SQL Server：GROUP BY多個相似/匹配行的段

Question

我需要能夠做一個group by但分段。所以在下面的例子中，有4個「組」代表一段旅程。

SELECT id, mode, distance 
FROM TravelLog; 

輸出：

----------------------------------------------------------------------------- 
    id mode  distance 
----------------------------------------------------------------------------- 
    1 on_foot 0.1 
    2 on_foot 0.2 
    3 on_foot 0.4 
    4 on_foot 0.6 
    5 on_foot 0.7 
    6 on_foot 0.8 
    7 in_vehicle 0.9 
    8 in_vehicle 2.0 
    9 in_vehicle 2.5 
    10 in_vehicle 3.0 
    11 in_vehicle 3.5 
    12 in_vehicle 4.0 
    13 in_vehicle 4.5 
    14 on_foot 4.6 
    15 on_foot 4.7 
    16 on_foot 4.8 
    17 on_foot 4.9 
    18 in_vehicle 5.5 
    19 in_vehicle 6.0 
    20 in_vehicle 6.5 
----------------------------------------------------------------------------- 

期望的結果：

----------------------------------------------------------------------------- 
    start_id end_id start_distance end_distance mode 
----------------------------------------------------------------------------- 
    1   6  0.1    0.8    on_foot 
    7   13  0.9    4.5    in_vehicle 
    14   17  4.6    4.9    on_foot 
    18   20  5.5    6.5    in_vehicle 
----------------------------------------------------------------------------- 

Answer 1

您需要確定相鄰的 「模式」 的羣體。一種方法是使用行號的區別：

select min(id) as start_id, max(id) as end_id, 
     min(distance) as start_distance, max(distance) as end_distance, 
     mode 
from (select t.*, 
      row_number() over (order by id) as seqnum, 
      row_number() over (partition by mode order by id) as seqnum_m 
     from t 
    ) t 
group by mode, (seqnum - seqnum_m); 

理解爲什麼在行數之差的工作原理是有點棘手。如果您運行子查詢並盯着結果，那麼您可能會弄清楚爲什麼區別標識組。