Result of np.sum() fails isinstance(s, int)

[clorton]

It is reasonable to use np.sum() to determine the required capacity of a LaserFrame. However, the result of np.sum() is of type numpy.int64 which does not pass the test of isinstance(s, int).

That means the following test fails and it is a hassle to have to cast the result of np.sum() to a Python int to create a LaserFrame:

https://github.com/InstituteforDiseaseModeling/laser/blob/main/src/laser_core/laserframe.py#L59

The test should include NumPy integer types, e.g.

isinstance(capacity, [int, np.integer])

This is probably true for the test of initial_count as well.