Handling tag data in a table

[elettra84]

Hi all,

I went through the posts but couldn't find an answer to the following question (apologies if I missed it).

I have a table that has a field containing a list of tags (I can format this string as I want)  but I would like to aggregate data by tag rather than bu the whole string.

Example, I have a table with 2 rows:

ID-1	<tag1>,<tag2>
ID-2	<tag2>

I would like data to be aggregated this way:

<tag1> 1 record

<tag2> 2 records

Instead of this way:

 <tag1>,<tag2> 1 record

<tag2> 1 records

Is there a way I can do that?

Many thanks

[v-eachen-msft · Accepted solution]

Hi @elettra84 ,

You could open query editor and unpivot the tag columns to get an extra column. Then pivot this column to get IDs without aggregation and remove blank tags.

Here is my test file for your reference.

[AllisonKennedy]

Is there a clear delimiter that separates the tags? If so, you could add a step in Power Query Editor that does a split column by delimiter and expand the advanced options to split into rows rather than columns.

[elettra84]

I was thinking about this solution, however in that table I also have metrics that shouldn't be duplicated as it will double/triple count them. What do you suggest?

[AllisonKennedy]

@elettra84  One option in that case is to split the table into 2 - what I sometimes do in this case is right click on the table query and choose 'reference' to create the additional tables. Keep the Key ID column in both tables, but only keep the Tags in one table and only the other info in the other table. Relate them in the data model and carry on as normal. 

Does that make any sense?