(1/2) Shiny Apps for demystifying statistical models and methods 🚀
This is a cool website that explains different statistical concepts with the use of interactive Shiny Apps. Ben Prytherch made this website from the Department of Statistics at Colorado State University.
The {geotargets} extends the {targets} package to work with geospatial data formats. This release provides support for {terra} formats.
It simple would not have been possible to have this package without Eric Scott @LeafyEricScott and Andrew Brown @humus_rocks - it's been a really fun project to work on together, looking forward to future iterations!
If you are somewhere between disillusioned & enraged with the direction that StackOverflow has taken, but like me you still periodically have the “answer programming questions on the internet" itch, here's a small suggestion that I’ve found rewarding.
Follow the GitHub issues of an open source project you use a lot. I 💯 guarantee you that they periodically get issues that are just confused users asking for help, or “bug reports” that are just a simple misunderstanding of the tool…
📣 I launched my first newsletter! It's free and I'll be using it to send a little email note any time I publish a blog post, publication, talk, or project on my #quarto site.
(1/4) TIL about the plotnine library- the grammar of graphics in Python 🚀
I had never heard about the Plotnine library until I came across the Posit Plotnine contest (see the link below). The plotnine is a Python implementation of a grammar of graphics based on the ggplot2 library.
For #API client packages, I want to make sure args are in the form the API expects. Other packages ({checkmate}, {vctrs}) didn't QUITE do what I want, so I https://xkcd.com/927/ 'ed it and made my own!
There's still a lot missing, and I'll likely tweak the errors to make things clearer, but it was done-enough to put it into the world. Enjoy!
For model calibration (esp via logistic regression), does anyone know of a statistical investigation of the properties of the resulting calibrated predictions?
IOW, if we use predictions from one model as inputs to another model, do we know the probability distribution of the final predictions?
It would enable the relatively easy creation of serious games and interactive exercises to learn R or statistics (or with the {rock} package, qualitative research); combining R with a simple game/adventure engine.
I just published a very basic introduction to using Boston's #Bluebikes bikeshare system data in #RStats. IMHO this is a great, underutilized dataset and is especially well-suited for student projects!
I've always wanted to try to get axis text directly on my plots like I've seen in the New York Times and other places. Finally figured it out with the help of the {ggh4x} package. Code here: https://rfor.us/axistext#rstats
The End To End Data Science With R is a new book by Rene Essomba. The book, as the name implies, focuses on the core data science applications using R ❤️. This book covers the following topics:
✅ Exploratory data analysis
✅ Data visualization
✅ Supervised learning
✅ Unsupervised learning
✅ Time series
✅ Natural language processing
✅ Image classification
Check out another fireside chat hosted by Audrey Yeo, featuring Heather Turner and Abhishek Ulayil. This week's chat is on building foundations for R’s future as an accessible and diverse collaboration.
Heather is an R Foundation member with a strong track record promoting diversity in contributions to R, while Abhishek has recently converted the R Journal content to more accessible web content, and keynote speaker at useR! 2024.
A new release of #rstats broom is on CRAN! v1.0.6 includes several changes to well-used tidiers from the package, e.g. for lm(), gam(), and survfit() output.
Next Tuesday I'll be part of a Fireside Chat alongside #useR2024 keynote Abhishek Ulayil on the topic of "Building foundations for R’s future as an accessible and diverse collaboration".