03-08-2016 09:21 PM - last edited on 03-08-2016 09:31 PM by Reeza
When I used the proc surveylogistic in SAS, the number of observations deleted is larger than the number of obervations with missing values, so I want to print the excluded observations to check where is wrong with the data. I searched for several hours but did not find how to do that. Could anyone help me?
03-08-2016 09:30 PM
I don't think there's an easy way, but you could relatively easy move that into a proc print.
proc print data=have; where nmiss(numeric variable list)>0 or cmiss(character variable list)>0; var numeric_variable_list character_variable_list; run;
03-08-2016 10:15 PM
Thanks! I just found the reason is that I put more variables in the class statement than I put in the model statement, and SAS will remove all the observations with missing values in the class or model statement. That is why the number of observations deleted due to missing values is larger than the number of variables with missing values in the model.....