voi: a generic package to calculate the expected value of information

• A common interface for several methods to calculate the

  • Expected Value of (Partial) Perfect Information and the

  • Expected Value of Sample Information

• A project of the ConVOI Group: the Collaborative Network for Value of Information

Comparison with other packages

voi is pure “command-based” R, with no web interface like SAVI.

• The R commands in voi are clean and consistent: they all have the same basic interface, so you can switch between computational methods easily.

• Outputs are all in “tidy” data frames for consistency, and to facilitate post-processing and plotting with modern tools such as ggplot2.

EVPPI computation

• voi includes all the EVPPI computation methods that are in SAVI (GAM and Gaussian process regression), and includes the INLA method from BCEA.

• Some other nonparametric regression methods ("earth", "bart") are included in voi, which may improve efficiency for multiparameter EVPPI computation problems with large numbers of parameters.

EVSI computation

• voi is the first package to implement a range of EVSI computation methods: nonparametric regression, moment matching and importance sampling. A simple model for the expected net benefit of sampling is also included.

In summary

• voi will not benefit you if you want a web interface, or if you just need single-parameter EVPPI and are happy with SAVI/BCEA.

• voi will benefit you if you want to calculate EVSI, or multiparameter EVPPI with large numbers of parameters.

Installation

Stable version

install.packages("voi")

Development version

remotes::install_github("chjackson/voi")

User guide

voi for Value of Information calculation: package overview

Source code

Github repository