streaming-window-functions
4 items tagged with "streaming-window-functions"
Diagrams
Session Window Function in Streaming Analytics
Under streaming analytics context, a session window function is usually used to group events that arrive around similar times and to filter out periods of time where there is no events. Events can belong to only one session window. Session window function usually takes at least two parameters: Window timeout- A session window starts when the first event occurs and if new events occur within the timeout (compared with last ingested event), it will be included in the window otherwise the window will be closed at the timeout. Max window duration - If the events keep appearing, the window will close at the max window duration and a new window will start.
Sliding Window Function in Streaming Analytics
Under streaming analytics context, a sliding window function segments a data stream into time segments that can overlap with each other. The window only changes when an event enters or exits the window. Events can belong to more than one sliding window. Each window must include at least one event otherwise the sliding window will be emitted from output.
Hopping Window Function in Streaming Analytics
Under streaming analytics context, a hopping window function segments a data stream into time segments that can overlap with each other. Events can belong to more than one window. This diagram shows hopping window function with 20 seconds as window duration and 10 seconds as hop size. If the hop size is same as window duration, it generates the same results as a tumbling window function.
Tumbling Window Function in Streaming Analytics
Under streaming analytics context, a tumbling window function segments a data stream into distinct time segments and then do aggregation against them. Events can only belong to one single tumbling window and each tumbling window doesn't overlap with others. This diagram shows tumbling window function with 20 seconds as window duration.