Hi community,
I'll try to outline what I'd like to do, give you the two ideas I had and would appreciate your feedback on those and also other solutions, I didn't think about.
Let's say I have data about a big farm that I want to display in a nice table. I want the user to be able to choose what the data should be segmented by and also what kind of analysis should be done on the data. Here an example of a workflow.
- The user chooses the segmentation "building" and the analysis "count" . Then the table should contain the total count of animals as a column per building (stable 1, stable 2, barn, farmouse, ...) in the rows.
- Now if the user changes the analysis to "size", the row segmentation (i.e. by building) should stay the same, but I want a column for the median size and one for the mean size of the animals instead of showing the total count.
- If now the user changes the segmentation from "building" to "species", then the columns should stay the same but now I get the analysis per species ("pigs", "horses", "chicken", "dogs", ...)
Obviously that can be done by just creating a myriad of tabs, each featuring exactly the wanted table, but that's not pretty and a poor user experience. So here my ideas to do it more elegantly:
- Build the myriad of tables (one for each cross product of segmentation and analysis) on one page, above each other and then link a selection to bookmarks of the respective views. Probably a nice user experience, but doesn't feel very "clean".
- Use the matrix visual and put field parameters in rows and columns that need to be linked to the correct measures in an elaborate way to ensure the functionality described above. Feels cleaner but also a lot of effort and might require defining a number of conditional filtering and auxilliary tables and the like
Any comments are much appreciated!
