Gensim Word2Vec 模型更新后如何正确提取词向量进行PCA降维

from gensim.models import Word2Vec
from sklearn.decomposition import PCA
import pandas as pd
import numpy as np

# 示例语料库
corpus = [
    ["the", "quick", "brown", "fox", "jumps", "over", "the", "lazy", "dog"],
    ["the", "dog", "barks", "at", "the", "cat"],
    ["a", "cat", "chases", "a", "mouse"],
    ["fox", "cat", "dog", "mouse", "barks", "jumps", "chases", "lazy", "quick", "brown"]
]

# 初始化并训练Word2Vec模型
model = Word2Vec(corpus, min_count=1, vector_size=5)

# 旧版代码中提取词向量的方式 (在新版gensim中会导致错误)
# X = model[model.wv.vocab] 

# 尝试使用 model.wv.key_to_index (会出错，因为它是一个字典，不是向量数组)
# X = model.wv.key_to_index 
# pca = PCA(n_components=2)
# result = pca.fit_transform(X) # 示例错误: TypeError: A sparse matrix was passed, but dense data is required.

from gensim.models import Word2Vec
from sklearn.decomposition import PCA
import pandas as pd
import numpy as np

# 示例语料库 (与上文相同)
corpus = [
    ["the", "quick", "brown", "fox", "jumps", "over", "the", "lazy", "dog"],
    ["the", "dog", "barks", "at", "the", "cat"],
    ["a", "cat", "chases", "a", "mouse"],
    ["fox", "cat", "dog", "mouse", "barks", "jumps", "chases", "lazy", "quick", "brown"]
]

# 初始化并训练Word2Vec模型
# 注意：min_count和vector_size的设置在实际应用中应遵循最佳实践 (详见下文)
model = Word2Vec(corpus, min_count=1, vector_size=5)

# 正确地获取所有词向量，用于PCA
X = model.wv.vectors

# 打印X的形状以验证
print(f"提取的词向量形状: {X.shape}") 
# 预期输出示例: 提取的词向量形状: (12, 5) (其中12是词汇量，5是向量维度)

# 应用PCA降维
pca = PCA(n_components=2)
result = pca.fit_transform(X)

# 将PCA结果转换为DataFrame
pca_df = pd.DataFrame(result, columns=['x', 'y'])
print("\nPCA 降维结果 DataFrame (前5行):")
print(pca_df.head())

# 如果需要获取对应词汇，可以使用 model.wv.index_to_key
# 词汇的顺序与 model.wv.vectors 中的向量顺序一致
words = model.wv.index_to_key
print(f"\n模型中的词汇 (前5个): {words[:5]}")