How to adjust the size of the dots in the legend of a Seaborn scatterplot?

I know that the s argument in searbons scatterplot allows to control the size of the dots. For instance:

import seaborn as sns
import pandas as pd
import matplotlib.pyplot as plt
df = pd.DataFrame()
df['Y_LOCATION'] = [1, 1, 1, 2, 2, 4]
df['X_LOCATION'] = [1, 2, 3, 4, 3, 1]
df['VALUE'] = [-0.45, -0.14, -0.12, -0.36, -0.48, -0.20]
sns.set(rc={'figure.figsize':(40,20)})
sns.set(font_scale=2)
sns.scatterplot(df.Y_LOCATION, df.X_LOCATION, df.VALUE, s=800, palette = "Greens_r")

However, this parameter appears to have no impact on the size of the dots shown in the legend. How can these be adjusted?

Solution

You can increase the size of those dots by getting the handles of the dots in the legend, and calling set_sizes() on them. They are [36] by default, by multiplying that by 10, it increases the area (not diameter) by a factor of 10.

Example:

ax = sns.scatterplot(x = df.Y_LOCATION, y = df.X_LOCATION, hue = df.VALUE, s=800, palette = "Greens_r")
handles, labels = ax.get_legend_handles_labels()
for dot in handles:
    dot.set_sizes(dot.get_sizes() * 10)
plt.legend(handles, labels)

Documentation on PathCollection.set_sizes(): https://matplotlib.org/stable/api/collections_api.html#matplotlib.collections.PathCollection.set_sizes