Skip to content

Preprocessing

AbstractNormalization

Bases: OneToOneFeatureMixin, TransformerMixin, BaseEstimator

Abstract base class for normalization techniques. All normalization classes should inherit from this class and implement the transform and inverse_transform methods.

Source code in asf/preprocessing/performace_scaling.py 
 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
class AbstractNormalization(OneToOneFeatureMixin, TransformerMixin, BaseEstimator):
    """
    Abstract base class for normalization techniques. All normalization classes
    should inherit from this class and implement the `transform` and `inverse_transform` methods.
    """

    def __init__(self) -> None:
        super().__init__()

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "AbstractNormalization":
        """
        Fit the normalization model to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            AbstractNormalization: The fitted normalization instance.
        """
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        raise NotImplementedError

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the input data.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        raise NotImplementedError

fit(X, y=None, sample_weight=None)

Fit the normalization model to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
AbstractNormalization AbstractNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "AbstractNormalization":
    """
    Fit the normalization model to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        AbstractNormalization: The fitted normalization instance.
    """
    return self

inverse_transform(X)

Inverse transform the input data.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
45
46
47
48
49
50
51
52
53
54
55
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the input data.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    raise NotImplementedError

transform(X)

Transform the input data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
33
34
35
36
37
38
39
40
41
42
43
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    raise NotImplementedError

BoxCoxNormalization

Bases: AbstractNormalization

Normalization using Box-Cox transformation.

Source code in asf/preprocessing/performace_scaling.py 
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
class BoxCoxNormalization(AbstractNormalization):
    """
    Normalization using Box-Cox transformation.
    """

    def __init__(self) -> None:
        super().__init__()

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "BoxCoxNormalization":
        """
        Fit the Box-Cox transformer to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            BoxCoxNormalization: The fitted normalization instance.
        """
        self.box_cox = PowerTransformer(method="yeo-johnson")
        self.box_cox.fit(X.reshape(-1, 1))
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data using Box-Cox transformation.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        return self.box_cox.transform(X.reshape(-1, 1)).reshape(-1)

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data back to the original scale.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        X = self.box_cox.inverse_transform(X.reshape(-1, 1)).reshape(-1)
        return X

fit(X, y=None, sample_weight=None)

Fit the Box-Cox transformer to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
BoxCoxNormalization BoxCoxNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "BoxCoxNormalization":
    """
    Fit the Box-Cox transformer to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        BoxCoxNormalization: The fitted normalization instance.
    """
    self.box_cox = PowerTransformer(method="yeo-johnson")
    self.box_cox.fit(X.reshape(-1, 1))
    return self

inverse_transform(X)

Inverse transform the data back to the original scale.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
489
490
491
492
493
494
495
496
497
498
499
500
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data back to the original scale.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    X = self.box_cox.inverse_transform(X.reshape(-1, 1)).reshape(-1)
    return X

transform(X)

Transform the input data using Box-Cox transformation.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
477
478
479
480
481
482
483
484
485
486
487
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data using Box-Cox transformation.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    return self.box_cox.transform(X.reshape(-1, 1)).reshape(-1)

DummyNormalization

Bases: AbstractNormalization

Normalization that does not change the data.

Source code in asf/preprocessing/performace_scaling.py 
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
class DummyNormalization(AbstractNormalization):
    """
    Normalization that does not change the data.
    """

    def __init__(self) -> None:
        super().__init__()

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "DummyNormalization":
        """
        Fit the DummyNormalization model to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            DummyNormalization: The fitted normalization instance.
        """
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data (no change).

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        return X

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data (no change).

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        return X

fit(X, y=None, sample_weight=None)

Fit the DummyNormalization model to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
DummyNormalization DummyNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "DummyNormalization":
    """
    Fit the DummyNormalization model to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        DummyNormalization: The fitted normalization instance.
    """
    return self

inverse_transform(X)

Inverse transform the data (no change).

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
438
439
440
441
442
443
444
445
446
447
448
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data (no change).

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    return X

transform(X)

Transform the input data (no change).

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
426
427
428
429
430
431
432
433
434
435
436
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data (no change).

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    return X

InvSigmoidNormalization

Bases: AbstractNormalization

Normalization using inverse sigmoid scaling.

Source code in asf/preprocessing/performace_scaling.py 
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
class InvSigmoidNormalization(AbstractNormalization):
    """
    Normalization using inverse sigmoid scaling.
    """

    def __init__(self) -> None:
        super().__init__()

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "InvSigmoidNormalization":
        """
        Fit the InvSigmoidNormalization model to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            InvSigmoidNormalization: The fitted normalization instance.
        """
        self.min_max_scale = MinMaxScaler(feature_range=(1e-6, 1 - 1e-6))
        self.min_max_scale.fit(X)
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data using inverse sigmoid scaling.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        X = self.min_max_scale.transform(X.reshape(-1, 1)).reshape(-1)
        return np.log(X / (1 - X))

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data back to the original scale.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        X = scipy.special.expit(X)
        return self.min_max_scale.inverse_transform(X.reshape(-1, 1)).reshape(-1)

fit(X, y=None, sample_weight=None)

Fit the InvSigmoidNormalization model to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
InvSigmoidNormalization InvSigmoidNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "InvSigmoidNormalization":
    """
    Fit the InvSigmoidNormalization model to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        InvSigmoidNormalization: The fitted normalization instance.
    """
    self.min_max_scale = MinMaxScaler(feature_range=(1e-6, 1 - 1e-6))
    self.min_max_scale.fit(X)
    return self

inverse_transform(X)

Inverse transform the data back to the original scale.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
339
340
341
342
343
344
345
346
347
348
349
350
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data back to the original scale.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    X = scipy.special.expit(X)
    return self.min_max_scale.inverse_transform(X.reshape(-1, 1)).reshape(-1)

transform(X)

Transform the input data using inverse sigmoid scaling.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
326
327
328
329
330
331
332
333
334
335
336
337
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data using inverse sigmoid scaling.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    X = self.min_max_scale.transform(X.reshape(-1, 1)).reshape(-1)
    return np.log(X / (1 - X))

LogNormalization

Bases: AbstractNormalization

Normalization using logarithmic scaling.

Source code in asf/preprocessing/performace_scaling.py 
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
class LogNormalization(AbstractNormalization):
    """
    Normalization using logarithmic scaling.
    """

    def __init__(self, base: float = 10, eps: float = 1e-6) -> None:
        """
        Initialize LogNormalization.

        Args:
            base (float, optional): Base of the logarithm. Defaults to 10.
            eps (float, optional): Small constant to avoid log(0). Defaults to 1e-6.
        """
        super().__init__()
        self.base = base
        self.eps = eps

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "LogNormalization":
        """
        Fit the LogNormalization model to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            LogNormalization: The fitted normalization instance.
        """
        if X.min() <= 0:
            self.min_val = X.min()
        else:
            self.min_val = 0
            self.eps = 0

        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data using logarithmic scaling.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        X = X - self.min_val + self.eps
        return np.log(X) / np.log(self.base)

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data back to the original scale.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        X = np.power(self.base, X)
        if self.min_val != 0:
            X = X + self.min_val - self.eps
        return X

__init__(base=10, eps=1e-06)

Initialize LogNormalization.

Parameters:

Name Type Description Default
base float

Base of the logarithm. Defaults to 10.

 10
eps float

Small constant to avoid log(0). Defaults to 1e-6.

 1e-06
Source code in asf/preprocessing/performace_scaling.py 
172
173
174
175
176
177
178
179
180
181
182
def __init__(self, base: float = 10, eps: float = 1e-6) -> None:
    """
    Initialize LogNormalization.

    Args:
        base (float, optional): Base of the logarithm. Defaults to 10.
        eps (float, optional): Small constant to avoid log(0). Defaults to 1e-6.
    """
    super().__init__()
    self.base = base
    self.eps = eps

fit(X, y=None, sample_weight=None)

Fit the LogNormalization model to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
LogNormalization LogNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "LogNormalization":
    """
    Fit the LogNormalization model to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        LogNormalization: The fitted normalization instance.
    """
    if X.min() <= 0:
        self.min_val = X.min()
    else:
        self.min_val = 0
        self.eps = 0

    return self

inverse_transform(X)

Inverse transform the data back to the original scale.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
219
220
221
222
223
224
225
226
227
228
229
230
231
232
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data back to the original scale.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    X = np.power(self.base, X)
    if self.min_val != 0:
        X = X + self.min_val - self.eps
    return X

transform(X)

Transform the input data using logarithmic scaling.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
206
207
208
209
210
211
212
213
214
215
216
217
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data using logarithmic scaling.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    X = X - self.min_val + self.eps
    return np.log(X) / np.log(self.base)

MinMaxNormalization

Bases: AbstractNormalization

Normalization using Min-Max scaling.

Source code in asf/preprocessing/performace_scaling.py 
 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
class MinMaxNormalization(AbstractNormalization):
    """
    Normalization using Min-Max scaling.
    """

    def __init__(self, feature_range: tuple[float, float] = (0, 1)) -> None:
        """
        Initialize MinMaxNormalization.

        Args:
            feature_range (tuple[float, float], optional): Desired range of transformed data. Defaults to (0, 1).
        """
        super().__init__()
        self.feature_range = feature_range

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "MinMaxNormalization":
        """
        Fit the Min-Max scaler to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            MinMaxNormalization: The fitted normalization instance.
        """
        self.min_max_scale = MinMaxScaler(feature_range=self.feature_range)
        self.min_max_scale.fit(X.reshape(-1, 1))
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data using Min-Max scaling.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        return self.min_max_scale.transform(X.reshape(-1, 1)).reshape(-1)

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data back to the original scale.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        return self.min_max_scale.inverse_transform(X.reshape(-1, 1)).reshape(-1)

__init__(feature_range=(0, 1))

Initialize MinMaxNormalization.

Parameters:

Name Type Description Default
feature_range tuple[float, float]

Desired range of transformed data. Defaults to (0, 1).

 (0, 1)
Source code in asf/preprocessing/performace_scaling.py 
63
64
65
66
67
68
69
70
71
def __init__(self, feature_range: tuple[float, float] = (0, 1)) -> None:
    """
    Initialize MinMaxNormalization.

    Args:
        feature_range (tuple[float, float], optional): Desired range of transformed data. Defaults to (0, 1).
    """
    super().__init__()
    self.feature_range = feature_range

fit(X, y=None, sample_weight=None)

Fit the Min-Max scaler to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
MinMaxNormalization MinMaxNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "MinMaxNormalization":
    """
    Fit the Min-Max scaler to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        MinMaxNormalization: The fitted normalization instance.
    """
    self.min_max_scale = MinMaxScaler(feature_range=self.feature_range)
    self.min_max_scale.fit(X.reshape(-1, 1))
    return self

inverse_transform(X)

Inverse transform the data back to the original scale.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
103
104
105
106
107
108
109
110
111
112
113
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data back to the original scale.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    return self.min_max_scale.inverse_transform(X.reshape(-1, 1)).reshape(-1)

transform(X)

Transform the input data using Min-Max scaling.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data using Min-Max scaling.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    return self.min_max_scale.transform(X.reshape(-1, 1)).reshape(-1)

NegExpNormalization

Bases: AbstractNormalization

Normalization using negative exponential scaling.

Source code in asf/preprocessing/performace_scaling.py 
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
class NegExpNormalization(AbstractNormalization):
    """
    Normalization using negative exponential scaling.
    """

    def __init__(self) -> None:
        super().__init__()

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "NegExpNormalization":
        """
        Fit the NegExpNormalization model to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            NegExpNormalization: The fitted normalization instance.
        """
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data using negative exponential scaling.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        return np.exp(-X)

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data back to the original scale.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        return -np.log(X)

fit(X, y=None, sample_weight=None)

Fit the NegExpNormalization model to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
NegExpNormalization NegExpNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "NegExpNormalization":
    """
    Fit the NegExpNormalization model to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        NegExpNormalization: The fitted normalization instance.
    """
    return self

inverse_transform(X)

Inverse transform the data back to the original scale.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
389
390
391
392
393
394
395
396
397
398
399
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data back to the original scale.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    return -np.log(X)

transform(X)

Transform the input data using negative exponential scaling.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
377
378
379
380
381
382
383
384
385
386
387
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data using negative exponential scaling.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    return np.exp(-X)

SqrtNormalization

Bases: AbstractNormalization

Normalization using square root scaling.

Source code in asf/preprocessing/performace_scaling.py 
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
class SqrtNormalization(AbstractNormalization):
    """
    Normalization using square root scaling.
    """

    def __init__(self, eps: float = 1e-6) -> None:
        """
        Initialize SqrtNormalization.

        Args:
            eps (float, optional): Small constant to avoid sqrt(0). Defaults to 1e-6.
        """
        super().__init__()
        self.eps = eps

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "SqrtNormalization":
        """
        Fit the SqrtNormalization model to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            SqrtNormalization: The fitted normalization instance.
        """
        if X.min() < 0:
            self.min_val = X.min()
            X = X + self.min_val + self.eps
        else:
            self.min_val = 0
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data using square root scaling.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        X = X + self.min_val + self.eps
        return np.sqrt(X)

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data back to the original scale.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        X = np.power(X, 2)
        if self.min_val != 0:
            X = X - self.min_val - self.eps
        return X

__init__(eps=1e-06)

Initialize SqrtNormalization.

Parameters:

Name Type Description Default
eps float

Small constant to avoid sqrt(0). Defaults to 1e-6.

 1e-06
Source code in asf/preprocessing/performace_scaling.py 
240
241
242
243
244
245
246
247
248
def __init__(self, eps: float = 1e-6) -> None:
    """
    Initialize SqrtNormalization.

    Args:
        eps (float, optional): Small constant to avoid sqrt(0). Defaults to 1e-6.
    """
    super().__init__()
    self.eps = eps

fit(X, y=None, sample_weight=None)

Fit the SqrtNormalization model to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
SqrtNormalization SqrtNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "SqrtNormalization":
    """
    Fit the SqrtNormalization model to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        SqrtNormalization: The fitted normalization instance.
    """
    if X.min() < 0:
        self.min_val = X.min()
        X = X + self.min_val + self.eps
    else:
        self.min_val = 0
    return self

inverse_transform(X)

Inverse transform the data back to the original scale.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
284
285
286
287
288
289
290
291
292
293
294
295
296
297
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data back to the original scale.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    X = np.power(X, 2)
    if self.min_val != 0:
        X = X - self.min_val - self.eps
    return X

transform(X)

Transform the input data using square root scaling.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
271
272
273
274
275
276
277
278
279
280
281
282
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data using square root scaling.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    X = X + self.min_val + self.eps
    return np.sqrt(X)

ZScoreNormalization

Bases: AbstractNormalization

Normalization using Z-Score scaling.

Source code in asf/preprocessing/performace_scaling.py 
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
class ZScoreNormalization(AbstractNormalization):
    """
    Normalization using Z-Score scaling.
    """

    def __init__(self) -> None:
        super().__init__()

    def fit(
        self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
    ) -> "ZScoreNormalization":
        """
        Fit the Z-Score scaler to the data.

        Args:
            X (np.ndarray): Input data.
            y (np.ndarray, optional): Target values. Defaults to None.
            sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

        Returns:
            ZScoreNormalization: The fitted normalization instance.
        """
        self.scaler = StandardScaler()
        self.scaler.fit(X.reshape(-1, 1))
        return self

    def transform(self, X: np.ndarray) -> np.ndarray:
        """
        Transform the input data using Z-Score scaling.

        Args:
            X (np.ndarray): Input data.

        Returns:
            np.ndarray: Transformed data.
        """
        return self.scaler.transform(X.reshape(-1, 1)).reshape(-1)

    def inverse_transform(self, X: np.ndarray) -> np.ndarray:
        """
        Inverse transform the data back to the original scale.

        Args:
            X (np.ndarray): Transformed data.

        Returns:
            np.ndarray: Original data.
        """
        return self.scaler.inverse_transform(X.reshape(-1, 1)).reshape(-1)

fit(X, y=None, sample_weight=None)

Fit the Z-Score scaler to the data.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required
y ndarray

Target values. Defaults to None.

 None
sample_weight ndarray

Sample weights. Defaults to None.

 None

Returns:

Name Type Description
ZScoreNormalization ZScoreNormalization

The fitted normalization instance.
Source code in asf/preprocessing/performace_scaling.py 
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
def fit(
    self, X: np.ndarray, y: np.ndarray = None, sample_weight: np.ndarray = None
) -> "ZScoreNormalization":
    """
    Fit the Z-Score scaler to the data.

    Args:
        X (np.ndarray): Input data.
        y (np.ndarray, optional): Target values. Defaults to None.
        sample_weight (np.ndarray, optional): Sample weights. Defaults to None.

    Returns:
        ZScoreNormalization: The fitted normalization instance.
    """
    self.scaler = StandardScaler()
    self.scaler.fit(X.reshape(-1, 1))
    return self

inverse_transform(X)

Inverse transform the data back to the original scale.

Parameters:

Name Type Description Default
X ndarray

Transformed data.

 required

Returns:

Type Description
ndarray

np.ndarray: Original data.
Source code in asf/preprocessing/performace_scaling.py 
154
155
156
157
158
159
160
161
162
163
164
def inverse_transform(self, X: np.ndarray) -> np.ndarray:
    """
    Inverse transform the data back to the original scale.

    Args:
        X (np.ndarray): Transformed data.

    Returns:
        np.ndarray: Original data.
    """
    return self.scaler.inverse_transform(X.reshape(-1, 1)).reshape(-1)

transform(X)

Transform the input data using Z-Score scaling.

Parameters:

Name Type Description Default
X ndarray

Input data.

 required

Returns:

Type Description
ndarray

np.ndarray: Transformed data.
Source code in asf/preprocessing/performace_scaling.py 
142
143
144
145
146
147
148
149
150
151
152
def transform(self, X: np.ndarray) -> np.ndarray:
    """
    Transform the input data using Z-Score scaling.

    Args:
        X (np.ndarray): Input data.

    Returns:
        np.ndarray: Transformed data.
    """
    return self.scaler.transform(X.reshape(-1, 1)).reshape(-1)

get_default_preprocessor(categorical_features=None, numerical_features=None)

Creates a default preprocessor for handling categorical and numerical features.

Parameters:

Name Type Description Default
categorical_features Optional[Union[List[str], Callable]]

List of categorical feature names or a callable selector. Defaults to selecting object dtype columns.

 None
numerical_features Optional[Union[List[str], Callable]]

List of numerical feature names or a callable selector. Defaults to selecting numeric dtype columns.

 None

Returns:

Name Type Description
ColumnTransformer ColumnTransformer

A transformer that applies preprocessing pipelines to categorical and numerical features.
Source code in asf/preprocessing/sklearn_preprocessor.py 
 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
def get_default_preprocessor(
    categorical_features: Optional[Union[List[str], Callable]] = None,
    numerical_features: Optional[Union[List[str], Callable]] = None,
) -> ColumnTransformer:
    """
    Creates a default preprocessor for handling categorical and numerical features.

    Args:
        categorical_features (Optional[Union[List[str], Callable]]):
            List of categorical feature names or a callable selector. Defaults to selecting object dtype columns.
        numerical_features (Optional[Union[List[str], Callable]]):
            List of numerical feature names or a callable selector. Defaults to selecting numeric dtype columns.

    Returns:
        ColumnTransformer: A transformer that applies preprocessing pipelines to categorical and numerical features.
    """
    if categorical_features is None:
        categorical_features = make_column_selector(dtype_include=object)

    if numerical_features is None:
        numerical_features = make_column_selector(dtype_include="number")

    preprocessor = ColumnTransformer(
        [
            (
                "cat",
                make_pipeline(
                    SimpleImputer(strategy="most_frequent"),
                    OneHotEncoder(sparse_output=False, handle_unknown="ignore"),
                ),
                categorical_features,
            ),
            (
                "cont",
                make_pipeline(SimpleImputer(strategy="median"), StandardScaler()),
                numerical_features,
            ),
        ]
    )

    preprocessor.set_output(transform="pandas")
    return preprocessor