The goal of tobit.mean is to calculate descriptive statistics of censored vectors (for example, concentrations below a limit of quantification). The arithmetic mean, geometric mean, standard deviation, and coefficient of variation are estimated by maximum likelihood with the Tobit model, so censored values are accounted for rather than dropped or replaced.
You can install the development version of tobit.mean from GitHub with:
# install.packages("pak")
pak::pak("humanpred/tobit.mean")Calculate the mean of a vector where values at or below 0.5 are censored:
library(tobit.mean)
tobit.mean::mean_tobit(c(0, 1, 4, 10), lower = 0.5, distribution = "t")