A package is introduced that provides the weighted smooth backfitting estimator for a large family of popular semiparametric regression models. This family is known as generalized structured models, comprising, for example, generalized varying coefficient model, generalized additive models, mixtures, potentially including parametric parts. The kernel-based weighted smooth backfitting belongs to the statistically most efficient procedures for this model class. Its asymptotic properties are well-understood thanks to the large body of literature about this estimator. The introduced weights allow for the inclusion of sampling weights, trimming, and efficient estimation under heteroscedasticity. Further options facilitate easy handling of aggregated data, prediction, and the presentation of estimation results. Cross-validation methods are provided which can be used for model and bandwidth selection.[^1]
Supplementary materials are available in addition to this article. It can be downloaded at RJ-2021-042.zip
Text and figures are licensed under Creative Commons Attribution CC BY 4.0. The figures that have been reused from other sources don't fall under this license and can be recognized by a note in their caption: "Figure from ...".
For attribution, please cite this work as
Roca-Pardiñas, et al., "The R Journal: Package wsbackfit for Smooth Backfitting Estimation of Generalized Structured Models", The R Journal, 2021
BibTeX citation
@article{RJ-2021-042,
author = {Roca-Pardiñas, Javier and Rodríguez-Álvarez, María Xosé and Sperlich, Stefan},
title = {The R Journal: Package wsbackfit for Smooth Backfitting Estimation of Generalized Structured Models},
journal = {The R Journal},
year = {2021},
note = {https://doi.org/10.32614/RJ-2021-042},
doi = {10.32614/RJ-2021-042},
volume = {13},
issue = {1},
issn = {2073-4859},
pages = {314-334}
}