I am afraid not to a satisfactory level. @JDogg016 suggested (see Jun 5 post in thread) creating another dimension for invoice numbers, I could then sum the numbers. The problem was my counts were on order numbers,invoice numbers, Product/Part,customers. There were so many dimensions to add it made build time go from 45 mins to 120 minutes.
I also considered creating a custom SQL data set which just the counts with all other dimension attributes to link to. The problem is our users will come up with weird and wonderful analysis requirements regarding counting and I cannot factor in all of them, so I would be constantly updating this table. (i.e extra data set which has each Order,Invoice,customer,product/part dimension value and a count next to them. This can be Summed as a measure). Not to mention, as soon as you add time to the mix those counts in the data set would become problematic.
I consider this probably my #1 issue with Phocas, dynamic distinct counts would solve so many headaches on the data.
Counts do exist in Phocas if you look at chart gauges, but this is not enough as I need them in a grid.
At the moment to reduce dev time, I am telling the business that the easiest way to do this is to export a list to Excel and then do the count analysis there. This is not ideal, and really hope this option becomes available soon as the distinct counting of orders, invoices, customers is pretty basic and has been available in all other BI solutions for many years. Don’t get me wrong I think Phocas is one of the best BI platforms around, this though is a big problem for the analysis we are doing.
Sorry I couldn’t give you a solution.