Bars for one metric, one bar per model and scope. Defaults to the .overall
rows, which is the comparison people usually want: pooled versus subgroup
versus individual.
Usage
plot_metrics(
x,
metric = NULL,
model = NULL,
scope = NULL,
overall = TRUE,
type = c("overall", "boxplot", "density", "time"),
...
)Arguments
- x
An idiographic fit.
- metric
Metric column to plot. Defaults to
rmsefor regression andaccuracyfor classification. Distribution and density views also accept a vector of metric names.- model, scope
Optional filters.
- overall
Logical. Plot the aggregate rows (default) or every subject.
- type
Overall points (default), learner distributions (
"boxplot"or"density"), or rolling-fold trajectories ("time").- ...
Passed to base plotting functions.
