1. Preface

This article is about the learning and using of Matplotlib library, including the introduction, installation and drawing of Matplotlib. It can also be used as the manual of Matplotlib drawing, which can be viewed by browsing when necessary.

2. Introduction to Matplotlib

Inspired by Matlab (drawing way is very similar), Matplotlib is composed of various visualization classes, and its internal structure is very complex. In order to facilitate the user's use, Matplotlib.pyplot is provided as a sub Library of python drawing, which provides commands for drawing all kinds of visualization graphics.

In general, we import pyplot as follows

import matplotlib.pyplot as plt

Some of the figures drawn by Matplotlib are shown below:

3. installation

The installation of Matplotlib is very simple. As long as the python environment and pip installation tools are installed locally, you can install them with the following command:

pip install matplotlib

4. Drawing basis

4.1 plot() function

plot(x,y,format_string,**kwargs) is the basic function of pyplot plot plot. The meaning of each parameter is as follows:

• x:X-axis data, list or array (single curve optional)
• y:Y-axis data, list or array
• Format string: control the format string of the curve, which is used to fill in the curve information (optional), including color, style and mark three characters. The corresponding optional parameters are as follows:

Character type	parameter	Explain
Color character	'b'	blue
	'g'	green
	'r'	Red
	'c'	Cyan green
	'm'	Magenta
	'y'	yellow
	'k'	black
	'w'	white
	'#000000'	Color acquisition by RGB
	'0.6'	Gray value
Style character	'-'	Solid line
	'--'	Broken line
	'-.'	Dot marking
	':'	Dotted line
	'' ''	Wireless bar
Marked character	'.'	Dot marker
	'o'	Solid circle mark
	'^'	Upper triangle mark

import matplotlib.pyplot as plt
import numpy as np

x = np.arange(10)
plt.plot(x,x*2,'b-.',x,x*3,'r*',x,x*4,'y-.^')  # 1: Blue Solid dot mark, 2: Red ` * 'mark, 3: yellow dotted triangle mark
plt.show()

• **kwargs: more curve parameters (x,y,format_string)

4.1.1 Chinese display

The pyplot label does not support Chinese fonts. You can modify the fonts by adding the fontproperties property to the label. fontsize sets the font size. First, you need to ensure that there is a corresponding font. If you are not prompted, you need to download and install it. For details, please refer to This article.

The Chinese display example is as follows:

import matplotlib.pyplot as plt
import numpy as np

x = np.arange(10)
plt.xlabel('I am a horizontal axis.',fontproperties='Simhei',fontsize=10)
plt.ylabel('I am a vertical axis.',fontproperties='Simhei',fontsize=10)
plt.plot(x,x*2,'b-.')
plt.show()

plt.rcParams['font.sans-serif'] = ['SimHei']

4.1.2 text display

In addition to drawing, we also need to make some explanatory text information for the image, generally including the following text display functions:

function	Explain
plt.xlabel()	x axis text
plt.ylabel()	y axis text
plt.title()	Image title text
plt.text()	Add text anywhere
plt.annotate()	Add notes with arrows

Here is a more detailed example:

import numpy as np
import matplotlib.pyplot as plt

x = np.arange(0,10,0.1)
plt.xlabel('I am a horizontal axis.',fontsize=15,color='r')
plt.ylabel('I am a vertical axis.',fontsize=15,color='b')
plt.title('Sinusoidal curve $y=sin(2\pi x)$',fontsize=12)
plt.text(5,1,'(5,1)',fontsize=12)  # Add text information at (5,1)
plt.annotate('Look here',xy=(3,-1),xytext=(4,-1.5),  
             arrowprops={'facecolor':'k','shrink':0.1,'width':1.5})  # Set indicator arrow
plt.ylim(-2,2)  # Set the upper and lower limit of y-axis

plt.plot(x,np.sin(2*np.pi*x),'b-.')
plt.grid(True)  # show grid
plt.show()

5. Drawing of common charts

The previous article has explained the general curve drawing, this chapter describes several other commonly used chart drawing methods, each chart will give a simple drawing example (for quick query drawing), and then explain the function parameter information in detail and give a richer drawing method and effect.

5.1 histogram

5.1.1 simple example

• Effect
  
• Code

  import matplotlib.pyplot as plt
  import numpy as np
  
  x = np.random.randint(0,50,50)  
  bins = np.arange(0,51,10)  # Set histogram distribution interval [0-50], 10 as spacing
  
  plt.hist(x,bins,color="darkgreen")
  plt.title('Student achievement distribution chart',fontproperties='SimHei', fontsize=15)
  plt.xlabel('achievement',fontproperties='SimHei', fontsize=15)
  plt.ylabel('Number',fontproperties='SimHei', fontsize=15)
  plt.xlim(-10,60)  # Set the range of x-axis, and make some space
  
  plt.show()
  

5.1.2 function definition

The histogram drawing function in Matplotlib is plt.hist(), which is defined as follows:

def hist(x, bins=None, range=None, density=False, weights=None, cumulative=False, bottom=None, histtype='bar', 
align='mid', orientation='vertical', rwidth=None, log=False, color=None, label=None, stacked=False, *, data=None, **kwargs)[source]¶

Parameter (bold is common parameter)	Meaning
x	Input data, accept array or list
bins	Interval distribution setting (set a total interval and interval length)
range	The upper and lower limits of bins, if not set, default to x.min(),x.max()
density	Display statistical frequency, default is False, if True, display frequency of y-axis
histtype	Histogram type, bar (default), barred, step (ladder), stepfilled
align	Control the horizontal distribution of histogram, left, mid (default), right

5.1.3 histogram of normal distribution

• Effect
  
• Code

  import numpy as np
  import matplotlib.pyplot as plt
  
  # Fixing random state for reproducibility
  np.random.seed(20200306)
  
  mu, sigma = 50, 20  # Mean and variance
  x = mu + sigma * np.random.randn(10000)
  
  # the histogram of the data
  plt.hist(x, 50, density=True, facecolor='g', alpha=0.75)
  
  plt.xlabel('Smarts')
  plt.ylabel('Probability')
  plt.title('Histogram of IQ')
  plt.text(25, .023, r'$\mu=50,\ \sigma=20$')
  plt.xlim(-40, 160)
  plt.ylim(0, 0.03)
  plt.grid(True)
  plt.show()
  

5.2 pie chart

5.2.1 simple example

• Effect
  

• Code

  import matplotlib.pyplot as plt
  
  plt.rcParams['font.sans-serif'] = ['SimHei'] 
  labels = ('Chinese','Mathematics','English?','Physics')
  x = [20,30,25,25]  # 100% distribution
  explode = (0.2,0.1,0,0)  # Distance from center
  
  plt.pie(x,explode=explode,labels=labels,autopct='%.2f',
  shadow = False, startangle=45)
  plt.show()
  

5.2.2 function definition

The function to draw pie chart in Matplotlib is plt.pie(). The complete function definition and parameters are as follows:

def pie(x, explode=None, labels=None, colors=None, autopct=None,
pctdistance=0.6, shadow=False, labeldistance=1.1, startangle=None,
radius=None, counterclock=True, wedgeprops=None, textprops=None,
center=(0, 0), frame=False, rotatelabels=False, hold=None, data=None)

Parameter (bold is common parameter)	Meaning
x	(each block), if sum(x) > 1, sum(x) normalization will be used
explode	Distance from center
labels	(each piece) explanatory text displayed on the outside of the pie chart
colors	Custom color, list, such as ['red', 'yellow']
autopct	Control the percentage setting in pie chart. You can use format string or format function
pctdistance	Similar to labeldistance, specify the position scale of autopct, with the default value of 0.6
shadow	Draw a shadow under the pie. Default: False, i.e. no shadow
labeldistance	The drawing position of the label mark, relative to the scale of the radius, is 1.1 by default. If < 1, it will be drawn on the inside of the pie chart
startangle	The starting drawing angle, the default drawing is from the positive direction of x-axis and counterclockwise, if set = 90, then from the positive direction of y-axis
radius	Control pie radius, default is 1
counterclock	Specifies the pointer direction; Boolean value, optional parameter, default: True, i.e. counter clockwise. Change the value to False to clockwise
wedgeprops	Dictionary type, optional parameter, default value: None. The parameter dictionary is passed to the wedge object to draw a pie chart. For example: widget props = {'linewidth': 3} set widget linewidth to 3
textprops	Format labels and scale text; dictionary type, optional parameter, default value: None. Dictionary parameter passed to the text object.
center	Floating point type list, optional parameter, default value: (0,0). Icon center position
frame	Boolean type, optional parameter, default: False. If true, draw the axis frame with the table.
rotatelabels	Boolean type, optional parameter, default: False. If True, rotate each label to the specified angle

5.2.3 add legend

• Effect
  

• Code

  import matplotlib.pyplot as plt
  
  plt.rcParams['font.sans-serif'] = ['SimHei'] 
  labels = ('Chinese','Mathematics','English?','Physics')
  x = [20,30,25,25]  # 100% distribution
  explode = (0.2,0.1,0,0)  # Distance from center
  
  plt.pie(x,explode=explode,labels=labels,autopct='%.2f%%',  # Added percent sign
  shadow = False, startangle=45)
  
  plt.axis('equal')  # straighten
  plt.legend(loc="upper right", fontsize = 15, bbox_to_anchor=(1.1,1.2),borderaxespad=0.3)  
  # LOC: 'upper right' in the upper right corner
  # Bbox to anchor: top margin
  # Borderexpad: inside margin of legend
  plt.show()
  

5.3 scatter diagram

When we classify a large number of data, we can intuitively feel that there are several types by using scatter diagram.

The scatter chart itself is also a collection of data points, so it can be implemented by using plot function or plt.scatter

5.3.1 simple example

• Effect
  
• Code

  import matplotlib.pyplot as plt
  import numpy as np
  
  plt.plot(np.random.randint(0,100,50),np.random.randint(0,100,50),'o')
  plt.title('Scatter')
  
  plt.show()
  

• Effect
  
• Code

  import matplotlib.pyplot as plt
  import numpy as np
  
  x = np.random.randn(100)
  y = np.random.randn(100)
  
  plt.scatter(x,y, alpha=0.4)
  plt.title('plt.scatter')
  
  plt.show()
  

5.3.2 function definition

The plt.scatter() function is defined as follows:

def scatter(x, y, s=None, c=None, marker=None, cmap=None, norm=None, vmin=None, vmax=None,
alpha=None, linewidths=None, verts=None, edgecolors=None, hold=None, data=None, **kwargs)

parameter	Meaning
x	Point abscissa
y	Point ordinate (size equal to x)
s	Tag size, specify the size of the point
c	Mark color (such as' red ',' r ')
maker	Marking style, such as' o 'for circle and' x 'for cross
alpha	transparency

5.3.3 multiple point sets

There may be two or more different types of points in the scatter diagram. We need to draw them and compare them. At this time, it is more convenient to use the plt.scatter() function.

• Effect
  

• Code

  import matplotlib.pyplot as plt
  import numpy as np
  
  x = np.random.randn(100)
  y = np.random.randn(100)
  colors = np.random.rand(100) 
  
  plt.scatter(x,y,c = colors, alpha=0.4)
  plt.title('plt.scatter')
  plt.colorbar()
  
  plt.show()
  

• Iris data scatter
  Iris is a common data used in machine learning classification problems, including 150 data, three categories, each data contains four attributes. Let's draw these three kinds of data with scatter diagram and see their distribution.

  • Effect
    

  • Code

    from sklearn.datasets import load_iris
    iris = load_iris()
    features = iris.data.T  # Transpose is used to obtain all features plt.scatter(features[0],features[1],alpha=0.4,c=iris.target) separately
    plt.xlabel('Calyx length(cm)',fontproperties='SimHei',fontsize=15)
    plt.ylabel('Calyx width(cm)',fontproperties='SimHei',fontsize=15)
    plt.title('Iris data distribution(According to the length and width of calyx)',fontproperties='SimHei',fontsize=20)
    plt.colorbar()
    plt.show()
    

6. summary

This paper briefly describes the use of Matplotlib, as well as the drawing of some common graphs. More content will be updated in the later use.