https://github.com/berndbischl/mlr
Tip revision: b251a1542cf77ae04eb0b76db947f58fc7aab3a9 authored by pat-s on 22 January 2021, 10:10:04 UTC
update update-tic
update update-tic
Tip revision: b251a15
generateLearningCurveData.Rd
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/generateLearningCurve.R
\name{generateLearningCurveData}
\alias{generateLearningCurveData}
\alias{LearningCurveData}
\title{Generates a learning curve.}
\usage{
generateLearningCurveData(
learners,
task,
resampling = NULL,
percs = seq(0.1, 1, by = 0.1),
measures,
stratify = FALSE,
show.info = getMlrOption("show.info")
)
}
\arguments{
\item{learners}{[(list of) \link{Learner})\cr
Learning algorithms which should be compared.}
\item{task}{(\link{Task})\cr
The task.}
\item{resampling}{(\link{ResampleDesc} | \link{ResampleInstance})\cr
Resampling strategy to evaluate the performance measure.
If no strategy is given a default "Holdout" will be performed.}
\item{percs}{(\link{numeric})\cr
Vector of percentages to be drawn from the training split.
These values represent the x-axis.
Internally \link{makeDownsampleWrapper} is used in combination with \link{benchmark}.
Thus for each percentage a different set of observations is drawn resulting in noisy performance measures as the quality of the sample can differ.}
\item{measures}{[(list of) \link{Measure})\cr
Performance measures to generate learning curves for, representing the y-axis.}
\item{stratify}{(\code{logical(1)})\cr
Only for classification:
Should the downsampled data be stratified according to the target classes?}
\item{show.info}{(\code{logical(1)})\cr
Print verbose output on console?
Default is set via \link{configureMlr}.}
}
\value{
(\link{LearningCurveData}). A \code{list} containing:
\itemize{
\item The \link{Task}
\item List of \link{Measure})\cr
Performance measures
\item data (\link{data.frame}) with columns:
\itemize{
\item \code{learner} Names of learners.
\item \code{percentage} Percentages drawn from the training split.
\item One column for each \link{Measure} passed to \link{generateLearningCurveData}.
}
}
}
\description{
Observe how the performance changes with an increasing number of observations.
}
\examples{
r = generateLearningCurveData(list("classif.rpart", "classif.knn"),
task = sonar.task, percs = seq(0.2, 1, by = 0.2),
measures = list(tp, fp, tn, fn),
resampling = makeResampleDesc(method = "Subsample", iters = 5),
show.info = FALSE)
plotLearningCurve(r)
}
\seealso{
Other generate_plot_data:
\code{\link{generateCalibrationData}()},
\code{\link{generateCritDifferencesData}()},
\code{\link{generateFeatureImportanceData}()},
\code{\link{generateFilterValuesData}()},
\code{\link{generatePartialDependenceData}()},
\code{\link{generateThreshVsPerfData}()},
\code{\link{plotFilterValues}()}
Other learning_curve:
\code{\link{plotLearningCurve}()}
}
\concept{generate_plot_data}
\concept{learning_curve}