In the current code, we run measurements, then perform evaluation sequentially. But in many cases, we don’t want to wait on the evaluation before we can start processing the next sample. The codes need to be improved so both evaluation and running of the next sample can happen concurrently. This needs to be done for both the simulated and the real sequence manager.
I think Simon suggested before we could have a global shared object, so evaluation can then update this objects used in the experiment mid-experiment.
In the current code, we run measurements, then perform evaluation sequentially. But in many cases, we don’t want to wait on the evaluation before we can start processing the next sample. The codes need to be improved so both evaluation and running of the next sample can happen concurrently. This needs to be done for both the simulated and the real sequence manager.
I think Simon suggested before we could have a global shared object, so evaluation can then update this objects used in the experiment mid-experiment.