norm2-package: Analysis of incomplete multivariate data under a normal model
In norm2: Analysis of Incomplete Multivariate Data under a Normal Model

Description Details Note Author(s) References

Functions for estimation and multiple imputation from incomplete multivariate data under a normal model

The norm2 package provides functions for analyzing incomplete multivariate data using techniques and algorithms described by Schafer (1997). The name of this package derives from the assumed model for the complete data, which is a multivariate normal model. The major functions are:

   emNorm        EM algorithm estimating model parameters
   mcmcNorm      MCMC algorithm for simulating parameters and missing values
   impNorm       Simulate or predict missing values
   loglikNorm    Loglikelihood function
   logpostNorm   Log-posterior density function
   miInference   Combine results from analyses after multiple imputation

The package also includes three datasets:

1
2
3

   cholesterol   Cholesterol levels for heart-attack patients
   flas          Foreign Language Attitude Scale
   marijuana     Changes in heart rate after marijuana use

Fortran source code written by the author for a much earlier version called norm was ported to an R package by Alvaro A. Novo and distributed through the Comprehensive R Archive Network (CRAN). The old package norm is still available on CRAN, but it has some major limitations (e.g., it does not work reliably when the number of variables exceeds 30) and the author does not recommend its use.

Joseph L. Schafer <Joseph.L.Schafer@census.gov>

Maintainer: Joseph L. Schafer <Joseph.L.Schafer@census.gov>

Schafer, J.L. (1997) Analysis of Incomplete Multivariate Data. London: Chapman & Hall/CRC Press.

For more information about functions in norm2, see User's Guide for norm2 in the library subdirectory doc.

norm2 documentation built on Feb. 12, 2021, 5:10 p.m.