def list_clustering(dataset_name):
'''
list clustering results associated with the dataset
:param dataset_name: name of a dataset
'''
path = config.get_result_file_path(dataset_name)
return [os.path.basename(u[:-1]) for u in glob.glob(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 save_result(result):
if isinstance(result, Result):
result.save()
else:
filepath = config.get_result_file_path(result['dataname'],
result['runname'],
create=True)
try:
fpath = os.path.join(filepath, 'result.txt')
with open(fpath, 'wt') as f:
json.dump(result, f)
except:
utils.remove_if_file_exit(fpath, is_dir=False)
raise
def load_result(dataname, runname):
fpath = os.path.join(config.get_result_file_path(dataname, runname),
'result.txt')
with open(fpath, 'rt') as f:
j = json.load(f)
d = dataset.load_local(dataname)
edges = d.get_edges()
nodes = set(edges['src']).union(set(edges['dest']))
result_nodes = set()
for l in j['clusters'].values():
result_nodes.update(l)
missed_nodes = nodes.difference(result_nodes)
j['clusters']['-9999'] = list(missed_nodes)
return Result(j)
def has_run(self, runame, dsname):
fpath = os.path.join(
config.get_result_file_path(dsname=dsname, runname=runame),
'result.txt')
return utils.file_exists(fpath)
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()
def has_result(dataname, runname):
fpath = os.path.join(config.get_result_file_path(dataname, runname),
'result.txt')
return utils.file_exists(fpath)