Proc MI variables specification problem

Haemoglobin17 — Mon, 19 Feb 2024 09:16:38 GMT

Hi Everyone,

I'm trying to impute 3 variables in my dataset: Country and Smoke (both binary variables) and bmi (categorical with 4 levels).

This is my code:

proc mi data= Have nimpute=5 out=Want seed=54321;
class Age AR Group bmi Country Smoke;
var Age AR Group bmi Country Smoke;
fcs logistic(Country= Age AR Group );
fcs logistic(Smoke=Age AR Group Country);
fcs discrim(bmi=Age AR Group Country Smoke/CLASSEFFECTS=INCLUDE);
run;

For some reason, it gives me error. Here's the log:

proc mi data= Have nimpute=5 out=Want seed=54321;
3307 class Age AR Group bmi Country Smoke;
3308 var Age AR Group bmi Country Smoke;
3309 fcs logistic(Country= Age AR Group );
3310 fcs logistic(Smoke=Age AR Group Country);
3311 fcs discrim(bmi=Ag AR Group Country Smoke/CLASSEFFECTS=INCLUDE);
3312 run;

WARNING: The covariates are not specified in an FCS discriminant method for variable Group,
only remaining continuous variables will be used as covariates with the default
CLASSEFFECTS=EXCLUDE option.
ERROR: There are no continuous variables in the VAR list to impute the variable Group with
an FCS discriminant method and the default CLASSEFFECTS=EXCLUDE option.
NOTE: The SAS System stopped processing this step because of errors.
WARNING: The data set WORK.MI_MVN may be incomplete. When this step was stopped there were 0
observations and 94 variables.
WARNING: Data set WORK.Want was not replaced because this step was stopped.
NOTE: PROCEDURE MI used (Total process time):
real time 9.00 seconds
cpu time 1.67 seconds

Important, I don't want to impute Group, because it doesn't have missing. I just want to use it to compute the 3 variables above. Do you have any idea on how to solve this problem? Thank you in advance 😉

topic Proc MI variables specification problem in Statistical Procedures

Proc MI variables specification problem