ROC曲线绘制和求分类四个指标（F1,recall,acc,pre）

简介

都用库函数实现,首先导入包。

import numpy as np
from sklearn.metrics import roc_curve, f1_score, recall_score, accuracy_score, precision_score

ROC

fpr, tpr, thresholds = roc_curve(true_labels, predicted_probabilities)
# 假设你有模型的预测概率和真实标签
predicted_probabilities = ...  # 模型的预测概率
true_labels = ...  # 真实标签

注意这里的predicted_probabilities是预测为正例的概率。不是0，1标签，也不是两个概率里的max值！

计算完后，fpr，tpr就是坐标数组。
之后使用使用 fpr、tpr 绘制 ROC 曲线即可。

# 绘制ROC曲线
plt.plot(fpr, tpr, label='ROC curve')
plt.plot([0, 1], [0, 1], 'k--')  # 绘制对角线
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('Receiver Operating Characteristic (ROC)')
plt.legend(loc='lower right')
plt.show()

四个指标（F1,recall,acc,pre）

# 计算F1得分
predicted_labels = (predicted_probabilities > 0.5).astype(int)  # 根据预测概率确定类别
f1 = f1_score(true_labels, predicted_labels)

# 计算召回率
recall = recall_score(true_labels, predicted_labels)

# 计算准确率
accuracy = accuracy_score(true_labels, predicted_labels)

# 计算正确率
precision = precision_score(true_labels, predicted_labels)

# 打印结果
print("F1 Score:", f1)
print("Recall:", recall)
print("Accuracy:", accuracy)
print("Precision:", precision)

注意，这里predicted_labels是0，1标签，不是概率！

完整代码

import numpy as np
from sklearn.metrics import roc_curve, f1_score, recall_score, accuracy_score, precision_score

# 假设你有模型的预测概率和真实标签
predicted_probabilities = ...  # 模型的预测概率
true_labels = ...  # 真实标签

# 计算ROC曲线
fpr, tpr, thresholds = roc_curve(true_labels, predicted_probabilities)

# 绘制ROC曲线
plt.plot(fpr, tpr, label='ROC curve')
plt.plot([0, 1], [0, 1], 'k--')  # 绘制对角线
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('Receiver Operating Characteristic (ROC)')
plt.legend(loc='lower right')
plt.show()

# 计算F1得分
predicted_labels = (predicted_probabilities > 0.5).astype(int)  # 根据预测概率确定类别
f1 = f1_score(true_labels, predicted_labels)

# 计算召回率
recall = recall_score(true_labels, predicted_labels)

# 计算准确率
accuracy = accuracy_score(true_labels, predicted_labels)

# 计算正确率
precision = precision_score(true_labels, predicted_labels)

# 打印结果
print("F1 Score:", f1)
print("Recall:", recall)
print("Accuracy:", accuracy)
print("Precision:", precision)