mlr3proba
mlr3proba is shedding weight as its survival work moves into sibling packages
Comparison · Analytics
modelbased vs probably
A side-by-side editorial comparison of modelbased and probably — release velocity, themes, recent moves, and the top alternatives to consider.
modelbased vs probably: at a glance
What is modelbased?
modelbased is turning marginal effects into a full contrast grammar
modelbased computes marginal means, contrasts, and slopes from fitted models, and it ships every one to two months with a consistent shape: new comparison types, broader model support, and steady renaming toward clearer vocabulary. The recent arc runs from marginal effects inequality measures through inequality ratios to an omnibus global test and a post_process argument for multi-step comparisons. Argument names have been settled along the way, with trend becoming slope and an alias left behind.
What is probably?
The package that made calibration a step instead of an afterthought.
probably started as a small utility for class predictions and equivocal zones, and version 1.0.0 turned it into tidymodels' calibration and uncertainty package: cal_plot_*, cal_estimate_*, cal_validate_* and cal_apply across binary, multiclass and regression problems, plus conformal prediction intervals. Since then the work has been consolidation — a large internal refactor with no API change, split conformal and conformal quantile regression, bound_prediction(), and required_pkgs() and butcher methods so conformal objects can be deployed and stripped.
modelbased vs probably: editorial side-by-side
modelbased
0.0ANALYTICS
modelbased is turning marginal effects into a full contrast grammar
◆ Current state
modelbased computes marginal means, contrasts, and slopes from fitted models, and it ships every one to two months with a consistent shape: new comparison types, broader model support, and steady renaming toward clearer vocabulary. The recent arc runs from marginal effects inequality measures through inequality ratios to an omnibus global test and a post_process argument for multi-step comparisons. Argument names have been settled along the way, with trend becoming slope and an alias left behind.
◆ Where it's heading
The package is building a compositional vocabulary rather than a fixed menu — contrasts of average slopes, contrasts across two numeric predictors, inequality summaries across all outcome categories, and now user-supplied post-processing of comparisons. Support quietly widens underneath, covering nestedLogit, brms finite mixtures, and offsets under population and average estimation. Plotting gets attention in proportion to how often these results are presented rather than tabulated, including collapse_by_group() for showing averaged raw data under mixed-model fits.
◆ Prediction
With post_process and omnibus tests both landed, the likely next step is making these composed comparisons easier to report — formatting or plotting methods for the multi-step results rather than new comparison types.
probably
0.0ANALYTICS
The package that made calibration a step instead of an afterthought.
◆ Current state
probably started as a small utility for class predictions and equivocal zones, and version 1.0.0 turned it into tidymodels' calibration and uncertainty package: cal_plot_*, cal_estimate_*, cal_validate_* and cal_apply across binary, multiclass and regression problems, plus conformal prediction intervals. Since then the work has been consolidation — a large internal refactor with no API change, split conformal and conformal quantile regression, bound_prediction(), and required_pkgs() and butcher methods so conformal objects can be deployed and stripped.
◆ Where it's heading
The recent releases are about making these objects survive leaving the session. butcher and required_pkgs() methods are what a model needs to be pinned, containerised and served, and their arrival alongside workflows adding a tailor postprocessing stage and vetiver adding probably support points the same way: calibration is being moved out of analysis scripts and into the deployed pipeline. The cal_*_none() reference implementations are the tell that calibration is now something people tune rather than apply once.
◆ Prediction
Expect the calibration functions to be reachable directly from a tuned workflow's postprocessing stage rather than applied to predictions afterwards, following the tailor integration that workflows just shipped.
Alternatives to modelbased and probably
Other Analytics products tracked by Sparkpulse, ranked by recent ship velocity. Each card links to a full editorial trajectory and lets you pivot into a head-to-head comparison with either modelbased or probably.
mlr3viz
mlr3viz keeps the ecosystem's plots working while the plots themselves move out
mlr3tuning
mlr3tuning is rebuilding its async machinery under a stable public surface
timetk
timetk swallowed anomalize whole, then went quiet for two years
parameters
easystats' parameters package absorbs one more model class every few weeks
tidyposterior
A finished Bayesian model-comparison package in pure maintenance mode
See all modelbased alternatives → · See all probably alternatives →
Recent activity from modelbased and probably
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
- 1mo agomodelbasedmodelbased 0.16.0 adds post-processing and omnibus contrast tests
- 3mo agomodelbasedmodelbased 0.15.0 contrasts average slopes across numeric predictors
- 5mo agomodelbasedmodelbased 0.14.0 renames trend to slope and adds collapse_by_group()
- 8mo agomodelbasedmodelbased 0.13.1 adds marginal group-level estimates and as.data.frame()
- 10mo agoprobablyConformal objects gain required_pkgs() and butcher methods
- 11mo agomodelbasedmodelbased 0.13.0 adds inequality ratios and slope marginalization
- 1y agoprobablyggplot2 test updates and a clearer validation-set error
- 1y agomodelbasedmodelbased 0.12.0 introduces marginal effects inequality measures
- 1y agoprobablyCalibration internals refactored; isotonic bootstrap bug fixed
- 2y agoprobablyFix grouping sensitivity to variable type
- 3y agoprobablySplit conformal and conformal quantile regression added
- 3y agoprobablyCalibration and conformal inference arrive in tidymodels ⚡
Frequently asked questions
What is the difference between modelbased and probably?
They serve adjacent needs but don't currently overlap on shipped themes. modelbased and probably are shipping at a similar cadence (velocity 0.0 vs 0.0, both within Sparkpulse's "active" band). See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Is modelbased better than probably?
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. modelbased and probably are shipping at a similar cadence (velocity 0.0 vs 0.0, both within Sparkpulse's "active" band). For your specific use case, the alternatives sections above list other Analytics products to evaluate alongside.
What are the best alternatives to modelbased?
Top modelbased alternatives in Analytics are ranked by recent ship velocity. Browse the "modelbased alternatives" section above for the current picks, or visit /alternatives/modelbased for the full list with editorial commentary on each.
What are the best alternatives to probably?
Top probably alternatives in Analytics are ranked by recent ship velocity. Browse the "probably alternatives" section above for the current picks, or visit /alternatives/probably for the full list with editorial commentary on each.