concepts.pdsketch.operator.OperatorApplier

class OperatorApplier

Bases: object

An operator applier is essentially a grounded operator, composed of an operator and its arguments.

Methods

iter_sub_operator_appliers()
pddl_str()
replace_arguments(argument_map)	Replace the arguments of the operator applier with a map from argument names to new arguments.

Attributes

is_macro	Whether this operator is a macro operator.
name	The name of the operator.
operator	The operator.
arguments	The arguments of the grounded operator.
regression_rule	The regression rule that generates this operator applier.

__init__(operator, arguments, regression_rule=None)

Initialize an operator applier.

Parameters:

• operator (Operator | MacroOperator)

• arguments (Sequence[str | Variable | UnnamedPlaceholder | TensorValue])

• regression_rule (RegressionRuleApplier | None)

__new__(**kwargs)

iter_sub_operator_appliers()

Return type:

Iterator[OperatorApplier]

pddl_str()

Return type:

str

replace_arguments(argument_map)

Replace the arguments of the operator applier with a map from argument names to new arguments.

Parameters:

argument_map (Mapping[str, Variable | UnnamedPlaceholder | TensorValue])

Return type:

OperatorApplier

arguments: Tuple[str | Variable | UnnamedPlaceholder | TensorValue, ...]

The arguments of the grounded operator.

property is_macro: bool

Whether this operator is a macro operator.

property name: str

The name of the operator.

operator: Operator | MacroOperator

The operator.

regression_rule: RegressionRuleApplier | None

The regression rule that generates this operator applier. It is only used in the context of goal regression search.