Scale setting problem

1. Preface

When we visualize data, we usually encounter an acceptance problem , We want to make X Axis or Y The axis represents more intuitive information , But obviously matplotlib The default scale of the library cannot meet our requirements . The default scale is shown in the figure below ：

Although this icon can also express the quarterly income , But the diagram is not intuitive . The biggest problem is the scale of the abscissa and ordinate , It has nothing to do with the data you want to respond to .

---

2. Solution

2.1 Solution 1

We use set_xticks Method to clean up meaningless scales , The method parameters are as follows ：

Parameter description ：

set_xticks(self, ticks, minor=False)
Parameters 1：ticks： Specify where the scale appears
Parameters 2：minor： Specifies whether it is a minor scale
Return value 1： contain XTick List of instances

---

Parameters, ：

1. Although the official document states ticks The parameter must be a list , However, according to the blogger's experiment, any iterative sequence can be passed to this parameter .
2. minor Parameters are very important , It is used to specify the minor scale , The so-called secondary scale is the scale without scale label , They have only one short stick on the picture , Won't have a label , therefore set_xticklabels Methods cannot label them by default .
3. Bear in mind set_xticklabels Method by default, the number of labels can be set to be the number of non secondary scales , Not the sum of minor and non minor scales .

---

The sample program for deleting meaningless scale is as follows ：

# Import necessary Libraries 
import matplotlib.pyplot as plt
import numpy as np
plt.rcParams['font.sans-serif'] = ['SimHei']  #  Settings support Chinese 

# Initialize data and frame 
height = np.random.randint(10, 40, 4)
figure, ax = plt.subplots()
X = [0, 2, 4, 6]
width = 0.6

rectangle = ax.bar(X, height, width=width, color='#000000')
texts_list = plt.bar_label(rectangle, fmt='%d Ten thousand yuan ')

# Set font and other properties 
for text in texts_list:
    text.set_fontsize(15)
    text.set_color('#FF0000')

ax.set_xticks(X)    # Delete meaningless scale 

ax.set_title('2020 Annual income ', fontsize=25)
ax.set_xlabel(' quarter ', fontsize=15)
ax.set_ylabel(' income ( Ten thousand yuan )', fontsize=15)
plt.show()

The result is as follows ：

Now we have deleted the meaningless scale , For the existing scale, we also hope that he will express the meaning of quarter , We need to use set_xticklabels Method , This method is used to set the scale label , Its parameters are as follows ：

Parameter description ：

set_xticklabels(self, labels, fontdict=None, minor=False, **kwargs)
Parameters 1：labels： Specify the scale name
Parameters 2：fontdict： Dictionary type , Specify font properties
Parameters 3：minor： Specifies whether it is a minor scale label
Return value 1： contain Text List of instances

---

Parameters, ：

1. By default minor=False, So you can only set the label of non minor scale , We can pass on minor=True, This sets the secondary scale label .
2. fontdict Parameters pass a dictionary , Used for global settings , But the general practice of bloggers is to return from this method Text Instance to set properties , So as to accurately set the properties .
3. Use only set_ticks Method after fixing the scale position , This method should be used . otherwise , The label may appear in an unexpected position .

---

The sample program for adding scale labels is as follows ：

# Import necessary Libraries 
import matplotlib.pyplot as plt
import numpy as np
plt.rcParams['font.sans-serif'] = ['SimHei']  #  Settings support Chinese 

# Initialize data and frame 
height = np.random.randint(10, 40, 4)
figure, ax = plt.subplots()
X = [0, 2, 4, 6]    # initialization X coordinate 
width = 0.6 # Initialize the column width 
xtick_label = [' quarter 1', ' quarter 2', ' quarter 3', ' quarter 4']  # initialization x Axis scale label 
xtickcolors_list = ['red', 'blue', 'green', 'purple']   # initialization x Axis scale label color 

rectangle = ax.bar(X, height, width=width, color='#000000') # drawing 
texts_list = plt.bar_label(rectangle, fmt='%d Ten thousand yuan ')   # Set bar chart labels 

ax.set_xticks(X)    # Delete meaningless scale 
ticklabels_list = ax.set_xticklabels(xtick_label, fontdict=dict(fontsize=15), rotation=30) # Add calibration label 

# Set the scale label color 
for ticklabel, color in zip(ticklabels_list, xtickcolors_list):
    ticklabel.set_color(color)

# Set attributes such as the font size of the histogram annotation 
for text, color in zip(texts_list, xtickcolors_list):
    text.set_fontsize(15)
    text.set_color(color)

ax.set_title('2020 Annual income ', fontsize=35)    # Set title 
ax.set_ylabel(' income ( Ten thousand yuan )', fontsize=20)    # Set up y Axis labels 
plt.show()

The result is as follows ：

---

Similarly, we can also set y Axis for better visualization , The method used is consistent . The example program is as follows ：

# Import necessary Libraries 
import matplotlib.pyplot as plt
import numpy as np
plt.rcParams['font.sans-serif'] = ['SimHei']  #  Settings support Chinese 

figure, ax = plt.subplots()

# Initialization data 
height = np.random.randint(10, 40, 4)   # Initialize column height data 
data_mean = height.mean()
X = [0, 2, 4, 6]    # initialization X coordinate 
width = 0.6 # Initialize the column width 
yticks_list = np.arange(0, 45, 5)   # initialization y scale 
yticks_label = [str(i) + ' Ten thousand yuan ' for i in yticks_list]  # initialization x Axis scale label 
xticks_label = [' quarter 1', ' quarter 2', ' quarter 3', ' quarter 4']  # initialization x Axis scale label 
xtickcolors_list = ['red', 'blue', 'green', 'purple']   # initialization x Axis scale label color 

rectangle = ax.bar(X, height, width=width, color='#000000', label='2020 Annual income ') # drawing 
ax.axhline(data_mean, label='2020 Average annual revenue ', linewidth=3)    # Draw an average 
texts_list = plt.bar_label(rectangle, fmt='%d Ten thousand yuan ')   # Set bar chart labels 

ax.set_xticks(X)    # Delete meaningless scale , Set up x Main scale 
xticklabels_list = ax.set_xticklabels(xticks_label, fontdict=dict(fontsize=15), rotation=30) # Add calibration label 

ax.set_yticks(yticks_list)  # Set up y scale 
ax.set_yticks([data_mean], minor=True)    # Set up y Minor scale 
major_yticklabels_list = ax.set_yticklabels(yticks_label, fontdict=dict(fontsize=15))   # Set up y Major scale labels 
minor_yticklabels_list = ax.set_yticklabels([str(data_mean) + ' Ten thousand yuan '], fontdict=dict(fontsize=15, color='#ff0000'), minor=True)   # Set up y Secondary scale label 


# Set the scale label color 
for ticklabel, color in zip(xticklabels_list, xtickcolors_list):
    ticklabel.set_color(color)

# Set attributes such as the font size of the histogram annotation 
for text, color in zip(texts_list, xtickcolors_list):
    text.set_fontsize(15)
    text.set_color(color)

figure.suptitle('2020 Annual income ', fontsize=25) # Set title 
figure.supylabel(' income ( Ten thousand yuan )', fontsize=20)    # Set up y Axis labels 

ax.set_xlim([-1, 7])    # Limit the view range 
ax.legend(title='2020 Annual income ')  # Set legend 
plt.show()

The result is as follows ：

Many other skills and solutions are used in the program , For details, please see Reference link .

---

3. Reference resources

1. Solution to the problem of cylinder annotation
2. linestyle Parameters,
3. capstyle Parameters,
4. joinstyle Parameters,
5. Solution to subgraph spacing problem
6. Subgraph setting headline solution

---