NEFCLASS/data_loading.py at main · dhecloud/NEFCLASS · 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
import os
import pandas as pd
import random
import numpy as  np

def shuffle_idxs(length):
    tmp =  list(range(length))
    random.shuffle(tmp)
    return tmp


def load_iris(args, path='data/iris.csv'):
    data = pd.read_csv(path)
    class_mapping = {}
    for c in data['Species'].unique():
        class_mapping[c] = len(class_mapping)
    data['class'] = [class_mapping[c] for c in data['Species']]

    targets = data['class'].to_numpy()
    data = data[['SepalLengthCm','SepalWidthCm','PetalWidthCm']].to_numpy()
    vars(args)['num_input_units'] = data.shape[1]
    vars(args)['output_units'] = len(class_mapping)

    if args.cv:
        return data, targets, np.max(data,axis=0), np.min(data,axis=0)

    else:
        train_data, train_targets = data[0::2], targets[0::2]
        test_data, test_targets = data[1::2], targets[1::2]
        return train_data, train_targets, test_data, test_targets, np.max(data,axis=0), np.min(data,axis=0)


def load_wine(args,path ='data/wine/'):
    feature_cols = ['fixed acidity', 'volatile acidity', 'citric acid', 'residual sugar', \
    'chlorides', 'free sulfur dioxide', 'total sulfur dioxide', 'density', 'pH', 'sulphates', 'alcohol']
    target_col = ['quality']

    red_data = pd.read_csv(os.path.join(path,'winequality-red.csv'), sep=';')
    white_data = pd.read_csv(os.path.join(path,'winequality-white.csv'),sep=';')
    data = red_data.append(white_data)
    classes = []
    for c in data['quality']:
        if c <= 4: #poor
            classes.append(0)
        elif c <= 6: #average
            classes.append(1)
        else:
            classes.append(2)
    data['class'] = classes

    targets = data['class'].to_numpy()
    data = data[feature_cols].to_numpy()
    vars(args)['num_input_units'] = data.shape[1]
    vars(args)['output_units'] = 3 #poor average excellent

    if args.cv:
        return data, targets, np.max(data,axis=0), np.min(data,axis=0)

    else:
        train_data, train_targets = data[0::2], targets[0::2]
        test_data, test_targets = data[1::2], targets[1::2]
        return train_data, train_targets, test_data, test_targets, np.max(data,axis=0), np.min(data,axis=0)


def load_breast_cancer(args,path ='data/breast_cancer/'):
    train_data = pd.read_csv(os.path.join(path, 'train_data.csv'))
    test_data = pd.read_csv(os.path.join(path, 'test_data.csv'))

    class_mapping = {}
    for c in train_data['Class'].unique():
        class_mapping[c] = len(class_mapping)
    train_data['class_mapped'] = [class_mapping[c] for c in train_data['Class']]
    test_data['class_mapped'] = [class_mapping[c] for c in test_data['Class']]
    vars(args)['num_input_units'] = train_data.shape[1]
    vars(args)['output_units'] = len(class_mapping)

    train_targets = train_data['class_mapped'].to_numpy()
    train_data = train_data.drop(['class_mapped', 'Class'], axis=1).to_numpy()
    test_targets = test_data['class_mapped'].to_numpy()
    test_data = test_data.drop(['class_mapped', 'Class'], axis=1).to_numpy()


    assert train_data.shape[0] == train_targets.shape[0]
    assert test_data.shape[0] == test_targets.shape[0]

    vars(args)['num_input_units'] = train_data.shape[1]
    vars(args)['output_units'] = len(class_mapping)

    return train_data, train_targets, test_data, test_targets, np.max(train_data,axis=0), np.min(train_data,axis=0)

def load_breast_cancer_wisconsin(args,path ='data/breast_cancer_wisconsin/'):

    data = pd.read_csv(os.path.join(path, 'breast-cancer-wisconsin.data'),\
                    header=None)
    feature_cols =['Clump Thickness', 'Uniformity of Cell Size', \
     'Uniformity of Cell Shape', 'Marginal Adhesion', \
     'Single Epithelial Cell Size', 'Bare Nuclei', 'Bland Chromatin',
      'Normal Nucleoli', 'Mitoses']
    data.columns = ['id', *feature_cols, 'Class']

    class_mapping = {}
    for c in data['Class'].unique():
        class_mapping[c] = len(class_mapping)
    data['class_mapped'] = [class_mapping[c] for c in data['Class']]

    for c in feature_cols:
        data[c] = data[c].replace('?', '-9999999')
        max = data[c].max()

        data[c]= data[c].replace('-9999999', max)


    data[feature_cols] = data[feature_cols].astype(float)
    vars(args)['num_input_units'] = data.shape[1]
    vars(args)['output_units'] = len(class_mapping)

    if args.cv:
        targets = data['class_mapped'].to_numpy()
        data = data[feature_cols].to_numpy()

        return data, targets, np.max(data,axis=0), np.min(data,axis=0)

    else:
        train_idxs = random.sample(list(range(data.shape[0])), k=int(9*data.shape[0]/10))
        test_idxs = [x not in train_idxs for x in list(range(data.shape[0])) ]
        train_data, train_targets = data.loc[train_idxs, feature_cols].to_numpy(), data.loc[train_idxs, 'class_mapped'].to_numpy()
        test_data, test_targets = data.loc[test_idxs, feature_cols].to_numpy(), data.loc[test_idxs, 'class_mapped'].to_numpy()


        assert train_data.shape[0] == train_targets.shape[0]
        assert test_data.shape[0] == test_targets.shape[0]

        return train_data, train_targets, test_data, test_targets, np.max(data,axis=0), np.min(data,axis=0)


if __name__ == '__main__':
    load_breast_cancer()