An unknown prior density *g*(*θ*) has yielded (unobservable) *Θ*_{1}, *Θ*_{2}, …, *Θ*_{N}, and each *Θ*_{i} produces an observation *X*_{i} from an exponential family. `deconvolveR`

is an R package for estimating prior distribution *g*(*θ*) from the data using Empirical Bayes deconvolution.

The current package is still under construction but will soon appear on CRAN along with a manuscript. Meanwhile, you can reproduce many examples by installing the package in R thus:

```
devtools::install_github("bnaras/deconvolveR")
library(deconvolveR)
vignette("deconvolution")
```