LoginSign Up
← Back to all sparks
L

lime

ANALYTICS
Velocity0.0

Local Interpretable Model-Agnostic Explanations (R port of original Python package)

github.com/tidymodels/lime ↗

The R port of LIME has shipped one commit in three years, and it was an xgboost compatibility patch

explainabilitymachine-learningmaintenance-moder-packagesmodel-agnostic
◆Current state
lime is the R implementation of local interpretable model-agnostic explanations, and it is effectively in preservation rather than development. Its last substantive feature release was 0.5.0 in 2019; 0.5.3 in 2022 recorded a maintainer handover and general upkeep; 0.5.4 in December 2025 exists solely to keep the package working across xgboost versions. Six releases span eight years, and only two of them contain features.
◆Where it's heading
The pattern is upstream-driven survival: every release since 0.5.0 responds to a change in something lime depends on — glmnet's namespace, order() semantics on data frames, xgboost's interface. The one deliberate change in that stretch was moving htmlwidgets, shiny and shinythemes to Suggests, which lightens installation for the majority of users who never open the interactive explainer. Nothing in the feed indicates work on the explanation method itself.
◆Prediction
Expect the next release, whenever it comes, to be another compatibility patch triggered by a dependency change rather than anything touching how explanations are computed. The three-year gaps make timing unpredictable.

◆Recent moves

  1. 8mo ago

    xgboost compatibility restored across versions

    Makes the package work across all xgboost versions after three years of silence. Narrow in scope, but xgboost is one of the model backends lime explains, so this is the difference between the package working and not working for a common use case.

    View source ↗
  2. 3y ago

    Maintainer handover and general upkeep

    Records that Emil Hvitfeldt has taken over maintenance, plus general upkeep. No functional change, though the handover is the clearest marker in the feed that the original author had moved on.

    View source ↗
  3. 5y ago

    Shiny dependencies moved to Suggests

    Fixes order() applied to data.frame objects and moves htmlwidgets, shiny and shinythemes to Suggests. The dependency move lightens installation for anyone not using the interactive explainer, which is most callers.

    View source ↗
  4. 6y ago

    Namespace fix for glmnet changes

    A one-line namespace fix following changes in glmnet. Representative of what most releases in this feed are: reacting to a dependency rather than changing lime.

    View source ↗
  5. 7y ago

    gower_pow added and lambda aligned with the Python implementation

    The last release with real feature work. explain() forwards ... to the underlying predict() method, explain.data.frame() gains gower_pow to reshape the gower distance, and the lambda in the local model fit is changed to match the Python implementation — which means explanations produced before and after this release can differ. Also warns and falls back when quantile binning is not feasible.

    View source ↗
  6. 8y ago

    h2o support, NA handling, and date columns held during permutation

    The broadest release in the window: h2o joins the supported backends, NA values become acceptable in both training and explanation data, Date and POSIXt columns are held constant during permutation so explanations are read at a fixed timepoint, and plot_explanations() gives an overview across large explanation sets. This is the release where lime became usable on realistic data rather than clean examples.

    View source ↗