mojolearn.MaxAbsScaler#

class mojolearn.MaxAbsScaler(*, copy=True)[source]#

Bases: _PrepBase

Scale each feature by its maximum absolute value.

Follows scikit-learn’s MaxAbsScaler: each column is divided by the largest absolute value among its non-NaN entries, so the result lies in [-1, 1]. NaN values are kept as NaN by transform. A column of zeros is scaled by one. Input and output are float32.

Parameters:

copy (bool, default True) – Accepted for scikit-learn compatibility; transform always returns a new array.

Variables:

  • max_abs_ (array of shape (n_features,)) – Per-column maximum absolute value.

  • scale_ (array of shape (n_features,)) – Per-column divisor (max_abs_, with zeros replaced by one).

  • n_features_in_ (int) – Number of columns seen in fit.

  • n_samples_seen_ (int) – Number of rows seen in fit.

  • numeric_mode_ (str) – Numeric mode the estimator was fitted in; transform runs in the same mode.

Examples

>>> import numpy as np, mojolearn
>>> X = np.random.default_rng(0).standard_normal((1000, 8)).astype(np.float32)
>>> Xt = mojolearn.MaxAbsScaler().fit_transform(X)
__init__(*, copy=True)[source]#

Methods

fit(X[, y])

fit_transform(X[, y])

Fit to X, then transform X.

get_params([deep])

Return the constructor parameters as a dict.

set_params(**params)

Set constructor parameters.

transform(X)