paleobuddy
paleobuddy can now simulate trait-dependent diversification, not just birth-death.
A side-by-side editorial comparison of modeltime.resample and parallelDist — release velocity, themes, recent moves, and the top alternatives to consider.
modeltime.resample exists to keep backtesting working as tidymodels shifts underneath it.
modeltime.resample runs time series cross-validation over modeltime models, returning per-resample predictions and accuracy plots. Version 0.3.0 is the substantive release in view: tune 2.0.0 compatibility, deterministic seeding via withr, guaranteed .predictions output, and clearer failures when resample fits break. The three releases before it are dependency chores.
parallelDist is in pure preservation mode — one build fix every few years.
parallelDist computes distance matrices across threads in C++ via RcppParallel and Armadillo. The feature set has been settled since 0.2.3 in 2018, which added hamming distance and cosine similarity; everything after that is compatibility work. The most recent release, 0.2.7, exists only to drop a C++11 pin that newer Armadillo versions no longer tolerate.
modeltime.resample runs time series cross-validation over modeltime models, returning per-resample predictions and accuracy plots. Version 0.3.0 is the substantive release in view: tune 2.0.0 compatibility, deterministic seeding via withr, guaranteed .predictions output, and clearer failures when resample fits break. The three releases before it are dependency chores.
Every entry here is compatibility work against something upstream — hardhat 1.0.0, workflows regression mode, then tune 2.0.0 twice. The 0.3.0 notes show a second concern emerging alongside it: making failures legible, with .notes on failed fits, actionable errors from unnest_modeltime_resamples(), and fallback logic when prediction columns go missing across versions. Reproducibility gets the same treatment through explicit seeding.
Expect the next release to track the next tidymodels breaking change, with any new work continuing on error reporting rather than resampling strategies.
parallelDist computes distance matrices across threads in C++ via RcppParallel and Armadillo. The feature set has been settled since 0.2.3 in 2018, which added hamming distance and cosine similarity; everything after that is compatibility work. The most recent release, 0.2.7, exists only to drop a C++11 pin that newer Armadillo versions no longer tolerate.
The package is being kept installable, not developed. The three most recent releases are a toolchain pin removal, a DESCRIPTION field removal, and a coercion change inherited from proxy — none originate from user-facing intent. Gaps of three to four years between releases are the norm now, and each one is triggered by something upstream breaking rather than by a roadmap.
The next release will almost certainly be another compatibility fix timed to whatever Armadillo, Rcpp or CRAN check policy changes next. Nothing in the entries points to new distance measures or API work.
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 modeltime.resample or parallelDist.
paleobuddy can now simulate trait-dependent diversification, not just birth-death.
geodist stays dependency-free and fast, and warns you when 'cheap' distances stop being honest.
errors keeps making uncertainty print the way each scientific field expects.
CMAQ went global in v5.5, and has been patching that surface ever since.
enpls has not changed its statistics since 2016 — only its website, twice.
grex is a lookup table with a version number — it ships when the annotation moves.
See all modeltime.resample alternatives → · See all parallelDist alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. modeltime.resample and parallelDist 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. modeltime.resample and parallelDist 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 modeltime.resample alternatives in Analytics are ranked by recent ship velocity. Browse the "modeltime.resample alternatives" section above for the current picks, or visit /alternatives/modeltime-resample for the full list with editorial commentary on each.
Top parallelDist alternatives in Analytics are ranked by recent ship velocity. Browse the "parallelDist alternatives" section above for the current picks, or visit /alternatives/paralleldist for the full list with editorial commentary on each.