haddock.fcc.cluster_fcc module

Asymmetric Taylor-Butina Disjoint Clustering Algorithm.

Authors:

RODRIGUES Joao TRELLET Mikael MELQUIOND Adrien

class haddock.fcc.cluster_fcc.Cluster(name, center)

Bases: object

Defines a Cluster. A Cluster is created with a name and a center (Element class)

add_member(element)

Adds one single element to the cluster.

center

members

name

populate()

Populates the Cluster member list through the neighbor list of its center.

class haddock.fcc.cluster_fcc.Element(name)

Bases: object

Defines a ‘clusterable’ Element

add_neighbor(neighbor)

Adds another element to the neighbor list

assign_cluster(clust_id)

Assigns the Element to Cluster. 0 if unclustered

cluster

name

neighbors

haddock.fcc.cluster_fcc.cluster_elements(element_pool, threshold)

Groups Elements within a given threshold together in the same cluster.

haddock.fcc.cluster_fcc.output_clusters(handle, clusters)

Outputs the cluster name, center, and members.

haddock.fcc.cluster_fcc.read_matrix(path, cutoff, strictness)

Reads in a four column matrix (1 2 0.123 0.456

)

and creates an dictionary of Elements.

The strictness factor is a <float> that multiplies by the cutoff to produce a new cutoff for the second half of the matrix. Used to allow some variability while keeping very small interfaces from clustering with anything remotely similar.

haddock.fcc.cluster_fcc.remove_true_singletons(element_pool)

Removes from the pool elements without any neighbor