Seaborn pairplot off-diagonal KDE with two classes

Question

I'm trying to look at a Seaborn pairplot for two different classes of variables and I'd like to see KDEs on the offdiagonals instead of scatterplots. The documentation has instructions on how to do a KDE for all of the data, but I want to see separate KDEs for each subclass of data. Suggestions welcome!

My code looks something like this:

plot = sns.pairplot(
    df,
    vars=labels,
    hue='has_accident',
    palette='Set1',
    diag_kind='kde',
)

which results in:

As you can see the data are sufficiently dense that it is difficult to see the difference in the red and blue data on the off diagonal.

Answer 1

You possibly mean something like this:

import seaborn as sns
import matplotlib.pyplot as plt

iris = sns.load_dataset("iris")

g = sns.PairGrid(iris, hue="species", hue_kws={"cmap": ["Blues", "Greens", "Reds"]})
g = g.map_diag(sns.kdeplot, lw=3)
g = g.map_offdiag(sns.kdeplot, lw=1)

plt.show()