splits#

Module containing splitting functionalities.

train_test_split(data: list[pandas.core.frame.DataFrame], test_split: float = 0.2, input_length: float = 0, split_axis: str = 'vertical') → tuple[list[pandas.core.frame.DataFrame], list[pandas.core.frame.DataFrame]]#

Splits a given dataframe in train, test and validations split.

Parameters:

  • data – List of time series.

  • test_split – percentage of data that will be used for the test split.

  • input_length – length of the input window. Defaults to 0.

  • split_axis – Axis along which the data will be split. Either “horizontal” or “vertical”.

Returns:

Tuple of train and test set.

Raises:

ValueError – if split_axis is not “horizontal” or “vertical”.

train_test_split_horizontal(data: list[pandas.core.frame.DataFrame], test_split: float) → tuple[list[pandas.core.frame.DataFrame], list[pandas.core.frame.DataFrame]]#

Splits a given dataframe in train, test and validations split.

Parameters:

  • data – List of time series.

  • test_split – percentage of data that will be used for the test split.

Returns:

Tuple of train and test set.

train_test_split_vertical(data: list[pandas.core.frame.DataFrame], test_split: float, input_length: float) → tuple[list[pandas.core.frame.DataFrame], list[pandas.core.frame.DataFrame]]#

Splits a given dataframe in train, test and validations split.

Parameters:

  • data – List of time series.

  • test_split – percentage of data that will be used for the test split.

  • input_length – length of the input window.

Returns:

Tuple of train and test set.