使用Python，PCA对iris数据集降维2维/3维并进行2D，3D散点图绘制（包括有图例&无图例，有标题Label&无标题Label）

这篇博客将介绍如何使用Python，PCA对iris数据集降维2，3并进行2D,3D散点图绘制（包括有图例&无图例，有标题Label&无标题Label）。
着重介绍怎么一次性添加多类型的图例到图表，通过显式获取scatter。

# 方法1
scatter = ax.scatter(x_reduced[:, 0], x_reduced[:, 1], x_reduced[:, 2], c=iris.target)
# 方法2
scatter = plt.scatter(x_reduced[:, 0], x_reduced[:, 1], c=iris.target, marker="d")
# 添加图例名称到图标
plt.legend(handles=scatter.legend_elements()[0],
           labels=sp_names,
           title="species", loc="upper right")
1
2
3
4
5
6
7
8

1. 效果图

对鸢尾花进行PCA降维成2维后进行绘制，Seaborn效果图如下：

对鸢尾花进行PCA降维成2维后进行绘制，Seaborn添加标题及散点拟合线 效果图如下：

对鸢尾花进行PCA降维成3维后进行绘制，Matplotlib3D效果图如下：

对鸢尾花进行PCA降维成3维简单绘制效果图如下：

对鸢尾花进行PCA降维成3维后进行绘制，添加中文标题，xyz轴描述及图例，不同类型用不同的样式mark 效果图如下：

2. 源码

# 使用PCA对鸢尾花特征进行降维2维/3维并绘制 2D,3D散点图（有图例&无图例，有标题Lable&无标题Lable）
# python iris_pca.py
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns
from mpl_toolkits.mplot3d import Axes3D
from sklearn import datasets
from sklearn.decomposition import PCA

# 支持中文
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False

iris = datasets.load_iris()
R = np.array(iris.data)

R_cov = np.cov(R, rowvar=False)

iris_covmat = pd.DataFrame(data=R_cov, columns=iris.feature_names)
iris_covmat.index = iris.feature_names
eig_values, eig_vectors = np.linalg.eig(R_cov)


def plot_2D_Seaborn():
    featureVector = eig_vectors[:, :2]
    featureVector_t = np.transpose(featureVector)
    # R is the original iris dataset
    R_t = np.transpose(R)
    newDataset_t = np.matmul(featureVector_t, R_t)
    newDataset = np.transpose(newDataset_t)

    # 可视化 绘制2D图
    # 鸢尾花数据创建dataframe
    df = pd.DataFrame(data=newDataset, columns=['PC1', 'PC2'])
    y = pd.Series(iris.target)
    y = y.replace(0, 'setosa')
    y = y.replace(1, 'versicolor')
    y = y.replace(2, 'virginica')
    df['Target'] = y
    # 绘制2维数据，fit_reg是否拟合线
    sns.lmplot(x='PC1', y='PC2', data=df, hue='Target', fit_reg=False, legend=True)
    # sns.lmplot(x='PC1', y='PC2', data=df, hue='Target', fit_reg=True, legend=True)
    plt.title("Seaborn鸢尾花降维2维图")  # 会被截取不全
    plt.show()


# PCA降2维
# 无图例 & 有图例
def plot_2D_PCA_Legend():
    # 进行PCA降维
    x_reduced = PCA(n_components=2).fit_transform(iris.data)
    y = pd.Series(iris.target)
    y = y.replace(0, 'setosa')
    y = y.replace(1, 'versicolor')
    y = y.replace(2, 'virginica')

    fig = plt.figure()
    ax = fig.add_subplot()
    # 2D散点图无图例
    ax.scatter(x_reduced[:, 0], x_reduced[:, 1], c=iris.target, marker="d")
    plt.show()

    print(np.unique(np.array(y)).tolist())
    print(np.unique(iris.target).tolist())
    # ax.legend(np.unique(iris.target).tolist())
    # ax.legend(np.unique(np.array(y)).tolist(),loc='upper right')

    # 2D散点图有标题，label，图例
    plt.figure(figsize=(8, 6))
    sp_names = np.unique(np.array(y)).tolist()
    scatter = plt.scatter(x_reduced[:, 0], x_reduced[:, 1], c=iris.target, marker="d")
    plt.title('鸢尾花降维2维图')
    plt.xlabel("PC1", size=18)
    plt.ylabel("PC2", size=18)
    # 添加图例名称到图标
    plt.legend(handles=scatter.legend_elements()[0],
               labels=sp_names,
               title="species", loc="upper right")
    plt.show()


# PCA降3维
# 无图例 & 有图例
def plot_3D_PCA_Legend():
    # 进行PCA降维
    x_reduced = PCA(n_components=3).fit_transform(iris.data)
    y = pd.Series(iris.target)
    y = y.replace(0, 'setosa')
    y = y.replace(1, 'versicolor')
    y = y.replace(2, 'virginica')

    fig = plt.figure()
    ax = Axes3D(fig)
    # 3D散点图无图例
    ax.scatter(x_reduced[:, 0], x_reduced[:, 1], x_reduced[:, 2], c=iris.target)
    plt.show()

    sp_names = np.unique(np.array(y)).tolist()
    fig = plt.figure()
    ax = Axes3D(fig)
    # 3D散点图有标题，label，图例
    scatter = ax.scatter(x_reduced[:, 0], x_reduced[:, 1], x_reduced[:, 2], c=iris.target)
    ax.set_title('鸢尾花降维3维图')
    ax.set_xlabel("PC1", size=18)
    ax.set_ylabel("PC2", size=18)
    ax.set_zlabel("PC3", size=18)
    # 添加图例名称到图标
    ax.legend(handles=scatter.legend_elements()[0],
              labels=sp_names,
              title="species", loc="upper right")
    plt.show()


# 不同的PCA降维
def plot_3D():
    featureVector = eig_vectors[:, :3]
    featureVector_t = np.transpose(featureVector)
    # R is the original iris dataset
    R_t = np.transpose(R)
    newDataset_t = np.matmul(featureVector_t, R_t)
    newDataset = np.transpose(newDataset_t)

    # 构建DataFrame
    df = pd.DataFrame(data=newDataset, columns=['PC1', 'PC2', 'PC3'])
    y = pd.Series(iris.target)
    y = y.replace(0, 'setosa')
    y = y.replace(1, 'versicolor')
    y = y.replace(2, 'virginica')
    df['Target'] = y

    # print(df.head(5))
    # 根据其中一列分组
    df = df.groupby("Target")
    # print(df.groups)  # key
    # print(df.get_group('setosa'))  # 某个value

    # matplot支持的点样式及点颜色
    marks = ['.', '*', 's', ',', 'o', 'v', '^', '<', '>', '1', '2', '3', '4', '8', 'p', 'P', 'h', 'H', '+', 'x', 'X',
             'D', 'd', '|', '_']
    colors = ['b', 'g', 'r', 'c', 'm', 'y', 'k', 'w']

    fig = plt.figure()
    ax = fig.add_subplot(projection='3d')
    ax.set_title('鸢尾花降维3维图')
    for i, key in enumerate(df.groups.keys()):
        val = df.get_group(key)
        ax.scatter(val["PC1"], val["PC2"], val["PC3"], c=colors[i % len(colors)], marker=marks[i % len(marks)],
                   label=key)
    ax.set_xlabel('PC1')
    ax.set_ylabel('PC2')
    ax.set_zlabel('PC3')
    ax.legend(loc='upper right')
    plt.show()


plot_2D_Seaborn()
plot_2D_PCA_Legend()
plot_3D_PCA_Legend()
plot_3D()

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161

参考

• https://datavizpyr.com/add-legend-to-scatterplot-colored-by-a-variable-with-matplotlib-in-python/
• https://blog.csdn.net/weixin_39531761/article/details/110717013
• https://www.jianshu.com/p/25a66dee6450