F1ScoreMacro

Bases: Metric

Source code in metrics_toolbox/metrics/classification/f1_score_macro.py 
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
class F1ScoreMacro(Metric):
    _name = MetricNameEnum.F1_SCORE
    _type = MetricTypeEnum.LABELS
    _scope = MetricScopeEnum.MACRO

    def __init__(self):
        """Initialize F1 score metric for classification."""

    def compute(
        self, y_true: np.ndarray, y_pred: np.ndarray, column_names: list[str] = None
    ) -> MetricResult:
        """Compute F1 score for label classification.

        Parameters
        ----------
        y_true : array-like of shape (n_samples, n_classes)
            True binary labels in one-hot encoded format.
        y_pred : array-like of shape (n_samples, n_classes)
            Predicted binary labels in one-hot encoded format.
        column_names : list[str], optional
            Class names corresponding to column indices.

        Returns
        -------
        MetricResult
            The computed F1 score metric result.
        """

        value = 0.0
        for i in range(len(column_names)):
            tp_c = sum((y_pred[:, i] == 1) & (y_true[:, i] == 1))
            fn_c = sum((y_pred[:, i] == 0) & (y_true[:, i] == 1))
            fp_c = sum((y_pred[:, i] == 1) & (y_true[:, i] == 0))
            f1_c = (
                2 * tp_c / (2 * tp_c + fn_c + fp_c)
                if (2 * tp_c + fn_c + fp_c) > 0
                else 0.0
            )
            value += f1_c
        value /= len(column_names)

        return MetricResult(
            name=self.name,
            scope=self.scope,
            type=self.type,
            value=value,
        )

__init__()

Initialize F1 score metric for classification.

Source code in metrics_toolbox/metrics/classification/f1_score_macro.py 
17
18
def __init__(self):
    """Initialize F1 score metric for classification."""

compute(y_true, y_pred, column_names=None)

Compute F1 score for label classification.

Parameters:
  • y_true (array-like of shape (n_samples, n_classes)) –

    True binary labels in one-hot encoded format.

  • y_pred (array-like of shape (n_samples, n_classes)) –

    Predicted binary labels in one-hot encoded format.

  • column_names (list[str], default: None ) –

    Class names corresponding to column indices.
Returns:
Source code in metrics_toolbox/metrics/classification/f1_score_macro.py 
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
def compute(
    self, y_true: np.ndarray, y_pred: np.ndarray, column_names: list[str] = None
) -> MetricResult:
    """Compute F1 score for label classification.

    Parameters
    ----------
    y_true : array-like of shape (n_samples, n_classes)
        True binary labels in one-hot encoded format.
    y_pred : array-like of shape (n_samples, n_classes)
        Predicted binary labels in one-hot encoded format.
    column_names : list[str], optional
        Class names corresponding to column indices.

    Returns
    -------
    MetricResult
        The computed F1 score metric result.
    """

    value = 0.0
    for i in range(len(column_names)):
        tp_c = sum((y_pred[:, i] == 1) & (y_true[:, i] == 1))
        fn_c = sum((y_pred[:, i] == 0) & (y_true[:, i] == 1))
        fp_c = sum((y_pred[:, i] == 1) & (y_true[:, i] == 0))
        f1_c = (
            2 * tp_c / (2 * tp_c + fn_c + fp_c)
            if (2 * tp_c + fn_c + fp_c) > 0
            else 0.0
        )
        value += f1_c
    value /= len(column_names)

    return MetricResult(
        name=self.name,
        scope=self.scope,
        type=self.type,
        value=value,
    )