Distributed Lag Interaction Model

Collection of functions for fitting and interpreting distributed lag interaction models (DLIM). A DLIM regresses a scalar outcome on repeated measures of exposure and allows for modification by a continuous variable. Includes a dlim() function for fitting, predict() function for inference, and plotting functions for visualization. Details on methodology are described in Demateis et al. (2024) .

dlim

Overview

This repository contains the source files for the dlim R package that implements distributed lag interaction models as described in the paper:

• Demateis D, Keller KP, Rojas-Rueda D, Kioumourtzoglou M-A, Wilson A (2024). “Penalized Distributed Lag Interaction Model: Air Pollution, Birth Weight and Neighborhood Vulnerability.” Environmetrics. (DOI: 10.1002/env.2843, arXiv preprint)

Simulation scripts from the manuscript can be found at https://github.com/ddemateis/dlim_simulations.

The dlim package estimates a distributed lag model with modification by a single continuous factor. If you are instead interested in a DLM with modification by a single categorical or binary variable see the bdlim package. If you are interested in distributed lag models with heterogeneity by with multiple modifiers see the heterogeneous distributed lag model in the dlmtree package.

Installation

Install from CRAN:

install.packages("dlim")

Install from GitHub:

remotes::install_github("ddemateis/dlim")

A vignette is available on the package website or in the package using the following code.

vignette('dlimOverview')