timeseriescv 0.1

Scikit-learn style cross-validation classes for time series data

Package implementing two cross-validation algorithms suitable to evaluate machine learning models based on time series datasets.

Installation

timeseriescv can be installed using pip:

>>> pip install timeseriescv

Content

For now the package contains two main classes handling cross-validation:

• PurgedWalkForwardCV: Walk-forward cross-validation with purging.

• CombPurgedKFoldCV: Combinatorial cross-validation with purging and embargoing.

Check their respective docstrings for more information.