TitanGraph graph_clustering

---

graph_clustering(self, edge_distance)

Performs graph clustering over an initial titan graph.

Parameters:

edge_distance : unicode

Column name for the edge distance.

Returns:

: _Unit

Performs graph clustering over an initial titan graph using a distributed edge collapse algorithm.

Examples

The data file sample_graph.txt is a file in the following format: src, dest, distance:

1, 2, 1.5f
2, 1, 1.5f
2, 3, 1.5f
3, 2, 1.5f
1, 3, 1.5f
3, 1, 1.5f

The edge column name should be passed in as an argument to the plug-in.

>>> import trustedanalytics as ta
>>> ta.connect()
>>> my_graph = ta.get_graph("mytitangraph")
>>> my_graph.graph_clustering("dist")

The expected output (new vertices) can be queried:

>>> my_graph.query.gremlin('g.V.map(\'id\', \'vertex\', \'_label\', \'name\',\'count\')')

Snippet output for the above query will look like this:

{u'results': [u'{id=18432, count=null, _label=null, vertex=29, name=null}', u'{id=24576, count=null, _label=null, vertex=22, name=null}', u'{id=27136, count=null, _label=2, vertex=null, name=21944_25304}'

where:

24576 - represents an initial node 27136 - represents a meta-node of 2 nodes (as per _label value)