STACAS
Single-cell batch correction that learned to use cell labels, then spent three releases chasing Seurat.
Comparison · Analytics
flashlight vs GeneNMF
A side-by-side editorial comparison of flashlight and GeneNMF — release velocity, themes, recent moves, and the top alternatives to consider.
Shared themes:r-package
flashlight vs GeneNMF: at a glance
What is flashlight?
flashlight hit 1.0 by giving away its SHAP feature and making most of its API internal.
flashlight computes model-agnostic interpretability output — variable importance, partial dependence and effect profiles, breakdown plots — for fitted models in R. Version 1.0.0 in October 2025 executed a contraction announced two years earlier: add_shap() is deprecated in favour of the separate kernelshap and fastshap packages, type = "shap" is gone from every light_* function that accepted it, and eight previously exported helpers became internal. The release also restored compatibility with ggplot2 v4.
What is GeneNMF?
GeneNMF rebuilt how it derives meta-programs, changing every result it had produced.
GeneNMF applies non-negative matrix factorization to single-cell expression data to find gene programs, then consolidates programs recurring across samples into meta-programs. Version 0.6.0 replaced the consolidation method: instead of reducing each program to a gene set and taking a consensus, it retains full gene weight vectors and compares them by cosine similarity. Later releases have built reporting and control around that core — a metaprogram composition matrix showing which samples contributed, custom signature databases for enrichment testing, and the ability to drop meta-programs from results.
flashlight vs GeneNMF: editorial side-by-side
flashlight
0.0ANALYTICS
flashlight hit 1.0 by giving away its SHAP feature and making most of its API internal.
◆ Current state
flashlight computes model-agnostic interpretability output — variable importance, partial dependence and effect profiles, breakdown plots — for fitted models in R. Version 1.0.0 in October 2025 executed a contraction announced two years earlier: add_shap() is deprecated in favour of the separate kernelshap and fastshap packages, type = "shap" is gone from every light_* function that accepted it, and eight previously exported helpers became internal. The release also restored compatibility with ggplot2 v4.
◆ Where it's heading
The package is narrowing rather than growing, and doing it on a published schedule — 0.9.0 listed the breaking changes, 1.0.0 applied them essentially unchanged. Handing SHAP computation to dedicated packages leaves flashlight as an effects-and-profiles visualization layer rather than an all-purpose interpretability toolkit. Removing the ability to rename result columns via options() points the same way: fewer configuration surfaces, a smaller contract to maintain.
◆ Prediction
With the deprecation list from 0.9.0 now fully applied, the next releases most likely remove the functions currently deprecated rather than adding capability, and continue tracking ggplot2. The entries show no new analysis method in progress.
GeneNMF
0.0ANALYTICS
GeneNMF rebuilt how it derives meta-programs, changing every result it had produced.
◆ Current state
GeneNMF applies non-negative matrix factorization to single-cell expression data to find gene programs, then consolidates programs recurring across samples into meta-programs. Version 0.6.0 replaced the consolidation method: instead of reducing each program to a gene set and taking a consensus, it retains full gene weight vectors and compares them by cosine similarity. Later releases have built reporting and control around that core — a metaprogram composition matrix showing which samples contributed, custom signature databases for enrichment testing, and the ability to drop meta-programs from results.
◆ Where it's heading
The package is moving from producing meta-programs to letting users interrogate and constrain how they were formed. Composition matrices, the drop function and downsampled similarity heatmaps all serve inspection rather than derivation. The parameters added alongside the 0.6.0 rewrite — specificity weighting, cumulative weight thresholds, confidence defined as the fraction of programs containing a gene — turn what were fixed internal choices into stated, tunable ones.
◆ Prediction
Recent releases have been fixes and compatibility work rather than method changes, so the core approach appears settled. The dependency on an RcppML version not on CRAN is the loose end most likely to force the next release.
Alternatives to flashlight and GeneNMF
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 flashlight or GeneNMF.
ggtrace
A debugger for ggplot2's internals, hardening its grip as the internals it traces keep moving.
kde1d
A univariate density estimator that added zero-inflated data and reopened its C++ API to do it.
svines
Stationary vine copulas for time series, released in lockstep with the rest of Nagler's vine stack.
gghighlight
A single-purpose ggplot2 extension that has spent six years tracking ggplot2 instead of growing.
rtrek
A Star Trek data package that became a Memory Alpha web client and has been patching scrapers ever since.
See all flashlight alternatives → · See all GeneNMF alternatives →
Recent activity from flashlight and GeneNMF
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
- 10mo agoflashlightSHAP handed off to kernelshap as the public API shrinks ⚡
- 11mo agoGeneNMFSingle-sample runs fixed; gene weight definition refined
- 1y agoGeneNMFMetaprogram composition exposed and custom signature DBs supported
- 1y agoGeneNMFSimilarity heatmap downsampling and meta-program removal
- 2y agoGeneNMFMeta-programs rebuilt on gene weight vectors and cosine similarity ⚡
- 2y agoGeneNMFFirst stable release published to CRAN
- 3y agoflashlightBreaking changes for 1.0.0 announced two years ahead
- 4y agoflashlightMaintenance release covering package generation
- 5y agoflashlightggpubr dependency removed
Frequently asked questions
What is the difference between flashlight and GeneNMF?
Both compete on the same themes — r-package — within Analytics. flashlight and GeneNMF 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 flashlight better than GeneNMF?
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. flashlight and GeneNMF 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 flashlight?
Top flashlight alternatives in Analytics are ranked by recent ship velocity. Browse the "flashlight alternatives" section above for the current picks, or visit /alternatives/flashlight for the full list with editorial commentary on each.
What are the best alternatives to GeneNMF?
Top GeneNMF alternatives in Analytics are ranked by recent ship velocity. Browse the "GeneNMF alternatives" section above for the current picks, or visit /alternatives/genenmf for the full list with editorial commentary on each.