Sorry. It is papersize= , I mixed up with pagesize=. options papersize=(10in 40in); Papersize= can enable you to define as wide paper as you want . Which means you can hold all these variables in a row . Xia Keshan ... View more

Hi Kannan, One way is to have another macro test if a data set exists, and if so, then count the number of observations. Then, if the conditions are met, call the macro that does the PROC. I've taken your data and simplified it a bit (using PROC MEANS, and creating a 0 obs data set).... Data merchant_1;   Input Item $ sales_amount;   Datalines;   shoe 50   fan  100   TV   500   SOFA 1000   ; run;   Data merchant_999;   Input item $ sales_amount;   Datalines;   LOCK 50   PC   100   MUG   500   LAPTOP 1000   ; run; /* create data set with 0 obs...*/ data MERCHANT_2;  set MERCHANT_1; if sales_amount < 1; run; %macro TAB; proc means data=MERCHANT_&seq ; var sales_amount; title1 "STATISTICS from data set MERCHANT_&seq"; run; %mend TAB; %macro REPORT(seq); %if %sysfunc(exist(MERCHANT_&seq)) %then %do; %let dsid = %sysfunc(open(MERCHANT_&seq));  *open the data set ; %let NumObs = %sysfunc(attrn(&dsid,nlobs)); *count number of logical observations ; %let RC = %sysfunc(close(&Dsid));   %IF &NUMOBS > 0 %then     %TAB;   %ELSE    %PUT "DATA SET MERCHANT_&seq HAS 0 OBS!!"; %end; %ELSE %put "DATA SET MERCHANT_&seq DOES NOT EXIST!!"; %mend REPORT; %report(1) %report(2) %report(3) %report(999) Hope this helps.. ... View more

TBH my first question would be why the need for 1540 Excel files with x number of observations.  ballardw's idea of using .xls (which is M$ old binary proprietary format) will indeed be a lot smaller.  However thats still a large amount of Excel files.  Is this some kind of data transfer or a review output?  If its data transfer then there are better methods (using CSV will drastically shrink your file size also and still be openable in Excel - though again without the formatting).  If its output though what you could do is go down the route of CSV, i.e. small data file, then have an Excel document with some built in VBA functions which load and present the data. The only other suggestion I can offer is to build the XLSX file yourself.  Within the XLSX (rename it zip and have a look), is a folder structure and some files.  One of the files relates to unique values.  This is their attempt to minimise the size.  So you could build that yourself. ... View more

You didn't post your output yet. Assuming the data has already been sorted. data have; input Emp_NO REPORTING_DIVISION $ DIVISIONAL_CODE $ MONTH     ; cards; 85776 B B 201201 85776 B B 201202 85776 B B 201203 85776 B B 201204 85776 B B 201205 85776 B B 201206 85776 B B 201207 85776 B B 201208 85776 B B 201209 85776 B B 201210 85776 B B 201211 85776 B B 201212 85776 B B 201301 85776 B B 201302 85776 B B 201303 85776 B B 201304 85776 B B 201305 85776 B B 201306 85776 B B 201307 85776 B B 201308 85776 B B 201309 85776 B B 201310 85776 B B 201311 85776 B B 201312 85776 B B 201401 85776 B B 201402 85776 B B 201403 85776 B B 201404 85776 B B 201405 85776 B B 201406 85776 C C 201407 85776 C C 201408 85776 C C 201409 86322 B B 201201 86322 B B 201202 86322 B B 201203 86322 B B 201204 86322 B B 201205 86322 B B 201206 86322 B B 201207 86322 B B 201208 86322 B B 201209 86322 B B 201210 86322 B B 201211 86322 B B 201212 86322 B B 201301 86322 B B 201302 86322 B B 201303 86322 B B 201304 86322 C C 201407 86322 C C 201408 86322 C C 201409 ; run; data want; do until(last.Emp_NO);   set have;   by Emp_NO; end; _X=REPORTING_DIVISION; do until(last.Emp_NO);   set have;   by Emp_NO; REPORTING_DIVISION=_X;     output; end; drop _X; run; Xia Keshan ... View more

Are you just trying to do something like?: data have;   format Hospital_Name $80.;   input Hospital_Name 1-80;   if countw(scan(Hospital_Name,2,'-')) eq 1 then    Hospital_Name=compress(Hospital_Name,'-');   else Hospital_Name=scan(Hospital_Name,2,'-');   cards; INDIA - ABC HEALTHCARE CHINA - EFF HEALTHCARE USA  - GHI HEALTHCARE UK - KLK HEALTHCARE HAPPY - HEALTHCARE WELL - HEALTHCARE WEALTH - HEALTHCARE FREE - HEALTHCARE AUS - HEALTHCARE ; ... View more

Customize a format for it. proc format; value fmt       0-1=[percent8.2]       other=[best8.] ; ... View more

I'll offer the comment that there's no gurantee that you can and it seems like a bit of a weird request. If your programs are data management, I'd expect you to be using SQL, PowerShell, Python or R, not DOS. If you have any statistical component to the programs then you'll most likely need Python or R, not DOS. I'd suggest hiring someone to do the conversion if you really need it done. If its an extremely small program you can post it here and see what people suggest. ... View more

Hi, Try License Assistance for SAS Customer Support and / or Contact SAS Institute Technical Support Regards, Amir. ... View more

You could try "repair" Base SAS(R) 9.3 Procedures Guide, Second Edition ... View more

This page addresses your issues http://www.sascommunity.org/wiki/Macro_Loops_with_Dates This section has three ways of getting an integer from a data reference: http://www.sascommunity.org/wiki/Macro_Loops_with_Dates#A_Date_is_a_Reference_to_an_Integer Ron Fehd  macro maven "A SAS date is an index into the associative array of dates." ... View more

You don't need a date constant using the interger value will work as well in your code. 18         %let ASATDATE=03-06-2013; 19         %LET LAP_DAT_IND = %sysfunc(inputN(&ASATDATE,mmddyy,10)); 20         %put NOTE: &=asatdate &=lap_dat_ind; NOTE: ASATDATE=03-06-2013 LAP_DAT_IND=19423 ... View more

Proc ASI2 is an interface to IDMS using that  Database. It will not run faster as allowed by the schema / network access statements. International Software Products - ASI2 and ASI2LR Sample Code IDMS is a network database (NOSQL). Accessing a network database can be superfast or very slow dependent of the quality of the access paths (set first-next). The LRF option when used does a translation to a relation approach at the cost of overhead. Your IDMS DBA should be able to  help you with performance issues    ... View more

Just to add to Cynthia's post: You can use a command like "filename dummy pipe '<zip command>'; " Beginning with SAS 9.4 there is now also a zip engine which you could use: SAS(R) 9.4 Statements: Reference ... View more