multiple_linear_regression/dn_scale_norm.py at main · alumia-code/multiple_linear_regression · 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
import numpy as np

class Normalizer():
    """A scaler that performs normalization to given range.
    Deals with multiple feature/multiple parameter arrays
    Expects a 3D array
    """
    def __init__(self):
        self.parameters = list() #stores max and min for each feature prior to transformation
        self.n_examples = 0
        self.n_features = 0
        self.norm_range = list()

    def fit(self, X: np.array, norm_range: list()):

        parameters = list()
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        if len(X_1.shape) > 1:
            n_examples = X_1.shape[0]
            n_features = X_1.shape[1]
        else:
            n_examples = 1
            n_features = X_1.shape[-1]

        #find maximum and minimum for each column, store it and perform transformation
        for i in range(n_features):
            temp_max = np.max(X_1[:, i:i+1], axis=None)
            temp_min = np.min(X_1[:, i:i+1],axis=None)
            parameters.append([temp_min, temp_max])

        self.n_examples = n_examples
        self.n_features = n_features
        self.parameters = parameters
        self.norm_range = norm_range

        return parameters

    def transform(self, X: np.array):
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        for i in range(self.n_features):
            X_1[:,i:i+1] = ((X_1[:, i:i+1] - self.parameters[i][0])/(self.parameters[i][1]-self.parameters[i][0]))*(self.norm_range[1]-self.norm_range[0])+self.norm_range[0]

        return X_1

    def inverse_transform(self, X: np.array):
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        for i in range(self.n_features):
            X_1[:,i:i+1] = ((X_1[:, i:i+1] - self.norm_range[0])/(self.norm_range[1]-self.norm_range[0]))*(self.parameters[i][1]-self.parameters[i][0])+self.parameters[i][0]

        return X_1

class MeanScaler():
    """A scaler that performs mean normalization.
    Deals with multiple feature/multiple parameter arrays
    Expects a 3D array
    """
    def __init__(self):
        self.parameters = list() #stores max and min for each feature prior to transformation
        self.n_examples = 0
        self.n_features = 0

    def fit(self, X: np.array):

        parameters = list()
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        if len(X_1.shape) > 1:
            n_examples = X_1.shape[0]
            n_features = X_1.shape[1]
        else:
            n_examples = 1
            n_features = X_1.shape[-1]

        #find maximum and minimum for each column, store it and perform transformation
        for i in range(n_features):
            temp_max = np.max(X_1[:, i:i+1], axis=None)
            temp_min = np.min(X_1[:, i:i+1],axis=None)
            temp_mean = np.mean(X_1[:, i:i+1],axis=None)
            parameters.append([temp_min, temp_max, temp_mean])

        self.n_examples = n_examples
        self.n_features = n_features
        self.parameters = parameters

        return parameters

    def transform(self, X: np.array):
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        for i in range(self.n_features):
            X_1[:,i:i+1] = (X_1[:, i:i+1] - self.parameters[i][2])/(self.parameters[i][1]-self.parameters[i][0])

        return X_1

    def inverse_transform(self, X: np.array):
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        for i in range(self.n_features):
            X_1[:,i:i+1] = X_1[:, i:i+1]*(self.parameters[i][1]-self.parameters[i][0])+self.parameters[i][2]

        return X_1

class Zscore():

    """A scaler that performs mean normalization.
    Deals with multiple feature/multiple parameter arrays
    Expects a 3D array
    """
    def __init__(self):
        self.parameters = list() #stores max and min for each feature prior to transformation
        self.n_examples = 0
        self.n_features = 0

    def fit(self, X: np.array):

        parameters = list()
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        if len(X_1.shape) > 1:
            n_examples = X_1.shape[0]
            n_features = X_1.shape[1]
        else:
            n_examples = 1
            n_features = X_1.shape[-1]

        #find maximum and minimum for each column, store it and perform transformation
        for i in range(n_features):
            temp_max = np.max(X_1[:, i:i+1], axis=None)
            temp_min = np.min(X_1[:, i:i+1],axis=None)
            temp_mean = np.mean(X_1[:, i:i+1],axis=None)
            temp_std = np.std(X_1[:, i:i+1],axis=None)
            parameters.append([temp_min, temp_max, temp_mean, temp_std])

        self.n_examples = n_examples
        self.n_features = n_features
        self.parameters = parameters

        return parameters

    def transform(self, X: np.array):
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        for i in range(self.n_features):
            X_1[:,i:i+1] = (X_1[:, i:i+1] - self.parameters[i][2])/(self.parameters[i][3])

        return X_1

    def inverse_transform(self, X: np.array):
        X_1 = X.copy()
        X_1 = X_1.astype(np.float32)

        for i in range(self.n_features):
            X_1[:,i:i+1] = X_1[:, i:i+1]*self.parameters[i][3]+self.parameters[i][2]

        return X_1