0408072 (391) total: 18.2s remaining: 13.5s
574: learn: 0.0234986 test: 0.0411469 best: 0.0408072 (391) total: 18.2s remaining: 13.4s
575: learn: 0.0234702 test: 0.0411520 best: 0.0408072 (391) total: 18.2s remaining: 13.4s
576: learn: 0.0234569 test: 0.0411627 best: 0.0408072 (391) total: 18.2s remaining: 13.4s
577: learn: 0.0234523 test: 0.0411589 best: 0.0408072 (391) total: 18.2s remaining: 13.3s
578: learn: 0.0234523 test: 0.0411589 best: 0.0408072 (391) total: 18.2s remaining: 13.3s
579: learn: 0.0234421 test: 0.0411547 best: 0.0408072 (391) total: 18.2s remaining: 13.2s
580: learn: 0.0234210 test: 0.0411532 best: 0.0408072 (391) total: 18.3s remaining: 13.2s
581: learn: 0.0234100 test: 0.0411497 best: 0.0408072 (391) total: 18.3s remaining: 13.1s
582: learn: 0.0234024 test: 0.0411522 best: 0.0408072 (391) total: 18.3s remaining: 13.1s
583: learn: 0.0233888 test: 0.0411439 best: 0.0408072 (391) total: 18.3s remaining: 13s
584: learn: 0.0233431 test: 0.0411634 best: 0.0408072 (391) total: 18.3s remaining: 13s
585: learn: 0.0233362 test: 0.0411649 best: 0.0408072 (391) total: 18.3s remaining: 12.9s
586: learn: 0.0233240 test: 0.0411577 best: 0.0408072 (391) total: 18.3s remaining: 12.9s
587: learn: 0.0233141 test: 0.0411644 best: 0.0408072 (391) total: 18.4s remaining: 12.9s
588: learn: 0.0233003 test: 0.0411649 best: 0.0408072 (391) total: 18.4s remaining: 12.8s
589: learn: 0.0232971 test: 0.0411663 best: 0.0408072 (391) total: 18.4s remaining: 12.8s
590: learn: 0.0232639 test: 0.0411543 best: 0.0408072 (391) total: 18.4s remaining: 12.7s
591: learn: 0.0232570 test: 0.0411546 best: 0.0408072 (391) total: 18.4s remaining: 12.7s
592: learn: 0.0232273 test: 0.0411604 best: 0.0408072 (391) total: 18.4s remaining: 12.6s
593: learn: 0.0232162 test: 0.0411559 best: 0.0408072 (391) total: 18.4s remaining: 12.6s
594: learn: 0.0232115 test: 0.0411574 best: 0.0408072 (391) total: 18.4s remaining: 12.5s
595: learn: 0.0232072 test: 0.0411538 best: 0.0408072 (391) total: 18.4s remaining: 12.5s
596: learn: 0.0232072 test: 0.0411537 best: 0.0408072 (391) total: 18.5s remaining: 12.5s
597: learn: 0.0231945 test: 0.0411459 best: 0.0408072 (391) total: 18.5s remaining: 12.4s
598: learn: 0.0231809 test: 0.0411507 best: 0.0408072 (391) total: 18.5s remaining: 12.4s
599: learn: 0.0231689 test: 0.0411433 best: 0.0408072 (391) total: 18.5s remaining: 12.3s
600: learn: 0.0231689 test: 0.0411433 best: 0.0408072 (391) total: 18.5s remaining: 12.3s
601: learn: 0.0231390 test: 0.0411442 best: 0.0408072 (391) total: 18.5s remaining: 12.2s
602: learn: 0.0231349 test: 0.0411407 best: 0.0408072 (391) total: 18.5s remaining: 12.2s
603: learn: 0.0231349 test: 0.0411407 best: 0.0408072 (391) total: 18.5s remaining: 12.1s
604: learn: 0.0231146 test: 0.0411388 best: 0.0408072 (391) total: 18.5s remaining: 12.1s
605: learn: 0.0231116 test: 0.0411401 best: 0.0408072 (391) total: 18.6s remaining: 12.1s
606: learn: 0.0231076 test: 0.0411367 best: 0.0408072 (391) total: 18.6s remaining: 12s
607: learn: 0.0230954 test: 0.0411485 best: 0.0408072 (391) total: 18.6s remaining: 12s
608: learn: 0.0230925 test: 0.0411454 best: 0.0408072 (391) total: 18.6s remaining: 11.9s
609: learn: 0.0230791 test: 0.0411360 best: 0.0408072 (391) total: 18.6s remaining: 11.9s
610: learn: 0.0230595 test: 0.0411343 best: 0.0408072 (391) total: 18.6s remaining: 11.8s
611: learn: 0.0230556 test: 0.0411345 best: 0.0408072 (391) total: 18.6s remaining: 11.8s
612: learn: 0.0230556 test: 0.0411345 best: 0.0408072 (391) total: 18.6s remaining: 11.8s
613: learn: 0.0230556 test: 0.0411345 best: 0.0408072 (391) total: 18.6s remaining: 11.7s
614: learn: 0.0230527 test: 0.0411293 best: 0.0408072 (391) total: 18.6s remaining: 11.7s
615: learn: 0.0230417 test: 0.0411298 best: 0.0408072 (391) total: 18.6s remaining: 11.6s
616: learn: 0.0230417 test: 0.0411298 best: 0.0408072 (391) total: 18.7s remaining: 11.6s
617: learn: 0.0230417 test: 0.0411298 best: 0.0408072 (391) total: 18.7s remaining: 11.5s
618: learn: 0.0230417 test: 0.0411298 best: 0.0408072 (391) total: 18.7s remaining: 11.5s
619: learn: 0.0230188 test: 0.0411246 best: 0.0408072 (391) total: 18.7s remaining: 11.4s
620: learn: 0.0230042 test: 0.0411349 best: 0.0408072 (391) total: 18.7s remaining: 11.4s
621: learn: 0.0229452 test: 0.0411614 best: 0.0408072 (391) total: 18.7s remaining: 11.4s
622: learn: 0.0229452 test: 0.0411614 best: 0.0408072 (391) total: 18.7s remaining: 11.3s
623: learn: 0.0229235 test: 0.0411794 best: 0.0408072 (391) total: 18.7s remaining: 11.3s
624: learn: 0.0229045 test: 0.0411779 best: 0.0408072 (391) total: 18.7s remaining: 11.2s
625: learn: 0.0228934 test: 0.0411928 best: 0.0408072 (391) total: 18.8s remaining: 11.2s
626: learn: 0.0228905 test: 0.0411897 best: 0.0408072 (391) total: 18.8s remaining: 11.2s
627: learn: 0.0228535 test: 0.0411927 best: 0.0408072 (391) total: 18.8s remaining: 11.1s
628: learn: 0.0228446 test: 0.0411892 best: 0.0408072 (391) total: 18.8s remaining: 11.1s
629: learn: 0.0228334 test: 0.0412008 best: 0.0408072 (391) total: 18.8s remaining: 11s
630: learn: 0.0227914 test: 0.0412199 best: 0.0408072 (391) total: 18.8s remaining: 11s
631: learn: 0.0227887 test: 0.0412212 best: 0.0408072 (391) total: 18.8s remaining: 10.9s
632: learn: 0.0227866 test: 0.0412224 best: 0.0408072 (391) total: 18.8s remaining: 10.9s
633: learn: 0.0227731 test: 0.0412255 best: 0.0408072 (391) total: 18.8s remaining: 10.9s
634: learn: 0.0227731 test: 0.0412254 best: 0.0408072 (391) total: 18.8s remaining: 10.8s
635: learn: 0.0227698 test: 0.0412264 best: 0.0408072 (391) total: 18.8s remaining: 10.8s
636: learn: 0.0227578 test: 0.0412281 best: 0.0408072 (391) total: 18.8s remaining: 10.7s
637: learn: 0.0227462 test: 0.0412299 best: 0.0408072 (391) total: 18.8s remaining: 10.7s
638: learn: 0.0227298 test: 0.0412359 best: 0.0408072 (391) total: 18.9s remaining: 10.7s
639: learn: 0.0227265 test: 0.0412362 best: 0.0408072 (391) total: 18.9s remaining: 10.6s
640: learn: 0.0227173 test: 0.0412427 best: 0.0408072 (391) total: 18.9s remaining: 10.6s
641: learn: 0.0227080 test: 0.0412492 best: 0.0408072 (391) total: 18.9s remaining: 10.5s
642: learn: 0.0226803 test: 0.0412547 best: 0.0408072 (391) total: 18.9s remaining: 10.5s
643: learn: 0.0226803 test: 0.0412546 best: 0.0408072 (391) total: 18.9s remaining: 10.4s
644: learn: 0.0226803 test: 0.0412546 best: 0.0408072 (391) total: 18.9s remaining: 10.4s
645: learn: 0.0226803 test: 0.0412546 best: 0.0408072 (391) total: 18.9s remaining: 10.4s
646: learn: 0.0226619 test: 0.0412528 best: 0.0408072 (391) total: 18.9s remaining: 10.3s
647: learn: 0.0226258 test: 0.0412442 best: 0.0408072 (391) total: 18.9s remaining: 10.3s
648: learn: 0.0226258 test: 0.0412442 best: 0.0408072 (391) total: 18.9s remaining: 10.2s
649: learn: 0.0225982 test: 0.0412501 best: 0.0408072 (391) total: 19s remaining: 10.2s
650: learn: 0.0225876 test: 0.0412521 best: 0.0408072 (391) total: 19s remaining: 10.2s
651: learn: 0.0225608 test: 0.0412576 best: 0.0408072 (391) total: 19s remaining: 10.1s
652: learn: 0.0225485 test: 0.0412531 best: 0.0408072 (391) total: 19s remaining: 10.1s
653: learn: 0.0225222 test: 0.0412586 best: 0.0408072 (391) total: 19s remaining: 10.1s
654: learn: 0.0225179 test: 0.0412576 best: 0.0408072 (391) total: 19s remaining: 10s
655: learn: 0.0224876 test: 0.0412733 best: 0.0408072 (391) total: 19s remaining: 9.97s
656: learn: 0.0224701 test: 0.0412828 best: 0.0408072 (391) total: 19s remaining: 9.93s
657: learn: 0.0224627 test: 0.0412879 best: 0.0408072 (391) total: 19s remaining: 9.89s
658: learn: 0.0223953 test: 0.0413157 best: 0.0408072 (391) total: 19s remaining: 9.85s
659: learn: 0.0223775 test: 0.0413136 best: 0.0408072 (391) total: 19s remaining: 9.81s
660: learn: 0.0223507 test: 0.0413068 best: 0.0408072 (391) total: 19.1s remaining: 9.77s
661: learn: 0.0223489 test: 0.0413082 best: 0.0408072 (391) total: 19.1s remaining: 9.73s
662: learn: 0.0223231 test: 0.0413021 best: 0.0408072 (391) total: 19.1s remaining: 9.69s
663: learn: 0.0222972 test: 0.0413074 best: 0.0408072 (391) total: 19.1s remaining: 9.66s
664: learn: 0.0222811 test: 0.0413179 best: 0.0408072 (391) total: 19.1s remaining: 9.62s
665: learn: 0.0222701 test: 0.0413186 best: 0.0408072 (391) total: 19.1s remaining: 9.58s
666: learn: 0.0222700 test: 0.0413185 best: 0.0408072 (391) total: 19.1s remaining: 9.54s
667: learn: 0.0222587 test: 0.0413207 best: 0.0408072 (391) total: 19.1s remaining: 9.5s
668: learn: 0.0222072 test: 0.0413528 best: 0.0408072 (391) total: 19.1s remaining: 9.47s
669: learn: 0.0221844 test: 0.0413457 best: 0.0408072 (391) total: 19.1s remaining: 9.43s
670: learn: 0.0221821 test: 0.0413476 best: 0.0408072 (391) total: 19.2s remaining: 9.39s
671: learn: 0.0221571 test: 0.0413421 best: 0.0408072 (391) total: 19.2s remaining: 9.36s
672: learn: 0.0221571 test: 0.0413420 best: 0.0408072 (391) total: 19.2s remaining: 9.32s
673: learn: 0.0221551 test: 0.0413436 best: 0.0408072 (391) total: 19.2s remaining: 9.28s
674: learn: 0.0221535 test: 0.0413451 best: 0.0408072 (391) total: 19.2s remaining: 9.25s
675: learn: 0.0221484 test: 0.0413470 best: 0.0408072 (391) total: 19.2s remaining: 9.21s
676: learn: 0.0221235 test: 0.0413521 best: 0.0408072 (391) total: 19.2s remaining: 9.17s
677: learn: 0.0221136 test: 0.0413546 best: 0.0408072 (391) total: 19.2s remaining: 9.13s
678: learn: 0.0220891 test: 0.0413595 best: 0.0408072 (391) total: 19.2s remaining: 9.1s
679: learn: 0.0220547 test: 0.0413647 best: 0.0408072 (391) total: 19.3s remaining: 9.06s
680: learn: 0.0220506 test: 0.0413640 best: 0.0408072 (391) total: 19.3s remaining: 9.02s
681: learn: 0.0220402 test: 0.0413719 best: 0.0408072 (391) total: 19.3s remaining: 8.98s
682: learn: 0.0220402 test: 0.0413718 best: 0.0408072 (391) total: 19.3s remaining: 8.94s
683: learn: 0.0220194 test: 0.0413860 best: 0.0408072 (391) total: 19.3s remaining: 8.91s
684: learn: 0.0220078 test: 0.0413798 best: 0.0408072 (391) total: 19.3s remaining: 8.87s
685: learn: 0.0220036 test: 0.0413795 best: 0.0408072 (391) total: 19.3s remaining: 8.83s
686: learn: 0.0220018 test: 0.0413810 best: 0.0408072 (391) total: 19.3s remaining: 8.79s
687: learn: 0.0219742 test: 0.0414012 best: 0.0408072 (391) total: 19.3s remaining: 8.76s
688: learn: 0.0219575 test: 0.0413994 best: 0.0408072 (391) total: 19.3s remaining: 8.72s
689: learn: 0.0219510 test: 0.0413937 best: 0.0408072 (391) total: 19.3s remaining: 8.68s
690: learn: 0.0219233 test: 0.0414096 best: 0.0408072 (391) total: 19.3s remaining: 8.65s
691: learn: 0.0219160 test: 0.0414349 best: 0.0408072 (391) total: 19.4s remaining: 8.61s
692: learn: 0.0218935 test: 0.0413990 best: 0.0408072 (391) total: 19.4s remaining: 8.58s
693: learn: 0.0218683 test: 0.0414124 best: 0.0408072 (391) total: 19.4s remaining: 8.54s
694: learn: 0.0218547 test: 0.0414094 best: 0.0408072 (391) total: 19.4s remaining: 8.51s
695: learn: 0.0218456 test: 0.0414138 best: 0.0408072 (391) total: 19.4s remaining: 8.47s
696: learn: 0.0218311 test: 0.0414178 best: 0.0408072 (391) total: 19.4s remaining: 8.43s
697: learn: 0.0218293 test: 0.0414194 best: 0.0408072 (391) total: 19.4s remaining: 8.4s
698: learn: 0.0218165 test: 0.0414314 best: 0.0408072 (391) total: 19.4s remaining: 8.36s
699: learn: 0.0217925 test: 0.0414367 best: 0.0408072 (391) total: 19.4s remaining: 8.32s
700: learn: 0.0217681 test: 0.0414500 best: 0.0408072 (391) total: 19.4s remaining: 8.29s
701: learn: 0.0217396 test: 0.0414564 best: 0.0408072 (391) total: 19.4s remaining: 8.25s
702: learn: 0.0217293 test: 0.0414574 best: 0.0408072 (391) total: 19.4s remaining: 8.21s
703: learn: 0.0217210 test: 0.0414636 best: 0.0408072 (391) total: 19.5s remaining: 8.18s
704: learn: 0.0216930 test: 0.0414780 best: 0.0408072 (391) total: 19.5s remaining: 8.14s
705: learn: 0.0216930 test: 0.0414780 best: 0.0408072 (391) total: 19.5s remaining: 8.11s
706: learn: 0.0216690 test: 0.0414699 best: 0.0408072 (391) total: 19.5s remaining: 8.07s
707: learn: 0.0216438 test: 0.0414761 best: 0.0408072 (391) total: 19.5s remaining: 8.04s
708: learn: 0.0216438 test: 0.0414761 best: 0.0408072 (391) total: 19.5s remaining: 8s
709: learn: 0.0216438 test: 0.0414760 best: 0.0408072 (391) total: 19.5s remaining: 7.96s
710: learn: 0.0216218 test: 0.0414402 best: 0.0408072 (391) total: 19.5s remaining: 7.93s
711: learn: 0.0216218 test: 0.0414402 best: 0.0408072 (391) total: 19.5s remaining: 7.89s
712: learn: 0.0216217 test: 0.0414402 best: 0.0408072 (391) total: 19.5s remaining: 7.86s
713: learn: 0.0216201 test: 0.0414417 best: 0.0408072 (391) total: 19.5s remaining: 7.82s
714: learn: 0.0215961 test: 0.0414547 best: 0.0408072 (391) total: 19.5s remaining: 7.79s
715: learn: 0.0215961 test: 0.0414547 best: 0.0408072 (391) total: 19.6s remaining: 7.75s
716: learn: 0.0215961 test: 0.0414547 best: 0.0408072 (391) total: 19.6s remaining: 7.72s
717: learn: 0.0215728 test: 0.0414473 best: 0.0408072 (391) total: 19.6s remaining: 7.68s
718: learn: 0.0215638 test: 0.0414496 best: 0.0408072 (391) total: 19.6s remaining: 7.65s
719: learn: 0.0215638 test: 0.0414495 best: 0.0408072 (391) total: 19.6s remaining: 7.61s
720: learn: 0.0215485 test: 0.0414479 best: 0.0408072 (391) total: 19.6s remaining: 7.58s
721: learn: 0.0215384 test: 0.0414556 best: 0.0408072 (391) total: 19.6s remaining: 7.54s
722: learn: 0.0215095 test: 0.0414800 best: 0.0408072 (391) total: 19.6s remaining: 7.51s
723: learn: 0.0214777 test: 0.0414746 best: 0.0408072 (391) total: 19.6s remaining: 7.47s
724: learn: 0.0214552 test: 0.0414677 best: 0.0408072 (391) total: 19.6s remaining: 7.44s
725: learn: 0.0214339 test: 0.0414775 best: 0.0408072 (391) total: 19.6s remaining: 7.41s
726: learn: 0.0214201 test: 0.0414631 best: 0.0408072 (391) total: 19.6s remaining: 7.37s
727: learn: 0.0214201 test: 0.0414631 best: 0.0408072 (391) total: 19.6s remaining: 7.34s
728: learn: 0.0214200 test: 0.0414633 best: 0.0408072 (391) total: 19.7s remaining: 7.3s
729: learn: 0.0213935 test: 0.0414892 best: 0.0408072 (391) total: 19.7s remaining: 7.27s
730: learn: 0.0213522 test: 0.0414794 best: 0.0408072 (391) total: 19.7s remaining: 7.24s
731: learn: 0.0213374 test: 0.0414876 best: 0.0408072 (391) total: 19.7s remaining: 7.21s
732: learn: 0.0213373 test: 0.0414874 best: 0.0408072 (391) total: 19.7s remaining: 7.17s
733: learn: 0.0213372 test: 0.0414875 best: 0.0408072 (391) total: 19.7s remaining: 7.14s
734: learn: 0.0213139 test: 0.0414932 best: 0.0408072 (391) total: 19.7s remaining: 7.1s
735: learn: 0.0212951 test: 0.0414749 best: 0.0408072 (391) total: 19.7s remaining: 7.07s
736: learn: 0.0212858 test: 0.0414729 best: 0.0408072 (391) total: 19.7s remaining: 7.04s
737: learn: 0.0212506 test: 0.0415497 best: 0.0408072 (391) total: 19.7s remaining: 7.01s
738: learn: 0.0212361 test: 0.0415580 best: 0.0408072 (391) total: 19.8s remaining: 6.98s
739: learn: 0.0212361 test: 0.0415580 best: 0.0408072 (391) total: 19.8s remaining: 6.94s
740: learn: 0.0212221 test: 0.0415640 best: 0.0408072 (391) total: 19.8s remaining: 6.91s
741: learn: 0.0211989 test: 0.0415699 best: 0.0408072 (391) total: 19.8s remaining: 6.88s
742: learn: 0.0211754 test: 0.0415827 best: 0.0408072 (391) total: 19.8s remaining: 6.84s
743: learn: 0.0211663 test: 0.0415822 best: 0.0408072 (391) total: 19.8s remaining: 6.81s
744: learn: 0.0211347 test: 0.0415880 best: 0.0408072 (391) total: 19.8s remaining: 6.78s
745: learn: 0.0211271 test: 0.0415911 best: 0.0408072 (391) total: 19.8s remaining: 6.75s
746: learn: 0.0211271 test: 0.0415911 best: 0.0408072 (391) total: 19.8s remaining: 6.71s
747: learn: 0.0211113 test: 0.0415900 best: 0.0408072 (391) total: 19.8s remaining: 6.68s
748: learn: 0.0211112 test: 0.0415901 best: 0.0408072 (391) total: 19.8s remaining: 6.65s
749: learn: 0.0210690 test: 0.0416256 best: 0.0408072 (391) total: 19.8s remaining: 6.62s
750: learn: 0.0210572 test: 0.0416548 best: 0.0408072 (391) total: 19.9s remaining: 6.58s
751: learn: 0.0210572 test: 0.0416547 best: 0.0408072 (391) total: 19.9s remaining: 6.55s
752: learn: 0.0210342 test: 0.0416702 best: 0.0408072 (391) total: 19.9s remaining: 6.52s
753: learn: 0.0210235 test: 0.0416637 best: 0.0408072 (391) total: 19.9s remaining: 6.49s
754: learn: 0.0210235 test: 0.0416636 best: 0.0408072 (391) total: 19.9s remaining: 6.45s
755: learn: 0.0210019 test: 0.0416690 best: 0.0408072 (391) total: 19.9s remaining: 6.42s
756: learn: 0.0209809 test: 0.0416816 best: 0.0408072 (391) total: 19.9s remaining: 6.39s
757: learn: 0.0209796 test: 0.0416832 best: 0.0408072 (391) total: 19.9s remaining: 6.36s
758: learn: 0.0209573 test: 0.0416891 best: 0.0408072 (391) total: 19.9s remaining: 6.32s
759: learn: 0.0209573 test: 0.0416890 best: 0.0408072 (391) total: 19.9s remaining: 6.29s
760: learn: 0.0209066 test: 0.0416567 best: 0.0408072 (391) total: 19.9s remaining: 6.26s
761: learn: 0.0208598 test: 0.0416884 best: 0.0408072 (391) total: 20s remaining: 6.23s
762: learn: 0.0208308 test: 0.0416847 best: 0.0408072 (391) total: 20s remaining: 6.2s
763: learn: 0.0208308 test: 0.0416847 best: 0.0408072 (391) total: 20s remaining: 6.17s
764: learn: 0.0208273 test: 0.0416846 best: 0.0408072 (391) total: 20s remaining: 6.13s
765: learn: 0.0208063 test: 0.0416919 best: 0.0408072 (391) total: 20s remaining: 6.1s
766: learn: 0.0208063 test: 0.0416919 best: 0.0408072 (391) total: 20s remaining: 6.07s
767: learn: 0.0208063 test: 0.0416918 best: 0.0408072 (391) total: 20s remaining: 6.04s
768: learn: 0.0207978 test: 0.0416913 best: 0.0408072 (391) total: 20s remaining: 6.01s
769: learn: 0.0207978 test: 0.0416913 best: 0.0408072 (391) total: 20s remaining: 5.98s
770: learn: 0.0207763 test: 0.0416975 best: 0.0408072 (391) total: 20s remaining: 5.95s
771: learn: 0.0207614 test: 0.0416939 best: 0.0408072 (391) total: 20s remaining: 5.92s
772: learn: 0.0207443 test: 0.0417200 best: 0.0408072 (391) total: 20s remaining: 5.88s
773: learn: 0.0207330 test: 0.0417237 best: 0.0408072 (391) total: 20s remaining: 5.85s
774: learn: 0.0207131 test: 0.0417280 best: 0.0408072 (391) total: 20.1s remaining: 5.82s
775: learn: 0.0207074 test: 0.0417350 best: 0.0408072 (391) total: 20.1s remaining: 5.79s
776: learn: 0.0206873 test: 0.0417457 best: 0.0408072 (391) total: 20.1s remaining: 5.76s
777: learn: 0.0206673 test: 0.0417123 best: 0.0408072 (391) total: 20.1s remaining: 5.73s
778: learn: 0.0206569 test: 0.0417056 best: 0.0408072 (391) total: 20.1s remaining: 5.7s
779: learn: 0.0206297 test: 0.0417148 best: 0.0408072 (391) total: 20.1s remaining: 5.67s
780: learn: 0.0206145 test: 0.0417117 best: 0.0408072 (391) total: 20.1s remaining: 5.64s
781: learn: 0.0205947 test: 0.0417073 best: 0.0408072 (391) total: 20.1s remaining: 5.61s
782: learn: 0.0205747 test: 0.0417127 best: 0.0408072 (391) total: 20.1s remaining: 5.58s
783: learn: 0.0205746 test: 0.0417125 best: 0.0408072 (391) total: 20.1s remaining: 5.55s
784: learn: 0.0205746 test: 0.0417125 best: 0.0408072 (391) total: 20.2s remaining: 5.52s
785: learn: 0.0205636 test: 0.0417414 best: 0.0408072 (391) total: 20.2s remaining: 5.49s
786: learn: 0.0204975 test: 0.0417541 best: 0.0408072 (391) total: 20.2s remaining: 5.46s
787: learn: 0.0204786 test: 0.0417591 best: 0.0408072 (391) total: 20.2s remaining: 5.43s
788: learn: 0.0204559 test: 0.0417716 best: 0.0408072 (391) total: 20.2s remaining: 5.4s
789: learn: 0.0204407 test: 0.0417743 best: 0.0408072 (391) total: 20.2s remaining: 5.37s
790: learn: 0.0204108 test: 0.0417802 best: 0.0408072 (391) total: 20.2s remaining: 5.34s
791: learn: 0.0203887 test: 0.0417930 best: 0.0408072 (391) total: 20.2s remaining: 5.31s
792: learn: 0.0203622 test: 0.0417823 best: 0.0408072 (391) total: 20.2s remaining: 5.28s
793: learn: 0.0203406 test: 0.0417949 best: 0.0408072 (391) total: 20.2s remaining: 5.25s
794: learn: 0.0203406 test: 0.0417949 best: 0.0408072 (391) total: 20.2s remaining: 5.22s
795: learn: 0.0203289 test: 0.0417862 best: 0.0408072 (391) total: 20.3s remaining: 5.19s
796: learn: 0.0203079 test: 0.0417925 best: 0.0408072 (391) total: 20.3s remaining: 5.16s
797: learn: 0.0202930 test: 0.0417911 best: 0.0408072 (391) total: 20.3s remaining: 5.13s
798: learn: 0.0202800 test: 0.0418005 best: 0.0408072 (391) total: 20.3s remaining: 5.1s
799: learn: 0.0202686 test: 0.0417921 best: 0.0408072 (391) total: 20.3s remaining: 5.07s
800: learn: 0.0202686 test: 0.0417921 best: 0.0408072 (391) total: 20.3s remaining: 5.04s
801: learn: 0.0202685 test: 0.0417922 best: 0.0408072 (391) total: 20.3s remaining: 5.01s
802: learn: 0.0202685 test: 0.0417921 best: 0.0408072 (391) total: 20.3s remaining: 4.98s
803: learn: 0.0202394 test: 0.0417983 best: 0.0408072 (391) total: 20.3s remaining: 4.95s
804: learn: 0.0202393 test: 0.0417980 best: 0.0408072 (391) total: 20.3s remaining: 4.92s
805: learn: 0.0202392 test: 0.0417979 best: 0.0408072 (391) total: 20.3s remaining: 4.89s
806: learn: 0.0202338 test: 0.0418049 best: 0.0408072 (391) total: 20.3s remaining: 4.86s
807: learn: 0.0202111 test: 0.0417860 best: 0.0408072 (391) total: 20.4s remaining: 4.84s
808: learn: 0.0201915 test: 0.0417917 best: 0.0408072 (391) total: 20.4s remaining: 4.81s
809: learn: 0.0201862 test: 0.0417921 best: 0.0408072 (391) total: 20.4s remaining: 4.78s
810: learn: 0.0201862 test: 0.0417921 best: 0.0408072 (391) total: 20.4s remaining: 4.75s
811: learn: 0.0201635 test: 0.0418165 best: 0.0408072 (391) total: 20.4s remaining: 4.72s
812: learn: 0.0201439 test: 0.0418289 best: 0.0408072 (391) total: 20.4s remaining: 4.69s
813: learn: 0.0201234 test: 0.0418414 best: 0.0408072 (391) total: 20.4s remaining: 4.66s
814: learn: 0.0200987 test: 0.0418315 best: 0.0408072 (391) total: 20.4s remaining: 4.63s
815: learn: 0.0200796 test: 0.0418370 best: 0.0408072 (391) total: 20.4s remaining: 4.61s
816: learn: 0.0200796 test: 0.0418370 best: 0.0408072 (391) total: 20.5s remaining: 4.58s
817: learn: 0.0200796 test: 0.0418370 best: 0.0408072 (391) total: 20.5s remaining: 4.56s
818: learn: 0.0200795 test: 0.0418370 best: 0.0408072 (391) total: 20.5s remaining: 4.53s
819: learn: 0.0200638 test: 0.0418348 best: 0.0408072 (391) total: 20.5s remaining: 4.5s
820: learn: 0.0200638 test: 0.0418348 best: 0.0408072 (391) total: 20.5s remaining: 4.48s
821: learn: 0.0200637 test: 0.0418347 best: 0.0408072 (391) total: 20.5s remaining: 4.45s
822: learn: 0.0200533 test: 0.0418629 best: 0.0408072 (391) total: 20.6s remaining: 4.42s
823: learn: 0.0200374 test: 0.0418748 best: 0.0408072 (391) total: 20.6s remaining: 4.4s
824: learn: 0.0200186 test: 0.0418409 best: 0.0408072 (391) total: 20.6s remaining: 4.37s
825: learn: 0.0199949 test: 0.0418318 best: 0.0408072 (391) total: 20.6s remaining: 4.35s
826: learn: 0.0199949 test: 0.0418317 best: 0.0408072 (391) total: 20.7s remaining: 4.32s
827: learn: 0.0199495 test: 0.0418581 best: 0.0408072 (391) total: 20.7s remaining: 4.29s
828: learn: 0.0199464 test: 0.0418561 best: 0.0408072 (391) total: 20.7s remaining: 4.26s
829: learn: 0.0199277 test: 0.0418618 best: 0.0408072 (391) total: 20.7s remaining: 4.24s
830: learn: 0.0199227 test: 0.0418686 best: 0.0408072 (391) total: 20.7s remaining: 4.21s
831: learn: 0.0199227 test: 0.0418685 best: 0.0408072 (391) total: 20.7s remaining: 4.18s
832: learn: 0.0199199 test: 0.0418677 best: 0.0408072 (391) total: 20.7s remaining: 4.15s
833: learn: 0.0199199 test: 0.0418677 best: 0.0408072 (391) total: 20.7s remaining: 4.13s
834: learn: 0.0199032 test: 0.0418696 best: 0.0408072 (391) total: 20.7s remaining: 4.1s
835: learn: 0.0198944 test: 0.0418723 best: 0.0408072 (391) total: 20.8s remaining: 4.07s
836: learn: 0.0198863 test: 0.0418770 best: 0.0408072 (391) total: 20.8s remaining: 4.05s
837: learn: 0.0198443 test: 0.0419084 best: 0.0408072 (391) total: 20.8s remaining: 4.02s
838: learn: 0.0198411 test: 0.0419096 best: 0.0408072 (391) total: 20.8s remaining: 3.99s
839: learn: 0.0198149 test: 0.0419100 best: 0.0408072 (391) total: 20.8s remaining: 3.96s
840: learn: 0.0197954 test: 0.0419221 best: 0.0408072 (391) total: 20.8s remaining: 3.94s
841: learn: 0.0197674 test: 0.0419280 best: 0.0408072 (391) total: 20.8s remaining: 3.91s
842: learn: 0.0197674 test: 0.0419280 best: 0.0408072 (391) total: 20.8s remaining: 3.88s
843: learn: 0.0197674 test: 0.0419280 best: 0.0408072 (391) total: 20.8s remaining: 3.85s
844: learn: 0.0197673 test: 0.0419279 best: 0.0408072 (391) total: 20.9s remaining: 3.82s
845: learn: 0.0197673 test: 0.0419279 best: 0.0408072 (391) total: 20.9s remaining: 3.8s
846: learn: 0.0197646 test: 0.0419239 best: 0.0408072 (391) total: 20.9s remaining: 3.77s
847: learn: 0.0197603 test: 0.0419331 best: 0.0408072 (391) total: 20.9s remaining: 3.74s
848: learn: 0.0197220 test: 0.0419674 best: 0.0408072 (391) total: 20.9s remaining: 3.71s
849: learn: 0.0197166 test: 0.0419717 best: 0.0408072 (391) total: 20.9s remaining: 3.69s
850: learn: 0.0196943 test: 0.0419903 best: 0.0408072 (391) total: 20.9s remaining: 3.66s
851: learn: 0.0196763 test: 0.0419886 best: 0.0408072 (391) total: 20.9s remaining: 3.63s
852: learn: 0.0196763 test: 0.0419885 best: 0.0408072 (391) total: 20.9s remaining: 3.6s
853: learn: 0.0196655 test: 0.0419983 best: 0.0408072 (391) total: 20.9s remaining: 3.58s
854: learn: 0.0196561 test: 0.0419915 best: 0.0408072 (391) total: 20.9s remaining: 3.55s
855: learn: 0.0196560 test: 0.0419915 best: 0.0408072 (391) total: 20.9s remaining: 3.52s
856: learn: 0.0196496 test: 0.0419969 best: 0.0408072 (391) total: 21s remaining: 3.5s
857: learn: 0.0196384 test: 0.0420063 best: 0.0408072 (391) total: 21s remaining: 3.47s
858: learn: 0.0196384 test: 0.0420063 best: 0.0408072 (391) total: 21s remaining: 3.44s
859: learn: 0.0196205 test: 0.0419727 best: 0.0408072 (391) total: 21s remaining: 3.42s
860: learn: 0.0195827 test: 0.0419574 best: 0.0408072 (391) total: 21s remaining: 3.39s
861: learn: 0.0195827 test: 0.0419574 best: 0.0408072 (391) total: 21s remaining: 3.36s
862: learn: 0.0195489 test: 0.0419766 best: 0.0408072 (391) total: 21s remaining: 3.34s
863: learn: 0.0195147 test: 0.0419603 best: 0.0408072 (391) total: 21s remaining: 3.31s
864: learn: 0.0194996 test: 0.0419719 best: 0.0408072 (391) total: 21s remaining: 3.28s
865: learn: 0.0194996 test: 0.0419719 best: 0.0408072 (391) total: 21s remaining: 3.25s
866: learn: 0.0194818 test: 0.0419838 best: 0.0408072 (391) total: 21s remaining: 3.23s
867: learn: 0.0194742 test: 0.0419860 best: 0.0408072 (391) total: 21.1s remaining: 3.2s
868: learn: 0.0194193 test: 0.0419942 best: 0.0408072 (391) total: 21.1s remaining: 3.17s
869: learn: 0.0194157 test: 0.0419940 best: 0.0408072 (391) total: 21.1s remaining: 3.15s
870: learn: 0.0194157 test: 0.0419939 best: 0.0408072 (391) total: 21.1s remaining: 3.12s
871: learn: 0.0194157 test: 0.0419939 best: 0.0408072 (391) total: 21.1s remaining: 3.09s
872: learn: 0.0193931 test: 0.0419859 best: 0.0408072 (391) total: 21.1s remaining: 3.07s
873: learn: 0.0193719 test: 0.0419818 best: 0.0408072 (391) total: 21.1s remaining: 3.04s
874: learn: 0.0193640 test: 0.0419850 best: 0.0408072 (391) total: 21.1s remaining: 3.02s
875: learn: 0.0193196 test: 0.0419922 best: 0.0408072 (391) total: 21.1s remaining: 2.99s
876: learn: 0.0193196 test: 0.0419922 best: 0.0408072 (391) total: 21.1s remaining: 2.96s
877: learn: 0.0193109 test: 0.0419860 best: 0.0408072 (391) total: 21.1s remaining: 2.94s
878: learn: 0.0193026 test: 0.0419886 best: 0.0408072 (391) total: 21.1s remaining: 2.91s
879: learn: 0.0192620 test: 0.0419868 best: 0.0408072 (391) total: 21.2s remaining: 2.88s
880: learn: 0.0192581 test: 0.0419840 best: 0.0408072 (391) total: 21.2s remaining: 2.86s
881: learn: 0.0192581 test: 0.0419840 best: 0.0408072 (391) total: 21.2s remaining: 2.83s
882: learn: 0.0192314 test: 0.0419890 best: 0.0408072 (391) total: 21.2s remaining: 2.81s
883: learn: 0.0192314 test: 0.0419890 best: 0.0408072 (391) total: 21.2s remaining: 2.78s
884: learn: 0.0192108 test: 0.0419886 best: 0.0408072 (391) total: 21.2s remaining: 2.75s
885: learn: 0.0192108 test: 0.0419886 best: 0.0408072 (391) total: 21.2s remaining: 2.73s
886: learn: 0.0192033 test: 0.0419934 best: 0.0408072 (391) total: 21.2s remaining: 2.7s
887: learn: 0.0191888 test: 0.0420049 best: 0.0408072 (391) total: 21.2s remaining: 2.68s
888: learn: 0.0191872 test: 0.0420080 best: 0.0408072 (391) total: 21.2s remaining: 2.65s
889: learn: 0.0191872 test: 0.0420080 best: 0.0408072 (391) total: 21.2s remaining: 2.63s
890: learn: 0.0191759 test: 0.0420143 best: 0.0408072 (391) total: 21.3s remaining: 2.6s
891: learn: 0.0191550 test: 0.0420064 best: 0.0408072 (391) total: 21.3s remaining: 2.57s
892: learn: 0.0191362 test: 0.0420214 best: 0.0408072 (391) total: 21.3s remaining: 2.55s
893: learn: 0.0191362 test: 0.0420214 best: 0.0408072 (391) total: 21.3s remaining: 2.52s
894: learn: 0.0191172 test: 0.0420284 best: 0.0408072 (391) total: 21.3s remaining: 2.5s
895: learn: 0.0191172 test: 0.0420284 best: 0.0408072 (391) total: 21.3s remaining: 2.47s
896: learn: 0.0191172 test: 0.0420283 best: 0.0408072 (391) total: 21.3s remaining: 2.44s
897: learn: 0.0191119 test: 0.0420453 best: 0.0408072 (391) total: 21.3s remaining: 2.42s
898: learn: 0.0190856 test: 0.0420517 best: 0.0408072 (391) total: 21.3s remaining: 2.39s
899: learn: 0.0190812 test: 0.0420583 best: 0.0408072 (391) total: 21.3s remaining: 2.37s
900: learn: 0.0190812 test: 0.0420583 best: 0.0408072 (391) total: 21.3s remaining: 2.34s
901: learn: 0.0190423 test: 0.0420896 best: 0.0408072 (391) total: 21.3s remaining: 2.32s
902: learn: 0.0189881 test: 0.0420945 best: 0.0408072 (391) total: 21.3s remaining: 2.29s
903: learn: 0.0189771 test: 0.0421044 best: 0.0408072 (391) total: 21.4s remaining: 2.27s
904: learn: 0.0189666 test: 0.0421141 best: 0.0408072 (391) total: 21.4s remaining: 2.24s
905: learn: 0.0189666 test: 0.0421141 best: 0.0408072 (391) total: 21.4s remaining: 2.22s
906: learn: 0.0189598 test: 0.0421185 best: 0.0408072 (391) total: 21.4s remaining: 2.19s
907: learn: 0.0189528 test: 0.0421196 best: 0.0408072 (391) total: 21.4s remaining: 2.17s
908: learn: 0.0189477 test: 0.0421349 best: 0.0408072 (391) total: 21.4s remaining: 2.14s
909: learn: 0.0189392 test: 0.0421417 best: 0.0408072 (391) total: 21.4s remaining: 2.12s
910: learn: 0.0189323 test: 0.0421407 best: 0.0408072 (391) total: 21.4s remaining: 2.09s
911: learn: 0.0189323 test: 0.0421407 best: 0.0408072 (391) total: 21.4s remaining: 2.07s
912: learn: 0.0189323 test: 0.0421406 best: 0.0408072 (391) total: 21.4s remaining: 2.04s
913: learn: 0.0189267 test: 0.0421459 best: 0.0408072 (391) total: 21.4s remaining: 2.02s
914: learn: 0.0189267 test: 0.0421458 best: 0.0408072 (391) total: 21.4s remaining: 1.99s
915: learn: 0.0189226 test: 0.0421523 best: 0.0408072 (391) total: 21.4s remaining: 1.97s
916: learn: 0.0189025 test: 0.0421681 best: 0.0408072 (391) total: 21.4s remaining: 1.94s
917: learn: 0.0188651 test: 0.0421544 best: 0.0408072 (391) total: 21.5s remaining: 1.92s
918: learn: 0.0188488 test: 0.0421497 best: 0.0408072 (391) total: 21.5s remaining: 1.89s
919: learn: 0.0188316 test: 0.0421612 best: 0.0408072 (391) total: 21.5s remaining: 1.87s
920: learn: 0.0187970 test: 0.0421429 best: 0.0408072 (391) total: 21.5s remaining: 1.84s
921: learn: 0.0187908 test: 0.0421499 best: 0.0408072 (391) total: 21.5s remaining: 1.82s
922: learn: 0.0187826 test: 0.0421440 best: 0.0408072 (391) total: 21.5s remaining: 1.79s
923: learn: 0.0187633 test: 0.0421599 best: 0.0408072 (391) total: 21.5s remaining: 1.77s
924: learn: 0.0187614 test: 0.0421597 best: 0.0408072 (391) total: 21.5s remaining: 1.75s
925: learn: 0.0187599 test: 0.0421617 best: 0.0408072 (391) total: 21.5s remaining: 1.72s
926: learn: 0.0187395 test: 0.0421571 best: 0.0408072 (391) total: 21.5s remaining: 1.7s
927: learn: 0.0186891 test: 0.0421608 best: 0.0408072 (391) total: 21.5s remaining: 1.67s
928: learn: 0.0186891 test: 0.0421608 best: 0.0408072 (391) total: 21.6s remaining: 1.65s
929: learn: 0.0186707 test: 0.0421715 best: 0.0408072 (391) total: 21.6s remaining: 1.63s
930: learn: 0.0186707 test: 0.0421715 best: 0.0408072 (391) total: 21.6s remaining: 1.6s
931: learn: 0.0186707 test: 0.0421715 best: 0.0408072 (391) total: 21.6s remaining: 1.58s
932: learn: 0.0186620 test: 0.0421993 best: 0.0408072 (391) total: 21.7s remaining: 1.55s
933: learn: 0.0186359 test: 0.0422068 best: 0.0408072 (391) total: 21.7s remaining: 1.53s
934: learn: 0.0186147 test: 0.0422149 best: 0.0408072 (391) total: 21.7s remaining: 1.51s
935: learn: 0.0185968 test: 0.0422138 best: 0.0408072 (391) total: 21.7s remaining: 1.48s
936: learn: 0.0185565 test: 0.0422222 best: 0.0408072 (391) total: 21.7s remaining: 1.46s
937: learn: 0.0184729 test: 0.0422338 best: 0.0408072 (391) total: 21.7s remaining: 1.44s
938: learn: 0.0184458 test: 0.0422323 best: 0.0408072 (391) total: 21.7s remaining: 1.41s
939: learn: 0.0184098 test: 0.0422640 best: 0.0408072 (391) total: 21.8s remaining: 1.39s
940: learn: 0.0184002 test: 0.0422724 best: 0.0408072 (391) total: 21.8s remaining: 1.36s
941: learn: 0.0183824 test: 0.0422875 best: 0.0408072 (391) total: 21.8s remaining: 1.34s
942: learn: 0.0183824 test: 0.0422875 best: 0.0408072 (391) total: 21.8s remaining: 1.32s
943: learn: 0.0183823 test: 0.0422876 best: 0.0408072 (391) total: 21.8s remaining: 1.29s
944: learn: 0.0183822 test: 0.0422876 best: 0.0408072 (391) total: 21.8s remaining: 1.27s
945: learn: 0.0183758 test: 0.0422873 best: 0.0408072 (391) total: 21.8s remaining: 1.25s
946: learn: 0.0183582 test: 0.0423100 best: 0.0408072 (391) total: 21.9s remaining: 1.22s
947: learn: 0.0183443 test: 0.0423138 best: 0.0408072 (391) total: 21.9s remaining: 1.2s
948: learn: 0.0183395 test: 0.0423283 best: 0.0408072 (391) total: 21.9s remaining: 1.18s
949: learn: 0.0183265 test: 0.0423350 best: 0.0408072 (391) total: 21.9s remaining: 1.15s
950: learn: 0.0183113 test: 0.0423304 best: 0.0408072 (391) total: 21.9s remaining: 1.13s
951: learn: 0.0183113 test: 0.0423304 best: 0.0408072 (391) total: 21.9s remaining: 1.1s
952: learn: 0.0182870 test: 0.0423317 best: 0.0408072 (391) total: 21.9s remaining: 1.08s
953: learn: 0.0182870 test: 0.0423317 best: 0.0408072 (391) total: 21.9s remaining: 1.06s
954: learn: 0.0182772 test: 0.0423220 best: 0.0408072 (391) total: 21.9s remaining: 1.03s
955: learn: 0.0182521 test: 0.0423098 best: 0.0408072 (391) total: 22s remaining: 1.01s
956: learn: 0.0182343 test: 0.0423061 best: 0.0408072 (391) total: 22s remaining: 987ms
957: learn: 0.0182278 test: 0.0423052 best: 0.0408072 (391) total: 22s remaining: 964ms
958: learn: 0.0182024 test: 0.0423040 best: 0.0408072 (391) total: 22s remaining: 940ms
959: learn: 0.0181989 test: 0.0423134 best: 0.0408072 (391) total: 22s remaining: 917ms
960: learn: 0.0181971 test: 0.0423129 best: 0.0408072 (391) total: 22s remaining: 893ms
961: learn: 0.0181799 test: 0.0423281 best: 0.0408072 (391) total: 22s remaining: 870ms
962: learn: 0.0181620 test: 0.0423297 best: 0.0408072 (391) total: 22s remaining: 846ms
963: learn: 0.0181561 test: 0.0423370 best: 0.0408072 (391) total: 22s remaining: 823ms
964: learn: 0.0181284 test: 0.0423758 best: 0.0408072 (391) total: 22s remaining: 800ms
965: learn: 0.0181268 test: 0.0423779 best: 0.0408072 (391) total: 22.1s remaining: 776ms
966: learn: 0.0181100 test: 0.0423933 best: 0.0408072 (391) total: 22.1s remaining: 753ms
967: learn: 0.0181100 test: 0.0423933 best: 0.0408072 (391) total: 22.1s remaining: 730ms
968: learn: 0.0180692 test: 0.0424024 best: 0.0408072 (391) total: 22.1s remaining: 706ms
969: learn: 0.0180692 test: 0.0424024 best: 0.0408072 (391) total: 22.1s remaining: 683ms
970: learn: 0.0180692 test: 0.0424024 best: 0.0408072 (391) total: 22.1s remaining: 660ms
971: learn: 0.0180692 test: 0.0424024 best: 0.0408072 (391) total: 22.1s remaining: 637ms
972: learn: 0.0180501 test: 0.0424188 best: 0.0408072 (391) total: 22.1s remaining: 614ms
973: learn: 0.0180501 test: 0.0424188 best: 0.0408072 (391) total: 22.1s remaining: 591ms
974: learn: 0.0180380 test: 0.0424235 best: 0.0408072 (391) total: 22.1s remaining: 568ms
975: learn: 0.0180222 test: 0.0424331 best: 0.0408072 (391) total: 22.2s remaining: 545ms
976: learn: 0.0180222 test: 0.0424331 best: 0.0408072 (391) total: 22.2s remaining: 522ms
977: learn: 0.0180205 test: 0.0424351 best: 0.0408072 (391) total: 22.2s remaining: 499ms
978: learn: 0.0180205 test: 0.0424351 best: 0.0408072 (391) total: 22.2s remaining: 476ms
979: learn: 0.0180205 test: 0.0424351 best: 0.0408072 (391) total: 22.2s remaining: 454ms
980: learn: 0.0180135 test: 0.0424305 best: 0.0408072 (391) total: 22.2s remaining: 431ms
981: learn: 0.0179882 test: 0.0424375 best: 0.0408072 (391) total: 22.3s remaining: 408ms
982: learn: 0.0179728 test: 0.0424357 best: 0.0408072 (391) total: 22.3s remaining: 385ms
983: learn: 0.0179648 test: 0.0424373 best: 0.0408072 (391) total: 22.3s remaining: 363ms
984: learn: 0.0179648 test: 0.0424373 best: 0.0408072 (391) total: 22.3s remaining: 340ms
985: learn: 0.0179648 test: 0.0424373 best: 0.0408072 (391) total: 22.3s remaining: 317ms
986: learn: 0.0179493 test: 0.0424531 best: 0.0408072 (391) total: 22.3s remaining: 294ms
987: learn: 0.0179253 test: 0.0424545 best: 0.0408072 (391) total: 22.4s remaining: 272ms
988: learn: 0.0179108 test: 0.0424672 best: 0.0408072 (391) total: 22.4s remaining: 249ms
989: learn: 0.0178963 test: 0.0424846 best: 0.0408072 (391) total: 22.4s remaining: 226ms
990: learn: 0.0178963 test: 0.0424846 best: 0.0408072 (391) total: 22.4s remaining: 203ms
991: learn: 0.0178963 test: 0.0424846 best: 0.0408072 (391) total: 22.4s remaining: 181ms
992: learn: 0.0178819 test: 0.0425012 best: 0.0408072 (391) total: 22.4s remaining: 158ms
993: learn: 0.0178819 test: 0.0425012 best: 0.0408072 (391) total: 22.4s remaining: 135ms
994: learn: 0.0178663 test: 0.0424699 best: 0.0408072 (391) total: 22.4s remaining: 113ms
995: learn: 0.0178303 test: 0.0424791 best: 0.0408072 (391) total: 22.4s remaining: 90.1ms
996: learn: 0.0178303 test: 0.0424791 best: 0.0408072 (391) total: 22.4s remaining: 67.5ms
997: learn: 0.0178165 test: 0.0424904 best: 0.0408072 (391) total: 22.5s remaining: 45ms
998: learn: 0.0178128 test: 0.0424959 best: 0.0408072 (391) total: 22.5s remaining: 22.5ms
999: learn: 0.0177898 test: 0.0424977 best: 0.0408072 (391) total: 22.5s remaining: 0us
bestTest = 0.04080722846
bestIteration = 391
Shrink model to first 392 iterations.
Feature #9 eliminated
Feature #10 eliminated
Feature #8 eliminated
Feature #1 eliminated
Feature #7 eliminated
Feature #3 eliminated
[3]:
{'selected_features': [0, 1, 2, 3, 4, 5, 6, 7, 8],
'eliminated_features_names': ['CN_1_num_1', 'CC_0_num_0'],
'loss_graph': {'main_indices': [0],
'removed_features_count': [0, 1, 2, 3, 4, 5, 6],
'loss_values': [0.04080722846067758,
0.04061368944936421,
0.04052306106280114,
0.04082481450966274,
0.04155470069446264,
0.04233382921819037,
0.04317638985682021]},
'eliminated_features': [9, 10],
'selected_features_names': ['num_0',
'num_1',
'num_2',
'num_3',
'num_4',
'num_5',
'num_c0_num_0',
'num_c1_num_1',
'CN_0_num_0']}
By default, the number of features to be selected will be set to 50% of the number of features, and then we check which number of features removed resulted in the lowest loss function value, starting from the loss when 0 features were removed up until 50% of the features were removed (we can check that using the summary generated by the CatBoost model, accessible using the get_summary() method). However, we can override this behavior by specifying the n_feat parameter, which indicates the exact number of features that should be selected by the class. The user can still look up the summary generated by CatBoost (get_summary() method) and look which combination of features resulted in the best loss (more details on this summary can be found in CatBoost’s official documentation here).
Notice that CatBoost logs information to the console during the run. We can suppress this output by setting the catboost_log parameter to False. Another parameter to look at is the catboost_plot, which when set to True (its default value is False) will use CatBoost’s plot feature. This feature will create an interactive plot of the training results, as well as plot with the loss values for each number of features removed. The behavior of this plot might be different when running in a script environment when compared to running it in a Python Notebook environment.
[4]:
feat_sel = CatBoostSelection(catboost_log=False, catboost_plot=True)
feat_sel.fit(df=df, label_col=label_col)
feat_sel.get_summary()
No columns specified for imputation. These columns have been automatically identified:
[]
[4]:
{'selected_features': [0, 1, 2, 3, 4, 5, 6, 7, 8, 10],
'eliminated_features_names': ['CN_1_num_1'],
'loss_graph': {'main_indices': [0],
'removed_features_count': [0, 1, 2, 3, 4, 5, 6],
'loss_values': [0.03994256086264185,
0.039916477674306304,
0.0400362168295295,
0.04027958782317799,
0.04094085129490312,
0.04166788709875054,
0.043032430295762404]},
'eliminated_features': [9],
'selected_features_names': ['num_0',
'num_1',
'num_2',
'num_3',
'num_4',
'num_5',
'num_c0_num_0',
'num_c1_num_1',
'CN_0_num_0',
'CC_0_num_0']}
We can also change the algorithm used to set the importance of each feature. As previously mentioned, there are three approaches implemented by CatBoost. By using our CatBoostSelection class, these algorithms can be changed using the algorithm parameter, which accepts the following string values:
‘predict’: uses the RecursiveByPredictionValuesChange algorithm;
‘loss’: uses the RecursiveByLossFunctionChange algorithm (default value);
‘shap’: uses the RecursiveByShapValues algorithm;
Another important feature is the steps parameter, which determines how many different CatBoost models are trained before defining the importance of each feature. Higher values results in more accurate results at the cost of a longer running time for the fit() method.
Finally, we can also automatically save the summary generated by the feature selection process executed by CatBoost using the save_json and json_summary parameters. By default, save_json is set to False, which means that no JSON files are saved. By setting it to True, the summary will be saved in the file specified by json_summary.
[5]:
feat_sel = CatBoostSelection(
catboost_log=False,
catboost_plot=False,
steps=4,
algorithm="shap",
save_json=True,
json_summary="json_files/cb_feat_summary.json"
)
feat_sel.fit(df=df, label_col=label_col)
feat_sel.get_summary()
No columns specified for imputation. These columns have been automatically identified:
[]
[5]:
{'selected_features': [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
'eliminated_features_names': [],
'loss_graph': {'main_indices': [0, 2, 4, 5],
'removed_features_count': [0, 1, 2, 3, 4, 5, 6],
'loss_values': [0.042581531841316564,
0.04300461125060388,
0.04292228797527912,
0.04295873268921599,
0.04305697687862006,
0.043266739258661785,
0.045862933645388584]},
'eliminated_features': [],
'selected_features_names': ['num_0',
'num_1',
'num_2',
'num_3',
'num_4',
'num_5',
'num_c0_num_0',
'num_c1_num_1',
'CN_0_num_0',
'CN_1_num_1',
'CC_0_num_0']}
After using the .fit() method, we will have access to a summary generated by the feature selection functionality of CatBoost (printed in the above cell). This is the same summary that is saved when using the save_json parameter.
In order to transform a dataset and select only the chosen features, we simply call the transform method:
[6]:
new_df = feat_sel.transform(df)
new_df.head()
[6]:
| num_0 | num_1 | num_2 | num_3 | num_4 | num_5 | num_c0_num_0 | num_c1_num_1 | CN_0_num_0 | CN_1_num_1 | CC_0_num_0 | label | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | -3.086844 | -3.307366 | -1.562467 | 0.250221 | 0.019486 | 1.773897 | -3.088523 | -3.294641 | val0_1 | val1_4 | val0_1 | 1 |
| 1 | -2.817753 | -1.305013 | -2.649276 | 1.577143 | 0.808751 | 2.885842 | -2.791640 | -1.301471 | val0_1 | val1_0 | val0_1 | 1 |
| 2 | 0.838206 | 1.005391 | -1.835377 | 2.740660 | 2.297497 | 0.647299 | 0.838768 | 1.004481 | val0_2 | val1_3 | val0_2 | 1 |
| 3 | -4.779331 | -1.087113 | -2.335707 | 4.325008 | 0.183701 | 2.754548 | -4.788262 | -1.085760 | val0_0 | val1_4 | val0_1 | 1 |
| 4 | -0.054659 | -3.051933 | -2.188465 | 0.813331 | 3.929342 | 2.225584 | -0.033824 | -3.055057 | val0_0 | val1_2 | val0_2 | 1 |
As with the other feature selection approach, we can still overwrite the CatBoostSelection class’s selected feature by setting the list of the selected features manually.
[7]:
features_manual = ["num_0", "num_1", "num_2"]
feat_sel.set_selected_features(features_manual)
new_df = feat_sel.transform(df)
new_df
[7]:
| num_0 | num_1 | num_2 | label | |
|---|---|---|---|---|
| 0 | -3.086844 | -3.307366 | -1.562467 | 1 |
| 1 | -2.817753 | -1.305013 | -2.649276 | 1 |
| 2 | 0.838206 | 1.005391 | -1.835377 | 1 |
| 3 | -4.779331 | -1.087113 | -2.335707 | 1 |
| 4 | -0.054659 | -3.051933 | -2.188465 | 1 |
| ... | ... | ... | ... | ... |
| 9995 | -3.165588 | -3.506656 | -2.577650 | 1 |
| 9996 | -2.533155 | -4.917804 | -2.034633 | 1 |
| 9997 | -1.443761 | 2.141603 | -2.699636 | 1 |
| 9998 | -0.732597 | 0.254555 | -2.290704 | 1 |
| 9999 | -1.932632 | -3.802342 | -1.526755 | 1 |
10000 rows × 4 columns
The user can also create their own CatBoost classifier and pass it to the CatBoostSelection class. To do so, just pass the object of the CatBoost classifier to the estimator parameter. By doing this, the cat_col parameter will be ignored, since it is only used when creating a default CatBoost classifier, which is not the case when an external classifier is provided by the user.
2 - DataFrame without column names
[8]:
df = create_dummy_dataset(
samples=10000,
n_features=6,
n_num_num=2,
n_cat_num=2,
n_cat_cat=1,
num_num_noise=[0.01, 0.02],
pct_change=[0.03, 0.05]
)
df.columns = [i for i in range(df.shape[1])]
label_col = 6
df
[8]:
| 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 2.079545 | -2.445094 | 2.341241 | 2.231165 | 4.678349 | -3.700149 | 1 | 2.075150 | -2.417959 | val0_2 | val1_0 | val0_1 |
| 1 | 2.780027 | 0.082871 | 2.923892 | 3.748558 | 0.181716 | -3.360609 | 1 | 2.775921 | 0.067200 | val0_2 | val1_0 | val0_1 |
| 2 | 0.869917 | 2.041848 | 3.359383 | 2.681346 | 3.295876 | 1.909756 | 0 | 0.860433 | 2.059067 | val0_1 | val1_0 | val0_0 |
| 3 | 2.667241 | 0.794927 | 1.881219 | 0.273481 | -0.339529 | 0.293166 | 0 | 2.662639 | 0.798612 | val0_2 | val1_1 | val0_1 |
| 4 | 2.204499 | -3.331591 | 1.779642 | 1.727541 | 2.620087 | -4.066566 | 1 | 2.204856 | -3.330251 | val0_2 | val1_1 | val0_0 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 9995 | 0.822905 | 2.107042 | 3.153906 | 3.739316 | 2.715528 | 0.834233 | 0 | 0.847458 | 2.124767 | val0_1 | val1_1 | val0_0 |
| 9996 | 1.552587 | -2.924511 | 6.640770 | 4.022040 | -0.505248 | -1.746004 | 1 | 1.572432 | -2.902834 | val0_1 | val1_0 | val0_0 |
| 9997 | 0.185203 | -2.535111 | 0.165340 | 0.143321 | 1.258967 | 1.244145 | 1 | 0.184503 | -2.553170 | val0_1 | val1_0 | val0_0 |
| 9998 | 1.683835 | 0.043373 | 3.990938 | 1.661648 | 3.014460 | 0.729575 | 0 | 1.671167 | 0.037742 | val0_1 | val1_1 | val0_0 |
| 9999 | 0.961729 | -1.382162 | 3.627225 | 2.287614 | -0.509653 | -1.746448 | 1 | 0.970809 | -1.368223 | val0_1 | val1_0 | val0_0 |
10000 rows × 12 columns
[9]:
feat_sel = CatBoostSelection(catboost_log=False)
feat_sel.fit(df=df, label_col=6 )
feat_sel.get_summary()
No columns specified for imputation. These columns have been automatically identified:
[]
[9]:
{'selected_features': [1, 2, 3, 4, 5, 7, 8],
'eliminated_features_names': ['0', '7', '10', '11'],
'loss_graph': {'main_indices': [0],
'removed_features_count': [0, 1, 2, 3, 4, 5, 6],
'loss_values': [0.04293644412698623,
0.042180607283099525,
0.04153902780042639,
0.041193942305112335,
0.04118253992855187,
0.04134915176067319,
0.04159032536730382]},
'eliminated_features': [0, 6, 9, 10],
'selected_features_names': ['1', '2', '3', '4', '5', '8', '9']}
[10]:
new_df = feat_sel.transform(df)
new_df.head()
[10]:
| 1 | 2 | 3 | 4 | 5 | 8 | 9 | 6 | |
|---|---|---|---|---|---|---|---|---|
| 0 | -2.445094 | 2.341241 | 2.231165 | 4.678349 | -3.700149 | -2.417959 | val0_2 | 1 |
| 1 | 0.082871 | 2.923892 | 3.748558 | 0.181716 | -3.360609 | 0.067200 | val0_2 | 1 |
| 2 | 2.041848 | 3.359383 | 2.681346 | 3.295876 | 1.909756 | 2.059067 | val0_1 | 0 |
| 3 | 0.794927 | 1.881219 | 0.273481 | -0.339529 | 0.293166 | 0.798612 | val0_2 | 0 |
| 4 | -3.331591 | 1.779642 | 1.727541 | 2.620087 | -4.066566 | -3.330251 | val0_2 | 1 |
3 - Regression Task
So far, we only showed examples of the CatBoostSelection for classification tasks. However, this class also works for regression tasks. First, let’s create a dummy regression dataset for the next example. For this, we’ll use the create_dummy_dataset function:
[12]:
df = create_dummy_dataset(
samples=1000,
n_features=6,
n_num_num=2,
n_cat_num=2,
n_cat_cat=0,
num_num_noise=[0.01, 0.02],
pct_change=[0.03, 0.05],
regression=True,
)
df
[12]:
| num_0 | num_1 | num_2 | num_3 | num_4 | num_5 | label | num_c0_num_0 | num_c1_num_1 | CN_0_num_0 | CN_1_num_1 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | -0.284073 | 0.409715 | 0.098934 | 1.158437 | -0.866076 | 0.277924 | -34.411949 | -0.274862 | 0.387955 | val0_0 | val1_2 |
| 1 | -0.930755 | -0.060777 | -0.188225 | -0.986624 | -0.318273 | -1.689217 | -188.627154 | -0.936006 | -0.046597 | val0_0 | val1_4 |
| 2 | 0.060595 | -1.227406 | -0.689698 | -0.992912 | -0.896072 | -0.054733 | -219.906838 | 0.067076 | -1.249945 | val0_0 | val1_1 |
| 3 | -1.713446 | -1.456382 | -0.606355 | 0.437594 | -0.279864 | -0.560286 | -290.449938 | -1.728779 | -1.451134 | val0_0 | val1_0 |
| 4 | -2.226068 | 0.653653 | -0.305475 | 0.206521 | -0.691982 | -0.499834 | -170.448948 | -2.220170 | 0.650969 | val0_0 | val1_3 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 995 | -0.179003 | -0.352460 | -1.009324 | -0.666459 | -0.499371 | 0.778567 | -83.523775 | -0.171964 | -0.337933 | val0_0 | val1_1 |
| 996 | -1.673827 | -0.137832 | 0.444258 | -1.433685 | 0.952260 | 0.136158 | -21.425478 | -1.665045 | -0.131123 | val0_0 | val1_2 |
| 997 | 0.766691 | 1.228732 | 1.598219 | -0.084453 | 1.330511 | 1.274605 | 376.549377 | 0.754177 | 1.206221 | val0_1 | val1_3 |
| 998 | 0.524016 | -1.908487 | -1.079144 | 0.280277 | -0.914423 | -0.643964 | -274.319408 | 0.526594 | -1.916078 | val0_1 | val1_0 |
| 999 | -0.837132 | -0.052062 | -0.291399 | -1.239365 | 1.126793 | 1.438376 | 100.351058 | -0.833215 | -0.058532 | val0_0 | val1_2 |
1000 rows × 11 columns
The CatBoostSelection class will automatically detect if a problem is a classification or regression task by looking at the label column: if the data type of the label column is a variation of the float data type, then the task is considered to be a regression. Otherwise, it is considered a classification task. Note that we can explicitly determine if we want to solve a classification or a regression task by setting the ‘regression’ parameter when instantiating the CatBoostSelection class: if the ‘regression’ parameter is set to True, then the task will be considered a regression task, and if set to False, it will be treated as a classification task. The default value of this parameter is None, and in this case, the task will be determined by looking at the data type of the label column, as previously mentioned. If we have a classification problem, but the label column is set with float values (1.0 for class 1, 2.0 for class 2, and so on), then we must set the ‘regression’ parameter to True.
[13]:
feat_sel = CatBoostSelection(verbose=False)
feat_sel.fit(df=df, label_col="label")
feat_sel.get_selected_features()
[13]:
['num_0',
'num_1',
'num_2',
'num_3',
'num_4',
'num_5',
'num_c0_num_0',
'num_c1_num_1',
'CN_0_num_0',
'CN_1_num_1']
The internal variable ‘regression’ will indicate if the task is a regression or a classification:
[14]:
feat_sel.regression
[14]:
True