python画混淆矩阵_matplotlib画混淆矩阵和正确率曲线

今天很兴奋，再来一篇庆祝一下~~~

混淆矩阵

找不到参看的那篇博客啦~~希望原博主不要讨伐我

#!/usr/bin/python3.5

# -*- coding: utf-8 -*-

import numpy as np

import matplotlib.pyplot as plt

plt.rcParams['font.sans-serif'] = ['FangSong'] #可显示中文字符

plt.rcParams['axes.unicode_minus']=False

classes = ['a','b','c','d','e','f','g']

confusion_matrix = np.array([(99,1,2,2,0,0,6),(1,98,7,6,2,1,1),(0,0,86,0,0,2,0),(0,0,0,86,1,0,0),(0,0,0,1,94,1,0),(0,1,5,1,0,96,8),(0,0,0,4,3,0,85)],dtype=np.float64)

plt.imshow(confusion_matrix, interpolation='nearest', cmap=plt.cm.Oranges) #按照像素显示出矩阵

plt.title('混淆矩阵')

plt.colorbar()

tick_marks = np.arange(len(classes))

plt.xticks(tick_marks, classes, rotation=-45)

plt.yticks(tick_marks, classes)

thresh = confusion_matrix.max() / 2.

#iters = [[i,j] for i in range(len(classes)) for j in range((classes))]

#ij配对，遍历矩阵迭代器

iters = np.reshape([[[i,j] for j in range(7)] for i in range(7)],(confusion_matrix.size,2))

for i, j in iters:

plt.text(j, i, format(confusion_matrix[i, j]),fontsize=7) #显示对应的数字

plt.ylabel('真实类别')

plt.xlabel('预测类别')

plt.tight_layout()

plt.show()

正确率曲线

fig ,ax= plt.subplots()

plt.plot(np.arange(iterations), fig_acc，'b')

plt.plot(np.arange(iterations), fig_realacc, 'r')

ax.set_xlabel('迭代次数')

ax.set_ylabel('正确率(%)')

labels = ["训练正确率", "测试正确率"]

# labels = [l.get_label() for l in lns]

plt.legend( labels, loc=7)

plt.show()

睡啦睡啦~~