Reputation: 339
I'm currently using StandardSQL in BigQuery, I tried to join two sets of table one of which is a pseudo-column table partitioned by day.
I tried to use this query below:
SELECT
DISTINCT DATE(create_time) AS date,
user_id,
city_name,
transaction_id,
price
FROM
table_1 a
LEFT JOIN (SELECT user_id, city_name FROM table_2) b
ON (a.user_id = b.user_id AND DATE(create_time) = _PARTITIONDATE)
I've tried this kind of JOIN (using _PARTITIONDATE) and worked out, but for this particular query I got an error message:
Unrecognized name: _PARTITIONDATE
Can anyone tell me why this happened, and how could I solve this? Thanks in advance.
Upvotes: 0
Views: 647
Reputation: 7058
The issue is that you are not selecting the _PARTITIONDATE
field from table_2
when joining it so it can't recognize it:
SELECT user_id, city_name FROM table_2
In order to solve it you can add it as follows:
SELECT
DISTINCT DATE(create_time) AS date,
user_id,
city_name,
transaction_id,
price
FROM
table_1 a
LEFT JOIN (SELECT _PARTITIONDATE AS pd, user_id, city_name FROM table_2) b
ON (a.user_id = b.user_id AND DATE(create_time) = pd)
Note that you'll need an alias such as pd
as it's a pseudocolumn
Probably it was working in the past if you were joining two tables directly such as in (you don't get selectivity benefits in that case):
FROM
table_1 a
LEFT JOIN table_2 b
ON (a.user_id = b.user_id AND DATE(create_time) = _PARTITIONDATE)
Upvotes: 1