geometry
geometry
Functions
| Name | Description |
|---|---|
| as_pairwise_dist | |
| delaunay_complex | |
| enclosing_radius | Returns the smallest ‘r’ such that the Rips complex on the union of balls of radius ‘r’ is contractible to a point. |
| flag_weight | Filter function factory method for constructing flag/clique filter functions. |
| lower_star_weight | Constructs a simplex-parameterized Callable that evaluates its lower star value based on x. |
| rips_complex | Constructs the Vietoris-Rips complex from x by unioning balls of diameter at most 2 * radius |
| rips_filtration | Constructs a p-dimensional rips filtration from x by unioning balls of diameter at most 2 * radius |
as_pairwise_dist
geometry.as_pairwise_dist(x)
delaunay_complex
geometry.delaunay_complex(x)
enclosing_radius
geometry.enclosing_radius(x)
Returns the smallest ‘r’ such that the Rips complex on the union of balls of radius ‘r’ is contractible to a point.
flag_weight
geometry.flag_weight(x, vertex_weights=None)
Filter function factory method for constructing flag/clique filter functions.
Parameters: x: point cloud, vector of pairwise weights, or square matrix. vertex_weights: optional weights to use for vertices. Defaults to None, which sets vertex weights to 0.
Returns: callable which takes as input a simplex or set of simplices and returns their clique weights.
lower_star_weight
geometry.lower_star_weight(x)
Constructs a simplex-parameterized Callable that evaluates its lower star value based on x.
Vertex labels are assumed to be 0-indexed for now.
If simplex-like, use 0-indexed vertex labels to index vertex values directly.
Otherwise assumes a 2d array of simplex labels is given and vectorizes the computation.
rips_complex
geometry.rips_complex(x, radius=None, p=1)
Constructs the Vietoris-Rips complex from x by unioning balls of diameter at most 2 * radius
Parameters: x: point cloud, pairwise distance vector, or distance matrix radius: scale parameter for the Rips complex. p: highest dimension of simplices to consider in the expansion.
Returns: rips complex, returned as a simplex tree
rips_filtration
geometry.rips_filtration(x, radius=None, p=1, **kwargs)
Constructs a p-dimensional rips filtration from x by unioning balls of diameter at most 2 * radius