Forum Discussion

ebecerra's avatar
ebecerra
Microsoft Employee
3 years ago
Solved

Recursively split string into columns

I want to create some sort of "tree map visual" from some data stored in a DB, let me explain a little bit. I have the following sample data:

 

MainPage

MainPage/Page1

MainPage/Page1/SubPage

MainPage/Page2

MainPage/Page3/Subpage/OtherSubpage/AndSoOn/AndSoForth/Until/DozensOfTimes

MainPage/Page1/Subpage2

 

It is a list of "Paths" of dynamic length and I have no way of knowing the max levels that a path can have. All of them start with MainPage. I would like to visualize those paths in a way in PBI (ideally in hierarchical way) to kind of view a map of the paths to see where things are more clustered than others. I have split text in PBI before into columns, but this problem would require me to keep splitting an unknown number of times until the longest path has been completely processed. Ideally I would like to get something like this:

 

 

 

What would be the best way to parse those paths to create some useful visual for how those pages are organized? What would be a good visualization to aid with this? Or does this sounds like a problem that PBI can't solve and I should look into some other programatic way to get that visual?

 

Thanks in advance!

  • ebecerra You can do a split and then Unpivot other rows in Power Query or you can use SUBSTITUE(MAX('Table'[Column]), "/", "|"). That will turn it into a path and you can use the PATH functions like PATHLENGTH.

3 Replies

  • Greg_Deckler's avatar
    Greg_Deckler
    Community Champion

    ebecerra You can do a split and then Unpivot other rows in Power Query or you can use SUBSTITUE(MAX('Table'[Column]), "/", "|"). That will turn it into a path and you can use the PATH functions like PATHLENGTH.

  • CNENFRNL's avatar
    CNENFRNL
    Community Champion

    You can easily map the path this way,

    let
        Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45W8k3MzAtITE9VitVBcPRBhCEWIf3g0iTsqo0whYxBqgtAPP+SjNQiGMcxLyU43z8PQrvlF5Vk6IfmlWTm6LvkV6XmFfunhWTmphbjsLwAYlUsAA==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Path = _t]),
        #"Split Path" = Table.TransformColumns(Source, {"Path", each let l1=Text.Split(_,"/"), l2={null} & List.RemoveLastN(l1) in Table.FromRows(List.Zip({l1,l2,{1..List.Count(l1)}}),{"Dir","Parent Dir","Depth"})}),
        #"Expanded Path" = Table.Distinct(Table.ExpandTableColumn(#"Split Path", "Path", {"Dir","Parent Dir","Depth"}))
    in
        #"Expanded Path"