The goal of testerror is to help produce modelled prevalence estimates for diseases based on positive test results in the inevitable presence of test error. This is particularly problematic when test results are based on the combination of many test results from a multiplex panel, in which case error in component tests compound to cause a great deal of uncertainty and in some situations significant bias.
Installation
The testerror package is hosted in the Bristol Vaccine Centre r-universe. Installation from there is as follows:
options(repos = c(
"bristol-vaccine-centre" = 'https://https://bristol-vaccine-centre.r-universe.dev/',
CRAN = 'https://cloud.r-project.org'))
# Download and install tableone in R
install.packages("testerror")You can install the development version of testerror from GitHub with:
# install.packages("devtools")
devtools::install_github("bristol-vaccine-centre/testerror")Further information
For examples and usage head to the main documentation website. An academic paper is in preparation which details the methodology implemented here.