re: Window Functions in Stream Analytics VIEW POST


From what I see in the blog what they call "sliding windows" are actually "hopping windows". Does it make sense?


you said "A sliding window moves along the time axis, grouping together events that happen within the window length s." and the Flink blog said "For example, we can compute every thirty seconds the number of cars passed in the last minute. Such windows are called sliding windows.". I see that they are the same. I understand sliding window with a second parameter on Flink like the Tumbling Hop window that you have mentioned. Based on this post ( "The sliding windows assigner assigns elements to windows of fixed length. Similar to a tumbling windows assigner, the size of the windows is configured by the window size parameter. An additional window slide parameter controls how frequently a sliding window is started. Hence, sliding windows can be overlapping if the slide is smaller than the window size. In this case elements are assigned to multiple windows."

I think the terminology is a bit confusing and the differences between the windows are subtle.

An additional window slide parameter controls how frequently a sliding window is started

By the definition used in my post a sliding window is started based on the data and not based on a fixed parameter.

For example, you could have windows of size 10 minutes that slides by 5 minutes

This corresponds to a hopping window of 10 minute size, hopping 5 minutes.

What I'm trying to say is what Flink calls sliding windows are hopping windows in my post.

Makes sense?

yes. it makes sense what you argued about hopping windows. I have decided to send a message on your post because I saw your table which does not include sliding windows for Flink. So, I guess I could conclude that because Flink has to types of Sliding window, the second type with a second parameter is actually a hopping window in your concept. Nice discussion by the way. Thanks

code of conduct - report abuse