numpy overrides cg.Point dtype?

[jGaboardi]

It appears that numpy is overriding libpysal.cg.Point objects and converting directly into arrays when storing the points in a numpy.array. I came across this issue (pysal/spaghetti#378) when implementing improved support for libpysal.cg geometries. Has anyone else seen this?

In [1]: %load_ext watermark                                                                                                     

In [2]: %watermark                                                                                                              
2020-01-12T15:31:01-05:00

CPython 3.7.3
IPython 7.10.2

compiler   : Clang 9.0.0 (tags/RELEASE_900/final)
system     : Darwin
release    : 19.2.0
machine    : x86_64
processor  : i386
CPU cores  : 4
interpreter: 64bit

In [3]: import libpysal                                                                                                         

In [4]: import numpy                                                                                                            

In [5]: import scipy                                                                                                            

In [6]: %watermark -iv                                                                                                          
numpy    1.17.3
scipy    1.3.3
libpysal 4.2.0

In [7]: points_in_list = [libpysal.cg.Point((1,1)), libpysal.cg.Point((2,2))]                                                   

In [8]: print(type(points_in_list), type(points_in_list[0]))                                                                    
<class 'list'> <class 'libpysal.cg.shapes.Point'>

In [9]: for dtype in (tuple, numpy.array):
    ...:     pts_containers = dtype(points_in_list)
    ...:     print(type(pts_containers), type(pts_containers[0]), type(pts_containers[0][0]))                                                                                                                   
<class 'tuple'> <class 'libpysal.cg.shapes.Point'> <class 'float'>
<class 'numpy.ndarray'> <class 'numpy.ndarray'> <class 'numpy.float64'>