SQL Select 最后 6 个月来自在列中保存年份和月份的表

Question

I have a table that saves year and month in columns and I need to get the top 5 grouping by an ID in the last 6 month.

I can easily get the information for a particular month an year but how can I get the same ranking of the top 5 for the last 6 month?

CREATE TABLE mytable (
table_id INT,  
some_id INT, 
some_other_id INT, 
mt_month INT, 
mt_year INT);

INSERT INTO mytable VALUES
("101","17","370","12","2021"),
("102","17","370","1","2021"),
("104","14","371","2","2021"),
("105","14","371","11","2021"),
("107","14","371","12","2021"),
("108","12","372","2","2021"),
("109","12","372","3","2021"),
("111","13","372","12","2021"),
("113","18","373","10","2021"),
("115","18","373","1","2021"),
("117","18","373","4","2021"),
("119","11","373","5","2021"),
("120","11","373","5","2021"),
("121","12","373","6","2021"),
("122","13","373","6","2021"),
("123","13","373","6","2021"),
("124","12","373","7","2021"),
("125","11","373","7","2021"),
("126","11","373","7","2021"),
("127","12","373","8","2021"),
("128","11","373","8","2021"),
("129","11","373","8","2021"),
("130","13","373","9","2021"),
("131","11","373","10","2021"),
("132","11","373","11","2021"),
("133","11","373","12","2021");

select some_id, min(mt_month), min(mt_year), count(table_id) as number from mytable group by some_id order by number DESC LIMIT 5;

https://www.db-fiddle.com/f/iRmvLZs9L8Tk1zJFgTMusX/0

Answer 1

Following approach should work:

1. Filter the data for the last 6 months in the subquery (can be done by concatenating year and month and applying the relevant filter)
2. Use your logic of grouping in the outer query.

Answer 2

This is what I came up with quickly. Made a variable to hold the wanted info and applied your same logic. Just specify time manually.

WITH a AS (
SELECT * FROM mytable
WHERE mt_year = 2021 and
mt_month BETWEEN 7 and 12 
)
select some_id, min(mt_month), min(mt_year), count(table_id) as number from a 
group by some_id
order by number DESC LIMIT 5;

This might not run but something like this should do the trick for february.

WITH a AS (
SELECT * FROM mytable
WHERE mt_year = 2020 and
mt_month BETWEEN 9 and 12 
),
b AS (  
SELECT * FROM mytable
WHERE mt_year = 2021 and
mt_month BETWEEN 1 and 2 
)
select some_id, min(mt_month), min(mt_year), count(table_id) as number from a 
UNION ALL
select some_id, min(mt_month), min(mt_year), count(table_id) as number from b
group by some_id order by number DESC LIMIT 5;

Easiest way is to use a regular date format that sql understands from the beginning.

Answer 3

The below query returns the last 6 months' values based on the current date. If I understand correctly you are trying something like:

SELECT table_id, some_id, some_other_id,mt_year,mt_month
FROM (
SELECT table_id, some_id, some_other_id, 
  mt_year , 
  mt_month ,
  @order_rank := IF(@current_month = mt_month,
  @order_rank + 1, 1) AS order_rank,
  @current_month := mt_month 
FROM mytable
ORDER BY   mt_year ,mt_month , some_id DESC ) ranked_orders
WHERE order_rank <= 5
and
mt_month BETWEEN MONTH((CURRENT_DATE() - INTERVAL 6 MONTH)) AND MONTH(CURRENT_DATE());

Demo: https://www.db-fiddle.com/f/iRmvLZs9L8Tk1zJFgTMusX/2

Try on your data and let me know.