Yellowbrick learningCurve: chage legend

Question

I want to plot the learning curve of train and validation set using yellowbrick learning curve. I am not using cross validation when plotting the learning curve, but a hold out validation set. Anyway, the legend is fixed to "Cross validation score".

Is there a way to replace it with another string?

Answer 1

Viz returns the axes object in the ax property, you can use it to set the label like that:

viz.ax.get_lines()[1].set_label('My custom label')

Example:

import numpy as np

from yellowbrick.datasets import load_energy
from yellowbrick.model_selection import ValidationCurve

from sklearn.tree import DecisionTreeRegressor

# Load a regression dataset
X, y = load_energy()

viz = ValidationCurve(
    DecisionTreeRegressor(), param_name="max_depth",
    param_range=np.arange(1, 11), cv=10, scoring="r2"
)

# Fit and show the visualizer
viz.fit(X, y)
viz.ax.get_lines()[1].set_label('My custom label')
viz.show()