I have a hierarchical dataset that is 5 levels. Essentially, levels 2 on up are totaling the child rows. The columns for each level are the same. There is a requirement, based on a user selection, to pick what high level to show. In other words, I may want to show level 5 on down or level 4 on down, etc, or just show level 1. I think I am always going to build the dataset for all 5 levels. What would be your recommendation on binding the selection to a grid? Do I create other datasets based on the top level selected that are clones of the 5 level dataset, or can you specify a subset of the 5 level dataset to bind to the grid? Should the dataset relations be recursive or parent-child? Also, since the column headers are the same for all levels, is there a way to not display the column headers on any child rows? A rough approximation of the grid would be:

                          Sales      Orders  

 Year 2013           $$$$      99999     

   Spring                $$$       9999    

     1st Quarter       $$$         999           

         Jan               $$$        999                

          Week 1        $$$        999              

          Week 2        $$$        999              

          Week 3        $$$        999       

          Week 4        $$$        999

But I may only want to display 1st Quarter on down or Jan on down etc.

Thanks,

Ron