The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends August 31st. Request your voucher.
Good morning, Thank you in advance for reading my case. I have a list of names and what these people have spent in several years.
is it possible to transform a table in powerquery to display the last 2 records for each name?
Logically, people who only have 1 record, show that record. Attached pbix example
attached pbix
https://www.dropbox.com/s/l8vhureetptsn0g/test%20%282%29.pbix?dl=0
Name Mount Year
andres | 1250 | 2020 |
andres | 2120 | 2021 |
andres | 4584 | 2022 |
andres | 5466 | 2023 |
felipe | 1365 | 2020 |
felipe | 1211 | 2021 |
jesus | 1520 | 2020 |
juan | 2000 | 2020 |
juan | 452 | 2021 |
pepe | 1000 | 2019 |
pepe | 1500 | 2021 |
pepe | 6554 | 2022 |
transform to:
Name Mount Year
andres | 4584 | 2022 |
andres | 5466 | 2023 |
felipe | 1365 | 2020 |
felipe | 1211 | 2021 |
jesus | 1520 | 2020 |
juan | 2000 | 2020 |
juan | 452 | 2021 |
pepe | 1500 | 2021 |
pepe | 6554 | 2022 |
Solved! Go to Solution.
let
Origen = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("bY7RCoMwDEX/Jc8+JLG3zG8pfRCWwWSIKP6/rq2uhT2VnpN7kxBoscWoI2Hm81GWgWIXaNrHOf0LVk74msaFJeFxfq62fYWimX/Z550TvUctJtv2FIA2gbtJRZsV5SAHremvX0VqUQ71gMtY23qHx38B530WPcV4AA==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Name = _t, Mount = _t, Year = _t]),
#"Tipo cambiado" = Table.TransformColumnTypes(Origen,{{"Mount", Int64.Type}}),
Personalizado1 = Table.Buffer(Table.Sort(#"Tipo cambiado",{{"Name", Order.Descending}})),
#"Grouped Rows" = Table.Group(Personalizado1, {"Name"}, {{"Record", each _, type table [Name=nullable text, Mount=nullable number, Year=nullable text]}}),
#"Added Custom" = Table.AddColumn(#"Grouped Rows", "Last2", each Table.LastN([Record], 2)),
#"Removed Other Columns" = Table.SelectColumns(#"Added Custom",{"Last2"}),
#"Expanded Last2" = Table.ExpandTableColumn(#"Removed Other Columns", "Last2", {"Name", "Mount", "Year"}, {"Name", "Mount", "Year"})
in
#"Expanded Last2"
¡wow! Thankyou very much... GSun!! Your are incredible.
let
Origen = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("bY7RCoMwDEX/Jc8+JLG3zG8pfRCWwWSIKP6/rq2uhT2VnpN7kxBoscWoI2Hm81GWgWIXaNrHOf0LVk74msaFJeFxfq62fYWimX/Z550TvUctJtv2FIA2gbtJRZsV5SAHremvX0VqUQ71gMtY23qHx38B530WPcV4AA==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Name = _t, Mount = _t, Year = _t]),
#"Tipo cambiado" = Table.TransformColumnTypes(Origen,{{"Mount", Int64.Type}}),
Personalizado1 = Table.Buffer(Table.Sort(#"Tipo cambiado",{{"Name", Order.Descending}})),
#"Grouped Rows" = Table.Group(Personalizado1, {"Name"}, {{"Record", each _, type table [Name=nullable text, Mount=nullable number, Year=nullable text]}}),
#"Added Custom" = Table.AddColumn(#"Grouped Rows", "Last2", each Table.LastN([Record], 2)),
#"Removed Other Columns" = Table.SelectColumns(#"Added Custom",{"Last2"}),
#"Expanded Last2" = Table.ExpandTableColumn(#"Removed Other Columns", "Last2", {"Name", "Mount", "Year"}, {"Name", "Mount", "Year"})
in
#"Expanded Last2"
If you sort by Name and then Year, you can add an index column for each Name group. Try and find this out yourself (on this forum or there are plenty of examples out there on the web. I will help if you get stuck).
Then filter the new Index column to keep only Index = 1 or 2