mi (version 0.10-2)

mi.binary: Elementary function: Bayesian logistic regression to impute a binary variable.

Description

Imputes univariate missing data using bayesglm, an R functions for generalized linear modeling with independent normal, t, or Cauchy prior distribution for the coefficients.

Usage

mi.binary(formula, data = NULL, start = NULL, maxit = 100,
  draw.from.beta = TRUE, missing.index = NULL, ...)
## S3 method for class 'mi.binary':
resid(object, y)
## S3 method for class 'mi.binary':
residuals(object, y)
## S3 method for class 'mi.binary,ANY':
plot( x, y, main=deparse( substitute( y ) ), gray.scale = FALSE, ...)

Arguments

formula

an object of class '"formula"' (or one that can be coerced to that class): a symbolic description of the model to be fitted. See bayesglm 'formula' for details.

data

A data frame containing the incomplete data and the matrix of the complete predictors.

start

Starting value for bayesglm.

maxit

Maximum number of iteration for bayesglm. The default is 100.

draw.from.beta

Draws from posterior distribution of the betas to add randomness.

missing.index

The index of missing units of the outcome variable.

...

Currently not used.

object

mi.binary object.

x

mi.binary object.

y

Observed values.

main

main title of the plot.

gray.scale

When set to TRUE, makes the plot into gray scale with predefined color and line type.

Value

• model
  A summary of the bayesian fitted model.
• expected
  The expected values estimated by the model.
• random
  Vector of length n.mis of random predicted values predicted by using the binomial distribution.

Details

In bayesglm default the prior distribution is Cauchy with center 0 and scale 2.5 for all coefficients (except for the intercept, which has a prior scale of 10). See also glm for other details.

References

Andrew Gelman and Jennifer Hill, Data Analysis Using Regression and Multilevel/Hierarchical Models, Cambridge University Press, 2007.

See Also

mi.info, mi.method, mi

Examples

# true data
  x <- rnorm(100,0,1) # N(0,1)
  y <- rbinom(100,1,invlogit(1+2*x)) # y ~ Bin(1,invlogit(1 + 2*x)
  # create artificial missingness on y
  y[seq(1,100,10)]<-NA
  dat.xy <- data.frame(x,y)
  # imputation
  mi.binary(y~x, data = dat.xy)