weixin_54097460: 如何绘制自己的pr曲线图呢

小皮每天进步一点点: 太厉害啦，小白福音~

kokolosaga: 修正了一下 from PIL import Image import numpy as np #读取原始数据 def read_image(): image = np.array(Image.open('C:/Users\gloriarains\Desktop\SOTY\soty\soty0000.bmp','r')) print(image.size) #Imagesize=image.size global length,width length=1600 width=1200 '''length=Imagesize[0] width=Imagesize[1]''' return image lst1 = read_image() #lst1=[] print(length,width) '''for i in range(width): lst1.append(lst[(length-1)*i:(length-1)*i+length-1]) ''' newimg = Image.new('RGB',(length,width)) '''for val,i in zip(lst1,range(0,1)): print(val," ",len(val)) print(i)''' #print(lst1[1279][719]) convert_data=np.ones((width,length,3)) for i in range(length): for j in range(width-1): convert_data[j][i]=list(lst1[i][j]) print(lst1[100][100]) #print(convert_data.size()) for i in range(width): for j in range(width-1): newimg.putpixel((i,j),(int(convert_data[i][j][0]),int(convert_data[i][j][1]),int(convert_data[i][j][2]))) newimg.show()

2301_78077620: 有一点：FP是假阳性，那么根据你前面 二、的内容，FP应该是瓜是好瓜，但是被预测为坏瓜吧？欢迎讨论

sigmoidAndRELU: 字字珠玑，写的很到位，把hook机制解释的清清楚楚，多多益善