How do we aggregate time series in Excel? - excel

Question

Given the following time series of cashflow, how can I aggreate them into a cumulative time series of cashflow in Excel, ideally by using array formula only and without VBA macro?
Specifically, I was given this time series of cashflow for each transaction:
Given the inputs (in column F) for the number of transactions in each period, I would like to be able to calculate the aggregated time series of total cashflow (in column G, highlighted in yellow), ideally by using array formula only and without VBA macro?
Note: Column H to J are for illustrations only to show how column G should be calculated, I don't want to have them in my final spreadsheet.
Thank you very much for your help!

Answer 1

I believe you can do it by formula - most easily by reversing the cash flows and multiplying by the current and previous 5 transactions:
=SUMPRODUCT(INDEX(F:F,MAX(ROW()-5,3)):F16*INDEX(C:C,MAX(11-ROW(),3)):$C$8)
in G3.
This is an ordinary non-array formula.

Answer 2

OK Put this array formula in G3:
=IFERROR(SUMPRODUCT(INDEX($B$2:$B$7,N(IF({1},MODE.MULT(IF(INDEX(F:F,MAX(ROW()-5,3)):F3>0,(ROW()-ROW(INDEX(F:F,MAX(ROW()-5,3)):F3)+1)*{1,1}))))),INDEX(INDEX(F:F,MAX(ROW()-5,3)):F3,N(IF({1},MODE.MULT(IF(INDEX(F:F,MAX(ROW()-5,3)):F3>0,(ROW(INDEX(F:F,MAX(ROW()-5,3)):F3)-MIN(ROW(INDEX(F:F,MAX(ROW()-5,3)):F3))+1)*{1,1})))))),0)
Being an array formula it must be confirmed with Ctrl-Shift-Enter instead of Enter when exiting edit mode. Then copy down.
Once Microsoft relaeases FILTER and SEQUENCE it can be shortened:
=IFERROR(SUMPRODUCT(INDEX($B$2:$B$7,FILTER(SEQUENCE(ROW()-MAX(ROW()-5,3)+1,,ROW()-MAX(ROW()-5,3)+1,-1),INDEX(F:F,MAX(ROW()-5,3)):F3>0)),FILTER(INDEX(F:F,MAX(ROW()-5,3)):F3,INDEX(F:F,MAX(ROW()-5,3)):F3>0)),0)

Answer 3

This can also be done in Power Query.
Please refer to this article to find out how to use Power Query on your version of Excel. It is available in Excel 2010 Professional Plus and later versions. My demonstration is using Excel 2016.
Steps are:
Load both tables being the time series of cash-flow and your 2-column output table to the power query editor, then you should have:
For the first table, merged the Period column with Cashflow column with semicolon ; as the delimiter;
Transpose the column/table, then merge the columns with comma , as the delimiter;
Add a custom column use this formula ="Connector" which will fill the column with the word Connector, then you should have:
For the second table, also add a custom column use the same formula ="Connector" which will fill the column with the word Connector;
Merge the second table with the first table using the Custom column as the connection, then expand the new column to show the Merged column from the first table, then you should have:
Remove the Custom column, then split the Merged column by delimiter comma , and put the results into Rows;
Split the Merged column again by delimiter semicolon ; to separate the Period and Cashflow from the first table;
Add a custom column to calculate the New Period being =[Period]+[Merged.1];
Add another custom column to calculate the Cashflow being =[#"# Tran"]*[Merged.2], then you should have something like the following:
Group/sum the Cashflow column by New Period.
Once done you can Close & Load the result to a new worksheet (by default). If you want to show the # Trans column in the final output, you can make a duplicate of your second table before making any changes, and then merge it with the final output table by the Period column to show the corresponding number of transactions.
Here are the power query M codes for the first table:
let
Source = Excel.CurrentWorkbook(){[Name="Tbl_CFS"]}[Content],
#"Changed Type" = Table.TransformColumnTypes(Source,{{"Period", Int64.Type}, {"Cashflow", Int64.Type}}),
#"Merged Columns1" = Table.CombineColumns(Table.TransformColumnTypes(#"Changed Type", {{"Period", type text}, {"Cashflow", type text}}, "en-AU"),{"Period", "Cashflow"},Combiner.CombineTextByDelimiter(";", QuoteStyle.None),"Merged"),
#"Transposed Table" = Table.Transpose(#"Merged Columns1"),
#"Merged Columns" = Table.CombineColumns(Table.TransformColumnTypes(#"Transposed Table", {{"Column1", type text}, {"Column2", type text}, {"Column3", type text}, {"Column4", type text}, {"Column5", type text}, {"Column6", type text}}, "en-AU"),{"Column1", "Column2", "Column3", "Column4", "Column5", "Column6"},Combiner.CombineTextByDelimiter(",", QuoteStyle.None),"Merged"),
#"Added Custom" = Table.AddColumn(#"Merged Columns", "Custom", each "Connector")
in
#"Added Custom"
And here are the codes for the second table:
let
Source = Excel.CurrentWorkbook(){[Name="Tbl_Total"]}[Content],
#"Changed Type" = Table.TransformColumnTypes(Source,{{"Period", Int64.Type}, {"# Tran", Int64.Type}}),
#"Added Custom" = Table.AddColumn(#"Changed Type", "Custom", each "Connector"),
#"Merged Queries" = Table.NestedJoin(#"Added Custom", {"Custom"}, Tbl_CFS, {"Custom"}, "Tbl_CFS", JoinKind.LeftOuter),
#"Expanded Tbl_CFS" = Table.ExpandTableColumn(#"Merged Queries", "Tbl_CFS", {"Merged"}, {"Merged"}),
#"Removed Columns" = Table.RemoveColumns(#"Expanded Tbl_CFS",{"Custom"}),
#"Split Column by Delimiter" = Table.ExpandListColumn(Table.TransformColumns(#"Removed Columns", {{"Merged", Splitter.SplitTextByDelimiter(",", QuoteStyle.Csv), let itemType = (type nullable text) meta [Serialized.Text = true] in type {itemType}}}), "Merged"),
#"Changed Type1" = Table.TransformColumnTypes(#"Split Column by Delimiter",{{"Merged", type text}}),
#"Split Column by Delimiter1" = Table.SplitColumn(#"Changed Type1", "Merged", Splitter.SplitTextByDelimiter(";", QuoteStyle.Csv), {"Merged.1", "Merged.2"}),
#"Changed Type2" = Table.TransformColumnTypes(#"Split Column by Delimiter1",{{"Merged.1", Int64.Type}, {"Merged.2", Int64.Type}}),
#"Added Custom1" = Table.AddColumn(#"Changed Type2", "New Period", each [Period]+[Merged.1]),
#"Added Custom2" = Table.AddColumn(#"Added Custom1", "Cashflow", each [#"# Tran"]*[Merged.2]),
#"Grouped Rows" = Table.Group(#"Added Custom2", {"New Period"}, {{"Sum", each List.Sum([Cashflow]), type number}})
in
#"Grouped Rows"
All steps are using built-in functions so should be straight forward and easy to execute. Let me know if there is any question. Cheers :)