SAS Support Communities

29 rename AGE_85_+_PCT=age_85_plus; ERROR: Missing numeric suffix on a numbered variable list (AGE_85_-_PCT). This code is not what was provided to you as a solution. You have to enclose the variable name inside     '  'n     as shown.       ... View more

You can read the header line and check. Say your input file is NEWCSV and expected structure is described by the SAS datasets TEMPLATE. * Read headers from new delimited file ; data new_names ; length name $50 ; infile newcsv obs=1 dsd dlm='09'x ; input name @@ ; run; * Get names from template structure ; proc transpose data=template(obs=0) out=old_names; var _all_; run; * Compare; proc compare data=old_names compare=new_names; run; PROC COMPARE will actually set a macro variable you can test to see if there are differences.  Or you can use some other methods to compare. ... View more

Hi , Thanks for infoo ,I have read that post but my issue still remains the same    column datatype in teradata is timestamp(0) and when i tried to cull data from TD to sas via pass through sql as   proc sql;  connect to teradata (user=&user password=&pwd mode=teradata); create table mysas.data1 as select * from connection to teradata(   /**********approaches i used so far****/   select employee_id , registration_date format =timestamp19.); fail select employee_id , cast(registration_date as mm-dd-yyyy hh:mm:ss.); fail  disconnect to teradata;  quit;   could you help me how to get this thing into SAS ... View more

Your first code is SAS code.    Your second set of code is Teradata code, use SQL Pass Through - code is literally passed to Teradata server.    When end using a libname connection in EG you need to use SAS code.  ... View more

Then we need to add an additional flag for the processing of each subscriber_num: data have; length yyyymm $6; input subscriber_num yyyymm id_flg; cards; 101 201512 1 101 201511 23 101 201510 223 102 201512 . 102 201511 1 102 201501 54 103 201512 . 103 201511 . 103 201510 1 ; run; data want (keep=subscriber_num var_m1_m4); set have; by subscriber_num; retain flag; if first.subscriber_num then flag = 1; if id_flg > 0 and flag then do; var_m1_m4 = intck('month',input(yyyymm!!'01',yymmdd8.),'30jan2016'd); if 1 <= var_m1_m4 <= 4 then do; output; flag = 0; end; end; run; ... View more

It would be helpful if you also attached the huang_bal_04a dataset. ... View more

I would use a fast export. ... View more

Yes RW9@ & KurtBremser... i agree there are multiple ways but couldn't bring them that time so used that.. actually i used earlier select a,b from (select id ,cat from T 1 where some condition) aliasname....as i forgot to put this alias name so it makes me to put question  here on discussion portal ...but thanks for the effrort you guys put ..i appreciate it .thanks.. ... View more

got it by my own ....thanks for your effort ... View more