How to select certain date from datetime datas using clickhouse? - clickhouse

Question

I have multiple date time in my clickhouse DB and I'm trying to use 'group by' option in my query to select the dates by certain interval. For example, my DB: (the column's type is DateTime64(3, 'UTC')
worktime
--------------------
2020-01-11 14:57:24.000
2020-01-11 16:57:24.000
2020-01-11 17:56:36.000
2020-01-11 18:04:36.000
2020-01-11 19:07:48.000
2020-01-11 20:46:48.000
2020-01-12 02:23:00.000
2020-01-12 02:35:00.000
2020-01-12 02:44:12.000
2020-01-12 02:46:24.000
2020-01-12 03:07:12.000
2020-01-12 03:26:00.000
2020-01-12 03:29:00.000
2020-01-12 03:37:00.000
2020-01-13 03:54:48.000
2020-01-13 04:15:24.000
2020-01-13 04:35:36.000
2020-01-13 04:46:12.000
2020-01-13 05:08:00.000
2020-01-13 05:13:00.000
2020-01-13 05:21:12.000
2020-01-13 05:32:24.000
2020-01-13 03:54:48.000
2020-01-13 04:15:24.000
2020-01-13 04:35:36.000
2020-01-13 04:46:12.000
2020-01-13 05:08:00.000
2020-01-14 05:13:00.000
2020-01-14 05:21:12.000
2020-01-14 05:32:24.000
2020-01-14 03:54:48.000
2020-01-15 04:15:24.000
2020-01-15 04:35:36.000
2020-01-15 04:46:12.000
2020-01-15 05:08:00.000
2020-01-16 05:13:00.000
2020-01-16 05:21:12.000
2020-01-16 05:32:24.000
2020-01-16 03:54:48.000
2020-01-16 04:15:24.000
2020-01-17 04:35:36.000
2020-01-17 04:46:12.000
2020-01-17 05:08:00.000
2020-01-17 05:13:00.000
2020-01-18 05:21:12.000
2020-01-18 05:32:24.000
and the result I want is grouping the dates by certain interval
first I tried without any interval
SELECT worktime FROM test2 GROUP BY worktime
and it did gave me the result of grouped dates, but I want to add certain interval (like 3 days, 4days, 12 hours..)
I googled and found a way to do with 'interval' function but it doesn't work like how I wanted it to work..
SELECT worktime FROM test2 GROUP BY worktime - interval 3 day as worktime
but I don't get the 3 days interval result :( (I get the same result as selecting all the dates)
I'm sorry my question is so vague, I'm new to clickhouse and I'm pretty much lost :(
I want the result of grouped dates with interval (2 days for example)
2020-01-11, 2020-01-13, 2020-01-15 . . .

Answer 1

Try this query:
SELECT
-- toStartOfInterval(worktime, INTERVAL 3 DAY) interval
toStartOfInterval(worktime, INTERVAL 2 DAY) interval
FROM (
/* emulate the test dataset */
SELECT toDateTime64(dt, 3) AS worktime
FROM (
SELECT arrayJoin([
'2020-01-11 14:57:24.000',
'2020-01-11 16:57:24.000',
'2020-01-11 17:56:36.000',
'2020-01-11 18:04:36.000',
'2020-01-11 19:07:48.000',
'2020-01-11 20:46:48.000',
'2020-01-12 02:23:00.000',
'2020-01-12 02:35:00.000',
'2020-01-12 02:44:12.000',
'2020-01-12 02:46:24.000',
'2020-01-12 03:07:12.000',
'2020-01-12 03:26:00.000',
'2020-01-12 03:29:00.000',
'2020-01-12 03:37:00.000',
'2020-01-13 03:54:48.000',
'2020-01-13 04:15:24.000',
'2020-01-13 04:35:36.000',
'2020-01-13 04:46:12.000',
'2020-01-13 05:08:00.000',
'2020-01-13 05:13:00.000',
'2020-01-13 05:21:12.000',
'2020-01-13 05:32:24.000',
'2020-01-13 03:54:48.000',
'2020-01-13 04:15:24.000',
'2020-01-13 04:35:36.000',
'2020-01-13 04:46:12.000',
'2020-01-13 05:08:00.000',
'2020-01-14 05:13:00.000',
'2020-01-14 05:21:12.000',
'2020-01-14 05:32:24.000',
'2020-01-14 03:54:48.000',
'2020-01-15 04:15:24.000',
'2020-01-15 04:35:36.000',
'2020-01-15 04:46:12.000',
'2020-01-15 05:08:00.000',
'2020-01-16 05:13:00.000',
'2020-01-16 05:21:12.000',
'2020-01-16 05:32:24.000',
'2020-01-16 03:54:48.000',
'2020-01-16 04:15:24.000',
'2020-01-17 04:35:36.000',
'2020-01-17 04:46:12.000',
'2020-01-17 05:08:00.000',
'2020-01-17 05:13:00.000',
'2020-01-18 05:21:12.000',
'2020-01-18 05:32:24.000']) AS dt))
GROUP BY interval
ORDER BY interval
/*
-- INTERVAL 3 DAY
┌────────────interval─┐
│ 2020-01-09 00:00:00 │
│ 2020-01-12 00:00:00 │
│ 2020-01-15 00:00:00 │
│ 2020-01-18 00:00:00 │
└─────────────────────┘
-- INTERVAL 2 DAY
┌────────────interval─┐
│ 2020-01-11 00:00:00 │
│ 2020-01-13 00:00:00 │
│ 2020-01-15 00:00:00 │
│ 2020-01-17 00:00:00 │
└─────────────────────┘
*/