保留重复项（Power Query 之 M 语言）

数据源：

“姓名”“基数”“个人比例”“个人缴纳”“公司比例”“公司缴纳”“总计”，共7列7行数据，其中姓名列，第1、2行与第6、7行内容重复

目标：

留下第1、2、6、7姓名列中内容重复的行

操作过程：

选取指定列》【主页】》【保留行】》【保留重复项】

 

M公式：

= let columnNames = {"指定列名"}, addCount = Table.Group(步骤名, columnNames, {{"Count", Table.RowCount, type number}}), selectDuplicates = Table.SelectRows(addCount, each [Count] > 1), removeCount = Table.RemoveColumns(selectDuplicates, "Count") in Table.Join(步骤名, columnNames, removeCount, columnNames, JoinKind.Inner)

说明：

= let columnNames = {"指定列名"}, addCount = Table.Group(步骤名, columnNames, {{"Count", Table.RowCount, type number}}), selectDuplicates = Table.SelectRows(addCount, each [Count] > 1), removeCount = Table.RemoveColumns(selectDuplicates, "Count") in Table.Join(步骤名, columnNames, removeCount, columnNames, JoinKind.Inner)

• 给指定的列名一个说法，叫作“columnNames”

 

= let columnNames = {"指定列名"}, addCount = Table.Group(步骤名, columnNames, {{"Count", Table.RowCount, type number}}), selectDuplicates = Table.SelectRows(addCount, each [Count] > 1), removeCount = Table.RemoveColumns(selectDuplicates, "Count") in Table.Join(步骤名, columnNames, removeCount, columnNames, JoinKind.Inner)

• 【分组依据】计算指定列里各个值出现次数，并将这结果命名为addCount

 

= let columnNames = {"指定列名"}, addCount = Table.Group(步骤名, columnNames, {{"Count", Table.RowCount, type number}}), selectDuplicates = Table.SelectRows(addCount, each [Count] > 1), removeCount = Table.RemoveColumns(selectDuplicates, "Count") in Table.Join(步骤名, columnNames, removeCount, columnNames, JoinKind.Inner)

• 【保留行】保留“Count”列中值大于1的行

 

= let columnNames = {"指定列名"}, addCount = Table.Group(步骤名, columnNames, {{"Count", Table.RowCount, type number}}), selectDuplicates = Table.SelectRows(addCount, each [Count] > 1), removeCount = Table.RemoveColumns(selectDuplicates, "Count") in Table.Join(步骤名, columnNames, removeCount, columnNames, JoinKind.Inner)

• 【删除列】将“Count”列删除

 

= let columnNames = {"指定列名"}, addCount = Table.Group(步骤名, columnNames, {{"Count", Table.RowCount, type number}}), selectDuplicates = Table.SelectRows(addCount, each [Count] > 1), removeCount = Table.RemoveColumns(selectDuplicates, "Count") in Table.Join(步骤名, columnNames, removeCount, columnNames, JoinKind.Inner)

• 将原表和只保留重复值的表进行【合并查询】，连接种类使用“内部”

 

将所有步骤拆分如图所示。

 

其中核心部分：

addCount = Table.Group( 更改的类型,"姓名", {{"Count", Table.RowCount, type number}}),

selectDuplicates = Table.SelectRows(addCount, each [Count] > 1),

removeCount = Table.RemoveColumns(selectDuplicates, "Count"),

最后一步 = Table.Join(更改的类型, "姓名", removeCount, "姓名", JoinKind.Inner)

最终效果：

数据只剩下姓名列中重复的四行数据

 

多说一句：

好吧，我承认，我被这个公式惊到了！这其实已经不是一个简单的公式，而是一段M代码，这不是有let开头，in结尾么……