Date:         Mon, 17 May 1999 08:34:03 EDT
Reply-To:     Jamstat@AOL.COM
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Jamstat@AOL.COM
Subject:      Interleaving data sets and deleting observations
Content-Type: text/plain; charset="us-ascii"

Hi SAS-L'ers,
 
I have a problem that I don't know quite how to handle.  I have two large SAS
data sets (over a 5 million records each).  The first data set has previous
months data.  The second data set has the current months transactions.  Both
data sets are sorted by an account number and date of the tranaction.  The
problem I am having is that I need to combine the two data sets but if the
current data set has a transaction type of CLSD then all of transactions for
that account that are older get deleted.
 
         CURRENT
 
ACCT   DATE      TYPE    VAR4-VAR67
2      990401    CHRG    .....
2      990411    BILL
2      990415    PAID
2      990420    CHRG
3      990402    CHRG
3      990403    CHRG
3      990422    PAID
4      990415    OPEN
4      990422    CHRG
5      990411    PAID
5      990412    CLSD
6      990401    BILL
6      990411    CHRG
6      990412    PAID
6      990413    CLSD
6      990422    OPEN
6      990423    CHRG
7      990402    OPEN
 
          PREVIOUS
 
ACCT   DATE      TYPE    VAR4-VAR67
1      990101    OPEN     ....
1      990115    CHRG
1      990201    BILL
1      990301    BILL
2      990301    OPEN
3      990303    OPEN
3      990315    CHRG
3      990322    BILL
5      990311    OPEN
5      990322    CHRG
6      990311    OPEN
6      990330    CHRG
 
        RESULT
 
ACCT   DATE      TYPE    VAR4-VAR67
1      990101    OPEN     ....
1      990115    CHRG
1      990201    BILL
1      990301    BILL
2      990301    OPEN
2      990401    CHRG
2      990411    BILL
2      990415    PAID
3      990303    OPEN
3      990315    CHRG
3      990322    BILL
3      990402    CHRG
3      990403    CHRG
3      990422    PAID
4      990415    OPEN
4      990422    CHRG
6      990422    OPEN
6      990423    CHRG
7      990402    OPEN
 
Because there are so many transactions, I need to make this as efficient as
possible.  I'm using SAS 6.09 MVS and the datasets are all on tape.  I
figured out how to do it with2 passes off each data set but I would like to
do this in fewer passes.  Any suggestions would be greatly appreciated.
 
Thanks in advance,
Jarrie