Discretisation

class arm_preprocessing.discretisation.Discretisation

Bases: object

Discretisation class.

discretise(method='equal_width', num_bins=10, columns=[], information=None)

Discretise the dataset using the specified method.

Parameters:
  • data (pd.DataFrame) – Dataset.

  • method (str) – Discretisation method (‘equal_width’, ‘equal_frequency’, ‘kmeans’).

  • num_bins (int) – Number of bins.

  • columns (list) – List of columns to discretise.

Raises:
  • ValueError – Invalid data type.

  • ValueError – Invalid discretisation method.

  • ValueError – Columns not specified.

  • ValueError – Column type is not numerical.

Returns:

Discretised dataset.

Return type:

pd.DataFrame