def as_unweighted_undirected(data, newname, description="", overide=False):
'''
convert the dataset to undirected unweighted dataset
:param data: :py:class:`gct.Dataset`
:param newname: the name of the new dataset
:param description: discription
:param overide: When true and the named dataset already exists, it will be deleted
:rtype: :py:class:`gct.Dataset`
'''
edges = data.get_edges()[['src', 'dest']]
if data.has_ground_truth():
gt = data.get_ground_truth()
else:
gt = None
return Dataset(name=newname,
description=description,
groundtruthObj=gt,
edgesObj=edges,
directed=False,
weighted=False,
overide=overide)
def from_snap(name, graph, description="", overide=False):
'''
create a datast from a SNAP graph
:param name: identifier of the dataset
:param graph: a SNAP graph
:param description: discription
:param overide: When true and the named dataset already exists, it will be deleted
:rtype: :py:class:`gct.Dataset`
'''
import snap
if isinstance(graph, snap.PUNGraph):
directed = False
elif isinstance(graph, snap.PNGraph):
directed = True
else:
raise Exception("Unkown graph type: " + str(type(graph)))
lst = []
for EI in graph.Edges():
lst.append([EI.GetSrcNId(), EI.GetDstNId()])
df = pd.DataFrame(lst, columns=['src', 'dest'])
return Dataset(name,
edgesObj=df,
weighted=False,
directed=directed,
description=description,
overide=overide)
def from_igraph(name, graph, data='weight', description="", overide=True):
'''
create a datast from `iGraph <http://igraph.org/>`_ graph
:param name: identifier of the dataset
:param graph: a igraph graph
:param data: the name of the edge data which is taken as weights. Ignore for unweighted graph.
:param description: discription
:param overide: When true and the named dataset already exists, it will be deleted
:rtype: :py:class:`gct.Dataset`
'''
directed = graph.is_directed()
weighted = graph.is_weighted()
lst = []
for e in graph.es:
lst.append(e.tuple)
df = pd.DataFrame(lst, columns=['src', 'dest'])
if weighted:
w = graph.es[data]
df['weight'] = w
return Dataset(name,
edgesObj=df,
weighted=weighted,
directed=directed,
description=description,
overide=overide)
def load_snap_dataset(name, overide=False):
if not overide and local_exists(name):
return load_local(name)
else:
conf = _DATASET_[name]
edges = conf.get_edges()
gt = conf.get_ground_truth()
description = conf.description
weighted = conf.weighted
directed = conf.directed
return Dataset(name,
description=description,
groundtruthObj=gt,
edgesObj=edges,
directed=directed,
weighted=weighted,
overide=overide)
def as_mirror_edges(data, newname, description="", overide=False):
edges1 = data.get_edges()
edges2 = edges1.copy()
edges2['src'] = edges1['dest']
edges2['dest'] = edges1['src']
edges = pd.concat([edges1, edges2], axis=0)
if data.has_ground_truth():
gt = data.get_ground_truth()
else:
gt = None
return Dataset(name=newname,
description=description,
groundtruthObj=gt,
edgesObj=edges,
directed=data.is_directed(),
weighted=data.is_weighted(),
is_edge_mirrored=True,
overide=overide)
def from_edgelist(name,
edgelist,
groundtruth=None,
directed=False,
description="",
overide=True):
"""
create a graph from edge list.
:param name: identifier of the dataset
:param edgelist: a 2d list (list of list) or a 2d numpy ndaray in [[src node, target node, weight],...] format.
Or a dataframe that has columns of "src","dest","weight".
Weight is optional, if missing it is an unweighted graph.
:param groundtruth: None or a 2d list (list of list) or a 2d numpy ndaray in [[node, cluster],...] format.
Or a dataframe that has columns of "node","cluster".
:param directed: this is a directed graph
:param description: discription
:param overide: When true and the named dataset already exists, it will be deleted
:rtype: :py:class:`gct.Dataset`
"""
assert len(edgelist) > 0, "Error, empty edgelist"
if isinstance(edgelist, pd.DataFrame):
firstrow = edgelist.iloc[0]
else:
firstrow = edgelist[0]
if len(firstrow) == 2:
weighted = False
elif len(firstrow) == 3:
weighted = True
else:
raise Exception("Format not right")
return Dataset(name=name,
edgesObj=edgelist,
groundtruthObj=groundtruth,
weighted=weighted,
directed=directed,
description=description,
overide=overide)
def from_networkx(name,
graph,
weighted=False,
data='weight',
default=1,
description="",
overide=True):
'''
create a datast from networkx graph
:param name: identifier of the dataset
:param graph: a networkx graph
:param weight: is it a weighted graph?
:param data: the name of the edge data which is taken as weights.
:param default: default weight if networkx edge data is missing.
:param description: discription
:param overide: When true and the named dataset already exists, it will be deleted
:rtype: :py:class:`gct.Dataset`
'''
directed = graph.is_directed()
lst = []
if weighted:
for e in graph.edges(data=data, default=default):
lst.append(e)
else:
for e in graph.edges():
lst.append(e)
return Dataset(name,
edgesObj=lst,
weighted=weighted,
directed=directed,
description=description,
overide=overide)