LabelPropertyOutliers#

class LabelPropertyOutliers[source]#

Find outliers labels with respect to the given properties.

The check computes several label properties and then computes the number of outliers for each property. The check uses IQR to detect outliers out of the single dimension properties.

Parameters

label_propertiesList[Dict[str, Any]], default: None

List of properties. Replaces the default deepchecks properties. Each property is a dictionary with keys 'name' (str), method (Callable) and 'output_type' (str), representing attributes of said method. ‘output_type’ must be one of:

'numeric' - for continuous ordinal outputs.
'categorical' - for discrete, non-ordinal outputs. These can still be numbers, but these numbers do not have inherent value.
'class_id' - for properties that return the class_id. This is used because these properties are later matched with the VisionData.label_map, if one was given.

For more on image / label properties, see the guide about Data Properties.

n_show_topint , default: 5

number of outliers to show from each direction (upper limit and bottom limit)

iqr_percentiles: Tuple[int, int], default: (25, 75)

Two percentiles which define the IQR range

iqr_scale: float, default: 1.5

The scale to multiply the IQR range for the outliers detection

__init__(label_properties: Optional[List[Dict[str, Any]]] = None, n_show_top: int = 5, iqr_percentiles: Tuple[int, int] = (25, 75), iqr_scale: float = 1.5, **kwargs)[source]#

__new__(*args, **kwargs)#

Methods

`LabelPropertyOutliers.add_condition`(name, ...)	Add new condition function to the check.
`LabelPropertyOutliers.clean_conditions`()	Remove all conditions from this check instance.
`LabelPropertyOutliers.compute`(context, ...)	Compute final result.
`LabelPropertyOutliers.conditions_decision`(result)	Run conditions on given result.
`LabelPropertyOutliers.config`([...])	Return check configuration (conditions' configuration not yet supported).
`LabelPropertyOutliers.draw_image`(data, ...)	Return an image to show as output of the display.
`LabelPropertyOutliers.from_config`(conf[, ...])	Return check object from a CheckConfig object.
`LabelPropertyOutliers.from_json`(conf[, ...])	Deserialize check instance from JSON string.
`LabelPropertyOutliers.get_default_properties`(data)	Return default properties to run in the check.
`LabelPropertyOutliers.initialize_run`(...)	Initialize the properties state.
`LabelPropertyOutliers.metadata`([with_doc_link])	Return check metadata.
`LabelPropertyOutliers.name`()	Name of class in split camel case.
`LabelPropertyOutliers.params`([show_defaults])	Return parameters to show when printing the check.
`LabelPropertyOutliers.remove_condition`(index)	Remove given condition by index.
`LabelPropertyOutliers.run`(dataset[, model, ...])	Run check.
`LabelPropertyOutliers.to_json`([indent, ...])	Serialize check instance to JSON string.
`LabelPropertyOutliers.update`(context, batch, ...)	Aggregate image properties from batch.

Examples#

ImagePropertyOutliers.update

LabelPropertyOutliers.add_condition