DESCRIPTION

Package: vibr
Title: Variable Importance in Black-Box Regression
Version: 1.0.5
Date: 1-sep-2023
Authors@R: 
    person(given = "Alexander",
           family = "Keil",
           role = c("aut", "cre"),
           email = "alex.keil@nih.gov",
           comment = c(ORCID = "0000-0002-0955-6107"))
Author: Alexander Keil [aut, cre]
Maintainer: Alexander Keil <alex.keil@nih.gov>
Description: This package implements two variable importance estimators for 
    prediction of binary or continuous target variables. Namely, based on a 
    black-box prediction/regression model fit to the data, this approach yields 
    g-computation and augmented inverse probability weighted estimates of 
    variable importance, both described by Diaz et al (2014), and the targeted 
    maximum likelihood approach described in Diaz and van der Laan (2018). 
    This package relies on black-box prediction from the \code{sl3} package, 
    which, among other things, uses super learning (stacking) to make 
    predictions.
    References: 
    1) Keil, A.P. and O'Brien, K.M. Stat Bioscience (In revision)
    2) Diaz, I., Hubbard, A., Decker, A., & Cohen, M. (2015). 
    Variable importance and prediction methods for longitudinal problems with missing variables. 
    <https://doi.org/10.1371/journal.pone.0120031>
    3) Diaz, I. and van der Laan, M. Stochastic Treatment Regimes. In Targeted Learning in Data Science: 
    Causal Inference for Complex Longitudinal Studies by Mark van der Laan and Sherri Rose.
Depends: R (>= 3.5), sl3
Imports: mvtnorm, qgcomp, Rsolnp, R6, data.table, future, MASS, polspline, assertthat, ggplot2, devtools
Suggests: knitr, markdown
Remotes: github::tlverse/sl3
VignetteBuilder: knitr
License: GPL (>= 3)
Encoding: UTF-8
LazyData: true
Roxygen: list(markdown = TRUE)
RoxygenNote: 7.2.3