Sql 为ColumnB中的每个唯一值计算ColumnA中的不同值

我有一个装运编号表（可以重复）和一个连接的起始-目的地字段（可以重复）。我有一个select列，它为每个装运编号选择所有的始发地目的地。我想添加第三列，用于获取每个唯一装运编号的唯一原产地目的地的数量

我试着做了一个实验

Select [OD], [ShipmentNo], [UniqueODs] = (COUNT([OD])) 

也试过了

 Select [OD], [ShipmentNo], [UniqueODs] =(COUNT([ShipmentNo]) 

而不是[OD]。在我的附加列中，所有内容都返回1

 SELECT     [OD],
            [ShipmentNo],
            [UniqueODs] = (COUNT([OD]))


             FROM tblShipments

             GROUP BY MVMT_NO, [OD Pair]

             ORDER BY MVMT_NO, [OD Pair]

预期结果：

[Origin-Destination]                [ShipmentNo]             [UniqueODs]
Michigan-California                 A484310                  1
Michigan-California                 A484311                  1
Texas-California                    A484312                  2
Michigan-California                 A484312                  2

[OD]                    [ShipmentNo]              [UniqueODsForSN]
Michigan-California     A484310                   1
Michigan-California     A484311                   1
Texas-California        A484312                   2
Michigan-California     A484312                   2

目前的结果：

[Origin-Destination]                [ShipmentNo]             [UniqueODs]
Michigan-California                 A484310                  1
Michigan-California                 A484311                  1
Texas-California                    A484312                  1
Michigan-California                 A484312                  1

我希望两行上A484312的输出都是2，而不是1

预期结果：

[Origin-Destination]                [ShipmentNo]             [UniqueODs]
Michigan-California                 A484310                  1
Michigan-California                 A484311                  1
Texas-California                    A484312                  2
Michigan-California                 A484312                  2

[OD]                    [ShipmentNo]              [UniqueODsForSN]
Michigan-California     A484310                   1
Michigan-California     A484311                   1
Texas-California        A484312                   2
Michigan-California     A484312                   2

---

我想您需要一个窗口功能：

SELECT OD, ShipmentNo, SUM(COUNT(*)) OVER (PARTITION BY ShipmentNo) as UniqueODs
FROM tblShipments
GROUP BY MVMT_NO, [OD Pair]
ORDER BY MVMT_NO, [OD Pair];

嗯。如果当前计数为

1

，则不需要聚合：

SELECT OD, ShipmentNo, COUNT(*) OVER (PARTITION BY ShipmentNo) as UniqueODs
FROM tblShipments
ORDER BY MVMT_NO, [OD Pair];

编辑：

如果你真的想要：

SELECT OD, ShipmentNo, COUNT(*) OVER (PARTITION BY ShipmentNo) as UniqueODs
FROM tblShipments
GROUP BY MVMT_NO, [OD Pair]
ORDER BY MVMT_NO, [OD Pair];

---

我想您需要一个窗口功能：

SELECT OD, ShipmentNo, SUM(COUNT(*)) OVER (PARTITION BY ShipmentNo) as UniqueODs
FROM tblShipments
GROUP BY MVMT_NO, [OD Pair]
ORDER BY MVMT_NO, [OD Pair];

嗯。如果当前计数为

1

，则不需要聚合：

SELECT OD, ShipmentNo, COUNT(*) OVER (PARTITION BY ShipmentNo) as UniqueODs
FROM tblShipments
ORDER BY MVMT_NO, [OD Pair];

编辑：

如果你真的想要：

SELECT OD, ShipmentNo, COUNT(*) OVER (PARTITION BY ShipmentNo) as UniqueODs
FROM tblShipments
GROUP BY MVMT_NO, [OD Pair]
ORDER BY MVMT_NO, [OD Pair];

---

您应该提供一个要处理的示例数据。您应该提供一个要处理的示例数据。到此为止。它确实得到了ShipmentNo的ODs数量，但没有得到ShipmentNo的唯一*ODs数量。有什么想法吗？@NthngLeftToBurn。您问题中的示例数据不包含任何重复项。但请查看最后一次编辑。这已接近尾声。它确实得到了ShipmentNo的ODs数量，但没有得到ShipmentNo的唯一*ODs数量。有什么想法吗？@NthngLeftToBurn。您问题中的示例数据不包含任何重复项。但请查看最后一次编辑。