Post

🙇‍♀️

This is so friggin good 🤩 Thanks for sharing, it’s a wonderful resource!

Clear and thought-provoking as always. Thanks. I'll share the article in my other social networks.

Why stop at reproducible charts? Make the whole analysis fully reproducible, including getting the data, processing etc, and of course the graph. (I am a big fan of quarto notebooks for this.)

📌

this is really detailed, and explosive regarding various use case around visualization, thanks.

🫡

📌

Read this, and recommend to your students. Lots of good stuff here. 👍

I love that you mention Tufte :) See also "information is beautiful" - informationisbeautiful.net

This is a great article. I love me a good data visualization! Ed Tufte is one of my idols!

Thank you thank you thank you thank you...

A case in point from the oldest known statistical graphic: van Langren's 1644 line of twelve estimates for the Toledo-Rome longitude. Plotted, the disagreement is the finding - they scatter across more degrees than the real gap, and every one overshoots. A table hides that.

Bookmarked. Worth a line on the origin: the first known graph of statistical data is Michael van Langren's 1644 plot of twelve rival estimates for the longitude between Toledo and Rome, laid out on a single line. The scatter was the argument - and every one of the twelve was too wide.

Bookmarked. The theory I'd put early in any guide is Jacques Bertin's Semiologie graphique (1967): every mark breaks into seven visual variables - position, size, shape, value, hue, orientation, texture - and only some carry order. Most bad charts encode a quantity in one that can't rank.

Great guide. John Snow's cholera map deserves its place with an asterisk: he drew the famous map after the outbreak, for the report — and noted the epidemic was already fading when the pump handle came off. The map didn't stop the outbreak; it won the argument afterward.

Nice. The case study I'd add is Nightingale's 1858 polar-area diagram of Crimean mortality: wedges sized by deaths, shaded by cause, showing preventable disease dwarfing battle wounds. She wasn't illustrating a report, she was lobbying Parliament — and became the RSS's first woman member that year.

Nice. A founding oddity of the field: Playfair invented the bar chart in 1786 because of missing data. He had no time series for Scotland's trade, only one year, so each partner became a bar - and he wrote in the text that it was an inferior substitute for a proper line chart.

Oh that's wonderful! Very happy to hear that :)

From our perspective (Hydrogeology /Geotechnical) Data analysis/ modeling/simulation/ visualisation/animation etc 🔍 ResearchGate: www.researchgate.net/profile/Dr-Y... 📚 Academia: latrobe.academia.edu/YohannesYihd... 🐦Twitter (X): twitter.com/Yohannesyihd...

Thank you so much ☺️☺️☺️

Now with an audio voiceover! 🎧

Excellent posts - this will be a great resource for my maps and data viz course next semester!

Thank you!! ☺️

Thanks for sharing.

Thank you for including colorblind accessibility!