E-means clustering for R: a fast mini-batch alternative to k-means - coatless-rpkg/emeans

A new release of prrd arrived at CRAN today, and has already been built for r2u and on r-universe. The prrd package facilitates the _parallel running [of] reverse dependency [checks]_ when preparing R packages. It is used extensively for releases I make of Rcpp, RcppArmadillo, RcppEigen, BH, and others. The key idea of prrd is simple, and described in some more detail on its webpage and its GitHub repo. Reverse dependency checks are an important part of package development that is easily done in a (serial) loop. But these checks are also generally _embarassingly parallel_ as there is no or little interdependency between them (besides maybe shared build depedencies). See the (somewhat dated) screenshot (running six parallel workers, arranged in a split byobu session). This release updates internals to allow for greater concurrency. SQLite is used as the backend, both for the job queue (managed via the liteq package) as well as for storing actual reverse-dependency test results. It is well-known that SQLite copes quite well with multiple _readers_. Multiple concurrent _writers_ are more challenging. We tended to get by via `BEGIN EXCLUSIVE` pragma use, but still get occassional issue even at a somewhat moderate concurrency levels. After some back and forth and trying a few approaches, more recently we established one first solution relying on very explicity (OS-level, filesystem-based) locking via package filelock. That worked. We mentioned this in an old liteq issue that had circled around the same question with a number of posts over the years, but no resolution. This comment seemingly awoke Gábor, and he nailed this years-open issue with a fine pull request deploying `BEGIN IMMEDIATE` and well as a loop around `tryCatch` to not immediately fail. I borrowed/ported that approach to prrd and we now have a second solution (and no longer need filelock). On a somewhat better server with enough cores and memory, I am now running eighteen concurrent workers—watched via three multi-pane byobu windows. Works like a charm without issues. At last! A big thanks to Gábor for finding another solution, and especially that only relies on SQLite. The release is summarised in the NEWS entry: > #### Changes in prrd version 0.1.0 (2026-11-07) > > * Concurrent operations are now supported via updated sqlite semantics borrowed with appreciation from liteq > > * This follows an interim solution relying on filelock use; however this later package is not needed so no increase in dependencies > > Courtesy of my CRANberries, there is also a diffstat report for this release. This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. If you like this or other open-source work I do, you can sponsor me at GitHub.

Sparse Bayesian factor models for multivariate data, built around a flexible Generalized Lower Triangular (GLT) process prior on the loading matrix by Canale and Frühwirth-Schnatter (2026) &lt;<a href="https://doi.org/10.48550%2FarXiv.2610.09762" target="_top">doi:10.48550/arXiv.2610.09762</a>&gt;. The number of active factors, their sparsity pattern, and the loadings themselves are learned jointly via a Markov chain Monte Carlo sampler implemented in C++ with 'Rcpp' and 'RcppArmadillo'.

Sparse Bayesian factor models for multivariate data, built around a flexible Generalized Lower Triangular (GLT) process prior on the loading matrix by Canale and Frühwirth-Schnatter (2026) &lt;<a href="https://doi.org/10.48550%2FarXiv.2610.09762" target="_top">doi:10.48550/arXiv.2610.09762</a>&gt;. The number of active factors, their sparsity pattern, and the loadings themselves are learned jointly via a Markov chain Monte Carlo sampler implemented in C++ with 'Rcpp' and 'RcppArmadillo'.

Armadillo is a powerful and expressive C++ template library for linear algebra and scientific computing. It aims towards a good balance between speed and ease of use, has a syntax deliberately close to Matlab, and is useful for algorithm development directly in C++, or quick conversion of research code into production environments. RcppArmadillo integrates this library with the R environment and language–and is widely used by (currently) 1370 other packages on CRAN, downloaded 49.4 million times (per the partial logs from the cloud mirrors of CRAN), and the CSDA paper (preprint / vignette) by Conrad and myself has been cited 735 times according to Google Scholar. This versions updates to the 15.6.2 upstream Armadillo release made yesterday. It, along with 15.6.1 release we incorporated earlier, addresses only minor issues and should not present any user-facing changes. As usual, we did run full reverse-depdency checks against the over 1300 ackages, and did not find anything. As usual, we aim to accommodate the desired CRAN cadences for uploads and try not to exceed six in six months—so we did not ship 15.6.1-1 there. The package has also already been updated for Debian, built for r2u and r-universe, and will build shortly at CRAN for the different binary releases. All changes since the last CRAN release follow. > #### Changes in RcppArmadillo version 15.6.2-1 (2026-10-16) > > * Upgraded to Armadillo release 15.6.2 (Medium Roast Cortado) > > * Faster inplace addition of `eye()` expressions > > > #### Changes in RcppArmadillo version 15.6.1-1 [github-only] (2026-10-02) > > * Upgraded to Armadillo release 15.6.1 (Medium Roast Cortado) > > * Fix for failures in `pinv()` when using the (non-default) "std" method > Courtesy of my CRANberries, there is a diffstat report relative to previous release. More detailed information is on the RcppArmadillo page. Questions, comments etc should go to the rcpp-devel mailing list off the Rcpp R-Forge page. This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. If you like this or other open-source work I do, you can sponsor me at GitHub.

Provides high-performance implementations of Small Area Estimation (SAE) methods leveraging C++ ('Rcpp', 'RcppArmadillo') and 'OpenMP' multi-threading. Supports standard area-level Fay-Herriot models (Fay and Herriot, 1979 &lt;<a href="https://doi.org/10.1080%2F01621459.1979.10482505" target="_top">doi:10.1080/01621459.1979.10482505</a>&gt;), Spatial Fay-Herriot models (Pratesi and Salvati, 2008 &lt;<a href="https://doi.org/10.1002%2Fenv.861" target="_top">doi:10.1002/env.861</a>&gt;), Spatio-Temporal Fay-Herriot models (Marhuenda et al., 2013 &lt;<a href="https://doi.org/10.1016%2Fj.csda.2013.01.016" target="_top">doi:10.1016/j.csda.2013.01.016</a>&gt;), and unit-level Battese-Harter-Fuller models (Battese et al., 1988 &lt;<a href="https://doi.org/10.1080%2F01621459.1988.10478561" target="_top">doi:10.1080/01621459.1988.10478561</a>&gt;). Features include empirical best linear unbiased prediction (EBLUP), analytical and bootstrap Mean Squared Error (MSE) estimation, automatic handling of unsampled domains, and modern S3 diagnostic methods.

Provides high-performance implementations of Small Area Estimation (SAE) methods leveraging C++ ('Rcpp', 'RcppArmadillo') and 'OpenMP' multi-threading. Supports standard area-level Fay-Herriot models (Fay and Herriot, 1979 &lt;<a href="https://doi.org/10.1080%2F01621459.1979.10482505" target="_top">doi:10.1080/01621459.1979.10482505</a>&gt;), Spatial Fay-Herriot models (Pratesi and Salvati, 2008 &lt;<a href="https://doi.org/10.1002%2Fenv.861" target="_top">doi:10.1002/env.861</a>&gt;), Spatio-Temporal Fay-Herriot models (Marhuenda et al., 2013 &lt;<a href="https://doi.org/10.1016%2Fj.csda.2013.01.016" target="_top">doi:10.1016/j.csda.2013.01.016</a>&gt;), and unit-level Battese-Harter-Fuller models (Battese et al., 1988 &lt;<a href="https://doi.org/10.1080%2F01621459.1988.10478561" target="_top">doi:10.1080/01621459.1988.10478561</a>&gt;). Features include empirical best linear unbiased prediction (EBLUP), analytical and bootstrap Mean Squared Error (MSE) estimation, automatic handling of unsampled domains, and modern S3 diagnostic methods.

Implements the Bayesian Exchangeable Non-Exchangeable (EXNEX) framework for right-censored log-normal survival data in basket trials. Based on 'Rcpp' and 'RcppArmadillo', the package provides a fast Gibbs sampler supporting EXNEX, complete pooling, and no pooling models to facilitate methodological comparisons and simulation studies.

Implements the Bayesian Exchangeable Non-Exchangeable (EXNEX) framework for right-censored log-normal survival data in basket trials. Based on 'Rcpp' and 'RcppArmadillo', the package provides a fast Gibbs sampler supporting EXNEX, complete pooling, and no pooling models to facilitate methodological comparisons and simulation studies.

Fast, documented implementations of robust estimation, testing, dimension reduction, classification, and clustering methods for high-dimensional elliptically symmetric data. Computational kernels use 'Rcpp' and 'RcppArmadillo'. The package follows methods reviewed in Feng (2026), "High-Dimensional Data Analysis for Elliptically Symmetric Distributions" &lt;<a href="https://github.com/flnankai/HDElliptical/releases" target="_top">https://github.com/flnankai/HDElliptical/releases</a>&gt;.

Fast, documented implementations of robust estimation, testing, dimension reduction, classification, and clustering methods for high-dimensional elliptically symmetric data. Computational kernels use 'Rcpp' and 'RcppArmadillo'. The package follows methods reviewed in Feng (2026), "High-Dimensional Data Analysis for Elliptically Symmetric Distributions" &lt;<a href="https://github.com/flnankai/HDElliptical/releases" target="_top">https://github.com/flnankai/HDElliptical/releases</a>&gt;.

Armadillo is a powerful and expressive C++ template library for linear algebra and scientific computing. It aims towards a good balance between speed and ease of use, has a syntax deliberately close to Matlab, and is useful for algorithm development directly in C++, or quick conversion of research code into production environments. RcppArmadillo integrates this library with the R environment and language–and is widely used by (currently) 1331 other packages on CRAN, downloaded 48.5 million times (per the partial logs from the cloud mirrors of CRAN), and the CSDA paper (preprint / vignette) by Conrad and myself has been cited 727 times according to Google Scholar. This versions updates to the 15.6.0 upstream Armadillo release made yesterday. It extends solver options for poorly conditioned systems, and brings some updates and extension to the `cube` data type. For this release, we once again ran the usual complete reverse-dependency check which came back spotless, and did CRAN so no email exchange needed despite nearly 1300 reverse dependencies (but it ended up taking more than a single business day). Still, automation can be helpful when used with a well-maintained software stack. The package has also already been updated for Debian, built for r2u and r-universe, and will build shortly at CRAN for the different binary releases. All changes since the last CRAN release follow. > #### Changes in RcppArmadillo version 15.6.0-1 (2026-09-07) > > * Upgraded to Armadillo release 15.6.0 (Medium Roast Cortado) > > * Expanded `solve()` with `solve_opts::scale_thresh` option to widen detection of poorly conditioned systems > > * Expanded `trans()` and `.t()` to handle cubes > > * Added `permute()` to rearrange dimensions of cubes (generalised transpose) > > * Added `cubemul()` for batched matrix multiplication of cube slices > > Courtesy of my CRANberries, there is a diffstat report relative to previous release. More detailed information is on the RcppArmadillo page. Questions, comments etc should go to the rcpp-devel mailing list off the Rcpp R-Forge page. This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. If you like this or other open-source work I do, you can sponsor me at GitHub.

A new minor release of prrd arrived at CRAN this morning: the a first release in two and a half years. prrd facilitates the _parallel running [of] reverse dependency [checks]_ when preparing R packages. It is used extensively for releases I make of Rcpp, RcppArmadillo, RcppEigen, BH, and others. The key idea of prrd is simple, and described in some more detail on its webpage and its GitHub repo. Reverse dependency checks are an important part of package development that is easily done in a (serial) loop. But these checks are also generally _embarassingly parallel_ as there is no or little interdependency between them (besides maybe shared build depedencies). See the (dated) screenshot (running six parallel workers, arranged in a split byobu session). This release updates continuous intgegration files, switches to Authors@R, and robustifies one SQLite aspect. The release is summarised in the NEWS entry: > #### Changes in prrd version 0.0.7 (2026-08-27) > > * Updates to DESCRIPTION have been made as CRAN requirements change > > * The continuous integration setup was updated several times > > * The database connection now uses sqliteSetBusyHandler > > Courtesy of my CRANberries, there is also a diffstat report for this release. This post by Dirk Eddelbuettel originated on his Thinking inside the box blog. If you like this or other open-source work I do, you can sponsor me at GitHub.

Estimates regression coefficients in proportional hazards models for interval-censored multistate data. Individuals may be observed at irregular times and their states may be partially observed. Allowable transitions and transition-specific covariate effects can be specified. The numerical estimation is implemented in 'C++' using 'RcppArmadillo'. The method implemented in this package is described in You, Liu, and Krischer (2024) &lt;<a href="https://doi.org/10.1002%2Fsim.10079" target="_top">doi:10.1002/sim.10079</a>&gt;.

Provides density, distribution function, quantile function, and random generation for the Tweedie distribution under the compound Poisson-Gamma parameterisation with power parameter in (1, 2). The density is evaluated using the series expansion of Dunn and Smyth (2005) &lt;<a href="https://doi.org/10.1007%2Fs11222-005-4070-y" target="_top">doi:10.1007/s11222-005-4070-y</a>&gt;, implemented in C++ via 'Rcpp' and 'RcppArmadillo' for performance. A constructor compatible with the distributional package is also provided for use in tidy modelling workflows.

Provides a comprehensive toolset for estimating Latent Profile Analysis (LPA) models that are robust to multivariate outliers and missing data. By integrating a high-performance 'C++' engine via 'RcppArmadillo' with a Full Information Maximum Likelihood (FIML) approach and Huber weighting, it reliably extracts latent profiles even in complex datasets. It supports multiple geometric variance-covariance models, along with functions for bootstrapped likelihood ratio tests and plotting. For methodological details on the Bootstrapped Likelihood Ratio Test, see Nylund et al. (2007) &lt;<a href="https://doi.org/10.1080%2F10705510701575396" target="_top">doi:10.1080/10705510701575396</a>&gt;. For robust clustering methods, see Garcia-Escudero et al. (2010) &lt;<a href="https://doi.org/10.1007%2Fs11634-010-0064-5" target="_top">doi:10.1007/s11634-010-0064-5</a>&gt;.