python画P-R曲线，并计算AUPR值-实战代码

小小的马儿

已于 2022-06-19 22:53:06 修改

阅读量3.5k

点赞数

分类专栏：数据处理文章标签：可视化 python pandas numpy 机器学习

于 2021-04-26 16:02:47 首次发布

本文链接：https://blog.csdn.net/qq_41011835/article/details/116159383

版权

数据处理专栏收录该内容

4 篇文章 0 订阅

订阅专栏

def draw_PR(dictNameYtruePred,title=None,savePath = './picture/',dpi = 600):
    '''
    :param dictNameYtruePred:
    :param title:
    :param savePath:
    :param dpi:
    #输入实例
        #     label = [1, 1, 0, 0,0,0,0, ] # y 是 a的值，x是各个元素的索引
        #     probability = [0.8, 0.3, 0.1, 0.2,0.3,0.4,0.6,]
        #     draw_PR({'name': [label, probability]})
    '''
    colors = cycle(['sienna','seagreen','blue', 'red', 'darkorange',  'gold', 'orchid',  'gray'])
    plt.figure()  # 指定大小分辨率等，相当于画纸
    for toolsName,y_trueANDpred_proba ,color in zip(dictNameYtruePred.keys(),dictNameYtruePred.values(),colors):
        precision, recall, thresholds = precision_recall_curve(y_trueANDpred_proba[0], y_trueANDpred_proba[1], pos_label=1)
        aupr = auc( recall,precision)#计算面积的
        plt.plot(precision, recall, color=color, label=str(toolsName)+'(AUPR = %0.3f)' % aupr, linestyle='--', LineWidth=3)#横纵坐标的取值，颜色样式等

    plt.xlim([-0.05, 1.05])#x坐标轴
    plt.ylim([-0.05, 1.05])#y坐标轴
    plt.xlabel('Recall', size=25)#x标签
    plt.ylabel('Precision', size=25)#y标签
    plt.legend(loc="lower right", ncol=1, fontsize=22)#加上图例
    plt.title(title, size=25)
    if not os.path.exists(savePath):
        os.makedirs(savePath)
    plt.savefig(savePath+str(1), dpi=dpi)
    plt.show()
if __name__ == '__main__':
    label1 = [1,1, 1, 1, 1,1,1 ,1,1,0, 0, 0, 0, 0, ]
    probability1 = [0.1,0.8, 0.3, 0.1, 0.6,0.5,0.8,0.9,0.4 ,0.2,0.3,0.4,0.6,0.1]
    label2 = [1, 1, 1, 1,1,1 ,1,1,0, 0, 0, 0, 0, 0,0,]
    probability2 = [0.9, 0.3, 0.8, 0.6,0.5,0.8,0.9,0.4 ,0.2,0.3,0.4,0.6,0.1,0.3,0.9]
    draw_PR({'name1': [label1, probability1],
             'name2': [label2, probability2]

             })

注意导入所需要的对应的包。

结果：因为数据量较少，所以显示的粗糙