Date:         Sun, 20 May 2007 09:27:57 -0400
Reply-To:     John James <free562james@YAHOO.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         John James <free562james@YAHOO.COM>
Subject:      Re: Lag function

Thanks for the reply. I would ideally like to use a data step to accomplish
this. If anybody else has other suggestions, let me know.

On Sun, 20 May 2007 08:09:29 -0400, Richard A. DeVenezia
<rdevenezia@WILDBLUE.NET> wrote:

>John James wrote:
>> Dear SAS-L users,
>>
>> I use the lag function to create the  second, third and fourth lag of
>> variable X for each firm (y) in my sample across time (t). However, on
>> certain occassions the lag variable created corresponds to a different
>> firm. Is there a neat way to set these observations equal to missing?
>>
>> this is the code i am using
>> proc sort data=a;
>> by y t;
>> run;
>>
>> data a;
>> set a;
>> xlag2=lag2 (x);
>> xlag3=lag3 (x);
>> xlag4=lag4 (x);
>> run;
>
>There is probably a Stat procedure for doing such.
>
>Here is some data step code that uses macro to manage a lag array in a by
>group setting.
>
>-----------------------------
>data foo;
>  do group = 1 to  4;
>    do seq = 1 to 10;
>      x = group*100 + seq;
>      output;
>    end;
>  end;
>run;
>
>%macro lagArray(by=,var=,first=,last=);
>  %local i;
>
>  array &var.lag[&first:&last] &var.lag&first-&var.lag&last ;
>  %do i = &first %to &last;
>    &var.lag&i = lag&i(&var);
>  %end;
>
>  if first.&by then &by.seq=1; else &by.seq+1;
>
>  do _i_ = max(&by.seq,&first) to &last;
>    &var.lag[_i_] = .;
>  end;
>
>  drop _i_ &by.seq;
>
>%mend;
>
>options mprint;
>
>proc format;
>  value missword .='-----' other=[z5.];
>run;
>
>data scan;
>  set foo;
>  by group;
>
>  %lagArray(by=group,var=x,first=2,last=5);
>run;
>
>proc print;
>  by group;
>  id group;
>run;
>-----------------------------
>
>Richard A. DeVenezia
>http://www.devenezia.com/