Grafana provides a set of statistical reducer functions that allow you to summarize time series data into a single value. These are essential when you want to display summaries of time-based data, such as displaying a single number that summarises a changing numerical value over a specific time period, or when defining alert conditions.
Reducer functions "reduce" a set of values to one representative value using a statistical method. This is useful when:
You want to summarize a time series (e.g., show the average CPU usage over the last hour).
You’re using panels that show a single value per metric (like a stat panel or a table cell).
You’re adjusting the visual summary of a time series chart (e.g., when grouping by time intervals or choosing how to aggregate multiple series).
Grafana Reducer Functions commonly used in BAM dashboards
Function | Description |
|---|---|
Low (Min) | The smallest value in the series over the selected time range. |
Mean (Average) | The arithmetic average of all values in the time range. |
High (Max) | The largest value in the series. |
Time Weighted Mean | Averages values weighted by how long each value persists. More accurate when data points are spaced unevenly. |
Difference | The difference between the last and first value in the time range. |
Range | The span between the highest and lowest values ( |
How Reducer Functions Work in Different Panel Types
Time Series Charts:
Time series charts typically plot all data points over time, but reducer functions can still influence what’s visualized:If multiple series or multiple values per time bucket are returned, Grafana may use a reducer to aggregate them (e.g., average, min, or max) before plotting.
The resolution of the chart (based on time interval or query configuration) determines how data is grouped, and the selected reducer affects what value represents each interval on the Y-axis.
Reducers can also apply in legends, tooltips, or hover summaries, showing a single number summarizing the series across the selected time range.
Stat / Gauge / Table Panels:
These panels reduce a time series to a single value per metric, based on the chosen reducer function. For example, a stat panel might display the maximum memory usage over the past 30 minutes, while a table cell might show the average response time for each endpoint.