Are you doing #GeoSpatial #Statistics in #RStats ?

A common operation is to get the adjacency matrix, or neighbour list of the regions using {spdep} and {sf}.

But: often there are islands or unconnected regions. This is a problem for the statistical models! And it's really annoying to fix when you have many regions.

Presenting {sfneighbourconnector}: an #rstudio add-in to manually add connections to the neighbour list through a point-and-click interface. https://github.com/vankesteren/sfneighbourconnector

Quick #RStats #RStudio #Quarto question: With the latest version of RStudio, is it still necessary to download quarto separately or is it included?

Disclaimer: I read the websites of both and it wasn't clear to me, possibly due to a lack of deductive skills. I could find out myself by uninstalling and reainstalling, but I'm too lazy to mess up my setup.

I write a lot of multi-script analysis flows and I number the scripts like the lines in BASIC scripts I wrote for the C64.

010_phenotype_definition.R
020_regression_analysis.R
030_result_reformatting.R

That way I can stick additional scripts in-between if I realize I need them.

From Excel to R

Transition from spreadsheets to scripting with this 2-day intro to #R & #RStudio for #beginners.

31 Mar–1 Apr
Panum, University of Copenhagen
Register: https://heads.ku.dk/calendar/2025/courses/from-excel-to-r-spring-2025/

Wrote this whole #emacs config because of the increasing presence of pay-to-use workbenches in #research. I find it more complete and flexible than #jupyter or #rstudio. Give it a go!

GitHub - lf-araujo/workbenchless: Single-file Emacs configuration for a powerful scientific Notebook system that works flawlessly over ssh.
https://github.com/lf-araujo/workbenchless

I love seeing all these #positronIDE posts with tips on settings etc. Really helpful stuff!

What I feel in missing, is a recording of how to efficiently work with the debugger. I'm very comfy in the #rstudio debugger, but can't seem to get the hang of Positron's.

The latest Rstudio 2024.12.0 adds a ProjectId field to .Rproj file. I don't see any comment on the release notes. Could someone at @Posit clarify how is this used and how it is calculated?
We got different ids on the same project by different team members.
It would help to decide if I add this for all the projects or not. #rstudio
Edit: found answer, see message below

@LinuxTorvaldIsGod I know that #GTK is very popular among developers on the #Linux platform, but isn't it possible to find alternatives for all of these applications [1]?

The #KDE / #QT ecosystem also has excellent applications (#Krita, #LibreOffice QT, #KDEnlive, #Shotcut, #QGIS, #DigiKam, #OBSStudio, #Friction, #PDF4QT, ...). There are many #Electron-based applications (#Chromium, #RStudio, #Slack, ...) and those applications using other toolkits (#Blender, ...)

[1] https://github.com/valpackett/awesome-gtk

If you use #RStats {targets} a lot, within #RStudio or #PositronIDE , you might want to try out the new(ish) `tar_assign` for workflows instead of `tar_plan` or making your own lists of targets.

Because `tar_assign` takes a wrapped code block (using `{ ... }` for multiline statements), the interpreter actually **sees** the variables properly, and can do tab completion on the variable names!

It's so nice.

https://docs.ropensci.org/tarchetypes/reference/tar_assign.html

Dear scientists, especially the R-community:

The NZ government wants to scrap all soc social sciences and humanities.

I probably don't have to explain why this is a totally bonkers milei-ish idea. (maybe just add the detail that Māori researchers will be overproportionally affected by these budget cuts, but this is probably anyway a feature of the plan)

I address the #RStats community explicitely, since R was developed in Aotearoa. And here in the Fediverse we are *many* R nerds!

So, @kjhealy or other kiwis, is there something the international academic community could do to express our thoughts and support your struggle?

https://mastodon.social/@kjhealy/113645641303523196

I'm starting a new research project and I want to use #python to force myself to learn new skills. What is the current best practice for creating reproducible reports in the python world? I dislike jupyter notbooks with a passion, so I'm leaning towards #quartopub. What would be the best IDE to manage a project running on a remote machine? Would #rstudio server work for that or should I use something else?

Made a new map of the Salish Sea region
Always fun experimenting with #RStudio and #rayshader

I love the topography of #britishcolumbia and #washington

We are thrilled to announce #RStudio IDE and Posit Workbench 2024.09.0!

• Bundled #Quarto v1.5
• Managed credential support for #AWS, #Azure, #Databricks, and #Snowflake
• The “Save as Image” dialog now
controls whether plots are scaled according to the current display’s DPI

Learn more in the blog post: https://posit.co/blog/rstudio-2024-09-0-whats-new/

I just realized that #Positron implemented the feature I requested in #RStudio...10 years ago! A fixed-size graphics device. Set your dimensions and plot interactively knowing the figure will look the same when you send to a file! Thanks @juliasilge @jmcphers and team!

Now, is there some documentation of the graphics device? Can we set the dimensions programmatically or in config? Should I mess with the .ps.graphics helpers for now?

Ctrl-Shift-L
Ctrl-Shift-D
Ctrl-Shift-B
Ctrl-Shift-T
Ctrl-Shift-E

The Vulcan nerve pinch of package development in #RStudio !

I wrote down my notes on getting RStudio Server to run on a SLURM instance via Singularity - it's as fiddly as it sounds

https://philippbayer.github.io/blerg/posts/2024_09_06_running_rstudio_server_on_singularity/