Python 如何在决策树中计算精确召回率？

我尝试在标准数据集“iris.csv”中预测

我有一个错误

File "C:/Users/Ultra/PycharmProjects/poker_ML/decision_tree.py", line 20, in <module>
    average_precision = average_precision_score(testY, y_score)
  File "C:\Users\Ultra\PycharmProjects\poker_ML\venv\lib\site-packages\sklearn\metrics\ranking.py", line 241, in average_precision_score
    average, sample_weight=sample_weight)
  File "C:\Users\Ultra\PycharmProjects\poker_ML\venv\lib\site-packages\sklearn\metrics\base.py", line 74, in _average_binary_score
    raise ValueError("{0} format is not supported".format(y_type))
ValueError: multiclass format is not supported

文件“C:/Users/Ultra/PycharmProjects/poker\u ML/decision\u tree.py”，第20行，在
平均精度=平均精度分数（测试，y分数）
文件“C:\Users\Ultra\PycharmProjects\poker\ML\venv\lib\site packages\sklearn\metrics\ranking.py”，第241行，平均精度分数
平均值，样本重量=样本重量）
文件“C:\Users\Ultra\PycharmProjects\poker\ML\venv\lib\site packages\sklearn\metrics\base.py”，第74行，二进制平均分数
raise VALUERROR（“{0}格式不受支持”。格式（y_类型））
ValueError:不支持多类格式

如何计算3类的精确召回率？ sklearn中决策树的精确召回工作原理。也许我在计算“y_分数”时出错了？

根据

平均精度_分数

无法处理多类分类

相反，您可以这样使用：

# Decision tree
...
y_pred = decision.predict(testX)
y_score = decision.score(testX, testY)
print('Accuracy: ', y_score)

# Compute the average precision score
from sklearn.metrics import precision_score
micro_precision = precision_score(y_pred, testY, average='micro')
print('Micro-averaged precision score: {0:0.2f}'.format(
      micro_precision))

macro_precision = precision_score(y_pred, testY, average='macro')
print('Macro-averaged precision score: {0:0.2f}'.format(
      macro_precision))

per_class_precision = precision_score(y_pred, testY, average=None)
print('Per-class precision score:', per_class_precision)

请注意，您需要指定如何平均分数。如果您的数据集显示标签不平衡（而

iris

没有），这一点尤其重要

# Decision tree
...
y_pred = decision.predict(testX)
y_score = decision.score(testX, testY)
print('Accuracy: ', y_score)

# Compute the average precision score
from sklearn.metrics import precision_score
micro_precision = precision_score(y_pred, testY, average='micro')
print('Micro-averaged precision score: {0:0.2f}'.format(
      micro_precision))

macro_precision = precision_score(y_pred, testY, average='macro')
print('Macro-averaged precision score: {0:0.2f}'.format(
      macro_precision))

per_class_precision = precision_score(y_pred, testY, average=None)
print('Per-class precision score:', per_class_precision)