Date:         Mon, 7 Jun 1999 20:47:34 -0400
Reply-To:     Raynald Levesque <rlevesque@VIDEOTRON.CA>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         Raynald Levesque <rlevesque@VIDEOTRON.CA>
Subject:      Re: Restructuring
Comments: To: Thomas Otter <Thomas.Otter@WU-WIEN.AC.AT>
Content-Type: text/plain; charset="iso-8859-1"

Hi Otter,
 
I believe the following syntax solves your problem:
 
 
compute id=$casenum.
 
* NOTE: change all 5's in the script by the number of objects that you have.
vector obj = obj_1 to obj_5.
 
* convert each record to multiple records.
loop obj_idx = 1 to 5.
 compute xvalue = obj(obj_idx) .
 xsave outfile = 'c:\temp\temp1.sav' / keep = id obj_idx xvalue /compressed.
end loop.
execute.
get file = 'c:\temp\temp1.sav' .
 
* rank and sort ratings of each object by id.
rank variables = xvalue by id
 / ties = low / rank into xrank .
sort cases by id xrank.
 
* convert from multiple records back to single records.
vector r(5) rank_n(5).
compute r(xrank)  =obj_idx.
aggregate outfile = * /presorted / break = id
 /rank_n1 to rank_n5 = max(r1 to r5).
 
formats rank_n1 to rank_n5 (f8.0).
 
Regards
 
Raynald Levesque rlevesque@videotron.ca
 
----- Original Message -----
From: Thomas Otter <Thomas.Otter@WU-WIEN.AC.AT>
Newsgroups: bit.listserv.spssx-l
To: <SPSSX-L@UGA.CC.UGA.EDU>
Sent: Monday, June 07, 1999 4:28 PM
Subject: Restructuring
 
 
> Dear restructuring experts,
>
>
> I have data like:
>
>    obj_1   obj_2   obj_3  obj_4 ... obj_k
>
> 1   x11    x12     x13 .........     x1k
> 2
> 3
> 4
> 5
> 6
> .
> .
> .
> .
> n                                    xnk
>
> The x values can take on any real valued number.
> I need to restructure the data into
>
>    first     second    third   fourth ... k_th
>
> 1   y(obj_2) y(obj_k)
> 2
> 3
> 4
> 5
> 6
> .
> .
> .
> .
> n
>
> where the y(obj_?) simply indicate that it was e.g. obj_2 that
>  received the lowest x-value from person 1 and obj_k receiving
> the second lowest x-value from person 1 and so
> on.
>
>
> Please let me know if you need more information.
> Thank you in advance,
>
> Thomas Otter