Forum Discussion

kimbtoth's avatar
kimbtoth
Frequent Visitor
2 years ago

Calculate the difference in a value from two rows

I have a spreadsheet that displays a score for each screening completed by a patient (identified by unique PAT_ID column) over time. NOTE - All screenshots are de-identified.

I am trying to calculate the difference in score (MEAS_VALUE column) for (MaxDate - MinDate) per each PAT_ID.

I was able to group by PAT_ID and limit to rows > 1 and calculate the difference in date duration since I included in on the row. I also was able to use Table-Buffer to sort PAT_ID and  DATE_RECORD Desc) but I can't figure out how to calculate the difference in score (MEAS_VALUE column)  across different rows but only when the PAT_ID is the same.

I tried reviewing previously submitted questions/responses but I'm not getting it.  Thanks for any help!
Screenings Raw DataScreenings Grouped

 

 

 

Need Help

2 Replies

  • Greg_Deckler's avatar
    Greg_Deckler
    Community Champion
    2 years ago

    kimbtoth Not sure of the Power Query solution, maybe ImkeF can help. Here is a DAX solution as a fall back if you just need to get it done. See my article on Mean Time Between Failure (MTBF) which uses EARLIER: http://community.powerbi.com/t5/Community-Blog/Mean-Time-Between-Failure-MTBF-and-Power-BI/ba-p/339586.
    The basic pattern is:
    Column = 
      VAR __Current = [Value]
      VAR __PreviousDate = MAXX(FILTER('Table','Table'[Date] < EARLIER('Table'[Date])),[Date])

      VAR __Previous = MAXX(FILTER('Table',[Date]=__PreviousDate),[Value])
    RETURN
      __Current - __Previous

  • ronrsnfld's avatar
    ronrsnfld
    Super User
    2 years ago

    You don't show what you want for results, but a simple way of calculating the MEAS_VAL difference between the latest and the earliest dates is within the Table.Group function:

    let
    Source = Excel.CurrentWorkbook(){[Name="Table14"]}[Content],
    #"Changed Type" = Table.TransformColumnTypes(Source,{
        {"PAT_ID", type text}, {"SCREENING_NAME", type text}, 
        {"DATE_RECORDED", type date}, {"ENTRY_USER_ID", type text}, 
        {"MEAS_VALUE", Int64.Type}}),
    
    #"Grouped Rows" = Table.Group(#"Changed Type", {"PAT_ID"}, {
        {"Diff" , 
            each 
                Table.Max(_,"DATE_RECORDED")[MEAS_VALUE] - 
                Table.Min(_,"DATE_RECORDED")[MEAS_VALUE], 
                type number}})
in
    #"Grouped Rows"

     

    Given this data:

     

    The above code produces: