custom_metric

CustomMetric

Class for interacting with custom metrics in NannyML Cloud.

create classmethod

create(name: str, description: str, problem_type: Literal['REGRESSION'], *, loss_function: TCustomMetricSource, aggregation_function: TCustomMetricSource, lower_value_limit: Optional[float] = None, upper_value_limit: Optional[float] = None) -> CustomRegressionMetricDetails

create(name: str, description: str, problem_type: Literal['BINARY_CLASSIFICATION', 'MULTICLASS_CLASSIFICATION'], *, calculation_function: TCustomMetricSource, estimation_function: Optional[TCustomMetricSource] = None, lower_value_limit: Optional[float] = None, upper_value_limit: Optional[float] = None) -> CustomClassificationMetricDetails

create(name: str, description: str, problem_type: ProblemType, calculation_function: Optional[TCustomMetricSource] = None, estimation_function: Optional[TCustomMetricSource] = None, loss_function: Optional[TCustomMetricSource] = None, aggregation_function: Optional[TCustomMetricSource] = None, lower_value_limit: Optional[float] = None, upper_value_limit: Optional[float] = None) -> TCustomMetricDetails

Create a new custom metric.

delete classmethod

delete(metric_id: str) -> TCustomMetricDetails

Delete a custom metric.

get classmethod

get(metric_id: int) -> TCustomMetricDetails

Get details of a custom metric.

Parameters:

Name	Type	Description	Default
metric_id	int	Unique identifier of the custom metric.	required

Returns:

Type	Description
TCustomMetricDetails	Details of the custom metric.

list classmethod

list(name: Optional[str] = None, problem_type: Optional[ProblemType] = None) -> List[CustomMetricSummary]

List defined custom metrics.

Parameters:

Name	Type	Description	Default
name	Optional[str]	Optional name filter.	None
problem_type	Optional[ProblemType]	Optional problem type filter.	None

Returns:

Type	Description
List[CustomMetricSummary]	List of custom metrics that match the provided filter criteria.

CustomMetricDetails

Bases: TypedDict

Details of a custom metric.

Attributes:

Name	Type	Description
description	str	User-defined description of the custom metric

CustomMetricSummary

Bases: TypedDict

Summary of a custom metric.

Attributes:

Name	Type	Description
id	int	Unique identifier of the custom metric (generated by NannyML Cloud when a custom metric is created).
name	str	User-defined name of the custom metric.
problemType	ProblemType	Type of problem the custom metric can be applied to.
createdAt	datetime	Timestamp when the model was created.