You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The current math.EwmaRate implementation does not require a warmup period, allowing a rate to be available after a single sample is added, which can lead to unexpected results, as described in #5394. Let's move warmup support from that PR into EwmaRate itself, to avoid problems with other users of this type.
The text was updated successfully, but these errors were encountered:
@jhalterman the implementation in the PR compares the firstUpdate time with the sliding window, whereas the VividCortex approach is to keep a count of "warmup samples" and to compare with the WARMUP_SAMPLES const. Were you thinking that math.EwmaRate should use the current approach or switch to the count approach?
The count approach looks like a cleaner approach for math.EwmaRate, otherwise there would need to be a way to pass in the first update and sliding window times.
Describe the bug
The current
math.EwmaRate
implementation does not require a warmup period, allowing a rate to be available after a single sample is added, which can lead to unexpected results, as described in #5394. Let's move warmup support from that PR intoEwmaRate
itself, to avoid problems with other users of this type.The text was updated successfully, but these errors were encountered: