mi: multitple imputation for proportions

In yuliasidi/bin2mi: Binary Multiple Imputations

Description

employs one/two-stage multiple imputation for proportions

Usage

mi(dt, n_mi, m_mi = 0, mu_k = 1, sd_k = 0, y.m = "y.m",
  phat_out = TRUE)

Arguments

dt	dataframe/tibble
n_mi	positive integer, number of multiply imputed subject level values
m_mi	non-negative integer, Default: 0, number of muttiply imputed values for multitplier k
mu_k	numeric, Default: 1, mean value for normal distribution of multiplier k
sd_k	numeric, Default: 0, standard deviation value for normal distribution of multiplier k
y.m	character, incomplete binary column, Defualt = 'y.m'
phat_out	logic, Default TRUE

Details

DETAILS

Value

tibble, summary per imputation or if phat_out=FALSE, returns imputed data (available only for one-stage imputation)

See Also

tibble map sym unnest

Examples

dt <- tibble::tibble(y = rbinom(100,1,0.6))
dt$y.m <- c(rep(NA, 10), dt$y[11:100])
dt$r <- ifelse(is.na(dt$y.m)==TRUE, 1, 0)
 mi(dt, 5, y.m = 'y.m')
 mi(dt, 2, 10, mu_k = 1.3, sd_k = 0.1, y.m = 'y.m')

yuliasidi/bin2mi documentation built on March 11, 2021, 8:10 p.m.