actually fix numpy str type deprecation

[hellkite500]

The fix in #623 served only to mute the module attribute/reference issue of numpy.str being deprecated. The numpy.dtype.str isn't a suitable replacement of numpy.str, as the str attribute here, according to the docs, is

attribute

dtype.str
The array-protocol typestring of this data-type object.

which is not a compatible string datatype. The below error seen when using this as a dtype has been similarly reported while trying to run t-route since merging #623.

It seems that native python types are reasonably supported, though they may not be the most efficient. Using str for the astype arg should work, as seen below, this converts to a unicode string.

Successfully installed numpy-1.25.2
(venv) nelsfrazier@Lynkers-MBP t-route % python
Python 3.11.4 (main, Jul 25 2023, 17:36:13) [Clang 14.0.3 (clang-1403.0.22.14.1)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> test = np.zeros(2)
>>> test
array([0., 0.])
>>> test.dtype
dtype('float64')
>>> test.astype(np.dtype.str)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: Cannot interpret '<attribute 'str' of 'numpy.dtype' objects>' as a data type
>>> test.astype(str)
array(['0.0', '0.0'], dtype='<U32')

I believe this should close #616 with the default unicode string conversion.