def setUpClass(self):
self.karate = nx.karate_club_graph()
self.measures_no_part = gm.calculate_global_measures(self.karate)
self.totalpart = {x: x for x in list(self.karate.nodes)}
self.measures_part = gm.calculate_global_measures(
self.karate, partition=self.totalpart)
self.extra_measure = {'hat': 'cap'}
def test_from_existing(self):
assert (gm.calculate_global_measures(
self.karate,
partition=self.totalpart,
existing_global_measures=self.measures_no_part) ==
self.measures_part)
measures_with_extra = self.measures_no_part
measures_with_extra.update(self.extra_measure)
assert (gm.calculate_global_measures(
self.karate,
partition=self.totalpart,
existing_global_measures=measures_with_extra) ==
self.measures_part.update(self.extra_measure))
def calculate_global_measures(self,
force=False,
seed=None,
partition=True):
'''
Calculate global measures `average_clustering`,
`average_shortest_path_length`, `assortativity`, `modularity`, and
`efficiency` of G. The resulting dictionary of global measures can be
accessed and manipulated as ``G.graph['global_measures']``
Parameters
----------
force : bool
pass True to recalculate any global measures that have already
been calculated
partition : bool
The "modularity" measure evaluates a graph partition.
pass True to calculate the partition of each graph using
:func:`BrainNetwork.partition`. Note that this won't recalculate
a partition that exists.
If False, modularity may not be calculated.
Returns
-------
dict
a dictionary of global network measures of G
See Also
--------
:func:`calculate_global_measures`
'''
if partition:
self.partition()
partition = nx.get_node_attributes(self, name='module')
else:
partition = None
if force:
global_measures = calculate_global_measures(self,
partition=partition)
self.graph['global_measures'] = global_measures
else:
global_measures = calculate_global_measures(
self,
partition=partition,
existing_global_measures=self.graph.get('global_measures'))
self.graph['global_measures'].update(global_measures)
return self.graph['global_measures']