pgf

pgf is an R implementation of the parametric g formula, first developed by Robins in the mid 1980s. The purpose of this program is to simulate the follow-up of a cohort study with complex longitudinal data under different exposure regimes. pgf accomplishes this in two steps:

• Fit empirical models to the observed complex longitudinal dataset.

• Simulate follow-up from each model, but setting the exposure/censoring values to desired levels.

You can install the latest development version from github with

```R
install.packages("devtools")
devtools::install_github("ainaimi/pgf")
```

If you encounter a bug or other issue, please file a request on github.

Learning pgf

A vignette for the pgf package is currently in preparation. To learn about the parametric g formula more generally, please refer to the following citations:

• Naimi et al (2016)
• Daniel et al (2013)
• Robins and Hernan (2009)

This work is supported by NIH Grant R01 HD093602.