Difference between dates stored in Same column

Hi Experts
I am actually stuck with a logic problem on a SQL query in Oracle.  
I need the difference between two dates (that are in date datatypes) that are in same column.
I tried using the ROW_NUMBER () OVER (PARTITION BY) concept but its yielding incorrect values at some places. The "Days in each status" output column is what I am looking for.  The status id is the column which has few set of statuses which get repeated for each unique person.

STATUS_CHANGE_DATE      status_id      Days in each status
            
4/9/2009 6:51      220      0 (there is no value above it so it should be 0)
4/6/2009 7:57      114      4 (diff between the 4/9 and 4/6 incl both the days)
4/6/2009 7:56      6      1 (diff between 4/6 =0 but should consider it as1)
4/6/2009 7:56      5      1
3/30/2009 14:30      111      8 (diff between 4/6 and t3/30)
3/26/2009 9:04      208      5 (diff between 3/30 and 3/26)
3/25/2009 16:17      49      2 (diff between 3/26 and 3/25)

Thanks
LVL 8
gvsbnarayanaAsked:
Who is Participating?
 
lwadwellConnect With a Mentor Commented:
gvsbnarayana,

Try this simple SQL first ...

BTW.  Are you sure the partitioning key is assignment_status_type_id?  These values must be the same across the multiple rows to form a partion.  Your examples in the original question had the values as different for each row.

lwadwell

SELECT assignment_status_id, 
       assignment_id,
       assignment_status_type_id,
       status_change_date, 
       LAG (status_change_date,1,null) OVER (PARTITION BY assignment_id ORDER BY status_change_date) prev_date
  FROM apps.irc_assignment_statuses
 WHERE assignment_id = 649650
 ORDER BY assignment_status_id, 
       status_change_date

Open in new window

0
 
lwadwellCommented:
Hi gvsbnarayana,

The LAG() function would be better than the ROW_NUMBER() for this I think.  What is your PARTITION BY column(s)?  The LAG() function allows you to look back 'n' rows (the other one is LEAD() which allows you to look forward 'n' rows).  You could do something like

SELECT status_change_date, LAG(status_change_date,1,null) OVER(partition by some_column ORDER BY status_change_date) as prev_date
FROM ...

lwadwell
0
 
gvsbnarayanaAuthor Commented:
My partition column is the status_id as per the above example I provided. I tried using the lag but no luck.
0
Free Tool: Site Down Detector

Helpful to verify reports of your own downtime, or to double check a downed website you are trying to access.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

 
lwadwellCommented:
lwadwell,

please show me your SQL.

lwadwell
0
 
gvsbnarayanaAuthor Commented:
SELECT   (SELECT user_status
            FROM apps.per_assignment_status_types
           WHERE assignment_status_type_id =
                                           a.assignment_status_type_id)
                                                                       status,
         (a.status_change_date - b.status_change_date) AS days,        
         a.arowno,
         b.browno
    FROM (SELECT assignment_status_id, status_change_date, assignment_id,
                 assignment_status_type_id,
                 ROW_NUMBER () OVER (PARTITION BY assignment_status_type_id ORDER BY status_change_date)
                                                                       arowno
            FROM apps.irc_assignment_statuses) a,
         (SELECT assignment_status_id, status_change_date, assignment_id,
                 assignment_status_type_id,
                 ROW_NUMBER () OVER (PARTITION BY assignment_status_type_id ORDER BY status_change_date)
                                                                       browno
            FROM apps.irc_assignment_statuses) b
   WHERE a.assignment_status_type_id = b.assignment_status_type_id
     AND a.arowno - 1 = b.browno
     AND b.assignment_id = 649650
     order by a.status_change_date
0
 
gvsbnarayanaAuthor Commented:
The Assignment Status Type id gets repeated to each Assignment id. The Assignment Status table holds the lookup values for the statuses table. So, per defintion, to form a partition I considered the  assignment status type_id as the partition key. I think that was the mistake I did.

But as per the above query I got the values for the prev value as you considered the assignment id itself and I made my modifications accordingly.

Awesome...
Thanks for the idea.  
0
 
gvsbnarayanaAuthor Commented:
Thanks for showing me the right direction
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.