It would be good to start hooking these up to the smee.TensorReporter and EnsembleAverageOp ways of computing things to ensure we have everything set up sooner rather than later -- it would make it easier to consider things like the pressure in msgpack reporting from the ground up. relying on the openmm CSVs may make it more difficult to extend to what we really want, which is interaction with smee, down the line.
Originally posted by @lilyminium in #131 (comment)
Originally posted by @lilyminium in #131 (comment)