When to use a plug-in aggregation function and when to use a plug-in custom view?
If you are not sure how to choose between custom plug-in aggregation functions and custom plug-in views, this entry explains the differences or advantages of one over the other in more detail. A plug-in custom aggregation function works like other aggregation functions such as count, sum, average or standard deviation and may appear in the select-clause and in the having-clause in EPL. A plug-in custom view can be a data window providing an expiry policy like a time window or length window, for example. Or instead a custom view can derive new information from a stream of events such as the results of a linear regression function (aka. derived-value view). A plug-in view is always attached to a certain type of event that is provided by a filtered event stream or a pattern or by another view. Plug-in views can receive only one type of input event (input stream). If the view is a data window view, the output event type is always the same event type as the input event type. For derived-val
