mathematical mystery....

[#1]

Here is the mystery

Gathering information from 5 excel list in different languages i need to do an overview of the situaiton.

First figure is the following formula

='Participants : ' &num(Count(DISTINCT [Adresse e-mail])) 

second graph...is the following...

=num(Count(DISTINCT [Adresse e-mail])) 

with DImensions Language.

guess what i have a différence of 63 participants....any clue...

[sunny_talwar]

Distinct is based on the dimension. When you use it in your pie chart, it can be double counted based on Addresse e-mail occurrence in two or more different language

[avinashelite]

I feel the text is including the [Adresse e-mail] where the language is null but in the pivot table it doesn't include the null hence you might be getting this difference

[Kushal_Chawda]

Your text box expression should be match when you will use aggr.

='Participants : ' & sum(aggr(Count(DISTINCT [Adresse e-mail]),Language))

[#5]

so that mean that somebody has been able to register in FRENCH for example AND in ENGLISH ...therefore is counted as 2 ??

is that right ?

[sunny_talwar]

Yes, because they are in FRENCH and ENGLISH, they are counted twice in the pie chart, but not in the text box object

[#7]

ok got your point...but does that means that some people where able to register in different languages then ??

[Kushal_Chawda]

Yes, might be possible. Also check if there are any User who does not have any Language

[ashishkalia]

Only two possibilities:-

1. single person applying for multiple courses
2. If in your field [Adresse e-mail] you by any chance have taken out the name and counting.

        i.e. email:- jean@example.com and field [Adresse e-mail] containing only jean its quit possible to have different users with same names.

Else not

[Peter_Cammaert]

You can do a simple check;

Create a Straight table with Dimension [Adresse e-mail] and expression =count(DISTINCT Language). Sort the expression value in descending order. The multiples should come out at the top. If there are any.

Best,

Peter