目录

一. 数据形式 (输入数据)

二.绘图(完整代码)

三. plt.plot() 函数 (调整图形)

1.plt.plot(x, y)

2.plt.plot(x, y, “格式控制字符串”)

2.1 “颜色”与”线型”

2.2″点型”

3. plt.plot(x, y, “格式控制字符串”, 关键字=参数)


一. 数据形式 (输入数据)

训练过程中每个epoch都输出当前轮结果,输出数据保存在.txt文件,形式如下:

因为只是举个例子,只用30张图跑了5个epoch,不过数值不重要!过程先搞明白。

#每个epoch都输出当前轮结果print("epoch[%d/%d],train_loss,%.4f,train_acc,%.4f,train_miou,%.4f,eval_loss,%.4f,eval_acc,%.4f,eval_miou,%.4f,lr,%.6f,time,%ds" %(epoch + 1, EPOCHES, epoch_loss, epoch_acc, epoch_miou, val_loss, val_acc, val_miou, learning_rate, time.time() - st_epoch))#输出结果形式epoch[1/5],train_loss,3.0900,train_acc,0.0190,train_miou,0.0046,eval_loss,3.0398,eval_acc,0.0438,eval_miou,0.0075,lr,0.000005,time,23sepoch[2/5],train_loss,2.9437,train_acc,0.0667,train_miou,0.0090,eval_loss,2.9367,eval_acc,0.1133,eval_miou,0.0082,lr,0.000005,time,25sepoch[3/5],train_loss,2.8345,train_acc,0.1875,train_miou,0.0141,eval_loss,2.8379,eval_acc,0.2588,eval_miou,0.0087,lr,0.000005,time,25sepoch[4/5],train_loss,2.7256,train_acc,0.3285,train_miou,0.0163,eval_loss,2.7290,eval_acc,0.4454,eval_miou,0.0072,lr,0.000005,time,24sepoch[5/5],train_loss,2.6142,train_acc,0.5298,train_miou,0.0167,eval_loss,2.6255,eval_acc,0.5811,eval_miou,0.0032,lr,0.000005,time,27s

二.绘图(完整代码)

  1. 逐行读取数据
  2. split()按所给符号分割
  3. 把所需列的数据按索引添加到对应的表里
  4. 用plt画图,需要什么形式的图,函数的参数可以参考第三部分函数介绍自行更改
#完整代码import matplotlib.pyplot as pltfile = open('log.txt')#打开文档lines = file.readlines() #读取文档数据#epoch = list(1, range(len(lines))+1) #epoch可以直接赋值,不放心的就用下面epoch的代码epoch = []train_loss = []val_loss = []for line in lines:# split用于将每一行数据用自定义的符号(我用的是逗号)分割成多个对象# 取分割后的第0列,转换成float格式后添加到epoch列表中epoch.append(str(line.split(',')[0]))# 取分割后的第2列,转换成float格式后添加到train_loss列表中train_loss.append(float(line.split(',')[2]))#取分割后的第8列,转换成float格式后添加到val_loss列表中val_loss.append(float(line.split(',')[8]))plt.figure() plt.title('loss during training')#标题plt.plot(epoch, train_loss, label="train_loss")plt.plot(epoch, val_loss, label="valid_loss")plt.legend()plt.grid()plt.show()

输出结果:

三. plt.plot() 函数 (调整图形)

matplotlib.pyplot模块下的一个函数,用于画图。它可以绘制点和线, 并且对其样式进行控制。

函数定义为plt.plot(*args, **kwargs)

import matplotlib.pyplot as plthelp(plt.plot) # 查看英文函数定义
  • *args, 可变位置参数, 以元组形式存放了很多无名参数.
  • **kwargs, 可变关键字参数, 以字典形式存放了很多关键字及参数. 调用时可传入
  • *args要放在**kwargs之前.
  • 本函数的*args, 允许传入多对x和y和一个可选的”格式控制字符串”.
  • 本函数的**kwargs, 允许传入多个可选的关键字参数.

1.plt.plot(x, y)

  • x为x轴数据, y为y轴数据。
  • x, y可传入(元组), [列表], np.array, pd.Series。
  • 也可同时传入多组。
import matplotlib.pyplot as pltimport numpy as npimport pandas as pd#示例一:x为x轴数据, y为y轴数据x=[3,4,5] # [列表]y=[2,3,2] # x,y元素个数N应相同plt.plot(x,y)plt.show()#示例二:x, y可传入(元组), [列表], np.array, pd.Seriesx=(3,4,5) # (元组)y1=np.array([3,4,3]) # np.arrayy2=pd.Series([4,5,4]) # pd.Seriesplt.plot(x,y1)plt.plot(y2)# x可省略,默认[0,1..,N-1]递增plt.show() # plt.show()前可加多个plt.plot(),画在同一张图上#示例三:可传入多组x, yx=(3,4,5)y1=np.array([3,4,3])y2=pd.Series([4,5,4])plt.plot(x,y1,x,y2) # 此时x不可省略plt.show()

结果示例:

2.plt.plot(x, y, “格式控制字符串”)

点和线的格式可以用“格式控制字符串”设置,”最多可以包括三部分, “颜色“, “点型“, “线型“。

2.1 “颜色”与”线型”

如果只控制”颜色”, 格式控制字符串还可以输入英文全称, 如”red”, 甚至是十六进制RGB字符串, 如”#FF0000″. python可用的”颜色”大全

============================================charactercolor============================================``'b'``blue 蓝``'g'``green 绿``'r'``red 红``'c'``cyan 蓝绿``'m'``magenta 洋红``'y'``yellow 黄``'k'``black 黑``'w'``white 白============================================
============================================characterdescription============================================``'-'``solid line style 实线``'--'`` dashed line style 虚线``'-.'`` dash-dot line style 点画线``':'``dotted line style 点线============================================
#示例import numpy as npimport pandas as pdimport matplotlib.pyplot as pltcolor=['b','g','r','c','m','y','k','w']linestyle=['-','--','-.',':']dic1=[[0,1,2],[3,4,5]]x=pd.DataFrame(dic1)dic2=[[2,3,2],[3,4,3],[4,5,4],[5,6,5]]y=pd.DataFrame(dic2)# 循环输出所有"颜色"与"线型"for i in range(2):for j in range(4):plt.plot(x.loc[i],y.loc[j],color[i*4+j]+linestyle[j]) plt.show()

输出结果:

2.2“点型”

============================================characterdescription============================================``'.'``point marker``','``pixel marker``'o'``circle marker``'v'``triangle_down marker``'^'``triangle_up marker``''``triangle_right marker``'1'``tri_down marker``'2'``tri_up marker``'3'``tri_left marker``'4'``tri_right marker``'s'``square marker``'p'``pentagon marker``'*'``star marker``'h'``hexagon1 marker``'H'``hexagon2 marker``'+'``plus marker``'x'``x marker``'D'``diamond marker``'d'``thin_diamond marker``'|'``vline marker``'_'``hline marker============================================
#示例import numpy as npimport pandas as pdimport matplotlib.pyplot as pltmarker=['.',',','o','v','^','','1','2','3','4','s','p','*','h','H','+','x','D','d','|','_','.',',']dic1=[[0,1,2],[3,4,5],[6,7,8],[9,10,11],[12,13,14],[15,16,17]]x=pd.DataFrame(dic1)dic2=[[2,3,2.5],[3,4,3.5],[4,5,4.5],[5,6,5.5]]y=pd.DataFrame(dic2)# 循环输出所有"点型"for i in range(6):for j in range(4):plt.plot(x.loc[i],y.loc[j],"b"+marker[i*4+j]+":") # "b"蓝色,":"点线plt.show()

3. plt.plot(x, y, “格式控制字符串”, 关键字=参数)

本函数的**kwargs, 允许传入多个可选的关键字参数

#示例y=[2,3,2,4,5] # 青色,线宽10,星星,点尺寸50,点填充绿色,点边缘宽度6,点边缘青色plt.plot(y,color="c",linewidth=10,marker="*",markersize=20, markerfacecolor="g",markeredgewidth=3,markeredgecolor="c")plt.show()

函数用法是总结的这篇文章和matplotlib.pyplot.plot()参数详解两篇

python中的颜色表