Utils Module

This module serves as a support for other modules. The following helper functions are defined:

exception scripts.utils.ModelNotFittedError

Bases: Exception

It is raised when a method or attribute is requested that requires the model to be trained (such as .predict() or .score())

scripts.utils.check_consistent_length(*arr)

Checks whether the provided set of nd-arrays have the same first dimension.

Raises

ValueError – If the len of input arrays is incosistent.

scripts.utils.generate_summary(**kwargs)

Generate summary for the given model. The summary is written to the specified filepath.

Parameters

kwargs (dict-like object) – Dict of parameters neeeded in order for summary to be built.

scripts.utils.get_data(raw=True, scaled=True, pca=False)

Loads transformed dara according to the given specificatiion.

Parameters

  • raw (bool) – Do you want to return the raw data.

  • scaled (bool) – Do you want to return data which were standard scaled - zero mean, unit variance.

  • pca (bool) – Do you want to return data which were transformed using pca.

Raises

AssertionError – If the specification of data to be returned does not match the available options.

scripts.utils.validate_feature_matrix(X)

Makes any inserted matrix with less than 3 dimensions into a 2d-feature matrix.

Parameters

X (numpy.ndarray) –

Returns

X – 2d array.

Return type

numpy.ndarray

Raises

AssetionError – If X is 3 dimensional.

scripts.utils.validate_target_vector(y)

Makes sure that the target vector is one dimensional.

Parameters

y (numpy.ndarray) – Target vector.

Returns

y – 1d target vector.

Return type

numpy.ndarray