如何在 Postgres SQL 中传播/分布一个数字

Question

I have a table as follows:

id    rank    user
1      1       1
2      1       1
3      1       2
4      2       1
5      2       1

I would like to spread the rank value for a given user between 1 (or 0) and a max value to leave space between values. For example, I would like to spread the rank for user 1 , with max value of 100 (in this case), the result should be:

id    rank    user
1      20      1
2      40      1
3      1       2
4      60      1
5      80      1

I thought to calculate an increment = maxValue / count(items) + 1 and then do many UPDATE queries using the increment for each new one, but that seems extremely inefficient. Is there a way to do it better, maybe with one big UPDATE query, something like:

UPDATE MyTable
SET rank = increment * ??? (?current position? in SELECT * FROM MyTable WHERE user = 1 ORDER BY rank)
WHERE user = 1

Answer 1

You could achieve this result by using a CTE to generate a row number ( rn ) for each entry for a user (in order of rank with id to order duplicates) and also a count ( cnt ) of the number of rows for that user, and then set the rank for that user to be:

rn * 100 / (cnt + 1)

As a query:

WITH CTE AS (
  SELECT id, user, 
         ROW_NUMBER() OVER (PARTITION BY user ORDER BY rank, id) AS rn,
         COUNT(*) OVER (PARTITION BY user) AS cnt
  FROM MyTable
  WHERE "user" = 1
)
UPDATE MyTable
SET rank = CTE.rn * 100 / (CTE.cnt + 1)
FROM CTE
WHERE MyTable.id = CTE.id

Resultant table:

id  rank    user
1   20      1
2   40      1
3   1       2
4   60      1
5   80      1

Demo on SQLFiddle