sorbetto.core package

class sorbetto.core.AbstractHomogeneousBinaryRelationOnPerformances(name: str | None = None)[source]

Bases: ABC

getDual()[source]

abstract isAntisymmetric() → bool[source]

abstract isAsymmetric() → bool[source]

abstract isEquivalence() → bool[source]

abstract isIrreflexive() → bool[source]

abstract isOrder() → bool[source]

abstract isPartialOrder() → bool[source]

abstract isPreorder() → bool[source]

abstract isReflexive() → bool[source]

abstract isSymmetric() → bool[source]

abstract isTotalOrder() → bool[source]

abstract isTransitive() → bool[source]

property name: str | None

class sorbetto.core.Entity(performance: TwoClassClassificationPerformance, name: str = 'ε', color: Any = None)[source]

Bases: object

property color: str | tuple[float] | list[float]

property name: str

property performance: TwoClassClassificationPerformance

The result of the evaluation of the entity, that is its performance.

Returns:: The entity’s performance
Return type:: TwoClassClassificationPerformance

class sorbetto.core.Importance(itn: float | int, ifp: float | int, ifn: float | int, itp: float | int, name: str = 'I')[source]

Bases: object

This class encodes some application-specific preferences. Currently, it is a random variable, called importance, that gives a positive value to each element of the sample space: tn (for true negative), fp (for false positive), fn (for false negative), and tp (for true positive).

See Piérard et al. [17] for more information on this topic.