tarexp.component.stopping module#

Stopping rules are a particular focus of TAR research.

class tarexp.component.stopping.StoppingRule(target_recall: float | None = None)[source]#

Bases: Component

checkStopping(ledger: Ledger, *args, **kwargs) bool[source]#
class tarexp.component.stopping.NullStoppingRule(target_recall: float | None = None)[source]#

Bases: StoppingRule

checkStopping(*args, **kwargs)[source]#
class tarexp.component.stopping.FixedRoundStoppingRule(max_round, *args, **kwargs)[source]#

Bases: StoppingRule

checkStopping(ledger, *args, **kwargs)[source]#
class tarexp.component.stopping.KneeStoppingRule(target_recall: float | None = None)[source]#

Bases: StoppingRule

See also

checkStopping(ledger: Ledger, *args, **kwargs) bool[source]#
class tarexp.component.stopping.BudgetStoppingRule(target_recall: float | None = None)[source]#

Bases: StoppingRule

See also

checkStopping(ledger: Ledger, *args, **kwargs) bool[source]#
class tarexp.component.stopping.ReviewHalfStoppingRule(target_recall: float | None = None)[source]#

Bases: StoppingRule

checkStopping(ledger: Ledger, *args, **kwargs) bool[source]#
class tarexp.component.stopping.BatchPrecStoppingRule(prec_cutoff=0.025, slack=1)[source]#

Bases: StoppingRule

checkStopping(ledger: Ledger, *args, **kwargs) bool[source]#
class tarexp.component.stopping.Rule2399StoppingRule(target_recall: float | None = None)[source]#

Bases: StoppingRule

checkStopping(ledger: Ledger, *args, **kwargs) bool[source]#
class tarexp.component.stopping.QuantStoppingRule(target_recall: float, nstd: float = 0)[source]#

Bases: StoppingRule

See also

checkStopping(ledger: Ledger, workflow, **kwargs) bool[source]#
class tarexp.component.stopping.CHMHeuristicsStoppingRule(target_recall: float, alpha=0.05)[source]#

Bases: StoppingRule

See also

checkStopping(ledger: Ledger, *args, **kwargs) bool[source]#