programming issue

[Niugg2010]

Below is my dataset

In the first row, if some column is blank, then to polulate that column with second row data;

simimally, for third row, if some column is blank then to polulate the column with the fourth row data.

the final dataset looks like this:

 

How can I write the code to get dataset B? Thanks.

The below code can get first dataset data;

/*Code for first dataset data*/

data a;
format patid event1 event2 event3 event4 event5 event6 event7 column8 column9 $8.;
infile datalines missover dsd ;
input patid event1 event2 event3 event4 event5 event6 event7 column8 column9;
datalines;
001, aa1,, aa3, aa4,, aa6, aa7,, aa9
001, bb1, bb2, bb3,,bb5, bb6, bb7, bb8,,
002,cc1,cc2,cc3,,cc5,cc6,,cc8,cc9
002, dd1,dd2,dd3,dd4,dd5,dd6,dd7,dd8,dd9
;

[art297]

Here is one way:

data want (drop=_:);
  merge have have(firstobs=2 drop=patid
    rename=(event1-event7=_e1-_e7 column8-column9=_c8-_c9));
  array now(*)  $ event1-event7 column8 column9;
  array next(*) $ _e1-_e7 _c8 _c9;
  do i=1 to dim(now);
    now(i)=coalescec(now(i),next(i));
  end;
run;

Art, CEO, AnalystFinder.com

[Niugg2010]

Thanks. Your method is novel.  The code is short, but learn a lot.  

I transposed the datasets  first, and dealt with the data, then transposed back.