topic Re: How do I simplify pulling multiple datasets (from a pubic library) into my work library? in SAS Programming

How do I simplify pulling multiple datasets (from a public library) into my work library?

jlsagisi — Fri, 26 Jun 2020 19:27:20 GMT

I'm learning SAS and I'm looking for ways to make my code better. For assignments, I start by pulling the files I need from a public library into my work library. Right now, I am writing out a DATA SET step for each file pull. For example, if I know I need to work on three files, I start my assignment with the code below:

data dataset1;
set public.dataset1;
run;

data dataset2;
set public.dataset2;
run;

data dataset3;
set public.dataset3;
run;

Is there a way to simplify this? And what are your thoughts on best practices, if any?

Re: How do I simplify pulling multiple datasets (from a public library) into my work library?

PaigeMiller — Fri, 26 Jun 2020 17:00:18 GMT

Since you specifically ask about "best practices", one has to ask why you want to do this. You can do a PROC MEANS (for example) or any other PROC on public.dataset1, just as you could on work.dataset1. The advantage of doing it on public.dataset1 is that it is one less step to program for each data set of interest. So I am asking you to consider simplifying your work flow by leaving the data sets in library PUBLIC. In your LIBNAME PUBLIC statement, you can add option ACCESS=READONLY so that you don't accidentally overwrite the data sets in PUBLIC. These are "best practices" in my opinion.

Nevertheless, you may have valid reasons to do this which you didn't state, in which case PROC COPY will do what you want, in bulk, all (or selected) datasets in PUBLIC get copied to WORK by writing one PROC COPY.

Re: How do I simplify pulling multiple datasets (from a pubic library) into my work library?

ballardw — Fri, 26 Jun 2020 15:05:02 GMT

In addition to @PaigeMiller's suggestion I would say that if your are going to want to use those data sets in another SAS session that you should copy them to permanent library instead of Work. When you close SAS then your work library is emptied (barring some interesting settings) and if you need the data then you start from scratch. If you create a local permanent library then you just make sure that library is available and you're good to go.

Of do you need to have new versions because the public data changes?

Re: How do I simplify pulling multiple datasets (from a pubic library) into my work library?

jlsagisi — Fri, 26 Jun 2020 15:12:28 GMT

Great question. I was told to do this so I don't accidentally make changes to a "master copy" in the public folder (which is used by more than person). And, I definitely see what you're saying. I was able to use PROC COPY. Thank you.

Re: How do I simplify pulling multiple datasets (from a pubic library) into my work library?

Tom — Fri, 26 Jun 2020 15:15:52 GMT

Use PROC COPY.

proc copy inlib=public outlib=work;
  select dataset1 dataset2 dataset3 ;
run;

Re: How do I simplify pulling multiple datasets (from a pubic library) into my work library?

Tom — Fri, 26 Jun 2020 15:19:27 GMT

@jlsagisi wrote:
Great question. I was told to do this so I don't accidentally make changes to a "master copy" in the public folder (which is used by more than person). And, I definitely see what you're saying. I was able to use PROC COPY. Thank you.

Just make sure to define PUBLIC as readonly and you will not be able to change them.

libname public 'path' access=readonly;

Re: How do I simplify pulling multiple datasets (from a pubic library) into my work library?

ballardw — Fri, 26 Jun 2020 16:12:03 GMT

@jlsagisi wrote:
Great question. I was told to do this so I don't accidentally make changes to a "master copy" in the public folder (which is used by more than person). And, I definitely see what you're saying. I was able to use PROC COPY. Thank you.

Depending on where your Public library is defined permissions or options should prevent that.

If you execute a LIBNAME statement you can add the option Access=readonly and then you cannot modify any of the data sets, intentionally or accidentally.

The statement would look something like:

libname mylib "<path to the library>" access=readonly;

Re: How do I simplify pulling multiple datasets (from a public library) into my work library?

PaigeMiller — Fri, 26 Jun 2020 17:01:07 GMT

@jlsagisi

Might I suggest that you go back to your ORIGINAL post and correct the spelling of "public" in the Subject line? (Or perhaps a moderator can do this). Thanks!

Re: How do I simplify pulling multiple datasets (from a public library) into my work library?

ballardw — Fri, 26 Jun 2020 19:27:49 GMT

@PaigeMiller wrote:

@jlsagisi

Might I suggest that you go back to your ORIGINAL post and correct the spelling of "public" in the Subject line? (Or perhaps a moderator can do this). Thanks!

Done