OHPL
A 2017 chemometrics method frozen in place, visited only when CRAN changes its documentation rules.
Comparison · Analytics
geoarrow-r vs kernelshap
A side-by-side editorial comparison of geoarrow-r and kernelshap — release velocity, themes, recent moves, and the top alternatives to consider.
geoarrow-r vs kernelshap: at a glance
What is geoarrow-r?
geoarrow tracks the GeoArrow spec and otherwise just keeps compiling
geoarrow gives R zero-copy access to geospatial data in the Arrow columnar format, with conversions to and from sf and wk. The package is thin by design — most of the work lives in vendored copies of geoarrow-c and nanoarrow — and its release notes reflect that: the substantive one in the window implemented GeoArrow 0.2 specification features, and everything since has been compiler warnings and a test fixed for a new sf version.
What is kernelshap?
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.
geoarrow-r vs kernelshap: editorial side-by-side
geoarrow-r
0.0ANALYTICS
geoarrow tracks the GeoArrow spec and otherwise just keeps compiling
◆ Current state
geoarrow gives R zero-copy access to geospatial data in the Arrow columnar format, with conversions to and from sf and wk. The package is thin by design — most of the work lives in vendored copies of geoarrow-c and nanoarrow — and its release notes reflect that: the substantive one in the window implemented GeoArrow 0.2 specification features, and everything since has been compiler warnings and a test fixed for a new sf version.
◆ Where it's heading
Development is downstream of two things the package does not control: the GeoArrow specification and the C libraries it vendors. When the spec added non-PROJJSON CRS types and spheroidal edge interpolations, the R package followed; when the geoarrow.box type appeared, it gained conversions to and from wk::rct(). Between those, releases exist to keep CRAN builds green. The type-selection work in 0.4.0 — letting callers force geoarrow.wkb output when converting from sf — is the only recent change driven by R-side ergonomics rather than upstream.
◆ Prediction
Given that every feature release so far has implemented a spec revision, the next one likely arrives when GeoArrow publishes its next set of extension types rather than on any schedule of its own.
kernelshap
0.0ANALYTICS
kernelshap makes permutation SHAP practical past eight features, then fixes the kernel weights it had wrong.
◆ Current state
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.
◆ Where it's heading
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.
◆ Prediction
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.
Alternatives to geoarrow-r and kernelshap
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 geoarrow-r or kernelshap.
fastplyr
A fast dplyr stand-in that keeps finding new places to skip work entirely.
trias
Belgium's invasive-species indicator toolkit is in steady refinement, one plotting edge case at a time.
icesSAG
The ICES stock assessment client took upload away in 2024 and spent two years giving it back.
dggridR
A discrete global grid generator grew cell traversal and became a usable spatial index.
vegan
Community ecology's standard toolkit is retiring the functions a generation of scripts was built on.
See all geoarrow-r alternatives → · See all kernelshap alternatives →
Recent activity from geoarrow-r and kernelshap
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
- 2mo agogeoarrow-rgeoarrow 0.4.3
- 8mo agogeoarrow-rgeoarrow 0.4.1
- 10mo agogeoarrow-rgeoarrow lets callers choose the geometry type when converting from sf
- 1y agokernelshapKernel weight bug fixed; parallelism moves to doFuture
- 1y agokernelshapSampling permutation SHAP with standard errors ⚡
- 1y agogeoarrow-rgeoarrow implements the GeoArrow 0.2 specification features
- 1y agokernelshapBackground data now optional; ranger survival support
- 2y agokernelshapFactor-valued predictions dropped
- 2y agogeoarrow-rgeoarrow 0.2.1
- 2y agokernelshapadditive_shap() explains additive models exactly
- 2y agogeoarrow-rLegacy geoarrow tagged before the rewrite
- 2y agokernelshapFaster on plain data.frames
Frequently asked questions
What is the difference between geoarrow-r and kernelshap?
They serve adjacent needs but don't currently overlap on shipped themes. geoarrow-r 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.
Is geoarrow-r better than kernelshap?
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. geoarrow-r 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.
What are the best alternatives to geoarrow-r?
Top geoarrow-r alternatives in Analytics are ranked by recent ship velocity. Browse the "geoarrow-r alternatives" section above for the current picks, or visit /alternatives/geoarrow-r for the full list with editorial commentary on each.
What are the best alternatives to kernelshap?
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.