npz `Loader` and `Saver`

[casabre]

Loading and saving from numpy npz files would be helpful.

The data should be stored in the i and q format in the numpy file, e.g.

np.savez_compressed("file.npz", i=np.real(data), q=np.imag(data))

file = np.load("file.npz")
restored_data = file["i"] + 1j * file["q"]

assert np.array_equal(data, restored_data)

Additionally, the following parameter for the Saver would be beneficial:

• store_astype which converts the data type before saving (defaults to np.float16)
• compress which would use the compressed version of npz