I have used the Parallel Chunk Loop to replace basic loops with great success. I would liketo extend this to a scenario where I am performing a "Group Loop." However, the groups are not all of the same size, so I cannot use the parallel Chunk Loop nodes (which use a fixed chunk size). Is there an equivelant parallelized Group Loop node?
No we currently only have one parallel looping set of nodes. However extending them (and especially having a group version) is on our feature list. I keep you posted when there are any news.
As a workaround. Instead of looping over the table, loop over a list of the grouping columns values (unique) with the parallel loop start node. Than use inside a table row to variable loop start and filer the whole data table using the generated flow variable. This will give you the group part of your table.
However extending them (and especially having a group version) is on our feature list. I keep you posted when there are any news.
Is there any news about a possible Parallel Group Chunk Loop?
I'm using your proposed workaround in a few workflows now, but it really turns out as a common pattern (particularly when used with the Image Processing extension), so it would be really useful to have such a node. It would save me a GroupBy and a Joiner node, would look much nicer, and would be easier to explain to others ;)
could you maybe share an example workflow with the required nodes/settings for your example? Unfortunately I cannot re-create your proposed workaround.
It’s been on the request list for seven years. As processing speeds gains on a single-core are limited and multi-core processors are increasingly prevalent, this is a much needed development.
+1 for this feature. Increasing speed and throughput by using the capabilities of modern computers is more important than the JavaScript UI.