sum if 2 conditions are met

[MZambo89]

Hi all.
I want to "sum" the field "value" or count the row id's (customers) who have logged in both in January and February for example 202401, 202402. If a customer has not logged in, I won't have any existing row. 

Hence, I want to sum those values when (period='202401' and value is 1 )and (period='202402' and value=1) then sum. Both statement needs to be there.
I got a field with the id (cust id), period (ex 202401, 202402 etc) and a field with value 1 only if exists.

The 1 means that the customer has visited the app during that particular month.

What would be the best way to tackle this? 

[Sayed_Mannan]

you can use set expression to tackle this, here is a sample based on above details:

// Define Variables for periods

SET vPeriod1 = '202401';
SET vPeriod2 = '202402';

// Create a flag for customers who logged in both periods

Sum({<cust_id = P({<period = {"$(vPeriod1)"}, value = {1}>} cust_id) * P({<period = {"$(vPeriod2)"}, value = {1}>} cust_id)>} value)

// P({<period = {"$(vPeriod1)"}, value = {1}>} cust_id) : this code will identify customers who logged in during first period.

//P({<period = {"$(vPeriod2)"}, value = {1}>} cust_id) : this code will identify customers who logged in during second period.

// * : this will ensure only customers who logged in both periods are considerd.

// Sum({<cust_id =......>} value): this will sums the values for the identified customers.

Hope this helps/

[MZambo89]

Hi Sayed. Thank you for explaining. I see the rationale, however, when I apply it will sum them up instead of counting them only 1x. 
I basically want to count the distinct id's or sum up the value 1 (each way will give the same result).

If I use this method: =Count(distinct{< period={'202401','202402'} >}single_id)
then it gives me the unique users whohave been either in one or the other but not in both... (the majority will be the same but not all of them). 

[MZambo89 · Accepted solution]

I sorted it 
count({$< cust_id=
P({1<dt ={'2024-04-01'},value={1}>})
* P({1<dt ={'2024-05-01'},value={1}>})
>}distinct cust_id)
Thank you for reminding me to use indirect set analysis! 
I got a bit rusty lately as I wasn't using Qlik too often.