#DotPlot
LightNews LightNews
mjskay.com
Dotplots

With good defaulta it beats the obvious alternatives, scales better than people think (like 10000+ data points) and isn't as sensitive to binwidth or bandwidth estimators

Density? Nope dotplot
Histogram? Nope dotplot
Violin? Nope dotplot
opinionhaver.bsky.social William B. Fuckley @opinionhaver.bsky.social · Apr 17
what's in your view the best data-visualization technique/type of plot/chart? in terms of concisely conveying information effectively and aiding substantive understanding. obviously depends hugely the use case, etc. but if you had to pick one in a vacuum?
April 18, 2025 at 12:47 AM
mjskay.com
Topline frequency-framed text forecast + a dotplot. Love it.
baseballot.bsky.social Nathaniel Rakich @baseballot.bsky.social · Jun 11
The 2024 presidential election between Biden and Trump is a tossup, according to 538's just-released forecast: projects.fivethirtyeight.com/2024-electio...
June 11, 2024 at 12:34 PM
mikaelniku.bsky.social
Pyysin tekoälyltä apua R-koodauspulmaan, kun en saanut aikaiseksi haluamaani kuvaajaa (boxplot + dotplot jossa datapisteet väritetty muuttujan perusteella). Mielenkiintoinen, vaikuttavakin kokemus! 👇
March 16, 2025 at 6:27 AM
mjskay.com
Just released a minor update to ggdist (v3.3.1) #rstats

Mostly minor stuff, but I did add a new bar dotplot layout based on feedback from @steveharoz.com: github.com/mjskay/ggdis...

Plus faster dotplot binning (especially for large n) using Rcpp (yay!)

Full changelog: mjskay.github.io/ggdist/news
A bar dotplot consisting of 7 categories, each containing four columns of bars. Each successive category has fewer dots. Within each category, the dots are split roughly 50-50 between two different colors. The plot is the result of this R code: data.frame( x = factor(c(rep(1:5, times = 5:1 * 11), 6, 6, 7)), g = c("a","b") ) %>% ggplot(aes(x, fill = g, group = NA, order = g)) + geom_dots(layout = "bar")
November 27, 2023 at 7:21 AM
mjskay.com
why do I have to help write VIS papers when I have all this cool new dotplot code to push

(tbf I am very excited about our papers this year)
March 29, 2026 at 4:22 PM
tommytang.bsky.social
chatomics! How to make a multi-group dotplot for single-cell RNAseq data divingintogeneticsandgenomics.com/post/how-to... #bioinformatics #rstats
December 5, 2024 at 3:15 PM
datavisfriendly.bsky.social
Dotplot
- plays smoothly with densities
- plays sweetly with violins
mjskay.com Matthew Kay @mjskay.com · Apr 18
Dotplots

With good defaulta it beats the obvious alternatives, scales better than people think (like 10000+ data points) and isn't as sensitive to binwidth or bandwidth estimators

Density? Nope dotplot
Histogram? Nope dotplot
Violin? Nope dotplot
opinionhaver.bsky.social William B. Fuckley @opinionhaver.bsky.social · Apr 17
what's in your view the best data-visualization technique/type of plot/chart? in terms of concisely conveying information effectively and aiding substantive understanding. obviously depends hugely the use case, etc. but if you had to pick one in a vacuum?
April 18, 2025 at 1:11 AM
mjskay.com
soon, soon... #rstats 📊
github.com/mjskay/ggdis...
Four dotplots in nested groups (two groups and two subgroups) with sub-axes on each dotplot labeling the count of the number of dots in each bin. Created using the "subguide" branch of the development version of ggdist.
January 12, 2024 at 6:32 PM
mjskay.com
Dotplots with exact or near-exact x position + automatic dot sizing are great EDA charts, they reveal artifacts histograms and densities miss. Often later I might use a histogram or density, but usually only after I've convinced myself they aren't lying to me using a dotplot.
April 18, 2025 at 12:54 AM
chenxinli2.bsky.social
New Catharanthus cultivar assembly coming up. x-axis is the chromosomes of a reference genome. y-axis is the contigs of the new assembly. Should scaffold to chromosomes nicely.
A dotplot showing alignment between 2 genomes.
February 14, 2025 at 8:52 PM
mjskay.com
Love this use of Sankeys!

Two quick thoughts:

1. Rotating the histograms could make comparison between pre and post easier.

2. Using dotplots instead of histograms allows seeing raw data and proportions simultaneously (plus allows coloring post by pre, keeping the Sankey color scheme throughout)
Vis of pre/post observations from an ordinal variable with 5 levels. First, a dotplot of pre observations, a Sankey diagram showing change from pre to post, then a dotplot of post values (colored by the pre value).
October 14, 2024 at 4:18 PM
mjskay.com
okay maybe it's like... *a little bit* cursed

geom_lava_lamp?
a weighted dotplot where each stack of dots contains dots of different heights, with the shortest dots on the bottom
March 9, 2024 at 6:12 PM
tjmahr.com
@mjskay.com i keep reading the ggdist notes and thinking "you're crazy for this one". you kick so much ass
a blurred dotplot
March 5, 2024 at 5:13 PM
mjskay.com
oh then you would hate ggdist::stat_mcse_dots()

mjskay.github.io/ggdist/artic...
Plot with four panels created using ggdist::stat_mcse_dots(). Each plot shows a dotplot of 100 quantiles calculated from samples of different sizes: 50, 200, 500, or 5000. The distribution is skewed, with a long right tail. Dots have a Gaussian blur applied using the Monte Carlo standard error for the estimate of the corresponding quantile. At low sample sizes the dotplot is quite blurry, particularly in the right tail; at n = 5000, only the rightmost few quantiles in the tail remain a little blurry.
December 30, 2024 at 8:56 PM
cedricscherer.com
🎉 I am thrilled to share my latest client project!

A set of data visualizations for the SPIEGEL bestseller book "Triggerpunkte" by @steffenmau.bsky.social, @thomaslux.bsky.social & @lwestheuser.bsky.social 🚀📚

Proud to contribute impactful visuals to such a remarkable work!
A copy of the book "Triggerpunkte" on my desktop next to a bottle of cola and my Mac keyboard. A scatter plot from the book showing the position of party editors and professional classes on a gradient of particularism vs universalism and state vs market with direct annotations and party editors being highlighted as larger, black points. Two connected dotplot / dumbbell charts from the book showing the distribution of sypathism with different groups for various groups of people. Two connected dotplot / dumbbell charts from the book showing the distribution of sypathism with different groups for various groups of people.
December 1, 2023 at 5:20 PM
lotp-hello.bsky.social
youtube.com/@dotplot-ish...
YouTubeの方も是非👍
度数分布
youtube.com
December 25, 2025 at 7:35 AM
kenveksler.bsky.social
so I guess rates and everything else waiting for the presser, bcs lets face it - that's a muted reaction to whatever was already baked in

Seemingly even the dotplot/SEP
September 16, 2026 at 6:05 PM
sponce1.bsky.social
📊 2025 #30DayChartChallenge | day 11 | distributions | stripes
.
🔗: stevenponce.netlify.app/data_visuali...
.
#rstats | #r4ds | #dataviz | #ggplot2
A dotplot showing the distribution of assist-to-turnover ratios for NBA teams in the 2024 season. Teams are sorted vertically by median ratio, with Jazz at the top and Celtics at the bottom. Purple circles represent players, with larger circles indicating more players with that specific ratio value. Most players cluster between ratios of 1.0-2.5. Alternating gray background stripes highlight each team row. A vertical dashed line marks the league median.
April 11, 2025 at 8:38 AM
datavisfriendly.bsky.social
Any #rstats font experts or friends out there?

I'm looking for fonts to use to try to reproduce the famous 1D dotplot of Michael Florent van Langren (1664), using common R font tools-- showtext 📦...

There seem to be at least 2 fonts-- the SmallCaps one used for most labels, script one for names
July 18, 2026 at 6:34 PM
mjskay.com
Mostly just me trying to make beeswarms / dotplots with exact x positions whose layouts I don't hate. Side benefit is they also allow stacking.

(+ some other minor improvements, including better automatic binwidth and faster layouts)
Stratified swarm dotplot with exact x positions and two groups stacked on top of each other. Low resolution of stratification (more regular row-like layout). Stratified beeswarm with exact x positions and two groups stacked on top of each other. Low resolution of stratification (more regular row-like layout). Stratified swarm dotplot with exact x positions and two groups stacked on top of each other. High resolution of stratification (less regular rows). Stratified beeswarm with exact x positions and two groups stacked on top of each other. High resolution of stratification (less regular rows).
March 29, 2026 at 6:17 PM
mjskay.com
Thanks! I'm fortunate to build on plenty of others who've thought about dotplot layout (e.g. Wilkinson) and beeswarm layout, plus my own touches :)
August 10, 2025 at 3:48 AM
drjohnrussell.com
Day 7 of #30DayChartChallenge is 'outliers', and I'm going slightly off script with this article, which makes an argument for a dataset that is nearly 50% outliers (more now)... The boxplot doesn't pick it up, but the dotplot does.

jse.amstat.org/v13n1/datase...

Code: github.com/drjohnrussel...
April 7, 2025 at 1:08 AM
mjskay.com
I work in #datavis, best known for uncertainty vis: see my work on election forecast vis (forecasts.cs.northwestern.edu), this talk (youtu.be/E1kSnWvqCw0), or my lab (mucollective.northwestern.edu)

I like writing #rstats 📦s: ggdist (mjskay.github.io/ggdist), tidybayes, ggblend, posterior::rvar...
Some examples from the three main families of ggdist geometries: slabinterval, dotsinterval, and lineribbon An example of an election forecast communicated using a quantile dotplot, from: Yang, Fumeng, Chloe Rose Mortenson, Erik Nisbet, Nicholas Diakopoulos, and Matthew Kay. "In Dice We Trust: Uncertainty Displays for Maintaining Trust in Election Forecasts Over Time." In Proceedings of the CHI Conference on Human Factors in Computing Systems, pp. 1-24. 2024.
September 7, 2024 at 2:24 PM
jgeller1phd.bsky.social
Those are cool! I have been wanting to try these out: dallasnova.rbind.io/post/creatin...
Simplifying Transparent Data Visualizations Using Faded Dotplots and Shadeplots | Dallas Novakowski
Building off of raincloud and fadecloud plots, I introduce two new styles for plotting data transparently: (1) the faded dotplot and (2) the shadeplot. Both leverage the r package ggdist to present st...
dallasnova.rbind.io
December 17, 2025 at 6:51 PM
mjskay.com
These were big motivators for ggdist::geom_dots(). It has a near-exact ("weave") and exact ("swarm") layout and automatic binwidth (optionally constrained by min/max size in data units or physical units). It is the product of years of me not wanting to futz with dotplot settings all the damn time.
August 9, 2025 at 9:21 PM