xactonomial

The goal of xactonomial is to use an exact (but computational and stochastic) method to compute a confidence interval and a function for calculation of p values in the k sample multinomial setting where interest is about a real-valued functional of the multinomial probabilities.

Installation

You can install the development version of xactonomial like so:

remotes::install_github("sachsmc/xactonomial")

Example

This is a basic example which shows you how to use the main function:

library(xactonomial)

psi_ba <- function(theta) {
   theta1 <- theta[1:4]
   theta2 <- theta[5:8]
   sum(sqrt(theta1 * theta2))
   }

data <- list(T1 = c(2,1,2,1), T2 = c(0,1,3,3))
xactonomial(psi_ba, data, psi_limits = c(0, 1), maxit = 5, chunksize = 20)