Yearly average of an aggregate measure, again

[Johannes_W]

 Dear all,

it seems this kind of question comes up often, but I can not seem to find a solution.

Here is my situation.

I have a list of prcesses that happen at a specific date. Each process has a number of events (some just a few, some hundreds) which may or may not fail.

Table1

Process	event_id	failed?
A	1	0
A	2	0
A	3	1
B	4	1
B	5	1
C	6	0
C	7	0
D	8	0
D	9	1
D	10	1
D	11	0

I also have a date table that tells me in which year the processes have run

A	2020
B	2020
C	2019
D	2019

I have created a measure that claculates for each process the percentage of failed events

%failed = DIVIDE( SUM('Table1'[failed]), COUNT('Table1'[event_id]) )

very naively, I just sum up the 1s and divide by the number of events.

This works nicely when I look at the individual processes.

Process	%failed	Year
A	33%	2020
B	100%	2020
C	0%	2019
D	50%	2019

However: what I would like to achieve is a measure that now gives me the average %failed over all processes each year

(2020 = 66.5%, 2019 = 25%)

When I simply slice by year, what I get is the total % of all EVENTS in that year, not the average percentage by process.

(2020 = 60% and 2019 = 40 %)

This doesn't help me, because as I said, the number of events can differ vastly thus ruining the statistics.

Any advice is apprechiated.

Thanks!

Joh

[Jihwan_Kim · Accepted solution]

Hi, @Johannes_W 

Please check the below picture and the sample pbix file's link down below.

When you put the Process column into the visualization, the correct result is still shown.

failed percentage =

VAR newtable =

SUMMARIZE (

Data,

Data[Process],

"@idcount", COUNT ( Data[event_id] ),

"@failed", SUM ( Data[failed] )

)

RETURN

AVERAGEX ( newtable, [@failed] / [@idcount] )

https://www.dropbox.com/s/6evisu06nqw6qlv/johannes.pbix?dl=0 

Hi, My name is Jihwan Kim.

If this post helps, then please consider accept it as the solution to help other members find it faster, and give a big thumbs up.

Linkedin: linkedin.com/in/jihwankim1975/

Twitter: twitter.com/Jihwan_JHKIM

[Johannes_W]

Dear Jihwan,

that works perfectly, thank you so much! I need to ponder on why it works though.

Thanks!

Joh