Skip to content

DecodingTreeNode

Bases: object

Source code in src\decoder\sc_decoding_tree.py 
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
class DecodingTreeNode(object):
    """"""
    def __init__(self, left_child: "DecodingTreeNode | None" = None, right_child: "DecodingTreeNode | None" = None):
        """"""
        self.left_child = left_child
        self.right_child = right_child

    def f(self, L: NDArray[np.float64]) -> NDArray[np.float64]:
        """"""
        sgn = lambda x: copysign(1, x) # signum function
        new_L = np.zeros(L.size//2, dtype=L.dtype)

        i = 0
        for beliefs in zip(L[:L.size//2], L[L.size//2:]): 
            belief = sgn(beliefs[0]) * sgn(beliefs[1]) * min(abs(beliefs[0]), abs(beliefs[1])) 
            new_L[i] = belief
            i += 1

        return new_L 

    def g(self, L: NDArray[np.float64], c: NDArray[np.uint8]) -> NDArray[np.float64]:
        """"""
        new_L = np.zeros(L.size//2, dtype=L.dtype)

        i = 0
        for beliefs in zip(L[:L.size//2], L[L.size//2:]):
            belief = beliefs[1] + beliefs[0] if c[i] == 0 else beliefs[1] - beliefs[0]
            new_L[i] = belief
            i += 1

        return new_L

__init__(left_child=None, right_child=None)

Source code in src\decoder\sc_decoding_tree.py 
 7
 8
 9
10
def __init__(self, left_child: "DecodingTreeNode | None" = None, right_child: "DecodingTreeNode | None" = None):
    """"""
    self.left_child = left_child
    self.right_child = right_child

f(L)

Source code in src\decoder\sc_decoding_tree.py 
12
13
14
15
16
17
18
19
20
21
22
23
def f(self, L: NDArray[np.float64]) -> NDArray[np.float64]:
    """"""
    sgn = lambda x: copysign(1, x) # signum function
    new_L = np.zeros(L.size//2, dtype=L.dtype)

    i = 0
    for beliefs in zip(L[:L.size//2], L[L.size//2:]): 
        belief = sgn(beliefs[0]) * sgn(beliefs[1]) * min(abs(beliefs[0]), abs(beliefs[1])) 
        new_L[i] = belief
        i += 1

    return new_L

g(L, c)

Source code in src\decoder\sc_decoding_tree.py 
25
26
27
28
29
30
31
32
33
34
35
def g(self, L: NDArray[np.float64], c: NDArray[np.uint8]) -> NDArray[np.float64]:
    """"""
    new_L = np.zeros(L.size//2, dtype=L.dtype)

    i = 0
    for beliefs in zip(L[:L.size//2], L[L.size//2:]):
        belief = beliefs[1] + beliefs[0] if c[i] == 0 else beliefs[1] - beliefs[0]
        new_L[i] = belief
        i += 1

    return new_L