Multiply Values from each Series - Manufacturing Yield by Station

Hello!  

 

I am trying to determine the overall yield of a manufacturing line.  The data columns include the date, the station, which is a column of row labels representing each station in the manufacturing line, the number produced, and the number of defects.

 

I am able to find the yield by station, by creating a yield function in Beastmode (1 - SUM(`Defect Quantity`) / SUM(`Production Quantity`)), then creating a graph with the station names along the x axis and the yield % along the y axis.  However, I am trying to find the overall yield for the line, which requires multiplying together the yield of each station.  I know its possible in an ETL, but I want the card to be responsive with other cards.  E.g., When I select a time period in another card, I would want the overall yield to update as well.  

 

Below is my attempt to create a Beastmode function, but the output is not correct.  For example, if Station A has 3 defects out of 10, and station B has 2 defects out of ten, the overall yield of the line would be (.70)*(.80) = .56.  However, this function would evaluate to (7+8)/(10+10)  = 0.75.  Any ideas of how to tackle this?

 

(CASE when `Station` = 'Station A' then 1 - SUM(`Defect Quantity`) / SUM(`Production Quantity`) else 1 End ) * (CASE when `Station` = 'Station B' then 1 - SUM(`Defect Quantity`) / SUM(`Production Quantity`) else 1 End )

Comments

  • MarkSnodgrass
    MarkSnodgrass Portland, Oregon 🟤

    In your sample scenario, it seems like the overall yield should be 75%, not 56%. I'm not clear as to why you would be multiplying the individual station yields together. If you thought of all the production as one station, you would have 7+8 and 10+10 which would give you 15/20 or 75%.