- feat: Allow passing parameters to
precrec::autoplot()
(egshow_cb
) when plottingBenchmarkResult
andResampleResult
objects, usingtype
=roc
orprc
. - refactor: Wrong
type
inautoplot
s now gives hints of which ones to use. - refactor: Update
EnsembleFSResult.autoplot
to use theactive_measure
field.
- Add plot for
LearnerSurvCoxPH
.
- Work with new bbotk 0.9.0 and mlr3tuning 0.21.0
- Add plots for
EnsembleFSResult
object.
- Work with new paradox version 1.0.0
- Add
"prediction"
plots for classification and regression learners.
- Fix snapshots for bbotk 0.7.3.
- Add
"incumbent"
plot forOptimInstanceSingleCrit
.
- Add
binwidth
argument to histogram plots. - The
"performance"
plot always connected the maximum performance values. Now the minimum values are connected when the measure is minimized.
- Add
theme
option toautoplot()
functions to supply aggplot2::theme()
. The default isggplot2::theme_minimal()
. - Remove
theme_mlr3()
. - Unify plot layouts.
- Remove support for passing extra arguments to
geom_
functions via...
. This behavior was not consistent across theautoplot()
functions.
- Improved documentation.
- Make checks run without suggested packages.
- Plots for survival objects moved to mlr3proba.
- Compatibility fixes.
- Compatibility fix for testthat.
- Compatibility fix for mlr3tuning.
- Fixed position of labels in barplot for
PredictionClassif
.
- Fixed another bug for ROC- and Precision-recall-curves (#79).
- Fixed a bug for ROC- and Precision-recall-curves (#72, #75).
- New autoplot for
LearnerClustHclust
.
- New autoplot for
TuningInstanceSingleCrit
from packagemlr3tuning
. - Fixed bugs in autoplot function for
BenchmarkResult
(#63, #65). - Fixed a bug in autoplot function for
PredictionClust
(#67).
- Fix test on solaris.
- New autoplot for
PredictionSurv
. - New autoplots for learners from package
glmnet
viaggfortify
. - Fixed ROC and PRC plots for resampling
"holdout"
(#54). - If possible, we now show confidence bounds for ROC and PRC plots (#55).
- Fixed a bug in autoplot function
TaskDens
(#57).
- All
autoplot.*()
functions now also have a generic S3plot()
sibling (#51). - New plots for cluster tasks, learners and predictions from
mlr3cluster
. - Fixed wrong labels for boxplots of
BenchmarkResult
.
- Compatibility with
mlr3
>= 0.6.0.
- Added plots for
TaskDens
andTaskSurv
from packagemlr3proba
. - Update documentation of
PredictionRegr
(#23) autoplot.BenchmarkResult()
: Support for learners with identical IDs (#19)- Fixed a bug in
plot_learner_prediction()
(#47)
- New plot: learner prediction for objects of class
ResampleResult
. Additionally, the helper functionplot_learner_prediction()
first performs aresample()
and then plots the result. - New plot: residual plot for objects of class
PredictionRegr
.
- Initial CRAN release