/w=3840,quality=90,fit=scale-down)
Histograms and kernel density estimates (KDEs) are popular ways to visualize the underlying distribution of some sample. Both of these approaches are lossy, and poor choices of the number of bins or bandwidth can introduce or obscure structure in the data. Empirical cumulative distribution functions (ECDFs) are non-lossy visualizations that allow quantiles to be easily read off the plot. See https://towardsdatascience.com/what-why-and-how-to-read-empirical-cdf-123e2b922480 for more examples.
We added ECDF plotting functionality to Makie.jl, an interactive plotting package in the Julia ecosystem:
