Date: Mon, 13 Jul 1998 14:26:02 -0500
Reply-To: "H. T. Cheng" <hcheng@ERAC.COM>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: "H. T. Cheng" <hcheng@ERAC.COM>
Subject: Re: Comparing observation variables within 1 dataset
Content-Type: text/plain; charset=ISO-8859-1
I have seen several very good answers to this thread, including SQL and
Proc Transpose solutions. Here comes another thought: If the original
author is interested in any kind of ranking of the current value among all
the values for a given a variable in the dataset, then Proc Rank might be
useful.
Regards!
Harry Cheng (hcheng@erac.com)
Enterprise Rent-A-Car
(314) 512-2096 (Work)
(314) 512-4096 (Fax)
----------
> From: Bernard Tremblay <bernard@CAPITALE.QC.CA>
> To: SAS-L@UGA.CC.UGA.EDU
> Subject: Re: Comparing observation variables within 1 dataset
> Date: Monday, July 13, 1998 5:16 AM
>
> Hi,
>
> First thought about it:
>
> Is your dataset small enough to be transposed and then load
> one variable in memory ?
>
> if yes, then you could :
>
> 1- Get the number of observations ;
> 2- Get the name and type of each variable from the
> dictionary (dictionary.columns);
> 3- transpose the dataset;
> 4- Merge dictionary info with the transposed dataset by _name_;
> 5- In a datastep have 2 arrays of size NOBS. One for numeric,
> one for character. Then loop thru your array to compare,
> if any differences then output an appropriate message...
> (The variable _name_ contains the name of the variable)
>
> 6- done!
>
>
> Second thought about it:
>
> May be one could loop with a macro that would create a new
> dataset similar to the original but the order of observations changed.
> then run a proc compare to compare both files and so on until all
> combination are done ???
>
> Happy coding!
>
> regards,
> Bernard Tremblay
> \\\|///
> \\ - - //
> ( @ @ )
> +------oOOo-(_)-oOOo----------+---------------------------------+
> | Bernard Tremblay | |
> | La Capitale | Tel: (418) 646-2401 |
> | | Fax: (418) 646-5960 |
> | | Int: bernard@capitale.qc.ca |
> +-----------------------------+---------------------------------+
> | Imaginasys enr | Res: (418) 878-4447 |
> | | Int: bertrem@quebectel.com |
> +---------------Oooo----------+---------------------------------+
> oooO ( )
> ( ) ) /
> \ ( (_/
> \_)
>
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~
> >>>Date: Sat, 11 Jul 1998 06:55:26 GMT
> >>>Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
> >>>From: Andreas Grueninger <grueni@STUTTGART.NETSURF.DE>
> >>>Subject: Re: Comparing observation variables within 1 dataset
> >>>
> >>>DATA test;
> >>> DO i=1 TO 5; a=i; OUTPUT; END;
> >>>RUN;
> >>>
> >>>DATA _null_;
> >>> SET test NOBS=nobs;
> >>>PUT 'Comparing value of record ' _n_ ;
> >>> DO i=_n_+1 TO nobs;
> >>> SET test (RENAME=(a=next)) POINT=i;
> >>>PUT +3 'with value of record ' i a= next=;
> >>> END;
> >>>RUN;
> >>>
> >>>On Thu, 09 Jul 1998 16:38:19 -0400, vermurra@email.uc.edu wrote:
> >>>
> >>>>Can anyone help me or point me in the right direction?
> >>>>
> >>>>I'm trying to locate discrepancies in a dataset by comparing
variables
> >>>>in one observation to the same variables in every other observation.
> >>>>
> >>>>In short, I want to compare var1 in obs1 to var1 in obs2, obs3, etc.
> >>>>Then I want to do the same for obs2, comparing it to obs3, obs4 etc.
> >>>
> >>>---------------------------
> >>>Andreas Grueninger
> >>>
> >>>PRIVAT: grueni@stuttgart.netsurf.de
> >>>OFFICIAL: grueninger@lfl.bwl.de
> >>>---------------------------
> >>>
| 
