How to get average value for each hourly increment that is split into 5 minute intervals - sql

Question

I have an AWS Redshift table that looks like this:
interval_date
interval_time
power
on_status
2022-05-01
00:00
2.65
Y
2022-05-01
00:05
3.92
Y
2022-05-01
00:10
2.05
Y
2022-05-01
00:15
1.85
Y
2022-05-01
00:20
5.92
Y
2022-05-01
00:25
7.52
Y
2022-05-01
00:30
9.84
Y
2022-05-01
00:35
6.84
N
2022-05-01
00:40
5.01
N
2022-05-01
00:45
4.70
N
2022-05-01
00:50
8.57
N
2022-05-01
00:55
1.94
N
2022-05-01
01:00
3.87
Y
The table continues with more timestamps going all the way until 11:55 PM for any given day up to the current day/time. I am trying to get the average value of power for each hourly interval (so the average for 12 AM should be the values from the previous day (4/30/2022) at 23:05 to current day (5/1/2022) at 00:00, 1 AM is 00:05 to 01:00, 2 AM is 01:05 to 02:00, etc) where the on_status equals Y.
I have a basic query that gets me the average for a whole day (for context, the interval_date would be parameterized).
SELECT AVG(power)
FROM table
WHERE on_status = 'Y'
AND interval_date = '2022-05-01';
I am unsure how to partition the interval_time column so that the values are averaged hourly. An idea of the final result I am looking for is:
interval_date
interval_time
power
on_status
2022-05-01
00:00
2.65
Y
2022-05-01
01:00
5.00
Y
2022-05-01
02:00
X
Y
2022-05-01
03:00
X
Y

Answer 1

You didn't specify the type for interval_time, so I'm assuming a string, you can parse it out with a case statement like this:
SELECT interval_date,
CASE WHEN SUBSTRING(interval_time,4,2)='00' THEN interval_time
WHEN SUBSTRING(interval_time,1,2)='23' THEN '00:00'
ELSE FORMAT(convert(int,SUBSTRING(interval_time,1,2))+1,'00')+':00'
END interval_time,
AVG(power)
FROM mytable
WHERE on_status = 'Y'
GROUP BY interval_date,
CASE WHEN SUBSTRING(interval_time,4,2)='00' THEN interval_time
WHEN SUBSTRING(interval_time,1,2)='23' THEN '00:00'
ELSE FORMAT(convert(int,SUBSTRING(interval_time,1,2))+1,'00')+':00'
END
Note that to get your target of 5.17, I had to comment out the on_status = 'Y' filter.
https://dbfiddle.uk/?rdbms=sqlserver_2019&fiddle=b604e4fe6696465aac75676e69b92a47