We should write out volume averages of various quantities at regular intervals. This is obviously useful for checking the growth rate of linear instabilities. But perhaps more importantly, it gives us a way to easily monitor energy and momentum conservation.
How to implement this in a flexible way (it should be easy to add new diagnostics) will require some thought. In addition, we need to think about which data format would be most suitable for this. I would be curious to see what an HDF5 implementation would look like.
We should write out volume averages of various quantities at regular intervals. This is obviously useful for checking the growth rate of linear instabilities. But perhaps more importantly, it gives us a way to easily monitor energy and momentum conservation.
How to implement this in a flexible way (it should be easy to add new diagnostics) will require some thought. In addition, we need to think about which data format would be most suitable for this. I would be curious to see what an HDF5 implementation would look like.