In UNIX, you can do something similar by setting an environment variable to the name of an existing directory: mkdir /tmp/saslib export SASREF=/tmp/saslib /sas/install/SASFoundation/9.4/sas In the SAS session, library SASREF will be available without executing an extra LIBNAME statement. ... View more

With the original dataset HAVE the grouping in the "data want1" and "data want2" steps is the same because variables a and x change their values in exactly the same observations. After sorting HAVE by x, however, the two formerly separated BY groups with a='A' and a='D' form a single BY group with respect to variable x (four consecutive observations with x=2). Hence the "data want2" step would select only one observation (with y=24) from that new group, not two (y=22, y=24) as before.   So, if the original grouping is what you want, but no variable like a exists, then you're in a situation where the NOTSORTED option (applied to x) comes to the rescue.   ... View more

I don't get such an error, Richard. ... View more

You need to provide a lot more detail about what you have and what you are trying to do. The NATURAL JOIN syntax in PROC SQL will autodetect matching variable names and use those as keys when joining datasets. proc sql; create table want as select * from a natural join b ; quit; ... View more

If the libraries are accessed through metadata, and we talk about metadata permissions, then someone decided that you have no business even knowing that the datasets exist.   Now, if you do not use the datasets through metadata, but have no read permission in the operating system, but do have read permission on the directory that constitutes the library, you can use (UNIX) a system command like ls -l to display the wanted information. ... View more

Some of the modeling procedures produce data set with very particular structures. If you are not intimately aware of what the structure is and how it is used for in model then no changes would be a good idea.   Changes to values to any variables or their names would almost certainly invalidate any result when used as Inmodel.   ... View more

If you mean that you generate code similar to this: data one; infile filename <fileoptions>; input <variables>; run; data two; set one; newvar = <some calculation>; end; data three; set two; othervar = <some calculation>; end; data four; set three; nowthatvar =<more calculations>; run; What I would suggest is after you testing the data from creating data set two to move the calculations in the data step that creates data set one. Then as you test each bit of code to move it to the earliest practical set.   It may be that the changes should all be moved into the data two step especially if you are replacing values in existing variables (data cleaning, standardizing or recoding perhaps). That way if data step two does have a problem then the data one is available to restart from the same point.   If any of these extra data steps only do Format or Label assignments then really they should be in an earlier data step. ... View more

Sorry, I just don't understand what you are trying to achieve.   I want to set values of Goals_Scored between 10 and 100 Can you give an example?    then if someone falls into Group (A, B, C, D, E)   What are Group (A, B, C, D, E)? ... View more

@edasdfasdfasdfa wrote: It is a .xlsx file. One approach: Export to a CSV file and then import that CSV. Then you will have the data step code in the LOG to read a csv file. You may want to use the GUESSINGROWS option to specify a large number of rows to examine before SAS sets the properties for the variables. Otherwise only 20 or so rows are used which is a frequent problem.   That code could be copied, pasted into the editor and edit the Informat and format statements.   I convert dozens of xlsx to csv files monthly to have control while reading the files because I can't trust my data sources not to change value codes, change column headings or change column order. ... View more

Hi @edasdfasdfasdfa ,   Have you tried a proc dataset ? Normally, the output report specifies the dataset owner. Hope this help. proc datasets library=mylib; contents data=mydata; run;   ... View more

@edasdfasdfasdfa: SQL doesn't understand the => operator, hence the syntax error.  Use either >= or the mnemonic GE instead, and it will go away.  Your outer CASE clause is not closed with ELSE. Though It's not a syntax error, you'll get a missing value as a result if no WHEN condition in that outer clause is true. SAS will warn you about it in the log, so pay attention or else your output data may be wrong. Kind regards Paul D.  ... View more

Macro variables resolve in double quotes, not in single quotes If you have an & or % in the path, SAS may think that's a macro variable/macro so you would use single quotes so it will not try to resolve the macro variable or macro.  If you're using SQL pass through or X commands, you may need specific quotes - follows the rules of the SQL or OS. Other than those cases, single and double quotes are usually interchangeable. @edasdfasdfasdfa wrote: So I notice that for csv files there are double quotes and for .txt single quotes around the path of the file in the infile statement. Why is this the case?   ... View more

Depends on context, but usually "PROC step" refers to any SAS procedure step.   Sometimes people compare PROC SQL vs DATA step for data management tasks.  But that can be a little narrow, as really you have PROC MEANS/SUMMARY for summarization, TRANSPOSE for "flipping" rows/columns, FREQ for counting, etc -- it's a larger suite of procedures to accomplish data-related goals. ... View more