MySQL: Count(*) NULL -> 0 with right join

Question

我即将完成一道 LeetCode 难题。假设 driver 或客户都未被禁止使用拼车服务，我们的想法是找到某个日期范围内的取消率（百分比）。

我运行遇到的问题是 10 月 2 日没有取消；使用内部联接时，该日期将被省略。使用右连接，该值返回为 NULL。我需要将此 NULL 值映射到 0。

问题：https://leetcode.com/problems/trips-and-users/

我的代码：


SELECT output2.day AS "Day", ROUND(output1.failed/output2.success,2) AS "Cancellation Rate"
FROM

((SELECT t1.request_at AS Day, COUNT(*) AS failed
FROM Trips T1
WHERE T1.Status != 'completed'
AND (T1.request_at BETWEEN '2013-10-01' AND '2013-10-03')
AND T1.client_id IN 
  (SELECT users_id 
   FROM users 
   WHERE role = 'client'
   AND banned = 'No')
AND t1.driver_id IN
    (SELECT users_id 
   FROM users U2 
   WHERE role = 'driver'
   AND banned = 'No')

GROUP BY t1.request_at
) AS output1

RIGHT JOIN

(SELECT t2.request_at AS Day, COUNT(*) AS success
FROM Trips T2
WHERE (T2.request_at BETWEEN '2013-10-01' AND '2013-10-03')
AND T2.client_id IN 
  (SELECT users_id 
   FROM users 
   WHERE role = 'client'
   AND banned = 'No')
AND t2.driver_id IN
    (SELECT users_id 
   FROM users 
   WHERE role = 'driver'
   AND banned = 'No')

GROUP BY t2.request_at
) AS output2

ON output1.Day = output2.Day)

GROUP BY output2.day

我的输出：

{"headers": ["Day", "Cancellation Rate"], "values": [["2013-10-01", 0.33], ["2013-10-02", null], ["2013-10-03", 0.50]]}

期望的输出：

{"headers": ["Day", "Cancellation Rate"], "values": [["2013-10-01", 0.33], ["2013-10-02", 0.00], ["2013-10-03", 0.50]]}

我看过各种社区推荐，例如使用函数 ISNULL 或 IFNULL。但是都返回了一个错误。有什么想法吗？

Answer 1

您可以在最外层的查询中使用 COALESCE() 将 NULL 值转换为 0s:

SELECT 
    output2.day AS "Day", 
    COALESCE(ROUND(output1.failed/output2.success,2), 0) AS "Cancellation Rate"
FROM ...

但是请注意，使用联接和条件聚合可以大大简化您的查询。像这样的东西应该接近你想要的：

select 
    t.request_at as day, 
    avg(t.status != 'completed') as cancellation_rate
from trips t
inner join users uc 
    on uc.users_id = t.client_id and uc.role = 'client' and uc.banned = 'no'
inner join users ud 
    on ud.users_id = t.client_id and ud.role = 'driver' and ud.banned = 'no'
where t.request_at between '2013-10-01' and '2013-10-03'
group by t.request_at

MySQL: Count(*) NULL -> 0 with right join

MySQL: Count(*) NULL -> 0 with right join

mysql

null

join

right-join