Skip to content

clsmetric

clsmetric

Module for clsmetric

Classes

ClsMetric 

ClsMetric(
    source_col,
    target_cols_prefix,
    function="accuracy",
    name_suffix="",
    use_multitarget=False,
    use_probabilities=False,
)

Bases: DfMetric

Let's use calculate arbitrary classification metrics as part of the DataframeAnalyzer(ResultAnalyzer).

Parameters

source_col: str name of the column with the class index or indexes (multitarget) target_cols_prefix: str starting name of the columns containing the output of the model function: str or dict, default "accuracy" function to calculate the metric. If str is given it must be a key of the cl_metric_dict. Otherwise the given dict is initialized with init_object and additional the key name must be available to define the output name in the result dict. name_suffix: str, default "" can be used to add a str to the dict output name (e.g. if the same metric should be applied to multiple columns) use_multitarget: bool, default false allows to be more than one class present per sample use_probabilities: bool, default false if this is set true then the values are not binarized before calculating the metric function

Source code in niceml/mlcomponents/resultanalyzers/dataframes/clsmetric.py 
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
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
def __init__(
    self,
    source_col: str,
    target_cols_prefix: str,
    function: Union[str, dict] = "accuracy",
    name_suffix: str = "",
    use_multitarget: bool = False,
    use_probabilities: bool = False,
):
    """
    Let's use calculate arbitrary classification
    metrics as part of the DataframeAnalyzer(ResultAnalyzer).

    Parameters
    ----------
    source_col: str
        name of the column with the class
        index or indexes (multitarget)
    target_cols_prefix: str
        starting name of the columns
        containing the output of the model
    function: str or dict, default "accuracy"
        function to calculate the metric.
        If str is given it must be a key of the `cl_metric_dict`.
        Otherwise the given dict is initialized with
        `init_object` and additional the key `name`
        must be available to define the
        output name in the result dict.
    name_suffix: str, default ""
        can be used to add a str to the dict
        output name (e.g. if the same metric should be applied
        to multiple columns)
    use_multitarget: bool, default false
        allows to be more than one class present per sample
    use_probabilities: bool, default false
        if this is set true then the values are
        not binarized before calculating the metric function
    """
    self.source_col = source_col
    self.target_cols_prefix = target_cols_prefix
    self.multi_target = use_multitarget
    self.name_suffix = name_suffix
    self.use_probabilities = use_probabilities
    if type(function) is str:
        self.func_name = function
        try:
            self.function = cl_metric_dict[function]
        except KeyError as e:
            raise Exception(
                f"Function with name {function} not "
                f"supported! Available: {list(cl_metric_dict.keys())}"
            ) from e
    else:
        self.func_name = function["name"]
        self.function = function["target"]
Functions