ShapeSignature

class ShapeSignature(registration_points: List[RegistrationPoint], hypergraph_incidence_matrix: ndarray)

A ShapeSignature represents a sort of “fingerprint” for a Shape.

Shape signatures speed up embedding and matching by providing a data structure for faster comparisons. If one shape is embedded in another, its signature can embed in the other’s (see can_embed_in()). If one shape’s signature can embed in another’s, it is plausible that that shape can embed in the other. This provides a faster way to filter possible embedding transformations.

Parameters:

registration_points – The registration points within the signature.
rp_connectivity – A dictionary mapping each registration points ID to the IDs of registration points it’s connected to by a line carrier.

Attributes

`centroid`	The centroid (average location) of this signature.
`direct_attribute_frequency`
`kdtree`
`locations`	An `(n, 3)` numpy array containing each of the `n` locations of this signature.
`neighbors_map`
`spoke_attribute_frequency`
`total_spokes`
`registration_points`	The registration points within the signature.

Methods

`can_embed_in`	Returns `True` if there is a sub-signature in `other` such that `self.signature.can_embed_in(sub_signature)` returns `True` (see `can_embed_in()`).
`explore_against`
`filter`
`filter_to_attributes`
`filter_to_rp`
`find_matches_for_other`
`transform`

can_embed_in(other: ShapeSignature) → bool: Returns True if there is a sub-signature in other such that self.signature.can_embed_in(sub_signature) returns True (see can_embed_in()).

explore_against(other: ShapeSignature) → Iterator[Tuple[int, RegistrationPoint]]

filter(predicate: Callable[[RegistrationPoint], Any]) → ShapeSignature

filter_to_attributes(spoke_attributes: Set[ElementAttributes], direct_attributes: Set[ElementAttributes]) → ShapeSignature

filter_to_rp(rp: RegistrationPoint)

find_matches_for_other(other: ShapeSignature, other_idxs: List[int], *, pre_filter: Callable[[int, int, int], bool] | None = None, post_filter: Callable[[Tuple[int, ...]], bool] | None = None) → Iterator[Tuple[int, ...]]

transform(transformation: Transformation) → ShapeSignature

property centroid: The centroid (average location) of this signature.

property direct_attribute_frequency: Dict[ElementAttributes, int]

property kdtree: KDTree

property locations: ndarray: An (n, 3) numpy array containing each of the n locations of this signature.

property neighbors_map: List[Set[int]]

registration_points: The registration points within the signature.

property spoke_attribute_frequency: Dict[ElementAttributes, int]

property total_spokes: int