How to Change Order of Items in Matplotlib Legend?

Legend is a method present in the Matplotlib package which represents the area to describe the elements in a plot/graph.

The order of labels in the legend region was based on the order we add the lines(data) to the plot.

Example:

Let us consider a DataFrame that stores student marks in weekly tests.

	Raju	Hari	Bablu	Dora
Test-1	8	6	9	10
Test-2	10	4	9	9
Test-3	7	6	9	10
Test-4	7	7	9	9
Test-5	10	6	9	10

The above table stores the 4 student marks in 5 tests. Now plot the graph with legend using the above data.

Python3

# import necessary packages
import pandas as pd
import matplotlib.pyplot as plt

# create a dataframe
Marks = pd.DataFrame({'Raju': [8, 10, 7, 7, 10],
                      'Hari': [6, 4, 6, 7, 6],
                      'Bablu': [9, 9, 9, 9, 9],
                      'Dora': [10, 9, 10, 9, 10]})

# plot marks of each student
plt.plot(Marks['Raju'], label="Raju Marks", color="Red")
plt.plot(Marks['Hari'], label="Hari Marks", color="Blue")
plt.plot(Marks['Bablu'], label="Bablu Marks", color="Yellow")
plt.plot(Marks['Dora'], label="Dora Marks", color="Black")

# labelling the axes
plt.xlabel("Tests")
plt.ylabel("Marks")

# add legend to plot
plt.legend()
plt.show()

Output

Change order of items in the legend

The above order of elements in the legend region can be changed by the gca method that uses another sub-method called get_legend_handles_labels method.

Syntax

handles, labels = plt.gca().get_legend_handles_labels()

These handles and labels lists are passed as parameters to legend method with order of indexes.

Consider the below example code for detailed understanding.

Example 1:

In this code, we used the same DataFrame we used in the above code. But used the above-specified methods to change the order of elements in the legend region.

Python3

# import necessary packages
import pandas as pd
import matplotlib.pyplot as plt

# create a dataframe
Marks = pd.DataFrame({'Raju': [8, 10, 7, 7, 10],
                      'Hari': [6, 4, 6, 7, 6],
                      'Bablu': [9, 9, 9, 9, 9],
                      'Dora': [10, 9, 10, 9, 10]})

# plot marks of each student
plt.plot(Marks['Raju'], label="Raju Marks", color="Red")
plt.plot(Marks['Hari'], label="Hari Marks", color="Blue")
plt.plot(Marks['Bablu'], label="Bablu Marks", color="Yellow")
plt.plot(Marks['Dora'], label="Dora Marks", color="Black")

# labelling the axes
plt.xlabel("Tests")
plt.ylabel("Marks")

# reordering the labels
handles, labels = plt.gca().get_legend_handles_labels()

# specify order
order = [2, 3, 1, 0]

# pass handle & labels lists along with order as below
plt.legend([handles[i] for i in order], [labels[i] for i in order])
plt.show()

Output

Example 2:

Below is the implementation of plotting a graph that represents student marks along with legend with specified order in legend region.

Python3

# import necessary packages
import pandas as pd
import matplotlib.pyplot as plt

# create a dataframe
Marks = pd.DataFrame({'Raju': [8, 10, 7, 7, 10],
                      'Hari': [6, 4, 6, 7, 6],
                      'Bablu': [9, 9, 9, 9, 9],
                      'Dora': [10, 9, 10, 9, 10]})

# plot marks of each student
plt.plot(Marks['Raju'], label="Raju Marks", color="Red")
plt.plot(Marks['Hari'], label="Hari Marks", color="Blue")
plt.plot(Marks['Bablu'], label="Bablu Marks", color="Yellow")
plt.plot(Marks['Dora'], label="Dora Marks", color="Black")

# labelling the axes
plt.xlabel("Tests")
plt.ylabel("Marks")

# reordering the labels
handles, labels = plt.gca().get_legend_handles_labels()

# specify order
order = [3, 2, 1, 0]

# pass handle & labels lists along with order as below
plt.legend([handles[i] for i in order], [labels[i] for i in order])
plt.show()

Output