Author | Lightnews

Reposted by SMT

Charlie Gao @shikokuchuo.net · 1d

I put out a patch release of mirai today. Version 2.5.2 really improves the OpenTelemetry integration so you can more easily see into your async workflows. Other key ecosystem packages will roll out with this enabled - next up: Shiny!

mirai.r-lib.org

#Rstats

Minimalist Async Evaluation Framework for R

Designed for simplicity, a mirai evaluates an R expression asynchronously, locally or distributed over the network. Built on nanonext and NNG for modern networking and concurrency, scales efficiently ...

mirai.r-lib.org

3 11

SMT @pp0196.bsky.social · 9h

rixpress is a R based proposition for so called multilingual pipelines (in reality most of the time why not just wrap what is needed via reticulate or similar for other languages) brodrigues.co/posts/2025-1...

Orchestrating Polyglot, Reproducible Data Science with Nix and {rixpress} – Econometrics and Free Software

brodrigues.co

2 1 2

Reposted by SMT

Josiah @josiahparry.com · 2d

Any examples of fitting a graph convolution network in {torch} or {luz} folks are aware of? #rstats

2 4 5

SMT @pp0196.bsky.social · 1d

Sometimes one need more control than just letting lexical scoping rule, one reason would be to avoid size effects in caller environment (using `source` with an environment parameter). Other than the more general questions around environment and Namespaces henrikbengtsson.github.io/environments...

How objects are found by package functions

henrikbengtsson.github.io

2

Reposted by SMT

Nick Eagles @nick-eagles.bsky.social · 2d

We recently checked out tidySingleCellExperiment at the R Stats Club @libdrstats.bsky.social, an R package part of a broader effort to bring tidy coding to genomics. Pretty useful!

youtu.be/ob_SQY4-YjQ?...

#RStats

[2025-07-11] Tidy manipulation of SingleCellExperiment objects

YouTube video by Leonardo Collado Torres

youtu.be

3 5

SMT @pp0196.bsky.social · 3d

As it happens, there is a R level equivalent of the C API finalization registration mechanism stat.ethz.ch/R-manual/R-d...
Never saw it in the wild, but it looks like some CRAN packages use it (expectedly lot of these packages manage external processes)
github.com/search?q=org...
#Rstats

R: Finalization of Objects

stat.ethz.ch

1 2

SMT @pp0196.bsky.social · 3d

#ClinicalGenomics #AfricanDiasporas
www.medrxiv.org/content/10.1...

Genome sequencing of 35,024 predominantly African ancestry persons addresses gaps in genomics and healthcare

Genetic variation is crucial in human development, disease susceptibility, and drug response. Despite populations of African descent having the highest degree of genetic variation, genetic research ha...

www.medrxiv.org

SMT @pp0196.bsky.social · 3d

If you add to this the terrible docs you can understand the nihilist and p(y)edantic (pun intended) type annotations drive

SMT @pp0196.bsky.social · 3d

Use any python library and you quickly find yourself buried under layers upon layers of encapsulated OOP inheritance crap, the tracebacks are order of magnitude worse than tidyverse NSE tracebacks before the rlang package (this is when the module author allows you to see traces)

1

Reposted by SMT

Peter Dalgaard @pdalgd.bsky.social · 6d

R version 4.5.2 "[Not] Part in a Rumble" (source version) has been released. (You can find it in cran.r-project.org/src/base/R-4/, or wait for CRAN to be updated.)

11 28

Reposted by SMT

Daniel Falbel @dfalbel.bsky.social · 7d

luz v0.5.1 is now on #rstats CRAN. Just a small bug fix related to forwarding `predict` parameters to the model. I also added a new Variational Autoencoder example to our examples gallery: mlverse.github.io/luz/articles...

Examples

mlverse.github.io

1 2 11

Reposted by SMT

Hacker News Companion @hncompanion.com · Oct 7

A significant OpenZL theme is genomic data compression. Users are eager for benchmarks against CRAM, FASTA, FASTQ, SAM, and VCF. It shows promise for tackling large datasets, especially challenging nanopore data. #Genomics 2/5

1 2 4

Reposted by SMT

Dirk Eddelbuettel @eddelbuettel.com · 10d

R^4 post #054: Faster r-ci Contimuous Integration via r2u Container
Accelerating Continous Integration at GitHub Actions
dirk.eddelbuettel.com/blog/2025/10...
#rstats

Screenshot one of two of blog post at https://dirk.eddelbuettel.com/blog/2025/10/27#054_faster_r-ci describing a container-based improvement for Continuous Integration at GitHub Actions

Screenshot two of two of blog post at https://dirk.eddelbuettel.com/blog/2025/10/27#054_faster_r-ci describing a container-based improvement for Continuous Integration at GitHub Actions

4 5

SMT @pp0196.bsky.social · 11d

#mojo vs #C in the #RStats convolve classic benchamrk. Looks like mojo goes brrr
github.com/sounkou-bioi...

2

SMT @pp0196.bsky.social · 12d

i ve try it out using uv , it appears that the free threaded python is a significant ABI change github.com/rstudio/reti...

find libpython for uv threaded python venvs by sounkou-bioinfo · Pull Request #1857 · rstudio/reticulate

github.com

SMT @pp0196.bsky.social · 12d

is there some container setup to test reticulate against specific CPython builds (i can see that the libpython is linked against in src) ? Disclosure i am not a (direct) reticulate even though i should become one pretty soon :D

1

SMT @pp0196.bsky.social · 12d

My template for any #RStats package that used LLM tools and hopefully somebody should trigger some lawsuit to enforce this interpretation in any project out there that uses LLM tools

LLM Usage Disclosure
Code and documentation in this project have been generated with the assistance of the github Copilot LLM tools. While we have reviewed and edited the generated content, we acknowledge that LLM tools were used in the creation process and accordingly (since these models are trained on GPL code and other commons + proprietary software license is fake anyway) the code is released under GPL-3. So if you use this code in any way, you must comply with the GPL-3 license.

1 2

SMT @pp0196.bsky.social · 12d

how does the python so called free threading interact with R ?

1

Reposted by SMT

Stephen Turner @stephenturner.us · 15d

BioProBench: Comprehensive Dataset and Benchmark in Biological Protocol Understanding and Reasoning arxiv.org/abs/2505.07889 🧬🖥️🧪
Code: github.com/YuyangSunshi...
Data: huggingface.co/datasets/Bio...

2 2

SMT @pp0196.bsky.social · 14d

je ne sais pas pourquoi, mais j ai l impression que c'est fait avec litedown

2

SMT @pp0196.bsky.social · 14d

Il manque plus que des diagrammes pour les personnes qui sont plus dans le visuel (comme moi)

1

Reposted by SMT

Maëlle Salmon @masalmon.eu · 14d

Did "Writing R Extensions" go to a beauty salon?!

cran.r-project.org/doc/manuals/...

#RStats

Writing R Extensions

cran.r-project.org

3 5 22

SMT @pp0196.bsky.social · 16d

Where are the (vendor agnostic) #Rstats case studies, packages, books, CRAN task views on Data engineering and ETL ? Maybe I am uninformed, but if I am not this gap is not sustainable : cannot expect people to go along with the instable equilibrium that is programming Diglossia (a.k.a "polyglot")

1

SMT @pp0196.bsky.social · 16d

Loom at the results of this search for repos, this is really problematic

github.com/search?q=org...

Build software better, together

GitHub is where people build software. More than 150 million people use GitHub to discover, fork, and contribute to over 420 million projects.

github.com

SMT @pp0196.bsky.social · 16d

We should really not leave data engineering to the snake side. The snake creep is both from the ETL and Data warehouse side, sandwiching the poor Rstat data analyst, not sustainable long term. We have similar issue in my field (Bioinformatics) where R is getting reduced to plotting for papers lol

1 2