Drawing library matplotlibmatplotlib quick start

Matplotlib What is it? ？

Matplotlib Is a comprehensive library , Used in Python Create static , Animated and interactive visual images .

at present （22 year 6 month ） The latest stable version is 3.5.2

install :

Use pip Installation

pip install matplotlib

Quick start

Let's first import matplotlib

import matplotlib as mplimport matplotlib.pyplot as pltimport numpy as np  # use numpy Generate the data , Demonstrate as an example .

Matplotlib In the figure (Figure)（ for example , window ,Jupyter Small parts, etc ） Draw data on , A graph can contain one or more axis fields (Axes).

Axes It can be based on x-y Coordinates the drawing area of the specified point （ Or in polar coordinates θ-r,3D In the picture x-y-z etc. ）. Create with Axes The easiest way to get a graph of is to use  pyplot.subplots. then , We can use Axes.plot stay Axes Draw some data on ：

x = [1,2,3,4]y = [1,4,3,2]fig,ax = plt.subplots() #  Create a diagram fig,  The default includes a axesax.plot(x,y) #  draw x-y Line chart of plt.show()  #  Show the drawing . Please note that , If you use save Save the picture , Need to be in show Save before 

The line chart will be displayed after running ：

chart （Figure） Structure

A picture has the following parts ： title (Title)、 legend （Legend）、x、y Axis labels (xlabel、ylabel) wait …

The following is a part by part introduction .

chart Figure

Complete image . This graph tracks all sub axis domains （Axes）---- A group of “ special ” Painting （ title , legend , Color bar, etc ）, Even nested subgraphs .

The easiest way to create a new diagram is to use pyplot：

fig = plt.figure()  #  Empty graph , No, Axesfig, ax = plt.subplots()  # Yes 1 individual Axesfig, axs = plt.subplots(2, 2)  #  Yes 2x2（ Two lines and two columns ）  individual Axes

Usually , Set the axis field （Axes） And Figure It is convenient to create together , But you can also add the axis domain manually later .

Axis domain Axes

Axis domain （Axes） Is attached to the graph （Figure） On , Contains graphics for plotting data .

It usually includes two shafts （Axis） object . The two axes provide scales （ticks） And labels (tick label), To provide a scale for the data in the axis . Every Axis domain There's also a title （ adopt  set_title（）） Set up ）、 One x label （ adopt  set_xlabel（））  Set up ） And a y label （ adopt  set_ylabel（））  Set up ）.

Axes  Class and its member functions use the OOP The main entry point of the interface , And most drawing methods are defined on it （ for example , As shown above , Use plot Method ）ax.plot()

Axis Axis

Axis setting scale （scale） And limits （limits） And generate scales （ticks, Mark on shaft ） And scale labels （ticklabels, String marking the scale ）ticks The position is determined by the positioner （Locator） determine ,ticklabel The string consists of Formatter （Formatter） Set up . Correct positioner （Locator） and Formatter （Formatter） The combination of can control the scale position and label very finely .

Artist

Artist Here it is translated into an artist or painter .

Basically , Everything visible on the graph is an artist （ Even graphics , Axis domain and Axis object ）. This includes Text 、Line2D 、 aggregate 、Patch etc. . When rendering graphics , All artists will be drawn to canvas On . Most artists are associated with axis fields ; Such artists cannot be shared by multiple axis fields , Nor can it move from one axis to another .

Input data type of drawing function

Drawing function receives  numpy.array or numpy.ma.masked_array As input , Or it can be passed to numpy.asarray The data of .pandas Data or numpy.matrix May not work properly . A common convention is to convert data into numpy.array. for example ：

b = np.matrix([[1, 2], [3, 4]])b_asarray = np.asarray(b) #  Use np.asarray() Convert it to np.array type 

Most methods can also resolve addressable objects , Such as  dict,np.recarray or pandas.DataFrame.

Matplotlib Allow the use of keyword parameters to generate images , Transfer and x,y Corresponding string .

np.random.seed(19680801)  # seed the random number generator.data = {'a': np.arange(50),        'c': np.random.randint(0, 50, 50),        'd': np.random.randn(50)}data['b'] = data['a'] + 10 * np.random.randn(50)data['d'] = np.abs(data['d']) * 100fig, axes = plt.subplots(2,1) # fig  Have 2 That's ok 1 List subgraphs , Store in axes Array （np.array type ） in .ax = axes[0]ax.scatter('a', 'b', c='c', s='d', data=data)ax.set_xlabel('entry a')ax.set_ylabel('entry b')#  Axis domain 2 , Remove the color c, shape s Parameters .ax2 = axes[1]ax2.scatter('a', 'b', data=data)plt.show()

The drawing is , The second one is that s,c Image after parameter ：

Encoding style Coding Styles

object-oriented （OO） and pyplot The function interface .

Basically, there are two uses Matplotlib Methods ：

• Create explicitly “ graphics （Figures）” and “ Axis domain （Axes）”, And call the method on it （“ object-oriented （OO） style ”）.

• rely on pyplot Automatically create and manage drawings and axes , And use pyplot Function to plot .

Use OO style ( I feel OO The style is better , Just in the axis field （Axes） Object can be set , Very clear )：

x = np.linspace(0, 2, 100)  #  Generate some data #  Use OO style , First generate two objects   Graph and axis field  （fig, ax）fig, ax = plt.subplots(figsize=(5, 2.7), layout='constrained')#  call ax Object method ax.plot(x, x, label='linear')  # Plot some data on the axes.ax.plot(x, x**2, label='quadratic')  # Plot more data on the axes...ax.plot(x, x**3, label='cubic')  # ... and some more.ax.set_xlabel('x label')  # Add an x-label to the axes.ax.set_ylabel('y label')  # Add a y-label to the axes.ax.set_title("Simple Plot")  # Add a title to the axes.ax.legend()  # Add a legend.plt.show() a legend.

Or use pyplot Function style ：

x = np.linspace(0, 2, 100)  # Sample data.plt.figure(figsize=(5, 2.7), layout='constrained')plt.plot(x, x, label='linear')  # Plot some data on the (implicit) axes.plt.plot(x, x**2, label='quadratic')  # etc.plt.plot(x, x**3, label='cubic')plt.xlabel('x label')plt.ylabel('y label')plt.title("Simple Plot")plt.legend();

（ Besides , There's a third way , Used in GUI Embed... In the application Matplotlib The situation of , Even for graphic creation . For more information , See the corresponding section in the Library ： Embed Matplotlib.

Common drawing types

From the homepage of the official website Get into Plot types, You can see how to draw commonly used different types of graphs .

You can see , Common line chart 、 Scatter plot 、 Histogram, etc . Click the corresponding figure to enter the corresponding case .

The main usage is basically like this , We'll talk about some style adjustments later .