hyponic.utils package

hyponic.utils.history

class hyponic.utils.history.History(**kwargs)

Bases: object

This class is used to track the history of the optimizer. It tracks the global best, current best, and the time of the epoch. Also, it provides methods to visualize the history of the optimizer.

get_current_best() → tuple

This method returns the current best

get_global_best() → tuple

This method returns the global best

get_history() → dict

This method returns the history of the optimizer

is_early_stopping(current_epoch, early_stopping: int)

This method checks if the early stopping condition is met

Parameters:

• current_epoch – current epoch

• early_stopping – number of epochs to wait before stopping the optimization

Returns:

boolean

update_history(epoch, epoch_time)

This method updates the history of the optimizer

static visualize(func)

Decorator to visualize the history of the optimizer

Parameters:

func – function to be decorated

visualize_fitness(**kwargs)

visualize_time(**kwargs)

hyponic.utils.problem_identifier

class hyponic.utils.problem_identifier.ProblemIdentifier(y)

Bases: object

This class is used to identify the problem type using dataset. It can be used for regression and classification only

get_problem_type()

This method identifies the problem type using number of classes and data type of y. If number of classes is 2 and data type of y is int64, then it is a binary classification problem If number of classes is more than 2 and data type of y is int64, then it is a multiclass classification problem If data type of y is not int64, then it is a regression problem

class hyponic.utils.problem_identifier.ProblemType(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)

Bases: Enum

This class is used to identify the problem type using dataset. It can be used for regression and classification only

BINARY_CLASSIFICATION = 2

MULTICLASS_CLASSIFICATION = 3

REGRESSION = 1