MySQL >、< 和按组缺失

Question

我在 MySQL 中有两个表，我将它们与以下属性进行比较：

tbl_fac : facility_id, chemical_id, criteria
             10      , 25         , 50
             10      , 26         , 60
             10      , 27         , 60
             11      , 25         , 30
             11      , 27         , 31 
              etc...

tbl_samp: sample_id, chemical_id, result
            5     ,    25         , 51
            5     ,    26         , 61
            6     ,    25         , 51
            6     ,    26         , 61
            6     ,    27         , 500

              etc.... 

这些表是由chemical_id（多对多----呃）连接的，有几千个facility_id，每个facility_id 有几百个chemic_id。还有几千个sample_id，每个sample_id 都有几百个chemic_id。总而言之，tbl_fac 中有大约 500,000 条记录，tbl_samp 中有 1,000,000 多条记录。

我正在尝试从该数据集中提取三组 sample_id：

第 1 组：任何 sample_id，其中 tbl_samp.result > tbl_fac.criteria（即结果超出标准）

第 2 组：tbl_samp.result

第 3 组：tbl_samp.result

问题来了：如何在一个查询中有效地获取所有三个组？

我试过了：

select * 
from tbl_fac 
left join tbl_samp 
    on tbl_fac.chemical_id = tbl_samp.chemical_id

但这只会产生整个数据集（而不是单个样本）缺失的值。我有一个 hackish 查询，它使用第三个表来连接 tbl_fac 和 tbl_samp，但它太丑了，我实际上很尴尬地发布它......

与往常一样，非常感谢您对此的想法！

干杯，

乔什

编辑：理想情况下，我希望返回 sample_id 和 Group - 每个样本 ID 只有一个 Group（我对数据的了解表明它们将始终属于上述三个类别之一)。

Answer 1

此答案假设tbl_fac 中的facility_id 和chemical_id 存在唯一约束，tbl_samp 中sample_id 和chemical_id 存在唯一约束。我所做的是一次一步地建立查询。这是否有效还有待观察。

第 1 组：任何 sample_id，其中 tbl_samp.result > tbl_fac.criteria（即结果超出标准）

SELECT tbl_samp.sample_id,
       'ResultsGreaterThanCriteria' AS samplegroup
FROM   tbl_fac
       INNER JOIN tbl_samp
         ON tbl_fac.chemical_id = tbl_samp.chemical_id
WHERE  tbl_samp.result > tbl_fac.criteria
GROUP  BY tbl_samp.sample_id

第 2 组：tbl_samp.result

SELECT tbl_samp.sample_id,
       'ResultLessThanCriteriaAndAllChems' AS samplegroup
FROM   tbl_fac
       INNER JOIN tbl_samp
         ON tbl_fac.chemical_id = tbl_samp.chemical_id
WHERE  tbl_samp.result < tbl_fac.criteria
       AND NOT EXISTS (SELECT *
                       FROM   tbl_fac tf
                              LEFT JOIN tbl_samp ts
                                ON tf.chemical_id = ts.chemical_id
                       WHERE  ts.chemical_id IS NULL
                              AND tbl_samp.sample_id = ts.sample_id)
GROUP  BY tbl_samp.sample_id

第 3 组：tbl_samp.result

SELECT tbl_samp.sample_id,
       'ResultsLessThanCriteriaWithMissingChems' AS samplegroup
FROM   tbl_fac
       INNER JOIN tbl_samp
         ON tbl_fac.chemical_id = tbl_samp.chemical_id
WHERE  tbl_samp.result < tbl_fac.criteria
       AND EXISTS (SELECT *
                   FROM   tbl_fac tf
                          LEFT JOIN tbl_samp ts
                            ON tf.chemical_id = ts.chemical_id
                   WHERE  ts.chemical_id IS NULL
                          AND tbl_samp.sample_id = ts.sample_id)
GROUP  BY tbl_samp.sample_id 

最后，将所有三个查询合并在一起，得到：

SELECT tbl_samp.sample_id,
       'ResultsGreaterThanCriteria' AS samplegroup
FROM   tbl_fac
       INNER JOIN tbl_samp
         ON tbl_fac.chemical_id = tbl_samp.chemical_id
WHERE  tbl_samp.result > tbl_fac.criteria
GROUP  BY tbl_samp.sample_id
UNION ALL
SELECT tbl_samp.sample_id,
       'ResultLessThanCriteriaAndAllChems' AS samplegroup
FROM   tbl_fac
       INNER JOIN tbl_samp
         ON tbl_fac.chemical_id = tbl_samp.chemical_id
WHERE  tbl_samp.result < tbl_fac.criteria
       AND NOT EXISTS (SELECT *
                       FROM   tbl_fac tf
                              LEFT JOIN tbl_samp ts
                                ON tf.chemical_id = ts.chemical_id
                       WHERE  ts.chemical_id IS NULL
                              AND tbl_samp.sample_id = ts.sample_id)
GROUP  BY tbl_samp.sample_id
UNION ALL
SELECT tbl_samp.sample_id,
       'ResultsLessThanCriteriaWithMissingChems' AS samplegroup
FROM   tbl_fac
       INNER JOIN tbl_samp
         ON tbl_fac.chemical_id = tbl_samp.chemical_id
WHERE  tbl_samp.result < tbl_fac.criteria
       AND EXISTS (SELECT *
                   FROM   tbl_fac tf
                          LEFT JOIN tbl_samp ts
                            ON tf.chemical_id = ts.chemical_id
                   WHERE  ts.chemical_id IS NULL
                          AND tbl_samp.sample_id = ts.sample_id)
GROUP  BY tbl_samp.sample_id 

Answer 2

SELECT
    sample_id,
    IF(result = criteria, -1,  /* unspecified behavior */
     IF(result > criteria, 1,
      IF(nb_chemicals = total_nb_chemicals, 2, 3))) AS grp

FROM (
    SELECT s.result, s.sample_id, f.criteria, f.chemical_id,
        COUNT(DISTINCT f.chemical_id) AS nb_chemicals
    FROM tbl_fac f JOIN tbl_samp s
        ON f.chemical_id = s.chemical_id
    GROUP BY s.sample_id
) t 

CROSS JOIN (
    SELECT COUNT(DISTINCT chemical_id) AS total_nb_chemicals
    FROM tbl_fac
) u

新解决方案：

SELECT
    s.sample_id,
    IF(s.result = f.criteria, -1,  /* unspecified behavior */
     IF(s.result > f.criteria, 1,
      IF(sample_nb_chemicals = total_nb_chemicals, 2, 3))) AS grp

FROM
    tbl_fac f JOIN tbl_samp s
    ON f.chemical_id = s.chemical_id

    JOIN (
        SELECT s.sample_id, 
               COUNT(DISTINCT f.chemical_id) AS sample_nb_chemicals
        FROM tbl_fac f JOIN tbl_samp s
             ON f.chemical_id = s.chemical_id
        GROUP BY s.sample_id
    ) u
       ON s.sample_id = u.sample_id

    CROSS JOIN (
        SELECT COUNT(DISTINCT chemical_id) AS total_nb_chemicals
        FROM tbl_fac
    ) v

GROUP BY sample_id, grp