A cloude decomposition implementation

Download package and execute either

python setup.py install

to install the package or

python setup.py develop

to install it in development mode.

from cloude import cloude_decomp
import numpy as np
mueller_matrix = np.array([[1, 0.0707, 0.0348, -0.0060],
 [0.0480, 0.4099, 0.0077, 0.0650],
 [0.0162, -0.0184, 0.2243, -0.3580],
 [0.0021, -0.0465, 0.3571, 0.1783]])
decomp_matrix = cloude_decomp(mueller_matrix)

The matrices are scaled by their respective eigenvalues. If you want to use them to fit to an experiment, normalize them by the M_00 matrix element (decomp_matrix[:, 0, 0]). To get the corresponding Jones matrix estimate you may use

from cloude import cloude_decomp_jones
decomp_matrix = cloude_decomp_jones(mueller_matrix)

Parameter usage is the same for cloude_decomp and cloude_decomp_jones functions. For an output of the eigenvalues use the output_eigenvalue parameter

decomp_matrix, ev = cloude_decomp(mueller_matrix, output_eigenvalue=True)

By default the matrix with the highest eigenvalue is returned, which is equivalent to

decomp_matrix = cloude_decomp(mueller_matrix, ev_mask=[1, 0, 0, 0])

You may tune the output as an arbitrary sum of decompositions by the eigenvalue mask

decomp_matrix = cloude_decomp(mueller_matrix, ev_mask=[1, 1, 1, 1])

returns the full sum of all coherent matrices, hence decomp_matrix == mueller_matrix.