I jumped from my chair

Could someone help me understand why most of Manhattan has no color in this map by @nytimes.com, showing Mamdani and Cuomo voters by 'commute type'? Is it not either public transit or cars, why is most blank? Can I find this broken down by type of transit (subway, buses) for these areas? #dataviz

"Their timing proved crucial. Had they started even a few months later, years of censorship data would have vanished". A nice profile of our project in The Frontline. This wasn't just #dataviz but also data acquisition, which is a painful process in India. frontline.thehindu.com/arts-and-cul...

A thread of some design/code highlights for me from our 'How Bangalore Uses the Metro' 📊 project. 1. The header sign is fully HTML/CSS. 'Go Down Take Up For Toilet' is real signage at one of our biggest stations (pic below) :P

Filed a Right to Information request with Bangalore's metro authority asking how many people traveled on the metro and got a file with 1.2M rows of data! We break our findings down in this story on how Bangalore uses the metro. w/ @vonter.bsky.social #dataviz diagramchasing.fun/2025/how-ban...

A great post from @djnavarro.net, providing an accessible explanation of why statistics that can appear to reveal our ignorance may actually do the opposite. Makes excellent use of #datavis to provide visual support for the argument. blog.djnavarro.net/posts/2025-0...

Experiments in visualizing crowds in Bangalore's metro station. The floor plans come in SVG formats, so I solved this by managing to constrain rendering clusters within a polygon with a certain color code (platform color is always consistent) :P

Today, @vonter.bsky.social and I launched cbfc.watch, an archive of over 100k censorship records for ~18k movies released in India since 2017. For the first time, you can search for any movie and see how they were modified before release. The site is fun to explore, but here's some #dataviz in it!

I've written a short blog post introducing "The Art of Visualization with ggplot2"! This book is all about the end-to-end process of creating charts with #ggplot2 📊 including data wrangling, sketching, and styling your plots! nrennie.rbind.io/blog/art-of-... #RStats #DataViz #TidyTuesday

The Art of Data Visualization with ggplot2: A free online book by @nrennie.bsky.social that guides us through the entire process of creating plots, including why certain decisions were made, using real datasets that have been part of #TidyTuesday. Very excited to get stuck into this one. #rstats

Big into doing long walks now. This is a 31 km walk I did from central Bangalore through the peripheries so I could get to the other end of a newly opened metro line and ride it back.

Frog calling patterns (as recorded by app users) for this week's #tidytuesday Highly recommend curating data if you haven't already - I thought it would be tricky but it was straightforward! Just follow these steps dslc-io.github.io/tidytuesdayR... Code: jessjep.github.io/blog/posts/t... #frogID

Hmm, is it possible for people like me who graduated with a degree in design but want to do a master's in something like statistics? I have a deep interest in data and some experience building and doing stuff, like at diagramchasing.fun and aman.bh but I would love to study the core subject more.

(kindly share) I'm selling photos to raise money private medical help for a severe - quite lifeshortening - immune damaging disease I have, #MECFS, outside the coverage of public health. There are some good nature photos in any event ;) More info here: payhip.com/MiskaKnapek/...

Would be interesting to see how you did that. I can see uses in my courses

Can a README file become a "dashboard" with #dataviz? Maybe. My web design module has students submitting daily journal entries as Github Issues, which means I can use the API and Github Actions to do all sorts of neat stuff. Simple processing and throw in some ASCII charts, you have a live dash.

Love the real-use examples!!

I'm collecting examples of beautiful/unique/well-made personal sites for a class on web design. Preferably sites of non-tech folks (writers, artists, designers), any suggestions? #indieweb

the 200th edition of. the Dataviz Dispatch from @datawrapper.de features a history of the series, how they choose figures, and why they link to paywalled articles. Well worth a read, as always: www.datawrapper.de/blog/data-vi... 📊

I've written an explainer blog post for my "data strips" #dataviz workbench web app, trying alternative 1-D distribution summary plots. rawdatastudies.com/2025/07/05/d...

Picked up some good readings while thinking about this. List: gist.github.com/thedivtagguy...

I especially enjoy (?) UEFA and their bizarre tournament structures, makes for a fun sankey:

In 2025 for a course on web design should someone be learning js/css/html specifics from scratch, or focus more on or design thinking/judgment/intuition + enough code to get you started. How would you rather learn? I am not against LLMs for code (only PMF, imo) so trying to understand the balance.

New: I recreated Tomas Wright’s 1742 astronomical treatise, Clavis Cælestis, and its accompanying sheets titled A Synopsis of the Universe. New posters are available in antique style and modern colors. Explore: www.c82.net/clavis-caele... How it was made: www.c82.net/blog/?id=99

Integrating the low-level WebGL library regl with Svelte works like a charm! At the same time this is the advanced data viz coding content that is so hard to find on the web. So I sat down for an hour and walked through a recent project using WebGL visualizations in Svelte. ➡️ youtu.be/rrdIP-oKyjE

Psst: If you're interested, I have a Discord server where we work on our dataviz and web dev projects in public to share progress, get feedback also have other people share their work! Process is much more interesting than the output, and this is my space to share more of that. Link below!

Our new project with 6k+ "pages" is fully-static. The source data is chunked and pre-processed, enabling fast lookup across different categories with fallbacks. We are able to do the same stuff we would need a database for by optimizing everything at build time. No devops headache needed later.

📊 LayerChart: Svelte 5 edition Released `2.0.0-next.2` that includes a massive migration to Svelte 5 state/snippets and MANY new features: ✔︎ Annotations with simplified chart integration ✔︎ Simplified patterns and Canvas support ✔︎ Radial BarChart ✔︎ Arc text and more (see 👇) next.layerchart.com

At the science communication panel I was on last Friday I mentioned the concept of an “attentive public”. If your audience is interested in your topic to there point where they are actively seeking it out, they’ll be enthusiastic learners and you don’t need to over-simplify. 📊 #dataviz

Found an old project from 4 years ago for which I mixed Harry Styles, Beatles songs and The Office quotes with Markov chains to send Rhea weekly newsletters of lyrical nonsense with matching GIFs. One of my first "big" Python projects 😛 Also, text from Markov chains is still funnier than GPT text.

This is true. I've also been pushed to realize this and now put extra effort into avoiding any sort of backend. Projects live longer, I can move on and there's no lock-in (just drop your static site somewhere else). Here's two projects and a WIP to showcase this.

A trick for fitting world maps on phones: Stack two azimuthal equal-area projections, clipped at 90° longitude, each one rotated 180° from the other. The bottom map covers the top's Antarctica, which reduces the total height and looks nice. As seen in www.nytimes.com/2025/04/25/c...

Want to create your own custom marks in @observablehq.com Plot, but don't want to go full D3? I've created a guide with 6 approaches of varying sophistication for specifying reusable custom marks: observablehq.com/@jwolondon/h...

In data visualization design, we often assume the goal is to minimise the effort required to 'see' the data. And that adding support a title and annotation is the route to memorability. But sometimes making something more effortful achieves memorability and understanding more effectively...

This year I'd like to spend more time learning how to make better tools. Viz is great, but making data accessible in this way feels rewarding. And absolutely crazy to me that SQL can now run in browsers, data of this scale can be processed reasonably fast. Opens up a lot of possibilities.

New project! Made a tool lets you explore, subset and summarize over 10M rows of data from the Time Use Survey of India in your browser. How people in different states, professions and groups spend their time? And then export your findings for #dataviz too! diagramchasing.fun/2025/time-us...

#rstats Version 0.9-3 of the {HistData} 📦 is now on #CRAN, largely a maintenance release. Another near record time from submission to accept (28 min). Already bumped the dev version, anticipating neat graphic examples from @drjohnrussell.com friendly.github.io/HistData/

I just got back from teaching a no-code #dataviz course, and the best side effect for me was learning and then teaching Orange. Data analysis is a blind spot for no-code teaching, and Orange handles it well. I compare common #rstats functions with Orange in this long post. aman.bh/blog/2025/vi...

Is there another simple online tool for stuff like this? Anything else I can add to make it useful for folks who teach #dataviz? Or any other small utilities too! If there's a need, I'd like to try my hand at making more stuff like this.

Made a little something to help students quickly convert between wide and long data for my (no-code) viz class. Helpful for when Datawrapper or Rawgraphs insist on a particular format for certain chart types. Built with #svelte and #rstats using webR! pivotteer.netlify.app

I've always found Shiny just a littttttlleeee bit too clunky for my taste and purposes (writing frontend UI in a very arm-twisting way) but maybe this is what I can do instead? Dashboards, APIs, analysis on the fly but Svelte + webr + whatever else? Maybe maybe.

webR is AMAZING!?! I'm running these interactive examples in a revealjs presentation all in the browser! Absolutely endless possibilities being able to write in Svelte, create UI like I normally would but have #rstats running for students to play with. Thats ggplot in the browser omg. #dataviz

Fantastic must read article by @moritzstefaner.bsky.social on the crisis (or not) in #dataviz Very well articulated and resonates deeply with me: truth-and-beauty.net/texts/crisis...

I made each of these graphs on ggplot2 and exported them to Adobe Illustrator to add further notes, titles, improve the legend, etc. You can find scripts to recreate each of them here: github.com/owid/noteboo... The README file has a table to help quickly find each one.

Love this interview so much. @fil.rezo.net is an inspiration to me and so many others. I'm always in awe of his work. "Now, we want to make maps quickly, and we have tools that are imposing their form. That means we’re assembling maps from existing parts, but the parts are more rigid" Great read.

At the bottom of this newsletter is a link to join 'the Bangalore amateur dataviz society', a Discord server where we're working on our #dataviz side projects with the garage doors up, in public. Sometimes the process is just as interesting as the outcome, come join us as we figure stuff out!

Published the first issue of my newsletter! I'll be using this space to share progress on ongoing side projects, readings and other web things that I found interesting that month. This time, it's on film censorship, weather and Dracula. Consider subscribing too :) buttondown.com/amanbh/archi...

Fav Svelte charting library just hit a milestone.

And I’d say that since that same time, the work of @jburnmurdoch.ft.com has had as large an influence — on both the field and the world — as any other’s.

Yesterday we launched the @vizchitra.bsky.social website and around ~100 showed up at the same time on the banner, made real-time interactive with partykit.io. Crazy nice feeling to see the lil cursors moving around all together on something I made (and things not breaking, phew). 🔗 vizchitra.com