Forum Discussion

anvikuttu's avatar
anvikuttu
Icon for Advocate I rankAdvocate I
5 years ago
Solved

Need help to get the conditional average in power query

Hi,

I am a newbie to PowerQuery. I need a M Query solution to the below problem

Suppose I have a Country column, Value column. Now I wanted to average the Value column not equal to 0, for each country in the country column.  Kindly see the below example.

 

Let me know if you have any questions.

 

Thank you...

Vj

Need Help
  • CNENFRNL's avatar
    CNENFRNL
    5 years ago

    Hi, Fowmy's solution surely works. When your dataset consists of 2k+ rows, you might try this,

    let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WCg1W0lEyNFCK1YGxTRFsIyQ2VIk3SBiJDVPuDVfiGIrKNERmQ1Q7OyLEwWwgMxYA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Country = _t, Value = _t]),

    #"Added Custom" = Table.AddColumn(Source, "Custom", each Table.Group(Source, {"Country"}, {{"grouped", each _}}){[Country = [Country]]}[grouped][Value]),
    Avg = Table.TransformColumns(#"Added Custom", {{"Custom", each List.Average(List.RemoveItems(List.Transform(_, Number.From), {0}))}})
in
    Avg
  • Jimmy801's avatar
    Jimmy801
    5 years ago

    Hello

     

    the simplest thing is to include all in ONE Group-Step... the average except 0 and All rows (to keep the value-column for later)

    Check it out

    let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WCg1W0lEyNFCK1YGxTRFsIyQ2VIk3SBiJDVPuDVfiGIrKNERmQ1Q7OyLEwWwgMxYA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Country = _t, Value = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"Value", type number}}),
    #"Grouped Rows" = Table.Group(#"Changed Type", {"Country"}, {{"Average", each List.Average(List.Select([Value], each _ >0)), type number}, {"AllRows", each _, type table [Country=text, Value=number]}}),
    #"Expanded AllRows" = Table.ExpandTableColumn(#"Grouped Rows", "AllRows", {"Value"}, {"Value"})
in
    #"Expanded AllRows"

     

    This function of the group-function makes the average without 0

    {"Average", each List.Average(List.Select([Value], each _ >0)), type number}

     

    Copy paste this code to the advanced editor in a new blank query to see how the solution works.

    If this post helps or solves your problem, please mark it as solution (to help other users find useful content and to acknowledge the work of users that helped you)
    Kudoes are nice too

    Have fun

    Jimmy

6 Replies

  • Fowmy's avatar
    Fowmy
    Icon for Super User rankSuper User
    5 years ago

    anvikuttu 

    Click Add Column < Custom and paste the below code to get the average excluding zero and by country.

    (r)=> 
List.Average(
Table.SelectRows(#"Changed Type", each [VALUE] <> 0 and [COUNTRY] = r[COUNTRY])[VALUE])

    ________________________

    If my answer was helpful, please consider Accept it as the solution to help the other members find it

    Click on the Thumbs-Up icon if you like this reply 🙂

    YouTube  LinkedIn

  • CNENFRNL's avatar
    CNENFRNL
    Icon for Community Champion rankCommunity Champion
    5 years ago

    Hi, Fowmy's solution surely works. When your dataset consists of 2k+ rows, you might try this,

    let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WCg1W0lEyNFCK1YGxTRFsIyQ2VIk3SBiJDVPuDVfiGIrKNERmQ1Q7OyLEwWwgMxYA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Country = _t, Value = _t]),

    #"Added Custom" = Table.AddColumn(Source, "Custom", each Table.Group(Source, {"Country"}, {{"grouped", each _}}){[Country = [Country]]}[grouped][Value]),
    Avg = Table.TransformColumns(#"Added Custom", {{"Custom", each List.Average(List.RemoveItems(List.Transform(_, Number.From), {0}))}})
in
    Avg
    • anvikuttu's avatar
      anvikuttu
      Icon for Advocate I rankAdvocate I
      5 years ago

      Hi CNENFRNL ,

      Could you please help me understand the 2nd and 3rd step?

      Also kindly let me know the best way to learn the power query... I followed a book to learn, but it looks like I don't know anything 😞

       

      Thanks and regards,

      Vj

      • CNENFRNL's avatar
        CNENFRNL
        Icon for Community Champion rankCommunity Champion
        5 years ago

        Hi, anvikuttu , in my code, step 2 and 3 finish the following tasks respectively,

        step 2 adds an extra column; in this column each row contains a list of all values of the same country;

        step 3 averages the list in step 2 excluding 0.

         

        As to the study of M language, I myself is also a learner for half of a year or so. I recommend such a roadmap,

        Study fundamental concepts of M, in particular, table/list/record, relationships and conversions from one to another;

        Spend some time studying codes by UI;

        Try to master some most frequently used functions such as Tabel.AddColumn, Table.TransformColumns, etc.

        Use advanced editor to practise user defined functions.

  • ziying35's avatar
    ziying35
    Icon for Impactful Individual rankImpactful Individual
    5 years ago

    anvikuttu 

    Try this:

    // output
let
    Source = Table.FromRecords(Json.Document(Binary.Decompress(Binary.FromText("i65Wcs4vzSspqlSyUgoNVtJRCkvMKU1VsjI0qNXBKWeKW84Ijxy6kd5I2vDIYVjnjdNIx1CipNA9hyKHZpuzI259yHIGtbEA", BinaryEncoding.Base64),Compression.Deflate))),
    group = Table.Group(Source, {"Country"}, {"Foo", each Table.SplitColumn(_, "Value", (val)=>{val, List.Average(List.RemoveItems([Value], {0}))}, {"Value", "Avg"}), type table}),
    result = Table.Combine(group[Foo])
in
    result