Question

我正在尝试编写一个查询，该查询将在多个列上执行COUNT，按其各自的列对它们进行分组。

以下是我现在的查询：

SELECT COUNT(*) as total,
COUNT(q3_pay) as q3_pay_total,
COUNT(q3_holidays) as q3_holidays_total,
COUNT(q3_vacation) as q3_vacation_total,
COUNT(q3_sick) as q3_sick_total,
COUNT(q3_insurance) as q3_insurance_total,
COUNT(q3_retirement) as q3_retirement_total,
COUNT(q3_fsa) as q3_fsa_total
FROM exit_responses

我需要找到两件事：

每列COUNT（列）的回复总数。
每列COUNT（列[x]）＆lt; - 示例

有没有办法让我在不为每列创建新的SQL查询的情况下执行此操作？

我目前还有其他查询，例如：

SELECT COUNT(*) as total, q2 FROM exit_responses
GROUP BY q2

谢谢大家：）。

Answer 1

您需要提供表格的定义和一些示例数据行。

话虽如此，以下内容应该有效。

SELECT COUNT(*) as total,
sum(nvl(q3_pay,0)) as q3_pay_total,
sum(nvl(q3_holidays,0)) as q3_holidays_total,
...
FROM exit_responses

不确定这是否是您要找的。

Answer 2

这取决于您的数据，如果这些是您想要的，但您可以做两件事：

GROUP BY所有不同的列，然后使用COUNT(DISTINCT col)获取该列中不同值的数量
GROUP BY任何合适的列，然后SUM(IF(col = 'xy',1,0))计算值达到某个条件的频率

Answer 3

Count（）将为每列返回相同的数字。你需要做的是每个的SUM（）。确保将每个值的值更改为1，具体取决于内容。（您可以在总和中使用案例来执行此操作）

Answer 4

正如其他人所说，请发布DDL或表格以及一些样本行和预期的输出。

嗯，我不认为这是你想做的事情？

select 'Q3'                   as quarter,
       count(*)               as total,
       count(q3_pay)          as q3_pay_total,
       count(q3_holidays)     as q3_holidays_total,
       count(q3_vacation)     as q3_vacation_total,
       count(q3_sick)         as q3_sick_total,
       count(q3_insurance)    as q3_insurance_total,
       count(q3_retirement)   as q3_retirement_total,
       count(q3_fsa)          as q3_fsa_total
  from exit_responses
union all
select 'Q2'                   as quarter,
       count(*)               as total,
       count(q2_pay)          as q2_pay_total,
       count(q2_holidays)     as q2_holidays_total,
       count(q2_vacation)     as q2_vacation_total,
       count(q2_sick)         as q2_sick_total,
       count(q2_insurance)    as q2_insurance_total,
       count(q2_retirement)   as q2_retirement_total,
       count(q2_fsa)          as q2_fsa_total
  from exit_responses

我真的必须为此编写8个不同的SQL查询吗？

4 个答案: