machine_id with an active baseline is provided, regimes are classified relative to the baseline. Without a baseline, z-scores are computed from the series itself and regimes are set to UNKNOWN.
Regime Types
array
required
Array of data points, each with
timestamp (ISO 8601) and value (number). Minimum 12 points, maximum 8640 points. Timestamps must be sorted ascending with no duplicates.string
default:"none"
Time interval between points. One of:
yearly, monthly, weekly, daily, hourly, minutely, secondly, microsecond, none.integer
default:95
Detection sensitivity (0-99). Higher values flag more anomalies.
number
Maximum fraction of points that can be flagged. Must be between 0 and 0.5 (exclusive).
string
Machine identifier for baseline comparison. If a baseline exists for this machine, regimes are classified using the baseline’s mean and standard deviation.
boolean
default:true
Include raw energy scores in the response.
boolean
default:false
Include 192-dimensional embedding vectors per point.
integer
Detected seasonality period (0 if none detected).
number[]
The model’s expected value for each point.
number[]
Upper bound margin for each point.
number[]
Lower bound margin for each point.
boolean[]
Whether each point is an anomaly.
boolean[]
Whether each point is a negative anomaly.
boolean[]
Whether each point is a positive anomaly.
number[]
Severity score from 0.0 to 1.0 for each point.
string[]
Per-point regime classification:
HEALTHY, ACTIVE, TRANSITION, SHOCK, or UNKNOWN.number[]
Raw prediction error (energy) per point.
number[]
Normalized energy relative to baseline (or series mean if no baseline).
number[][]
192-dimensional embedding vectors per point (null unless
include_embeddings is true).string
Model checkpoint version identifier.
string
Engine identifier (e.g.,
canary-v1).number
Inference wall time in milliseconds.