NEWS.md

In mlr-org/mlr3proba: Probabilistic Supervised Learning for 'mlr3'

mlr3proba 0.7.5

• fix: allow cloning of measures objects
• New PipeOps: PipeOpTaskSurvRegrPEM, PipeOpPredRegrPEM
• New pipeline (reduction method): pipeline_survtoregr_pem

mlr3proba 0.7.4

• fix + update MeasureSurv: survival measure labels are now printed and the obs_loss property is now supported
• feat: add na.rm parameter to msr("surv.calib_index") to avoid NaN scores

mlr3proba 0.7.3

• feat: added new calibration measure => msr("surv.calib_index")
• refactor + feat: autoplot.PredictionSurv
• The default "calib" plot uses the survival matrix directly now which is faster
• "dcalib" has extra barplot + better documentation
• Added new type = "scalib" which constructs the smoothed calibration plots as in Austin et al. (2020)
• BREAKING CHANGE: "preds" is now called "isd" (individual survival distribution). row_ids can now be used to filter the observations for which you draw the survival curves.

mlr3proba 0.7.2

• fix: lrn("surv.coxph") is now trained with model=TRUE which fixes an issue with using observation weights stackoverflow link.
• cleanup: remove tsk("unemployment") and associated files
• cleanup: remove unused references

mlr3proba 0.7.1

• cleanup: removed all PipeOps and pipelines related to survival => regression reduction techniques (see #414)
• fix: $predict_type of survtoclassif_disctime and survtoclassif_IPCW was prob (classification type) and not crank (survival type)
• fix: G(t) is not filtered when t_max|p_max is specified in scoring rules (didn't influence evaluation at all)
• docs: Clarified the use and impact of using t_max in scoring rules, added examples in scoring rules and AUC scores
• feat: Added new argument remove_obs in scoring rules to remove observations with observed time t > t_max as a processing step to alleviate IPCW issues. This was before 'hard-coded' which made the Integrated Brier Score (msr("surv.graf")) differ minimally from other implementations and the original definition.

mlr3proba 0.7.0

• Add mlr3pipelines to Imports and set minimum latest version from CRAN (0.7.0)
• Refactor code to minimize namespace calling and imports such as mlr3pipelines:: or R6::
• Doc updates: add experimental badge in a some PipeOps + add references in others
• Add argument scale_lp for AFT distrcompose pipeop + respective pipeline

mlr3proba 0.6.9

• New PipeOps: PipeOpTaskSurvClassifIPCW, PipeOpPredClassifSurvIPCW
• New pipeline (reduction method): pipeline_survtoclassif_IPCW
• Improved the way Integrated Brier score handles the times argument and the t_max, especially when the survival matrix has one time point (column)
• Improved documentation of integrated survival scores
• Improved documentation of all pipelines
• Temp fix of math-rendering issue in package website
• Add experimental lifecycle badge for 3 pipelines (survtoregr, distrcompositor and probregr) - these are currently either not supported by literature or tested enough.

mlr3proba 0.6.8

• Rcpp code optimizations
• Fixed ERV scoring to comply with mlr3 dev version (no bugs before)
• Skipping survtoregr pipelines due to bugs (to be refactored in the future)

mlr3proba 0.6.7

• Deprecate crank to distr composition in distrcompose pipeop (only from lp => distr works now)
• Add get_mortality() function (from survivalmodels::surv_to_risk()
• Add Rcpp function assert_surv_matrix()
• Update and simplify crankcompose pipeop and respective pipeline (no response is created anymore)
• Add responsecompositor pipeline with rmst and median

mlr3proba 0.6.6

• Small fixes and refactoring to the discrete-time pipeops

mlr3proba 0.6.5

• Add support for discrete-time survival analysis
• New PipeOps: PipeOpTaskSurvClassifDiscTime, PipeOpPredClassifSurvDiscTime
• New pipeline (reduction method): pipeline_survtoclassif_disctime

mlr3proba 0.6.4

• Add useR! 2024 tutorial
• Lots of refactoring, improving code quality, migration to testthat v3, etc. (thanks to @m-muecke)

mlr3proba 0.6.3

• Add new tasks from survival package: veteran, pbc, mgus, gbsg
• Refined docs for loaded tasks, task$help() works as it should now
• All loaded tasks have now complete cases (no features with missing values)
• Refined docs for task generators
• Added task generator from coxed package
• Added new methods for TaskSurv: cens_prop(), admin_cens_prop(), dep_cens_prop() and prop_haz()

mlr3proba 0.6.2

• Updates in surv.cindex measure
• added p_max (same as surv.graf)
• refactor cutoff to t_max

mlr3proba 0.6.1

• Compatibility with upcoming 'paradox' release.
• Fix bug when applying t_max in surv.graf and similar time-integrated scores.

mlr3proba 0.6.0

• Optimized surv.logloss and calib_alpha measures (bypassing distr6)
• Update/refine all measure docs (naming conventions from upcoming scoring rules paper) + doc templates
• fix very rare bugs in calib_alpha, surv.logloss and surv.graf (version with proper = FALSE)

mlr3proba 0.5.9

• Fix several old issues (#348, #301, #281)
• distrcompositor and crankcompositor deprecated functions were completely removed

mlr3proba 0.5.8

• Fix Breslow issues (phash warning and Inf lp predictions)

mlr3proba 0.5.7

• Add breslow function for estimating the cumulative baseline hazard of proportional hazard models
• Add PipeOpBreslow to wrap a survival learner and generate distr predictions from lp predictions
• Add option breslow estimator option in distrcompositor

mlr3proba 0.5.6

• Add extend_quantile to autoplot.PredictionSurv for type = "dcalib", which imputes NAs with the maximum observed survival time
• Fixes default in autoplot.PredictionSurv, now "calib"
• Update msr("surv.dcalib") default for truncate to Inf

mlr3proba 0.5.5

• Add $reverse() method to TaskSurv, which returns the same task but with 1-status.
• Add reverse parameter to TaskSurv$kaplan() method, which calculates Kaplan-Meier on the censoring distribution of the task (1-status).

mlr3proba 0.5.4

• Fix bottlenecks in Dcalib and RCLL

mlr3proba 0.5.3

• Add support for learners that can predict multiple posterior distributions by using distr6::Arrdist

mlr3proba 0.5.2

• Add plot_probregr for plotting probabilistic regression distribution predictions
• Fix big bug in surv.rcll creating erroneous results as distributions were not being subsetted correctly

mlr3proba 0.5.1

• Add regr.logloss

mlr3proba 0.5.0

• Possibly small breaking change, renamed PipeOpProbregrCompositor to PipeOpProbregr and default distribution now "Uniform".
• Renamed probregrcompositor pipeline to probregr and default distribution now "Uniform".
• Bumped minor version to highlight all fixes in measures below
• Bugfix in setting Uno's AUC parameters
• Bugfix in RCLL when no censoring
• Bugfix in all learners when making single predictions

mlr3proba 0.4.17

• Fix bug in surv.rcll when individual scores are NA.

mlr3proba 0.4.16

• Remove theme_mlr3 from plots.
• Fix bug in surv.logloss when using IPCW = TRUE

mlr3proba 0.4.15

• Fix bug in surv.cindex caused when probability of censoring was 0. Added eps parameter to control this.

mlr3proba 0.4.14

• Fix bug in converting distr6 distributions to matrices when creating PredictionSurv
• Fix bug in RCLL

mlr3proba 0.4.13

• Fix minor bug in 'SG' cindex method

mlr3proba 0.4.12

• Fix bug in survavg pipeline causing unequal weights not to be applied

mlr3proba 0.4.11

• Fix minor bug in scoring rules

mlr3proba 0.4.10

• Added ERV parameter to scoring rule measures to return more interpretable scoring rules. Explained Residual Variation is the percentage decrease between a scoring rule comparing a Kaplan-Meier baseline to the learner of interest.

mlr3proba 0.4.9

• Fixed bug in surv.logloss causing IPCW weighting to not be applied correctly

mlr3proba 0.4.8

• Bug fixes in AUC measures

mlr3proba 0.4.7

• Add right-censored log loss
• Fix bug in {rpart} where model was being discarded when set to be kept. Parameter model now called keep_model.

mlr3proba 0.4.6

• Patch for upstream breakages
• Add TaskSurv$kaplan method
• {survivalmodels} now imported (previously suggested)

mlr3proba 0.4.5

• Improved reduction from survival matrix predictions to ranking predictions
• Fixed cindex bug when all predictions equal
• Fix for valgrind

mlr3proba 0.4.4

• Minor change to how distributions are created to better support improper distributions
• Fixed bug in simsurv task that made it impossible to predict the target

mlr3proba 0.4.3

• Massive speed-up in distrcompositor PipeOp/pipeline
• More informative error given if $distr called for a learner that does not support this return type
• Fix massive bottleneck in scoring rule measures
• Add Density coercions as_task_dens and as_prediction_dens
• Measures now use parameter sets like learners. This streamlines the interface but unfortunately means ids can no longer be set dynamically.
• Add parameters t_max and p_max to Graf, Schmid and Integrated Log-loss as an alternative to times. t_max is equivalent to times = seq(t_max) and p_max is the proportion of censoring to integrate up to in the dataset.
• Fix bug in Rcpp code that was causing erroneous values for calculating the cindex in datasets greater than 20,000 observations.

mlr3proba 0.4.2

• Patch for linux

mlr3proba 0.4.1

• Remove mlr3extralearners from Suggests
• Add response to as_prediction_surv
• Now exported a couple cpp functions and assert_surv
• mlr3 is now in Depends not imports
• distr predictions are now internally stored as matrices to significantly reduce prediction object sizes
• Tasks now support strata property

mlr3proba 0.4.0

• Deprecated measures from 0.2.0 have now been deleted.
• IPCW measures such as surv.graf, surv.schmid, and surv.intlogloss now allow training data to be passed to the score function with task and train_set to allow the censoring distribution to be estimated on the training data. This is automatically applied for resample and benchmark results.
• IPCW measures such as surv.graf, surv.schmid, and surv.intlogloss now include a parameter proper to determine what weighting scheme should be applied by the estimated censoring distribution, The current method (Graf, 1999) proper = FALSE, weights observations either by their event time or 'current' time depending if they're dead or not, the new method proper = TRUE weights observations by event time. The proper = TRUE method is strictly proper when censoring and survival times are independent and G is estimated on large enough data. The proper = FALSE method is never proper. The default is currently proper = FALSE to enable backward compatibility, this will be changed to proper = TRUE in v0.6.0.
• The rm_cens parameter in surv.logloss has been deprecated in favour of IPCW. rm_cens will be removed in v0.6.0. If rm_cens or IPCW are TRUE then censored observations are removed and the score is weighted by an estimate of the censoring distribution at individual event times. Otherwise if rm_cens and IPCW are FALSE then no deletion or weighting takes place. The IPCW = TRUE method is strictly proper when censoring and survival times are independent and G is estimated on large enough data. The ipcw = FALSE method is never proper.
• Add surv.dcalib for the D-Calibration measure from Haider et al. (2020).

mlr3proba 0.3.2

• Patched bug causing "interval2" task type not to work
• Fixed bug causing pipelines not to function correctly in $aggregate

mlr3proba 0.3.1

• Reverted removal of "interval2"

mlr3proba 0.3.0

• Commonly used survival quantities have been added as active bindings to TaskSurv including times (observed survival times), status (observed survival indicator), unique_times (set of sorted unique outcome times), unique_event_times (set of sorted unique failure times), risk_set (set of observations alive 'just before' a given time)
• "interval2" censoring type has been removed from TaskSurv as this is covered by the other types
• Default values have now been given to the time and event arguments in TaskSurv
• PredictionDens can now include distr return type (equivalent to learner$model)

mlr3proba 0.2.6

• Minor internal fixes

mlr3proba 0.2.5

• PipeOpCrankCompositor updated to fix bottleneck in computation via mean. Now Inf or NA is replaced by 0 for response and imputed with the median for crank
• Bug in distr predict types fixed that lead to fitting degenerate distributions and returning incorrect values for mean survival time and crank

mlr3proba 0.2.4

• CRITICAL BUG FIX - compose_crank was previously returning ranks with the reverse ordering so that higher ranks implied higher risk not lower.

mlr3proba 0.2.3

• All learners that previously lived in the mlr3learners organisation are now in the mlr3extralearners repository.
• Fixed bottleneck in MeasureSurvLogloss
• Bugfix in MeasureSurvCalibrationAlpha
• Patch for valgrind
• TaskDens now inherits from TaskUnsupervised which means target/truth has been removed. No specification of a target column is required, instead a one-column matrix-like object or numeric vector should be passed to the task backend and the density will be estimated for this column, or two columns and one set as weight.
• Fixed bug in load_eruption to fix name of data columns
• Added calibration plot for comparing average predicted survival distribution to Kaplan-Meier to mlr3viz
• Removed unneccessary pracma dependency in learners
• Fix in PipeOpDistrCompositor, previously base distribution was only using the first predicted distribution, now the baseline is taken by averaging over all predictions with uniform weights

mlr3proba 0.2.2

• Default kernel for LearnerDensityKDE is now Epan to reduce imports
• Minor internal patches for mlr3 0.6.0
• Bug fix in MeasureSurvCalibrationBeta now returns NA not error if lp predict type not available

mlr3proba 0.2.1

• Removed PredictionRegr causing masking issues with {mlr3}
• Bug fix in PipeOpDistrCompositor causing some cdf predictions to be lost
• Internal fixes for mlr3pipelines: public train and predict methods to private
• Added four datasets and tasks: grace, actg, gbcs, whas
• Add overwrite to crankcompositor pipeop and pipeline
• Bug fix in surv.kaplan crank prediction

mlr3proba 0.2.0

Added Functionality

• MeasureSurvCindex added. Generalises all c-index measures with a fast C++ implementation
• Akritas estimator added to mlr3learners/mlr3learners.proba
• Added scoring rule MeasureSurvSchmid
• Addd calibration measures MeasureSurvCalibrationBeta and MeasureSurvCalibrationAlpha
• surv.brier alias added for surv.graf
• response parameter added to PipeOpCrankCompositor and crankcompositor to now optionally fill response predict type with same values as crank
• Added PipeOpProbregrCompostior and compose_probregr for composition to distr return type from (a) regression learner(s) predicting response and se
• Added PipeOpSurvAvg and surv_averager pipeline for weighted model averaging of distr, lp, crank, and response predictions.

Deprecated Functionality

• The following measures are deprecated use MeasureSurvCindex instead with following parameters: MeasureSurvBeggC, use defaults; MeasureSurvHarrellC, use defaults; MeasureSurvUnoC, use weight_meth = 'G/2'; MeasureSurvGonenC, use weight_method = 'GH'
• MeasureSurvGrafSE, MeasureSurvLoglossSE, MeasureSurvIntLoglossSE, MeasureSurvRMSESE, MeasureSurvMSESE, and MeasureSurvMAESE all deprecated and will be deleted in v0.4.0. Use msr("surv.graf", se = TRUE) instead (for example).
• Measures renamed such that surv.nagelkR2 is now surv.nagelk_r2, analogously for all R2, AUC, TPR, and TNR measures. Old constructors will be deleted in v0.4.0.
• Renamed distrcompose and crankcompose to distr_compose and crank_compose. Old ids will be deleted in v0.4.0.

Edited Functionality

• Measures renamed such that surv.nagelkR2 is now surv.nagelk_r2, analogously for all R2, AUC, TPR, and TNR measures. Old constructors will be deleted in v0.4.0.
• MeasureSurvGraf and MeasureSurvIntLogloss now have much faster C++ implementation

Moved Functionality

• LearnerSurvGlmnet, LearnerSurvCVGlmnet, LearnerSurvXgboost and LearnerSurvRanger have been moved to mlr-org/mlr3learners

• LearnerSurvGBM has been moved to https://www.github.com/mlr3learners/mlr3learners.gbm

• LearnerSurvMboost, LearnerSurvGlmBoost, LearnerSurvGamboost, LearnerSurvBlackboost have been moved to https://www.github.com/mlr3learners/mlr3learners.mboost

mlr3proba 0.1.6

• Early release due to backward compatibility error introduced by an upstream dependency
• Minor updates to mboost family of learners: added gehan family, fixed parameters for cindex, added support for: weights, response predict type, importance, selected_features
• Minor internal changes
• All density learners except LearnerDensHist and LearnerDensKDE have been moved to the mlr3learners org
• The following survival learners have been moved to the mlr3learners org, LearnerSurv: Flexible, ObliqueRSF, Penalized, RandomForestSRC
• Bugfix in LearnerSurvXgboost previously lp was erroneously returned as exp(lp)
• Now licenced under LPGL-3

mlr3proba 0.1.5

• LearnerSurvParametric and LearnerSurvNelson moved to mlr3learners/mlr3learners.survival repo
• LearnerSurvCoxboost and LearnerSurvCVCoxboost moved to mlr3learners/mlr3learners.coxboost repo
• LearnerSurvSVM moved to mlr3learners/mlr3learners.survivalsvm repo
• In the next release, all learners except for LearnerSurvKaplan, LearnerSurvCoxPH, and LearnerDensHist will be moved to the mlr3learners org
• Minor internal changes

mlr3proba 0.1.4

• Density estimation has now been added to mlr3proba, see TaskDens, LearnerDens, PredictionDens, and MeasureDens.
• Added mlr_tasks_faithful and mlr_tasks_precip for density task examples
• Added mlr_task_generators_simdens for generating density tasks
• Added learners for density estimation, see mlr3::mlr_learners$keys("^dens") for the full list
• In line with mlr3 0.1.7, public methods train_internal, predict_internal, score_internal are now private methods .train,.predict,.score
• Converted to roxygen2 R6 documentation

mlr3proba 0.1.3

• Changed lp in surv.parametric to include the intercept, which is in line with survival::survreg. Now exp(pred$lp) is equal to the predicted survival time for AFTs
• Moved mboost to suggests
• Added response predict type, which predicts the time until event. Currently only supported for AFT models in surv.parametric
• Added measures for response predict type: MeasureSurvMAE, MeasureSurvMAESE, MeasureSurvMSE, MeasureSurvMSESE, MeasureSurvRMSE, MeasureSurvRMSESE

mlr3proba 0.1.2

• Fixed error in r-patched-solaris
• Added mode option to crankcompositor
• Fixes bug resulting from R62S3 incompatibility

mlr3proba 0.1.1

• Added method argument to integrated scores and added weighting by bin-width
• Added notes to IGS documentation regarding default methods and comparison to other packages
• Added method to MeasureSurvIntegrated constructor and fields
• Fixed mistake in documentation of: TaskSurv, MeasureSurvUnoC
• Added missing LearnerSurvRpart parameter parms and cost
• Fixed errors in r-patched-solaris and r-devel debian-clang

mlr3proba 0.1.0

• Initial upload to CRAN.

mlr-org/mlr3proba documentation built on April 12, 2025, 4:38 p.m.