Beast Mode not calculating for Current Year to Date

user18178
user18178 Member
in Magic ETL

I am stumped on this one...

 

I created a beast mode to calculate the number of rejections as a percentage of sales.  Specific lines of data have reason codes to flag what type of transaction it is.  So F, H, etc are specific type of rejection lines.  The "0" is a normal sales transaction.  The beast mode below calculates properly when the date range is set to "All Time" or  "Previous Year".  For some reason it does not calculate when it is set to "Current Year".  I have checked my source data and the date is in the correct format.  I am not seeing a reason it should not calculate for "Current Year" (2019).  

 

(ABS(SUM(case when `Code` = 'F' then `quantity` end))
+
ABS(SUM(case when `Code` = 'H' then `quantity` end))
+
ABS(SUM(case when `Code` = 'I' then `quantity` end))
+
ABS(SUM(case when `Code` = 'J' then `quantity` end))
+
ABS(SUM(case when `Code` = 'T' then `quantity` end)))
/
(SUM(case when `Code` = 0 then `quantity` end))

Best Answer

  • ST_Superman
    ST_Superman Domo Employee
    Answer ✓

    Try adding some else 0 clauses to your case statments:

    (ABS(SUM(case when `Code` = 'F' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'H' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'I' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'J' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'T' then `quantity` ELSE 0 end)))
    /
    (SUM(case when `Code` = 0 then `quantity` ELSE 0 end))

     

    Essentially, if you are missing any of these Codes for current year data, then this beastmode will stop calculating.  By adding the "ELSE 0" you are giving the beastmode a way forward when the particular code does not exist in the dataset

Answers

  • ST_Superman
    ST_Superman Domo Employee
    Answer ✓

    Try adding some else 0 clauses to your case statments:

    (ABS(SUM(case when `Code` = 'F' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'H' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'I' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'J' then `quantity` ELSE 0 end))
    +
    ABS(SUM(case when `Code` = 'T' then `quantity` ELSE 0 end)))
    /
    (SUM(case when `Code` = 0 then `quantity` ELSE 0 end))

     

    Essentially, if you are missing any of these Codes for current year data, then this beastmode will stop calculating.  By adding the "ELSE 0" you are giving the beastmode a way forward when the particular code does not exist in the dataset

  • user18178
    user18178 Member

    This worked. Thank you for the explanation. Good to know that the beast mode will stop calculating at any part of the formula once it does not find that data in the time frame of the data set. I assumed that the formula in total would never equal zero so I dropped the "else 0" off.  I now know that is not the case.  Thanks again. 

Categories