Idiomatic Windowing on Streams

[Richard Warburton]

Hi,

Sometimes its useful to be able to operate on a time-window over a stream
of data with a defined encounter order. So for an element at index i, I
want to perform an operation on a stream of that element and the next n
elements. The most trivial example I can think of is a simple moving
average of data points.

I'm currently struggling to find a way to idiomatically express this in the
Streams framework.  My current best effort is to take a stream of indices
and map those indices to a substream of my original stream of data and I
can then do some reduction over those elements.  So here's an SMA for
example:

double[] means =
    IntStream.range(0, numbers.size() - N)
             .mapToDouble(i ->
                numbers.stream()
                       .substream(i, i + N)
                       .mapToInt(x -> x)
                       .average()
                       .getAsDouble())
             .toArray();

Is there a better approach than this?

regards,

  Dr. Richard Warburton

  http://insightfullogic.com
  @RichardWarburto <http://twitter.com/richardwarburto>