sql请教如何解决这个问题？如何从饼图元数据库更改维度标签？

Question

我一直在尝试在元数据库问题中构建一个饼图，以放入我的仪表板。我有 player_id 和 updated_at 列，并且我执行了对 return 每天 updated_at 数量的咨询。我这样做有一个问题，我不能使用 order by，我不知道为什么，但这是行不通的。所以，我想知道我错在哪里。

SQL:

SELECT * FROM (
SELECT date(updated_at, '%Y-%m-%d') AS d, count(DISTINCT player_id) AS c FROM matchmaking_stats_summary WHERE updated_at between '2020-02-01 00:00:00' and '2020-02-01 23:59:59' union 
SELECT date(updated_at, '%Y-%m-%d') AS d, count(DISTINCT player_id) AS c FROM matchmaking_stats_summary WHERE updated_at between '2020-02-02 00:00:00' and '2020-02-02 23:59:59' union
SELECT date(updated_at, '%Y-%m-%d') AS d, count(DISTINCT player_id) AS c FROM matchmaking_stats_summary WHERE updated_at between '2020-02-03 00:00:00' and '2020-02-03 23:59:59' union
SELECT date(updated_at, '%Y-%m-%d') AS d, count(DISTINCT player_id) AS c FROM matchmaking_stats_summary WHERE updated_at between '2020-02-04 00:00:00' and '2020-02-04 23:59:59' union
SELECT date(updated_at, '%Y-%m-%d') AS d, count(DISTINCT player_id) AS c FROM matchmaking_stats_summary WHERE updated_at between '2020-02-05 00:00:00' and '2020-02-05 23:59:59' 
) AS x
ORDER BY d asc

列updated_at类型是时间戳。一个值的例子是：2020-02-01 00:47:32 另一个问题是为什么我不能在 Metabase 中设置 Dimession 和 Meassure，我想将 Dimession 标签设置为相应的日期间隔和 Meassure 标签中的计数。可能吗？

谢谢，祝你有个愉快的一天。

Answer 1

为什么不使用 group by？

SELECT date(updated_at, '%Y-%m-%d') AS d, 
       count(DISTINCT player_id) AS c
FROM matchmaking_stats_summary
WHERE updated_at >= '2020-02-01 00:00:00' and 
      updated_at < '2020-02-06 00:00:00' 
GROUP BY date(updated_at, '%Y-%m-%d') ;

Answer 2

我一直在更改我的 sql 代码，我终于找到了一种方法来处理日期间隔和玩家数量作为饼图元数据库维度和度量。

select date(updated_at), count(DISTINCT updated_at) as jogadores
    from matchmaking_stats_summary
        group by date(updated_at)
        order by date(updated_at) asc;

非常感谢。

sql请教如何解决这个问题？如何从饼图元数据库更改维度标签？

How to fix this order by sql consult? How to change dimension label from pie chart metabase?

sql

metabase

date