Package: SDModels Title: Spectrally Deconfounded Models Version: 2.0.2 Authors@R: c( person("Markus", "Ulmer", email = "markus.ulmer@stat.math.ethz.ch", role = c("aut", "cre", "cph"), comment = c(ORCID = "0000-0001-7783-8475")), person("Cyrill", "Scheidegger", role = "aut", comment = c(ORCID = "0009-0005-2851-1384"))) Description: Screen for and analyze non-linear sparse direct effects in the presence of unobserved confounding using the spectral deconfounding techniques (Ćevid, Bühlmann, and Meinshausen (2020), Guo, Ćevid, and Bühlmann (2022) ). These methods have been shown to be a good estimate for the true direct effect if we observe many covariates, e.g., high-dimensional settings, and we have fairly dense confounding. Even if the assumptions are violated, it seems like there is not much to lose, and the deconfounded models will, in general, estimate a function closer to the true one than classical least squares optimization. 'SDModels' provides functions SDAM() for Spectrally Deconfounded Additive Models (Scheidegger, Guo, and Bühlmann (2025) ) and SDForest() for Spectrally Deconfounded Random Forests (Ulmer, Scheidegger, and Bühlmann (2025) ). License: GPL-3 Imports: DiagrammeR, future.apply, future, ggplot2, igraph, ggraph, gridExtra, Rdpack, tidyr, fda, grplasso, rlang, progressr Suggests: parallelly, plotly, datasets, rpart, knitr, rmarkdown, ranger, HDclassif, qpdf, cli, testthat (>= 3.0.0) RdMacros: Rdpack Encoding: UTF-8 RoxygenNote: 7.3.3 VignetteBuilder: knitr URL: https://www.markus-ulmer.ch/SDModels/ BugReports: https://github.com/markusul/SDModels/issues Config/testthat/edition: 3 Config/pak/sysreqs: cmake libfontconfig1-dev libfreetype6-dev libglpk-dev make libicu-dev libuv1-dev libxml2-dev libx11-dev Repository: https://markusul.r-universe.dev Date/Publication: 2025-12-12 16:40:09 UTC RemoteUrl: https://github.com/markusul/sdmodels RemoteRef: HEAD RemoteSha: c1dd77fba4993f957620093e052f91616cb7f98e NeedsCompilation: no Packaged: 2026-10-08 09:04:34 UTC; root Author: Markus Ulmer [aut, cre, cph] (ORCID: ), Cyrill Scheidegger [aut] (ORCID: ) Maintainer: Markus Ulmer