Help with query: Get MAX value of rows that previous row has certain value

oracle

My table TEST has three columns:

xID          xDATE       xVALUE
01134289    19/11/2009   1     
01134289    28/12/2009   1     
01134289    31/03/2010   0     
01134289    10/06/2010   0     
01134289    28/10/2010   0     
01134289    30/03/2012   1     
01134289    14/05/2014   1     
01134289    17/07/2014   0     
01134289    30/03/2015   1     
01134289    14/05/2015   1     
01134289    17/07/2015   0

I need help for a query to select the MAX xDATE of rows where xVALUE=1, but only for those rows which previous one row has xVALUE=0.

In this example, there are two rows with xValue=1 and previous rows has xValue=0:

xID          xDATE       xVALUE
...
01134289    28/10/2010   0     
01134289    30/03/2012   1     
...
01134289    17/07/2014   0     
01134289    30/03/2015   1

The query should output:

xID          xDATE       xVALUE
01134289    30/03/2015   1

Best Answer

Make use of LAG analytic function

SELECT xID, MAX(xDATE),xValue FROM
(
SELECT t.*,
       CASE
            WHEN xvalue = 1 AND LAG(xvalue) OVER(
                 PARTITION BY xid
                 ORDER BY xdate
            ) = 0 THEN 1
            ELSE 0
       END
  AS cond
FROM test t
) s WHERE s.cond = 1
GROUP BY xID,xValue

Demo

Related Solutions

How to select rows from a hierarchical query with the lowest level

If you have an hierarchical query that produces the whole tree under the root node, that also has a level column computed, you can wrap it in a derived table or cte and use the window aggregate:

WITH query AS
  ( SELECT <columns list>, level
  -- your query here
  ) ,
cte AS 
  ( SELECT <columns list>, level,
           MIN(the_level) OVER () AS min_level
    FROM query
    WHERE <conditions>
  )
SELECT *
FROM cte
WHERE min_level = level ;

Max from 2 columns with priority date col then the seq col Oracle

You don't mention DBMS, so I'll assume it is one that supports window functions (DB2, Oracle, SQL-server, Postgres, etc)

select personid, xdate, seqnumber
from (
     select personid, xdate, seqnumber
          , row_number() over (partition by personid
                               order by xdate desc, seqnumber desc) as rn
     from T
) as X
where rn = 1;

Edit: added sample data and result (http://sqlfiddle.com/#!4/73ad2/4)

create table T
(   personid   int
,  xdate      date
,  seqnumber  int );

ALTER SESSION SET NLS_DATE_FORMAT = 'YYYY-MM-DD';

insert into t values (123, '2012-01-01', 0);
insert into t values (123, '2011-01-01', 0);
insert into t values (136, '2012-01-02', 1);
insert into t values (136, '2012-01-02', 0);
insert into t values (136, '2012-01-01', 0);
insert into t values (168, '2012-01-04', 2);
insert into t values (168, '2012-01-04', 1);
insert into t values (168, '2012-01-04', 0);

select personid, xdate, seqnumber
from (
    select personid, xdate, seqnumber
         , row_number() over (partition by personid
                              order by xdate desc, seqnumber desc) as rn
    from T
) x
where rn = 1;

PERSONID    XDATE   SEQNUMBER
123     January, 01 2012 00:00:00+0000  0
136     January, 02 2012 00:00:00+0000  1
168     January, 04 2012 00:00:00+0000  2

Best Answer

Related Solutions

How to select rows from a hierarchical query with the lowest level

Max from 2 columns with priority date col then the seq col Oracle

Related Question