epiflows
epiflows has shipped four releases in eight years, none of which changed the code.
A side-by-side editorial comparison of bittermelon and kernelshap — release velocity, themes, recent moves, and the top alternatives to consider.
bittermelon is growing from binary bitmaps toward greyscale and color glyphs
The package manipulates bitmap fonts and bitmaps in R, and the recent direction is widening what a glyph can be. v2.3.1 adds greyscale font support by reading the experimental yaff levels property into pixmap glyphs with alpha derived from the level fraction, and makes writing a multi-colored glyph an error rather than silently clamping it. Earlier releases built out the same surface: bm_extract(), an interactive pixel picker, data frame coercion, and format support for other packages' bitmap types.
kernelshap makes permutation SHAP practical past eight features, then fixes the kernel weights it had wrong.
kernelshap computes model-agnostic SHAP values in R through Kernel SHAP, permutation SHAP and an exact additive explainer. Version 0.8.0 added a sampling permutation-SHAP algorithm with standard errors and early stopping, lifting the practical feature ceiling past what the exact method allows. Version 0.9.0 then corrected a bug in how kernel weights were computed — exact Kernel SHAP now agrees with exact permutation SHAP — and moved parallelism from foreach to doFuture.
The package manipulates bitmap fonts and bitmaps in R, and the recent direction is widening what a glyph can be. v2.3.1 adds greyscale font support by reading the experimental yaff levels property into pixmap glyphs with alpha derived from the level fraction, and makes writing a multi-colored glyph an error rather than silently clamping it. Earlier releases built out the same surface: bm_extract(), an interactive pixel picker, data frame coercion, and format support for other packages' bitmap types.
Two things are converging. The representation is moving beyond one-bit glyphs — pixmaps, color, and now greyscale levels with transparency — while the API gets stricter about the boundary between them, refusing to quietly downcast a color glyph on write. Separately, the package has been shedding weight: the embedded monobit copy was removed once it outgrew CRAN size limits, pushing format breadth onto a user-installed dependency.
Expect further greyscale and color work now that levels parsing exists — most likely writing that format rather than only reading it, since the write path is where the current release chose to draw a hard error instead.
kernelshap computes model-agnostic SHAP values in R through Kernel SHAP, permutation SHAP and an exact additive explainer. Version 0.8.0 added a sampling permutation-SHAP algorithm with standard errors and early stopping, lifting the practical feature ceiling past what the exact method allows. Version 0.9.0 then corrected a bug in how kernel weights were computed — exact Kernel SHAP now agrees with exact permutation SHAP — and moved parallelism from foreach to doFuture.
Two concerns drive this package: making exact methods reach further, and being demonstrably right. The first shows in the additive explainer, the optional background dataset and the sampling permutation algorithm; the second in unit tests written against Python's shap, credited fixes from outside contributors, and a willingness to ship a correctness fix that changes numbers people have already published. Speed work runs continuously underneath — direct solves replacing the Moore-Penrose pseudo-inverse, roughly 10% less memory.
The 0.6.0 and 0.7.0 notes each promised a stable 1.0.0 that has not arrived; with the weighting bug fixed and parallelism reworked, a 1.0 release is the most plausible next step.
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 bittermelon or kernelshap.
epiflows has shipped four releases in eight years, none of which changed the code.
The LPJmL toolkit finally reads NetCDF, closing a gap against the format its field uses.
survminer is back after a five-year gap, and spending it cutting dependencies loose.
A stock-assessment plotting package is growing a table engine to match its figures.
xts is finished software, and its releases now track R's C API more than user requests.
A thin R wrapper around a Java seasonal-adjustment tool, slowly absorbing the setup work.
See all bittermelon alternatives → · See all kernelshap alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. bittermelon and kernelshap 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.
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. bittermelon and kernelshap 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.
Top bittermelon alternatives in Analytics are ranked by recent ship velocity. Browse the "bittermelon alternatives" section above for the current picks, or visit /alternatives/bittermelon-r for the full list with editorial commentary on each.
Top kernelshap alternatives in Analytics are ranked by recent ship velocity. Browse the "kernelshap alternatives" section above for the current picks, or visit /alternatives/kernelshap for the full list with editorial commentary on each.