kmeans 算法，距离改用余弦相似度， Python 实现。

@SleipniR 先谢谢大佬，我咋打不开链接呢。

好像需要梯子：

from sklearn.cluster import k_means_
from sklearn.metrics.pairwise import cosine_similarity, pairwise_distances
from sklearn.preprocessing import StandardScaler


def create_cluster(sparse_data, nclust = 10):

# Manually override euclidean
def euc_dist(X, Y = None, Y_norm_squared = None, squared = False):
#return pairwise_distances(X, Y, metric = 'cosine', n_jobs = 10)
return cosine_similarity(X, Y)
k_means_.euclidean_distances = euc_dist

scaler = StandardScaler(with_mean=False)
sparse_data = scaler.fit_transform(sparse_data)
kmeans = k_means_.KMeans(n_clusters = nclust, n_jobs = 20, random_state = 3425)
_ = kmeans.fit(sparse_data)
return kmeans.labels_

@SleipniR 大佬，我看了下这个算法
def euc_dist(x, y=None):
return cosine_similarity(x, y)
k_means_.euclidean_distances = euc_dist
这个算法在这步改成用余弦相似度。计算聚类中心是怎么计算的我点进源码也没看明白。能帮我解释一下他是如何计算类中心的吗？谢谢大佬