metatools
SDTM supplemental-qualifier merging got sturdier, then the package went quiet for two years.
Comparison · Analytics
mirai vs xplainfi
A side-by-side editorial comparison of mirai and xplainfi — release velocity, themes, recent moves, and the top alternatives to consider.
mirai vs xplainfi: at a glance
What is mirai?
mirai removed its dispatcher process and added memory backpressure to the queue.
The async evaluation framework releases roughly monthly and moves fast at the architecture level. In 2.7.0 the dispatcher stopped being a separate process and became a thread, after its loop had already been rewritten in C inside nanonext one release earlier. The same release added an opt-in memory budget for queued task payloads and try_mirai(), which returns NULL immediately rather than blocking when that budget is exhausted.
What is xplainfi?
xplainfi treats feature importance as an estimate with error bars, not a number.
xplainfi implements feature importance methods for mlr3 — perturbation-based PFI, CFI and RFI, refit-based LOCO and WVIM, and SAGE. Its defining choice is that importance scores come with inference attached: several confidence-interval methods, including the Nadeau-Bengio correction and a distribution-free option added in 1.1.0. It declared itself released at 1.0.0 in January 2026.
mirai vs xplainfi: editorial side-by-side
mirai
2.5ANALYTICS
mirai removed its dispatcher process and added memory backpressure to the queue.
◆ Current state
The async evaluation framework releases roughly monthly and moves fast at the architecture level. In 2.7.0 the dispatcher stopped being a separate process and became a thread, after its loop had already been rewritten in C inside nanonext one release earlier. The same release added an opt-in memory budget for queued task payloads and try_mirai(), which returns NULL immediately rather than blocking when that budget is exhausted.
◆ Where it's heading
Two threads of work run together: cutting overhead out of the task path — thread-based dispatcher, in-process transport for synchronous daemons, lower per-element dispatch cost in mirai_map() — and making the framework safe to embed in an event loop, where blocking the host R thread is not acceptable. Deployment reach is growing too, with http_config() launching remote daemons over HTTP APIs and auto-configuring for Posit Workbench. Each release pins a minimum nanonext version, so the two packages advance as one unit.
◆ Prediction
With backpressure in place but opt-in, the open question these notes leave is whether a default memory budget arrives; continued overhead reduction and Shiny-facing non-blocking paths are the safer bet.
xplainfi
2.5ANALYTICS
xplainfi treats feature importance as an estimate with error bars, not a number.
◆ Current state
xplainfi implements feature importance methods for mlr3 — perturbation-based PFI, CFI and RFI, refit-based LOCO and WVIM, and SAGE. Its defining choice is that importance scores come with inference attached: several confidence-interval methods, including the Nadeau-Bengio correction and a distribution-free option added in 1.1.0. It declared itself released at 1.0.0 in January 2026.
◆ Where it's heading
Two lines of work run in parallel. The statistical side keeps adding inference options — variance corrections, conditional predictive impact, and the Lei et al. observation-wise loss-difference test — while the computational side attacks the cost of refit-based methods, most recently with a batch_size argument that parallelises refits and a default of one refit per resampling iteration. Support for pre-trained learners in 1.1.0 removes the refit requirement entirely in some workflows.
◆ Prediction
The stated reasoning that budget is better spent on resampling iterations than repeated refits suggests n_repeats may be removed from WVIM and LOCO outright, as the release notes hint.
Alternatives to mirai and xplainfi
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 mirai or xplainfi.
marquee
marquee is filling in the typographic details — outlines, border types, real font metrics for underlines.
logrx
A clinical-script logger that stopped shipping after its 0.2 line, changelogs made of merged PRs.
lobstr
R's object inspector is losing its view of the internals as CRAN closes off the private C API.
GSODR
A weather-station data client that broke one return type to hand back distances instead of bare IDs.
giscoR
giscoR's 1.0 moved its dataset index into the cache, so new Eurostat releases arrive without a package update.
See all mirai alternatives → · See all xplainfi alternatives →
Recent activity from mirai and xplainfi
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
- 20d agoxplainfiRefits parallelise; repeated refits deprioritised in favour of resampling
- 25d agomiraiMap collection and daemon lifecycle fixes
- 2mo agomiraiAgent skill ships in-package; HTTP headers take over auth
- 3mo agomiraiDispatcher becomes a thread, and the queue gains a memory budget ⚡
- 5mo agomiraiParallel RNG seeding leaves experimental status
- 5mo agoxplainfiPre-trained learners supported; distribution-free inference added
- 6mo agomiraiRemote daemons over HTTP, and a C dispatcher loop
- 6mo agoxplainfiVersion bumped to mark the package as released
- 8mo agomiraiTelemetry span timing and daemon-switch fix
- 9mo agoxplainfiConfidence intervals arrive for feature importance scores
Frequently asked questions
What is the difference between mirai and xplainfi?
They serve adjacent needs but don't currently overlap on shipped themes. mirai and xplainfi are shipping at a similar cadence (velocity 2.5 vs 2.5, 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 mirai better than xplainfi?
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. mirai and xplainfi are shipping at a similar cadence (velocity 2.5 vs 2.5, 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 mirai?
Top mirai alternatives in Analytics are ranked by recent ship velocity. Browse the "mirai alternatives" section above for the current picks, or visit /alternatives/mirai for the full list with editorial commentary on each.
What are the best alternatives to xplainfi?
Top xplainfi alternatives in Analytics are ranked by recent ship velocity. Browse the "xplainfi alternatives" section above for the current picks, or visit /alternatives/xplainfi for the full list with editorial commentary on each.