concepts.dm.pdsketch.executor.GeneratorManager#

class GeneratorManager[source]#

Bases: object

The GeneratorManager is used to manage calls to generators in the function dmoain. It is particularly useful for keep tracking of historical values generated by the generators.

Methods

`call`(g, max_generator_trials, args[, ...])	Call a generator.
`export_generator_calls`()	Export the generator calls.
`export_generator_stats`([divide_by])	Export the generator statistics.
`mark_success`(assignment_dict)	Mark the values in an assignment dictionary as successful.

Attributes

`store_history`	Whether to store the historical values generated by the generators.
`executor`	The executor.
`generator_calls`	Mappings from generator names to the list of calls made to the generator, including a tuple of the arguments and a tuple of the return values.
`generator_calls_successful`	Mappings from generator names to the list of Boolean values indicating whether the generated values lead to successful solution.

__init__(executor, store_history=True)[source]#

Initialize the generator manager.

Parameters:

executor (PDSketchExecutor) – the executor.
store_history (bool) – whether to store the historical values generated by the generators.

__new__(**kwargs)#

call(g, max_generator_trials, args, constraint_list=None)[source]#

Call a generator.

Parameters:

g (Generator) – the generator.
max_generator_trials (int) – the maximum number of trials to generate values.
args (Tuple[Any, ...]) – the arguments of the generator.
constraint_list (List[Constraint] | None) – the list of constraints to be satisfied by the generated values. This will be passed to the generator function if the list contains more than one constraint.

Yields:

A tuple of (index, generated value). The index is a tuple of (generator_name, value_index).

Return type:

Iterator[Tuple[Tuple[str, int], Any]]

export_generator_calls()[source]#

Export the generator calls.

Returns:: a dictionary mapping from generator names to the list of calls made to the generator, including a tuple of the arguments and a tuple of the return values, and a Boolean value indicating whether the generated values lead to successful solution.
Return type:: Dict[str, List[Tuple[Tuple[Any, …], Tuple[Any, …], bool]]]

export_generator_stats(divide_by=1)[source]#

Export the generator statistics.

Returns:: a string containing the generator statistics.
Parameters:: divide_by (float)
Return type:: str

mark_success(assignment_dict)[source]#

Mark the values in an assignment dictionary as successful.

Parameters:: assignment_dict (Dict[int, Assignment]) – the assignment dictionary.

executor: PDSketchExecutor#: The executor.

generator_calls: Dict[str, List[Tuple[Tuple[Any, ...], Tuple[Any, ...]]]]#: Mappings from generator names to the list of calls made to the generator, including a tuple of the arguments and a tuple of the return values.

generator_calls_successful: Dict[str, List[bool]]#: Mappings from generator names to the list of Boolean values indicating whether the generated values lead to successful solution.

property store_history: bool#: Whether to store the historical values generated by the generators.