Add warmup counter to EWMA rate

[slimjim777]

What this PR does

Adds a warm-up period to the EWMA rate so that the rate for the first N samples is returned as 0. The approach mirrors the approach in VividCortex, although there the recommendation is to initialize the moving rate as the mean of the first 60 samples.

Which issue(s) this PR fixes or relates to

Fixes #5443

Checklist

• Tests updated.
• Documentation added.
• CHANGELOG.md updated - the order of entries should be [CHANGE], [FEATURE], [ENHANCEMENT], [BUGFIX].
• about-versioning.md updated with experimental features.

[CLAassistant]

All committers have signed the CLA.

[slimjim777]

@jhalterman I took a shot at #5443 and the implementation is in between the VividCortex approach (with a counter) and the current approach with using the sliding window time. This returns 0 during the first 9 ticks, whereas the sliding window time does that for the first 60s.

I amended the existing test for math.EWMARate to expect 0 for the first 9 samples, but I suspect that there other tests that will fail as a result of this change.

[slimjim777]

Amended the warm-up sample count to be 60 so it's aligned with the sliding window time of 60s.

[jhalterman]

Hi @slimjim777 Thanks for the PR and apologies for the delay!

My main question is, can we make the warmup size configurable? We could create an EWMA with warmup via something like NewEWMARateWithWarmup.

[slimjim777]

My main question is, can we make the warmup size configurable? We could create an EWMA with warmup via something like NewEWMARateWithWarmup.

I’ll add that

[slimjim777]

Hi @slimjim777 Thanks for the PR and apologies for the delay!

My main question is, can we make the warmup size configurable? We could create an EWMA with warmup via something like NewEWMARateWithWarmup.

I've added a NewEWMARateWithWarmup method.

(It could have been done with adding the functional options approach or a builder pattern, but I kept it as a separate constructor)