I am trying to understand what does usertype= do in PROC X13 regression statement ? What happens behind the scenes when i specifie diffrent options ?
For example consider this code:
data SALESDATA;
set SASHELP.AIR;
where date<="01dec1959"d;
format date ddmmyy10.;
run;
proc sql noprint;
select avg(day(intnx('MONTH',date,0, 'e'))) into :lom_avg trimmed
from SASHELP.AIR;;
quit;
%put &=lom_avg;
data REGRESSORS(keep=date lom_full lom_centered );
set SASHELP.AIR;
lom_full = day(intnx('MONTH',date,0, 'e'));
lom_centered=lom_full-&lom_avg;
format date ddmmyy10.;
run;
* option 1;
proc x13 data=SALESDATA date=DATE interval=MONTH plots=none;
transform function=log;
regression PREDEFINED =LOM ;
arima model=((0,1,1)(0,1,1));
x11 trendma = 13 ;
output out=out_predef a1 b1 c1 d1 d10 d11 d12;
run;
* option 2;
proc x13 data=SALESDATA auxdata=REGRESSORS date=DATE interval=MONTH plots=none;
transform function=log;
regression uservar=lom_centered / usertype=lom;
arima model=((0,1,1)(0,1,1));
x11 trendma = 13 ;
output out=out_user_lom_c a1 b1 c1 d1 d10 d11 d12;
run;
* option 3;
proc x13 data=SALESDATA auxdata=REGRESSORS date=DATE interval=MONTH plots=none;
transform function=log;
regression uservar=lom_centered / usertype=user;
arima model=((0,1,1)(0,1,1));
x11 trendma = 13 ;
output out=out_user_usr_c a1 b1 c1 d1 d10 d11 d12;
run;
option 1 and 2 yield the same results (for table D11) but it's diffrent from option 3 (usertype=user), and I'd like to understand what's happening in the background of each option of usertype=user.
Can you, please, share what the usertype=user option does?
Checking the documentation of the the Census Bureau (https://www2.census.gov/software/x-13arima-seats/x-13-data/documentation/docx13as.pdf) revealed nothing too.
