power query or DAX: move data when doublets/triplets in a column

[Sober2010]

Hi,

I have extracted price data via a simple sql.

Som itemID have more than one price. In my sql extract I get a new line for each ItemID, which means I get Double or triple lines with same ItemID but different prices. I don't have knowledge to correct my sql so I only get one line per ItemID

I hope that I may be able to do the adjustments in Power query or with DAX.

I need one line per ItemID, and the data moved as shown in the figure.

Does anyone have a hint ?

[Anonymous · Accepted solution]

Hi @Sober2010 

You can put the following code to advanced editor in power query

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WMjQ0VNJRykvMTQXRhgYGENJAKVYnWsnIyAgqaQSXNMEhByRMoVLGxsZQKWNMM01MTKCSJpiSpqamUElTuKQxVjlTZLega4O5JRYA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Item = _t, Itemname = _t, Qty = _t, Price = _t]),
    #"Changed Type" = Table.TransformColumnTypes(Source,{{"Item", Int64.Type}, {"Itemname", type text}, {"Qty", Int64.Type}, {"Price", Int64.Type}}),
    #"Unpivoted Columns" = Table.UnpivotOtherColumns(#"Changed Type", {"Item", "Itemname"}, "Attribute", "Value"),
    #"Grouped Rows" = Table.Group(#"Unpivoted Columns", {"Item", "Itemname", "Attribute"}, {{"Count", each _, type table [Item=nullable number, Itemname=nullable text, Attribute=text, Value=number]}}),
    #"Added Custom" = Table.AddColumn(#"Grouped Rows", "Custom", each Table.AddIndexColumn([Count],"Index",1,1)),
    #"Removed Columns" = Table.RemoveColumns(#"Added Custom",{"Count"}),
    #"Expanded Custom" = Table.ExpandTableColumn(#"Removed Columns", "Custom", {"Value", "Index"}, {"Value", "Index"}),
    #"Inserted Merged Column" = Table.AddColumn(#"Expanded Custom", "Merged", each Text.Combine({[Attribute], Text.From([Index], "en-US")}, " "), type text),
    #"Removed Columns1" = Table.RemoveColumns(#"Inserted Merged Column",{"Attribute", "Index"}),
    #"Pivoted Column" = Table.Pivot(#"Removed Columns1", List.Distinct(#"Removed Columns1"[Merged]), "Merged", "Value")
in
    #"Pivoted Column"

Output

Best Regards!

Yolo Zhu

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[mussaenda]

Hi @Sober2010,

Can you please enlighten us why this step is needed?

I think it would be better if you will have a dim table for items and this table on your screenshot will be your fact table.

[Sober2010]

Hi Mussaenda. 

My thoughts was to built a table with only uniqe values in this way - so I could built relations with a fact table

I'm probably down a wrong path.  Those repeated items cause however problems for me.

I will try to split the sql extract up in several tables instead for one.