SAS macros %MISSING and %MISSDESCRIPTION for diagnosing and replacing missing values

Download the SAS macros

On this page you can download two SAS macros for diagnosing and replacing missing values in data sets. The macros were created during the conception of a SAS macro package for the development and validation of forecast models based on logistic regression [3], in which they are available in a limited version of the current macros. They were modified and extended as part of a diploma thesis [1] and further research work [2]. Both macros are available for SAS version 9.

• %MISSING (zip): Macro for performing single and multiple imputation based on different methods
• %MISSDESCRIPTION (zip): Macro for describing a data set with regard to missing values - output of the missing data pattern and (usual) description of all variables

The required sub-macros are downloaded with both links. The detailed documentation within the macro programmes should make it easy to use them.

 

Literature

[1]	Brodrecht K.: Implementation of different replacement methods for missing values in SAS. Diploma thesis at the Ulm University of Applied Sciences, Medicine Documentation and Computer Science (2005)
[2]	Hohl K., Muche R., Ring C., Ziegler C.: Missing values in the (regression) analysis of data sets: two SAS macros. 9th Conference of SAS Users in Research and Development, Shaker Verlag, Aachen (2005)
[3]	Muche R., Ring C., Ziegler C.: Development and validation of forecasting models based on logistic regression. Shaker Verlag Aachen (2005)