I don't have an environment where I could try it myself but just looking at your code, the referenced usage note and doing some further Internet research I can't see anything wrong in your code.  I couldn't find any confirmed solution anywhere how to send a single .sas7bdat as email attachment.  I believe what would add value to the forums here: Contact SAS Tech Support directly and then post the final answer/resolution back here and mark it as solution/answer as provided by SAS Tech Support. ... View more

@shrikrishna wrote: The requirement is that I need all data sorted in specified way (taken care of) in excel file. There is no way around it. people I am sending requirement to like excel format way too much. 10000 usernames are unique, and data cannot be grouped as they need day by day each instance in the file. (Legal requirement) I dare to doubt that you've got a requirement that leads to an Excel file with 1000 tabs or even to 1000 Excel files. That's just not manageable.  People normally like all the data within a single tab with columns that allow to apply filters. Or even data surfaced as pivot table in a single tab.    SAS can load data in any sort order into an Excel tab as long as there is some logic to it.   If you really believe you're asked to create 1000 tabs or 1000 separate Excel files then I'd strongly recommend you confirm this with your requestor once more before you proceed spending time on such code. ... View more

Works for me. Verify that both of your variables are of type numeric. data hs300_0604; if=2; dlastprice=1;output; if=4; dlastprice=3;output; run; proc sql noprint; select avg(IF)-avg(dlastprice) into :lpadj from hs300_0604 ; quit; %put &=lpadj; 38 %put &=lpadj; LPADJ= 1   Update: As a reaction on what @yabwon posted (worth testing) the code also works for me within a macro. data hs300_0604; if=2; dlastprice=1;output; if=4; dlastprice=3;output; run; %macro doit(param1,param2); proc sql noprint; select avg(&param1)-avg(&param2) into :lpadj from hs300_0604 ; quit; %put &=lpadj; %mend; options mprint; %doit(if,dlastprice); MPRINT(DOIT): proc sql noprint; MPRINT(DOIT): select avg(if)-avg(dlastprice) into :lpadj from hs300_0604 ; MPRINT(DOIT): quit; NOTE: PROCEDURE SQL used (Total process time): real time 0.00 seconds cpu time 0.01 seconds LPADJ= 1   ... View more

SQL is not very suitable for transposing data. Using SAS Proc Transpose or a data step is what you should be using. The only reason for a SQL would be if processing needs to get pushed to a database. If so then you would need to consult the documentation for the database and use explicit database specific SQL if and as available (like the PIVOT command in Oracle). ... View more

You need to tell us more about your tables. When people call a table large then some mean 100K rows and others 100M rows.  What are the volumes (number of rows) of t2 and t4? How many distinct values for RN do you have in t4? What is the datatype of RN and if character what's the variable length. Where is the data stored (SAS table on disk or in some database like Oracle or both)? A SAS data step hash lookup would likely perform better. Points 1-3 above are for making a decision if your table t4 could fit into memory, point 4 is for making a decision if processing needs to get pushed to a database side for processing (which a data step hash lookup wouldn't). Also don't use select * but list the variables you need explicitly, like: select t2.name, t2.rn, t4.xxxx, ....       ... View more

Did you already search the forums and assess discussions like this one? https://communities.sas.com/t5/SAS-Studio/Is-there-a-way-to-set-the-excel-sensitivity-label-by-a-SAS-proc/m-p/912951  ... View more

Below should work. data want(drop=_:); set have; retain _rdate; if response ne 'NE' and not missing(date) then _rdate=date; date=coalesce(date,_rdate); run; ... View more

@BW18 wrote: There are no leading zeros in the CorpIDs. They range from 4001-4975.  However, I closed out SAS and reimported the text file containing CorpAddresses. I performed the transformation to a character variable for CorpIDs and it appeared to work this time. Following this, I was able to perform the merge successfully. Thank you! Good that you've got something working. Just as a comment: Ideally you read the source string already into a character variable as part of the import step.  ... View more

@HeatherNewton wrote: Hi Patrick, Sorry. let me clarify. I didnt get customer to go to SAS tech. I took your advice to reach out to them to further resolve the whatever issue as we managed to conclude using the code you gave me to find pd and pd_char and confirm it is odbc and not my data issue. So I told them I proved by those code and request to look into it.   Sorry please dont get upset. I really appreciate your help. OK I will work with them further. Thanks a milion @HeatherNewton  My advice was for your customer to directly raise this issue with SAS TS. You don't have environment access and though piping all the communication via you as the middle-man just complicates things and will add confusion. I still suggest you share with SAS TS the link to the discussion here as this will help them to better understand the situation and where things are at. ... View more

@HeatherNewton wrote: I have now reached out to SAS technical they are asking why is it that the pd_char has the cast function whereas the pd does not?   please kindly let me know thanks,     ??? I can't know what you've told SAS TS that makes them ask this question. We've used the cast function on the DB side to convert the numerical value to a string on the DB side to figure out in what precision the values are stored on the DB side.  It appears you went against my advice to have the customer raise the issue with SAS TS. There are now too many middle-men in this chain. I'm not going to answer any further questions here. Let SAS TS take the lead. I suggest you share with SAS TS the link to the discussion here so they can understand what already has been discussed and looked at. ... View more

@BW18  Please share the output running below code: proc contents data=N.CorpAdresses(keep=CorpID); run; proc contents data=N.StoreInformation(keep=CorpID); run;   Do your CorpId values in the data set where the variable is of type character have any leading zeros?   ... View more

Please provide more SAS related details so we know this is not spam. ... View more