|
|
马上注册，结交更多好友，享用更多功能^_^
您需要 登录 才可以下载或查看，没有账号？立即注册
x
先说问题：
一直报错，
ImportError: cannot import name 'Lda2Vec' from 'gensim.models' (C:\ProgramData\anaconda3\lib\site-packages\gensim\models\__init__.py)
ModuleNotFoundError: No module named 'pyLDAvis'
具体经过：
最近想通过python做主题分析，一开始使用的LDA进行分析，发现结果很差不是自己想要的。代码和结果如下：
import jieba
from gensim import corpora, models
from gensim.models import CoherenceModel
import matplotlib.pyplot as plt
# 读取停用词
stopwords_path = 'C:\\Users\\11564\\Desktop\\实战\\scu_stopwords.txt'
stopwords = [line.strip() for line in open(stopwords_path, 'r', encoding='utf-8').readlines()]
# 读取文本数据
text_path = 'C:\\Users\\11564\\Desktop\\实战\\滑雪场.txt'
with open(text_path, 'r', encoding='utf-8') as f:
text = f.read()
# 分词和去停用词处理
words = [word for word in jieba.cut(text) if word not in stopwords]
# 将分词后的文本转换为词袋模型
dictionary = corpora.Dictionary([words])
corpus = [dictionary.doc2bow(words)]
# 通过计算困惑度确定最佳主题数
coherence_values = []
model_list = []
for num_topics in range(1, 20):
model = models.ldamodel.LdaModel(corpus=corpus, id2word=dictionary, num_topics=num_topics)
model_list.append(model)
coherence_model_lda = CoherenceModel(model=model, texts=[words], dictionary=dictionary, coherence='c_v')
coherence_lda = coherence_model_lda.get_coherence()
coherence_values.append(coherence_lda)
# 找到最佳主题数
best_coherence = max(coherence_values)
best_index = coherence_values.index(best_coherence)
best_model = model_list[best_index]
best_num_topics = best_index + 1
# 可视化主题数和困惑度之间的关系
x = range(1, 20)
plt.plot(x, coherence_values)
plt.xlabel('Number of Topics')
plt.ylabel('Coherence Score')
plt.show()
# 输出最佳主题数和主题分布
print('The best number of topics: ', best_num_topics)
topics = best_model.print_topics(num_words=10)
for topic in topics:
print(topic)
输出结果：
The best number of topics: 14
(0, '0.046*"，" + 0.020*"\n" + 0.019*"的" + 0.013*"。" + 0.007*"去" + 0.007*"不错" + 0.006*"好" + 0.006*"玩" + 0.005*"有" + 0.005*"！"')
(1, '0.097*"，" + 0.043*"的" + 0.036*"。" + 0.033*"\n" + 0.013*"不错" + 0.013*"去" + 0.012*"好" + 0.012*"有" + 0.011*"玩" + 0.009*"是"')
(2, '0.070*"，" + 0.028*"的" + 0.024*"\n" + 0.015*"。" + 0.007*"也" + 0.007*"好" + 0.007*"去" + 0.006*"有" + 0.006*"！" + 0.006*"人"')
(3, '0.065*"，" + 0.032*"的" + 0.026*"\n" + 0.018*"。" + 0.010*"不错" + 0.009*"也" + 0.009*"好" + 0.009*"去" + 0.007*"有" + 0.006*"玩"')
(4, '0.108*"，" + 0.050*"的" + 0.028*"\n" + 0.027*"。" + 0.011*"好" + 0.010*"去" + 0.009*"！" + 0.009*"玩" + 0.009*"也" + 0.008*"教练"')
(5, '0.071*"，" + 0.029*"的" + 0.026*"\n" + 0.021*"。" + 0.009*"去" + 0.008*"不错" + 0.008*"玩" + 0.007*"也" + 0.007*"好" + 0.006*"！"')
(6, '0.077*"，" + 0.041*"的" + 0.036*"\n" + 0.027*"。" + 0.011*"不错" + 0.010*"好" + 0.009*"人" + 0.009*"去" + 0.009*"有" + 0.009*"！"')
(7, '0.048*"，" + 0.020*"的" + 0.016*"\n" + 0.015*"。" + 0.007*"去" + 0.006*"好" + 0.005*"教练" + 0.005*"也" + 0.005*"玩" + 0.004*" "')
(8, '0.068*"，" + 0.032*"的" + 0.021*"\n" + 0.017*"。" + 0.007*"不错" + 0.007*"去" + 0.006*"好" + 0.006*"玩" + 0.006*"也" + 0.005*"都"')
(9, '0.037*"，" + 0.015*"的" + 0.012*"\n" + 0.009*"。" + 0.005*"好" + 0.005*"也" + 0.004*"不错" + 0.004*"去" + 0.004*"！" + 0.004*"有"')
(10, '0.135*"，" + 0.057*"的" + 0.043*"\n" + 0.031*"。" + 0.017*"好" + 0.013*"不错" + 0.013*"！" + 0.013*"去" + 0.013*"也" + 0.012*"玩"')
(11, '0.123*"，" + 0.040*"\n" + 0.035*"的" + 0.033*"。" + 0.016*"不错" + 0.011*"好" + 0.011*"也" + 0.010*"玩" + 0.009*"！" + 0.009*"去"')
(12, '0.053*"，" + 0.035*"的" + 0.022*"\n" + 0.019*"。" + 0.008*"去" + 0.007*"玩" + 0.007*"不错" + 0.007*"好" + 0.006*"也" + 0.006*"是"')
(13, '0.065*"，" + 0.030*"\n" + 0.029*"的" + 0.024*"。" + 0.010*"去" + 0.008*"有" + 0.007*"玩" + 0.007*"好" + 0.007*"不错" + 0.007*"是"')
结果很差，然后可视化也不是自己想要的图。于是，换成了：
import gensim
import os
from gensim.models import Lda2Vec
from gensim.models.ldamulticore import LdaMulticore
from gensim.models.word2vec import LineSentence
from gensim.utils import simple_preprocess
from gensim.parsing.preprocessing import STOPWORDS
from gensim.corpora import Dictionary
from gensim.models.coherencemodel import CoherenceModel
import pyLDAvis.gensim_models
import logging
logging.basicConfig(format='%(asctime)s : %(levelname)s : %(message)s', level=logging.INFO)
# 定义读取文件路径
input_folder = "C:/Users/11564/Desktop/实战/滑雪场"
stopwords_file = "C:/Users/11564/Desktop/实战/scu_stopword.txt"
# 加载停用词表
with open(stopwords_file, 'r', encoding='utf-8') as f:
stopwords = f.read().splitlines()
# 去除停用词和长度小于3的词语
def preprocess(text):
result = []
for token in simple_preprocess(text):
if token not in stopwords and len(token) > 3:
result.append(token)
return result
# 读取文件
class TextIterator:
def __init__(self, folder_path):
self.folder_path = folder_path
def __iter__(self):
for file_name in os.listdir(self.folder_path):
file_path = os.path.join(self.folder_path, file_name)
for line in open(file_path, 'r', encoding='utf-8'):
yield preprocess(line)
# 构建LDA2Vec模型
sentences = TextIterator(input_folder)
dictionary = Dictionary(sentences)
model = Lda2Vec(
corpus=[dictionary.doc2bow(sent) for sent in sentences],
num_topics=50,
id2word=dictionary,
chunksize=5000,
passes=10,
alpha=0.5,
eta=0.5,
iterations=200,
random_state=42,
batch_size=128,
workers=4
)
# 保存模型
model.save("lda2vec.model")
# 计算模型的困惑度和一致性得分
perplexity = model.log_perplexity([dictionary.doc2bow(sent) for sent in sentences])
coh_score = CoherenceModel(model=model, texts=sentences, dictionary=dictionary, coherence='c_v').get_coherence()
# 可视化结果
vis_data = pyLDAvis.gensim_models.prepare(model, [dictionary.doc2bow(sent) for sent in sentences], dictionary)
pyLDAvis.show(vis_data)
# 展示主题和对应的词汇
for topic in model.topic_word_.T:
words = ", ".join([model.id2word[idx] for idx in topic.argsort()[-5:]])
print(words)
print("Perplexity: ", perplexity)
print("Coherence score: ", coh_score)
问题就来了，一直报错，
ImportError: cannot import name 'Lda2Vec' from 'gensim.models' (C:\ProgramData\anaconda3\lib\site-packages\gensim\models\__init__.py)
ModuleNotFoundError: No module named 'pyLDAvis'
明明已经pip install 上面的模块和模型了，就是不行，也upgrade最新版了，甚至把gensim也upgrade了，就是不行，不管怎么尝试都不行。有没有大佬能出出主意啊。
这个问题可能是由于使用的gensim版本与pyLDAvis版本不兼容导致的。您可以尝试卸载已安装的pyLDAvis和gensim，并重新安装旧一些的版本：
``` pip uninstall pyldavis gensim
pip install pyLDAvis==2.1.2 gensim==3.8.3 -i https://mirrors.aliyun.com/pypi/simple
```
然后再尝试运行代码。另外，确保在安装这些包之前，先关闭所有Python解释器和编辑器，并在重新打开它们之前重启计算机。
|
|
( 粤ICP备18085999号-1 | 粤公网安备 44051102000585号)
狗仔卡
发表于 2023-4-2 19:42:44
置顶卡
千斤顶
显身卡
发表于 2023-4-2 19:47:19
楼主
，新问题又来了，又读不到后续的模块了，明明就在那，大小写都是对的