Your best bet would probably be to restructure your data, so that all the different donut types were in one column and each of their $ and Qtys were in two separate columns. The following KB article provides further information on how to arrange your data:
There is also a reshaper tool that can be found here:
Hope this helps!
Hello Tracy, thanks for your reply. Our data is in such a "multiple response" format so that we may use any donut type or variant or both as a filter for cross tabulations. are there any other ways of achieving this objective?