mi.mixed {mi}R Documentation

Two-stage elementary function: linear regression to impute a variable that can be positive or zero.

Description

Imputes univariate missing data using a two stage imputation process. The first regression model is fit to all the data for which the variable is observed, and the second model is fit to all the data for which the variable is observed and positive.

Usage

  mi.mixed(formula, data = NULL, start = NULL, n.iter = 100, draw.from.beta = FALSE, ...)

  ## S3 method for class 'mi.mixed':
  coef(object, ...)
  ## S4 method for signature 'mi.mixed':
  sigma.hat(object, ...)
  ## S3 method for class 'mi.mixed':
  fitted(object, ...)
  ## S3 method for class 'mi.mixed':
  residuals(object, y, ...)
  ## S4 method for signature 'mi.mixed':
  imputed(object, y)
  ## S3 method for class 'mi.mixed':
  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.
n.iter Maximum number of iteration for bayesglm. The default is 100.
draw.from.beta Draws from posterior distribution of the betas to add randomness.
... Currently not used.
object mi.mixed object.
x mi.mixed 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.

Details

Imputation variable that can be positive or zero. See Gelman and Hill, Data Analysis Using Regression and Multilevel/Hierarchical Models, (2206), Chapter 25.

Value

model$model.1 A summary of the fitted model for the all the data for which y is observed.
model$model.2 A summary of the fitted model for the all the data for which y is observed and positive.
expected The expected values estimated by the model
random Vector of length n.mis of random predicted values to impute the missing data. Whether missing values are positive random is positive, otherwise is equal to zero.

Author(s)

Masanao Yajima yajima@stat.columbia.edu, M.Grazia Pittau grazia@stat.columbia.edu, Andrew Gelman gelman@stat.columbia.edu

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<-rnorm(100,(1+2*x),1.2) # y ~ 1 + 2*x + N(0,1.2)
  y[y<0]<-0
  # create artificial missingness on y
  y[seq(1,100,10)]<-NA
  dat.xy <- data.frame(x,y)
  # imputation
  mi.mixed(list("1*(y>0)~x","y~x"), data = dat.xy)

[Package mi version 0.02-02 Index]