HyperOptUtils¶
-
class
mastml.hyper_opt.HyperOptUtils(param_names, param_values)[source]¶ Bases:
objectHelper class providing useful methods for other hyperparameter optimization classes.
- Args:
param_names: (list), list containing names of hyperparams to optimize
param_values: (list), list containing values of hyperparams to optimize
- Methods:
- _search_space_generator : parses GridSearch param_dict and checks values
- Args:
- params: (dict), dict of {param_name : param_value} pairs.
- Returns:
- params_: (dict), dict of {param_name : param_value} pairs.
- _save_output : saves hyperparameter optimization output and best values to csv file
- Args:
savepath: (str), path of output directory
data: (dict), dict of {estimator_name : hyper_opt.GridSearch.fit()} object, or equivalent
- Returns:
- None
- _get_grid_param_dict : configures the param_dict for GridSearch
- Args:
- None
- Returns:
- param_dict: (dict), dict of {param_name : param_value} pairs.
- _get_randomized_param_dict : configures the param_dict for RandomSearch
- Args:
- None
- Returns:
- param_dict: (dict), dict of {param_name : param_value} pairs.
- _get_bayesian_param_dict : configures the param_dict for BayesianSearch
- Args:
- None
- Returns:
- param_dict: (dict), dict of {param_name : param_value} pairs.