Anirban, if you'll take that spreadsheet and create packaged workbook getting as close as you can to a working solution, I'll be happy to help you finish it. Also, experiment with 'Only Relevant Values' in the filter drop-down menu.
Thanks Shawn and Matthew for the quick response.
I have attached 2 worksheets with 2 diff approaches.
cascading_filter_1 - with cascading filters
cascading_filter_2 - with dashboard actions.
In 1, aspect2 should be a cascading filter from aspect1. But it should show all the relevant aspects EXCEPT the one selected in 1st filter. It's like, if I select brand, the 2nd filter should show what all other aspects have a combination with brand and allow me select one of them and this cascading effect goes on.
I could fo this with parameter also, but could not find a way to make parameter dynamic from my product type filter. Because aspect names can differ based on a product type I select.
In 2, I tried to do the same with dashboard action. If u select any aspect in first section, second section narrows down according to my rule. But the problem comes with 3rd as it shows again the aspect I selected in 1st.
Any help is appreciated.
Workbooks.zip 25.6 KB
If Ville's post doesn't resolve your problem, let us know how we can help
Thanks Ville for you quick response.
As I understand your data sources, it's like a->(d,e,f) and d->(a1,a2,a3) like a hierarchy.
But my problem is, it's all in the same column and all are aspects of a product with no hierarchy.
I can derive and denormalize the structure, but the it's a huge data volume and aspects are different for different product types. Also number of aspect can vary from 3 to 13 based on product types.
If I translate your solution to my data sheet, the probably a sample data just for level_nr = 1 would be as below:
Please let me know if my understanding is correct.
If not, could you explain your solution a little bit? :-)
Just to clarify the above, in my data sheet there are 4 rows for level_nr = 1 (one for each attribute).
When I generated all combinations within level_nr = 1, then I was able to achieve the result as mentioned by Ville.
But I was looking for solution without having to do so as my data is huge and number of attributes can vary and can be entirely different depending on products.
Attaching a modified workbook.
Dashboard 2 - filters act properly as expected because of denormalization of data. All my measure will come from a diff data source based on level_nr.
Dashboard 1 - my original one with data as is, exclusion dashboard action does not work after 2nd level.
Is there any way to achieve the effect of dashboard 2 without having to create all combinations of data?
If not through dashboard, then cascading parameter also could work. But I could not find a way to create cascading parameter dynamically.
Any help is appreciated.
cascading_filter_2_new.twbx.zip 29.1 KB