I woul like to clarify some points:   - is the look_up table named  work.datekey ? - is the date you want to check is on that work.datekey table and is named DATE ? - is the transaction table named as work.payments ?   Now back to my oroginal proposed code - addapted to the above:   proc sql;     create table want as select       a.*, b.*               /* change to selet desired variables,  the asterisk (*) is to select all variables */      from work.DateKey (where=(date='03JAN2016'd)) as a      left  join work.payments as b      on (a.date_key = b.datekey)   ; quit;   In case it does not meet your expectations please clarify where and why ?   The error you got: ERROR: Column Date could not be found in the table/view identified with the correlation name B. is because the DATE variable is not in work.payments table (as B)  but in work.DateKey table (as A)   Don't hesitate to ask more if need. ... View more

try:   proc sql;     create table want as select       a.*, b.*   /* change to desired variables */      from lookup_table(where=(date='03JAN2016'd)) as a      left  join transaction_table as b      on (a.date_key = b.datekey)   ; quit;   Is that what you want ? if not addapt it to your needs. ... View more

You can concatenate the two tables before exporting:   data both;   set  ae  cm; run; proc export data=both; OUTFILE= "c:\ae_and_cm.csv" DBMS=CSV REPLACE; PUTNAMES=Yes; RUN;   Assuming both tables have same variables, otherwise it does not make sense to me. ... View more

When you create a new SAS dataset from another SAS table, all attributes of the varibales are inherited, including length, labels, formats, type etc. - from the input table ... View more

What OS and what SAS version do you use ?   Try change a line to:    %str(proc sql;);       instead: proc sql;   I have tried using %do ... %end in a proc sql, similar to your code, and had no ERROR like your ... View more

Do I understand right:   if TRANDATE was in a SAS date (numeric) format then you will use :  Where = (Trandate >= '01Jan2016'd);   As TRANDATE is a character variable you can use:   where = (input(trandate,yymmdd10.) >= '01JAN2016'd);     ... View more

I know it from MF environment. ... View more

That's right - WANT is rhe output dataset with the new variable CONTINENT; ... View more