zoo.automl.search package

Submodules

zoo.automl.search.RayTuneSearchEngine module

class zoo.automl.search.RayTuneSearchEngine.RayTuneSearchEngine(logs_dir='', resources_per_trial=None, name='', remote_dir=None)

Bases: zoo.automl.search.abstract.SearchEngine

Tune driver

compile(input_df, model_create_func, search_space, recipe, feature_transformers=None, future_seq_len=1, validation_df=None, mc=False, metric='mse', metric_mode='min')

Do necessary preparations for the engine :param input_df: :param search_space: :param num_samples: :param stop: :param search_algorithm: :param search_algorithm_params: :param fixed_params: :param feature_transformers: :param model: :param validation_df: :param metric: :return:

get_best_trials(k=1)

Get the best trials from . :param k: trials to be selected :return: the config of best k trials

run()

Run trials :return: trials result

test_run()

zoo.automl.search.abstract module

class zoo.automl.search.abstract.BayersianOpt

Bases: object

exception zoo.automl.search.abstract.GoodError

Bases: Exception

class zoo.automl.search.abstract.GridSearch(values)

Bases: object

class zoo.automl.search.abstract.RandomSample(func)

Bases: object

class zoo.automl.search.abstract.SearchEngine

Bases: abc.ABC

Abstract Base Search Engine class. For hyper paramter tuning.

get_best_trials(k)

Get the best trials from . :param k: trials to be selected :return: the config of best k trials

run()

Run the trials with searched parameters :return:

class zoo.automl.search.abstract.TrialOutput(config, model_path)

Bases: object

Module contents