I am going to analyze a batch of SAS program file and I am stucked in getting the last modified time of program files. I have thought about X command but it was too inefficient. I just find when I use infile statement:   data test; infile 'D:\test.txt' truncover; input ; run;   Log shows the last modified time: NOTE: The infile 'D:\test.txt' is:         Filename=D:\test.txt,         RECFM=V,LRECL=32767,File Size (bytes)=7,         Last Modified=2021/1/26 15:25:48,         Create Time=2021/1/26 15:25:42 As you can see, log window shows the infomation of file as a NOTE. However, my wish output is a variable filled with Last Modified Time. Is there some option to get it while using `infile` statement? Surely, Other efficient ways are welcomed, too. ... View more

As far as I know, there are three kinds of ways to add comment in SAS:    *Any comment text here; %*Any comment text here; /*Any comment text here*/   This afternoon, I am kind of excited of finding the fourth way to add comment by coincidence. It is:  comment Any comment text here;   As you can see, the first word comment is a keyword here to trigger the following text becoming comment text. I have tried several programs:   /*comment of macro in open code*/ %put This is %sysfunc(date(),e8601da.); comment %put This is %sysfunc(date(),e8601da.); /*comment of macro*/ comment %cmprs(test); /*comment after normal statement*/ data _null_; comment Hi there.; run;   They all behaves like a normal comment way. Only one point, there mustn't exist a semicolon in the comment text. I think there is enough exploration in SAS. I search in the help document and find nothing. My friend tells me this may be a pre-experience features, How do you know about it?   Please share your idea. ... View more

I am doing something needs more than once query in  proc sql  so I try to write code perform like nest function in  data step . Here is an example: Data test have 2 variables, Group and Num. data test; input Group$ Num; cards; A 10 A 30 B 10 B 40 C 30 C 30 ; run; Now I am looking for which  Group  has the biggest mean value. So firstly I compute mean for each  Group , secondly get the max value of these means, Finnaly select the  Group  whose mean has the same value of the result in step 2. proc sql; select Group, avg(Num) as Mean from test group by Group having Mean = ( select max(mean) from ( select avg(Num) as mean from test group by Group ) ) ; quit; Yes, I get the right answer, it's "C". But I don't like this method, it is too lengthy. The following code is wrong in syntax: proc sql; select Group, avg(Num) as Mean from test group by Group having Mean = max(avg(Num)) ; quit; But it is much beautiful, and shorter, too. Do you have any better way to do this? ... View more