Hi @Agustín Muñoz ,

First of all, if I understand your question correctly, you want to calculate the (average) throughput between the first occurrence of either A or B to the first occurrence of C or D? I'm a bit confused by what you mean with 'A and B to C and D'.

Second, your code snippet could work but I see two issues:

1. Both THEN statements are different, so you're averaging days (I guess) with hours
2. The ELSE 0 part messes up the result of AVG. It's best to return NULL instead (these values get ignored by AVG, whereas 0 is considered as a valid value, hence lowering the result).

I hope these two hints already help. Otherwise please provide clarification on what you are exactly looking for.

Hi!

I am sorry, It could be a bit complex.

I am going to give you an example: imagine that we have 3 different activities:

• Order confirmation
• Order confirmation - Automatism
• Order recieved

The KPI I want to calculate is the average duration from the first occurrence of 'Order confirmation' or 'Order confirmation - Automatism' to the first 'Order received'.

For this, I need a formula that calculate the duration between the first "Order confirmation" or "Order confirmation - Automatism" (which one occurs first) and first "Order recieved", but I dont know how to get it.

Thank you sou much,

Agustín

My first thought is a construct like this:

AVG(DATE_DIFF(dd,

PU_FIRST(case_table,activity_table.timestamp, activity_table.name IN ('Order confirmation ','Order confirmation - Automatism')),

PU_LAST(case_table,activity_table.timestamp, activity_table.name = 'Order recieved')

)

Could you try this and me let/us know?

Thank you so much!!! It works perfectly.

Best regards,

Agustín