topic Re: first. and last. usage in sas in SAS Programming

first. and last. usage in sas

buddha_d — Mon, 08 May 2017 20:26:13 GMT

Hi All:

Could some one explain first.variable and last. variable concept for me ?

eg: the data ids2 has been sorted using proc sort,

data dupes;
set ids2;
by id;
if not (first.id and last.id) then output;
run;

if not (first.id and last.id) then output; --> what doest this code do?

is there a paper about first. and last. usage in SAS?

Re: first. and last. usage in sas

Reeza — Mon, 08 May 2017 20:44:05 GMT

I would recommend starting with the documentation.

http://support.sas.com/documentation/cdl/en/lrcon/69852/HTML/default/viewer.htm#n138da4gme3zb7n1nifpfhqv7clq.htm

The section related to First/Last is under the section titled:

How the DATA Step Identifies BY Groups

Re: first. and last. usage in sas

LinusH — Mon, 08 May 2017 20:44:27 GMT

This means if first. and last. isn't the same observation then output. This will be true for all observations in a BY group with more than one observation.
There are tons of papers. Search support.sas.com and global forum proceedings as a start.

Re: first. and last. usage in sas

Astounding — Mon, 08 May 2017 20:44:37 GMT

The net effect: any ID values that have just a single observation get deleted. It's a topic that is well worth studying. Here is one place to start:

http://www2.sas.com/proceedings/forum2007/222-2007.pdf

You may need to read a few background pages before it gets to the point of adding the BY statement in the DATA step.

Re: first. and last. usage in sas

art297 — Mon, 08 May 2017 20:46:03 GMT

You said it was sorted, I'll presume it was sorted by id.

In that case, using ID as the by variable, first.id will be equal to 1 when, and only when, it is the first record for that ID.

Similarly, last.id will be equal to 1 when, and only when, it is the last record for that ID.

As such, think about the statement you asked about: if not (first.id and last.id) then output;

It's opposite if (first.id and last.id) then output; would only output records where there was only one record for an ID. For any IDs that had multiple records, it would never be true for any of the records. As such, if not (first.id and last.id) then output; will output all records where there ARE multiple records for a given ID.

Art, CEO, AnalystFinder.com

Re: first. and last. usage in sas

buddha_d — Mon, 08 May 2017 20:51:13 GMT

A big thanks to all the following people

LinusH, Reeza, Astounding & art297.

The explanantion certainly helped me understand the concept. Thanks a lot.