Class "bounded-continuous"
The bounded-continuous class inherits from the continuous-class and is intended for variables whose observations
fall within open intervals that have known boundaries. Although proportions satisfy this definition, the
proportion-class should be used in that case. At the moment, a bounded continuous variable is modeled as if it were
simply a continuous variable, but its mi-methods impute the missing values from a truncated normal distribution using
the rtruncnorm function in the truncnorm package. Note that the default transformation is the identity
so if another transformation is used, the bounds must be specified on the transformed data. Aside from these facts, the rest of the
documentation here is primarily directed toward developers.
Objects can be created that are of bounded-continuous class via the
the missing_variable generic function by specifying type = "bounded-continuous"
as well as lower and / or upper
The bounded-continuous class inherits from the continuous class and is intended for variables that are supported on a known
interval. Its default transformation function is the identity transformation and its imputation_method must be
"ppd". It has two additional slots:
a numeric vector whose length is either one or the value of the n_total slot giving the upper bound for
every observation; NAs are not allowed
a numeric vector whose length is either one or the value of the n_total slot giving the lower bound for
every observation; NAs are not allowed
Ben Goodrich and Jonathan Kropko, for this version, based on earlier versions written by Yu-Sung Su, Masanao Yajima, Maria Grazia Pittau, Jennifer Hill, and Andrew Gelman.
# STEP 0: GET DATA
data(CHAIN, package = "mi")
# STEP 0.5 CREATE A missing_variable (you never need to actually do this)
lo_bound <- 0
hi_bound <- rep(Inf, nrow(CHAIN))
hi_bound[CHAIN$log_virus == 0] <- 6
log_virus <- missing_variable(ifelse(CHAIN$log_virus == 0, NA, CHAIN$log_virus),
type = "bounded-continuous", lower = lo_bound, upper = hi_bound)
show(log_virus)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.