Estimation of Meanimiles

Provides a comprehensive suite of estimation tools for meanimiles, a general class of (risk) functionals. This package includes nonparametric estimators for univariate meanimile evaluation, copula-based estimation for portfolio risk aggregation (full parametric, semiparametric, and nonparametric), and novel estimators for meanimiles in regression settings. Following the articles D. Debrauwer, I. Gijbels, and K. Herrmann (2025) , D. Debrauwer and I. Gijbels (2026) .

Meanimiles

This R package provides tools for the statistical estimation of meanimiles, a general and flexible class of (risk) functionals that unifies quantiles, expectiles, and extremiles.

In particular, nonparametric estimators for univariate meanimile are implemented, as introduced and studied in:

• Debrauwer, D., Gijbels, I., & Herrmann, K. (2025). On a general class of functionals: Statistical inference and application to risk measures. Electronic Journal of Statistics, 19(1), 2456-2510. doi: 10.1214/25-EJS2391.

Furthermore, the package includes several copula-based estimation procedures for portfolio risk aggregation. These accommodate parametric and nonparametric methods for both margins and dependence structures, as investigated in:

• Debrauwer, D. & Gijbels, I. (2026). Copula-based estimation of meanimiles of aggregated risks. Metrika. doi: 10.1007/s00184-026-01022-9.

Finally, novel estimators for meanimiles in regression settings are implemented as well.