Use .map and .map_dataframe with Seaborn object interface

[leodarrigade]

Hi,

I am cross-posting here from Stackoverflow as I did not get any answer there.

I would like to apply the methods seaborn.FacetGrid.map and seaborn.FacetGrid.map_dataframe to a facet plot I build using the object interface. I cannot find a way to do it, any help would be appreciated.

For example, this is my figure

import numpy as np
import pandas as pd

import matplotlib.pyplot as plt
import seaborn as sns   
import seaborn.objects as so

df = pd.DataFrame(np.arange(5*3).reshape(5,3), columns=["A","B","C"])
df["C"] = df["C"].apply(lambda x: x > 5)

fig = plt.figure()
fg = ((so.Plot(df, x="A", y= "B", group = "C")
       .add(so.Line(linewidth = 0.9)))
       .facet(col = "C")
       .on(fig)
       .plot())

And this is the function I would like to apply on each axe of the FacetGrid. This function write a text on each axe displaying the number of data used for the plot.

def show_counts(color, data):
    ax = plt.gca()
    ax.text(0.5, 0.95, f'Count: {len(data)}', ha='center', transform=ax.transAxes)

fg.map_dataframe(show_counts)

However, from the last line of code, I get the following error :

AttributeError: 'Plotter' object has no attribute 'map_dataframe'

Would there be a way to access the data of each axe and apply a function to it ?

[mwaskom]

Hi, unfortunately this is not possible, the objects interface does not use FacetGrid internally; they're completely separate systems.

[leodarrigade]

Ok thanks