Make sure you run the SQL after you have created the data set with the variables you want to rename. ... View more

This will work. proc sql; select cat(name, ' = ', cats(name, '_3mo' )) into :renstr separated by ' ' from dictionary.columns where libname = 'WORK' and memname='FMIN'; quit; data new; set fmin (rename = (&renstr)); run; ... View more

Do you have SAS ODBC? I would think VISIO should be able to reverse engeneer from that. I don't have it here, so I cannot try it. You will not get any links between tables unless it uses column names. ... View more

Sorry I don't have any recent information. I used to run an SQL describe to generate the code for a table, then load that into ERWin. I no longer have a copy and the last few shops I have worked at did not have accessible ER tools. So I am just working from memory here. ... View more

Well first off &dates is not assigned so that code is not included. There is no code shown to define the file referenced by &dates so we cannot know what happened there. ... View more

Or just for fun. data start; input sys_id company_name $ company_identifer $ state $; cards; 1 ABC 245679762 MI 2 XYZ 876234T87 OH 3 WALDEN 88723O124 MA 4 GOOGLE 77823I457 CA 5 TALI 44321H887 AZ run; data look1 (index = (frst6)); input frst6 $; cards; 245679 88723O 77823I 44321H run; data end; set start; drop frst6; frst6 = substr(company_identifer, 1,6); set look1 key = frst6/unique; if (not _iorc_ = %sysrc(_sok)) then do; delete; _error_ = 0; end; run; ... View more

How about if substr(company_identifer , 1, 6) in ( &id ) then output; just create your list for ID quoted with a comma delimiter. ... View more

The error is Oracle http://www.orafaq.com/wiki/ORA-01858 You are converting an improper date. ... View more

While many ERD tools can reverse engineer table structures into an ER, SAS does not store foreign keys in the table structure so reslationships are not preserved. Years ago I wrote a set of macros to work with ERwin so some of it can be done. There is just no easy solution. ... View more

Since you include no code, and do not specify what the put command is doing I can only assume you are using a numeric format on a character value. A lot more info would be needed to get specific. ... View more

To keep it numeric floor(x/100); ... View more

I suggest you Google "Clinical endpoint" The listings etc. would totally depend on the clinical area and team you are working with. ... View more

ROFLMAO ... View more

In EG you would need to do : data CM; input subjid $ cmtrt $; cards; 001 ASIPIRIN 001 ZOFRAN 002 DIGOXIN 002 CARDIZEM run; data AE; input subjid $ aetrt $; cards; 001 HEADACHE 001 NAUSEA 002 VOMITING 003 INDIGESTION run; proc sort data = cm; by subjid; run; proc sort data = ae; by subjid; run; data cm; set cm; by subjid; if first.subjid then cnt = 0; cnt + 1; run; data ae; set ae; by subjid; if first.subjid then cnt = 0; cnt + 1; run; to get an identifyer on each row. Do an OUTER join to get the matching (on subjid and cnt) create a calculated column for SubjID "coalescec(CM.subjid ,AE.subjid )" ... View more

One of my favorite interview questions back when I did a lot of hiring. You don't even need a macro. The ODS steps could be outside the data _null_ data _null_; set sashelp.vmember( where = (libname = upcase('&libname') and memtype = 'DATA')) end = eof; if _n_ = 1 then call execute("ODS Listing Close; ods rtf nogtitle file='C:\report1.rtf' style=profile ;"); call execute("proc print data = " || memname || " noobs; run;"); if EOF then call execute("ods rtf close;"); run; ... View more