The feed of updates to Georgios Karamanis
This week’s data is the Palomar spectroscopic survey of nearby galaxy nuclei by Ho, Filippenko and Sargent. For each galaxy it records the brightness of emission lines from gas in the core. I used the ratios of four of those lines to recreate a BPT diagram . Point shapes show each galaxy’s Hubble type, and the small bar chart counts them. Most are spirals. Code on GitHub
This week’s data is Lesotho’s wool trade from UN Comtrade, pulled with the comtradr package. I kept raw wool (commodity code 5101) and the two importers that take almost all of it, China and South Africa. Values are monthly and reported by the importing country. Lesotho is landlocked and its wool has gone out through South Africa for decades, sold at auction in Port Elizabeth. In 2018 the…
This week’s data is Australian tourism from the ecotourism R package. It has quarterly counts of domestic overnight trips by region, from 2014 to 2022. I kept the holiday trips, assigned each region to a Greater Capital City Statistical Area, and calculated each area’s share of the national total in the first and third quarters. The two maps are mosaic cartograms, built with the Python popgrid…
This week’s data is a collection of near-death experience accounts from the Near Death Experience Research Foundation ( NDERF ). I scraped the archive and kept the roughly 5,570 entries that had a valid date, then counted them by the day and month each person gave as the date of their experience. Most of the calendar is low and even, but a few dates stick out: the 1st and the 18th of January, and…
This week’s data is a set of body measurements for 93 penguins, spanning 18 species and six genera. It comes from the AVONET database of bird traits (Tobias et al. 2022). I plotted the distribution of ten measurements for each genus, split by sex. Each triangle is one penguin. Some genera are well sampled, while others come down to a handful of birds. Code on GitHub
Code on GitHub
This week’s data is UFC fight and athlete data from the fightr R package , which collects it from UFC athlete profiles, UFC Stats, Kaggle and the Octagon API . I plotted how fights end in each weight division, from lightest to heaviest. The pattern is hard to miss: the knockout share climbs steadily with weight, from 14% in women’s strawweight to 52% in men’s heavyweight, while the lighter…
This week’s data is the Wreck Inventory of Ireland Database ( WIID ), a record of thousands of known and suspected shipwrecks in Irish waters, maintained by the National Monuments Service. I mapped every wreck that has coordinates over a bathymetric basemap. They cluster along the busy east and south coasts and the northern approaches, where many ships were lost to mines and submarines in both…
This week’s data is two papal encyclicals: Leo XIII ’ s Rerum Novarum (1891) and Leo XIV ’ s Magnifica Humanitas (2026). I tokenized both texts, removed stop words, and computed each word’s rate within its own encyclical. The chart shows the ten words from each encyclical with the largest difference in rate between the two. A word’s side and length show which encyclical favors it, and by how much.…
This week’s data comes from the UK birth registries, covering baby names from 1997 to 2024. England and Wales come from the ONS , Scotland from the NRS , and Northern Ireland from NISRA . I looked at names given to both boys and girls and picked the twelve whose split between the two shifted the most over the period, keeping only names with at least 20 babies a year. Each small chart is one name,…
This week’s data comes from Wikipedia’s List of films based on video games . I picked 15 of them, from the 1993 Super Mario Bros. to the 2025 A Minecraft Movie, and built an animated timeline that reveals each film in turn with its worldwide box office, its multiple of the production budget, and its Rotten Tomatoes score. The multiple is worldwide gross over the midpoint of the production budget.…
This week’s data comes from the European Parenting Leave Policies ( EPLP ) Dataset , which covers maternity, co-parent, and parental leave regulations across 21 European countries from 1970 to 2024. The chart shows how many weeks of maternity leave birth mothers are entitled to before and after childbirth, split into mandatory (crosshatch) and voluntary (outline) portions. Countries are sorted by…
This week’s data comes from Sustainable Energy for All ( SE4ALL ) , and covers the share of total final energy that comes from traditional biomass such as firewood, charcoal, crop residue and dung burned in open fires and inefficient stoves, from 1990 to 2010. I mapped it with geofacet : each country becomes a small stripe chart of its biomass share over the two decades, arranged on a world grid,…
This week’s dataset comes from Crossref , a DOI registration agency for scholarly content. The data covers metadata quality statistics, namely how consistently publishers provide structured information like references, abstracts, author identifiers, and funding details alongside their DOIs. I built a table with the gt package showing the top 20 countries by journal article DOI volume, grouped by…
This week’s dataset comes from Wikidata, collected by Ahmad Barclay for the Twin cities explorer . It covers sister city agreements, formal bonds between cities for cultural exchange and cooperation. I built a scrollytelling map focused on Rio de Janeiro, which has more sister city agreements than any other city in the dataset. The story starts with all 5 470 cities in the Wikidata dataset shown…
This week’s dataset covers Italian industrial production statistics from Istat. I focused on passenger cars and extended the series using a table from the Wikipedia article about the automotive industry in Italy . I annotated the chart with events from the same Wikipedia page to give the numbers some context. Code on GitHub
This week I looked at US import taxes on agricultural goods, showing the highest rate within each product category. The rates are shown as a percentage of the product’s value, using the standard tariff that applies to all trading partners. Tobacco is taxed at 350%, peanuts (Oil seeds) at 164%, peanut butter (Vegetable & fruit preparations) at 132%, all products with a long history of protection.…
This year’s personal theme is Uppsala transportation Code on GitHub Day 1, Part-to-Whole Day 2, Pictogram Day 3, Mosaic Day 4, Slope Day 5, Experimental Day 7, Multiscale Day 8, Circular Day 9, Wealth Day 12, FlowingData Day 13, Ecosystems Day 14, Trade Day 15, Correlation Day 16, Causation Day 17, Remake Day 19, Evolution Day 20, Global Change Day 21, Historical Day 22, New Tool Day 23, Seasons…
This week’s dataset comes from the WHO Global Health Expenditure Database ( GHED ), covering health spending by purpose across countries and years. I focused on preventive care spending as a share of current health expenditure for European countries, 2016 to 2023. Rather than showing raw values, where a shared y-axis gets dominated by a few high-spending outliers, I calculated the year-on-year…
This week’s dataset comes from the at-sea seabird observation records held by the Museum of New Zealand Te Papa Tongarewa, built largely from the handwritten logbooks of Captain John Arthur Francis Jenkins (1928–1989). Over nearly two decades of voyages in the Southern Ocean, Captain Jenkins recorded 6 583 bird sightings, far more than any other contributor to the Australasian Seabird Mapping…
This week’s dataset comes from RepairMonitor, which has been logging repairs at volunteer-run Repair Cafés worldwide since 2015. At these community events, volunteer fixers help people repair broken household items, recording a repairability score for each. The chart shows smoothed repairability scores by item age at the time of repair, with older items on the left and electric and non-electric…
This week’s #TidyTuesday dataset comes from the Centre for Marine Applied Research, which has been recording daily ocean temperatures along Nova Scotia’s coast for seven years at depths of 2, 5, 10, 15, 20, 30, and 40 metres. The visualization maps each month as a radial plot, where rings represent depth (shallower on the outside, deeper toward the centre) and color shows mean temperature. Summer…
This week’s #TidyTuesday dataset is a million digits of π, which is far more than anyone will ever use. Five digits is enough for everyday math. NASA uses only 16 to navigate Voyager 1 across the solar system with sub-centimeter error. Thirty-eight would let you calculate the circumference of the observable universe to within the width of a hydrogen atom. A hundred is the ceiling for any practical…
This week’s #TidyTuesday data comes from the Norwegian Veterinary Institute, which tracks salmon farming mortality across Norway’s 13 coastal production areas. At its worst, one in six farmed salmon died before reaching harvest. Roughly 38% of deaths are from infectious disease, another third from physical trauma during de-licing. The rest have no clear cause. The map shows mean monthly mortality…
This week’s #TidyTuesday data comes from Adam Kucharski’s CAPphrase survey, which asked people to translate probability phrases into numbers. I look at “ Highly likely” and “ Highly unlikely”, comparing responses by English fluency level. Each bar is a probability value from 0 to 100%, and its hight shows how many people gave that estimate. Darker bars mean a higher proportion of responses at that…
This week’s #TidyTuesday is about Hermann’s tortoises in North Macedonia, where males now vastly outnumber females and the population is collapsing toward extinction. The data come from a 2026 paper in Ecology Letters: Sex Ratio Bias Triggers Demographic Suicide in a Dense Tortoise Population by Arsovski et al., covering 16 years of capture surveys on the Golem Grad island plateau in Lake Prespa.…
This week’s #TidyTuesday is about grants awarded by Science Foundation Ireland ( SFI ) from 2005 to 2024. The chart shows total committed funding per year as a horizontal bar. Used {tidytext} to tokenize proposal titles and compute TF-IDF scores, a measure of how distinctive a word is to a given year relative to all other years. The top 3 words (or more if tied) are fitted inside each bar using…
New Zealand once had over 20 sheep per person. Animated this week’s #TidyTuesday data showing how sheep numbers, human population, and their ratio have changed from 1960 to 2024. The flock has been shrinking while the population keeps growing! Code on GitHub
Mapped all the 2026 Winter Olympics venues across Italy for this week’s #TidyTuesday. Started simple but kept adding stuff because I couldn’t stop myself, had way too much fun working through it all! Code on GitHub
The Edible Plant Database ( EPD ) grew out of the GROW Observatory, a European Citizen Science project about growing food, soil moisture sensing and land monitoring. For this week’s plot, I sampled some plants with a growing temperature range greater than 2 degrees Celsius, sorted them by the midpoint of their range, and tried out bitmap fonts and color gradients. Minimum and maximum temperatures…
Brazil’s Cadastro Nacional da Pessoa Jurídica ( CNPJ ) maintains a comprehensive registry of all legally established companies in the country. This visualization breaks down the composition of companies by owner qualification, the legal role or qualification of the person operating the business, and their company sizes. The width of each bar is scaled logarithmically to account for the vastly…