A CodeSights flagship: 6,377 tree taxa with every measurement their Wikipedia articles state and every fact the knowledge graph carries, built two-source because each alone is half a dataset. From prose: 4,408 maximum heights (median 20 m, ceiling held by the coast redwood at the 120 m physics guard), 1,326 trunk diameters crowned correctly by El Arbol del Tule's Taxodium mucronatum at 11.42 m after the extractor learned that a canopy 12 m wide is not a trunk, 315 maximum ages topped by the bristlecone pine's 5,062 dated years (clonal colonies rejected: Pando is a root system, not a tree, and the Norway spruce article praising the bristlecone very nearly aged the spruce), 1,577 leaf habits, 4,585 native ranges and USDA hardiness zones where stated. From Wikidata via QLever: family and genus for 6,372 rows (Myrtaceae leads with 654, then Fagaceae, Dipterocarpaceae, Fabaceae, Pinaceae, Rosaceae), IUCN status for 4,329 including 191 critically endangered, and the taxonomy-history columns: 5,029 described years, floored at 1753 because binomial nomenclature begins with Species Plantarum, whose half-century histogram shows the 1950s as the busiest describing era (1,211 taxa), and 1,186 describing botanists resolved through Wikidata's mul-label migration, which had hidden Carl Linnaeus himself until the en-or-mul fix put him where domain knowledge demands: first, with 240 trees. Year and author read as qualifiers on the taxon-name statement, where they actually live. Every row is a verified taxon (198 impostor articles like Tree measurement dropped by rank check). Eighteen acceptance tests: four species anchors, physics guards on every numeric column, the Linnaean floor, and three story tests, the ten tallest must be conifers and eucalypts, the oldest must be a pine, Linnaeus must lead.
6,377 rows in one CSV file, with 17 columns in total: Tree, Tree_ID, Common_Name, Rank, Family, Genus, Max_Height_m, Trunk_Diameter_m, Max_Age_Years, Leaf_Habit, Native_To, USDA_Zone_Min, USDA_Zone_Max, IUCN_Status, Described_Year, Described_By, Tree_URL.
It is built by programmatically scraping en.wikipedia.org, last pulled on 2026-10-13. Datasets are versioned; older versions stay downloadable.
Yes. Every CodeSights dataset is completely free as a CSV download; a free account is all it takes. Anyone can preview the data without signing in.
Yes. The exact Python scraper that built it is viewable on the dataset page by any signed-in member, so every number is reproducible.
Automated scraping leaves room for error and the underlying sources change over time, so no version is guaranteed accurate or complete. If a number matters, verify it against the original source.