Plot a Model's Learning Curve

Draw training and validation losses recorded by a rtemis Supervised object using the ECharts renderer in rtemis.draw.

Usage

plot_learning(x, ...)

Arguments

x

rtemis::Supervised: Fitted regression or classification result.

...

Additional arguments to rtemis.draw::draw_learning_curve().

Value

htmlwidget: ECharts learning curve.

Details

The method uses get_learning_curve() to extract algorithm-specific progress, losses, and the selected step, then rtemis.draw::draw_learning_curve() to render them. Forest curves are averaged at each step over available trees. Models without recorded learning losses produce an informative error. The default title names the algorithm; pass title = NULL to omit it or title = "..." to supply another title.

Examples

if (FALSE) { # interactive()
model <- train(
  mtcars[, c("wt", "mpg")],
  hyperparameters = setup_LINAD(max_leaves = 3L),
  verbosity = 0L
)
plot_learning(model)
}