My data looks like,
Table - usr_weight
user_id | weight | log_time |
---|---|---|
1. | 10 | 2021-11-30 10:29:03 |
1. | 12 | 2021-11-30 12:29:03 |
1. | 11 | 2021-11-30 14:29:03 |
1. | 18 | 2021-12-01 08:29:03 |
1. | 12 | 2021-12-15 13:29:03 |
1. | 14 | 2021-12-15 17:29:03 |
Here, I have duplicates for each date with different time. So, group date and return the record with max time for each date.
Query
select weight, log_time from usr_weight where user_id = 1 group by DATE(log_time)
Here, I get 1 record for each date, but the row is not by max(log_time).
CodePudding user response:
Using ROW_NUMBER
we can try:
WITH cte AS (
SELECT *, ROW_NUMBER() OVER (PARTITION BY DATE(log_time)
ORDER BY log_time DESC) rn
FROM usr_weight
WHERE user_id = 1
)
SELECT user_id, weight, log_time
FROM cte
WHERE rn = 1;
Here is an old school join way of doing this:
SELECT uw1.user_id, uw1.weight, uw1.log_time
FROM usr_weight uw1
INNER JOIN
(
SELECT DATE(log_time) AS log_time_date, MAX(log_time) AS max_log_time
FROM usr_weight
WHERE user_id = 1
GROUP BY DATE(log_time)
) uw2
ON uw2.log_time_date = DATE(uw1.log_time) AND
uw2.max_log_time = uw1.log_time
WHERE
uw1.user_id = 1;