casm.learn.fit_and_select

casm.learn.fit_and_select(input, save=True, verbose=True, read_existing=True, hall=None)

Parameters

• input (dict) – The input settings as a dict

• save (boolean, optional, default=True) – Save a pickle file containing the training data and scoring metric. The file name, which can be specified by input[“fit_data_filename”], defaults to “fit_data.pkl”.

• verbose (boolean, optional, default=True) – Print information to stdout.

• read_existing (boolean, optional, default=True) – If it exists, read the pickle file containing the training data and scoring metric. The file name, which can be specified by input[“fit_data_filename”], defaults to “fit_data.pkl”.

• hall (deap.tools.HallOfFame, optional, default=None) – A Hall Of Fame to add resulting individuals to

Returns

• (fdata, estimator, selector)

• fdata (casm.learn.FittingData) – A FittingData instance containing the problem data.

• estimator (estimator object implementing ‘fit’) – The estimator specified by the input settings.

• selector (selector object implementing ‘fit’ and having either a) – ‘get_support()’ or ‘get_halloffame()’ member The feature selector specified by the input settings.