qfmix: Sieve Quantile-Function Mixtures and Distributional Synthetic Controls

Approximates a target quantile function by a weighted mixture of quantile basis functions, with the weights estimated by a sieve generalized method of L-moments (Alvarez and Orestes 2024). L-moment matching reduces to a convex quadratic program over a choice of weight sets (unconstrained, non-negative, simplex, ridge); inference uses the numerical bootstrap. A distributional synthetic-control wrapper builds counterfactual quantile functions with confidence bands.

Main entry points

• qfmix() – fit a quantile-function mixture.

• qfmix_boot() – numerical-bootstrap inference (weights / quantiles).

• dsc() – distributional synthetic control.

• qfmix_basis() – construct a quantile basis.

References

Alvarez, L. A. F. and Orestes, V. M. (2024). Quantile Mixture Models: Estimation and Inference. Working paper.

Hosking, J. R. M. (1990). L-moments: analysis and estimation of distributions using linear combinations of order statistics. JRSS-B 52, 105–124.

Gunsilius, F. F. (2023). Distributional synthetic controls. Econometrica 91, 1105–1117.

See also

Useful links:

• https://github.com/kvenkita/qfmix

• Report bugs at https://github.com/kvenkita/qfmix/issues

Author

Maintainer: Kailas Venkitasubramanian kailasv@gmail.com [copyright holder]