Open GabrielSartori opened 5 years ago
ternary plots are great if you happen to have compositions with exactly three components, otherwise you have to look at the recommendations that Wilke gives for proportional data. There are really tough tradeoffs here; maintaining the visual metaphor of compositionality for data with lots of components causes other problems (too many colors, small slivers of area in a pie or bar chart, etc.)
What to do when i have composition data? What is the best practise?