Creating a New CaseID
Locked
 
|
*Dear Listers,
I am trying to apply a new definition of case. The data were provided in a manner where a unqiue CASE_ID equals one Case. However, I would like a case to include specific types of court orders - represented by VAR CODESCX, where (CODESCX=1) is the start of a case. Therefore, when the first order of a CASE_ID is 1, the following SYNTAX works; however, in some cases, a new case may start with another CODESCX value besides 1. In this situtation, the RPUID has a new CASE_ID number and the first CODESX has a value other than a 1. My question is, how do I start a NEW Case ID in this circumstance. As always, your assistance is greatly appreciated. I have included a sample dataset and syntax for your information. Cheers, Damir DATA LIST FREE / RPUID (F4) CASE_ID (A17) CASE_FILE_DATEX (DATE11) COURT_ORDER_DATEX (DATE11) CODESCX (F3). BEGIN DATA. 6294 2005MHXXXX1 18-Apr-2005 18-Apr-2005 1.00 6294 2005MHXXXX1 18-Apr-2005 11-May-2005 2.00 6294 2005MHXXXX2 27-May-2005 27-May-2005 3.00 6294 2005MHXXXX2 27-May-2005 01-Jun-2005 4.00 6294 2005MHXXXX2 27-May-2005 14-Jul-2005 4.00 6294 2005MHXXXX2 27-May-2005 27-Sep-2005 6.00 6294 2005MHXXXX2 27-May-2005 27-Sep-2005 6.00 6294 2005MHXXXX2 27-May-2005 05-Dec-2005 2.00 6294 2005MHXXXX2 27-May-2005 11-Jan-2006 6.00 6294 2005MHXXXX2 27-May-2005 11-Jan-2006 6.00 6294 2005MHXXXX2 27-May-2005 07-Mar-2006 6.00 6294 2005MHXXXX2 27-May-2005 20-Mar-2006 2.00 6294 2005MHXXXX2 27-May-2005 25-Apr-2006 2.00 6294 2005MHXXXX2 27-May-2005 19-Jul-2006 6.00 6294 2006MHXXXX3 12-Oct-2006 12-Oct-2006 1.00 6294 2006MHXXXX3 12-Oct-2006 28-Oct-2006 2.00 6294 2006MHXXXX4 28-Oct-2006 28-Oct-2006 3.00 6294 2006MHXXXX4 28-Oct-2006 02-Dec-2006 6.00 6294 2006MHXXXX5 29-Dec-2006 29-Dec-2006 1.00 6294 2006MHXXXX5 29-Dec-2006 04-Jan-2007 2.00 6294 2007MHXXXX6 31-Jan-2007 31-Jan-2007 3.00 6294 2007MHXXXX6 31-Jan-2007 09-Feb-2007 4.00 6294 2010MHXXXX8 31-Jan-2010 09-Feb-2010 6.00 1234 2012MHXXXX9 31-Jan-2012 09-Feb-2012 1.00 END DATA. DATASET NAME EXAMPLE. SORT CASES BY RPUID (A) COURT_ORDER_DATEX (A) CODESCX (A). DO IF CODESCX=1. COMPUTE CASEIDNX=$CASENUM. END IF. EXECUTE. IF MISSING (CASEIDNX) CASEIDNX = lag (CASEIDNX). VARIABLE LABELS CASEIDNX 'Unique CASES X ID (SPSS)'. EXECUTE. SORT CASES BY RPUID CASE_ID COURT_ORDER_DATEX (A). LIST RPUID CASE_ID CASE_FILE_DATEX COURT_ORDER_DATEX CODESCX CASEIDNX. EXECUTE. |
Administrator
|
SORT CASES BY RPUID (A) COURT_ORDER_DATEX (A) CODESCX (A).
MATCH FILES / FILE * / BY RPUID COURT_ORDER_DATEX / FIRST=@TOP@. DO IF @TOP@ . + COMPUTE CASEIDNX=$CASENUM. ELSE . + COMPUTE CASEIDNX = lag (CASEIDNX). END IF. EXECUTE. VARIABLE LABELS CASEIDNX 'Unique CASES X ID (SPSS)'. ......
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me. --- "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis." Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?" |
|
|
Thank you David,
I have slightly revised your syntax - which as always is more to the point - compared to my long round-about syntax! However, I am still encountering the same delimma. When the "new case" begins with CODESCX value other than a 1, the $casenum assigns the same number to the following CASE_ID, although, in this case the CASE_ID indicates its a new case. From the output (using the syntax below): I am trying to create CASEIDNX with the following values for the specific row numbers (RPUID and CASE_ID). Thanks, Damir David's proposed syntax with my minor revision: SORT CASES BY RPUID (A) COURT_ORDER_DATEX (A) CODESCX (A). MATCH FILES / FILE * / BY RPUID CASE_ID / FIRST=@TOP@. DO IF @TOP@ . + COMPUTE CASEIDNX=$CASENUM. ELSE . + COMPUTE CASEIDNX = lag (CASEIDNX). END IF. EXECUTE. VARIABLE LABELS CASEIDNX 'Unique CASES X ID (SPSS)'. RPUID CASE_ID CASE_FILE_DATEX COURT_ORDER_DATEX CODESCX CASEIDNX 1234 2012MHXXXX9 31-JAN-2012 09-FEB-2012 1 1.00 6294 2005MHXXXX1 18-APR-2005 18-APR-2005 1 2.00 6294 2005MHXXXX1 18-APR-2005 11-MAY-2005 2 2.00 6294 2005MHXXXX2 27-MAY-2005 27-MAY-2005 3 2.00 6294 2005MHXXXX2 27-MAY-2005 01-JUN-2005 4 2.00 6294 2005MHXXXX2 27-MAY-2005 14-JUL-2005 4 2.00 6294 2005MHXXXX2 27-MAY-2005 27-SEP-2005 6 2.00 6294 2005MHXXXX2 27-MAY-2005 27-SEP-2005 6 2.00 6294 2005MHXXXX2 27-MAY-2005 05-DEC-2005 2 2.00 6294 2005MHXXXX2 27-MAY-2005 11-JAN-2006 6 2.00 6294 2005MHXXXX2 27-MAY-2005 11-JAN-2006 6 2.00 6294 2005MHXXXX2 27-MAY-2005 07-MAR-2006 6 2.00 6294 2005MHXXXX2 27-MAY-2005 20-MAR-2006 2 2.00 6294 2005MHXXXX2 27-MAY-2005 25-APR-2006 2 2.00 6294 2005MHXXXX2 27-MAY-2005 19-JUL-2006 6 2.00 6294 2006MHXXXX3 12-OCT-2006 12-OCT-2006 1 3.00 6294 2006MHXXXX3 12-OCT-2006 28-OCT-2006 2 3.00 6294 2006MHXXXX4 28-OCT-2006 28-OCT-2006 3 3.00 6294 2006MHXXXX4 28-OCT-2006 02-DEC-2006 6 3.00 6294 2006MHXXXX5 29-DEC-2006 29-DEC-2006 1 4.00 6294 2006MHXXXX5 29-DEC-2006 04-JAN-2007 2 4.00 6294 2007MHXXXX6 31-JAN-2007 31-JAN-2007 3 4.00 6294 2007MHXXXX6 31-JAN-2007 09-FEB-2007 4 4.00 6294 2010MHXXXX8 31-JAN-2010 09-FEB-2010 6 5.00 |
|
|
Dear Listers,
Sorry, but upon further review of my post and explanation, I noticed that I should further clarify the following. The new definition of case can inlude a maximum of two unique CASE_IDs; therefore, I am looking to identify all those cases where subsequent CASE_ID (third) starts with a CODESCX that does NOT equal 1. Sorry about the confusion my description may have caused. Damir |
«
Return to SPSSX Discussion
|
1 view|%1 views
| Free forum by Nabble | Edit this page |