Reputation: 47
I have a field customer_id and I need to track the number of unique users and repeat users. For example the table is as below:
customer_id |
---|
11 |
22 |
33 |
11 |
44 |
22 |
Here, the no. of unique users is 4 (11,22,33,44) and number of repeat users are 2 (11,22).
I am calculating unique users as COUNTD([customer_id]).
How can I calculate repeat users? It is basically the distinct count of the values which appear more than once. I tried with the following expression:
COUNTD(IF COUNT([customer_id]) > 1
THEN [customer_id]
END)
but I'm getting an error: Cannot mix aggregate and non-aggregate arguments comparisons or results in IF expressions
How else can I calculate the repeat users?
Thanks in advance.
Upvotes: 0
Views: 1313
Reputation: 26218
A simple alternative to Fabio's answer can also do the job. Just create a calculated field
COUNT([customer id]) >1
and add this to filter shelf.
You can filter out false candidates to remove unique users and taking returning customers only.
Upvotes: 1
Reputation: 3167
According to your filter needs, you can rely on LOD using FIXED/INCLUDE:
{ FIXED [Customer Id] : if sum({ FIXED [Customer Id] : COUNT([Customer Id])}) > 1 then 1 end }
Basically, in the inner LOD you count the occourrences, and then you just take in consideration records having 2+ (>1) of them:
Upvotes: 1