textrecipes
Text features finally stay sparse all the way to the model.
A side-by-side editorial comparison of loo and tidytext — release velocity, themes, recent moves, and the top alternatives to consider.
loo keeps rewriting the diagnostics Bayesian modellers read off model comparison
loo computes leave-one-out cross-validation and model comparison for Bayesian models in the Stan ecosystem. Two releases in this window changed what users actually read: 2.7.0 replaced the fixed Pareto-k thresholds with sample-size-dependent ones and dropped the middle category, and 2.10.0 reshaped loo_compare's output into a data.frame with new uncertainty columns. The releases between are diagnostic robustness fixes and moment-matching corrections.
Finished, widely taught, and shipping roxygen fixes.
tidytext is the package that made unnest_tokens() and the tidy-data approach to text analysis standard, and it has reached the point where its releases contain nothing to announce. The last three are roxygen package anchors, alt text on vignette figures, and a single bug fix in one stm tidier. The most recent substantive changes were in 0.4.0 and 0.3.3 — stm tidiers for high FREX and lift words, a labels function for scale_x_reordered(), and support for tidying STM models that use content covariates.
loo computes leave-one-out cross-validation and model comparison for Bayesian models in the Stan ecosystem. Two releases in this window changed what users actually read: 2.7.0 replaced the fixed Pareto-k thresholds with sample-size-dependent ones and dropped the middle category, and 2.10.0 reshaped loo_compare's output into a data.frame with new uncertainty columns. The releases between are diagnostic robustness fixes and moment-matching corrections.
The package is being brought in line with the current PSIS literature rather than extended with new features, and the practical effect is that the numbers practitioners quote in papers keep changing meaning. Work is increasingly delegated to posterior for shared computations, and the project has added contributor process, benchmarks and a published AI contribution policy.
Expect further work on comparison diagnostics — the p_worse and diag_* columns are new enough that their defaults and documentation will likely be revised next.
tidytext is the package that made unnest_tokens() and the tidy-data approach to text analysis standard, and it has reached the point where its releases contain nothing to announce. The last three are roxygen package anchors, alt text on vignette figures, and a single bug fix in one stm tidier. The most recent substantive changes were in 0.4.0 and 0.3.3 — stm tidiers for high FREX and lift words, a labels function for scale_x_reordered(), and support for tidying STM models that use content covariates.
The direction is stability, and the release triggers are external. quanteda releases force updates to the dfm tidiers, a Matrix release forces another, tokenizers deprecating its tweet tokenizer forces removal of the tweet-specific functions here, and CRAN's Rd anchor requirement produces a release of its own. Nothing in the recent stream suggests new capability is planned, and for a package this embedded in teaching material that is a defensible position rather than a problem.
The entries do not support predicting new features. The likely next release is another compatibility update prompted by quanteda, stm, or a CRAN documentation requirement.
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 loo or tidytext.
Text features finally stay sparse all the way to the model.
The package that made calibration a step instead of an afterthought.
workflowsets keeps widening what counts as a model worth comparing.
The tidymodels pipeline grew a third stage, and it happens after the model runs.
Posit's MLOps package went quiet for two years, then came back to keep up with recipes.
patchwork stopped being a ggplot composer and became a page composer.
See all loo alternatives → · See all tidytext alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. loo is currently shipping more aggressively (velocity 2.5 vs 0.0), with 0 editorial sparks in the last 30 days against 0. 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. loo is currently shipping more aggressively (velocity 2.5 vs 0.0), with 0 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other Analytics products to evaluate alongside.
Top loo alternatives in Analytics are ranked by recent ship velocity. Browse the "loo alternatives" section above for the current picks, or visit /alternatives/loo for the full list with editorial commentary on each.
Top tidytext alternatives in Analytics are ranked by recent ship velocity. Browse the "tidytext alternatives" section above for the current picks, or visit /alternatives/tidytext for the full list with editorial commentary on each.