Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 
Tee_dubs
Contributor III
Contributor III

Clicking on a chart to show a related dimension

I have an odd question, I think I am starting to find the limitations of Qliksence but I will ask anyway.

I have created a scatter plot which shows me exactly what I want to see. "Events" in the lifecycle of our product.

What I want is when I select an "event", that it will select the "event category" to show all instances of the event rather than a single event itself.

The work around I am using is to make the selection on the scatter plot, then select the white (or possible) values in an "event Category" filter pane, then deselect the "event". This works fine for me as a developer but the usability for end users is poor and confusing.

Is there any way to do this?
Labels (1)
2 Solutions

Accepted Solutions
rwunderlich
Partner Ambassador/MVP
Partner Ambassador/MVP

You could concatenate EventId & Category as your dimension and then ignore the EventId in your chart dimension and measures.  Attached is an example that demonstrates this.  For a more robust example see https://qlikviewcookbook.com/2018/02/qdg-guru-day-inspiration/

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

View solution in original post

rwunderlich
Partner Ambassador/MVP
Partner Ambassador/MVP

The example helps. Seems like all you are missing is how to get the set expression into your x-axis expression.  Use only() like this:

RangeMin(12/(650/ only({<TASK_COUNT=>}ACCOMP_FC)), 12/(5000 / only({<TASK_COUNT=>}ACCOMP_FH)),only({<TASK_COUNT=>}ACCOMP_MO))

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

View solution in original post

6 Replies
rwunderlich
Partner Ambassador/MVP
Partner Ambassador/MVP

You could concatenate EventId & Category as your dimension and then ignore the EventId in your chart dimension and measures.  Attached is an example that demonstrates this.  For a more robust example see https://qlikviewcookbook.com/2018/02/qdg-guru-day-inspiration/

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

Tee_dubs
Contributor III
Contributor III
Author

This is coooool!

concatenating the dimension works well. At least it selects the "event category" automatically, this means it is a single click to get what I want .

The measures involved, however, are quite complicated to wrap in the "if" syntax.

here is the example:

measure 1:

RangeMin(12/(((median(COUNT_DISTINCT_of_flight_id)) / ACCOMP_FC)), 12/(((median(SUM_of_Flight_Hours)) / ACCOMP_FH)),ACCOMP_MO) -

I cant quite get the syntax right to add the set analysis...

 

Measure 2 :

((Sum(Aggr(Sum(TOTAL <TASKNUMBER> TASKMH), EventID, EventCat)) +
Sum(Aggr(Sum(TOTAL <TASKNUMBER> ACCESSMH), EventID, EventCat)) +
Sum(Aggr(Sum(TOTAL <TASKNUMBER> PREPMH), EventID, EventCat)))
* 75 * 2)/24

rwunderlich
Partner Ambassador/MVP
Partner Ambassador/MVP

I would think that you could just add {<EventId=>} into your aggregation functions.  But it also could be that your starting premise is more complex than it needs to be. Any chance you could upload a sample QVF?  You can put some dummy data in or drop the sensitive fields from the model.

-Rob

Tee_dubs
Contributor III
Contributor III
Author

 
Tee_dubs
Contributor III
Contributor III
Author

There are over 17000 events. They reoccur on a schedule. It looks daunting but using other filters it makes sense. I have removed other identifying details.

rwunderlich
Partner Ambassador/MVP
Partner Ambassador/MVP

The example helps. Seems like all you are missing is how to get the set expression into your x-axis expression.  Use only() like this:

RangeMin(12/(650/ only({<TASK_COUNT=>}ACCOMP_FC)), 12/(5000 / only({<TASK_COUNT=>}ACCOMP_FH)),only({<TASK_COUNT=>}ACCOMP_MO))

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