mi.info: Function to create information matrix for missing data imputation

Description

Produces matrix of information needed to impute the missing data. After the information is extracted user has the option of changing the default.

Usage

mi.info(data, threshold  = 0.99999)
  ## S3 method for class 'mi.info':
print(x, \dots)
  ## S3 method for class 'mi.info':
show(object)

Arguments

data

dataframe or matrix of dataset with missing data coded as NAs.

threshold

Threshold value for correlation to be considered a problem.

An object of a class mi.info.

object

An object of a class mi.info.

...

Currently not used.

Value

infoinformation matrix llll{ -name: Name of variable -imp.order: Imputation Order -nmis: Number of missing -type: Type of variable -var.class: Class of input variable -level: Levels in the input varialbe -include: Include in the imputation process or not -is.ID: Is ID variable or not -all.missing: All observation missing or not -collinear: Collineared variables -determ.pred: Deterministic predictor -imp.formula: Imputation formula -params: Parameters for the imputation model -other: Currently not used }

References

Yu-Sung Su, Andrew Gelman, Jennifer Hill, Masanao Yajima. (2011). Multiple Imputation with Diagnostics (mi) in R: Opening Windows into the Black Box. Journal of Statistical Software 45(2).

Examples

Run this code

data(CHAIN)
  info.CHAIN <- mi.info(CHAIN)
  
  info.CHAIN$imp.order # imputation order
  
  info.CHAIN$imp.formula # imputation formula
  info.CHAIN[["age.W1"]]$imp.formula  #imputation formula for specific variable