利用G25祖源计算器坐标画PCA散点图

利用G25祖源计算器坐标画PCA散点图

概述

---

编程语言：python3.x
模块：numpy sklearn matplotlib
可选：jupyter
整体思路：将G25给出的25维坐标降维并画图

二维PCA散点图

---

步骤：

1. 先把自己手中的G25坐标数据集整理成csv
2. 读取csv文件（此处我直接用的numpy读取，用csv模块或pandas也可）
3. 将读取的数组利用sklearn降维（此处的n_components表示降维后的维度，既然是二维图，所以是2）
4. 用matplotlib画散点图

代码：

import numpy as np
from sklearn.decomposition import PCA
import matplotlib.pyplot as plt

with open('samples.csv', encoding = 'utf-8') as f:
    data = np.loadtxt(f, float, delimiter = ",")
pca = PCA(n_components = 2)
new_data = pca.fit_transform(data)
label = ['Alex_scaled',
		'bao_scaled',
		'xiulan_scaled',
		'Robert_scaled',
		'Foxy_wg_scaled',
		'Haolin_scaled',
		'Leona_scaled',
		'penn_scaled',
		'shi_mf_scaled',
		'LeonaQi_scaled',
		'ENF','NEA','EEA']
for i in range(13):
    plt.scatter(new_data[i][0], new_data[i][-1], s = 40 ,label = label[i])
plt.legend(loc = 'best')
plt.show()

结果展示：

三维PCA散点图

---

步骤：

步骤与二维类似，只是需要把n_components改成3，然后matplotlib部分需要画三维图

代码：

import numpy as np
from sklearn.decomposition import PCA
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D

with open('samples.csv', encoding = 'utf-8') as f:
    data = np.loadtxt(f, float, delimiter = ",")
pca = PCA(n_components = 3)
new_data = pca.fit_transform(data)
label = ['Alex_scaled',
		'bao_scaled',
		'xiulan_scaled',
		'Robert_scaled',
		'Foxy_wg_scaled',
		'Haolin_scaled',
		'Leona_scaled',
		'penn_scaled',
		'shi_mf_scaled',
		'LeonaQi_scaled',
		'ENF','NEA','EEA']
fig = plt.figure()
ax = Axes3D(fig)
for i in range(13):
    ax.scatter(new_data[i][0], new_data[i][1], new_data[i][2], s = 40 ,label = label[i])
plt.legend(loc = 'best')
plt.show()

结果展示：

这时候估计有小伙伴就会说了：“这看着不膈应吗？”
所以这个时候…就需要获取每个点的坐标。

获取三维PCA坐标

---

话不多说，直接上代码：

代码：

import numpy as np
from sklearn.decomposition import PCA

with open('samples.csv', encoding = 'utf-8') as f:
    data = np.loadtxt(f, float, delimiter = ",")
pca = PCA(n_components = 3)
new_data = pca.fit_transform(data)
label = ['Alex_scaled',
		'bao_scaled',
		'xiulan_scaled',
		'Robert_scaled',
		'Foxy_wg_scaled',
		'Haolin_scaled',
		'Leona_scaled',
		'penn_scaled',
		'shi_mf_scaled',
		'LeonaQi_scaled',
		'ENF','NEA','EEA']

for i in range(13):
    print(f'{label[i]}:{new_data[i][0]},{new_data[i][1]},{new_data[i][2]}')

结果展示：

Alex_scaled:-0.011596488722119198,-0.00721039268000285,0.004327757597152235
bao_scaled:0.008957327165311457,0.03157204787328956,-0.012162365253179541
xiulan_scaled:-0.038263037685343365,-0.009423923747343946,0.009725629103724714
Robert_scaled:-0.028957259795948216,-0.024755176390934306,-0.009760572109930386
Foxy_wg_scaled:-0.05695083943398266,-0.039729613258914734,-0.014264926832824069
Haolin_scaled:-0.04331299000524389,-0.0002329777672932537,0.010328443388716371
Leona_scaled:-0.04650055185985425,0.0050266575879064045,0.003363030038096107
penn_scaled:-0.0464100888097683,-0.007910106367791488,-0.01116300525568197
shi_mf_scaled:-0.041521117170468275,-0.0019015953462750332,0.0030836026394054556
LeonaQi_scaled:-0.04471519716020188,0.0038319892067295172,0.002849597471080112
ENF:0.4295182339250529,-0.005889585418883738,0.0013563877559191943
NEA:-0.025410190741495663,0.0637904527870023,-0.0044124624225301305
EEA:-0.05483779970593805,-0.00716777647748854,0.016728883880051944