I have an interesting problem that has come up. I am dealing with several large blended datasets, one of which contains an entry for each index, and the others with different subsets thereof.
My issue is that to properly filter based on those subsets, it seems I need to be at the level of detail of each index in the primary dataset, but I cannot do that without losing the ability to build some of the charts I need (such as a bubble chart, which if given the detail of each index, simply generates one bubble for each index, defeating the purpose.
No, what I'm attempting to do, and getting tantalizingly close to succeeding at, is to filter from the secondary dataset (using parameters to select what to filter upon), but I keep getting stuck in trying to do so. The closest I have been able to get is this:
But because the Primary Filter is a calculated field using elements from a secondary data source, I can't simply set it as a filter for "True" due to the fact that it has an aggregation from the secondary data source.
Some other notes that will hopefully help solve this:
- The correct Number of Records for the different parameter settings are supposed to be: All - 10, No - 3, Yes - 2. The biggest challenge is to make the Yes show two, rather than three, which it will do if you aren't careful and use "Letter" as a level of detail.
- Remember that F1 cannot be used as a level of detail. If that were the case, it would be a simple problem, but several elements of the dashboard need to be bubble charts, which fail when doing so.
- Joining the data would prove difficult due to its size and the fact that it exists as published datasources on our server which cannot currently be joined. But if this does turn out to be the only solution, then that's what must be done.
- I'm using 2018.1 right now, but if a solution exists via Extensions in 2018.2, that's definitely something that can be done.
Thank you all for your help, and if you need any clarification, just let me know.
FilterTest.twbx 13.0 KB