python-multi-array exports std::shared_ptr<boost::multi_array<T, N>> for

T: bool, uint8, uint16, uint32, uint64, int8, int16, int32, int64, float32 and floa64
N: 1 to 8.

The library serves a powerful tool to cooperate python and its extension modules written in C++. You can create arrays and set values in python, make heavy calculation in C++ (OpenMP and every other C++ tools are available,) and get back to python and display the results using matplotlib.

The array is allocated via multi_array.make.

>>> import numpy, multi_array
>>> multi_array.make((4, 2), numpy.float32)
<multi_array.shared_float_matrix object at 0x10aeb3f50>)

The array itself has simple I/O APIs:

>>> x = multi_array.make((4, 2), numpy.float32)
>>> x.num_dimensions()
2
>>> x.num_elements()
8
>>> x.shape()
(4, 2)
>>> x[1, 1] = 3
>>> x[1, 1]
3.0

and conversion APIs with numpy.ndarray.

>>> x.set(scipy.rand(4, 2))
>>> type(x.get())
<type 'numpy.ndarray'>
>>> x.get()
array([[ 0.91382688, 0.374331 ],
 [ 0.43389955, 0.5571261 ],
 [ 0.6937117 , 0.40599877],
    [ 0.80906659, 0.75029951]], dtype=float32)

$ make

will build multi_array.so in the same directory.

Full documentation is available on GitHub Wiki.

• NumPy
• boost 1.63.0+ ... v1.63.0 or later is required to use Boost.Python NumPy Extension.

If you have any question about the library, feel free to contact me: https://github.com/yokaze