MCPcopy Create free account
hub / github.com/chengsen/PyTorch_TextGCN / __init__

Method __init__

build_graph.py:92–108  ·  view source on GitHub ↗
(self, dataset)

Source from the content-addressed store, hash-verified

90
91class BuildGraph:
92 def __init__(self, dataset):
93 clean_corpus_path = "data/text_dataset/clean_corpus"
94 self.graph_path = "data/graph"
95 if not os.path.exists(self.graph_path):
96 os.makedirs(self.graph_path)
97
98 self.word2id = dict() # 单词映射
99 self.dataset = dataset
100 print(f"\n==> 现在的数据集是:{dataset}<==")
101
102 self.g = nx.Graph()
103
104 self.content = f"{clean_corpus_path}/{dataset}.txt"
105
106 self.get_tfidf_edge()
107 self.get_pmi_edge()
108 self.save()
109
110 def get_pmi_edge(self):
111 pmi_edge_lst, self.pmi_time = get_pmi_edge(self.content, window_size=20, threshold=0.0)

Callers

nothing calls this directly

Calls 3

get_tfidf_edgeMethod · 0.95
get_pmi_edgeMethod · 0.95
saveMethod · 0.95

Tested by

no test coverage detected