合并重复信息

Question

问题在于，WO：W142930 Leg 0重复多次，这造成了太多混乱。 我可以使用excel中的数据透视表进行清理，但是我宁愿学习如何在SQL中进行操作。 我不知道是否有可能，但我想在每条腿的顶部看到总的“ ACT_RUN_HRS”。

我编写的代码：

select
w.base_id as WO_ID
, w.sub_id as LEG
, w.desired_qty as WO_QTY
, OPR.ACT_RUN_HRS
, OPR.ACT_SETUP_HRS
, cast ((OPR.ACT_RUN_HRS + OPR.ACT_SETUP_HRS) AS decimal(15,4)) AS TOTAL_RUN_HOURS
, cast(w.close_date as date) AS CLOSE_DATE
, cast(cast( wb.bits as varbinary(100)) as varchar(100)) as SPEC
from work_order w
left join workorder_binary wb on
w.type = wb.workorder_type and
w.base_id = wb.workorder_base_id and
w.lot_id = wb.workorder_lot_id and
w.split_id = wb.workorder_split_id and
w.sub_id = wb.workorder_sub_id
left join operation opr
ON w.type = opr.workorder_type and
w.base_id = opr.workorder_base_id and
w.lot_id = opr.workorder_lot_id and
w.split_id = opr.workorder_split_id and
w.sub_id = opr.workorder_sub_id
left join PART p0 on p0.id = w.PART_ID and '0' = w.SUB_ID
left join CUSTOMER c on c.id = p0.user_1
WHERE year(w.CLOSE_DATE) >= 2017
and year(w.CLOSE_DATE) <= 2018
and w.STATUS='C'
and w.TYPE = 'W'
and w.BASE_ID LIKE 'W%'
and w.BASE_ID = 'w142930'
Order by w.base_id, w.sub_id

谢谢

Answer 1

更新

我更新了答案，将每个字段（w.base_id，w.sub_id，w.desired_qty和wb.bits）替换为强制转换为varchar（100），并且我修改了GROUP BY子句以包含相同的强制类型转换。让我知道这是否适合您。

更新2

我将cast(wb.bits as varchar(100))替换为cast(cast(wb.bits as varbinary(100)) as varchar(100))

更新3

如@DeathMagnetic所指出的，ORDER BY列必须出现在SELECT或GROUP BY子句中

答案

您正在寻找SUM()和GROUP BY。类似于以下内容的东西应该可以工作（但是，由于我没有使用您的架构设置数据库，因此我没有对此进行测试）：

SELECT
    cast(w.base_id as varchar(100)) as WO_ID,
    cast(w.sub_id as varchar(100)) as LEG,
    cast(w.desired_qty as varchar(100)) as WO_QTY,
    SUM(OPR.ACT_RUN_HRS),
    SUM(OPR.ACT_SETUP_HRS),
    SUM(cast ((OPR.ACT_RUN_HRS + OPR.ACT_SETUP_HRS) AS decimal(15,4))) AS 
TOTAL_RUN_HOURS
    cast(w.close_date as date) AS CLOSE_DATE,
    cast(cast(wb.bits as varbinary(100)) as varchar(100)) as SPEC
FROM
    work_order w
LEFT JOIN
    workorder_binary wb ON
        w.type = wb.workorder_type and
        w.base_id = wb.workorder_base_id and
        w.lot_id = wb.workorder_lot_id and
        w.split_id = wb.workorder_split_id and
        w.sub_id = wb.workorder_sub_id
LEFT JOIN
    operation opr ON
        w.type = opr.workorder_type and
        w.base_id = opr.workorder_base_id and
        w.lot_id = opr.workorder_lot_id and
        w.split_id = opr.workorder_split_id and
        w.sub_id = opr.workorder_sub_id
LEFT JOIN
    PART p0 ON
        p0.id = w.PART_ID and
        '0' = w.SUB_ID
LEFT JOIN
    CUSTOMER c ON
        c.id = p0.user_1
WHERE
    year(w.CLOSE_DATE) >= 2017
    and year(w.CLOSE_DATE) <= 2018
    and w.STATUS = 'C'
    and w.TYPE = 'W'
    and w.BASE_ID LIKE 'W%'
    and w.BASE_ID = 'w142930'
GROUP BY
    cast(w.base_id as varchar(100)),
    cast(w.sub_id as varchar(100)),
    cast(w.desired_qty as varchar(100)),
    cast(w.close_date as date),
    cast(cast(wb.bits as varbinary(100)) as varchar(100))
ORDER BY
    cast(w.base_id as varchar(100)), cast(w.sub_id as varchar(100))

通常：

• 使用汇总函数（COUNT，SUM，AVERAGE等）时，返回的每个单列必须 是汇总或属于{ {1}}子句。假设您仅按GROUP BY和w.base_id分组，并且对于相同的w.sub_id和w.desired_qty，您对w.base_id有两个不同的值-您会选择哪个想回来吗？因此，您可能需要按w.sub_id分组或使用一些汇总函数来告诉它如何选择（w.desired_qty，MAX等）
• MIN子句在WHERE子句之前执行，因此它只会对您的GROUP BY返回的行求和/求平均/聚合。如果您想根据这些汇总信息进一步过滤（例如，返回所有带有WHERE的人），则应使用SUM(OPR.ACT_RUN_HOURS) > 10子句（其功能与{ {1}}子句，但在HAVING之后而不是之前运行）

Answer 2

您要与GROUP BY子句一起使用所谓的聚合函数。它的发音听起来很有用。我确定您可以在Google上搜索更多信息以了解它们，但是对于您的特定问题，您可以SUM所需的值，并按要求和的值分组。像这样：

select
w.base_id as WO_ID
, w.sub_id as LEG
, w.desired_qty as WO_QTY
, SUM(OPR.ACT_RUN_HRS) AS SUM_ACT_RUN_HRS
, SUM(OPR.ACT_SETUP_HRS) AS SUM_ACT_SETUP_HRS
, SUM(cast ((OPR.ACT_RUN_HRS + OPR.ACT_SETUP_HRS) AS decimal(15,4))) AS 
TOTAL_RUN_HOURS
, cast(w.close_date as date) AS CLOSE_DATE
, cast(cast( wb.bits as varbinary(100)) as varchar(100)) as SPEC
from work_order w
left join workorder_binary wb on
w.type = wb.workorder_type and
w.base_id = wb.workorder_base_id and
w.lot_id = wb.workorder_lot_id and
w.split_id = wb.workorder_split_id and
w.sub_id = wb.workorder_sub_id
left join operation opr
ON w.type = opr.workorder_type and
w.base_id = opr.workorder_base_id and
w.lot_id = opr.workorder_lot_id and
w.split_id = opr.workorder_split_id and
w.sub_id = opr.workorder_sub_id
left join PART p0 on p0.id = w.PART_ID and '0' = w.SUB_ID
left join CUSTOMER c on c.id = p0.user_1
WHERE year(w.CLOSE_DATE) >= 2017
and year(w.CLOSE_DATE) <= 2018
and w.STATUS='C'
and w.TYPE = 'W'
and w.BASE_ID LIKE 'W%'
and w.BASE_ID = 'w142930'
GROUP BY w.base_id, w.sub_id, w.desired_qty, w.close_Date, wb.bits
Order by w.base_id, w.sub_id

应该做你想做的。您可能需要在分组依据列中包含来自选择项的转换，以获得确切的结果，但我认为并非如此。