Distributed Lag Linear and Non-Linear Models in R: The Package dlnm

By Antonio Gasparrini. Journal of Statistical Software, 2011

763 citations recorded by the source corpus.

Distributed lag non-linear models (DLNMs) represent a modeling framework to flexibly describe associations showing potentially non-linear and delayed effects in time series data. This methodology rests on the definition of a crossbasis, a bi-dimensional functional space expressed by the combination of two sets of basis functions, which specify the relationships in the dimensions of predictor and lags, respectively. This framework is implemented in the R package dlnm, which provides functions to perform the broad range of models within the DLNM family and then to help interpret the results, with an emphasis on graphical representation. This paper offers an overview of the capabilities of the package, describing the conceptual and practical steps to specify and interpret DLNMs with an example of application to real data.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/0f3283467a68bb835a3b30679d57acced10d08d8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Ben Armstrong
2. Nicholas A. S. Hamm
3. Antonio Gasparrini

Citations in the corpus

• Multivariate meta-analysis for non-linear and other multi-parameter associations (2012, neutral): “0) [74] is used to specify the basis matrices for the quadratic spline for temperature and to predict and plot the results of a fitted model.”
• A machine learning method to estimate PM2.5 concentrations across China with remote sensing, meteorological and land use information. (2018, methodology): “…(Chen et al., 2018), where cb_TEMP , cb_RH , cb_BP and cb_WS are mean temperature, relative humidity, barometric pressure and wind speed on the current day and previous two days (lag 0-2 days) fitted using crossbasis() function of DLNM with 3 df (Gasparrini, 2011; Gasparrini, 2014), respectively.”
• Modeling exposure–lag–response associations with distributed lag non-linear models (2013, neutral): “The DLNM modeling framework is fully implemented in the package dlnm [25], by using the expressly extended version 2.”
• Reducing and meta-analysing estimates from distributed lag non-linear models (2013, methodology): “The software implementation of this methodology in the R package dlnm has been previously described [15].”