try:
pf = open(path, "r+")
result = [] # 存放文档分布
lines = pf.readlines()
for line in lines:
line = line.strip()
nums = line.split(" ")
result.append([float(i) for i in nums])
pf.close()
return result
except IOError:
print("文件无法正常打开")
return None
if __name__ == "__main__":
csv_filename = "data10.csv"
doc = du.getDocAsWordArray(csv_filename)
r_word_list, model = lda_model(doc, 10)
doc_topic = list(model.doc_topic_)
topic_word = list(model.topic_word_)
writeResult(doc_topic, "10_lda_doc_topic.txt")
"""
writeResult(doc_topic, "12_lda_doc_topic.txt")
title = "LDA"
print("正在降维")
d2_data = dr.dimension_down(doc_topic)
cp.paintModelPoint(d2_data, title)
"""
# 将结果转为可进行评估的方式,未被聚类的设置为-1
result = [-1] * whole_num
for i in range(whole_num):
for x in range(len(cluster_result)):
if i in cluster_result[x]:
result[i] = x
break
return result
if __name__ == "__main__":
k = 12
filename = "data1322.csv"
# 4 diping 3 8,12 diping 5
doc = du.getDocAsWordArray(filename, 5)
cluster_result = clusterResult_qt(k, doc)
former = du.getFormerCategory(filename)
# print(former)
# 将数据转为能直接进行准确度判定的形式
c_r_result = []
f_r = []
for i in range(len(cluster_result)):
if cluster_result[i] != -1:
c_r_result.append(cluster_result[i])
f_r.append(former[i])
# print(len(c_r_result))
dis_sum = sum(topic_distribution)
for i in range(self.k):
topic_distribution[i] /= dis_sum
for i in range(self.k):
doc_topic_dis[i] += topic_distribution[i] / bit_count
distribution.append(doc_topic_dis)
return distribution
# test---------------------------test
# 返回文档doc
if __name__ == "__main__":
file_name = "data10.csv"
document = du.getDocAsWordArray(file_name)
doc_for_test = {}
for name_simple in document:
doc_for_test[name_simple] = document[name_simple]
topic_num = 100
start = time.perf_counter()
model = BtmModel(topic_num, doc_for_test)
model.buildModel()
end = time.perf_counter()
print("time used:{}".format(end - start))
print("输出主题——词")
# 获取lda模型和词袋
print("创建主题模型")
word_list, r_model = ldaa.lda_model(doc, k, iterator)
# 获取文档——主题分布
doc_topic = r_model.doc_topic_
# 转为普通list进行聚类
doc_topic_list = np.array(doc_topic).tolist()
estimator = kmn.kMeansByFeature(topic, doc_topic_list)
labels = estimator.labels_
return list(labels)
if __name__ == "__main__":
r_k = 8
r_topic = 8
file_name = "data8.csv"
doc = du.getDocAsWordArray(file_name, 3)
# 获取标签信息
former = du.getFormerCategory(file_name)
model = w2v.load_model_binary(r"E:\学校\快乐推荐\word2vec\api_saveVec")
result = clusterResult_gibbs(r_k, r_topic, model, doc)
result_list = cr.printResult(r_topic, result, former)
print(result_list)