Clean up emp::DataFile

[mercere99]

Currently DataFile will collect data from DataNode objects and print it as columns in a file.

Several improvements are needed:

• [x] Collection should be doable from any function (and other types should be recognized)
• [x] Conditions should be added, for example to make it easy to print every so many updates. This goal can be accomplished by running a function that returns a bool; printing occurs only if it's true.
• [ ] Output formats other than CSV should be easy to produce. HTML tables being one.
• [ ] Files should be able to be stored in memory and not always automatically printed unless flushed. Ideally we should have a DataTable object that a DataFile can optionally work with.

[emilydolson]

What is the thinking behind the timing functions going in World_file rather than in DataFile? On the one hand I see the point that not all DataFiles need to worry about timing. But on the other hand, I think that there are a lot of situations outside of evolution where timing would be relevant (e.g. physics). Also, having the timing functions be in their own class makes inheritance messy if you want to write a class that inherits from DataFile but has timing functions.

The reason I ask is that I'm currently putting together a function to output the data necessary to make a Muller plot. This requires printing out multiple rows per update (one per active taxon). While this situation seems common enough that it's useful to include in the library, it seems rare enough that it shouldn't just be part of DataFile. So inheriting from DataFile seems like the right way to do it. But then if it inherits from World_file too, that's the dreaded multiple inheritance diamond and I suspect we don't want to do that.