Remove duplicate row, as well as row that was duplicated.

HI Andrew__K ,

Steps taken:

1. Unique Row Identifier: Combination of all columns delimeted by "|" in the Table Query (Your input)

2. Created a new query (RowsToRemove) referencing this

3. Grouped this query by Serial along with a summarized table of all rows and a count of all distinct rows. If distinct count > 1 then there are duplicates.

4. Filtered for rows >1 . Expanded the table.

5. Merged Table and Rows to delete as a new query on the unique identifier (Output - shown below). The rows which don't have a unique idenitifer are the rows which didn't have any duplicates in the raw data and are retained.

Copy and paste the following queries in the advanced editor of a blank query:

Table:

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("pdKxDoIwEAbgV7l0UoKhV0B0VOtGlChODQOBDiRESK2Jvr1tIqOD3NY/93+XDqcUk7XVLGRQmKF9NhaK4gQuS/1oTDfabrj76KcwVa/adHXvXqV5QfkeNatCxZIIRSS4iN0AeZaiZ5ezvB1KyHf7Y+6yiBCnTsY59x2crQVJxySdkHRK0muSzkh6Q9JbikZO0n/cGiwCqXttdRssf2xazf8I6Wzxe7bVBw==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [#"(blank)" = _t, #"(blank).1" = _t, #"(blank).2" = _t, #"(blank).3" = _t, #"(blank).4" = _t, #"(blank).5" = _t]),
    #"Promoted Headers" = Table.PromoteHeaders(Source, [PromoteAllScalars=true]),
    #"Changed Type" = Table.TransformColumnTypes(#"Promoted Headers",{{"Date", type date}, {" Product PPN ", Int64.Type}, {"Description ", type text}, {"Prod Date", type date}, {" Serial", Int64.Type}, {"Trx Type", Int64.Type}}),
    #"Unique Row Identifier" = Table.AddColumn(#"Changed Type", "Unique Row Identifier", each Text.Combine({Text.From([Date], "en-CA"), Text.From([#" Product PPN "], "en-CA"), [#"Description "], Text.From([Prod Date], "en-CA"), Text.From([#" Serial"], "en-CA"), Text.From([Trx Type], "en-CA")}, "|"), type text)
in
    #"Unique Row Identifier"

RowstoRemove:

let
    Source = Table,
    #"Grouped Rows" = Table.Group(Source, {" Serial"}, {{"Data", each _, type table [Date=nullable date, #" Product PPN "=nullable number, #"Description "=nullable text, Prod Date=nullable date, #" Serial"=nullable number, Trx Type=nullable number, Unique Row Identifier=text]}, {"Distinct Row Count", each Table.RowCount(Table.Distinct(_)), Int64.Type}}),
    #"Filtered Rows" = Table.SelectRows(#"Grouped Rows", each ([Distinct Row Count] = 2)),
    #"Removed Other Columns" = Table.SelectColumns(#"Filtered Rows",{"Data"}),
    #"Expanded Data" = Table.ExpandTableColumn(#"Removed Other Columns", "Data", {"Date", " Product PPN ", "Description ", "Prod Date", " Serial", "Trx Type", "Unique Row Identifier"}, {"Date", " Product PPN ", "Description ", "Prod Date", " Serial", "Trx Type", "Unique Row Identifier"})
in
    #"Expanded Data"

Output:

let
    Source = Table.NestedJoin(Table, {"Unique Row Identifier"}, RowsToRemove, {"Unique Row Identifier"}, "RowsToRemove", JoinKind.LeftOuter),
    #"Expanded RowsToRemove" = Table.ExpandTableColumn(Source, "RowsToRemove", {"Unique Row Identifier"}, {"Unique Row Identifier.1"}),
    #"Filtered Rows" = Table.SelectRows(#"Expanded RowsToRemove", each ([Unique Row Identifier.1] = null))
in
    #"Filtered Rows"

thank you! Using your suggestions as well as others, I ended up grouping a merge of PPN-DATE-SERIAL with a row count and all rows column, filtering >1 on the row count and then expanding the all rows column.