Skip to main content
Announcements
Introducing Qlik Answers: A plug-and-play, Generative AI powered RAG solution. READ ALL ABOUT IT!
cancel
Showing results for 
Search instead for 
Did you mean: 
TimoHart
Contributor III
Contributor III

"Read" and "Filter" table by other tables Column Name (Depending on cursuor position)

Hello, 

I will need to filter the Actions based on my cursor positon within the Statustable. E.g if I click a date within column TOSV34 the table below supposed to only show Action 34, in case I click within column TOSV100 respectively only action 100. 

Best Regards 

Timo 

Labels (4)
13 Replies
TimoHart
Contributor III
Contributor III
Author

Well ... sort of . The order consists out of several ACTIONS (15,30,34.. about #25+ in total) each Action has a Status and Planing Dates. The planning date is NOT the date of the status. Its a milestone-date to be reached. If the target date is overdue - the status is not showing a certain String, then this Target Plan Date( Milestone) is meant to change its color.  The Planning Dates (Milestones) are coded with the Actions number and eventually listed in a column. 

Each order number is simplified  by ist root 14 digits . So I will be able to collect all Actions for this simplified "OrderNumber" ( Orderkey) 

Each Milestone Action will have Tasks (Open Points)Collected in a seperate DB and keyed with the Orderkey and the action Number. Click on the Milestone of a Order will list All open Tasks of this Order + Action. 

If there is a open Task the Color of the Milestone Date suppose to also change. 

I guess I need to refer to your wording 2steps above :  "If it were not associated tables else they would be joined these empty cells could be avoided - but in my experiences it's further not an expedient way" 

--> I need to move forward ... and try + Error 😉

😅 Possibly I will do same experience ... 

 

 

marcus_sommer

It doesn't sound like a special difficulty else like a normal task. I could imagine just to join/map the plan-dates against current-dates and calculating there differences and further flagging/clustering/scoring them as well as the above mentioned counting/ordering/ranking them in regard to the order and order-line.

Further I think I would concatenate these plan-date to the current-data by harmonizing fields and values as much as possible and also adding a source-field with values like 'current' and 'plan'. Because like already hinted it are mainly the same data - only the direction is different. Also checking both sides against each other to flag which plan-tasks are already fulfilled and which not maybe useful to select easily completed tasks or open ones and also those which were not planned or changed/discarded ... whatever.

All this may cause a bunch of various transformation-steps but are not very complex. If these kinds of information are needed the work has to be done - in this way or another - but each other method will be harder ... 

TimoHart
Contributor III
Contributor III
Author

Ok... thank you Marcus for this possitive feedback - even I think you described the final solution already i am not able to accept. Sorry but I need to come back to the original topic - and point out this "minor" issue within the overall task.   "How to get the columns name in a MAP/Join?"

Best Regards 

Timo 

 

marcus_sommer

I suggest that you start in a parallel project with the hinted approach of loading the data just as stream and not transforming them into crosstable-logic - for both sides - the current- and plan-data by harmonizing as much as possible field-names and data-structures + running them against each other to join/map the opposite information + evaluating them and also flagging them.

It might not fulfill your requirements from the current point of view but I could imagine that it would be helpful to adjust your view in a way to find a working solution.