[Feature] Multiple rugplots on same plot

Hi there,

Is there a way to add multiple rugplots to the same figure? This would be useful when having unique values on one axis and multiple grouping variables to color by. I understand this is achievable with shapes/colors for groups of low cardinality but I believe can quickly get confusing. Currently, when calling rugplot twice the second rugplot draws over the first one.

Example:

import string
import random
import numpy as np
import pandas as pd
import seaborn as sns

N = 20
groups_a = random.choices(string.ascii_letters[0:5], k=N)
groups_b = random.choices(string.ascii_uppercase[0:5], k=N)
Y = np.random.random(N)
X = range(N)
df = pd.DataFrame.from_dict({"X":X, "Y":Y, "A":groups_a,"B":groups_b})

sns.scatterplot(data=df, x="X", y="Y")
sns.rugplot(data=df, x="X", hue="A", linewidth=12)
sns.rugplot(data=df, x="X", hue="B", linewidth=12)

This yields:

Expected output would be something like:

(of course with more modifications of cmaps etc. to make the plot more clear)

[jeskowagner]

Yes I understand that this might be misusing rug plots a bit much. I will think about alternative angles to this problem then. Thanks for your help!

[mwaskom]

ideally one would have some sort of spacing in between the two rugs

I think you’re asking for too much from a rug plot here, which is meant to show the marginal distribution of observations, and not additional categorical information. I’m still not exactly sure what you’re hoping for, but I think the answer is going to require a bespoke plot with a fair amount of customization. At this point, I’d suggest defining a clear problem statement and asking on StackOverflow, since the original feature request

Is there a way to add multiple rugplots to the same figure?

is already satisfied with existing functionality.