tvalentyn commented on code in PR #28290:
URL: https://github.com/apache/beam/pull/28290#discussion_r1321867089
##########
sdks/python/apache_beam/testing/analyzers/perf_analysis_utils.py:
##########
@@ -170,6 +170,10 @@ def find_latest_change_point_index(metric_values:
List[Union[float, int]]):
if not change_points_indices:
return None
change_points_indices.sort()
+ # If the latest change point is the last index of the metric_values,
+ # then consider it as no change point since it might be a flake.
+ if len(metric_values) - 1 == change_points_indices[-1]:
Review Comment:
i think we are talking about # of runs, not days.
> this could be an edge case where the edivisive
Do we understand why this is happening if the spike is specifically on last
run but not other runs ?
Overall, sg if this workaround works.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
