def persistent(self):
if not self.is_edges_persistent():
filepath = config.get_data_file_path(self.name, 'edges.parq')
self.persistent_edges(filepath)
if self.has_ground_truth():
for k, v in self.get_ground_truth().items():
if not v.is_persistent():
filepath = config.get_data_file_path(
self.name, 'gt_{}.parq'.format(k))
v.persistent(filepath)
def testFromSnap(self):
name = "testFromSnap"
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = snap.GenRndGnm(snap.PNGraph, 100, 1000)
d = convert.from_snap(name, G, overide=True)
print ("testFromSnap", d)
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = snap.GenRndGnm(snap.PUNGraph, 100, 1000)
d = convert.from_snap(name, G, overide=True)
print ("testFromSnap", d)
def load_local(name):
'''
load a local dataset
:param name: name of a dataset
'''
path = config.get_data_file_path(name, create=False)
if not utils.file_exists(path):
raise Exception("path not exists: " + path)
with open(os.path.join(path, 'meta.info')) as f:
meta = json.load(f)
edges = meta['parq_edges']
gt = None
if meta["has_ground_truth"]:
gt = {k: Clustering(v) for k, v in meta['parq_ground_truth'].items()}
additional_meta = None if not "additional" in meta else meta['additional']
is_edge_mirrored = meta['is_edge_mirrored']
return Dataset(name=meta['name'],
description=meta['description'],
groundtruthObj=gt,
edgesObj=edges,
directed=meta['directed'],
weighted=meta['weighted'],
overide=False,
additional_meta=additional_meta,
is_edge_mirrored=is_edge_mirrored)
def local_exists(name):
'''
check if dataset 'name' exists locally.
:param name: name of a dataset
:rtype: bool
'''
path = config.get_data_file_path(name)
return utils.file_exists(path)
def remove_local(name, rm_graph_data=True, rm_clustering_result=True):
'''
remove a local dataset
:param name: name of a dataset
:param rm_clustering_result: remove local graph data
:param rm_clustering_result: remove clustering results associated with the graph.
'''
if rm_graph_data:
path = config.get_data_file_path(name)
utils.remove_if_file_exit(path, is_dir=True)
if rm_clustering_result:
path = config.get_result_file_path(name)
utils.remove_if_file_exit(path, is_dir=True)
def testFromIGraph(self):
name = "testFromIGraph"
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = igraph.Graph.Full(5, directed=False)
d = convert.from_igraph(name, G)
print (d)
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = igraph.Graph.Full(5, directed=True)
d = convert.from_igraph(name, G)
print (d)
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = igraph.Graph.Full(5, directed=False)
G.es['weight'] = [1] * G.ecount()
d = convert.from_igraph(name, G)
print (d)
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = igraph.Graph.Full(5, directed=True)
G.es['weight'] = [1] * G.ecount()
d = convert.from_igraph(name, G)
print (d)
def testFromNextworkx(self):
name = "testFromNextworkx"
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = nx.complete_graph(5)
d = convert.from_networkx(name, G, weighted=False, overide=True)
print (d)
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = nx.complete_graph(5).to_directed()
d = convert.from_networkx(name, G, weighted=False, overide=True)
print (d)
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = nx.complete_graph(5)
G.add_weighted_edges_from((u, v, 1) for u, v in nx.complete_graph(5).edges())
d = convert.from_networkx(name, G, weighted=True, overide=True)
print (d)
utils.remove_if_file_exit(config.get_data_file_path(name), is_dir=True)
G = nx.complete_graph(5).to_directed()
G.add_weighted_edges_from((u, v, 1) for u, v in nx.complete_graph(5).edges())
d = convert.from_networkx(name, G, weighted=True, overide=True)
print (d)
def update_meta(self):
if not self.is_anonymous():
meta_file = config.get_data_file_path(self.name, 'meta.info')
json.dump(self.get_meta(), open(meta_file, 'wt'))
def __init__(self,
name=None,
description="",
groundtruthObj=None,
edgesObj=None,
directed=False,
weighted=False,
overide=False,
additional_meta=None,
is_edge_mirrored=False):
assert edgesObj is not None
self.name = name
self.description = description
self.additional_meta = additional_meta
self.logger = utils.get_logger("{}:{}".format(
type(self).__name__, self.name))
self.directed = directed
self.weighted = weighted
self.is_edge_mirrored = is_edge_mirrored
self.parq_edges = None
if name:
assert name
self.file_edges = config.get_data_file_path(self.name, 'edges.txt')
self.file_pajek = config.get_data_file_path(self.name, 'pajek.txt')
self.file_hig = config.get_data_file_path(self.name, 'pajek.hig')
self.file_scanbin = config.get_data_file_path(self.name, 'scanbin')
self.file_anyscan = config.get_data_file_path(
self.name, 'anyscan.txt')
self.file_snap = config.get_data_file_path(self.name, 'snap.bin')
self.file_mcl_mci = config.get_data_file_path(self.name, 'mcl.mci')
self.file_mcl_tab = config.get_data_file_path(self.name, 'mcl.tab')
self.file_topgc = config.get_data_file_path(self.name, 'topgc.txt')
self.file_mirror_edges = config.get_data_file_path(
self.name, 'edges_mirror.txt')
if self.is_weighted():
self.file_unweighted_edges = self.file_edges
else:
self.file_unweighted_edges = config.get_data_file_path(
self.name, 'unweighted_edges.txt')
self.set_ground_truth(groundtruthObj)
self.set_edges(edgesObj)
if name:
is_persistent = self.is_edges_persistent(
) and self.is_ground_truth_persistent()
self.home = config.get_data_file_path(name, create=False)
if utils.file_exists(self.home):
if overide:
utils.remove_if_file_exit(self.home, is_dir=True)
elif is_persistent:
pass
else:
raise Exception(
"Dataset {} exists at {}. Use overide=True or load it locally."
.format(name, self.home))
if not is_persistent:
utils.remove_if_file_exit(config.get_result_file_path(
self.name),
is_dir=True)
utils.create_dir_if_not_exists(self.home)
self.persistent()
self.update_meta()