Just sharing an observation and suggestion for a change in the behaviour of the Collect statement.
In my app I have a number of Collect statements of the following form:
Collect( '[ChemRec].[Recommendation]', Filter( colMyRecommendation, IsBlank(ID) ) )
Where the filter returns no records, I would expect the Collect statement to effectively be ignored - to simply not run at all rather than run by saying 'Add 0 rows to the collection' and initiate communication to the datasource.
When stringing a few of these commands together, peformance took a realy hit. Ten such collections took over a minute to run when there was no data returned by the filters (some Collect statements were run with Concurrent but others are ineterdependent so must run sequentially).
By wrapping all the Collect statements within If statements that first check the local collection to see if it actually contains entries (as shown below) peformance increased dramatically and the same set of commands take around a second to run.
If( CountRows( Filter( colMyRecommendation, IsBlank(ID) ) ) > 0, Collect( '[ChemRec].[Recommendation]', Filter( colMyRecommendation, IsBlank(ID) ) ) );
Solved! Go to Solution.
Hi!
Thank you for the feedback! I have shared your feedback with the engineering team for further consideration.
Thanks!
Hi!
Thank you for the feedback! I have shared your feedback with the engineering team for further consideration.
Thanks!
User | Count |
---|---|
140 | |
133 | |
80 | |
75 | |
74 |
User | Count |
---|---|
210 | |
197 | |
70 | |
66 | |
55 |