Dividing in SQL

gbennett

I currently have this code: categorizing some leads.

Select

    `Year-Month` 

,`Lead Source`

   ,`Product Line`

   ,`Product Models`

   ,`Country`

   ,`Early Stage No Movement`

   ,`Stage 0 to Early`

   ,`Early to Lost`

   ,`Early to Won`

   ,`Early to Late`

   ,`Late to Lost`

   ,`Late to Won`

,Sum(`Early to Lost`+`Early to Won`+`Early to Late`) AS 'Total Low to High'

,SUM(`Early Stage No Movement`+`Stage 0 to Early`+`Early to Lost`+`Early to Won`+`Early to Late`+`Late to Lost`+`Late to Won`) AS 'Total'

FROM `aggregation_categorization`

GROUP BY

`Year-Month`

,`Lead Source`

,`Product Line`

,`Product Models`

,Country

This runs fine, the next thing I am trying to do is add in the %.

SELECT *

,`Total Low to High`/`Total` AS '% Of Low to High'

FROM `formulas`

This works fine as well, but when I place it in a sumo table aggregating it does work. Any suggestions would be awesome.

Thank you.

jaeW_at_Onyx

Given the math you've put on the screen, there is no reason to run this ETL as a SQL transform. Just keep your data unaggregated.

Be careful with

Sum(`Early to Lost`+`Early to Won`+`Early to Late`) AS 'Total Low to High'

for any row, if any column contains the value null, the entire result will be null. this is the appropriate adjustment.

Sum(ifnull(`Early to Lost`,0) +ifnull(`Early to Won`,0)+ifnull(`Early to Late`,0) ) AS 'Total Low to High'

to do the division you're describing, just calculate the ratio in a beast mode (you may have to build the beast mode in analyzer, then share it to the dataset in order to reference it in a sumo card.

Personally i recommend you don't use the Sumo card and just use the Pivot Table card in Analyzer.