I have a standard deviation aggregate from Birt's end calculating on a dataset. I then need to group by a selection of four Id's so that I end up with four standard deviations for different timers. I've set up the group-by options on Birt, but it's not working, it's still as if there's no grouping in place.
Obviously there is a large scope for what may be the problem here, so I can only really leave it to experienced users to inquire what settings I have in place. I also must add that I can't do this group-by from the SQL end, it has to be done 开发者_StackOverflow中文版in Birt. But it does work properly in SQL
Appreciate the help.
Assuming I have understood correctly, I think the answer is that you should have added an aggregation to the table object on your report, rather than a computed column to your dataset.
If so, the answer is to delete the standard deviation computed column data item from its cell in the group header or footer in your report (whichever you are using to display the group results), then right-click on the empty cell and select Insert > Aggregation. Give the aggregation an appropriate name (eg. grp_StdDev), specify the datatype as Float, the function as STDDEV, the expression as your dataset timing item and the Aggregate On as the Group for timer ID.
OK the new aggregation and preview the results - the new aggregation should now vary between groups, with the correct standard deviation for the timings for each timer.
精彩评论