4

您好我正在尝试使用以下数据绘制召回精度曲线:

      Recall    Precision
0.88196 0.467257
0.898501    0.468447
0.89899 0.470659
0.900789    0.471653
0.900922    0.472038
0.901012    0.472359
0.901345    0.480144
0.901695    0.482353
0.902825    0.482717
0.903261    0.483125
0.905152    0.483621
0.905575    0.485088
0.905682    0.486339
0.906109    0.488117
0.906466    0.488459
0.90724 0.488587
0.908989    0.488875
0.909941    0.489362
0.910125    0.489493
0.910314    0.490196
0.910989    0.49022
0.91106 0.490786
0.911137    0.496624
0.91129 0.496891
0.911392    0.497301
0.911392    0.499379
0.911422    0.5
0.911452    0.503783
0.911525    0.515829

源代码:

import random
import pylab as pl
from sklearn import svm, datasets
from sklearn.metrics import precision_recall_curve
from sklearn.metrics import auc

##Load Recall
fname = "recall.txt"
fname1 = "precision.txt"

recall = []
precision = []

with open(fname) as inf:
    for line in inf:
        recall.append(float(line))

with open(fname1) as inf:
    for line in inf:
        precision.append(float(line))

area = auc(recall, precision)
print("Area Under Curve: %0.2f" % area)

pl.clf()
pl.plot(recall, precision, label='Precision-Recall curve')
pl.xlabel('Recall')
pl.ylabel('Precision')
pl.ylim([0.0, 1.05])
pl.xlim([0.0, 1.0])
pl.title('Precision-Recall example: AUC=%0.2f' % area)
pl.legend(loc="lower left")
pl.show()

我得到 AUC = 0.01 下的面积是正常的吗?

在此处输入图像描述

4

1 回答 1

2

这似乎是正确的答案。

使用numpy.trapz(precission, recall)我得到AUC = 0.014036223712000031

于 2013-12-20T15:28:53.353 回答