I'm not sure that I quite got there, but maybe this can give ideas.
Anomaly count is:
WINDOW_SUM(IF [Signal]="Anomaly" THEN 1 ELSE 0 END)
I'm not sure, but I think you may need to adjust the Table Calculation
setting to look like the below. Hard to tell from this sample size.
You may need to play around with those settings.
This generally generates many copies of this count, as many
as there are marks. So to filter it down to just one copy,
you can use
set the tablecalc settings as shown,
then filter to just 1.
But not sure if this will negatively impact your other filters.
Please see workbook v10.5 attached in the Forum Thread.
285270control.twbx 24.9 KB
Nailed it! Thanks, Swaroop!
Would you (or anyone else?) happen to know to group Types A-D and count anomaly or in range signals within the timeframe? For instance, Type B had 1 in range and 3 anomalies for the 4 dates.
Not sure that I fully caught the gist, but since everything is a table calculation,
I think it all comes down to adding Type to a shelf
and adjusting the Table Calculation settings accordingly.
In the workbook v10.5 attached in the Forum Thread,
Please example in sheet "time series".
There is another sheet "group anom count"
were types with the same number of anomalies are counted.
285270control2.twbx 42.1 KB
Wasn't sure if you got the follow up so re-posted : Control Chart - Group and got it answered. Like your approach and it will definitely come in handy. Thanks again!
Thanks for the follow-up, didn't see it.
All the best.
Full verbiage : Control Chart - Group Behaviour and Sum by Type