tidytlg
A tables-listings-graphs package that reached CRAN and then went quiet.
Comparison · Analytics
crul vs tabnet
A side-by-side editorial comparison of crul and tabnet — release velocity, themes, recent moves, and the top alternatives to consider.
crul vs tabnet: at a glance
What is crul?
crul took mocking back from webmockr and made it a property of the client itself
crul is the R6-based HTTP client underneath much of rOpenSci's package stack, covering synchronous requests, three flavours of async, pagination and retries. Its 1.6.0 release in July 2025 changed where test mocking lives: each client — HttpClient, Async, AsyncVaried — now takes a mocking parameter at initialisation or per method, and the standalone mock() function is deprecated. Mocking used to be something webmockr switched on from outside; it is now a setting on the client.
What is tabnet?
A tabular deep-learning model in R that keeps widening what counts as a tabular task.
tabnet ports the TabNet attentive tabular architecture to R on torch, wired into tidymodels through parsnip so it slots into workflows, tuning, and case weights like any other engine. The model surface has grown well past plain supervised fitting: unsupervised pretraining, missing values in predictors, multi-outcome fitting, hierarchical multi-label classification, and built-in explainability via tabnet_explain(). The 0.9.x line has been consolidating rather than adding, with 0.9.0 finally making hierarchical classification work correctly by accounting for the ancestor matrix.
crul vs tabnet: editorial side-by-side
crul
0.0ANALYTICS
crul took mocking back from webmockr and made it a property of the client itself
◆ Current state
crul is the R6-based HTTP client underneath much of rOpenSci's package stack, covering synchronous requests, three flavours of async, pagination and retries. Its 1.6.0 release in July 2025 changed where test mocking lives: each client — HttpClient, Async, AsyncVaried — now takes a mocking parameter at initialisation or per method, and the standalone mock() function is deprecated. Mocking used to be something webmockr switched on from outside; it is now a setting on the client.
◆ Where it's heading
The async surface has been the growth area for years — retries reached Async, AsyncVaried, AsyncQueue and HttpRequest in 1.4, AsyncQueue gained the response accessors in 1.2, and 1.5.0 wired async requests up to webmockr. The 1.6.0 change reverses that direction of dependency, and it landed within a minute of webmockr's own release severing its tie to vcr. Read together, the rOpenSci HTTP stack is being deliberately untangled so each package can be used without the others.
◆ Prediction
With mock() deprecated rather than removed, the next major release is the likely point of deletion. Expect the remaining work to follow the same decoupling theme rather than adding request features.
tabnet
2.5ANALYTICS
A tabular deep-learning model in R that keeps widening what counts as a tabular task.
◆ Current state
tabnet ports the TabNet attentive tabular architecture to R on torch, wired into tidymodels through parsnip so it slots into workflows, tuning, and case weights like any other engine. The model surface has grown well past plain supervised fitting: unsupervised pretraining, missing values in predictors, multi-outcome fitting, hierarchical multi-label classification, and built-in explainability via tabnet_explain(). The 0.9.x line has been consolidating rather than adding, with 0.9.0 finally making hierarchical classification work correctly by accounting for the ancestor matrix.
◆ Where it's heading
Two threads run through the release history. The first is task surface — each minor version tends to admit a class of problem the model previously could not express, from missing data to hierarchy to imbalanced binary outcomes. The second is torch-level performance and correctness, visible in the torch_ignite_adam default that cut pretraining time roughly 30% and the fix for optimizers frozen after checkpointing on cuda and mps. Tidymodels integration is treated as a first-class obligation, with parsnip breaking changes tracked release by release.
◆ Prediction
The hierarchical path is the least finished: 0.5.0 introduced it and 0.9.0 only just made it effective, so the next releases most likely extend evaluation and explainability to hierarchical fits rather than adding another task type.
Alternatives to crul and tabnet
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 crul or tabnet.
Tplyr
Tplyr made clinical summary tables explain where every number came from.
rlistings
Clinical listings that keep inheriting their hardest problem — pagination — from the layer below.
hoardr
A cache-directory helper that has shipped nothing but CRAN-triggered patches for seven years.
gigs
gigs redesigned its whole conversion API for rOpenSci, then spent three releases getting the docs to build.
riem
A weather-data client that keeps rewriting its HTTP layer while slowly tightening its API.
Recent activity from crul and tabnet
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
- 21d agotabnetvip dependency moves to r-universe
- 2mo agotabnetHierarchical classification made effective, augment() added
- 6mo agotabnetentmax15 and sparsemax15 masks, AUM loss for imbalanced data
- 1y agocrulMocking becomes a client parameter, independent of webmockr ⚡
- 1y agotabnetBugfix release for R 4.5 and dials tuning
- 2y agocrulAsync requests become mockable through webmockr
- 2y agotabnetCase weights and warm-start parameters via parsnip
- 2y agocrulDocumentation fixes and test helper tweak
- 2y agotabnetHierarchical multi-label classification via data.tree ⚡
- 3y agocrulHTTP retries reach the async classes
- 3y agocrulClearer error for mismatched urls and disk lengths
- 4y agocrulAsyncQueue gains response accessors; results print as a summary
Frequently asked questions
What is the difference between crul and tabnet?
They serve adjacent needs but don't currently overlap on shipped themes. tabnet 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.
Is crul better than tabnet?
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. tabnet 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.
What are the best alternatives to crul?
Top crul alternatives in Analytics are ranked by recent ship velocity. Browse the "crul alternatives" section above for the current picks, or visit /alternatives/crul for the full list with editorial commentary on each.
What are the best alternatives to tabnet?
Top tabnet alternatives in Analytics are ranked by recent ship velocity. Browse the "tabnet alternatives" section above for the current picks, or visit /alternatives/tabnet for the full list with editorial commentary on each.