Function _get_parameter_type in line 117 of instantiation.py throws error No AE parameter type corresponding to <class 'numpy.float64'>. if the bounds of a range variable are calculated outside Ax using some Numpy function, which will return numpy.float64 as opposed to a Python float.

If possible, given how ubiquitous Numpy is, either (1) allowing Numpy primitives or (2) suggesting casting in the error message will make it easier for users to debug their application. It may be just me (I was not a CS major), but I couldn't figure out that a simple casting would solve the problem until digging into Ax's code with the debugger.