The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredCompete to become Power BI Data Viz World Champion! First round ends August 18th. Get started.
Hello,
I've searched in forums and googled but I haven't found answer how to replace value if the condition is based on one column grouping and value of another column. If to look the sample data (there are other columns as well which are not basis of the replacing)
Currency | Node | Type |
EUR | 180d | OIS |
EUR | 180d | IB |
EUR | 270d | OIS |
EUR | 270d | IB |
EUR | 270d | TR |
EUR | 1yr | SW |
GBP | 7d | IB |
USD | 3d | IB |
USD | 3d | TR |
What I need is that under each currency separately where Type "OIS" is reprsented everything else beside "TR" is changed to "OIS". If there's no "OIS", it should remain as it is. Result should be as follows:
Currency | Node | Type | Replace |
EUR | 180d | OIS | OIS |
EUR | 180d | IB | OIS |
EUR | 270d | OIS | OIS |
EUR | 270d | IB | OIS |
EUR | 270d | TR | TR |
EUR | 1yr | SW | OIS |
GBP | 7d | IB | IB |
USD | 3d | IB | IB |
USD | 3d | TR | TR |
I can't use group by since I need the table to remain in the same format since there are other fields which I use in latter steps based on this new field.
Thanks in advance.
Solved! Go to Solution.
let
Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45Wcg0NUtJRMrQwSAFS/p7BSrE6aIKeTkhiRuZYFEIFsSkMCUI2sLIISAaHg4XcnQKAHHMkjaHBLkCOMXYRkEGxAA==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Currency = _t, Node = _t, Type = _t]),
#"Changed Type" = Table.TransformColumnTypes(Source,{{"Currency", type text}, {"Node", type text}, {"Type", type text}}),
#"Grouped Rows" = Table.Group(#"Changed Type", {"Currency"}, {{"Rows", each _, type table [Currency=nullable text, Node=nullable text, Type=nullable text]}}),
#"Added Custom" = Table.AddColumn(#"Grouped Rows", "has OIS", each List.Contains([Rows][Type],"OIS")),
#"Expanded Rows" = Table.ExpandTableColumn(#"Added Custom", "Rows", {"Node", "Type"}, {"Node", "Type"}),
#"Added Custom1" = Table.AddColumn(#"Expanded Rows", "Replace", each if [Type]="TR" or [has OIS] = false then [Type] else "OIS"),
#"Removed Other Columns" = Table.SelectColumns(#"Added Custom1",{"Currency", "Node", "Type", "Replace"})
in
#"Removed Other Columns"
How to use this code: Create a new Blank Query. Click on "Advanced Editor". Replace the code in the window with the code provided here. Click "Done".
Great. I used it in my code from #"Changed Type" and it works. Thank you.
let
Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45Wcg0NUtJRMrQwSAFS/p7BSrE6aIKeTkhiRuZYFEIFsSkMCUI2sLIISAaHg4XcnQKAHHMkjaHBLkCOMXYRkEGxAA==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Currency = _t, Node = _t, Type = _t]),
#"Changed Type" = Table.TransformColumnTypes(Source,{{"Currency", type text}, {"Node", type text}, {"Type", type text}}),
#"Grouped Rows" = Table.Group(#"Changed Type", {"Currency"}, {{"Rows", each _, type table [Currency=nullable text, Node=nullable text, Type=nullable text]}}),
#"Added Custom" = Table.AddColumn(#"Grouped Rows", "has OIS", each List.Contains([Rows][Type],"OIS")),
#"Expanded Rows" = Table.ExpandTableColumn(#"Added Custom", "Rows", {"Node", "Type"}, {"Node", "Type"}),
#"Added Custom1" = Table.AddColumn(#"Expanded Rows", "Replace", each if [Type]="TR" or [has OIS] = false then [Type] else "OIS"),
#"Removed Other Columns" = Table.SelectColumns(#"Added Custom1",{"Currency", "Node", "Type", "Replace"})
in
#"Removed Other Columns"
How to use this code: Create a new Blank Query. Click on "Advanced Editor". Replace the code in the window with the code provided here. Click "Done".