SQL 如何仅当组遵循某些规则时才找到许多组总和的平均值

Question

这是我当前的代码：

SELECT
AVG(famTotal)
FROM `OmniHealth.new2015Data`, (
SELECT
SUM( TOTEXP15 ) as famTotal
FROM `OmniHealth.new2015Data` 
GROUP BY DUID
)
WHERE BMINDX53 BETWEEN 0 AND 25 AND
ADSMOK42 = -1 AND
FCSZ1231 = 7

我想做的是找到每个家庭的平均成本，其中家庭的所有成员的BMI都在0到25之间，不吸烟，并且家庭是任意大小的。

支出数据是按每个人计算的，所以我试图根据所有人的"住宅单元ID"(DUID(进行求和，然后平均每个家庭的总数，只要该家庭只有我在上一段中陈述的房产。

感谢您的回复！我是SQL的新手。

Answer 1

请考虑联接两个按计数匹配的聚合查询派生表，以将所有家庭成员与具有特定条件的所有家庭成员对齐。

SELECT AVG(t1.famTotal) as famTotal
FROM
  (SELECT DUID, Count(*) As GrpCount, SUM(TOTEXP15) as famTotal
   FROM `OmniHealth.new2015Data` 
   GROUP BY DUID) As t1
INNER JOIN
  (SELECT DUID, Count(*) As GrpCount
   FROM `OmniHealth.new2015Data` 
   WHERE BMINDX53 BETWEEN 0 AND 25 
     AND ADSMOK42 = -1 
     AND FCSZ1231 = 7
   GROUP BY DUID) As t2
ON t1.DUID = t2.DUID AND t1.GrpCount = t2.GrpCount

Answer 2

下面是 BigQuery Standard SQL

#standardSQL
SELECT DUID, AVG(TOTEXP15) AS famAverage
FROM `OmniHealth.new2015Data` 
GROUP BY DUID
HAVING MIN(BMINDX53) >=0 AND MAX(BMINDX53) <=25
AND MIN(ADSMOK42) = -1 AND MAX(ADSMOK42) = -1
AND MIN(FCSZ1231) = 7 AND MAX(FCSZ1231) = 7