.legend()

Published Oct 14, 2022Updated Mar 28, 2023

The pyplot.legend() function returns a legend on the axes of a plot or subplot.

Syntax

plot_instance.legend(handles, labels)

The .legend() method accepts zero or more parameters and can be called on a valid plot_instance of the pyplot module. When called with no parameters, .legend() will attempt to add the elements automatically. The following parameters can be used:

Parameter Name	Data Type(s)	Usage
`handles`	sequence of `Artist`	Utilized if more control is desired with each entry on the plot and how they are labeled.
`labels`	list of strings	Displays the labels in the order they are entered.
`loc`	int, string	Determines the position of the legend on the axes. Defaults to `best`.
`ncol`	int	Determines how many columns the legend will have. Defaults to 1.
`fontsize`	int, string	Changes the font size the labels. Defaults to 11.
`labelcolor`	list, string, `None`	Changes the color of the label(s). Defaults to `None`
`title`	string, `None`	Inserts a title for the legend. Defaults to `None`.
`title_fontsize`	int, string, `None`	Changes the font size of the legend `title`. Defaults to `None`.

Example

In the following example, .legend() will be shown using the iris data set which is available using seaborn (pandas .read_csv() can be used alternatively). The handles parameter will be demonstrated, using seaborn to import the dataset.

import matplotlib.pyplot as plt
import seaborn as sns

iris = sns.load_dataset('iris')

fig, ax = plt.subplots(figsize=(12,6))

sepal_lw = ax.scatter(iris['sepal_length'], iris['sepal_width'], label='Sepal')
petal_lw = ax.scatter(iris['petal_length'], iris['petal_width'], label='Petal')

plt.legend(handles=[sepal_lw, petal_lw]);