Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
6 years ago
Solved

Remove Duplicates Per Group (keep first) measure help

Hi All

 

Is it possible to create a measure in PBI that remove duplicate values per group, and keeps the first occurrence of the value? Perhaps not a measure but maybe a new table, I'm unsure of how to work around this. For context, the database I'm working with has a funky treatment of some data that has a function back-end but doesn't make sense when trying to visualise user behaviour.

 

Currently, the data looks like:

USERENTRY IDENTRYSTATUSTIME SUBMITTED (HH:mm:ss)
userA1unicornfail10:23:10
userA2forestpass10:30:49
userA1unicornfail10:30:49
userB1unicornfail13:40:22
userB1fairypass13:43:59

 

I want to clean it so it looks like:

USERENTRY IDENTRYSTATUSTIME SUBMITTED (HH:mm:ss)
userA1unicornfail10:23:10
userA2forestpass10:30:49
userB1unicornfail13:40:22
userB1fairypass13:43:59
...............

 

Note the row I want to remove has

  1. duplicated ENTRY from the first instance for userA and
  2. duplicated TIME SUBMITTED from the first instance for userA

Also, the ENTRY ID cannot be used.

Any pointers would be greatly appreciated 🙂 

DAX
groupby
Help required
Measures
Power Query
  • Anonymous's avatar
    Anonymous
    6 years ago

    Here's the M code that does what you want:

    let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WKi1OLXJU0lEyBOLSvMzk/KI8ICstMTMHJGhgZWRsZWigFKuDUGkEks8vSi0uATIKEouLIQqNDaxMLFEU4jISVaUTTpXGViZA+40wVALliyqR7AaqM7YyBZoYCwA=", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [User = _t, EntryID = _t, Entry = _t, Status = _t, Time = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"User", type text}, {"EntryID", Int64.Type}, {"Entry", type text}, {"Status", type text}, {"Time", type time}}),
    #"Added Custom" = Table.AddColumn(#"Changed Type", "FirstTimeEntry", 
        each List.Min( 
            Table.SelectRows(
                #"Changed Type",
                (r) => r[User] = [User] and r[Entry] = [Entry]
            )[Time] 
        ) = [Time]),
    #"Filtered Rows" = Table.SelectRows(#"Added Custom", each ([FirstTimeEntry] = true)),
    #"Removed Columns" = Table.RemoveColumns(#"Filtered Rows",{"FirstTimeEntry"})
in
    #"Removed Columns"

     

    Best

    D

5 Replies

    • Anonymous's avatar
      Anonymous
      Not applicable
      6 years ago

      Thanks Pragati11 , the buffer got me halfway there! Now it's just removing the wrong duplicate, hopefully the other reply will resolve this 🙂 

  • Anonymous's avatar
    Anonymous
    Not applicable
    6 years ago
    Yeah... That's a perfect job for Power Query. You can do it in DAX as well as a calculated table but it really should be performed in PQ as this is the data-munging tool. I can create some sample M code for you so that you can see how such cleaning is done...

    Best
    D
  • Anonymous's avatar
    Anonymous
    Not applicable
    6 years ago

    Here's the M code that does what you want:

    let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WKi1OLXJU0lEyBOLSvMzk/KI8ICstMTMHJGhgZWRsZWigFKuDUGkEks8vSi0uATIKEouLIQqNDaxMLFEU4jISVaUTTpXGViZA+40wVALliyqR7AaqM7YyBZoYCwA=", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [User = _t, EntryID = _t, Entry = _t, Status = _t, Time = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"User", type text}, {"EntryID", Int64.Type}, {"Entry", type text}, {"Status", type text}, {"Time", type time}}),
    #"Added Custom" = Table.AddColumn(#"Changed Type", "FirstTimeEntry", 
        each List.Min( 
            Table.SelectRows(
                #"Changed Type",
                (r) => r[User] = [User] and r[Entry] = [Entry]
            )[Time] 
        ) = [Time]),
    #"Filtered Rows" = Table.SelectRows(#"Added Custom", each ([FirstTimeEntry] = true)),
    #"Removed Columns" = Table.RemoveColumns(#"Filtered Rows",{"FirstTimeEntry"})
in
    #"Removed Columns"

     

    Best

    D

    • Anonymous's avatar
      Anonymous
      Not applicable
      6 years ago

      Thanks Anonymous, had to have a little play around and add in r[entry id] = [entry id] but it's working well now 👍 cheers again