Create instance for arrays

[dineshbvadhia]

I've got pond working but it is not recycling correctly plus I'm not convinced the instance creation is optimal for my use case. The pool is for 3 numpy arrays:

np.array([1, 2])
np.array([5, 7])
np.array([3, 9]

A dictionary is created to hold the arrays:

def createInstance(self) -> PooledObject:
	d = {0:np.array([1, 2]), 1:np.array([5, 7]), 2:np.array([3, 9]}
	return PooledObject(d)

To access a relevant array, index into the pooling_object (the dictionary). A taken array must not be recycled because it changes and instead a new array should be created.

The test results show the indexed array is being recycled instead of creating a new one. Both pooled_maxsize=3 and pooled_maxsize=1 return the same results.

Does this dictionary method recreate the taken array or the whole dictionary of arrays?

Is this the best way to create an instance for the 3 numpy arrays?