Add another implementation of Min

[rnjtranjan]

No description provided.

[rnjtranjan]

supposed to be faster but benchmark show slower:

minimum (window size 100): OK (0.32s)
4.98 ms ± 259 μs
minimum (window size 1000): OK (0.21s)
13.6 ms ± 818 μs
minimum2 (window size 1000): OK (0.22s)
73.5 ms ± 4.8 ms

[harendra-kumar]

This is a scan benchmark. Can we also try the fold benchmarks? Also try descending order vs ascending order.

We have descending order benchmarks under fold category, we can try the same for postscan as well:

        , benchWith sourceDescendingInt numElements
            "minimum descending (window size 1000)"
(Ring.slidingWindow 1000 Window.minimum)

Also, try for smaller window sizes e.g. 10 and 100 sized window.

[rnjtranjan]

for 10,100 size:
scan
minimum (window size 10): OK (0.15s)
4.92 ms ± 461 μs
minimum (window size 100): OK (0.16s)
4.98 ms ± 414 μs
minimum (window size 1000): OK (0.20s)
13.5 ms ± 722 μs
minimum2 (window size 10): OK (1.14s)
2.21 ms ± 103 μs
minimum2 (window size 100): OK (0.31s)
9.95 ms ± 550 μs
minimum2 (window size 1000): OK (0.22s)
72.9 ms ± 3.0 ms

[rnjtranjan]

Folds are performing better.
fold
minimum (window size 100): OK (0.23s)
3.66 ms ± 170 μs
minimum (window size 1000): OK (0.11s)
8.17 ms ± 679 μs
minimum descending (window size 1000): OK (0.22s)
7.11 ms ± 345 μs
minimum ascending (window size 1000): OK (0.18s)
11.9 ms ± 742 μs
minimum2 (window size 100): OK (0.21s)
818 μs ± 42 μs
minimum2 (window size 1000): OK (0.22s)
852 μs ± 42 μs
minimum2 descending (window size 1000): OK (0.13s)
4.04 ms ± 385 μs
minimum2 ascending (window size 1000): OK (0.13s)
4.13 ms ± 346 μs

[rnjtranjan]

Looks good

[rnjtranjan]

Fixes:#1591