We currently return model output data in lists. Returning as Numpy arrays could enhance downstream math/analysis operations and be more SciPy-congruent.
To consider:
- Zero-d method is an iterative calculation, so initializing Numpy arrays from the start without any requirement for array operations will increase runtime. Any change will need to be benchmarked to see when/if it results in an unacceptable change in runtime for the user.
- With a change in return type, this may be a breaking change.
We currently return model output data in lists. Returning as Numpy arrays could enhance downstream math/analysis operations and be more SciPy-congruent.
To consider: