Hi Everybody, I'm trying to do a report showing forecast for 2 different provinces, 3 different markets, data, including Results from 2014 and 2015, with Sale_Year and Sale_Month. the OUT gives something similar to this. PROVINCE MRKT PRODUCT Sale_Year Sale_Month COUNT_of_sale_dt ON CORP. STORES HP Comp 2014 1 588 ON CORP. STORES HP Comp 2014 2 492 ON CORP. STORES HP Comp 2014 3 616 ON CORP. STORES HP Comp 2014 4 643 ON CORP. STORES HP Comp 2014 5 565 ON CORP. STORES HP Comp 2014 6 599 ON CORP. STORES HP Comp 2014 7 759 ON CORP. STORES HP Comp 2014 8 802 here is the coding I'm using for the moment: PROC SQL;   CREATE TABLE work.QUERY_FOR_FILTER_FOR_TEST01 AS   SELECT t1.PROVINCE,           t1.MRKT,             t1.Sale_Year,           t1.Sale_Month,   /* COUNT_of_sale_dt */             (COUNT(t1.sale_dt)) AS COUNT_of_sale_dt       FROM SASUSER.FILTER_FOR_TEST01_SALES_COH_0000 t1       GROUP BY t1.PROVINCE,               t1.MRKT_L2,                 t1.Sale_Year,               t1.Sale_Month; QUIT; proc forecast data=SASUSER.QUERY_FOR_FILTER_FOR_TEST01 lead=12 out=prediction; var Count_of_Sale_dt; run; I obtain this result attached. How can I add coding which will give me the forecast per province/Market and Product as appearing bellow? Anybody? thanks. ... View more

Here is the coding that I need to use: PROC SQL;    CREATE TABLE WORK.FILTER_FOR_XXXX AS SELECT t1.sale_dt,           t1.DUE_DT,           t1.TOWN,           t1.POSTAL_CD,           t1.PROV_CD,           t1.COMPETITION,           t1.AGENT_CD,           t1.STORE,           t1.MRKT,           t1.AGENT_ID,      FROM WORK.FILTER_FOR_XXXX       WHERE TOWN = 'NY' AND t1.sale_dt > 19724       ORDER BY t1.sale_dt; QUIT; I have sale dates in variable t1.sale_dt appearing as for example 11JAN2014 (in French) etc.. I want to add to more variables one for the Month January (or 1) and the second for the Year 2014. Which exact coding should I add to have it please? thank you. ... View more