Multi-stream: sliding window average over a reading feed

A sliding window overlaps, so it cannot be a groupingBy. Stream the end index instead and take a subList behind it - and format with Locale.ROOT so the report reads the same on every machine.

Code

record Reading(int at, double celsius) {}

var readings = List.of(new Reading(1, 20.0), new Reading(2, 22.0), new Reading(3, 24.0),
                       new Reading(4, 30.0), new Reading(5, 26.0));
int size = 3;

IntStream.rangeClosed(size - 1, readings.size() - 1)
         .mapToObj(i -> {
             double avg = readings.subList(i - size + 1, i + 1).stream()
                     .mapToDouble(Reading::celsius).average().orElse(0);
             return String.format(Locale.ROOT, "at %ds avg(%d) = %.2f", readings.get(i).at(), size, avg);
         })
         .forEach(System.out::println);

Output

at 3s avg(3) = 22.00
at 4s avg(3) = 25.33
at 5s avg(3) = 26.67