June 27, 2026
2026 in 9 projects
A recap of 9 data visualization projects from the first half of 2026 — from D3.js in the AI era to new galleries, courses and tools.
A short, actionable data visualization tip in your inbox each week — on both design theory and the tools that bring it to life. Here are the past issues.
Join 19,000+ readersA recap of 9 data visualization projects from the first half of 2026 — from D3.js in the AI era to new galleries, courses and tools.
A brand-new Python package, nineJs, made a lot of noise this week. Here's what it does and why it's so exciting.
How to present a complicated chart without losing anyone.
Why, in 2026, your flow charts should literally flow — and how a touch of animation makes direction obvious.
AI removed the joy of coding for me. A personal reflection on building, pleasure, and what comes next.
Interactivity is the dimension changing everything in dataviz — and AI is finally making it accessible.
Want to break into dataviz? Build something and ship it. Here's a concrete idea: the Pacific Dataviz Challenge.
How a cold LinkedIn message to Cedric Scherer turned into a train to Berlin and a year building a ggplot2 dataviz course.
A raw 10-minute YouTube demo of building a chart in R with the help of agentic AI, plus a look at the levels of AI adoption.
How I grew the R Graph Gallery to 930k monthly page views, how it changed my career, and what side projects mean in the age of AI.
Mapping out four levels of AI adoption for data folks — from no AI, to chat, to agentic tools, personalization and skill files.
Interactivity in dataviz is now accessible thanks to AI, but like salt, a pinch elevates a chart while too much ruins it.
Why interpolating between two colors to build a gradient is far harder than it looks, and what to do instead.
To celebrate my 38th birthday, 38 charts that shaped my life in dataviz — from inspiring classics to data art and the AI era.
Two striking bivariate map examples, honest thoughts on whether they actually work, and how interactive legends can make them far clearer.
A tour of mynextchart.com, a huge AI-powered dataviz inspiration gallery, plus news of an upcoming talk on dataviz and AI.
Rebuilding a static chart as an interactive React and D3 version in 83 minutes, and what that reveals about AI's impact on dataviz.
Introducing D3 loves React, my cohort course for building bespoke, interactive web charts with D3.js and React.
Launching Data to Art, an online data art gallery, reflecting on the end of Shiny, and why now is the perfect time to learn D3.js.
Why bridge tools like Shiny and Streamlit are becoming obsolete in the AI era, and what to build with instead.
Get one short, actionable dataviz tip every Saturday. Join 19,000+ data people.
Why AI is changing which dataviz tools are worth learning, and how to plan your next tool now that the harder ones are more accessible.
Kicking off Dataviz Universe in 2026 with what's ahead: open-source projects, online courses, a plan to learn D3.js, and data-driven consulting.
A year-end chart remake: start from a standard Matplotlib line chart and improve it step by step with an interactive article.
An honest look at the less visible, harder side of being a dataviz freelancer, and why the path is not for everyone.
Why a chart's title deserves real attention, plus practical rules for writing titles that tell a story and hook your readers.
Looking back at 12 years of dataviz work, and sharing 13 free resources: graph galleries, inspiration, Data To Viz, a color-palette finder and more.
The messy-workflow feeling that haunted my early data-analyst years — and the three habits that got me out of it.
The five steps I'd follow to land a job in data: specify the role, learn with focus, build projects, show your work, and search.
How to help your audience adopt unconventional or complex charts by revealing them gradually, one layer at a time.
How the HSL color model lets you choose chart colors intuitively, instead of guessing RGB values like a robot.
The three valid reasons to use color in a chart: to encode values, to distinguish groups, and to emphasize what matters.
The most common pitfalls in data professional portfolios, from being too verbose to skipping testimonials, with tips to fix them.
What high-level vs low-level programming means, and how it helps you pick the right dataviz tool for the job.
A tour of six iconic dataviz projects everyone working with data should know, from Gapminder to the Climate Stripes.
Line charts are the safe default for time series, but niche alternatives like heatmaps, bump charts, and candlesticks earn their place too.
A beginner-friendly intro to D3.js: what the JavaScript library really is, why it is powerful but hard, and how React changed how you should learn it.
Why R still leads Python for data visualization, and how new matplotlib packages like bumplot are closing the gap.
A back-to-school checklist of skills every data analyst should master: cleaner code, reproducibility, Quarto reports, and Git and GitHub.
Wrapping up a year of the Dataviz Universe newsletter — looking back at what I built, thanking you all, and sharing what's coming next.
Ten personal career lessons, tracing my path from academia to dataviz solopreneur.
Get one short, actionable dataviz tip every Saturday. Join 19,000+ data people.
My six-step process for turning a raw dataset into a polished, well-annotated chart.
A little game with side-by-side chart pairs to help you sharpen the design instincts that make a dataviz look polished and trustworthy.
Too many overlapping points wreck a scatterplot. Here's how 2D density charts, hexbins, and contour plots fix the overplotting problem.
Why you should map data to a bubble's area and not its radius, plus how to size circles correctly in R, Python and D3.
Introducing the Dataviz Caveat Gallery and a set of visual flashcards to help you spot and avoid the most common chart-making mistakes.
The story behind ggplot2 [un]charted, a new R dataviz project I built with Cédric Scherer — from the R Graph Gallery to a Berlin train.
A choropleth map of Belgian unemployment where the histogram doubles as the legend, rebuilt in Matplotlib step by step.
Why every developer should use a code formatter, plus Air, a blazingly fast new R formatter, and the go-to options for Python and JavaScript.
A new chart type, the nested column chart, fixes the trade-off between stacked and grouped bar charts, with tutorials for Python, R and D3.js.
Static vs dynamic websites explained, and why Quarto often beats Shiny or Streamlit for sharing interactive dataviz.
Step 3 of the web essentials series: turn an HTML file into a live website for free by hosting it on GitHub Pages.
How R and Python let you build interactive charts with a single line of code, why those packages are really JavaScript wrappers, and how to share them.
A quick, beginner-friendly tour of how web charts are built, from HTML and CSS to SVG and JavaScript.
Eight alternatives to the classic line chart for visualizing temporal data, from streamgraphs and heatmaps to bump charts and calendar heatmaps.
Three projects make the Information Is Beautiful longlist, plus the animated Getting Older population project and the art vs. science tradeoff in charts.
Every map projection distorts the world. A quick look at why Mercator misleads us and how to choose a projection that fits your question.
Why Venn diagrams break down beyond three groups, and why the UpSet plot is a clearer alternative for showing set intersections.
A look at Trump's 2019 election map and why choropleth maps mislead, with bubble maps and color gradients as more honest alternatives.
Are no-code tools like Tableau, or code libraries like R, Python and D3.js, good enough to be "serious" about data visualization?
Why stacked area charts can be misleading, when they still work, and the alternatives worth reaching for instead.
Get one short, actionable dataviz tip every Saturday. Join 19,000+ data people.
Why legends add cognitive load, and two better alternatives: labeling groups directly on the chart and using color in the title.
Why dual Y-axis charts are misleading, plus an interactive playground showing how tweaking axis limits can tell any story you want.
Why annotations turn a good chart into a great one, with real examples showing how a bit of text guides readers and highlights the key result.
Why a high R² can be misleading, shown through an interactive scatterplot you can drag to watch correlation stay high on clearly curved data.
The story of my first impactful chart, and how side projects grew into a full-time dataviz career.
Most good charts share the same background rule: high luminance, low saturation, so colours support the data instead of competing with it.
Why error bars often mislead: they hide the underlying distribution and their meaning is ambiguous unless clearly specified.
A dataviz survey revealed how few people know powerful chart types like violin, ridgeline and dumbbell plots, and why growing your graphicacy matters.
Why the R community produces more polished charts than Python users, and how Matplotlib's steep learning curve helps explain the gap.
Why choosing between a treemap and a barplot comes down to one thing: the exact question you want your chart to answer.
A look at data art, where visuals are generated from real datasets, featuring Maayke Leenstra's mesmerizing river elevation maps.
A three-minute explainer on hierarchical edge bundling, the graph type that curves connections to follow a hierarchy and reduce clutter.
My mental model of the dataviz tool landscape — from Excel to Datawrapper, R, Python, JavaScript libraries and D3.js.
Why box plots silently hide each group's real distribution, and better alternatives like jittered points, violin plots, and raincloud plots.
The personal story behind my Productive R Workflow course: years of messy code, becoming a software engineer, and why good habits pay off.
The three things I wish I'd known as a young data analyst: clean R code, interactive reports, and collaborating with Git and GitHub.
How changing a histogram's bin size can hide or reveal the real story in your data, using 400,000 marathon finishing times as an example.
Discover webR and Pyodide, browser-based tech that runs R and Python code live so readers can tweak your analyses right from the web.
Meet Lumo, a free custom Quarto format that gives your R and Python reports a polished, ready-to-share look in seconds.
How I used AI, a boilerplate, and the React Graph Gallery to win the Pacific Dataviz Challenge in just 14.5 hours.
Get one short, actionable dataviz tip every Saturday. Join 19,000+ data people.
New Productive R Workflow updates: a student project gallery, copy-paste buttons on code chunks, and Lumo, a custom Quarto starter format.
My four-step process for turning a dataset into an impactful chart: find the options, pick the best, avoid pitfalls, and choose the right tool.
Get one short, actionable dataviz tip every Saturday. Join 19,000+ data people.