autoplot.PredictionClassif: Plots for Classification Predictions

In mlr3viz: Visualizations for 'mlr3'

Plots for Classification Predictions

Description

Visualizations for mlr3::PredictionClassif. The argument type controls what kind of plot is drawn. Possible choices are:

• "stacked" (default): Stacked barplot of true and estimated class labels.

• "roc": ROC curve (1 - specificity on x, sensitivity on y). Requires package precrec.

• "prc": Precision recall curve. Requires package precrec.

• "threshold": Systematically varies the threshold of the mlr3::PredictionClassif object and plots the resulting performance as returned by measure.

Usage

## S3 method for class 'PredictionClassif'
autoplot(
  object,
  type = "stacked",
  measure = NULL,
  theme = theme_minimal(),
  ...
)

Arguments

object	(mlr3::PredictionClassif).
type	(character(1)): Type of the plot. See description.
measure	(mlr3::Measure) Performance measure to use.
theme	(ggplot2::theme()) The ggplot2::theme_minimal() is applied by default to all plots.
...	(ignored).

Value

ggplot2::ggplot().

References

Saito T, Rehmsmeier M (2017). “Precrec: fast and accurate precision-recall and ROC curve calculations in R.” Bioinformatics, 33(1), 145-147. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1093/bioinformatics/btw570")}.

Examples

if (requireNamespace("mlr3")) {
  library(mlr3)
  library(mlr3viz)

  task = tsk("spam")
  learner = lrn("classif.rpart", predict_type = "prob")
  object = learner$train(task)$predict(task)

  head(fortify(object))
  autoplot(object)
  autoplot(object, type = "roc")
  autoplot(object, type = "prc")
}

mlr3viz documentation built on July 1, 2024, 5:06 p.m.