Anomaly Detection Model Nuggets

Anomaly Detection model nuggets contain all of the information captured by the Anomaly Detection model as well as information about the training data and estimation process.

When you run a stream containing an Anomaly Detection model nugget, a number of new fields are added to the stream, as determined by the selections made on the Settings tab in the model nugget. See the topic Anomaly Detection Model Settings for more information. New field names are based on the model name, prefaced by $O, as summarized in the following table.

Table 1. New field name generation
Field name	Description
$O-Anomaly	Flag field indicating whether or not the record is anomalous.
$O-AnomalyIndex	The anomaly index value for the record.
$O-PeerGroup	Specifies the peer group to which the record is assigned.
$O-Field-n	Name of the nth most anomalous field in terms of deviation from the cluster norm.
$O-FieldImpact-n	Variable deviation index for the field. This value measures the deviation from the field norm for the cluster to which the record is assigned.

Optionally, you can suppress scores for non-anomalous records to make the results easier to read. See the topic Anomaly Detection Model Settings for more information.