topic Re: Regression Imputation in SAS Procedures

Regression Imputation

Reg_symput — Sun, 26 Apr 2009 20:26:37 GMT

Hi guys,

I'm trying to find out about regression imputation using CALL SYMPUT macro.

Does anyone has a simple example to show, I check the web, but couldn't find a useful one.

I've data set missing values on a market research with variables car sales data,
Ford, GM, Honda sales data are missing more than 40% I've peoples income levels and spending avg complete data, but I don't know how to impute with regression imputation.

Could you please give a simple example?

Thanks,

Re: Regression Imputation

sbb — Sun, 26 Apr 2009 22:11:08 GMT

What is expected with the CALL SYMPUT specifically? If you are looking to input data from an external file to combine with your other SAS dataset(s), the SAS conference paper (link below) demonstrates using the DATA step and the INPUT statement to build a SAS file from a sequential input.

Otherwise, if you have some specific INPUT and OUTPUT sample conditions to offer, beyond what can be accomplished with the processes explained in the referenced paper, please re-post with some explanation of the SAS macro variable purpose (where you mention CALL SYMPUT).

Scott Barry
SBBWorks, Inc.

Hands-On Introduction to SAS® DATA Step Programming
Debbie Buck, D. B. & P. Associates, Houston, TX

http://www2.sas.com/proceedings/sugi30/134-30.pdf

Re: Regression Imputation

statsplank — Mon, 27 Apr 2009 18:10:26 GMT

Hi Reg_symput,

I think the following may be of help to you.

There is PROC MI to do imputation of missing data:

http://support.sas.com/documentation/cdl/en/statug/59654/HTML/default/statug_mi_sect001.htm

Example 54.3 Regression Method shows how to impute missing values using the regression method:

http://support.sas.com/documentation/cdl/en/statug/59654/HTML/default/statug_mi_sect040.htm