Aggregations

Aggregations are mathematical operations that summarize numeric values from your monitoring data. They transform raw events or records into meaningful metrics by computing values such as counts, averages, sums, and measures of data variability.

When you configure a chart, you select one or more aggregation functions to apply to specific numeric fields. Each aggregation produces a single summary value for the entire dataset, or one value per group when a Group by field is added.

Supported aggregations

The following aggregation types are available when you configure a chart.

Table 1. Aggregation types
Aggregation Description
Count Returns the total number of events or records.
Unique count Returns the number of distinct values in the selected field.
Sum Returns the total sum of numeric field values.
Average Returns the arithmetic mean of numeric values.
Min Returns the lowest numeric value in the selected field.
Max Returns the highest numeric value in the selected field.
Variance Measures how far numeric values deviate from their mean; higher values indicate greater variability.
Standard deviation Returns the square root of the variance and indicates how widely values vary around the average.
Latest Returns the most recent numeric value based on timestamp ordering. Requires a timestamp field.

Aggregations apply only to numeric fields. Null or missing numeric values are ignored. When a Group by field is used, each aggregation is calculated independently for each group.

Applying aggregations in a chart

To configure aggregations when you create or edit a chart:

  1. Open or create a chart.
  2. Select a monitoring source and choose a numeric field to analyze.
  3. Select an aggregation function such as Count, Average, Variance, or Standard deviation.
  4. (Optional) Add a Group by field to compute separate values for each category.
  5. Preview the result and save the chart.

You can add multiple aggregations to a chart. Each aggregation is applied independently to its selected field. For details about measurement types and visualization behavior, see Measurements.

Using Standard deviation and Variance

Standard deviation and variance help you understand how consistent or variable numeric values are in your monitoring data. Use these aggregations to highlight irregular patterns, changes in behavior, or differences across categories when grouped.

  • Variance measures how far values deviate from the average.
  • Standard deviation is the square root of variance and indicates the typical distance between each value and the average.

These aggregations are available for numeric fields only, ignore null or missing values, and are computed per group when a Group by field is applied.

Suppose that your monitoring source includes a numeric field such as rainForecast and a category field such as temperature. You want to see how much the forecast varies across temperature conditions.

Configure the chart as follows:

  • Aggregation: Standard deviation
  • Data item: rainForecast
  • Group by: temperature

The standard deviation for each temperature group is calculated separately. For example:

  • In the cold group, values such as 1 and 90 produce a higher standard deviation because the values differ significantly.
  • In the warm group, values such as 1 and 1 produce 0 because all values are the same.
  • In the hot group, a single value such as 1 also results in 0 because there is no variation.

Variance behaves the same way but reports the value in squared units. These aggregations help you spot where values remain stable and where they fluctuate across different groups of your data.

Using the Latest aggregation

The Latest aggregation returns the most recent numeric value based on the selected timestamp field.

  • In Metric charts, the default timestamp is used automatically.
  • In Periodic Metric and Periodic KPI charts, you can select which timestamp field to use.
  • Latest applies only to numeric fields.

When combined with a Group by field, Latest returns the most recent value within each group. A table visualization can help verify which record is considered “latest” during analysis.