The R Data Scientist 25-11-2025
R & AI, new packages & developments, geospatial projects
🌐 Community Roundups
Ready4R (2025-11-24): Giving Thanks For All the Videos (buttondown.com/ready4r). Weekly Ready for R newsletter covers Databot explores NHANES, PositConf talks on WebR/Pyodide, and Surveydown survey tooling
R Weekly 2025-W48 Multi-agents, linter, rbranding (rweekly.org). Multi-Agent orchestration in R, a new linter, and branding tools for R projects plus various packages and updates
Weekly Recap (Nov 21, 2025) (blog.stephenturner.us). Gemini 3, R updates, Antigravity, DuckDB encryption, R+AI talks, biology essays, and AI-driven coding advances across Python and R
2025-11-21 AI Newsletter (posit.co). Posit AI newsletter covers model releases, TOON tokenization, and ecosystem tools for R, Python, and LLMs with insights from Sara and Simon
Weekly Web Harvest for 2025-11-16 (bionicteaching.com). Weekly Web Harvest highlights data viz with R and ggplot2, plus quotes on AI use in law enforcement and cybersecurity insights
🤖 R & AI
Tool Calling UI in shinychat (shiny.posit.co). Shinychat adds rich tool-calling UI for R and Python, enabling visible tool calls and results in LLM-powered Shiny apps
kuzco | Computer Vision made easy (r-consortium.org). Kuzco uses LLMs for prompt-driven computer vision in R, returning JSON outputs and offering a Shiny UI and multi-model support
Introducing {geodl}: An R package for geospatial semantic segmentation (r-consortium.org). Geospatial semantic segmentation in R using geodl with torch, terra, and luz, featuring UNet variants and wall-to-wall spatial predictions
Privacy and AI Assistants (posit.co). Posit discusses Privacy and AI Assistants, with insights from Simon Couch and Sara, plus tools for R, Python, and LLM integration
🧰 Packages & Development
Setting up a local HPC cluster with SLURM for testing & learning (tomsing1.github.io). Local HPC cluster with SLURM using slurm-docker-cluster, Docker Compose, and R (clustermq) on macOS
Jarl: just another R linter (etiennebacher.com). Etienne Bacher introduces Jarl, a Rust-based R linter that parses code for patterns, offers automatic fixes, and supports CI workflows
{bidux} v0.3.3: Where Databases Meet Quick Decisions (jrwinget.com). Bidux v0.3.3 enables direct DBI-backed telemetry, quick suggestions, and Quarto-ready dashboards for R users
Créer des flowcharts cliniques sous R : exclusions, populations analysées et filtrages automatisés (delladata.fr). Flowcharts cliniques sous R with fc_filter, fc_split, et fc_draw pour documenter exclusions et populations analysées
(ICYMI) RPweave: Unified R + Python + LaTeX System using uv (thierrymoudiki.github.io). (ICYMI) RPweave: Unified R + Python + LaTeX System using uv
Two New tidymodels Packages (tidyverse.org). Two new tidymodels packages filtro and important introduce feature selection tools for tidymodels workflows in R
posit::conf(2025) Quarto talks (quarto.org). Videos of posit::conf(2025) talks showcase Quarto features, extensions, workflows, teaching, and publishing
🗺️ Geospatial Projects
Create custom GPS route maps in R (nrennie.rbind.io). Using R to create custom GPS route posters with gpx, sf, osmdata, ggplot2, and ggtext
Impact of Nightlife on Barcelona Districts (jmsallan.netlify.app). Explores nightlife distribution in Barcelona districts using 2019/2024 Open Data BCN data, with R (tidyverse, sf) visualizations and district population metrics
Day 24: The Mud Lakes of Nova Scotia (dewey.dunnington.ca). Using SedonaDB in R with sf, wk, geos, ggplot2 and patchwork to map 46 Mud Lake polygons in Nova Scotia
Understanding the Rise in Domestic Terrorism: Context Matters (stevenponce.netlify.app). Population-adjusted domestic terrorism trends reveal a 26% per-capita rise, using R, tidycensus, and ggplot2 visualization by Steven Ponce
EPSG:3035 (r.iresmi.net). Lambert azimuthal equal-area projection for Europe, with R and sf-based reprojection of the globe and ggplot2 visualization
Perseverance (r.iresmi.net). Perseverance voyage on Mars using R with sf, dplyr, ggplot2, glue; 30DayMapChallenge entry
📊 Statistical Methods
RWE Data Analysis Using Propensity Score Matching (PSM): Concept, Implementation, and Interpretation (mihiretukebede.com). RWE data analysis with Propensity Score Matching in R (MatchIt), balance diagnostics, and ATT estimation using Lalonde data
Neyman-Pearson Tests: An Episode in Anglo-Polish Collaboration: (3.2) (errorstatistics.com). N-P Tests outline steps for constructing tests, discuss historical context with Neyman and Pearson, and compare severity and probabilistic criteria
Classification error and relative rates (emilkirkegaard.com). Measuring classification error in race labeling with simulations in R; DeepFace, census name stats, and misclassification effects
Are R^2 values declining in every scholarly field? And if so, what could or should be done about that? (dynamicecology.wordpress.com). R^2 values trend across fields; ecology, accounting, and economics; preprint insights; causal inference limits
Which explanation is best? (larspsyll.wordpress.com). Explores inference to the best explanation, its justification, fallibility, and epistemic status within science and reasoning
📚 Academic Research
Single-Dataset Meta-Analysis For Many-Analysts And Multiverse Studies (arxiv:stat). Introduces single-dataset meta-analysis to avoid overconfident conclusions from many‑analyst multiverse studies by weighting dataset information once. Implementable using standard R meta packages, promoting reproducible workflows
Comparing Bayesian and Frequentist Inference in Biological Models: A Comparative Analysis of Accuracy, Uncertainty, and Identifiability (arxiv:q-bio). Compares Stan HMC Bayesian and frequentist bootstrap inference across biological ODE models, showing when each excels depending on data observability. Practical guidance for R analysts
ggskewboxplots: Enhanced Boxplots for Skewed Data in R (arxiv:stat). Introduces ggskewboxplots R package integrating skew-aware boxplot variants into ggplot2, reducing swamping and masking for skewed data. Valuable for R visualisation and robust EDA workflows
👋 Before you go...
I've got a big favor to ask - keeping Blaze running isn't expensive, but it does all add up, so I'm asking readers like you to help, if you can, by joining the Patreon page. Nothing flashy, just a way for folks who find value in these newsletters to chip in a little each month.
If you are getting value from blaze, checking this out would mean the absolute world. But if you can't contribute, no worries - the newsletters keep coming either way. Thanks for reading and being part of this nerdy corner of the internet. All the best for the coming week - Alastair.
Add a comment