Question

I’m working on a ECG data that basically has the following schema

Col 1 = TIMESTAMP
Col 2 = PATIENTID
Col 3 = ECGVALUE

Now, I’m trying to write a SQL statement that should be able to select all the rows that satisfy the following condition

Row index >= n and TIMESTAMP of xth row <= TIMESTAMP of nth row + offset

To explaing it further, let say I’ve following data in my database

1.1 ANON 1.1
1.3 ANON 2.3
3.5 ANON 4.3
5.0 ANON 6.5
6.3 ANON 7.5
7.9 ANON 8
8.6 ANON 9.4

Now, I want to select the data from 3rd row till 3 seconds of data has been collected, which means that my resultset should have

3.5 ANON 2.3 *//3rd row till TIMESTAMP <= 3.5 + 3 <= 6.5*
1.3 ANON 2.3
3.5 ANON 4.3
5.0 ANON 6.5
6.3 ANON 7.5

Last two rows are neglected as difference between TIMESTAMP of first and last cannont go beyond 3. So If I go back to my condition, which is

Row index >= n and TIMESTAMP of xth row <= TIMESTAMP of nth row + offset

Here,

n: nth row from where data must be selected
x: Any arbitary row in result set
offset: Maximum difference between first and last TIMESTAMP of result set.

I’ve written a working SQL statement for the above condition, but I think its not that much optimized as I’m new to SQL.

SELECT TIMESTAMP, ECGVALUE
FROM
(
    SELECT TIMESTAMP, ECGVALUE, ROW_NUMBER() OVER() AS RN
    FROM EKLUND.DEV_RAWECG
)
WHERE RN >= n AND TIMESTAMP <=
(
    SELECT TIMESTAMP FROM
    (
        SELECT TIMESTAMP, ROW_NUMBER() OVER() AS TM
    )
    WHERE TM = n
) + offset;

Answer 1

The code will look a lot cleaner if you use a common table expression, as follows:

WITH rownums(TIMESTAMP, ECGVALUE, RN) AS (
    SELECT TIMESTAMP, ECGVALUE, ROW_NUMBER() OVER(ORDER BY TIMESTAMP) AS RN
    FROM EKLUND.DEV_RAWECG
)
SELECT allrows.TIMESTAMP, allrows.ECGVALUE, allrows.RN
FROM rownums allrows
     CROSS JOIN rownums rown
WHERE rown.RN = n
      AND allrows.RN >= n
      AND rown.TIMESTAMP + offset >= allrows.TIMESTAMP;