In question related to Filtering list values
From the posted example it only returns the nodes which contains values which are matched by the filter.
It does not return a filtered list of the the date times as well just returns all. it it possible return as filtered [datetime] or do you have to go down the expensive way of making each datetime and use more resources? Currently have the expensive way implemented and it uses far to much memory
Hi @madhousez, yes the list won’t be the filtered list of datetime and will return all.
This is how filter works currently. You apply a filter on the to-be-reached node, if the node matches the filter you traverse the edge and pick it.
We can decide if we want to pick that node based on timestamp values and pick the node if any of the value in the list matches the filter. But after picking that node we can’t filter now (can’t decide to traverse the edge) and have to return a complete list of timestamps. This is because timestamps contains list of values and not UIDs of other nodes.
The client can post-process the data in your use case.
Thank.
We go back to my list of edges to timestamps then as there would be 1000’s of entries in one list * few 1000s nodes.
Trying to do a count on how frequently smoothing happens between a period
Yes, for counting you would have to create a separate node for each of the timestamp as count() works only on predicates and UIDs. I can’t see any better workaround.