Anomaly Chart View
The chart view displays sensor data, anomalies, expected values, and boundaries, helping users visually identify and analyse deviations.
Last updated
The chart view displays sensor data, anomalies, expected values, and boundaries, helping users visually identify and analyse deviations.
Last updated
The Anomaly Detection Chart View in AlphaX Cloud provides a comprehensive visual representation of your anomaly detection data. Here’s a detailed description of the components and functionalities available in this view.
1. Top Right Buttons
Settings Button: Clicking this brings up the Edit Modal, allowing users to adjust the anomaly detection settings, including parameters like sensitivity, detection mode, and more.
Share Button: Opens the Share Modal, enabling users to generate a public link or share the chart data with others.
2. Interactive Chart
The chart itself is highly interactive and provides several features for users to engage with their data:
Download Data: Right-clicking on the chart allows users to download the underlying data in various formats, such as CSV or XLS. This feature is useful for deeper offline analysis or integration with other tools.
Download Chart Image: Users can also download an image of the chart in different formats, including PNG, PDF, JPG, or SVG. This is ideal for reporting or sharing visual insights with stakeholders.
3. Time Range Zoom and Shortcuts
Time Range Zoom: Located across the bottom of the chart, these toggles allow users to zoom into specific time ranges. This feature enables a closer examination of the data during critical periods, helping users identify subtle trends or anomalies.
Time Range Shortcuts: Positioned at the top right of the chart, these shortcuts let users quickly switch between predefined time ranges, such as the last 24 hours, 48 hours, 7 days, or 30 days. This provides an easy way to adjust the view and focus on relevant data periods.
The system flags anomalies based on the defined sensitivity and max anomaly ratio settings. When the confidence score deviates significantly, it indicates a potential anomaly, which is visually marked on the chart for easy identification.
This chart view not only helps in visualizing real-time and historical data but also empowers users with tools to analyze, share, and report on detected anomalies, ensuring they can maintain optimal system performance and quickly respond to any irregularities.
Data Ranges
Measured Value: The primary line on the chart represents the actual sensor data being monitored, such as temperature, air quality, or other metrics relevant to the anomaly detection.
Confidence Score: Displayed as a secondary range, the confidence score indicates the system's certainty about the normalcy of the data at each point. When there is a significant change in the confidence score, the system flags it as an anomaly.
In the Batch Analysis Mode of AlphaX Cloud, the data visualization includes several key elements to help users understand the anomaly detection process. Here’s a breakdown of the data ranges and their significance:
1. Sensor (Measured) Value
Blue Dots: The blue dots on the chart represent the sensor (or measured) values. These are the actual data points collected from the sensor over time.
2. Anomaly Indicator
Red Dots: Any data point flagged as an anomaly is marked with a red dot. This indicates that the sensor value falls outside the expected range based on the detection model's parameters, specifically sensitivity and max anomaly ratio.
3. Expected Value
Yellow Line: The expected value is represented by a yellow line on the chart. This line shows the predicted value for the sensor data at each point in time, based on the model's calculations.
4. Expected Range (Boundaries)
Grey Boundaries Box: The shaded grey area represents the expected range or boundaries within which the sensor values are considered normal. This range is calculated dynamically to account for variations in data.
The boundaries are defined as follows:
High Boundary (UpperBoundary): Calculated as the expected value plus the average of the current and the previous two upper margins. This creates an upper limit for what is considered a normal value.
Low Boundary (LowerBoundary): Calculated as the expected value minus the average of the current and the previous two lower margins. This sets the lower limit for normal values.
These boundaries are crucial in determining whether a data point is anomalous. If the sensor value falls outside these grey boundaries, it may be marked as an anomaly, depending on the sensitivity and max anomaly ratio settings configured in the detection model.
By using these data ranges, the system accurately identifies outliers and ensures that only significant deviations from expected behavior are flagged as anomalies, helping users maintain robust monitoring and control over their IoT/OT systems.