standard deviation and variance aggregator

TouK · Dec 12, 2024 · acf33c6 · acf33c6
1 parent abc91a4
commit acf33c6
Show file tree

Hide file tree

Showing 9 changed files with 697 additions and 76 deletions.
diff --git a/docs/Changelog.md b/docs/Changelog.md
@@ -25,6 +25,7 @@
     * Added support for PerformSingleExecution action in DeploymentManager and in GUI
     * Improved scenario state management to include information about current and deployed versions and allow more customization
 * [#7184](https://github.com/TouK/nussknacker/pull/7184) Improve Nu Designer API notifications endpoint, to include events related to currently displayed scenario
+* [#7307](https://github.com/TouK/nussknacker/pull/7307) Added StddevPop, StddevSamp, VarPop and VarSamp aggregations
 
 ## 1.18
 

diff --git a/docs/scenarios_authoring/AggregatesInTimeWindows.md b/docs/scenarios_authoring/AggregatesInTimeWindows.md
@@ -81,6 +81,10 @@ Let’s map the above statement on the parameters of the Nussknacker Aggregate c
 * Set - the result is a set of inputs received by the aggregator. Can be very ineffective for large sets, try to use ApproximateSetCardinality in this case
 * CountWhen - accepts boolean values, returns how many of them are true
 * Average - computes average of values
+* StddevPop - computes population standard deviation
+* StddevSamp - computes sample standard deviation
+* VarPop - computes population variance
+* VarSamp - computes sample variance
 * ApproximateSetCardinality - computes approximate cardinality of a set using [HyperLogLog](https://en.wikipedia.org/wiki/HyperLogLog) algorithm. Please note that this aggregator treats null as a unique value. If this is undesirable and the set passed to ApproximateSetCardinality aggregator contained null (this can be tested with safe navigation in [SpEL](./Spel.md#safe-navigation)), subtract 1 from the obtained result.
 
 If you need to count events in a window, use the CountWhen aggregate function and aggregate by fixed `true` expression - see the table with examples below. Subsequent sections use the Count function on the diagrams as an example for the **aggregator** - it is the easiest function to use in the examples. Please note, however, that technically, we provide an indirect implementation of this aggregator.