Skip to content

Single

dstz.evpiece.single

contour_transformation(ev)

Transforms a belief distribution into a contour function.

This transformation calculates the plausibility of each individual (singleton) element in the Frame of Discernment. The resulting distribution assigns each singleton element a mass equal to its plausibility value.

Parameters:

Name Type Description Default
ev Evidence

The evidence distribution.

 required

Returns:

Name Type Description
Evidence

A new Evidence object representing the contour function.
Source code in dstz/evpiece/single.py 
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
def contour_transformation(ev):
    """Transforms a belief distribution into a contour function.

    This transformation calculates the plausibility of each individual
    (singleton) element in the Frame of Discernment. The resulting
    distribution assigns each singleton element a mass equal to its
    plausibility value.

    Args:
        ev (Evidence): The evidence distribution.

    Returns:
        Evidence: A new `Evidence` object representing the contour function.
    """
    fod = get_fod(ev)
    res = Evidence()
    for event in fod:
        res[Element({event})] = pl(Element({event}), ev)
    return res

get_fod(ev)

Computes the Frame of Discernment (FoD) for an evidence object.

The Frame of Discernment is the set of all possible outcomes, which is constructed here by collecting all unique singleton elements from all focal sets within the given evidence.

Parameters:

Name Type Description Default
ev Evidence

The evidence object.

 required

Returns:

Name Type Description
set

A set containing all unique singleton elements in the evidence.
Source code in dstz/evpiece/single.py 
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
def get_fod(ev):
    """Computes the Frame of Discernment (FoD) for an evidence object.

    The Frame of Discernment is the set of all possible outcomes, which is
    constructed here by collecting all unique singleton elements from all
    focal sets within the given evidence.

    Args:
        ev (Evidence): The evidence object.

    Returns:
        set: A set containing all unique singleton elements in the evidence.
    """
    res = set()
    for ele in ev.keys():
        for item in ele.value:
            res.add(item)
    return res

pignistic_probability_transformation(ev)

Transforms a belief distribution into a pignistic probability distribution.

This transformation, often called BetP, converts a belief mass assignment into a classical probability distribution. It does this by distributing the mass of each focal element equally among all the individual (singleton) elements it contains.

Parameters:

Name Type Description Default
ev Evidence

An evidence distribution (mass function).

 required

Returns:

Name Type Description
Evidence

A new Evidence object representing the transformed probability distribution.
Source code in dstz/evpiece/single.py 
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
def pignistic_probability_transformation(ev):
    """Transforms a belief distribution into a pignistic probability distribution.

    This transformation, often called BetP, converts a belief mass assignment
    into a classical probability distribution. It does this by distributing the
    mass of each focal element equally among all the individual (singleton)
    elements it contains.

    Args:
        ev (Evidence): An evidence distribution (mass function).

    Returns:
        Evidence: A new `Evidence` object representing the transformed
                  probability distribution.
    """
    res = Evidence()
    for key in ev:
        for simple in key.value:
            s_item = Element(simple)
            if s_item in res:
                res[s_item] += ev[key] / len(key.value)
            else:
                res[s_item] = ev[key] / len(key.value)
    return res

shafer_discounting(ev, alpha, fod=None)

Applies Shafer's discounting to an evidence object.

Shafer discounting models partial reliability of an information source by transferring a portion of belief mass from each focal element to the entire Frame of Discernment (representing uncertainty).

Where alpha represents the degree of trust (1 = fully trusted, 0 = not trusted at all).

Parameters:

Name Type Description Default
ev Evidence

The evidence object to be discounted.

 required
alpha float

The discount factor, a value between 0 and 1 representing the degree of trust in the evidence source.

 required
fod set

Precomputed frame of discernment. If None, will be computed.

 None

Returns:

Name Type Description
Evidence

A new, discounted Evidence object where belief masses have been adjusted according to the discounting factor.
Source code in dstz/evpiece/single.py 
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
def shafer_discounting(ev, alpha, fod=None):
    """Applies Shafer's discounting to an evidence object.

    Shafer discounting models partial reliability of an information source by
    transferring a portion of belief mass from each focal element to the
    entire Frame of Discernment (representing uncertainty).

    Where alpha represents the degree of trust (1 = fully trusted, 0 = not trusted at all).

    Args:
        ev (Evidence): The evidence object to be discounted.
        alpha (float): The discount factor, a value between 0 and 1 representing
                       the degree of trust in the evidence source.
        fod (set, optional): Precomputed frame of discernment. If None, will be computed.

    Returns:
        Evidence: A new, discounted `Evidence` object where belief masses have
                  been adjusted according to the discounting factor.
    """

    res = Evidence()
    # Get Frame of Discernment
    if not fod:
        fod = get_fod(ev)
    fod_element = Element(fod)

    # Apply discounting to each focal element: multiply by trust factor alpha
    for key, value in ev.items():
        res[key] = alpha * value

    # Transfer the untrusted portion (1-alpha) to the universal set
    if fod_element in res:
        res[fod_element] += (1 - alpha)
    else:
        res[fod_element] = (1 - alpha)

    return res

temperature_transformation(ev, t)

Apply temperature-based transformation to evidence values.

This function performs a temperature scaling operation on evidence values.

Parameters:

Name Type Description Default
ev Evidence

Input evidence object containing key-value pairs where values represent belief masses for corresponding keys.

 required
t float

Temperature parameter.

 required

Returns:

Name Type Description
Evidence

A new Evidence object with temperature-transformed and normalized values.
Source code in dstz/evpiece/single.py 
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
def temperature_transformation(ev, t):
    """
    Apply temperature-based transformation to evidence values.

    This function performs a temperature scaling operation on evidence values.

    Args:
        ev (Evidence): Input evidence object containing key-value pairs where
                      values represent belief masses for corresponding keys.
        t (float): Temperature parameter.

    Returns:
        Evidence: A new Evidence object with temperature-transformed and
                 normalized values.
    """
    res = Evidence()
    for key, value in ev.items():
        # Apply temperature scaling: scale each value by its normalized position
        # The denominator (2^len(key) - 1) represents the maximum possible value for this key
        res[key] = math.pow(value / (2 ** len(key) - 1), t) * (2 ** len(key) - 1)

    # Normalize the transformed values to ensure they sum to 1.0
    norm_factor = sum(res.values())
    for key, value in res.items():
        res[key] = value / norm_factor

    return res