automl-RENOIR/experiment3.py at master · JakubKlik/automl-RENOIR · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
from datetime import datetime

import sklearn
import autosklearn
import autosklearn.classification

from sklearn.datasets import make_blobs, make_classification
from sklearn.model_selection import train_test_split

from classifiers import BernoulliNB_S
from classifiers import DecisionTree_S
from classifiers import GaussianNB_S
from classifiers import KNearestNeighborsClassifier_S
from classifiers import LibLinear_SVC_S
from classifiers import LibSVM_SVC_S
from classifiers import RandomForest_S
from classifiers import SGD_S

from utils import load_data

import os
import warnings
warnings.simplefilter("ignore")

autosklearn.pipeline.components.classification.add_classifier(BernoulliNB_S)
autosklearn.pipeline.components.classification.add_classifier(DecisionTree_S)
autosklearn.pipeline.components.classification.add_classifier(GaussianNB_S)
autosklearn.pipeline.components.classification.add_classifier(KNearestNeighborsClassifier_S)
autosklearn.pipeline.components.classification.add_classifier(LibLinear_SVC_S)
autosklearn.pipeline.components.classification.add_classifier(LibSVM_SVC_S)
autosklearn.pipeline.components.classification.add_classifier(RandomForest_S)
autosklearn.pipeline.components.classification.add_classifier(SGD_S)

datasets = [
"abalone-17_vs_7-8-9-10",
"page-blocks0",
"yeast-0-2-5-6_vs_3-7-8-9",
"kc1",
"vowel0",
"pima"
]
for dataset in datasets:

    X, y, c = load_data(dataset)
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5, shuffle=True, random_state=1111)


    # now = datetime.now()
    # name = now.strftime("%d-%m-%Y_%H:%M:%S")
    name = dataset

    all_run_time = 60*60
    run_time = int(all_run_time*0.05)
    # run_time = 15

    if not os.path.exists("./logs/exp3_%s" % name):
        os.makedirs("./logs/exp3_%s" % name)

    clfs = [
    "BernoulliNB_S",
    "DecisionTree_S",
    "GaussianNB_S",
    "KNearestNeighborsClassifier_S",
    "LibLinear_SVC_S",
    "LibSVM_SVC_S",
    "RandomForest_S",
    "SGD_S"
    ]

    aclf = autosklearn.classification.AutoSklearnClassifier(
        time_left_for_this_task=all_run_time,
        per_run_time_limit=run_time,
        ml_memory_limit=4000,
        ensemble_size=10,
        ensemble_nbest=10,
        ensemble_memory_limit=4000,
        include_estimators=clfs,
        tmp_folder="./logs/exp3_%s/tmp_sampling" % name,
        delete_tmp_folder_after_terminate=False,
        exclude_preprocessors=['balancing','pca'],
        resampling_strategy='cv',
    	resampling_strategy_arguments={'folds': 5},
    )
    aclf.fit(X_train.copy(), y_train.copy(), dataset_name=dataset, metric=autosklearn.metrics.f1)

    aclf.refit(X_train.copy(), y_train.copy())

    y_pred = aclf.predict(X_test)

    with open("./logs/exp3_%s/logfile_sampling.txt" % name, 'at') as logfile:
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  METRICS  ------------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("ACC: %0.5f" % autosklearn.metrics.accuracy(y_pred, y_test), file=logfile)
        print("F-1: %0.5f" % autosklearn.metrics.f1(y_pred, y_test), file=logfile)
        print("PRC: %0.5f" % autosklearn.metrics.precision(y_pred, y_test), file=logfile)
        print("REC: %0.5f" % autosklearn.metrics.recall(y_pred, y_test), file=logfile)
        print("BAC: %0.5f" % autosklearn.metrics.balanced_accuracy(y_pred, y_test), file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  MODELS  -------------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print(aclf.show_models(), file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  CV RESULTS  ---------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print(aclf.cv_results_, file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  STATS  --------------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print(aclf.sprint_statistics(), file=logfile)


    clfs = [
    # "adaboost",
    "bernoulli_nb",
    "decision_tree",
    # "extra_trees",
    "gaussian_nb",
    # "gradient_boosting",
    "k_nearest_neighbors",
    # "lda",
    "liblinear_svc",
    "libsvm_svc",
    # "multinomial_nb",
    # "passive_aggressive",
    # "qda",
    "random_forest",
    "sgd",
    ]


    aclf = autosklearn.classification.AutoSklearnClassifier(
        time_left_for_this_task=all_run_time,
        per_run_time_limit=run_time,
        ml_memory_limit=4000,
        ensemble_size=10,
        ensemble_nbest=10,
        ensemble_memory_limit=4000,
        include_estimators=clfs,
        tmp_folder="./logs/exp3_%s/tmp_normal" % name,
        delete_tmp_folder_after_terminate=False,
        exclude_preprocessors=['balancing','pca'],
        resampling_strategy='cv',
    	resampling_strategy_arguments={'folds': 5},
    )
    aclf.fit(X_train.copy(), y_train.copy(), dataset_name=dataset, metric=autosklearn.metrics.f1)

    aclf.refit(X_train.copy(), y_train.copy())

    y_pred = aclf.predict(X_test)

    with open("./logs/exp3_%s/logfile_normal.txt" % name, 'at') as logfile:
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  METRICS  ------------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("ACC: %0.5f" % autosklearn.metrics.accuracy(y_pred, y_test), file=logfile)
        print("F-1: %0.5f" % autosklearn.metrics.f1(y_pred, y_test), file=logfile)
        print("PRC: %0.5f" % autosklearn.metrics.precision(y_pred, y_test), file=logfile)
        print("REC: %0.5f" % autosklearn.metrics.recall(y_pred, y_test), file=logfile)
        print("BAC: %0.5f" % autosklearn.metrics.balanced_accuracy(y_pred, y_test), file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  MODELS  -------------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print(aclf.show_models(), file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  CV RESULTS  ---------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print(aclf.cv_results_, file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print("---  STATS  --------------------------------------------------------------------------------------------------------", file=logfile)
        print("--------------------------------------------------------------------------------------------------------------------", file=logfile)
        print(aclf.sprint_statistics(), file=logfile)