Skip to main content
Announcements
Accelerate Your Success: Fuel your data and AI journey with the right services, delivered by our experts. Learn More
cancel
Showing results for 
Search instead for 
Did you mean: 
jw8s
Contributor II
Contributor II

Distinct counts with condition

Hi. This is my source data

Person Registration_date Activity
X 01.01.2023 A
X 01.01.2023 B
X 01.01.2023 B
X 01.01.2023 C
Y 01.01.2023 A
Y 01.01.2023 B
Z 01.01.2023 A
Z 01.01.2023 C
W 02.01.2023 A
W 02.01.2023 D

 

I need to create a pivot table where I have Registration_date in rows, Activity in columns and an expression that counts unique persons for every registration date in a way that it does not take into account which activity was performed (meaning that all counts in a row should be the same for every column). See below

  A B C D
01.01.2023 3 3 3 3
02.01.2023 1 1 1 1


Can you please help me with this?

Labels (5)
1 Solution

Accepted Solutions
HeshamKhja1
Partner - Creator II
Partner - Creator II

You can use a combination of TOTAL and Distinct. Check expression below:

Count(TOTAL <Registration_date> distinct Person)

The Total <d1>  tells Qlik to get total based on d1.

I think you are asking a similar question to: https://community.qlik.com/t5/QlikView-App-Dev/Ignore-Dimension-in-calculation-Pivot-table/td-p/1128...

View solution in original post

7 Replies
Vegar
MVP
MVP

Use the  measure  

=count([Person])

to calculate the no of persons.

I notice that person X performed activity B twice on the same day. If you only want to calculate it once then you can use the measure 

=count(Distinct [Person])

jw8s
Contributor II
Contributor II
Author

Unfortunately =count(Distinct [Person]) would result in the following table where there is a different value for each activity

  A B C D
01.01.2023 3 2 2 0
02.01.2023 1 0 0 1


I would like to get the same value in every column where it represents the number of persons in the respective registration date no matter which activity was performed. 

HeshamKhja1
Partner - Creator II
Partner - Creator II

@jw8s , just our of curiosity, if all columns will have the same values, why do you need multiple columns? Won't one column give the insight?

jw8s
Contributor II
Contributor II
Author

My calculation consists of multiple parts and part of it is division. I was able to successfully write the other parts but I am struggling with this.

HeshamKhja1
Partner - Creator II
Partner - Creator II

You can use a combination of TOTAL and Distinct. Check expression below:

Count(TOTAL <Registration_date> distinct Person)

The Total <d1>  tells Qlik to get total based on d1.

I think you are asking a similar question to: https://community.qlik.com/t5/QlikView-App-Dev/Ignore-Dimension-in-calculation-Pivot-table/td-p/1128...

jw8s
Contributor II
Contributor II
Author

@HeshamKhja1 lets say I had an additional flag column in my dataset. How can I add a condition {<flag={'TRUE'}>}? I tried Count(TOTAL {< flag={'TRUE'}, Registration_date >} distinct Person) but that did not do the trick

HeshamKhja1
Partner - Creator II
Partner - Creator II

Maybe like this:

Count(TOTAL <Registration_date> distinct {<flag = {'TRUE'}>} Person)