Forum Discussion

jose_cruzrodz's avatar
jose_cruzrodz
Frequent Visitor
3 years ago
Solved

Filter records by second column, based on primary column answer

Greetings,

 

Here is my sample data:

 

Expected result:

 

 

 

 

 

 

What I need to accomplish is look at Column E first. If that column has a value of "Yes", then I look at Column B of that same record/row and grab its value. Once I get the target value of Column B, I will filter the table to show only all rows/records that match that collected value of Column B. From my example table, rows 6 and 10 had "Yes" as an answer in Column E, so, I grabbed their related values of Column B ("456" and "789", respectivaly). With those values at hand, I set my filter in Column B to just show all records that match those values. Column E is used to identify the target value of Column B, the filter itself is in Column B. All this must be done in POWER QUERY.

 

I have played with Custom Columns or Advanced Filters, but I cannot get it to work. Can someone help me with achieving what I need?

 

 

Thanks,

José

Need Help
  • Syk's avatar
    Syk
    3 years ago

    First thing that comes to my head is this..
    Duplicate the Table and remove all but columns B and E and filter it to just "Yes"

    Then, merge this table back to your main table using column B

    Expand column E(from table 2) and rename it to make it a little more clear.
    Use this new column and only filter for "Yes"

     

     

3 Replies

  • Syk's avatar
    Syk
    Resident Rockstar
    3 years ago

    First thing that comes to my head is this..
    Duplicate the Table and remove all but columns B and E and filter it to just "Yes"

    Then, merge this table back to your main table using column B

    Expand column E(from table 2) and rename it to make it a little more clear.
    Use this new column and only filter for "Yes"

     

     

    • jose_cruzrodz's avatar
      jose_cruzrodz
      Frequent Visitor
      3 years ago

      It is definitely not the best solution in terms of performance and data duplication (on the Power Query side), but it certainly worked and achieved what I needed! Right now, the raw data is about 5K records so I can live with this for now. At least I can generate the preliminary report while I find the permanent solution. You are definitely a life saver! I absolutely appretiate your help. Thanks a lot!!!

  • Vijay_A_Verma's avatar
    Vijay_A_Verma
    Most Valuable Professional
    3 years ago

    Insert this code where #"Changed Type" should be replaced with your previous step

    = Table.Combine(List.Select(Table.Group(#"Changed Type", {"ColumnB"}, {{"All", each _}})[All], (x)=>List.Contains(x[ColumnE], "Yes")))

    Complete test code in action for testing

    let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WCk8sSS1S0lEyNDIGksGlxYl5QDo7JSs7BUgrxeoglJiYmoGUlKSWpQLptNTUwtRCmJrgxLwUAkoIGBOZWoyizNzCEkg65iWCFKWkpaDZhCqdkoZkQmJeZUlGZl46LkNiAQ==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [ColumnA = _t, ColumnB = _t, ColumnC = _t, ColumnD = _t, ColumnE = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"ColumnA", type text}, {"ColumnB", Int64.Type}, {"ColumnC", type text}, {"ColumnD", type text}, {"ColumnE", type text}}),
    #"Grouped Rows" = Table.Combine(List.Select(Table.Group(#"Changed Type", {"ColumnB"}, {{"All", each _}})[All], (x)=>List.Contains(x[ColumnE], "Yes")))
in
    #"Grouped Rows"