train

SPAREICE.train(data, iwp_inputs=None, ice_cloud_inputs=None, iwp_model=None, ice_cloud_model=None, processes=None, cv_folds=None)[source]

Train SPARE-ICE with data

This trains the IWP regressor and ice cloud classifier.

Parameters:

  • data – A pandas.DataFrame object with the required input fields.

  • iwp_inputs – A list with the input field names for the IWP regressor. If this is None, the IWP regressor won’t be trained.

  • ice_cloud_inputs – A list with the input field names for the ice cloud classifier. If this is None, the ice cloud classifier won’t be trained.

  • iwp_model – Set this to your own sklearn estimator class.

  • ice_cloud_model – Set this to your own sklearn estimator class.

  • processes – Number of processes to parallelize the regressor training. If not set, the value from the initialization is taken.

  • cv_folds – Number of folds used for cross-validation. Default is 5. The higher the number the more data is used for training but the runtime increases. Good values are between 3 and 10.

Returns:

None