查询中的累积函数不会分组

Question

我正在尝试重置每种insurance_status的累积计数。现在，我的代码恢复了累计计数，但是当insurance_status更改时并没有重新开始。 如何设置查询的分组依据或重置变量以获取所需的查询。

我试图在代码中最后一个命令之前添加group by，但是它不起作用。

SELECT t.insurance_status, t.day, t.vendor_count,
       @running_total:=@running_total + t.vendor_count AS cumulative_sum
FROM (SELECT insurance_status,
             DATE_FORMAT(date, '%Y %m') as day,
             count(insurance_status) as vendor_count
      FROM vendor_report_jll
      WHERE project_id in (100, 92, 45, 91, 86, 87, 88, 40, 101, 90, 98)
      GROUP BY insurance_status,day
     ) t JOIN
     (SELECT @running_total := 0) AS t2
ORDER BY  t.insurance_status;

Answer 1

您需要另一个变量来记住保险状态：

SELECT t.insurance_status, t.day, t.vendor_count,
       (@sum := if(@is = t.insurance_status, @sum + t.vendor_count,
                   if(@is := t.insurance_status, t.vendor_count, t.vendor_count
                     )
                  )
       ) AS cumulative_sum
FROM (SELECT insurance_status,
             DATE_FORMAT(date, '%Y %m') as day,
             count(insurance_status) as vendor_count
      FROM vendor_report_jll
      WHERE project_id in (100, 92, 45, 91, 86, 87, 88, 40, 101, 90, 98)
      GROUP BY insurance_status, day
      ORDER BY insurance_status, day  -- just being explicit
     ) t CROSS JOIN
     (SELECT @sum := 0, @is := -1) params;

您需要在单个语句中完成所有分配。这是由于MySQL的规则所致-它不能保证SELECT中表达式的计算顺序，因此您不能在一个表达式中分配变量，而在另一个表达式中使用它们。