Return value from prior date

OK, jwin2424 . I think I understand your requirements. You want to get previous row group by one or two fields.

Here's a blog about this which really helps. Please refer to

Value from previous row – Power Query, M language – Trainings, consultancy, tutorials (exceltown.com)

The second part from this blog(Value from previous row, without writing of M code) can fix this.

I made a sample for you also, you can download it.

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WMtE31DcyMDJR0lHyCvXRBTMMjYyBpKmpUqwOSIERQoEfVgXGhBSYEFJgSkiBBUyBr2MkVgWGWEwwIeQLE0K+MCHkCxNCvjAh5AuYglgA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [#"Run Date" = _t, #"Scheduling Month" = _t, #"Order Line ID" = _t, #"Order #" = _t]),
    #"Grouped Rows" = Table.Group(Source, {"Order Line ID"}, {{"allrows", each _, type table [Run Date=nullable text, Scheduling Month=nullable text, Order Line ID=nullable text, #"Order #"=nullable text]}}),
    #"Added Custom" = Table.AddColumn(#"Grouped Rows", "Custom", each Table.AddIndexColumn([allrows],"index",1,1)),
    #"Removed Columns" = Table.RemoveColumns(#"Added Custom",{"allrows"}),
    #"Expanded Custom" = Table.ExpandTableColumn(#"Removed Columns", "Custom", {"Run Date", "Scheduling Month", "Order #", "index"}, {"Run Date", "Scheduling Month", "Order #", "index"}),
    #"Duplicated Column" = Table.DuplicateColumn(#"Expanded Custom", "index", "index - Copy"),
    #"Added to Column" = Table.TransformColumns(#"Duplicated Column", {{"index - Copy", each _ + 1, type number}}),
    #"Merged Queries" = Table.NestedJoin(#"Added to Column", {"index", "Order Line ID"}, #"Added to Column", {"index - Copy", "Order Line ID"}, "Added to Column", JoinKind.LeftOuter),
    #"Expanded Added to Column" = Table.ExpandTableColumn(#"Merged Queries", "Added to Column", {"Scheduling Month"}, {"Scheduling Month.1"})
in
    #"Expanded Added to Column"

Best Regards,

Stephen Tao

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