我正在尝试对当前行（在 SQL Server 中）的列求和。我该怎么做呢？
select t1.CounterTime,
t1.StartTime,
t1.EndTime,
isNull(t1.value, 0) as value1,
-- How do I make Total1 the sum of t1.value over all previous rows?
sum( isNull(t1.value, 0) ) over (partition by t1.CounterTime order by t1.CounterTime) as Total1
from SomeTable t1
order by t1.CounterTime
但我得到了partition by错误的...
╔═══╦═════════════════════════╦═════════════════════════╦═════════════════════════╦════════╦════════╗
║ ║ CounterTime ║ StartTime ║ EndTime ║ value1 ║ Total1 ║
╠═══╬═════════════════════════╬═════════════════════════╬═════════════════════════╬════════╬════════╣
║ 1 ║ 2015-03-17 12:00:00.000 ║ 2015-03-17 00:00:00.000 ║ 2015-03-18 00:00:00.000 ║ 0 ║ 0 ║
║ 2 ║ 2015-03-18 12:00:00.000 ║ 2015-03-18 00:00:00.000 ║ 2015-03-19 00:00:00.000 ║ 0 ║ 0 ║
║ 3 ║ 2015-03-19 12:00:00.000 ║ 2015-03-19 00:00:00.000 ║ 2015-03-20 00:00:00.000 ║ 422 ║ 422 ║
║ 4 ║ 2015-03-20 12:00:00.000 ║ 2015-03-20 00:00:00.000 ║ 2015-03-21 00:00:00.000 ║ 989 ║ 989 ║
║ 5 ║ 2015-03-21 12:00:00.000 ║ 2015-03-21 00:00:00.000 ║ 2015-03-22 00:00:00.000 ║ 1162 ║ 1162 ║
║ 6 ║ 2015-03-22 12:00:00.000 ║ 2015-03-22 00:00:00.000 ║ 2015-03-23 00:00:00.000 ║ 524 ║ 524 ║
╚═══╩═════════════════════════╩═════════════════════════╩═════════════════════════╩════════╩════════╝
应该：
╔════════╗
║ Total1 ║
╠════════╣
║ 0 ║
║ 0 ║
║ 422 ║
║ 1411 ║
║ 2573 ║
║ 3097 ║
╚════════╝
Try this
删除“分区依据”仅使用“排序依据”
select t1.CounterTime,
t1.StartTime,
t1.EndTime,
isNull(t1.value, 0) as value1,
-- How do I make Total1 the sum of t1.value over all previous rows?
sum( isNull(t1.value, 0) ) over (order by t1.CounterTime) as Total1
from SomeTable t1
order by t1.CounterTime
本文内容由网友自发贡献，版权归原作者所有，本站不承担相应法律责任。如您发现有涉嫌抄袭侵权的内容，请联系:hwhale#tublm.com(使用前将#替换为@)
