Open nathanrogersgoogle opened 5 years ago
We also discussed grouping this data by process and thread, which would necessitate changing the histogram-span in multi-event-sub-view to a full histogram-set-view. That happens to include controls such as the statistic picker. Instead of generating a single Histogram, the multi-event-sub-view would generate one histogram per thread. Each histogram would contain diagnostics such as 'process' and 'thread', which the histogram-set-view would use to group the histograms into a hierarchy, with a single Histogram at the root.
https://github.com/catapult-project/catapult/blob/master/tracing/tracing/ui/analysis/multi_event_sub_view.html#L61 https://github.com/catapult-project/catapult/blob/master/tracing/tracing/value/ui/histogram_span.html https://github.com/catapult-project/catapult/blob/master/tracing/tracing/value/ui/histogram_set_view.html
What we have now is just:
avg | 15,438,354 ns count | 12 max | 23,905,125 ns min | 6,185,208 ns std | 7,705,601 ns
Having medians and percentiles (and maybe a way to configure adding/removing stats from that view) would be nice.