Georgios Karamanis
LightNews LightNews
banner
karaman.is
Georgios Karamanis
@karaman.is
2.7K followers 180 following 300 posts
Dataviz designer, psychiatrist, PhD student

https://karaman.is
Posts Replies Media Videos
karaman.is
For this week's #TidyTuesday, I visualized James Bond's drinking habits across the 12 novels using a waffle chart, based on the 2013 study "Were James Bond's drinks shaken because of alcohol induced tremor?"

Code: github.com/gkaramanis/t...

#RStats #dataviz
Waffle chart showing James Bond's weekly alcohol consumption across 12 novels. Each gold square represents one unit of alcohol. The chart is faceted by book title and year of publication, with consumption rates ranging from 52.4 units per week (Goldfinger) to 132 units per week (You Only Live Twice). Numbers labelled on each facet show the weekly consumption total.
January 11, 2026 at 3:25 PM
karaman.is
Genuary 2026, Day 5: Write "Genuary". Avoid using a font

#RStats code: github.com/gkaramanis/a...

#rtistry #genuary #genuary5 #creativecoding
Text-based generative art showing 'genuary' with simulated lighting and shadow effects. Letters appear white while the background shows banded color gradients and directional shadows.
January 5, 2026 at 2:09 PM
karaman.is
For this week's #TidyTuesday, I visualized the themes of Charles Dickens's 'A Christmas Carol.' Used ggpage for a page-by-page layout of the novel with color-coded words.

Code: github.com/gkaramanis/t...

#RStats #dataviz
Bar chart and page-by-page visualization of 'A Christmas Carol' by Charles Dickens. The bar chart shows the percentage of words related to wealth, joy, and cold in each chapter. Joy-related words (e.g., "merry," "happy," "love") have the highest percentage in all chapters, with Chapter 5 standing out at 1.20%. Cold-related words (e.g., "cold," "frost," "dark") are most prominent in Chapter 1 (0.47%), while wealth-related words (e.g., "money," "poor," "generous") appear consistently but at lower percentages. The page-by-page layout displays the novel's text with color-coded words, highlighting the thematic analysis.
January 4, 2026 at 3:13 PM
karaman.is
Genuary 2026, Day 4: Lowres

Low-res version of day 1: bsky.app/profile/kara...

#RStats code: github.com/gkaramanis/a...

#rtistry #genuary #genuary4 #creativecoding
A grid of 10 by 10 square tiles, each filled with either black or white, arranged in a spiral-like pattern. Some tiles are overlaid with smaller squares of the opposite color
January 4, 2026 at 8:58 AM
karaman.is
Genuary 2026, Day 3: Fibonacci forever

#RStats code: github.com/gkaramanis/a...

#rtistry #genuary #genuary3 #creativecoding
A generative art plot showing a grid of colored tiles arranged in a spiral pattern. The colors of the tiles are determined by the Fibonacci sequence, with five repeating hues assigned based on each tile’s position relative to Fibonacci-based breaks in the spiral. Some smaller tiles are randomly sampled and overlaid in contrasting colors.
January 4, 2026 at 8:52 AM
karaman.is
Genuary 2026, Day 2: Twelve principles of animation

#RStats code: github.com/gkaramanis/a...

#rtistry #genuary #genuary2 #creativecoding
January 2, 2026 at 2:40 PM
karaman.is
Genuary 2026, Day 1: One color, one shape

Code: github.com/gkaramanis/a...

Inspiration: bsky.app/profile/roni...

#RStats #genuary #genuary1 #creativecoding
A generative art piece featuring an 80 by 80 grid of small squares, arranged in a spiral pattern radiating from the center. The squares alternate between black and white. Some squares are randomly highlighted with smaller, contrasting tiles.
January 2, 2026 at 5:47 AM
karaman.is
For last week's #TidyTuesday, I created a map of endangered languages around the world. Built entirely in R/ggplot2: transforming spatial coordinates, identifying hotspots and composing multiple charts!

Code: github.com/gkaramanis/t...

Inspiration: www.reddit.com/r/MapPorn/co...

#RStats #dataviz
Map and bar chart showing the global distribution of 4 209 endangered languages. The world map displays threatened, shifting, moribund, and nearly extinct languages as colored points, with yellow areas highlighting regions of high language clustering. A notable concentration of approximately 700 endangered languages appears in Papua New Guinea and the Indonesian portion of New Guinea, shown in detail in an inset map (bottom right). The bar chart (bottom left) breaks down the 4 209 languages by endangerment status: 1 629 threatened, 1 835 shifting, 434 moribund, and 311 nearly extinct.
December 29, 2025 at 11:47 AM
karaman.is
For this week's #TidyTuesday, an interactive Shiny app to explore roundabouts around the world. Select a city from the table to zoom to its roundabouts on the map, search by location, and discover which cities have the most roundabouts.

Code: github.com/gkaramanis/t...

#RStats #dataviz #Shiny
December 21, 2025 at 3:16 PM
karaman.is
For this week’s #TidyTuesday, I had fun making a custom table with the reactable R package to show the car attributes in the new Qatar Cars dataset.

Code: github.com/gkaramanis/t...

#RStats #dataviz
December 14, 2025 at 4:33 PM
karaman.is
For this week’s #TidyTuesday, I visualized Zurich’s rising summer temperatures. In the past 11 years, avg temperatures rose above 19°C eight times, and highs surpassed 30°C every summer. The Böögg can’t predict the weather, let alone the warming

Code: github.com/gkaramanis/t...

#RStats #dataviz
Line and point chart visualizing Zurich’s summer temperatures from 1923 to 2025. For each year, the highest, lowest, and mean summer temperatures are shown. A dotted line marks 19 °C. Recent years show an increase in both average and maximum summer temperatures, with highs above 30 °C every year at least since 2015.
December 3, 2025 at 12:25 PM
karaman.is
For this week’s #TidyTuesday, a two-panel chart showing World Bank's SPI scores for Europe & Central Asia. Most countries in the region have high SPI scores, but a few lag behind.

Code: github.com/gkaramanis/t...

#RStats #dataviz
Two-panel chart visualizing Statistical Performance Index (SPI) scores from 2016 to 2023. SPI is a World Bank measure of national statistical system capacity, combining data use, services, products, sources, and infrastructure. The left panel shows smoothed lines for each world region, highlighting Europe & Central Asia, which is near the top, just below North America. The right panel displays small multiples for each Europe & Central Asia country, with individual line plots showing that most countries maintain high SPI scores, while a few, such as Uzbekistan, Moldova, and Turkmenistan, consistently score lower.
November 30, 2025 at 7:31 PM
karaman.is
For this week's #TidyTuesday, a horizontal bar chart showing the number of exclamation marks per 1 000 words in each Sherlock Holmes book. “The Adventure of the Dying Detective” has the highest rate with 7.72

Code: github.com/gkaramanis/t...

#RStats #dataviz
A horizontal bar chart visualizes the number of exclamation marks per 1 000 words in each Sherlock Holmes book. Each bar’s width represents the total word count for the book, and the bars are stacked vertically with book titles labeled on the left. "The Adventure of the Dying Detective" has the highest rate at 7.72 exclamation marks per 1 000 words, followed by “The Adventure of the Red Circle” (6.23) and "The Adventure of the Solitary Cyclist" (6.06). "The Musgrave Ritual" has the lowest rate at 0.92. The chart uses a minimal black-and-white color scheme. The title reads “Sherlock!” and the subtitle explains the metric and bar width.
November 23, 2025 at 6:29 AM
karaman.is
For this week's #TidyTuesday a faceted scatterplot showing TB case notification rate vs mortality rate by WHO region in 2023.

Code: github.com/gkaramanis/t...

#RStats #dataviz
A faceted scatter plot showing tuberculosis (TB) case notification rate (new, relapse, and unknown previous treatment cases per 100,000 population per year) and TB mortality rate (deaths per 100,000 population per year) by WHO region for 2023. Each facet represents a WHO region, with countries displayed as points sized according to population. Labeled countries are those with the highest case notification rates, mortality rates, or population sizes. The Western Pacific and African regions have the highest case notification rates, accounting for nearly 70% of global reported TB cases. Mortality rates are notably high in several African countries, indicating ongoing challenges with TB diagnosis and treatment access worldwide. Overall, case notification rates show a slight increase from 2022, while global TB deaths have decreased by about 5%.
November 16, 2025 at 12:55 PM
karaman.is
I visualized the Flint water crisis data for this week's #TidyTuesday showing how sample exclusions masked lead contamination.

Code: github.com/gkaramanis/t...

#RStats #dataviz
Three horizontal beeswarm plots showing lead concentrations in water samples from Flint, Michigan, measured in parts per billion (ppb). Each plot displays individual water samples as dots along the x-axis. A shaded grey area extends from 0 to 15 ppb, marking the federal action level threshold. A dashed dark red line marks the 90th percentile for each dataset. Grey dots indicate samples below the 90th percentile, while bright red dots indicate samples at or above the 90th percentile. The top panel, labeled "Official data (with exclusions)," shows 69 samples after Michigan officials removed two "outlier" samples. The 90th percentile line sits at 11.4 ppb, below the 15 ppb federal action level. Most samples cluster between 0 and 20 ppb, with a few extending beyond 40 ppb. The middle panel, labeled "Official data (complete)," displays all 71 samples including the two removed samples. The 90th percentile line rises to 18 ppb, now exceeding the federal action level. The distribution is similar but shows higher values in the upper range, with samples extending past 100 ppb. The bottom panel, labeled "Citizen science data," presents 271 samples collected by Virginia Tech researchers and Flint residents. The 90th percentile line reaches 26.6 ppb, substantially above the federal action level. This dataset shows a wider spread with more high-value samples, including several exceeding 100 ppb, revealing significantly worse contamination than official reports suggested.
November 10, 2025 at 6:46 PM
karaman.is
Day 5 of #30DayMapChallenge, Earth

Nature Index 2025 for countries/territories

Code: github.com/gkaramanis/3...

#dataviz #spatial #Rstats #map #science #research
Map of the world showing countries shaded by their share of articles published in scientific journals, according to the Nature Index for the period August 1, 2024, to July 31, 2025. Countries with higher shares are shown in darker colors. The five countries with the highest shares are labeled with their names and values (China, 36 826.66; United States of America, 21 980.49; Germany, 4 916.4; United Kingdom, 3 985.63; Japan, 3 344.12). The map uses a Van der Grinten IV projection and includes a legend for share values (log scale). The background is light gray, and the map has a minimalist design.
November 8, 2025 at 2:52 PM
karaman.is
Day 4 of #30DayMapChallenge, My data

Maps showing the locations of affiliations for me and my coauthors across six of our research papers.

Code: github.com/gkaramanis/3...

#dataviz #spatial #Rstats #map #science #research
A grid of small maps, each representing the geographic locations of affiliations for the author and coauthors of six research papers. Each map highlights a specific region with bright orange star-shaped markers, indicating the locations of the affiliations. The maps are cropped to focus on the relevant areas. The background is light gray, and the maps feature white landmasses with gray borders. Captions below each map display the title and publication year of the corresponding paper.
November 5, 2025 at 6:12 AM
karaman.is
Day 3 of #30DayMapChallenge, Polygons
Uppsala University, founded in 1477, is the oldest university in the Nordic countries. Here are the 30 buildings marked as 'university' and labeled with a name on OpenStreetMap

Code: github.com/gkaramanis/3...

#dataviz #spatial #Rstats #science #research
Visualization of Uppsala University buildings in Sweden. Each building is represented as a polygon, highlighted in orange with black borders, overlaid on a grey map of surrounding structures. The title 'Uppsala University buildings' is centered at the top, and the caption reads 'Source: OpenStreetMap · Graphic: Georgios Karamanis'. The layout features 30 panels, each focusing on a specific building.
November 3, 2025 at 9:08 PM
karaman.is
Day 2 of #30DayMapChallenge, Lines

Uppsala University researchers have collaborated with hundeds of research institutions worldwide. More than 14,000 publications have been co-authored with Karolinska Institutet alone.

Code: github.com/gkaramanis/3...

#dataviz #spatial #Rstats #science #research
A world map in Robinson projection showing Uppsala University's research collaboration network. The map highlights countries containing the top 100 institutions that have co-authored publications with Uppsala University. Red points mark each collaborating institution, with point size proportional to the number of co-authored works. The top collaborating institutions per continent are labeled, including their collaboration counts. The most frequent collaborators include Karolinska Institutet in Europe with 14 598 publications, Lund University with 6 971, Stockholm University with 6 107, and Centre National de la Recherche Scientifique with 5 316. In North America, top partners include the United States Department of Energy with 2 335 publications, Harvard University with 1 976, and Stanford University with 1 517. In Asia, the Chinese Academy of Sciences leads with 1 847 publications, followed by Peking University with 847. The University of Melbourne represents Oceania with 822 publications. Countries with collaborations are shown in white against a light gray background.
November 2, 2025 at 7:05 PM
karaman.is
I created a heatmap for this week's #TidyTuesday showing UK literary prize demographics with data from @post45data.bsky.social. White British authors account for 61% of winners vs. 46% shortlisted. Men win 56% vs. 46% shortlisted.

Code: github.com/gkaramanis/t...

#RStats #dataviz
A heatmap showing gender and ethnic representation in UK literary prizes, split into two panels for shortlisted authors and winners. The visualization displays counts across ethnicity categories (African, Asian, Black British, Caribbean, Irish, Jewish, Non-UK White, Non-White American, and White British) on the y-axis and gender categories (Man, Non-binary, Transman, and Woman) on the x-axis. Each cell shows the count with white text on a colored background, where darker colors indicate higher counts. White British men show the highest counts in both panels (120 shortlisted, 149 winners), followed by White British women (123 shortlisted, 105 winners). Non-UK White authors also show substantial representation (39 men and 87 women shortlisted; 40 men and 32 women winners). Most other ethnic groups show single-digit counts, with several cells having zero representation. The data reveals that White British authors increase from 46% of shortlisted to 61% of winners, while women decrease from 54% of shortlisted to 43% of winners.
November 2, 2025 at 5:11 PM
karaman.is
Day 1 of #30DayMapChallenge, Points

There are almost 120,000 research organizations, in more than 17,000 locations, registered in the Research Organization Registry @researchorgs.bsky.social

#Rstats #dataviz #spatial #map #science #research
A world map in a Robinson projection showing the locations of research organizations registered in the Research Organization Registry (ROR). Countries are outlined in light gray, excluding Antarctica. Points on the map represent the number of organizations at each location, with larger and darker points indicating higher numbers. Labels for locations with 300 or more organizations are displayed, such as city names, in dark text with a white background. The map has a clean, minimalistic design with a white background and a gradient color palette for the points. The title reads 'Research organizations of the world,' and the subtitle explains that the size of the points corresponds to the number of organizations in each location.
November 1, 2025 at 12:41 PM
karaman.is
#Makingof of this week’s #TidyTuesday
October 26, 2025 at 6:56 PM
karaman.is
For this week's #TidyTuesday, rainfall from 12 UK weather stations 2012-2024. I’ve been wanting to try the idea of using letter heights to show data for some time, now possible thanks to {flexfont} by @coolbutuseless.fosstodon.org.ap.brid.gy

Code: github.com/gkaramanis/t...

#RStats #dataviz
A data visualization showing median annual rainfall at UK weather stations from 2012 to 2024. The chart displays twelve weather stations (Ballypatrick Forest, Bradford, Camborne, Cardiff Bute Park, Chivenor, Dunstaffnage, Eskdalemuir, Hurn, Lerwick, Newton Rigg, Stornoway Airport, and Tiree) arranged in two columns. For each station, the years 2012-2024 are shown along the x-axis, with rainfall amounts in millimeters on the y-axis (ranging from 0 to 200+ mm). Eskdalemuir recorded the highest rainfall (213.3 mm in 2020), while 2022 was notably dry across most stations. Each station's name is visualized by repeating its letters across the years, with letter height corresponding to that year's median rainfall. The most recent complete station name appears in dark blue, while earlier repetitions are shown in light blue. Blue bars at the bottom of each panel represent the year markers. A horizontal red line at the top of each panel indicates the maximum rainfall recorded for that station during the period.
October 26, 2025 at 2:30 PM
karaman.is
For this week's #TidyTuesday, I explored global differences in severe food insecurity between women and men. In 78 countries, women face higher rates, with the largest gaps seen in Pakistan and Afghanistan.

Code: github.com/gkaramanis/t...

#RStats #dataviz
Dot and rug plot with colored background bars showing the difference in prevalence of severe food insecurity between women and men across countries, ordered by difference. Each country is represented by a black point along the y-axis, with the x-axis showing the difference in percentage points. Vertical background bars are colored on a gradient from green to purple, indicating the value of the difference: purple for countries where women have higher rates, green for countries where men have higher rates. Rugs above the plot highlight positive and negative differences. Country names are labeled for the largest differences. For example, women have higher rates in countries such as Pakistan, Afghanistan, and Peru, while men have higher rates in countries including Jamaica, Libya, and Djibouti.
October 19, 2025 at 1:47 PM
karaman.is
For this week's #TidyTuesday, EuroLeague basketball club championships from 1958 to 2024, across four competition formats. Used Wikipedia data instead of the provided dataset 😅

Code: github.com/gkaramanis/t...

#RStats #dataviz
A horizontal tile plot showing the top 10 EuroLeague basketball clubs by championships from 1958 to 2025. Each club is displayed in a separate row, with colored tiles marking years won and gray tiles marking runner-up years. The x-axis shows years from 1960 to 2025, grouped by competition format: FIBA European Champions Cup, FIBA European League, FIBA EuroLeague, and EuroLeague (ECA). Real Madrid leads with 11 titles and 10 runner-up finishes. Darker tile shades indicate more cumulative titles. The plot background is light gray, with club names and title counts on the left.
October 12, 2025 at 2:33 PM