How to use the datebucket filter - snowflake-cloud-data-platform

Question

Trying to use the :datebucket filter but it doesn't seem to work.
select date, address from database.table where address = 'xyz' group by :datebucket(date)
This returns the error that date isn't in the group by statement, but it is. If it add it separately to the group by statement, it just groups by the individual date instead of respecting the date bucket selection.
Not finding anything in the Snowflake documentation about how this filter is suppose to work, just that it exists.

Answer 1

In this site: https://www.webagesolutions.com/blog/querying-data-in-snowflake was example like this about databucket function
SELECT COUNT(ORDER_DATE) as COUNT_ORDER_DATE, ORDER_DATE
FROM ORDERS
GROUP BY :datebucket(ORDER_DATE), ORDER_DATE
ORDER BY COUNT_ORDER_DATE DESC;
So could your query work if it was modified like this:
SELECT
date,
address
FROM
database.table
WHERE
address = 'xyz'
GROUP BY :datebucket(date), date

Answer 2

Datebucket is truncating the date, to buckets. But you have selected the raw date.
This is like grouping by decade '60,'70,'80 of what great years, but want the actual year.
SELECT column1 as year,
truncate(year,-1) as decade
FROM VALUES (1),(2),(3),(14),(15),(16),(27),(28),(29);
gives:
YEAR
DECADE
1
0
2
0
3
0
14
10
15
10
16
10
27
20
28
20
29
20
so if I try select
SELECT column1 as year
FROM VALUES (1),(2),(3),(14),(15),(16),(27),(28),(29)
GROUP BY truncate(year,-1)
ORDER BY 1;
gives the error
Error: 'VALUES.COLUMN1' in select clause is neither an aggregate nor in the group by clause. (line 15)
So if we move the decade into the selection, it makes sense:
SELECT truncate(column1,-1) as decade
FROM VALUES (1),(2),(3),(14),(15),(16),(27),(28),(29)
GROUP BY decade
ORDER BY 1;
and we get the
DECADE
0
10
20
So the problem is not :datebucket(date) but the fact while :datebucket(date) and date are related, from the perspective of GROUPING they are unrelated.

Answer 3

I've been trying to use datebucket(date) and daterange, and I also needed the results in a Snowflake graph.
It was a bit trick, because the value returned by datebucket(date) is actually a truncated date based on the selected date part. For that, I had to convert it to a char, and it worked!
select
to_char(:datebucket(start_time), 'YYYY.MM.DD # HH24') as start_time_bucket,
sum(credits_used) as credits_used
from snowflake.account_usage.warehouse_metering_history wmh
where
start_time = :daterange
group by :datebucket(start_time)
And if you're an ACCOUNTADMIN, you can now use the query to get the total credits usage by date :)
Last, to answer the main query by Tony, the query should be:
select date, address
from database.table
where address = 'xyz'
group by :datebucket(date), date, address
// or
select :datebucket(date), address
from database.table
where address = 'xyz'
group by :datebucket(date), address

Answer 4

Try adding the :datebucket(date) in the select part as well (not only in group by). Also, you will probably need an aggregate function for the field address (for example any_value(address):
select :datebucket(date), any_value(address)
from database.table
where address = 'xyz'
group by :datebucket(date)