iOpt.method Package

listener

method

optim_task

class iOpt.method.optim_task.TypeOfCalculation(value)

Базовые классы: enum.Enum

An enumeration.

FUNCTION = 1

CONVOLUTION = 2

class iOpt.method.optim_task.OptimizationTask(problem: Problem, perm: np.ndarray(shape=1, dtype=int) = None)

Базовые классы: object

calculate(data_item: iOpt.method.search_data.SearchDataItem, function_index: int, calculation_type: iOpt.method.optim_task.TypeOfCalculation = TypeOfCalculation.FUNCTION) → iOpt.method.search_data.SearchDataItem

Compute selected function by number

process

search_data

class iOpt.method.search_data.SearchDataItem(y: Point, x: np.double, function_values: np.ndarray(shape=1, dtype=FunctionValue) = [<iOpt.trial.FunctionValue object>], discrete_value_index: int = 0)

Базовые классы: iOpt.trial.Trial

The SearchDataItem class is intended for storing search information, which is an interval with a right point included, as well as links to neighbouring intervals. SearchDataItem is an inheritor of the Trial class

__init__(y: Point, x: np.double, function_values: np.ndarray(shape=1, dtype=FunctionValue) = [<iOpt.trial.FunctionValue object>], discrete_value_index: int = 0)

Constructor of SearchDataItem class

Параметры

• y – trial point in the original N-dimensional search area.

• x – Mapping the trial point y to the segment [0, 1].

• function_values – Vector of function values (objective and constraint functions).

• discrete_value_index – Discrete parameter.

get_x() → numpy.float64

Obtain the right point of the search interval where \(x\in[0, 1]\)

Результат

Value of the right point of the interval.

get_y() → iOpt.trial.Point

Provide an N-dimensional trial point of the original search area

Результат

N-dimensional trial point value.

get_discrete_value_index() → int

Obtain a discrete parameter

Результат

Discrete parameter value.

set_index(index: int)

Specify the index value of the last executed constraint for the index scheme

Параметры

index – Restriction index.

get_index() → int

Get the index value of the last executed constraint for the index scheme

Результат

Index value.

set_z(z: numpy.float64)

Allow you to specify a function value for a given index.

Параметры

z – Function value.

get_z() → numpy.float64

Get the value of the function for a given index

Результат

Function value for index.

set_left(point: iOpt.method.search_data.SearchDataItem)

Set the left interval for the source interval

Параметры

point – Left interval.

get_left() → iOpt.method.search_data.SearchDataItem

Get the left interval for the original interval

Результат

Left interval value.

set_right(point: iOpt.method.search_data.SearchDataItem)

Set the right interval for the original interval

Параметры

point – Right interval.

get_right() → iOpt.method.search_data.SearchDataItem

Get the right interval for the original interval

Результат

Right interval value.

__lt__(other) → bool

The method overrides the < comparison operator for two intervals

Параметры

other – Second interval.

Результат

The value is true - if the right point of the initial interval is less than the

the right point of the second interval, otherwise - false.

class iOpt.method.search_data.CharacteristicsQueue(maxlen: int)

Базовые классы: object

The CharacteristicsQueue class is designed to store a prioritised queue of characteristics with preempting

__init__(maxlen: int)

Constructor of the CharacteristicsQueue class

Параметры

maxlen – Maximum queue size.

Clear()

Clear the queue

insert(key: numpy.float64, data_item: iOpt.method.search_data.SearchDataItem)

Add search interval with specified priority. The priority is the value of the characteristic on this interval

Параметры

• key – Priority of the search interval.

• data_item – Insertion interval.

get_best_item()

Get the interval with the best characteristic

Результат

Tuple: interval with the best characteristic, priority of the interval in the queue.

is_empty()

Check for queue emptiness.

Результат

True if the queue is empty, otherwise false.

get_max_len() → int

Get the maximum queue size.

Результат

Value of maximum queue size.

get_len() → int

Get the current queue size

Результат

Value of the current queue size.

class iOpt.method.search_data.SearchData(problem: iOpt.problem.Problem, maxlen: Optional[int] = None)

Базовые классы: object

The SearchData class is used to store the set of all intervals, the original task and the priority queue of global characteristics

__init__(problem: iOpt.problem.Problem, maxlen: Optional[int] = None)

Constructor of SearchData class

Параметры

• problem – Information about the original task.

• maxlen – Maximum queue size.

clear_queue()

Clear the characteristic queue

insert_data_item(new_data_item: iOpt.method.search_data.SearchDataItem, right_data_item: Optional[iOpt.method.search_data.SearchDataItem] = None)

Add a new trial interval to the list of all trials performed and prioritised characteristic queue

Параметры

• new_data_item – New trial interval.

• right_data_item – The covering interval, is the right interval for the newDataItem.

insert_first_data_item(left_data_item: iOpt.method.search_data.SearchDataItem, right_data_item: iOpt.method.search_data.SearchDataItem)

Allow a pair of trial intervals to be added to the first iteration of the GSA.

Параметры

• left_data_item – Left interval for right_data_item.

• right_data_item – Right interval for left_data_item.

find_data_item_by_one_dimensional_point(x: numpy.float64) → iOpt.method.search_data.SearchDataItem

Find the covering interval for the obtained point x

Параметры

x – Right point of the interval.

Результат

Right point of the covering interval.

get_data_item_with_max_global_r() → iOpt.method.search_data.SearchDataItem

Obtain the interval with the best value of the global characteristic

Результат

Value of the interval with the best global characteristic.

refill_queue()

Refill the queue of global characteristics, for example, when it is empty or when the Lipschitz constant estimation is changed

get_count() → int

Get the current number of intervals in the list

Результат

Value of the number of intervals in the list.

get_last_item() → iOpt.method.search_data.SearchDataItem

Get the last added interval to the list

Результат

Value of the last added interval.

get_last_items(N: int = 1) → list[SearchDataItem]

Get the last added intervals to the list.

Результат

Values of the last series of added intervals.

searchdata_to_json(mode='full') → json

Save the optimization process to a file

Параметры

• mode – „full“ - save all optimization information

• file_name – file name.

json_to_searchdata(data, mode='full')

Load the optimization process from a file

Параметры

file_name – file name.

class iOpt.method.search_data.SearchDataDualQueue(problem: iOpt.problem.Problem, maxlen: Optional[int] = None)

Базовые классы: iOpt.method.search_data.SearchData

The SearchDataDualQueue class is incherited of the SearchData class. It is intended

for storing a set of all intervals, the initial task and two priority queues for global and local characteristics

__init__(problem: iOpt.problem.Problem, maxlen: Optional[int] = None)

Constructor of SearchDataDualQueue class

Параметры

• problem – Information on the initial task.

• maxlen – Maximum queue size.

clear_queue()

Clear the characteristic queues

insert_data_item(new_data_item: iOpt.method.search_data.SearchDataItem, right_data_item: Optional[iOpt.method.search_data.SearchDataItem] = None)

Add a new trial interval to the list of all trials performed

and priority queues of global and local characteristics

Параметры

• new_data_item – New trial interval.

• right_data_item – The covering interval, is the right interval for the new_data_item.

get_data_item_with_max_global_r() → iOpt.method.search_data.SearchDataItem

Obtain the interval with the best value of the global characteristic

Результат

Value of the interval with the best global characteristic.

get_data_item_with_max_local_r() → iOpt.method.search_data.SearchDataItem

Obtain the interval with the best value of the local characteristic

Результат

Value of the interval with the best local characteristic.

refill_queue()

Refill the queues of global and local characteristics, e.g.,

when they are emptied or when the Lipschitz constant estimation is changed

sol_value

class iOpt.method.sol_value.SolutionValue(calculations_number: int = - 1, holder_constants_estimations: numpy.float64 = - 1.0)

Базовые классы: iOpt.trial.FunctionValue