Is it possible to LEFT JOIN a #variable table in SQL - mysql

Question

Summary
I am attempting to LEFT JOIN on a filtered TABLE in SQL.
To do so I am first declaring a temporary table in an # variable, then attempting to join on it later.
Unfortunately I am not having much luck doing this, it massively speeds up my query when limiting to such a resultset.
Other Routes Tried
I initially was trying to conduct this on the WHERE, however this was preventing rows from occurring where there were no matching events (I am counting the events by intervals_days so I need to return regardless of whether there's matching in the other).
After I realised my mistake, I moved to the ON. I have not seen examples of this done, and I have a feeling doing a FIND_IN_SET here on a SELECT would not be performant?
QUERIES
The Initial Filter
DECLARE #filtered_events TABLE (id INT, eventable_id INT, eventable_type VARCHAR(255), occurred_at TIMESTAMP, finished_at TIMESTAMP)
INSERT INTO #filtered_events
SELECT
e.id, e.eventable_id, e.eventable_type, e.occurred_at, e.finished_at
FROM
units_events as ue
LEFT JOIN
events as e
ON
e.id = ue.event_id
WHERE
ue.unit_id
IN
(1,2,3);
Ignore the (1,2,3) here, these values are added to the query dynamically.
My Attempted Use Of It
SELECT
i.starts_at as starts_at,
i.ends_at as ends_at,
count(fe.id) as count
FROM
intervals_days as i
LEFT JOIN
#filtered_events as fe
ON
( fe.occurred_at >= starts_at AND fe.occurred_at < ends_at )
WHERE
( starts_at >= '15-01-2019' AND ends_at < NOW() )
GROUP BY
starts_at
ORDER BY
starts_at
DESC;
These queries are within one SQL document, one above the other with the terminating ; semicolon on each.
I expect this to output what my lower query outputs (a grouped resultset by the intervals_days rows)- however with the benefit of my LEFT JOIN query being conducted on a much smaller sample.

Answer 1

Sorry, no such syntax.
MySQL has no concept of arrays, either.
Nor can you use DECLARE outside of a Stored Routine.

Answer 2

As Shadow said in the comments, with MySQL I would use a subquery like this:
SELECT
i.starts_at as starts_at,
i.ends_at as ends_at,
count(fe.id) as count
FROM
intervals_days as i
LEFT JOIN
(
SELECT
e.id,
e.eventable_id,
e.eventable_type,
e.occurred_at,
e.finished_at
FROM units_events as ue
LEFT JOIN events as e
ON e.id = ue.event_id
WHERE ue.unit_id IN (1,2,3)
) fe
ON ( fe.occurred_at >= starts_at AND fe.occurred_at < ends_at )
WHERE ( starts_at >= '15-01-2019' AND ends_at < NOW() )
GROUP BY starts_at
ORDER BY starts_at
DESC;