asked on

Consolidate Codes and Dates in proper order

I need to group a person with enterdates,entercodes,withdrawal dates and withdrawal codes in correct sequential (date) order. One person may have 10 records with sequential dates. See file. The 'Days Attended' field may not look right if you calculate the date diff...that's ok, I just need to put the number on the correct line for now. P.S. I don't create these files, just inherit them...
Thank you!

I started the query below, but don't have a clue how to put the E code & W codes on one line (see Results table example in xls file).

select a.lastname,a.firstname,a.middlename, effdate,entercode,leavecode,
DATEDIFF(day,effdate,
   (Select MIN(effdate) from Table1 b
    where b.lastname=a.lastname and b.firstname=a.firstname and b.effdate > a.effdate)) NoDays
from Table1 a
order by a.lastname,a.firstname,a.middlename

Open in new window

Enterdateleavedates.xls

patrikt

I see your problem. There are two solution.
1) Use "fake" agregate. If you GROUP BY a.lastname,a.firstname,a.middlename, effdate you can use MAX(entercode), MAX(leavecode), MAX(DATEDIFF.....) because only other value is NULL and agregate skips nulls.

2) Better one. Convert subquery to JOIN and extract leave code there. I'll try to write it for you, give me a sec.

Patrik

patrikt

There is my blind shot code.
Without testing I can only guess that it works. Test it and come back with result if you want.

BTW: Structure you have looks wery strange. Start thinking about database redesign if you want to avoid problems in future.

Patrik

select a.lastname,a.firstname,a.middlename, a.effdate,a.entercode,b.effdate), b.leavecode,
DATEDIFF(day,b.effdate,a.effdate) AS NoDays
from Table1 a
JOIN Table1 b
	ON  b.lastname=a.lastname and b.firstname=a.firstname and b.effdate =(SELECT MIN(effdate) FROM Table1 c WHERE c.effdate>a.effdate)
order by a.lastname,a.firstname,a.middlename

Open in new window

lbarnett419

ASKER

I tried the query(THANK YOU!) & changed the DATEDIFF(day,b.effdate,a.effdate) to (day,a.effdate,b.effdate) so the days would be positive days. Either way, I am only getting a subset of the records for each person. The John Smith record originally had:
LN      FN      EffDate      EnterCode      LeaveCode
Abbott      Joseph      11/13/2006      E      NULL
Abbott      Joseph      12/19/2006      NULL      W20
Abbott      Joseph      2/6/2007      E      NULL
Abbott      Joseph      2/16/2007      NULL      W10
Abbott      Joseph      2/19/2007      E      NULL
Abbott      Joseph      3/22/2007      NULL      W20

The query result was:
LN      FN      EffDate      EnterCode      EffDate      LeaveCode      NoofDays
Smith      John      2/16/2007      NULL      2/19/2007      NULL      3
Any ideas?

patrikt

I'll look on it.
But you send wrong example, mixed Joseph Abbot with John Smith.

Patrik

patrikt

I think mz query is ok, but restriction on start row is missing.
I suppose Enter row has LeaveCode NULL and Leave row has EnterCode NULL.
Try this one:

select a.lastname,a.firstname,a.middlename, a.effdate,a.entercode,b.effdate), b.leavecode,
DATEDIFF(day,b.effdate,a.effdate) AS NoDays
from Table1 a
JOIN Table1 b
	ON  b.lastname=a.lastname and b.firstname=a.firstname and a.LeaveCode IS NULL AND b.effdate =(SELECT MIN(effdate) FROM Table1 c WHERE c.effdate>a.effdate AND c.LeaveCode IS NOT NULL)
order by a.lastname,a.firstname,a.middlename

Open in new window

ASKER CERTIFIED SOLUTION

patrikt

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

patrikt

Please write comment why it is graded B. What was wrong on solution?