Most Basic Bubble Chart
A bubble plot is basically a scatterplot with an additional dimension: size of points. Using seaborn library, a bubble plot can be constructed using the scatterplot() function. In the example, the following parameters are used to build a basic bubble plot:
data: Input data structurex: The data position on the x axisy: The data position on the y axissize: Grouping variable that will produce points with different sizesalpha: Transparancy ratio
Here, the gapminder data set is used and the relationship between life expectancy (y) and gdp per capita (x) of world countries is represented. The population of each country is represented through dot size.
Note: the gapminder dataset can be installed with pip install gapminder
# libraries
import matplotlib.pyplot as plt
import seaborn as sns
from gapminder import gapminder # import data set
# Control figure size for this notebook:
plt.rcParams['figure.figsize'] = [8, 8]
# data
data = gapminder.loc[gapminder.year == 2007]
# use the scatterplot function to build the bubble map
sns.scatterplot(data=data, x="gdpPercap", y="lifeExp", size="pop", legend=False, sizes=(20, 2000))
# show the graph
plt.show()
Control Bubble Size
You can control the sizes of bubbles using the sizes argument. You should pass the minimum and maximum size to use such that other values are normalized within this range.
# libraries
import matplotlib.pyplot as plt
import seaborn as sns
from gapminder import gapminder # import data set
# data
data = gapminder.loc[gapminder.year == 2007]
# use the scatterplot function
sns.scatterplot(data=data, x="gdpPercap", y="lifeExp", size="pop", alpha=0.5, sizes=(20, 800))
# show the graph
plt.show()
Bubble color
We added a third dimension to our scatterplot with the sizes of bubbles. Now, we will add a forth dimension by mapping colors to variables. Here, the continent of each country is used to control bubble color and passed to the function with the hue argument.
# use the scatterplot function
sns.scatterplot(data=data, x="gdpPercap", y="lifeExp", size="pop", hue="continent", alpha=0.5, sizes=(20, 400))
# show the graph
plt.show()
Control Aestethics
You can set the theme using the set_style() function of matplotlib and change the color palette of bubbles passing the palette argument to the scatterplot() function. It is also recomended to take the legend outside of the plot using the legend() function.
# set seaborn "whitegrid" theme
sns.set_style("darkgrid")
# use the scatterplot function
sns.scatterplot(data=data, x="gdpPercap", y="lifeExp", size="pop", hue="continent", palette="viridis", edgecolors="black", alpha=0.5, sizes=(10, 1000))
# Add titles (main and on axis)
plt.xlabel("Gdp per Capita")
plt.ylabel("Life Expectancy")
# Locate the legend outside of the plot
# plt.legend(bbox_to_anchor=(1, 1), loc='upper left', fontsize=17)
# show the graph
plt.show()
