scatter plots in seaborn/matplotlib with point size and color given by continuous dataframe column

Question

I would like to make a scatter plot in seaborn/matplotlib where the size of points is determined by a (continuous) value in a dataframe, and the color of points is also determined by the continuous value of another column in dataframe. In ggplot, the way to do it is:

ggplot(iris) + geom_point(aes(x=Sepal.Width, y=Sepal.Length, size=Petal.Width, color=Petal.Length))

(color/size here are continuous not categorical values)

what's the syntax for this in seaborn/matplotlib?

Answer 1

The following reproduces the code diagram from the question. Optaining a legend is a bit cumbersome, because we have to manually define some proxy artists to put to the legend and remove the first automatic legend entry which is generated via the seaborn style.

import seaborn as sns
import matplotlib.pyplot as plt
iris = sns.load_dataset("iris")

plt.scatter(iris.sepal_width, iris.sepal_length, 
            c = iris.petal_length, s=(iris.petal_width**2)*60, cmap="viridis")
ax = plt.gca()

plt.colorbar(label="petal_length")
plt.xlabel("sepal_width")
plt.ylabel("sepal_length")

#make a legend:
pws = [0.5, 1, 1.5, 2., 2.5]
for pw in pws:
    plt.scatter([], [], s=(pw**2)*60, c="k",label=str(pw))

h, l = plt.gca().get_legend_handles_labels()
plt.legend(h[1:], l[1:], labelspacing=1.2, title="petal_width", borderpad=1, 
            frameon=True, framealpha=0.6, edgecolor="k", facecolor="w")

plt.show()

Note that the size argument s denotes the area of the dots. So in order to have the diameter be proportional to the quantitiy to show, it has to to be squared.

Answer 2

Here is how I would solve this using Altair

from altair import Chart
import seaborn as sns
iris = sns.load_dataset("iris")
c = Chart(iris)
c.mark_circle().encode(
    x='sepal_width',
    y='sepal_length',
    color='petal_length',
    size='petal_width',
)

For such plots, Altair might be a great choice. Quoting from their website:

Altair is a declarative statistical visualization library for Python, based on Vega-Lite. With Altair, you can spend more time understanding your data and its meaning.