Hi - I’m in the process of trying to build a pivot table of sorts where the user can select which columns (called Measures in my app) they want and which rows (called Dimensions) they want to aggregated values for. Here it is so far:

Users can add/remove Dimensions and Measures using the configuration panel to the right of the table. Additionally, they can edit the names of measures and reorder them using drag and drop functionality. So far so good.

To get to the pivot table functionality, a user needs to create a ‘Set’ within a panel called Data Hub:

Here’s the ‘Set’ Table Type:

Note the Assigned Dimensions and Assigned Measures list fields. These are options sets - here’s the Measures one as an example:

The upload file option in the Add Set form, allows for data to be imported - here’s the ‘Data’ type:

When a Set is created, the data is imported and Measures and Dimensions are created - Here’s the ‘Measures’ type:

Onto the Pivot table… I’m basically using nested RGs. Here’s the first row:

It’s of type, ‘text’ and because I dont know of the selected first dimension, I need to have a data source for each of the 5 current dimensions.

Following on from this, for each column value, I need to have a condition for each dimension. Here’s the ‘Share’ field:

This is manageable but the next RG then has to cater for not knowing the two selected dimensions. So I have to cater for that. Here’s are the (now 20) sources:

and the Share field needs to be the same (as do all the other available measures):

This is just about manageable. But the 3rd RG (for the 3rd selected dimension) will have 64 options. I’ve started this but quickly realised this is not sustainable as there around 40 dimensions available (I can restrict to just 5 as any more than that seems too many but need to cater for all) and about the same measures.

Anyone got any (constructive) comments?!?

Thanks

Stewart