How to use proc genmode with a large sample size?

[Question]

Hi,

I would like to us proc genmode with repeated measures. but with a good sample size (1,000,000 id )?

How does it work? When I put subject=Customer_id, it doesn't seem to work as I guess because my sample size is big?

proc genmod data=work.modelling_dataset;

class id;

model target=no_trans week;

repeated subject = id / type=exch corrw; run; quit;

Thank You

[Ksharp]

1) try PROC HPMIXED 

2) try PROC GEE 

[Question]

Thank You Ksharp for your response!

How do I score a new dataset using the results of the model below. Thank you


proc hpmixed data=work.MODELLIND_DATASET ;
class customer_id P_Band Region;
model Target =P_Band no_transacted Region/ s;
random customer_id / s;
output out=work.P_pred pred=p resid=r;

run;

##- Please type your reply above this line. Simple formatting, no
attachments. -##

[Ksharp]

One thing I could think is appending the TEST data at the end of TRAIN data

( target variable should be missing in your TEST data).

proc append base=MODELLIND_DATASET data=TEST force;run;

proc hpmixed data=MODELLIND_DATASET;

..............

output ......

And use OUTPUT to score the TEST data.

And hope you can get the predicted value in output dataset.