generalized_additive_models.Linear#

class generalized_additive_models.Linear(feature=None, *, penalty=1, by=None, constraint=None)#

A linear term.

Examples

>>> linear = Linear(feature=0)
>>> linear.num_coefficients
1

Fitting and transforming a Linear term extracts the relevant column:

>>> import numpy as np
>>> X = np.arange(24).reshape(8, 3)
>>> linear.fit_transform(X)
array([[ 0],
       [ 3],
       [ 6],
       [ 9],
       [12],
       [15],
       [18],
       [21]])

Linear terms have a standard quadratic (l2) penalty:

>>> linear.penalty_matrix()
array([[1.]])

The square root of the penalty matrix is returned, so we must square it to get back the given penalty:

>>> Linear(feature=0, penalty=5).penalty_matrix()**2
array([[5.]])

It also works for pandas DataFrames:

>>> import pandas as pd
>>> df = pd.DataFrame({'a':[1, 2, 3], 'b':[4, 5, 6]})
>>> Linear('b').fit_transform(df)
array([[4],
       [5],
       [6]])
>>> Linear('b', by='a').fit_transform(df)
array([[ 4],
       [10],
       [18]])
>>> Linear('a', by='b').fit_transform(df)
array([[ 4],
       [10],
       [18]])

__init__(feature=None, *, penalty=1, by=None, constraint=None)#

Initialize a Linear term.

Parameters:

feature (str or int, optional) – The feature name or index associated with the term. The default is None.
penalty (float, optional) – Regularization penalty. The default is 1.
by (str or int, optional) – The feature name or index associated with a multiplicative term. The default is None.
constraint (str or None, optional) – Either None, or increasing or decreasing. The default is None.

Examples

>>> linear_term = Linear(0, penalty=2)
>>> linear_term
Linear(feature=0, penalty=2)
>>> Linear(0, penalty=2) == Linear(0, penalty=2)
True
>>> Linear(0, penalty=2) == Linear(0, penalty=3)
False
>>> Linear(0, penalty=2).is_redundant_with_respect_to(Linear(0, penalty=3))
True

Methods

`__init__`([feature, penalty, by, constraint])	Initialize a Linear term.
`fit`(X)	Fit to data.
`fit_transform`(X[, y])	Fit to data, then transform it.
`get_metadata_routing`()	Get metadata routing of this object.
`get_params`([deep])	Get parameters for this estimator.
`is_redundant_with_respect_to`(other)	Check if a Term is redundant with respect to another.
`penalty_matrix`()	Return the penalty matrix for the term.
`set_output`(*[, transform])	Set output container.
`set_params`(**params)	Set the parameters of this estimator.
`transform`(X)	Transform the input.

Attributes

`name`	Name of the term.
`num_coefficients`	Number of coefficients for the term.

fit(X)#

Fit to data.

Parameters:: X (np.ndarray or pd.DataFrame) – A dataset of shape (num_samples, num_features).

name = 'linear'#: Name of the term.

property num_coefficients#: Number of coefficients for the term.

penalty_matrix()#: Return the penalty matrix for the term.

transform(X)#

Transform the input.

Parameters:: X (np.ndarray) – An ndarray with 2 dimensions of shape (n_samples, n_features).
Returns:: X – An ndarray for the term.
Return type:: np.ndarray

Examples

>>> linear = Linear(1)
>>> X = np.eye(3)
>>> linear.fit_transform(X)
array([[0.],
       [1.],
       [0.]])