Thanks, I'll try all 3 responses (in SAS-L) at work today.

Cheers

Doug

"pudding man" <pudding_man@MAIL.COM> wrote in message news:20040129164901.5339.qmail@mail.com... > Doug is headed in the right direction. This is not > difficult to do with array processing. > > If I had a moderate volume of data, I might structure > something like (tested): > > data new(keep = id: item:); > *set old; > retain id 1 id2; > array rels(144) relate_1 - relate_144; > array item(12); > do i = 1 to dim(rels); rels(i) = i; end; *** just for test ***; > > j = 1; > do k = 1 to 144; > item(j)=rels(k); > j + 1; > if mod(k, 12) = 0 then do; id2 + 1; output; j = 1; end; > end; > run; > > proc print; run; > > If I had tons of data, I might consider something > with PEEK n' POKE: > > data new(keep = id: item:); > *set old; > retain id 1 id2; > array rels(144) relate_1 - relate_144; > array item(12); > do i = 1 to dim(rels); rels(i) = i; end; *** just for test ***; > > do j = 1 to dim(rels) by 12; > id2 + 1; > call poke(peekc(addr(rels(j)), 8 * 12),addr(item(1)), 8 * 12); > output; > end; > run; > > proc print; run; > > This latter should run a good bit faster ... > > Hope it hep' s... > > Prost, > Puddin' > > ******************************************************* > *** Puddin' Man *** Pudding_Man@mail.com ******** > *******************************************************; > > Guantanamo, anyone? > > "The power of the executive to cast a man into prison > without formulating any charge known to the law, and > particularly to deny him the judgment of his peers, is in > the highest degree odious, and the foundation of all > totalitarian government whether Nazi or Communist." > - Winston Churchill, 1943 > > ----- Original Message ----- > From: dg <queanbeyan@HOTMAIL.COM> > Date: Thu, 29 Jan 2004 22:00:33 +1100 > To: SAS-L@LISTSERV.UGA.EDU > Subject: simple output depending on variable > > I have a dataset that contains id, relate_1 to relate_144. Each value of > relate_1 to relate_144 is numeric. Id = character. > > I would like to create a new dataset that has the following items; > id, id2, item1 - item12. > > where > id =id, id2=1, item1-item12 = relate_1 to relate_12 for the relate_1 to > relate_12 variables, > id=id, id2=2, item1-item12 = relate_13 to relate_24 for the next 12 relate > variables etc down to ..... > > id=id, id2=12, item1-item12=relate 131 to relate_144 for the last 12 relate > variables > and so on. > > This is what I have tried.. > data new; > set old; > array rels(144) relate_1 - relate_144; > array item(12); > do h=1 to 12; > do i = h to 144 by 12; > item(h)=rels(i); > output; > end; > end; > run; > > but this doesn't give me what I want. I suppose I want a vertical slice > after each 12 relate variables and the number of the slice in a vertical > file. > > Any help appreciated. > > Doug > > > > -- > _______________________________________________ > Get your free email from http://www.mail.com