DetectAnomalies

DetectAnomalies(node, dciId, startTime, endTime, threshold) ⇒ ScoredDciValue

Detect anomalies for DCI within given period

Table 1. Parameters
node	Node	Node object
dciId	Integer	DCI id
startTime	Integer	Period start time
endTime	Integer	End time is optional (default is current time)
threshold	Number	Threshold is optional (default is 0.75)

Return

Returns null on error and an array of ScoredDciValue on success. Each element contains the original DCI value, its timestamp, and an anomaly score between 0.0 (normal) and 1.0 (highly anomalous). Values with a score above the threshold parameter are considered anomalies.

The threshold parameter (default 0.75) controls sensitivity: lower values detect more anomalies (higher sensitivity), higher values detect only the most significant deviations.

// Detect anomalies in last 24 hours with default threshold (0.75)
anomalies = DetectAnomalies($node, dciId, time() - 86400);

// Use lower threshold for higher sensitivity
anomalies = DetectAnomalies($node, dciId, time() - 86400, time(), 0.50);
if (anomalies != null)
{
    for (v : anomalies)
        println("Score: " .. v.score .. " Value: " .. v.value .. " Time: " .. v.timestamp);
}