LAE

Welcome to the LAE community!  Please feel free to start a discussion in the discussion tab or join in a conversation.

Discussions

Members

Resources

Events

 View Only

  • 1.  Data to names node or Agg Ex?

    Employee
    Posted 10-15-2018 23:03

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: I'mNewHere

    Hi Folks,

    Trying to change my dataset from


    lsexample.jpg

    to a cross reference of dates and states, with the calculated FCR result as the values.

    I have an agg ex node right now producing the results by state, but I don't have a time dimension, which is what I'm trying to change. to avoid confusion, I've pulled the month number form the date, so myDate is 01 for Jan, etc

    R7 = sum(int('Repeat Calls 7'))
FNN7 = sum(int('Eligible FCR Calls - FNN 7'))
FCR7 = double(FNN7 - R7) / double(FNN7)
emit referencedFields(1,{{^GroupBy^}}), FCR7
where lastInGroup

Group by 'state'
    If you could, please use laymen terms, as I'm pretty new to lavastorm


    I read this thread as a simple inverse of what I'm trying to do, but without the calculation step, but I couldnt figure out whether to use the data to names as it didn't appear to have input fields for a calculation.
    http://community.lavastorm.com/threa...-Column-Values


  • 2.  RE: Data to names node or Agg Ex?

    Employee
    Posted 10-16-2018 03:06

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: awilliams1024

    Hi,

    I'm not sure of the requirements for your final results but the attached example may give you some pointers. It shows two use cases.
    - The first performs the calculation by date only and then transposes the data
    - The second calculates the measure by state and date and transposes the data. (I assume this is what you wanted).

    For both use cases the month number is converted back to a month name string.
    Also, the calculation in the Agg Ex forces the FCR7 value to 0 if both the FNN7 and R7 are 0 (I'm not sure of the definition of the measure) but you may want a different outcome. The key issue here is to trap the case where the R7 value is 0, else you will get a divide by zero error.

    FCR7_Timeline.brg

    Regards,

    Adrian


  • 3.  RE: Data to names node or Agg Ex?

    Employee
    Posted 10-16-2018 07:31

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: stonysmith

    One comment "in simple terms"

    Unlike Excel, our Pivot node requires that all computations be done BEFORE you pivot.
    So, it's best to use an AGG (or Agg-Ex) to compute the values, then use Pivot to rotate it.

    ===
    I note that you're using sum() and count()
    You need to upgrade that to groupSum() and groupCount()
    Gives exactly the same result, but the first two are being phased out.


  • 4.  RE: Data to names node or Agg Ex?

    Employee
    Posted 10-16-2018 07:59

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: awilliams1024

    Good catch Stony!
    Yes, the sum() and count() functions are deprecated in the Lavastorm LAE 6.1 product but are not supported in the next major release of the product, which is part of the Data3Sixty(TM) platform and is now named Data3Sixty Analyze.

    We recommend that all users review their graphs and update them to use the groupSum() and groupCount() functions before attempting to import the graphs into Data3Sixty Analyze.


  • 5.  RE: Data to names node or Agg Ex?

    Employee
    Posted 10-16-2018 16:49

    Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

    Originally posted by: I'mNewHere

    Thank you folks,

    Your info was extremely useful to someone learning as I am.

    awilliams, your different variations were helpful as the group by date allowed me to do some validations in addition to providing the group by state result I was after.

    Thank you very much!