运用DeepWalk 对图进行嵌入并使用逻辑回归进行链路预测的 Python 代码

以下是使用DeepWalk进行图嵌入和逻辑回归进行链路预测的Python代码：

首先，安装依赖库：

!pip install networkx
!pip install gensim
!pip install scikit-learn

然后，导入必要的库：

import networkx as nx
import numpy as np
from gensim.models import Word2Vec
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import roc_auc_score, average_precision_score
from sklearn.model_selection import train_test_split

接下来，定义函数进行图嵌入：

def graph_embedding(G, dimensions=64, walk_length=30, num_walks=200, workers=4):
    walks = []
    for node in G.nodes():
        for i in range(num_walks):
            walks.append(nx.random_walk(G, node, walk_length=walk_length))
    embeddings = Word2Vec(walks, size=dimensions, window=10, min_count=0, sg=1, workers=workers)
    return embeddings

然后，定义函数进行链路预测：

def link_prediction(embeddings, edges):
    X = np.array([np.concatenate((embeddings[str(edge[0])], embeddings[str(edge[1])])) for edge in edges])
    y = np.array([edge[2] for edge in edges])
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
    clf = LogisticRegression(random_state=0)
    clf.fit(X_train, y_train)
    y_pred = clf.predict_proba(X_test)[:, 1]
    roc = roc_auc_score(y_test, y_pred)
    ap = average_precision_score(y_test, y_pred)
    return roc, ap

最后，使用示例：

# 构建图
G = nx.karate_club_graph()

# 添加随机标签
for node in G.nodes():
    G.nodes[node]['label'] = np.random.randint(2)

# 构建边列表
edges = []
for edge in G.edges():
    label = 1 if G.nodes[edge[0]]['label'] == G.nodes[edge[1]]['label'] else 0
    edges.append((edge[0], edge[1], label))

# 进行图嵌入
embeddings = graph_embedding(G)

# 进行链路预测
roc, ap = link_prediction(embeddings, edges)

# 输出结果
print('ROC AUC score:', roc)
print('Average Precision score:', ap)

代码说明：

该示例使用了karate_club_graph构建了一个简单的图，并随机为节点添加了标签。然后，使用graph_embedding函数进行图嵌入，生成节点的向量表示。接着，使用link_prediction函数进行链路预测，其中使用了逻辑回归模型。最终，输出ROC AUC和Average Precision两个评价指标的得分。