topic Re: Count in Set Analysis in App Development

Count in Set Analysis

Pericator — Mon, 27 May 2019 13:53:41 GMT

Hello,

I got a table formatted like:

Date, CustomerID, Login

2019-05-27, 1, 1

2019-05-27, 2, 1

2019-05-27, 3, 0

2019-05-26, 1, 1

So I know for every Date, for every Customer if he had a Login at that day or not. What I want to do now is to show how many distinct Customers had a Login every Day in a selected TimeRange.

I tried like this:

count(distinct {<count(distinct DATE) = count(Login)>}customerID). I figured out, why this doesnt work, by now. Set Analysis works as an imagined selection, so Qlik can't count an imagined Selection on multiple table rows. Is there any other solution to my problem?

I thought about building an extra table just for this metric. My problem is: The user can't custom select a TimeRange then, to count the customers, who logged in every day at the specific TimeRange.

Thanks for your Help in Advance

Percicator

Re: Count in Set Analysis

Ricardo_Gerhard — Mon, 27 May 2019 13:40:33 GMT

Just to understand your table structure:

Date, CustomerID, Login

2019-05-27, 1, 1

2019-05-27, 2, 1

2019-05-27, 3, 0

2019-05-27, 1, 1

Date: the first login? the last user login?

CustomerID: hope it´s the login ID of the customer.

Re: Count in Set Analysis

Pericator — Mon, 27 May 2019 13:55:00 GMT

I got a cross table of Date and CustomerID.
So I combined every Date, lets say since 2019-01-01, with every CustomerID. And the Login is a flag, 1 for had a login, and 0 for no Login at that day by that Customer.

Re: Count in Set Analysis

rwunderlich — Mon, 27 May 2019 14:24:44 GMT

I believe this will work:

=count({<CustomerID=E({<Login={0}>})>}DISTINCT CustomerID)

-Rob
http://masterssummit.com
http://qlikviewcookbook.com
http://www.easyqlik.com