iiinova — where your digital self keeps evolving

Leonardo Hansa @handle.invalid · 12/02/2026

Strange but efficient way of updating values on a list in R. Makes sense after you're explained how object copies work. kaizen-r.github.io/posts/2026-0... #rstats

050

Leonardo Hansa @handle.invalid · 19/05/2025

While attending online meetings, today I'm listening to Marin Marais by Brandon Acker and Craig Trompeter in the background. It's the only thing worth it in today's meetings.

010

Leonardo Hansa @handle.invalid · 30/03/2025

On our next podcast episode, @joscani.muestrear-no-es-pecado.es and I talk about how to correct your model's optimism using bootstrap. You train your model on each sample and compare its fitness metric against the original sample. The average of differences is your correction. Soon on Spotify

001

Leonardo Hansa @handle.invalid · 30/03/2025

Have you ever done a bootstrap plot? I don't know if it's valid, or even useful, but I just have. Just for fun. Sepal Length bootstrap density #rstats #dataviz

281

Leonardo Hansa @handle.invalid · 18/03/2025

For quick tests these days I'm using R GUI. I've just realised one of the most accessible buttons is the `Print` button. Who on Earth (and when) could be interested on printing a script? #rstats #20thCentury

230

Leonardo Hansa @handle.invalid · 28/02/2025

Is this a barplot? Where all bars have the same length and you need to read the labels and quantify their difference with your imagination? Source: Lego . com #dataviz

000

Leonardo Hansa @handle.invalid · 15/12/2024

Advise requested -I haven't yet used #NotebookLM. Somehow I find it a super powerful tool, though I haven't still figured out what for 😅 Any use case you're willing to share and you consider inspiring? 🙏

010

Leonardo Hansa @handle.invalid · 06/12/2024

I can't get used to indicating the percentage of a percentile as a percentage... Yeah, I know the sentence makes no sense, but I've always used proportions, a number from 0 to 1, not 0 to 100. Creating confidence intervals with 2.5 and 97.5 instead of 0.025 and 0.0975 seems like cheating. #stats

000

Leonardo Hansa @handle.invalid · 04/12/2024

Fiddling around with #aRtsy library(aRtsy) set.seed(6) canvas_slime(colors = colorPalette("vrolik4"))

020

Leonardo Hansa @handle.invalid · 04/12/2024

Loving the #aRtsy project for creating beautiful random images with #rstats github.com/koenderks/aR...

121

Leonardo Hansa @handle.invalid · 28/11/2024

During my first steps with #rstats, @posit.co's cheatsheets helped me out a lot easing the learning process. One of the (many) things I still struggle with on #Python is #dataviz, because of my love towards #ggplot2. Perhaps the matplotlib cheatsheets can be the key matplotlib.org/cheatsheets/

200

Leonardo Hansa @handle.invalid · 20/11/2024

Great tutorial by @albertrapp.bsky.social on #ggplot2 and some more #rstats extensions on how to do the plot on the image. Very well guided and lots of tiny tips on how to improve your #dataviz www.youtube.com/watch?v=e0_C...

150

Leonardo Hansa @handle.invalid · 20/11/2024

I didn't know you could assign a class to an error in #rstats, using the rlang package. This allows you managing your tryCatch based on different error classes. PS. Yeah, I guess for #Python users this is like an obvious thing.

Code for assigning a class to an error in R via the rlang package

040

Leonardo Hansa @handle.invalid · 20/11/2024

I'm starting to apply cli #rstats package after @hfrick.bsky.social's talk at Spanish R Conf. Question on best practices: which is the best approach to stopping execution and sending nice messages via cli? Should I use `cli::cli_alert_danger()` and then `stop()`?

First page on cli R package documentation

Example of error messages with cli R package

150

Leonardo Hansa @handle.invalid · 18/11/2024

I'm loving the officer #rstats package. Yeah, I know, Quarto slides blah blah... Look, sometimes (not often fortunately) I'm asked to do a PowerPoint. The officeverse is the way. ardata-fr.github.io/officeverse/

Text introduction to officeverse, a set of R packages oriented to work with Microsoft's Office package.

030

Leonardo Hansa @handle.invalid · 14/11/2024

Last week, my teammate Paula and I presenting our 'Confession: we develop software in R and Excel', our talk at the Spanish R Conference. Slides (in Spanish): ebiquityds.github.io/CongresoRSev... #rstats #Excel

181

Leonardo Hansa @handle.invalid · 11/11/2024

Last week, 14th Spanish R Conference. 14 years of #rstats gatherings, 2 days of talks and workshops. Inspiring. Awesome. Thanks a lot to the organizers!

Picture of 14th Spanish R Conference attendees

0117

Leonardo Hansa @handle.invalid · 25/10/2024

People seem to take a break from statistics on Google while Christmas

Google Trends for "Statistics" worldwide during the last three years. Seasonality with decrease during Christmas

020