添加具有相同列值的行的平均时间差的列答案

【问题标题】：Add Column with average time difference of rows with same column values添加具有相同列值的行的平均时间差的列
【发布时间】：2023-01-07 00:49:16
【问题描述】：

我得到了一个这样的 PowerQuery 表：

ID	A	B	C	COUNT	Timestamp
1	a1		c1	0	2017-05-10 09:55:28
	a3	b	c2		2017-05-10 10:12:54
2	a2		c3	2	2017-05-10 10:19:47
	a2	b	c4		2017-05-10 10:20:24
	a2	b	c5		2017-05-10 10:21:50
3	a3		c6	1	2017-05-10 10:31:02
	a3	c	c7		2017-05-10 10:31:02

在这里，列 COUNT 检查列值 ID 是否非空，如果是，它计算具有相同列值 A 的行这样列值 B 等于“b”。

现在，我想添加一个列 AVG_TIME_DIFF，它再次检查列值 ID 是否为非空，如果是：

检查COUNT是否等于0，在这种情况下返回“0”
else 获取具有相同列值的所有行的时间戳 A这样列值 B 等于“b”以及行本身，按时间顺序对它们进行排序（即COUNT+1 多行）并返回以秒为单位的平均时差
否则结果为空

在上面的示例表中，结果应该是

ID | A  | B | C  | COUNT |      Timestamp      | AVG_TIME_DIFF
 1 | a1 |   | c1 | 0     | 2017-05-10 09:55:28 | 0
   | a3 | b | c2 |       | 2017-05-10 10:12:54 |
 2 | a2 |   | c3 | 2     | 2017-05-10 10:19:47 | 62
   | a2 | b | c4 |       | 2017-05-10 10:20:24 |
   | a2 | b | c5 |       | 2017-05-10 10:21:50 |
 3 | a3 |   | c6 | 1     | 2017-05-10 10:31:02 | 1088
   | a3 | c | c7 |       | 2017-05-10 10:31:02 |

可以根据需要向上或向下舍入秒的小数部分。

【问题讨论】：

标签： powerquery

【解决方案1】：

尝试

let Source = Excel.CurrentWorkbook(){[Name="Table1"]}[Content],
#"Added Index" = Table.AddIndexColumn(Source, "Index2", 0, 1, Int64.Type),
#"Changed Type" = Table.TransformColumnTypes(#"Added Index",{{"ID", Int64.Type}, {"A", type text}, {"B", type text}, {"C", type text}, {"COUNT", Int64.Type}, {"Timestamp", type datetime}}),
#"Filtered Rows" = Table.SelectRows(#"Changed Type", each ([ID] <> null)),
#"Removed Other Columns" = Table.SelectColumns(#"Filtered Rows",{"ID", "A"}),
#"Added Custom" = Table.AddColumn(#"Removed Other Columns", "Custom",  (x)=>Table.SelectRows(#"Changed Type", each [A]=x[A])),
#"Added Custom2" = Table.AddColumn(#"Added Custom","AVG_TIME_DIFF2", each
    let a= Table.Sort([Custom],{{"Timestamp", Order.Ascending}}),
    b=  Table.AddIndexColumn(a, "Index", 0, 1, Int64.Type),
    c = Table.AddColumn(b, "Custom", each b{[Index]-1}[Timestamp]),
    d = Table.AddColumn(c, "Times", each [Timestamp]-[Custom],type duration),
    e = Table.Skip(d,1),
    f = List.Average(e[Times])
    in f, type duration),
#"Removed Columns" = Table.RemoveColumns(#"Added Custom2",{"ID", "A"}),
#"Expanded Custom" = Table.ExpandTableColumn(#"Removed Columns", "Custom", {"ID", "A", "B", "C", "COUNT", "Timestamp", "Index2"}, {"ID", "A", "B", "C", "COUNT", "Timestamp", "Index2"}),
#"Sorted Rows" = Table.Sort(#"Expanded Custom",{{"Index2", Order.Ascending}}),
#"Added Custom1" = Table.AddColumn(#"Sorted Rows", "AVG_TIME_DIFF", each if [ID]=null then null else if [AVG_TIME_DIFF2]=null then 0 else Duration.TotalSeconds([AVG_TIME_DIFF2]),type number),
#"Removed Columns1" = Table.RemoveColumns(#"Added Custom1",{"AVG_TIME_DIFF2"}),
#"Changed Type1" = Table.TransformColumnTypes(#"Removed Columns1",{{"Timestamp", type datetime}})

在 #“改变类型1”

【讨论】：