Class OverlappingClustering
object --+
|
Clustering --+
|
OverlappingClustering
- Known Subclasses:
-
Extension of Clustering that allows for overlapping clusters.
With overlapping clusters, a single vertex can be the member of
multiple clusters (or even none of them). Therefore, each item of the
membership vector is a list, set or tuple containing the cluster indices
for all vertices.
Members of an individual cluster can be accessed by the
[] operator:
>>> cl = Clustering([(0,), (0,), (0,1), (1,), (1,), ()])
>>> cl[0]
[0, 1, 2]
>>> cl[1]
[2, 3, 4]
The membership vector can be accessed by the membership
property:
>>> cl.membership
[frozenset([0]), frozenset([0]), frozenset([0,1]), frozenset([1]), frozenset([1]), frozenset([])]
The number of clusters can be retrieved by the len
function:
>>> len(cl)
2
|
|
__getitem__(self,
idx)
Returns the members of the specified cluster. |
|
|
|
|
__init__(self,
membership,
params={})
Constructor. |
|
|
|
|
sizes(self,
*args)
Returns the size of given clusters. |
|
|
|
Inherited from Clustering:
__len__,
size,
size_histogram
Inherited from object:
__delattr__,
__getattribute__,
__hash__,
__new__,
__reduce__,
__reduce_ex__,
__repr__,
__setattr__,
__str__
|
__getitem__(self,
idx)
(Indexing operator)
|
|
Returns the members of the specified cluster.
- Parameters:
idx - the index of the cluster - Returns:
- the members of the specified cluster as a list
- Raises:
IndexError - if the index is out of bounds - Overrides:
Clustering.__getitem__
|
__init__(self,
membership,
params={})
(Constructor)
|
|
Constructor.
- Parameters:
membership - the membership list -- that is, the cluster index in which each
element of the set belongs to.params - additional parameters to be stored in this object's dictionary. - Overrides:
object.__init__
|
|
Returns the size of given clusters.
- Parameters:
idxs - the cluster indices in which we are interested. If
None, defaults to all clusters - Overrides:
Clustering.sizes
|