How do you draw a number
so that people read it right?
One paper's figures: eight of them, 62 panels, nineteen kinds of chart. From plain bars to UMAP, nearly everything in use today is in there.
Over these 20 lessons you redraw all nineteen from the real data — learning what each one shows, what it hides, and how a figure misleads without changing a single number. Python runs right inside this page. There is nothing to install.
How you'll learn
Eight figures, 62 panels, nineteen kinds — counted and turned into the syllabus. What a real paper actually uses is what you learn.
Not a stand-in dataset. The paper's own GSE285342 mouse RNA-seq and the TCGA lung data, exactly as published. In the volcano plot, Ddx54 really does go down.
Every lesson has a laboratory widget. Truncate the axis, swap the colour map, move the threshold — the same data becomes a different story, and you watch it happen.
Truncated axes, dual y-axes, cherry-picked examples, 3-D bars. You switch on each trick yourself — then leave with five questions that catch them.