What is the proper design pattern for shiny modules with multiple outputs? So far, I went with 1 module = 1 output strategy. But it seems kinda wasteful to duplicate collecting from sql db when the second output is a derivative of the first one (dataset grouped and summarized on a higher level). Another option would be to create second module which takes the output of the first one and creates the derivative. This on the other hand (somewhat unnecesarily?) increases number of modules in app.
What would be the correct (intended by shiny developers) strategy here?