Four penguins newly measured, typed by hand, each compared with the average profile of the known species: the nearest one, how far, and where they sit. On the Palmer penguins (palmerpenguins, CC0): 333 penguins of three species.
The penguins’ four measures correlated two by two, over all penguins then within each species, where one relation changes sign; a heatmap clusters the measures by how alike they vary. On the Palmer penguins (palmerpenguins, CC0): 333 penguins of three species.
The shop’s customers counted per city and drawn where the cities are, then counted into the regions of France, each coloured by its revenue per customer. Needs the geo module. On the synthetic shop (180 customers in twelve cities) and the product’s map of the French regions.
Four years of daily temperatures: every day on a calendar, the seasons month by month, then the hot days counted per year. A workshop on dates: turn a day into its month or its year, then aggregate. Data: Météo-France, daily climatological data, station Paris-Montsouris, 2022 to 2025 (Licence Ouverte 2.0).
The pieces of a trial’s results in one flow (baseline table, survival curves, median survival, survival at 12 and 24 months) and a Word report that assembles them, each section introduced and captioned. Needs the statistics module. On a small two-arm clinical trial: 90 patients.
Four measures cannot be drawn: three ways to put the penguins on two axes (PCA, an MDS that also weighs island and sex, UMAP), and how to judge a map. Needs the statistics module and the python-ml environment. On the Palmer penguins (palmerpenguins, CC0): 333 penguins of three species.
From dates to time to event, then Kaplan-Meier curves per arm with the log-rank test, the median survival and the share still alive at 6, 12 and 24 months. Needs the statistics module. On a small two-arm clinical trial: 90 patients.
Four measures per penguin, grouped in three without looking at the species; then the groups set against the species they should find. Needs the statistics module. On the Palmer penguins (palmerpenguins, CC0): 333 penguins of three species.
Gapminder’s famous bubbles: income against life expectancy in 2007, each continent’s life expectancy since 1952, and the countries that gained the most. On the Gapminder data: 142 countries, 1952-2007.
Before judging a treatment, the arms must look alike at randomisation: a Table 1, the age spread, and a test per characteristic. Needs the statistics module. On a small two-arm clinical trial: 90 patients.
