用TSNE进行数据降维并展示聚类结果

TSNE提供了一种有效的数据降维方式，让我们可以在2维或3维的空间中展示聚类结果。

# -*- coding: utf-8 -*-

from __future__ import unicode_literals

from sklearn.manifold import TSNE

import pandas as pd

import matplotlib.pyplot as mp

inputfile = 'data/consumption_data.xls'

outputfile = 'tmp/data_type3.xls'

data = pd.read_excel(inputfile, index_col='Id')

data_zs = 1.0 * (data - data.mean()) / data.std()

from sklearn.cluster import KMeans

k = 3

iteration = 800

model = KMeans(n_clusters=k, n_jobs =4, max_iter=iteration)  # n_jobs 并发数设置为4

model.fit(data_zs)  # 开始聚类

r1 = pd.Series(model.labels_).value_counts()  # 统计各个类别的数目

r2 = pd.DataFrame(model.cluster_centers_)  # 找出聚类中心

r = pd.concat([data,pd.Series(model.labels_,index = data.index)],axis=1)

r.columns = list(data.columns) + [r'聚类类别']

tsne = TSNE()

tsne.fit_transform(data_zs)  # 进行数据降维

tsne = pd.DataFrame(tsne.embedding_, index=data_zs.index)

mp.rcParams['font.sans-serif'] = ['SimHei']

mp.rcParams['axes.unicode_minus'] = False

d = tsne[r[r'聚类类别'] == 0]

mp.plot(d[0], d[1], 'r.')

d = tsne[r[r'聚类类别'] == 1]

mp.plot(d[0], d[1], 'go')

d = tsne[r[r'聚类类别'] == 2]

mp.plot(d[0], d[1], 'b*')

mp.show()

运行结果：