gemseo / mlearning / transformers / power

yeo_johnson module¶

A Yeo-Johnson power transformation.

Transform \(x\) as:

\[\begin{split}y = & ((x + 1)^\lambda - 1) / \lambda, \text{ for } x \geq 0, \lambda \neq 0 \\\\ & \log(x + 1), \text{ for }x \geq 0, \lambda = 0 \\\\ & -\frac{(1-x)^{2 - \lambda} - 1}{2-\lambda}, \text{for} x < 0, \lambda \neq 2\\\\ & -log(1-x), \text{ for } x < 0, \lambda = 2\end{split}\]

Dependence¶

This transformation algorithm relies on the PowerTransformer class of scikit-learn.

class gemseo.mlearning.transformers.power.yeo_johnson.YeoJohnson(name='', standardize=True)[source]¶

Bases: Power

A Yeo-Johnson power transformation.

Parameters:

name (str) –
A name for this transformer. If None, use the class name.

By default it is set to “”.
standardize (bool) –
Whether to apply zero-mean, unit-variance normalization to the transformed output.

By default it is set to True.

compute_jacobian(data)¶

Compute the Jacobian of transform().

Parameters:: data (ndarray) – The data where the Jacobian is to be computed, shaped as (n_observations, n_features) or (n_features, ).
Returns:: The Jacobian matrix, shaped according to data.
Return type:: NoReturn

compute_jacobian_inverse(data)¶

Compute the Jacobian of the inverse_transform().

Parameters:: data (ndarray) – The data where the Jacobian is to be computed, shaped as (n_observations, n_features) or (n_features, ).
Returns:: The Jacobian matrix, shaped according to data..
Return type:: NoReturn

duplicate()¶

Duplicate the current object.

Returns:: A deepcopy of the current instance.
Return type:: Transformer

fit(data, *args)¶

Fit the transformer to the data.

Parameters:

data (ndarray) – The data to be fitted, shaped as (n_observations, n_features) or (n_observations, ).
args (float | int | str) –

Return type:

None

fit_transform(data, *args)¶

Fit the transformer to the data and transform the data.

Parameters:

data (ndarray) – The data to be transformed, shaped as (n_observations, n_features) or (n_observations, ).
args (float | int | str) –

Returns:

The transformed data, shaped as data.

Return type:

ndarray

inverse_transform(data, *args, **kwargs)¶

Force a NumPy array to be 2D and evaluate the function f with it.

Parameters:

data (ndarray) – A 1D or 2D NumPy array.
*args (Any) – The positional arguments.
**kwargs (Any) – The optional arguments.

Returns:

Any kind of output; if a NumPy array, its dimension is made consistent with the shape of data.

Return type:

Any

transform(data, *args, **kwargs)¶

Force a NumPy array to be 2D and evaluate the function f with it.

Parameters:

data (ndarray) – A 1D or 2D NumPy array.
*args (Any) – The positional arguments.
**kwargs (Any) – The optional arguments.

Returns:

Any kind of output; if a NumPy array, its dimension is made consistent with the shape of data.

Return type:

Any

CROSSED: ClassVar[bool] = False¶: Whether the fit() method requires two data arrays.

property is_fitted: bool¶: Whether the transformer has been fitted from some data.

lambdas_: ndarray¶: The parameters of the power transformation for the selected features.

name: str¶: The name of the transformer.

property parameters: dict[str, bool | int | float | ndarray | str | None]¶: The parameters of the transformer.