sklearn.plot_tree 如何可视化分类任务的 class_labels？

Posted 2023-03-12

【中文标题】sklearn.plot_tree 如何可视化分类任务的 class_labels？

【英文标题】：sklearn.plot_tree how to visualize class_labels for classification task?

【发布时间】：2022-01-13 08:49:42

【问题描述】：

我最近在决策树上完成了我的试用代码。它工作得很好，除了一件事。正在绘制的树中不包含类名。我是不是做错了什么？

请看下面的代码和数据集的图片。

#Import Data#

import pandas as pd

data_set = pd.read_excel(r"C:\Users\User\Desktop\Tree.xlsx")

print(data_set.head())

#Set Features and Training Targets#

features_names=["Money","Debt"]
target_names=["Mood1", "Mood2", "Mood3"]

features = data_set[features_names]
targets = data_set[target_names]

print(features)
print(targets)

#Set Training Set and Test Set#

train_features = features[:10]
train_targets = targets[:10]

test_features = features[10:]
test_targets = targets[10:]

print (train_features)
print (train_targets)

print(test_features)
print(test_targets)

#Estimating Tree#

from sklearn.tree import DecisionTreeRegressor

dt = DecisionTreeRegressor(max_depth = 3)
dt = dt.fit(train_features, train_targets)

print(dt.score(train_features, train_targets))
print(dt.score(test_features, test_targets))

#Plotting the Tree#

from sklearn import tree
import matplotlib.pyplot as plt

tree.plot_tree(dt, feature_names=features_names, class_names=target_names, filled = True)
plt.show()

【参考方案1】：

在回归任务中，可视化标签可能不起作用； documentation 声明 class_name 参数是“仅与分类相关”。

在这种情况下，您的目标变量Mood 可以是分类的，在单个列中表示它的值。完成后，您可以设置

tree.plot_tree(clf, class_names=True)

用于类名的符号表示

或

class_names = ['setosa', 'versicolor', 'virginica']

tree.plot_tree(clf, class_names=class_names)

具体的类名。

完整示例

import numpy as np
from matplotlib import pyplot as plt

from sklearn.model_selection import train_test_split
from sklearn.datasets import load_iris
from sklearn.tree import DecisionTreeClassifier
from sklearn import tree

iris = load_iris()
X = iris.data
y = iris.target
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)

clf = DecisionTreeClassifier(max_leaf_nodes=3, random_state=0)
clf.fit(X_train, y_train)

# Symbolic class name representation
tree.plot_tree(clf, class_names=True)

# Specific class name representation
class_names = iris['target_names']

tree.plot_tree(clf, class_names=class_names)

【讨论】：

正确，但 OP 使用 DT regressor，而不是分类器。如果这也适用于回归设置，请做出相应的说明。

好吧，我已经开始对原始帖子发表评论说“此选项不可用于回归”，但后来我意识到我对此不确定，所以我放弃评论；但看来您确实已经确认了:) 我建议您使用此说明更新您的帖子（包括相关的文档链接）。