Factor Groups

class nettlesome.groups.FactorGroup(factors=())

Terms to be used together in a comparison.

add(other)

Combine all Factors into a single FactorGroup.

Return type: Optional[FactorGroup]

add_or_raise_error(other)

Combine all Factors into a single FactorGroup.

Return type: FactorGroup

consistent_with(other, context=None)

Find whether two sets of Factors can be consistent.

Works by first determining whether one Factor potentially contradicts() another, and then determining whether it’s possible to make context assignments match between the contradictory Factors.

Parameters

other (Optional[Comparable]) –
context (Optional[ContextRegister]) – correspondences between Factors in self and other that can’t be changed in seeking a way to interpret the groups as consistent

Return type

bool

Returns

whether unassigned context factors can be assigned in such a way that there’s no contradiction between any factor in self_factors and other_factors, given that some Factors have already been assigned as described by matches.

contradicts(other, context=None)

Find whether two sets of Factors can be contradictory.

Parameters

other (Optional[Comparable]) – a second set of Factors with context factors that are internally consistent, but may not be consistent with self_factors.
context (Optional[ContextRegister]) – correspondences between Factors in self and other that can’t be changed in seeking a contradiction

Return type

bool

Returns

whether any Factor assignment can be found that makes a Factor in the output of other contradict a Factor in the output of self.

>>> from nettlesome import Statement, Entity
>>> nafta = FactorGroup([
... Statement(predicate="$country1 signed a treaty with $country2",
...        terms=[Entity(name="Mexico"), Entity(name="USA")]),
... Statement(predicate="$country2 signed a treaty with $country3",
...        terms=[Entity(name="USA"), Entity(name="Canada")]),
... Statement(predicate="$country3 signed a treaty with $country1",
...    terms=[Entity(name="USA"), Entity(name="Canada")])])
>>> brexit = FactorGroup([
... Statement(predicate="$country1 signed a treaty with $country2",
...         terms=[Entity(name="UK"), Entity(name="European Union")]),
... Statement(predicate="$country2 signed a treaty with $country3",
...         terms=[Entity(name="European Union"), Entity(name="Germany")]),
... Statement(predicate="$country3 signed a treaty with $country1",
...     terms=[Entity(name="Germany"), Entity(name="UK")], truth=False)])
>>> nafta.contradicts(brexit)
True

drop_implied_factors()

Reduce group by removing redundant members implied by other members.

Return type: FactorGroup
Returns: new group with any redundant items remomved

explanations_contradiction(other, context=None)

Find contexts that would cause self to contradict other.

In this example, by adding a context parameter to this method’s comparison of two FactorGroups for contradiction, we can narrow down how nettlesome discovers analogies between the Entity objects. The result is that nettlesome finds only two Explanations for how a contradiction can exist.

>>> from nettlesome import Statement, Entity
>>> nafta = FactorGroup([
... Statement(predicate="$country1 signed a treaty with $country2",
...     terms=[Entity(name="Mexico"), Entity(name="USA")]),
... Statement(predicate="$country2 signed a treaty with $country3",
...     terms=[Entity(name="USA"), Entity(name="Canada")]),
... Statement(predicate="$country3 signed a treaty with $country1",
...    terms=[Entity(name="USA"), Entity(name="Canada")])])
>>> brexit = FactorGroup([
... Statement(predicate="$country1 signed a treaty with $country2",
...     terms=[Entity(name="UK"), Entity(name="European Union")]),
... Statement(predicate="$country2 signed a treaty with $country3",
...     terms=[Entity(name="European Union"), Entity(name="Germany")]),
... Statement(predicate="$country3 signed a treaty with $country1",
...     terms=[Entity(name="Germany"), Entity(name="UK")], truth=False)])
>>> explanations_usa_like_uk = nafta.explanations_contradiction(
...     brexit,
...     context=([Entity(name="USA")], [Entity(name="UK")]))
>>> len(list(explanations_usa_like_uk))
2