L-moment ratio diagrams are the swiss jacknife for distribution identification, and is in general a very useful tool for statistical exploratory analysis.

This requires:

• Calculation of the L-moment ratio curves for specific families of distributions, e.g. as $\tau_r(\tau_k | \mathcal{F})$.
• Non-parametric L-ratio curve estimation, either using repeated measurements, or some bootstrapping method
• Plotting both parametric curves (lineplot), and optionally a non-parametric curve (scatterplot), simultaneously, including helpful legend- and axis labels. Preferrably, this should be backend-agnostic, with (for now) one backend implementation in contrib.matplotlib .

Some examples for inspiration:

• M.C. Peel et al. (2001):
• J.R.M. Hosking (2015):
• J. Galeano-Brajones et al. (2023):