How to modify the kernel density estimate line in a sns.histplot

Question

I am creating a histrogram (frecuency vs. count) and I want to add kernel density estimate line in a different colour. How can I do this? I want to change the colour for example

sns.histplot(data=penguins, x="flipper_length_mm", kde=True)

Example taken from https://seaborn.pydata.org/generated/seaborn.histplot.html

Answer 1

histplot's line_kws={...} is meant to change the appearance of the kde line. However, the current seaborn version doesn't allow changing the color that way, probably because the color goes together with the hue parameter (although hue isn't used in this case).

import seaborn as sns

penguins = sns.load_dataset('penguins')
ax = sns.histplot(data=penguins, x="flipper_length_mm", kde=True,
                  line_kws={'color': 'crimson', 'lw': 5, 'ls': ':'})

In seaborn's github, it is suggested to draw the histplot and the kdeplot separately. For both to match in the y-direction, it is necessary to use histplot with stat='density' (the kdeplot doesn't have a parameter to use histplot's default stat='count').

penguins = sns.load_dataset('penguins')
ax = sns.histplot(data=penguins, x="flipper_length_mm", kde=False, stat='density')
sns.kdeplot(data=penguins, x="flipper_length_mm", color='crimson', ax=ax)

If the count statistics is really needed, an alternative is to change the line color via matplotlib:

penguins = sns.load_dataset('penguins')
ax = sns.histplot(data=penguins, x="flipper_length_mm", kde=True)
ax.lines[0].set_color('crimson')