Python's built-in plt is one of the most commonly used libraries for deep learning. It must be supported by pictures when publishing articles. plt is one of the necessary skills for deep learning. As an introduction to deep learning, you only need to master some basic drawing operations, and just look at the function API when you need it.
1 Import the plt library (the name is long, a bit hard to remember)
import matplotlib.pyplot as plt
Draw a picture at random first, save it and try the water:
plt.figure(figsize=(12,8), dpi=80)
plt.plot([1,2,6,4],[4,5,6,9])
plt.savefig('./plt_png/test1.png') #必须要放在show()前面，show()之后会自动释放图表内存
plt.show()
2 Create data and draw a standard graph
import random
#创建一个小时内温度随时间变化的曲线（以分钟计算，温度在15到20之间）
#1.准备数据
x = range(60)
y = [random.uniform(15,20)for i in x]
#2.创建画布
plt.figure( figsize=(12,8), dpi=80 )
plt.plot(x,y)
#3 设置刻度及步长
z = range(40)
x_label = ['11:{}'.format(i) for i in x]
plt.xticks( x[::5], x_label[::5])
plt.yticks(z[::5]) #5是步长
#4 添加网格信息
plt.grid(True, linestyle='--', alpha=0.5) #默认是True，风格设置为虚线，alpha为透明度
#5 添加标题（中文在plt中默认乱码，不乱码的方法在本文最后说明）
plt.xlabel('Time')
plt.ylabel('Temperature')
plt.title('Curve of Temperature Change with Time')
#6 保存图片，并展示
plt.savefig('./plt_png/test1.2.png')
plt.show()
3 Draw a graph of two curves
#3 创建另一个曲线的数据
y_another = [random.uniform(5,10)for i in x]
#输入两个曲线的信息
plt.figure( figsize=(12,8), dpi=80 )
plt.plot(x, y, color='r', linestyle='--', label = 'ShangHai')
plt.plot(x, y_another, color='g', linestyle='-.', label = 'BeiJing')
#显示图例
plt.legend() #默认loc=Best
#设置刻度及步长
z = range(40)
x_label = ['11:{}'.format(i) for i in x]
plt.xticks( x[::5], x_label[::5])
plt.yticks(z[::5]) #5是步长
#添加网格信息
plt.grid(True, linestyle='--', alpha=0.5) #默认是True，风格设置为虚线，alpha为透明度
#添加标题
plt.xlabel('Time')
plt.ylabel('Temperature')
plt.title('Curve of Temperature Change with Time')
plt.savefig('./plt_png/test1.3.png')
plt.show()
4 Draw different pictures in multiple drawing areas
#4 创建多个绘图区
#输入两个曲线的信息
#plt.figure( figsize=(12,8), dpi=80 )
figure, axes = plt.subplots( nrows=1, ncols=2, figsize=(12,8), dpi=80 )
axes[0].plot(x, y, color='r', linestyle='--', label = 'ShangHai')
axes[1].plot(x, y_another, color='g', linestyle='-.', label = 'BeiJing')
#显示图例
axes[0].legend() #默认loc=Best
axes[1].legend()
#设置刻度及步长
z = range(40)
x_label = ['11:{}'.format(i) for i in x]
#axes[0].set_xticks( x[::5], x_label[::5]) #set_xticks()不支持字符串，只支持布尔值
#axes[1].set_yticks(z[::5]) #5是步长 #set_yticks()不支持字符串，只支持布尔值
axes[0].set_xticks( x[::10]) #设置步长
axes[0].set_xticklabels(x_label[::5]) #设置字符串名
axes[0].set_yticks(z[::5])
axes[1].set_xticks( x[::10]) #设置步长
axes[1].set_xticklabels(x_label[::5]) #设置字符串名
axes[1].set_yticks(z[::5])
#添加网格信息
axes[0].grid(True, linestyle='--', alpha=0.5) #默认是True，风格设置为虚线，alpha为透明度
axes[1].grid(True, linestyle='--', alpha=0.5)
#添加标题
axes[0].set_xlabel('Time')
axes[0].set_ylabel('Temperature')
axes[0].set_title("Curve of ShangHai's Temperature Change with Time")
axes[1].set_xlabel('Time')
axes[1].set_ylabel('Temperature')
axes[1].set_title("Curve of BeiJing's Temperature Change with Time")
plt.savefig('./plt_png/test1.4.png')
plt.show()
5 Draw a graph of simple mathematical functions
import numpy as np
#准备数据
x = np.linspace(-1,1,1000) #在-1至1之间等距生成1000个数
y = 2*x*x
#绘制画布
#折线图plot
plt.figure(figsize=(12,8), dpi=80)
plt.plot( x, y, color='b', label='Y=2*X^2' )
plt.legend()
plt.grid(True, linestyle='--', alpha=0.5)
plt.savefig('./plt_png/test1.5.png')
plt.show()
6 Scatter chart, histogram, double histogram, histogram, pie chart
6.1 Scatter chart
#散点图scatter
x = np.linspace(-1,1,20)
y = 2*x*x
plt.figure(figsize=(12,8), dpi=80)
plt.scatter( x, y, color='b', label='Y=2*X^2' ) #主要是这里plot换成scatter，下面的根据需要修改
plt.legend()
plt.grid(True, linestyle='--', alpha=0.5)
plt.savefig('./plt_png/test1.6.png')
plt.show()
6.2 Histogram
#柱状图bar
x = range(8)
y = [100,200,400,350,450,350,250,300]
plt.figure(figsize=(20,8), dpi=80)
plt.bar( x, y, width=0.4, color=['r', 'k', 'y', 'g', 'b', 'c', 'm', 'k'], label='Bar_x_y' ) #这里是bar()函数
plt.legend()
plt.grid(True, linestyle='--', alpha=0.5)
#修改x刻度名字
plt.xticks(x, ['class1','class2','class3','calss4','class5','class6','class7','calss8'])
#设置xy标签
plt.xlabel('Classes')
plt.ylabel('Numbers')
plt.title('the Numbers of Classes')
plt.savefig('./plt_png/test1.7.png')
plt.show()
6.3 Contrast histogram (double histogram)
#对比柱状图bar
x = range(8)
y = [100,200,400,350,450,350,250,300]
z = [200,300,350,250,450,450,500,250]
plt.figure(figsize=(20,8), dpi=80)
plt.bar( x, y, width=0.2, color='r', label='Bar_1' )
#加一个柱状图，[i+0.2 for i in x]为间距生成式
plt.bar( [i+0.2 for i in x], z, width=0.2, color='b', label='Bar_2' )
plt.legend()
plt.grid(True, linestyle='--', alpha=0.5)
#修改x刻度名字
plt.xticks([i+0.1 for i in x], ['class1','class2','class3','calss4','class5','class6','class7','calss8'])
#设置xy标签
plt.xlabel('Classes')
plt.ylabel('Numbers')
plt.title('the Numbers of Classes')
plt.savefig('./plt_png/test1.8.png')
plt.show()
6.4 Histogram
#直方图histogram
x_1 = range(80)
number = np.random.randint(50,200,[80])
print(max(number),min(number))
bins = int((max(number) - min(number))//10)
plt.figure(figsize=(20,8), dpi=80)
plt.hist( number, bins=bins, color='b', label='Histogram' ) #y轴可以是频数，也可以是频率
plt.legend()
plt.grid(True, linestyle='--', alpha=0.5)
plt.xticks(range(min(number), max(number), 10))
plt.savefig('./plt_png/test1.9.png')
plt.show()
6.5 Pie Chart
#饼图pie ，超过9个类别不适合用饼图
import matplotlib.pyplot as plt
x3 = range(8)
y3 = [100,200,400,350,450,350,250,300]
z3 = ['class1','class2','class3','calss4','class5','class6','class7','calss8']
plt.figure(figsize=(20,8), dpi=80)
plt.pie( y3, labels=z3, colors=['r', 'k', 'y', 'g', 'b', 'c', 'm', 'k'], autopct='%1.2f%%' )
plt.legend()
plt.title('the Numbers of Classes')
plt.axis('equal') #调节饼图长轴短轴的比例，以及整图布局
plt.savefig('./plt_png/test1.10.png')
plt.show()
Note: plt can not display the solution in Chinese
Add the following two lines of code when importing plt
import matplotlib.pyplot as plt
plt.rcParams['font.sans-serif'] = ['SimHei'] #显示中文
plt.rcParams['axes.unicode_minus']=False #用来正常显示负号