--- title: Calculating Averages, Sums, Minimums and Maximums description: A snippet for when you want the aggregate floatValues by sum, average, minimum or maximum in TelemetryDeck. lead: A snippet for when you want the aggregate floatValues by sum, average, minimum or maximum. order: 1 testedOn: TelemetryDeck API v3 searchEngineTitle: How to aggregate floatValues by sum, average, minimum or maximum searchEngineDescription: This is a quick overview on how to aggregate floatValues by sum, average, minimum or maximum using the TelemetryDeck Query Language. --- When you're sending numerical values to TelemetryDeck's `floatValue` signal property, you usually want to aggregate them somehow. This is how to get the average value of a `floatValue` metric for a specific signal type: ```json { "granularity": "day", "queryType": "timeseries", "filter": { "dimension": "type", "type": "selector", "value": "YOUR_SIGNAL_TYPE" }, "baseFilters": "thisApp", "appID": "YOUR_APP_ID", "aggregations": [ { "fieldName": "floatValue", "name": "Mean", // "doubleSum", "doubleMin", "doubleMax" also work here "type": "doubleMean" } ] } ``` Replace `YOUR_APP_ID` with your app id and `YOUR_SIGNAL_TYPE` with your signal type. ## How it works ### Granularity The granularity is set to `day` in this example. This means that the data will be aggregated by day. You can also use `hour` or `month` as a granularity to get averages for months or hours. Use granularity `all` to produce one single number for the whole displayed time range. (The displayed time range is usually set in the UI). ### Filter The filter is used to filter the data by app id and signal type. You can also add more filters here, for example to filter by a specific version of your app. ### Aggregations The aggregation is set to `doubleMean` in this example. This means that the average value of the `floatValue` metric will be calculated. You can also use `doubleSum` to get the sum of all values or `doubleMin` and `doubleMax` to get the minimum or maximum value. [Read more on aggregators in TQL](/docs/tql/aggregators/).