csr 或者coo .todense()返回矩阵
>>> print(data)
(0, 0) -1.0
(0, 1) -0.64
(0, 2) -0.86437
(0, 3) -1.0
(0, 4) -0.37037
(0, 5) 0.155011
...
>>> m = data.todense()
>>> type(m)
<class 'numpy.matrixlib.defmatrix.matrix'>
>>> m
matrix([[-1.,-0.64,-0.86437, ... ,-0.425532,1.,-0.82064 ],
[-0.999528,-1.,-0.515396, ... ,0.106383,1.,-0.59106 ],
[-0.999529,-1.,-0.515396, ... ,0.106383,0.979475,-0.873068],
...,
[-0.998776,-1.,-0.159091, ... ,0.787234,1.,-0.784216],
[-0.997679,-1.,-0.159091, ... ,0.787234,0.982601,-0.737859],
[-0.999076,-1.,-0.159091, ... ,0.787234,1.,-0.660596]])