Insert new data BigQuery - sql

Question

I have this table in BigQuery:
+------------+---------+------+
| date | country | sum |
+------------+---------+------+
| 2020-01-01 | UK | 10 |
| 2020-01-01 | Spain | 34 |
| 2020-01-01 | Germany | 78 |
| 2020-01-01 | France | 81 |
+------------+---------+------+
With the UI of BigQuery, I edited the schema and created a new metric: AVG. I would like to insert this new info in this table from another table like this:
INSERT dataset.table_old (AVG)
SELECT AVG(m) FROM table_m
If I do that I get it this:
+------------+---------+------+------+
| date | country | sum | AVG |
+------------+---------+------+------+
| 2020-01-01 | UK | 10 | NULL |
| 2020-01-01 | Spain | 34 | NULL |
| 2020-01-01 | Germany | 78 | NULL |
| 2020-01-01 | France | 81 | NULL |
| NULL | NULL | NULL | 28 |
| NULL | NULL | NULL | 7 |
| NULL | NULL | NULL | 10 |
| NULL | NULL | NULL | 41 |
+------------+---------+------+------+
How could I get the correct table with the correspondent match?
Thanks!

Answer 1

Solution for your problem in that you need to apply UPDATE command rather than INSERT command. If you apply INSERT then it will treat it as a new Records.
Expected Solution:-
Step 1: You need to Identify over which field you are getting AVG(m) from table table_m, that can be joined with table table_old. It Could be (date, country or sum) based on your business requirement.
Step 2: Once you Joining Field in above step, then you can write following UPDATE query:-
UPDATE dataset.table_old a
SET AVG = (SELECT AVG(m) FROM table_m b where a.JOINING_Coll=b.JOINING_Coll GROUP BY Field)
WHERE 1=1
Important part is a.JOINING_Coll=b.JOINING_Coll where you need to identify which filed will be used to join 2 tables and perform UPDATE.

Answer 2

You could use a BQ MERGE statement.
add a new column avg in table_od with NULL value
Then do a MERGE operation below
Assuming you need to calculate the average value in table_m for each date and country
MERGE INTO table_old AS dest
USING(
SELECT
a.date,
a.country,
a.sum,
b.avg
FROM table_old AS a
LEFT JOIN
(
SELECT
AVG(m) as avg
FROM table_m
GROUP BY date, country
) AS b
USING (date, country)
) AS source
ON source.date = dest.date AND source.coutry = dest.country
WHEN MATCHED THEN UPDATE SET
date = dest.date, country = dest.country, sum = dest.sum, avg = dest.avg
For more information about BQ MERGE statement