cover.coverage
cover.coverage(subsets, ind=None, weights=None)
Returns each elements coverage by the sets given by ind.
The coverage of a element \(e \in U\) in universe \(U\) is defined as the weighted sum: \[ \mathrm{coverage}(e) = \sum_{s \in S} w(s) \cdot \mathbf{1}(e \in s) \] where \(w(s)\) is the weight of set \(s \in S\). In the unweighted setting, the coverage is of an element is equal to number of sets containing it.
This function specialized the coverage computation efficiently for a variety of common sparse matrix formats., and can be used to quickly verify a family of sets satisfies the covering property.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
subsets |
sparray | (n x J) sparse matrix describing a set family of J sets covering n elements. |
required |
ind |
Optional[np.ndarray] | indices indicating which subsets to restrict too. By default, all sets are considered. | None |
Returns
| Type | Description |
|---|---|
| np.ndarray | the coverage of all elements in the sets specified by ind, or all sets otherwise. |
See Also
- valid_cover