kd.metrics.RocAuc#

class kauldron.metrics.RocAuc( *, logits: typing.Annotated[typing.Any, <object object at 0x7824c478ba80>] = '__KEY_REQUIRED__', labels: typing.Annotated[typing.Any, <object object at 0x7824c478ba80>] = '__KEY_REQUIRED__', mask: typing.Annotated[typing.Any, <object object at 0x7824c478ba80>] | None = None, unique_labels: typing.List[int] | None = None, multi_class_mode: str = 'ovr', )[source]

Bases: kauldron.metrics.base.Metric

Area Under the Receiver Operating Characteristic Curve (ROC AUC).

logits

The logits to evaluate.

Type:: Any

labels

The groundtruth labels.

Type:: Any

mask

Sample weights.

Type:: Any | None

unique_labels

If we are testing on a small subset of data and by chance it does not contain all classes, we need to provide the groundtruth labels separately. In case None, unique_labels will be determined from the labels.

Type:: List[int] | None

multi_class_mode

One-vs-Rest (“ovr”) or One-vs-One (“ovo”)

Type:: str

logits: Annotated[Any, <object object at 0x7824c478ba80>] = '__KEY_REQUIRED__'

labels: Annotated[Any, <object object at 0x7824c478ba80>] = '__KEY_REQUIRED__'

mask: Annotated[Any, <object object at 0x7824c478ba80>] | None = None

unique_labels: List[int] | None = None

multi_class_mode: str = 'ovr'

class State( labels: Int['*b 1'], probs: Float['*b n'], mask: Bool['*b 1'] | Float['*b 1'], *, parent: _MetricT = _EMPTY_TYPE.EMPTY, )[source]

Bases: kauldron.metrics.base_state.CollectingState[RocAuc]

RocAuc state.

labels: Int['*b 1']

probs: Float['*b n']

mask: Bool['*b 1'] | Float['*b 1']

compute() → float[source]: Returns the concatenated values.

merge( other: kauldron.metrics.base_state._SelfT, ) → kauldron.metrics.base_state._SelfT[source]

Returns a new state that is the accumulation of self and other.

Parameters:: other – A State whose intermediate values should be accumulated onto the values of self.
Returns:: A new State that accumulates the value from both self and other.

replace(**updates): Returns a new object replacing the specified fields with new values.

get_state( logits: jaxtyping.Float[Array, '*b n'] | jaxtyping.Float[ndarray, '*b n'], labels: jaxtyping.Int[Array, '*b 1'] | jaxtyping.Int[ndarray, '*b 1'], mask: jaxtyping.Bool[Array, '*b 1'] | jaxtyping.Bool[ndarray, '*b 1'] | jaxtyping.Float[Array, '*b 1'] | jaxtyping.Float[ndarray, '*b 1'] | None = None, ) → kauldron.metrics.classification.RocAuc.State[source]

empty() → kauldron.metrics.base.Metric.State[source]