Trying to figure out but.... What is a "type chart" or a "number chart"? I am not sure if you mean a OLAP table, or a Line Chart

I would also like to know that.

Do you mean a number component and a column chart. And facing issue if you selected a dimension for column chart?

Thank you for your answers. Please excuse me, with "type" chart I referred to the Single KPI component of type number.

The problem occurs within the column chart, where I selected month as dimension. However, all eventtime values are of month Juli 2022, so this should not cause any problem.

How does it affect if you use July 2022 and the days? Not only July 2022.

Can you show a picture of the OLAP Table? I would put everything as a dimension besides the KPI.

Dear Nihat,

thank you for the hint. Changing the formatting has no influence on the result.

The first OLAP table shows the dimension "eventtime in month" with the mentioned KPI, while the second shows all Dimensions, as you suggested.

Greetings,

Celina

Didn't you say that the single KPI shows 8? The sum of the column KPI is 10 like the single KPI, am I wrong?

What does happen if you just use

PU_COUNT_DISTINCT(

 DOMAIN_TABLE("S2A_AKTIVITAETEN_DETAILS_xlsx_Sheet1"."CASEKEY","S2A_AKTIVITAETEN_DETAILS_xlsx_Sheet1"."EVENTTIME"),

 "S2A_AKTIVITAETEN_DETAILS_xlsx_Sheet1"."AKTIVITAET","S2A_AKTIVITAETEN_DETAILS_xlsx_Sheet1"."AKTIVITAET" = 'Eskalation durchgeführt')

?

By using the KPI formula on any single kpi component (in my case type number), I receive 8 as a result.

Applying the same formula on any chart or table component, I receive a 10.

I am just wondering why the same formula has different results for the two different component types.

Using the formula without the sum just returns the single values for all table rows:

Damit wollte ich überprüfen, ob es beim CASEKEY 94 den EVENTTIME nicht unterscheiden kann, da dieser ja mit DISTINCT eine 1 ergeben müsste. Könnte daran liegen, dass wir dann 3 cases haben, wodurch die Summe dann 10 ergibt, weil wir die einzelnen cases + Eventtime anschauen.

Ok Celina, congratulations your question intrigued me so much I couldn't sleep 😁

I have reproduced the behaviour and found some clues. I can't fully explain it, but I hope with my input you or other Celopeer will be able to find it.

Let's go:

• I made a copy of your data, and uploaded
• I put you PU_COUNT_DISTINCT into a KPI
• Verified that in a Number widget, you get 8, and in a column chart, 10
• Verified that, if I put the KPI in a OLAP table, I get also 10
• Now, IF I add a GLOBAL clause around your KPI, then I get 8

Notice here... if you add the PU_COUNT_DICTINCT column ... it sums 10

so what happens is that 8 is the total of different case+timestamp....

and 10 is the total sum of the individual count distinct per case in a month ... so 10.

HTH

Yes, I also came to that conclusion..

Thank you for your work!

Yep, they are different things... the number is the total distict values. In total. Like "tell me all the distinct values in the table of the key case_id + eventide.

In OLAP and Column, what you get is, per each key (caseID+timestamp) the number of distinct values... and then grouped all from the same month using a sum.

Like "tell me per each case how many different values they have for each key (case id plus eventide)"... and then group it per month.

Dear Guillermo,

thank you so much for your effort, help, and explanation.

Greeting

Celina